GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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

Vapor Permeability of Roof Membrane

ljung | Posted in General Questions on

Modern Shopdominium. 32′ x 80′. All 7/8s” 24 gauge corrugated steel run vertically on walls and roof. No rain screen because the corrugated will self drain and vent. Shop on first floor treated as a completely separate building for insulation and heating purposes. Truss created attic room, 17′ wide x 8.5′ high x 80′ long. 8/12 pitch.  No vertical walls on exterior of attic room.  3/4″ plywood sheathing on trusses. Attic room completely conditioned. No vents. Montana Zone 6B.

Now the explanation, which I assume will result in me being shamed and ostracized!

I get it. Really, I do. I’ve read all of buildingscience’s and other’s articles on the subject I could find. I understand sorption and dew points and all of that. I understand about insulation ratios and venting. I know what – in an ideal situation – I’m supposed to do. Insulate the roof above with board insulation and/or insulate under the sheathing to hit the ratio to avoid sorption. Etc., etc., etc. But I simply cannot – or will not – do it.

First, outside insulation is probably not a good idea with the corrugated.
Second, I can’t afford the ideal solution.
Third, much of the discussions is a bit like debating the number of angels that can dance on the head of a pin. Lots of great theory, but a bit short on reality. Blower door tests? Really? You guys must have made a heck of a lot more money than I did in my 7 decades!

I’m not going to stack layers of board insulation on top of the sheathing. Why? Because the corrugated metal producers warn me that on an 8/12 pitch roof, the length of the screws necessary to reach through 11 inches of rigid insulation and into the plywood and trusses will allow too much “sliding” of the heavy metal, and trying to put some sort of a damn or ledger at the end of the run of the metal to prevent the sliding of a 20′ length of steel is just asking for trouble. So use shingles? I don’t want to. I don’t like the look. There. I said it. I like the look of a Scottish or New Zealand black barn. Yeah. I know. Montana’s climate is different than either of those countries.

Moreover, the cost of top mounted insulation in that quantity is nothing less than enormous. It’s approaching silly, really. Coughing up the cash in today’s ruined economic climate to reach r48 in the roof of a cathedral-like space is nuts.

So just vent the space and blow in insulation? Nope. Don’t want to. I like open ceilings and the very modern look of exposed trusses.

Insulating the underneath of the sheathing from the inside isn’t any more feasible. I have 2” x 8” top chords in the truss. To increase the depth of that chord in order to house the necessary amount of insulation to insure that the dew point isn’t reached at the plywood (the first cold plane) will cost far too much money. “Silly fool”, you say. “You’ll pay one way or another.” Maybe.

Yes, even keeping the interior at a max of 68 degrees, and the humidity at 30%, the moisture may (ok, probably will) condense on and in the plywood. But the rockwool insulation won’t absorb or adsorb the moisture. So I will have sort of damp plywood for several months. But not soaking wet. I’ll run the HVAC etc. systems. Or I’ll just sprinkle of bunch of those “Do not eat!” desiccants I find in beef jerky around. I like beef jerky.

The climate numbers support my assumption. Even though I am in Zone 6B here in Montana, the 28 degree median or average or degree days or whatever it’s called in the three critical winter months won’t result in a soaked structure.

No vapor barrier against the underneath of the insulated truss bays, and wall finish comprising spaced MDF 2′ x 4′ panels on 1.5″ hold-offs on the ceiling (with 3/8s” spacing between the panels) will allow for the structure to dry to the inside. Yes, I know – I’ll probably only end up with an r 30-something. OK. No codes and no inspections where I live. And cord after cord of firewood on the property. More than I can ever use. I do appreciate the irony of writing these sentences on green building site. Yet another reason to hold me in disdain.

What’s that? At least spray closed cell foam for an inch or two under the sheathing before you batt the rest of the 8”. Nope. Too expensive around here. And even if I did, I could never achieve a high enough r value in the interior of ceiling alone to prevent the warm interior air from condensing against the cold plywood. And I just am not going to couple internal insulation with rigid foam on the deck.

Deride me all you want. The building is now costing me 45% more than I planned last Fall. If it rots and falls apart, I’ll never admit to you guys.

So please save the lectures and the condemnations and answer the following: it is better to install an impermeable membrane on the roof plywood (e.g., Ice and Water Shield), which would prevent the structure from drying to the outside, or I should install a permeable membrane on the plywood, which would keep out liquid water and (with the corrugations acting as vent channels) but would allow the structure to dry to the outside AND the inside. And what product should I use? (I want a self-adhesive membrane to cover the entire roof.)

I realize that discussions of these topics have become battles between religions. And many of you may think that I am condemned to hell (or at least purgatory) for even considering this approach. But I’m 70, and it’s too late to seek penance anyway.

So, help me out here! Thanks.

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. kbentley57 | | #1

    Maybe a good cup of coffee and a fireplace is what's needed for a few minutes. I think you're having those pre-execution jitters! Everything will be OK, just step back for a moment. Nobody's going to call you out for not building an insulation igloo, and I rarely see anyone on this site berated for making decisions based on costs. It's a pretty friendly place.

    There's really no benefit to having a vapor permeable membrane above the roof sheathing, unless you have an actual air space, that has air flow. The ribs on the standing seam don't really provide much if any air flow, and the metal itself is impermeable.

    1. ljung | | #3

      Thanks for the considered reply. See my reply to Akos below.

  2. Expert Member
    Akos | | #2

    Unvented roofs in zone 6 means either spray foam or exterior insulation.

    Exterior insulation doesn't have to be rigid.

    For example, you can install your trusses, sheath with OSB. Tape the seams of the OSB as your main air barrier and connect it to your wall air barrier. Cover the roof sheathing with a vapor barrier. The vapor barrier could be roofing craft paper, roofing underlayment or peel and stick membrane.

    Over this install horizontal 2x6 purlins on edge. Fill the top chord of the truss with R24 batts, same batts for the purlins.

    Underlayment over the purlins and install the metal roof directly on the purlins.

    This gives you an ~R50 roof with 50/50 exterior/interior insulation for condensation control without using any expensive insulation materials. Quick to build, all you need is a bit of extra lumber for the purlins.

    Can probably even save some cost by bumping up the spacing of your trusses as the roof is resting on the purlins.

    1. ljung | | #4

      The trusses are already on 48" centers, which requires that each truss heel rest on a 6"x6" post, with 2"x6" infill studs between the posts on 16" centers. See pict. Due to the wind load (110 mph for 3 seconds) and the snow loads (100 inches), I can't install an additional purlin structure over the top of the roof sheathing. The truss design is already maxed out. There are already 8" x ~45" purlins installed every 24" between the top chords of the trusses, and 2"x12" x ~45" cross joists installed every 24" between the bottom chords of the trusses.

      So that leaves the foam option. Does the following make sense?

      Assume I spray 1.5" of closed cell polyiso (which is really expensive in this locale) between the truss bays and against the underneath of the sheathing, followed by as much rockwool as I can lay in the bay between the trusses (under the foam). What's that result - only about an R33? On really cold days, interior air at 33% relative humidity will migrate around and through the rockwool and hit the interior surface of the foam. The rockwool will not absorb the water vapor. Since the foam is pretty much impermeable, the humid air would not be able to permeate the foam and condense on the underneath of the plywood sheathing or migrate into the plywood itself. In that case, since the sheathing does not absorb or adsorb moisture, there is no reason the sheathing must dry to the outside. Right? I might as well use an impermeable product like Ice and Water Shield or Titanium by Corning.

      But how good of a seal does the polyiso make against the underside of the plywood sheathing. If there is a gap between the polyiso and the sheathing or if the seal fails, isn't it likely that the humid interior air will condense on and transfer into the sheathing plywood? If I use an impermeable membrane on top of the sheathing, won't I have increased the possibility of rot by trapping water in the plywood? Would a semi-permeable be better?

      While the corrugated sheathing on the roof doesn't allow tremendous volumes of air to scrub the exterior surface of the membrane/plywood, from what I understand, there is some venturi effect which, with coupled with convective drive, will draw air from the bottom of the corrugated to the ridge line, thus permitting some drying to the outside, particularly in hot weather. If that's true, then what would be the disadvantage of using semi-permeable membrane, since it will still stop any water that sneaks past the corrugated?

      Even if the polyiso is correctly applied and forms a good seal against the sheathing, if the interior face of the polyiso becomes cold enough (since there is no exterior insulation), the interior air could condense on the interior surface of the foam. Would drying to the inside be sufficient to ensure that the interior of the bay doesn't become soaked?

      Thanks much guys.

      1. Expert Member
        Akos | | #6

        Cut and cobble polyiso under the roof deck is not the same as spray foam. This will never be air tight enough and you'll have moisture issues sooner or later. The only thing that works under the roof deck is closed cell spray foam. Rigid can be used over the roof deck.

        If the building is not a living space and you don't have a big moisture source, you can push the 50/50 insulation ratio a bit. How much is hard to guess, I would rather go for the right ratio than have to fix moisture issues.

        I would take your existing design and move the purlins above the roof deck. This gives you exactly the same structural support but saves you installing a hundred or so joist hangers. You can then insulate as per my previous post.

        Most batts are 4' long, so you can install them sideways between the trussess. You'll need some way of holding the insulation under the roof deck between the trusses. The simplest is insulation pins that you can adhere to the bottom of the roof deck.

        Since the roof will now be ~6" taller, you'll have to extend the siding up a bit to meet the new roof surface. This will keep the look of the building exactly the same.

        1. ljung | | #8

          Yes, I understand about cut and cobble and rejected that for the reason you mentioned. We originally had the purlins running across the trusses, like one would do with a pole barn. But the engineers and the truss designer nixed that idea due to the shear forces and the high winds. I'll check out insulation pins.

          1. Expert Member
            Akos | | #9

            There are many ways of dealing with both shear and uplift.

            If they really can't make it work, I would add in the extra lumber on top or go for exterior rigid insulation.

            Interior cut and cobble rigid unvented roof is a very risky assembly in any cold climate. There is a pretty good chance that the roof will fail.

        2. ljung | | #10

          Thanks much for your advice. Working with engineer now.

  3. Expert Member
    PETER Engle | | #5

    There is no real downside to using permeable underlayment on the roof. It might help, or not, but it won't hurt. As you have stated, there is some possibility of condensation on the inside of the spray foam. If there's enough condensation, it will drip back down through your insulation and ceiling finishes. When it's that bad, it can be bad. But if you don't have very high interior moisture levels, this is unlikely with your stackup. Don't be drying lumber or firewood inside the shop and you'll probably be OK. As far as drying to the interior, that will happen, but only when it is warmer outside than inside. There is no drying to the inside in winter, because the vapor drive is from warm to cold. Condensation doesn't dry out until spring.

  4. ljung | | #7

    Thanks Peter. Early in the project, I made the mistake of thinking of A building, rather than TWO distinct buildings, requiring two distinct treatments.

    The shop will remain at 48F when not in use, and unless I am spraying or gluing, I like to work when the shop is ~63F. But that's only two or three days per week in the winter. Looking at the charts, there is a huge difference in the insulation ratios required during those 28F degree days when the interior of the shop is kept at 48F rather than 68F.

    The attic room is separate living space, with normal temps and controlled humidity. Thinking of the structure as two buildings, the designs became much easier to consider and much cheaper to build. The shop will have a plain old house wrap unless I feel like springing for Prosoco liquid applied. The floor in the attic room will be heavily insulated, and probably with a vapor barrier to control the higher humidity shop air from hitting the insulated floor.

    The attic room - at this time, I think I will go with a semi-permeable on the roof deck (or maybe even a smart membrane), with polyiso and rockwool - with humidity carefully controlled. You are, of course, correct re the lack of drying in the winter. My hope, though, is that by keeping the humidity down, there will be far less chance of condensation on the interior plane of the polyiso.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |