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

6mil poly location in ceiling

user941025 | Posted in General Questions on

Hi all,

Another forum reader and I were discussing a detail last night and he suggested I open it up to the board, so have at it.

Between a recent bump of Thorsten Chlupp’s JLC article and a recent thread on avoiding adhesive between plywood sheathing and wall studs, I’ve come to the conclusion that for the small building that’s underway in my yard, I’ll see good airsealing results from 6mil poly, adhered at edges and openings, around the walls and another layer lapped over, across the ceiling.

The walls seem to fall within the rule of thumb: inside to out, they’re R19 batts inside 2×6 studs, plywood sheathing, 6mil poly air barrier, 5 1/2″ of layered-and-taped polyiso, rainscreen cladding. (R19 fiberglass, 6mil poly, R35.75 polyiso.)

The question is at the ceiling. To avoid running the 6mil poly in a manner that forms the shape of a wading pool between layers, the assembly would be, inside to out: R19 batts, 4″ of polyiso, then 6mil poly, and 5″ more of layered-and-taped polyiso. (R19 fiberglass, R26 polyiso, 6mil poly, R32.5 polyiso.)

There will be an ERV.

Condensation problem in the ceiling? Or, am I in the safe zone?

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

    It's possible that you'll get away with installing poly in the manner you propose, but it's quite risky. If you have exterior foam, you're not going to get any moisture accumulation in your sheathing layer, because it will always be above the dew point. So you don't need to stop any diffusion from the interior.

    If you ever have a flashing problem, however, and rain gets in your assembly from the exterior, you want your wall assembly to dry towards the interior. That's why the basic rule is: whenever you have exterior rigid foam, never install interior poly.

    If you're installing poly as an air barrier, you are about 20 years behind the times. Establish your air barrier at the plywood sheathing, or use the Airtight Drywall Approach.

  2. dickrussell | | #2

    MinDis, I agree with Martin on not using poly for the air barrier, for the reasons he stated. If you still prefer a sheet material that you can tape at overlapped seams for your air barrier, you could replace the poly with Certainteed's MemBrain. It's expensive, though.

  3. user941025 | | #3

    Humor me: what's stopping the assembly from drying to the interior? Inside the poly is plywood, fiberglass, drywall, paint.

    I should say outright that one rationale for poly in this case is that I can get a 12'w roll to cover each side more readily than I can with 10' Tyvek (which would also mean more seams and fussier lapping).

    I also note that Chlupp's 2009 article speaks highly enough of poly in dry conditions (and I recognize that "dry conditions" is a significant point--but I'm certainly not in a marine climate):

    Most of the houses I build have wood or vinyl siding; on those
    jobs we typically install a 6-mil poly barrier membrane, lapping
    the seams shingle style and sealing the laps with a nonhardening
    acoustical sealant.

    and in regard to taping the foam (which would be the first taped exterior layer if there's a combination of 1/2" corner plywood and 1/2"-less-thick-foam):

    . . . Covering the house with rigid foam board creates a potentially
    tight structure, but it’s the barrier membrane applied
    to the sheathing that does most of the work . . . .
    The earliest REMOTE walls used peel-and-stick membranes;
    this produced an incredibly airtight, waterproof
    shell but it was expensive. So most builders switched
    to less expensive materials — 6-mil poly or, in very wet
    zones, vapor-permeable drainage wraps such as Tyvek
    StuccoWrap or DrainWrap. Poly works fine in Fairbanks,
    which is fairly dry; in rainy southern Alaska, the Tyvek
    products are the usual substitute for peel-and-stick. The
    reason is that rainwater might travel through a nail hole and
    get trapped behind poly, whereas using a housewrap will
    allow for evaporation to the outside.
    . . . Our goal is an airtightness of 0.6 air
    changes per hour at a pressure of 50 pascals
    (ACH50) — the same standard required
    for a Passive House. We’ve achieved this
    with the 6-mil poly barrier.

    Finally, my expectation is that my roof, with overhangs, ought to keep bulk water well away from the stack of polyiso above my ceilings, if that roof performs as a roof should.

    Not exactly trying to "win" here, just asking if, based on these points, it might actually be an acceptable approach--and that (counterproductively) thinking of it as an antiquated approach might mean I miss an opportunity to seal well at one layer while that opportunity is still available to me. It certainly seems much more straightforward than later scrambling to create that barrier at the drywall.

  4. user-788447 | | #4

    Here are some clarifications so MD's question can be addressed more directly:

    The poly is being used for air sealing purposes. In MD's judgement taping the seams of the plywood is not a proving method yet.

    For the walls the poly is outside the sheathing so this does not interfere with drying towards the interior. Because I've seen this method of air sealing used (on GBA) when installing larsen trusses to retrofit existing structures it seemed like a valid approach for the walls specifically. Does a 20 year old approach necessarily mean it won't work?

    The roof/ceiling condition is the gray area. MD is sandwiching his air barrier poly to the outside of R-45 with only R-32.5 to the outside of the poly. This concerns me because it can't be determined if the poly will always be warm enough not to act as a condensation plane. This is the heart of his question.

    Unlike the wall assembly, the sheathing of the roof assembly has no outboard insulation. It will be a low slope shed roof with potential for a perimeter soffit vent but this method a venting may not establish air movement. Because the sheathing will be cold a vapor retarder I believe is needed at the ceiling plane.

    The roof structure as planned will be unique. The roof structure is freestanding and will clear the ceiling with its built up insulation package. In other words the roof, which bears on columns outside the walls, hovers above the ceiling assembly. The ceiling assembly is 2x6 joists topped with 9" of polyiso. Above that will hover an empty roof rafter cavity between 7.5" to 12" thick.

    It might be easier to run your poly along the bottom of your 2x6 ceiling joists and run about 6" or so down your interior walls, caulking it to the wall's top plate. As an Airtight Drywall Approach measure, ideally you would have caulked the outside of your top plate before putting on the plywood sheathing. Theoretically air could escape between the top plate and the exterior sheathing upward and outward but you can still go back in and caulk this intersection from the inside.

  5. jklingel | | #5

    FWIW, Thorsten no longer uses poly here, unless mandated to do so (within the city). He air seals plywood, which is on the outside of the inner wall of a double-stud wall. Tyvek goes on the outside of the outer wall. Dense packed cellulose in between. I have not checked to compare weather, but I'd be surprised if Minn is as dry as Frb, especially in the winter.

  6. user941025 | | #6

    Good to know, John--thanks. If I were air-sealing a dense-packed double wall, I'm pretty sure I'd be using Tyvek and not 6mil poly, too, whereas this barrier is just to the inside of two layers of polyiso.

  7. user941025 | | #7

    Still curious to see an analysis of this based on the actual details, as posted (beyond the two of us who were already present and discussing the question before I posed it here). Thoughts?

  8. user-988403 | | #8

    I do not think these assemblies (neither roof nor wall) are good ideas. I checked those quickly with an old fashion German diffusion analysis called "Glaser diagram". I might be "over" careful with this analysis because it does not account for any drying potential. To my knowledge the rule of thumb is: If the "glaser" analysis works, no further analysis are necessary for the diffusion problem. For the mentioned assemblies this is not the case.
    It is also not just about keeping the sheathing warm but also about the permeability profile of the assembly. For example the mentioned assemblies are fine in the "glaser" analysis without the poly (the way BSC recommends them).

  9. user941025 | | #9

    Philipp, thanks for posting--as always, I appreciate your time spent on these questions.

    But it's unclear upon what criteria your warning is based, other than the very low permeability of the 6mil poly (yes) yet as you note, there's no accounting in the Glaser diagram for drying potential (and since the assemblies should be able to dry to the inside, that issue seems cancelled out). It's therefore not clear to me what to take away from the analysis which you acknowledge is limited in its scope.

    Perhaps more clarification on what I'm missing in your analysis?

    Also--I am betting that the builder we've cited above has not made his change over to Tyvek since the 2009 article because of any sudden wholesale failure of all of those wall assemblies he had previously built.

    It appears that:
    a) the wall and ceiling can dry to the interior;
    b) the airsealing by poly technique can work effectively, even though it's out of fashion;
    c) the issue of the balance of insulation to the outside and inside of the ceiling poly still need exploration.

    And just briefly, I'd like to make a minor correction: J Chesnut's point that I don't find airsealing drywall a "proven method" could be corrected to say that it's something I'm less likely to believe I can do fully and effectively than wrapping this beast in poly that's caulked airtight. I have nothing against taping plywood seams.

  10. user-988403 | | #10

    MD, yes the analysis is limited but on the safe site. Drying potential is always limited especially to the interior. I like to have that as a security feature for any water that gets into your assemblies not planned for. The simplified analyses I do points out that there will most definitely be condensation in your assemblies. That means an analysis is needed if the drying potential will be enough to get rid of the moisture over a given period of time. That is what WUFI does. But as I said before I would not recommend a assembly that relies on a lot of drying potential to work. I would recommend taping the plywood with a product made for that purpose as a belt and suspender glue the edges of the plywood.

  11. user941025 | | #11

    Thanks, Philipp. I will chew on that.

    Any other analyses out there, based on the details as actually described? Martin, perhaps? Others?

  12. GBA Editor
    Martin Holladay | | #12

    Yes, you can use poly in that location if you want.

  13. user-988403 | | #13

    Martin, from your previous post I assume that you agree with me that the assemblies would be better of without the poly!? I agree that it might be enough drying potential but I definitely do not see the reason for that risk. The challenge I see with high performance walls is, that due to the use of multiple layers and materials simple rules of thumb according the perm ratings (e.g. outside 5x more permeable than inside for cold climates) or even dew point calcs looking at 1 location only, might not be enough. I would be curios to get the BSC guys opinion on the issue of permeability profile.
    I really appreciate your work on GBA

  14. user-788447 | | #14

    You are in a position designers sometimes find themselves in - having to move the project forward without having the definitive answers you want ; )

    I believe your specific scenario involves less risk than a true residence. Since yours is an outbuilding you will not be showering, cooking, or sleeping in the structure which are some of the activities that raise the humidity levels and influence the vapor drive from inside to out during the heating months.

    If you go with the poly maybe you don't want to fill the hut with plants in the winter. I think you said the hut was primarily intended for jewelry making. This should be verified but if this requires an open gas flame (or if you are going to space heat with an open gas flame) this could potentially produce humidity in a somewhat tight space.

    Because you are running an ERV during the winter you will be retaining much of the humidity you may be producing will in the hut.

  15. GBA Editor
    Martin Holladay | | #15

    Here's why I wrote that Minneapolis can install polyethylene in this location if he wants: he is choosing to install it directly adjacent to his polyisocyanurate insulation. Assuming he is using foil-faced polyiso, the foam is already a vapor barrier, so adding polyethylene beside the polyisocyanurate doesn't change the drying characteristics of the wall. I'll admit that my first response misunderstood the planned location for the polyethylene; that's why I originally raised a warning flag.

    In fact, his plan is similar to a PERSIST wall, except he wants to use polyethylene instead of peel-and-stick. I'll add my usual advice to those building a PERSIST wall -- it's usually better to resist the temptation to fill the stud bays with fluffy insulation. The fluffy insulation just makes your plywood sheathing colder. Empty stud bays are better from a building science perspective.

  16. user941025 | | #16

    Thanks, everybody.

    Philipp, thanks again for the caution. Safe advice is sage advice. But, weighing over my options and where the project is at, I'm going to go ahead with trying the poly.

    Martin, I may well choose to avoid batts, especially in the ceiling (where the balance between inside/outside the poly is less favorable).

    J, there's no gas-anything in the hut. FWIW, according to the documentation for the ERV, below a certain outside temperature it's exhaust-only, or at least it runs through exhaust-only segments in its cycle.

    You are in a position designers sometimes find themselves in - having to move the project forward without having the definitive answers you want ; )

    Yep! I've been in that spot for many moons. It's been a great learning experience. And thanks again, everybody.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |