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

Outsulation – best way to install on roof and walls?

dmagee | Posted in Green Building Techniques on


I’m building in Whistler BC, Canada (climate zone 6) and I am planning on adding 2″ of rigid insulation to the outside of all roof surfaces, and 1″ of rigid to all exterior walls (XPS rigid insulation).

The roof is an unvented cathedral ceiling with 10″ of spray foam insulation between the engineered rafters. The walls are 2×6 with spray foam insulation.

We plan on sheathing the roof with plywood, then installing 2″ of rigid insulation (taped seams), and then covering the rigid with another complete layer of plywood (then ice and water shield, with standing seam metal roofing). We plan on doing the same plywood/rigid/plywood rigid sandwich on the walls.

Is there any professional out there who can comment on the above plywood/rigid/plywood sandwich approach – and if this approach is a good solution vs leaving an air space and strapping the outsulation so there is an air channels between the finished siding or roofing and the outsulation?

Thanks in advance,

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

    I'm not a fan of your approach. I don't recommend encapsulating plywood sheathing on both sides with low-permeance foam insulation. I always tell builders, "If you want to install foam insulation, choose one side of the sheathing to install it. You can install it on the inside of the sheathing if you want, or on the outside of the sheathing -- just don't install it on both sides."

    It's usually a good idea to allow the plywood sheathing to be able to dry in at least on direction. So, if you like exterior foam, you should install cellulose between your studs. And if you like closed-cell spray foam on the interior side of your sheathing, then you should install mineral wool insulation on the exterior side of your sheathing.

    If you take my advice and switch to vapor-permeable insulation between your studs and rafters, note that your exterior foam insulation isn't thick enough to keep your plywood above the dew point during the winter. For more information on that issue, see Calculating the Minimum Thickness of Rigid Foam Sheathing.

  2. dmagee | | #2

    Thanks I appreciate your time and response!

    My main goal is a high performance building (high r-values) outsulation was a way to reduce thermal bridging as we have lots of structural posts/studs in the exterior walls. While high r-value is important to me, I also want to ensure I'm building a structure that will be healthy and not have moisture issues.

    I only allowed for 1" wall outsulation, so it seems like I have to stick with closed cell spray foam between my studs. I can't find any 1" mineral wool insulation board - do you know of any? 1.25" is the thinnest I can find.

    Also, I'd like to enlist further advice/consulting to ensure we're making smart investments w/r/t energy efficiency and not causing weaknesses. Thoughts on where I can find this?


  3. GBA Editor
    Martin Holladay | | #3

    Why can't you install exterior insulation that is thicker than 1 inch?

    Q. "Also, I'd like to enlist further advice/consulting to ensure we're making smart investments w/r/t energy efficiency and not causing weaknesses. Thoughts on where I can find this?"

    A. The Green Building Advisor web site.

  4. dmagee | | #4

    Hi Martin,

    Our home has post and beam features that assumed adding 1" of outsulation to our walls. Adding more than 1" would result in posts and blocking details being recessed vs proud.

    Your GBA article below shows the sandwich (ply/rigid/ply) approach over unvented ceilings, wouldn't the same approach work on the wall? With closed cell interior (wall and rafter), and rigid exterior there should be little to no air/moisture movement through the assembly.

    Your original recommendation makes sense to me to allow drying in at least one direction. But the best approach for my build still eludes me.

    Do you recommend Air Krete insulation as an alternative (less risky?) to closed cell spray foam?

    This is my first build so have many questions.


  5. dmagee | | #5

    Hi - alternatively, is it best to eliminate the wall outsulation and stick with 5.5" of spray foam (approx. r 35) and focus on the ceiling & increase to at least 4" of outsulation?


  6. GBA Editor
    Martin Holladay | | #6

    In the article you linked to, I don't think that I recommended a "foam sandwich" with vapor-impermeable foam on both sides of the roof sheathing. This is what I wrote:

    "To summarize, there are three ways to build an unvented roof assembly:
    Install spray foam (either open-cell foam or closed-cell foam) against the underside of the roof sheathing, and no other type of insulation. Be sure that the thickness of spray foam is adequate to meet minimum code requirements. ....
    Install rigid foam insulation above the roof sheathing and air-permeable insulation between the rafters. ...
    Install a layer of closed-cell spray foam against the underside of the roof sheathing, and fill the rest of the rafter cavity with an air-permeable insulation."

    Somewhere on the GBA site, I have undoubtedly contradicted myself, and given the OK to a "foam sandwich" design. But it's best to avoid a foam sandwich if you can.

    I have no experience with AirKrete insulation, but I understand that it is very crumbly.

    When you wrote, "Our home has post and beam features that assumed adding 1" of outsulation to our walls. Adding more than 1" would result in posts and blocking details being recessed vs proud," you made me worried. I hope you are not planning to have the post and beam frame visible on the exterior, like a medieval German building, with insulated infill panels between the posts. If that's what you are doing, it will be an air-sealing nightmare that will lead to rot.

  7. dmagee | | #7

    Hi - The posts are not continuous thru the walls. The are actually steel posts clad on inside and outside to look like continuous 12"x12" fir posts with triple pane windows in between. The steel post cavity will be spray foamed as well.

    Thanks for your advice - it seems like I'm best to avoid the xps outsulation on both the walls and roof but consider mineral wool sheathing on the walls if I do anything. Would mineral wool sheathing work for the roof as well? Can I sandwich mineral wool with ply - or is it best to just use 1x4 strapping and leave the air cavity?

    thanks again for all your advice.

  8. GBA Editor
    Martin Holladay | | #8

    The steel posts have to be detailed carefully to make sure they aren't thermal bridges. Ideally you have found a way to install at least R-20 on the exterior side of the steel posts -- right? I hope so.

    And you've also found a way to install a continuous air barrier on the exterior side of your framing, so there are no vertical cracks between your steel posts and adjacent wall sections -- right?

  9. GBA Editor
    Martin Holladay | | #9

    Q. "Would mineral wool sheathing work for the roof as well? Can I sandwich mineral wool with ply?"

    A. There's a difference between wall and roof assemblies. Usually, wall cladding is vapor-permeable (especially if you include a ventilated rainscreen). But most roof assemblies don't dry to the exterior.

    There are some types of commercial mineral wool insulation designed for flat roofs. If you wanted to use these panels, you would have to build ventilation channels above the insulation (but under the roofing) to allow the assembly to dry to the exterior. It would be complicated.

  10. dmagee | | #10

    Hi - My overall goal is high R-Value/energy efficient home. Perhaps the extra complexity of outsulation isn't the right approach at this stage. What do you think about this traditional high r-value solution:

    - 11-12" of closed cell spray foam insulation in the rafters (r60+)(unvented cathedral ceiling)
    - 5.5" of closed cell spray foam insulation in walls (r35+)

    This is well above recommended code r-values and doesn't introduce any unknown risks with the outsulation.

    I'd still like to reduce thermal bridging and that was my original goal with the rigid outsulation. Any alternative interior suggestions?


  11. GBA Editor
    Martin Holladay | | #11

    Q. "What do you think about this traditional high-R-value solution: 11-12 inches of closed-cell spray foam insulation in the rafters?"

    A. I think you should check with a spray-foam installer before you assume that this is possible. Most closed-cell foam manufacturers limit the maximum thickness of installed insulation to meet fire safety regulations, and the maximum thickness of these foams is generally significantly less than 11 inches. It can be difficult to test the performance of 11 or 12 inches of spray foam for fire safety, since existing test methods can't accommodate a materials test at that thickness; only an assembly test is possible.

    Q. "I'd still like to reduce thermal bridging and that was my original goal with the rigid outsulation. Any alternative interior suggestions?"

    A. It's possible to install rigid foam on the interior side of your framing, although the method has its own disadvantages. It's difficult to detail interior rigid foam at partition intersections and rim joists, and if you follow the method, you end up with cold wall sheathing, which raises the risk for moisture accumulation in the sheathing.

    The other approach to a high-R wall is the double-stud wall. To learn more about double-stud walls, see Monitoring Moisture Levels in Double-Stud Walls.

  12. Expert Member
    Dana Dorsett | | #12

    The lifecycle global warming impact of 11-12" of closed cell spray foam is HUGE, and an environmental disaster, and the the chromatic opposite of "green".

    This is becaus the most common blowing agent used is HFC245fa at more than 1000x CO2 global warming potential (GWP).

    Furthermore, it has to be applied in 2" lifts, with sufficient cooling time between lifts to avoid an immediate fire hazard during installation.

    And, the high R/inch is severely undercut by the high thermal bridging of the webbing of your engineered rafters.

    Open cell foam can be applied in 5-6" lifts, is blown with water (very low GWP), and is sufficiently permeable to water vapor that the roof deck can dry.

    If you want a higher performance water blown open cell foam, the 2lb density Icynene MD-R-200 is about R5/inch, and would be about 0.35 perms @ 11-12", (just slighlly tighter than a kraft facers on batt insulation), sufficiently vapor permeable for the assembly to dry toward the interior, but tigtht enough that you would not need any sort of interior side vapor retarder. It'll run about the same cost-wise or only very slightly less than the R6-7/inch HFC-blown stuff, but you'll have a MUCH more resilient assembly and with a much reduces environmental footprint. They have a slightly denser version (MD-R-210), but I would avoid it in his application, since it's too vapor-tight.

    With either half-pound or two-pound water blown open cell in the cavity it's fine to use any thickness of exterior foam. (But use polyiso or EPS, not XPS, eh? XPS is blown with HFC134a, with a substantially worse GWP than HFC134a, whereas EPS and polyiso are blown with pentane @ 7x CO2 GWP.) If you add rigid foam to the interior of this assembly it has to be at least semi-permeable to avoid the moisture trap- unfaced EPS is probably the best bet, foil or poly faced polyiso or EPS the worst.

    If you're concerned that half-pound open cell foam is TOO open, see figure 10 p28 (PDF pagination) of this document:

    That's the simulation in a Minneapolis (zone 6) climate, with a stackup of of R25 open cell foam (~7-8") with R24 fiberglass between the foam & gypsum, with 5-perm spray-applied vapor retarder (a single coat of latex ceiling paint is also about 5 perms.) With 2x the foam you have 2x the vapor retardency- the peaks will be lower. If you want to do better, Intello Plus or MemBrain smart-retarders, or 1-3" of Type-II EPS on the interior side would cut the vapor permeance further, while still having a much higher drying rate than the closed cell version. The closed cell example in the figure 10 graph is R25 or about 3.5-5", which runs about 0.3 perms, which is comparable to a full cavity fill of MD-R-200 water blown semi-open cell foam (R55-R60, center cavity).

    Best bang per Loony without setting the globe (or the building) on fire:

    Half-pound open cell or blown cellulose in the cavities, 3-6" of rigid polyiso above the roof deck with Intello or Membrain vapor retarder on the interior (or 2-3" iso above the deck, and 3" of EPS between the bottom & gypsum as a 1-perm vapor retarder.) You'll have a center-cavity value over R75, and at least R18 of thermal break over the rafters (R24 of thermal break if you do the 3"/3" solution.) The roof deck can dry, and you won't have cooked the planet with high GWP blowing agents.

  13. dmagee | | #13

    Thanks for the feedback.

    Trying to piece together a high r-value home seems to be filled with possible land mines and uncertainty. Taking into account the air quality of a home - filling the walls and ceiling with a petroleum based spray foam chemical mixture has always seemed risky, but a risk worth taking considering the energy benefits.

    I now see why new building techniques/approaches are difficult to gain traction in residential construction. It is pretty easy to fall back on tried-and-tested approaches that eliminate some of the risk.

    I need to do more research.


Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |