How to do a drywall air barrier with exposed beams?
This is for a new house to be constructed hopefully starting Spring 2019. We want to have a drywall job that provides an excellent air barrier preventing exfiltration of interior air into the ceiling. In a regular ceiling I’d think a continuous plane of drywall would do the job. However this is a cathedral ceiling built on timber kingpost trusses that are exposed to the interior. In this case the ceiling drywall is not continuous but broken into sections by the exposed timbers? Any suggestions how best to get a solid air barrier? Is the only way to do this a tedious caulking-gluing-sealing job between the timbers, drywall and other framing? Any other great ideas?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Dennis,
Can you describe the structure a bit more? What is the spacing of the timber trusses, and what is above them?
Perhaps not practical, but maybe you could apply the drywall and then spray foam from above.
Thanks for the comments so far.
Jon, I'm not sure how to apply insulation, especially spray, once the drywall is in place.
Malcolm, I've provided a clip of the current structure design. The timber trusses are about 4 feet apart. The house is a double stud wall. In order to get enough cavity for some serious cellulose in the cathedral ceiling, the roof structure is somewhat separated from the ceiling structure. The rafters are supported by a roof beam and the outer stud wall. The timber trusses support the roof beam as well as the ceiling structure and rest upon the inner stud wall (carrying the ceiling load as well as half the roof load). The outer studs are taller than the inner studs, but the walls are connected for fireblocking with 2x4 and plywood as shown (or some other better idea).
But I think my question really applies to any situation where someone has a ceiling with exposed beams. What are some good methods to ensure a good air barrier when the ceiling plane is segmented?
Dennis,
As a general proposition, the most effective way to seal a roof with exposed structural members is to have a continuous plane of material above them. That's in part why many timber frame houses have T&G or plywood roof decking which can be fastened to the tops of the trusses.
In your case I'd agree with Brian that sheathing and taping the top of your 2x4" strapping would be the safest and most effective approach. You could substitute rigid foam, but I don't know how you would then seal it and build the upper-roof structure. There do you stand to do that work?
I would try to get some sheet good (OSB, plywood or membrane) above your timber beams and tape the seams on that
Sealing drywall along long seams is a pain (speaking from personal experience). If you want it to last, it has to be very clean even joint, need to put a backer rod into the gap then caulk. I did most of this work myself and even then had to redo section where the joint failed after a couple of years (no matter how careful, human nature is to take shortcuts and will bite you in the $%^). The probability of this being done properly by a contractor is near zero.
Akos, thanks for your thoughts.
I've been under the impression that the standard tape and mud joints in drywall serve as a good air barrier and that it's only the edges of the entire drywall plane that would need further attention. Are you saying that every single joint between every sheet requires sealing treatment to the framing?
Along the lines of your suggestion, I have thought maybe to put a 1" taped layer of XPS above my drywall nailing framework, leaving a few small openings for blowing in cellulose, then closing up and taping those openings. This would make the XPS the primary air barrier rather than relying on the drywall layer.
Or instead of XPS just put down a 4 or 6 mil sheet of poly, with a few openings to blow cellulose and then seal them shut. I would think poly or XPS is OK if the ceiling can dry to the exterior, and in this design there is a 2" ventilation space under the roof sheathing.
I’d use polyiso and not XPS for this. Polyiso with a foil backing will be the easiest to tape. You could use 1/2” too if it’s intended mainly as an air barrier and not really for insulation.
Poly sheet would work too. You’d still have to detail everything at the edges which might be tricky if you have timber framing. You could caulk the poly to the timber pieces, that might be better than trying to tape it.
I’d agree that the drywall with taped and mudded seams is a good air barrier. The tricky part will be sealing any uneven edges or edges butting against rough surfaces like timber beams. The best/easiest option I can think of is to hit all the gaps with spray foam from above if the roof is open but that obviously won’t work if you only have access from below.
Bill
Dennis,
The problem is a the joint of the drywall to your timber beam. These are the ones that are a pain. Mudding there doesn't work as it will crack as the building moves, even a 1/16" crack over a 20' of length adds up to a very large hole.
Mud and taping air seals joints between drywall sheets without problems.
Foam/osb/plywood above your nailing framework would work great. I would stay away from poly as it is very easy to install badly. Make sure the joints get taped with a quality tape. I would vote for wood instead of foam as workers can easy walk on it without chance of damage.
Also be careful with the joint to your wall air barrier and the ridge beam (especially the ridge beam). Ridge is where I would go with the belts and suspenders approach and have at least two redundant air sealing mechanisms.
Edit. Just looked at your drawing and doesn't look like you have an exposed ridge beam, so just watch out for around the king post.
Hi Dennis,
Airtight drywall is a legitimate approach to air-sealing, but it is much more involved than hanging and taping drywall panels. Below are links to an article and a video series that will offer some good how-to information on the process.
When I saw your detail, I also wondered if you could sheath the top side of the timbers. I thought of ZIP System sheathing because I have seen it used for an air-tight ceiling in other projects, but now I wonder if some builders are just enamored with ZIP. Using taped rigid foam as you suggested would add R-value. In either case, you'd have to carefully tape the joints and you'd have some detailing to figure out where the rafters meet the bottom chord and the king post, but that doesn't seem any more or less troublesome than trying to detail the air-tight drywall installed between the timbers.
Let us know what you decide to do and here are the links:
https://www.greenbuildingadvisor.com/article/how-to-hang-airtight-drywall
https://www.greenbuildingadvisor.com/article/new-video-series-airtight-drywall
Thank you all for your comments. I think it’s pretty unanimous that the best way to go is a continuous plane of “something” and not rely on a group of disconnected drywall segments.
Malcolm rightly pointed out that we can theorize all day about using various materials but that there needs to be a practical way to build it. This is one thing that still troubles me a little and will require more thought – the feasibility of building it and then insulating it. Not impossible, just takes planning. Since I intend to do a lot of the work, that may make it easier -- some methods might be feasible for me that a tradesman doesn’t want to fool with.
For rigid insulation, I think Zephyr7 has the right idea to use polyiso for the better tape joints.
And thanks to Akos and Brian for stressing the importance of airtight techniques and potential trouble spots to pay extra attention. Akos is so right about the joint between drywall and timber. We built a passive solar timber frame in the 1980s (back then Tyvek was cutting edge, the crews that helped us had never worked with it before, nor seen exterior basement insulation). The timbers were not kiln dried and over the years there was plenty of shrinkage where they pulled away from the drywall and needed a little filling in. So I am no stranger to seeing cracks form along drywall and wood and don't want to repeat a scenario that might allow damaging exfiltration. The videos linked by Brian are helpful. I know I’ve seen them before but it was even more enlightening to see them again and be refreshed in what I’d forgotten.
At this point I’m leaning towards a taped continuous layer of polyiso above the 2x4 layer as the primary internal air barrier and also will do my best to seal the drywall layer since redundancy can be good. But before finalizing my plan I think I should also check with the cellulose insulation guy to make sure I’m not creating an impossible situation for him/her.
BTW, this house will be in Lancaster county, PA in climate zone 5 just barely.
Seems to me that the extra materials to do it this way, would be less than all the labor trying to get a quality sealing job another way.
That's good point. Especially as the continuity of the air-barrier at the end of the trusses and the exterior walls is unclear.
OK, so now I'm a little confused. I have been talking about an internal air barrier that prevents moist internal air from entering the wall and reaching a point where it might condense. I assumed I would also have an exterior air barrier, basically what is shown in Trevor Chadwick's drawing, to prevent external air from blowing into the insulation and undermining its R-value. Am I thinking wrong about this?
I also was originally thinking only drywall on the ceiling so that the ceiling could dry inward. Putting a layer of polyiso or XPS would mess that up. I was intending the ventilation channel would be 2" wide spacers on the rafters to which strips of thin OSB or plywood are attached and the whole thing sealed.
If you were doing the ceiling without the timbers breaking things up, wouldn't you just make a giant drywall plane in an airtight manner, have an exterior air barrier (per Chadwick dwg) and be done with it? But I've complicated things by allowing the timbers to be penetrations? And that's the issue I'm trying to resolve.
The drawing showed a vent channel between the insulation and sheathing I didn't see any exterior air barrier, and assumed you where only going to have one
I'm sure others will chime in that know far more than I do, but I don't think you need an air barrier on both sides of the insulation in your climate with dense cellulose
Deleted
Trevor,
You're right that I might not need so much air barrier. Actually it seems to me that the ventilation channel is comparable, or even better, than the typical flat vented attic with insulation on the attic floor. The insulation in the attic just lays there and is not enclosed in an air or vapor barrier. So maybe the ventilation channel uncaulked is just fine. I don't know. But one thing I was hoping to prevent was the moisture from humid summer air condensing in the insulation as it gets near the air conditioned ceiling. But again maybe I'm over-thinking this -- if it's OK that the insulation in the attic doesn't have air and vapor barriers over it then why should my assembly need it?
Deleted
I've pretty much decided on the Trevor Chadwick approach in post #11. It's always good to let ideas evolve. I started out with the idea of vent baffles in each rafter bay, which is a common practice. GBA has articles about doing this using spacers nailed inside top of the bays then attaching strips of some type of sheet material (XPS, OSB...) from below to fit between the bays. Caulking would be required to get air-tightness.
But if I'm already using sheathing and wood spacers (probably ripped 2x4s), then Trevor's idea makes a lot of sense. Rather than nailing spacers inside the bays and fitting slices of sheet goods and then trying to seal each joint, instead nail those sheets full size to the top of the rafters. This makes a nice flat surface on which to continue the house wrap, and this eliminates all that caulking work (although maybe I should tape the sheet joints just like on exterior wall sheathing?). The spacers are now nailed into the rafters sandwiching the house wrap and sheet material in-between. The roof is built on top of the spacers. There is virtually no extra material involved, far less labor in fitting pieces together, far less labor in getting a good air seal, and most certainly a better air and weather-tight solution.
Maybe I worry too much, but this roof will have ridge vents and I imagine occasionally some small amount of rain from gusty thunderstorms (or snow of the drifting variety) will make it through the ridge vent and into the ventilation channels. But with the proposed method I'm much more confident that invading precipitation will not get into the insulation but instead find another way out.
SIPs are certainly one approach and may make sense to some people based on budget and schedule. Don't SIPs use foam insulation? Some people prefer not to use any more foam insulation than necessary since it is not as "green" as other insulations. In my case I'm doing most of the work so don't need a solution that needs to go up quickly to save on labor costs. The structure that the GBA readers have helped me visualize seems to be the best price point for me.