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
Building Science

Ceiling Air-Sealing

Strategies for creating super-airtight ceiling assemblies

Of the thermal imaging photographs I’ve taken over the years, some of my favorites are of ceilings with incomplete air control. Sometimes the problem is where the ceiling air barrier meets a wall. Sometimes it’s holes cut in the ceiling for lighting, attic access, or even worse, ducts running through an unconditioned and vented attic. Occasionally it’s the entire ceiling that is missing an air barrier. Here are a few examples:

A ceiling air leak along an interior partition wall
A poorly air-sealed attic hatch
A large portion of this ceiling is leaking air during a blower door test.

Air leaking into or out of an attic space can contribute to comfort issues and increased energy costs, and it can affect indoor air quality. In more severe cases, ice dams and attic frost can result. Ceilings are arguably the most important area to get the air-sealing right, but one of the keys is continuity—which at the ceiling means making sure the air barrier is connected to the air barrier on the walls and that all penetrations are sealed.

Poor air-sealing and minimal insulation at the eave was the cause of this ice dam.
Attic frost can be caused by ceiling air leaks and high interior relative humidity. This attic was vented.

Continuity of the ceiling air control layer is often poorly executed, and construction sequencing is a big part of the problem. When building a new home, the floor system is built, the exterior and interior walls are stood, and then the roof is constructed. The mechanical, electrical, and plumbing contractors come in to do their work. Often, providing an air barrier for the ceiling becomes an afterthought. This results in holes in the ceiling that require some product to bridge the gaps and fill the…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial

26 Comments

  1. thomaskansas | | #1

    Really great and thorough article Randy. I'm most interested in the full floor system decked with Zip for my next spec house. I think it should be simple for the typical framer to execute and gets the house dried in and air tight quickly. In my mind, if one was planning to stick frame the roof, the only additional material is the cost of Zip as ceiling joists would be needed anyway. I wonder if a load bearing ridge would be required as the joists and rafters can no longer lap or if there is a reasonable method to fasten them through the decking. I suppose that's one reason to use trusses. I'd be very interested to hear from those who have used this assembly.

    1. Malcolm_Taylor | | #2

      thomaskansas,

      The floor system is very similar to one in a house where the roof descends right down to the same point but the space is occupied, so no load bearing ridge would be necessary, as long as sufficient mechanical connections were use to secure the rafter to the floor framing below. That’s usually done by adding a plate above the decking to fasten the hardware to.

      Apart from cost, one weakness of using a full floor system is you add an additional set of rim-joists to the walls, which have to be air-sealed and insulated.

      1. thomaskansas | | #4

        Malcom,

        Sorry, I'm not tracking with your comparison to an occupied attic. I would think in that scenario the rafters and joists typically lap because the subfloor only extends to a pony wall framed within the attic. In Graphic Guide to Frame Construction I see rafters sitting atop a plate for the purpose of extra insulation but they've tied the rafters and joists together with a gusset. Do you mean that a plate is used to have enough wood to fasten some kind of special hanger to?

        As for the extra set of rim joists, I imagine Zip continuing up the side of them so it can be taped to the Zip "floor decking".

        1. Malcolm_Taylor | | #5

          thomaskansas,

          Yes, very similar to the section you posted, but rather than gussets, hardware to attach the rafters to the plate, or a reverse birdsmouth cut into the rafters there. If hardware, specifying that connection may at worst require engineering, but that's still cheaper and easier than making the ridge bearing and having to carry those loads down to the foundation.

          The additional rim-joists aren't much of an issue, especially if the primary air-barrier is the exterior sheathing, but rim-joists are the most complex part of a wall assembly to insulate and air-seal. To me the equation is whether that outweighs having to do the same with penetrations in the ceiling if you don't include the additional floor system above. It's tempting to include service cavities in both walls and ceilings without looking at what they will actually contain. If it's just two or three electrical boxes and some wiring, the extra effort involved might not be justified.

          A couple of things I like about the system is it follows the usual sequence of construction, so you don't have the framers coming back a second time to install interior partitions, and once done that ceiling space should make the house a lot easier to renovate in the future.

          1. thomaskansas | | #9

            Malcom,

            I like the reverse birdsmouth idea. It would resist the outward thrust but unlike rafters with the reverse birdsmouth merely sitting on top of plates, the heel of the rafter would be supported if the rafters stack above the joists.

            Good point about how many penetrations are really in the ceiling but I'm tired of the air barrier not being in place until further in the build process and relying on fixtures as part of it. I couldn't agree more about staying in sequence and the remodelability. Having a deck to build a roof on also seems like a big plus.

  2. climbing_carpenter | | #3

    How did Steve control moisture with the Zip ceiling? Assuming loose cellulose?

    1. GBA Editor
      RANDY WILLIAMS | | #6

      The Zip sheathing is a class II vapor retarder that is very airtight. Moisture moving from the interior into the attic would be by way of diffusion, very slow. The vented and unconditioned attic space can easily handle this small amount of moisture. The attic is insulated using a blown mineral wool product called American Rockwool.

  3. kyle_r | | #7

    When constructing a floor system as a service cavity, the only weight it would need to bear is the Zip, drywall, and any services run through it? Assuming no interior load bearing walls.

    I wonder what kind of span you could get out of a typical TJI in that application.

    1. Malcolm_Taylor | | #8

      kyle_r,

      It would be interesting to see how the code or an engineer would view that structure. Depending on whether the roof is trussed or stick-framed, it would also need to take the live load of people and storage.

    2. Expert Member
      Michael Maines | | #10

      A 9 1/2" TJI 110--their smallest option--with a dead load of of 10 psf (a rough estimate, rounded up) and a live load of 10 psf (from the IRC), with deflection limited to 0.6" (a typical engineering value, less than the L/240 that the IRC allows) can span 24'.

      The loose-blown insulation above the Zip sheathing mostly precludes walking on it once it's insulated. If there were a duckboard (walking path) it would likely be supported by the roof trusses.

  4. Expert Member
    Michael Maines | | #11

    Thorough article, Randy.

  5. steve41 | | #12

    Great article Randy. I've seen the 2x4 strapping detail previously. I'm curious about the wiring clearances with this detail: wouldn't this be a code issue unless nail plates were installed? Or would the romex need to be stapled flat against the axial face to be code compliant?

    1. Malcolm_Taylor | | #13

      Steve,

      Our code requires protection if the cable is less than 32mm (1.25") from the face of the stud it goes through, so what was shown in the two photos would be fine here.

      From what I see IRC E3802.1 says the same.

      1. steve41 | | #17

        Thanks Malcolm. Maybe I'm splitting hairs here or looking at this the wrong way still. If 14-2 Romex is .360 wide, and face stapled to 1.5" thick furring, the greatest that the Romex could be from the furring face is roughly 1 1/8". Of course, this doesn't account for the staple, so realistically, probably best case the Romex is 1" from the furring face.

        IRC E3802.1: Where run parallel with the framing member or furring strip, the wiring shall be not less than 11/4 inches from the edge of a furring strip or a framing member such as a joist, rafter or stud or shall be physically protected.

        1. Malcolm_Taylor | | #20

          Steve,

          I think we are discussing different situations. I'm referring to where the cable crosses a 2"x4", not where it is fastened to its side. I don't see the latter in any of the photos. The wires are either stapled to the sheathing or framing above.

          In the detail of the larger photo you posted the cable is fastened to blocking at the fixture, not a " furring strip or a framing member such as a joist, rafter or stud" where it might get damaged by fasteners when the ceiling material was applied. If the blocking was seen as an issue by a BI it would be easy to fasten the cable to the underlying framing or sheathing as the rest is.

    2. GBA Editor
      RANDY WILLIAMS | | #16

      Malcolm got the answer, I just want to add I had a conversation with our local electrical inspector before wiring began. He didn't have any issues with the detail, but you might want to have a similar conversation with your AHJ just in case.

      1. steve41 | | #19

        Thanks Randy- this makes sense to get pre-approval. I do like this detail.

  6. ayerswc | | #14

    Thanks for the great Article, Randy. For you or anyone else that has any insights. With regard to the pictures with the intello product being used with the strapping. What is the order of process for this type of application?
    -Drape intello or similar over the finished walls before the trusses are set.
    -After trusses, but before electrical, install the intello on the ceiling with staples, room by room, connect the previously draped material, tape seams.
    -Add the ceiling strapping.
    -install electrical
    Questions: With the strapping it looks like in addition to the ceiling, strapping of the same depth will need to have been also fixed to the top of the interior wall system to provide sufficient base for drywall to be nailed to along the top of the walls? I see that the strapping runs perpendicular to the trusses. I expect that the strapping needs to be 16 OC if using 1/2 drywall?

    My project is in Zone 4, Virginia.
    Thanks!

    1. Expert Member
      PETER Engle | | #15

      We used strapping for my current build. Our house has roof trusses that (mostly) span wall-to-wall, so there weren't many interior walls to deal with. The sequence was similar to what you listed:
      -drape tape or membrane over the exterior walls before the trusses are set. We used Blueskin because of its adhesive surface.
      -If there are going to be load-bearing interior walls, drape tape or membrane over them, too.
      -after trusses, staple membrane to underside of trusses, then build interior walls. A double top plate provides nailing for the drywall .
      -Install strapping @ 16" on center

      There is an additional benefit of strapping the ceilings: Securely fastened strapping provides support for the weight of insulation on the membrane. We used R-60 Timberfill with trusses on 2' centers and there was no way that staples would have supported it.

    2. GBA Editor
      RANDY WILLIAMS | | #18

      Peter used pretty much the same approach as what we did with the Intello install. We used a piece of 12" wide tape that was stuck from the exterior wall sheathing and over the top plate, installed before trusses were set. There were no interior walls, all the bearing was on the exterior walls. Once the trusses were set and we were dried in, we ran the ceiling Intello and connected to the tape. Strapping was then completed, then interior walls built. The tough part was the plumbing vent, trying to tape the penetration through the Intello with only 1.5 inch of space to get my hand into. I could have also gone into the attic to tape. I've since changed this detail; we are using a gasket for that and any other penetrations that need to move into the attic, it's easier to tape an oversized square gasket than trying to tape to a round pipe.

      1. Expert Member
        PETER Engle | | #22

        Randy,
        Some sort of gasket seems like a great idea for the penetrations. We had exactly the same issue with our pipe and wire penetrations. You try to limit them, but there are always a few. We tried to tape from below, but ended up doing the most effective sealing from the attic.

        But much of this mayhem brings us back to the idea of building an airtight floor system above the ceilings. With that easily accessible rigid decking on the floor, sealing around the inevitable penetrations would be much easier. It would also make construction sequencing much easier, since interior walls are built earlier in the process.

        1. GBA Editor
          RANDY WILLIAMS | | #24

          We are figuring out so many improved ways to deal with making a ceiling airtight, most were probably first used in the 70's but have been "rediscovered" more recently. I do like that airtight floor system below the roof, especially for slab on grade homes.

    3. Expert Member
      Michael Maines | | #21

      Just to point out what is hopefully obvious, if the draped membrane has any chance of being walked on, secure it well or use an adhesive-backed membrane.

      1. Malcolm_Taylor | | #26

        Michael,

        Our code now mandates solid backing for any seam or penetration in a membrane (usually poly here) air-barrier. So blocking at vent pipes, seams aligned with stud locations, and for ceilings the draped membrane and that covering the ceiling are brought down and lapped on the upper top-plate of the walls.

  7. OverYonder | | #23

    I have a project going with this detail. SE specified 9 1/2" TJI 110, 24" oc. There will only be foot traffic during construction but my builder is concerned about the flex at the sheathing end seams compromising the air seal tape or flashing caulk. He has some big guys on his crew. He is suggesting a T&G sheathing such as Advantech XFactor or blocking the seams - his preference being the XFactor. We won't be able to do a blower door test until after the insulation sequence and a membrane is installed on an adjoining vaulted ceiling so I'd like to avoid trying to tune up this floor system for air leaks late in the project. Does he have a valid concern?

    1. GBA Editor
      RANDY WILLIAMS | | #25

      You can always use a plywood clip and move to a 5/8 standard sheathing product. The clips do bulge the tape, possibly damaging the tape if there's enough foot traffic. Repairs are pretty easy; someone needs to take the time to inspect after the roof framing is completed and retape the damaged areas.

      If you have the budget for XFactor or even a standard T&G subfloor sheathing, you won't be using the clips but still have the potential for foot traffic damaging the tape.

Log in or become a member to post a comment.

Related

Community

Recent Questions and Replies

  • |
  • |
  • |
  • |