Insulating and Venting a Cold-Climate Cathedral Ceiling
I’m building a high performance ICF home in Zone 5 with a single slope mono pitch (3:12) roof and a cathedral ceiling. The Roof layers are currently as follows:
Standing seam metal
5/8 Zip Roof Sheathing
We were planning to have an unvented assembly with 3″ of ccSPF applied directly to the underside of the sheathing (technically about R21, not counting the thermal bridging of course), followed by open cell or mineral wool to get us to R49 or R60 (we have very tall parallel chord trusses with plenty of space for insulation…and the downside that there will be more thermal bridging since the bays will never be fully insulated, although we are considering encapsulating them with a thin coating of ccSPF to help with that). We are planning to use a T&G wood ceiling attached to the bottom of the trusses with no drywall above it – we believe this should work because the air barrier is the 3″ of ccSPF on the underside of the roof, so we’re basically creating an unvented, conditioned “attic” in the 12″ or so space between the T&G ceiling and the SPF on the underside of the roof (please let me know if anybody disagrees and sees a problem with that design).
The roof will be light colored, fairly low slope and facing due north, in a high snow area in Zone 5. We will have 2′ overhangs at the ridge (high side) and 1′ overhangs at the eave (low side). While I believe that this assembly should technically work, I’m thinking that with such a simple roofline (no dormers, skylights, etc. – only roof penetrations will be 1 wood stove pipe and 2 PVC plumbing vents), it might make sense to include a ventilation channel below the sheathing to allow for some drying potential in case it were to ever get wet. With 3″ of ccSPF below I’m not too concerned about air or vapor leakage from the house, but eventually every roof leaks and I’m concerned that the sheathing would have virtually no potential for drying and the leak may go undiscovered for a long time increasing the risk for catastrophic failure. It might also help with ice damning (not a huge concern with standing seam) and if I ever need to repair or replace roof panels (and sheathing below), we wouldn’t have to tear out all of the insulation.
We’re considering adding vents in the soffits at the eave and ridge and building in ventilation channels (probably at least a 3″ space since we have plenty) all the way from the eave to the ridge using rigid foam or maybe fanfold stapled to the sides of the rafters and then keeping our plan of using a combination of 3″ of ccSPF applied directly to the back of the ventilation baffles, followed by air permeable insulation to get the rest of our R-Value.
I’m wondering if there are any downsides to adding this vent channel? It is a mixed-humid environment in summer, with high outdoor humidity levels which could unnecessarily introduce moist air to the underside of the sheathing in the vent channels. We also have a very tall detached garage very close to the house on the eave (low) side, which might block a lot of potential for wind exposure from that side which would limit the usefulness of the vents. I’m also concerned about truss uplift caused by basically putting all or part of the top chord in the ventilation channel outside of the building envelope while the rest of the truss members are inside the envelope. On a similar note, I’m concerned that the thermal bridging through the trusses might be made much worse by exposing the top chord and bottom of the roof sheathing to the exterior air in the vent channel rather than keeping them inside the conditioned envelope (at least inside the foam) so that the cold has to conduct from the air to the roof, through the ice and water shield, through the sheathing into the trusses).
I would greatly appreciate your thoughts on these options and proposed assemblies. Our main goal is to build the house in the most durable and energy efficient manner within the confines of our mono-pitch roof design.
Happy to provide drawings or answer any questions if I’ve omitted any details that we would help.
Thanks in advance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
The roof type that you are building is by nature unvented, so vent channels do not make much sense. Hybrid assemblies often get kind of weird.
You want the closed cell spray foam applied directly to the roof decking, do not leave an air space at this location.I would also be spray foaming the transition from wall to ceiling, so the foam is continuous from the roof decking to the top plate, to keep the air barrier continuous.
As far as future proofing against future leaks, this is a downside of this type of roof. Repairs will be costly. But if constructed well I do not see why you should have any concerns about leaks for many many years.
It's safer to assume about R6 per inch of closed cell spray foam, and also assume you will have some valleys in the the installation that will be less than the ideal thickness. For these reasons, if it's always safer to go a little over on the spray foam to avoid potential issues down the road. I like to shoot for a target of 50/50 on the ratio of closed cell to "other stuff" for moisture control, which would mean about R20 worth of closed cell spray foam for an R49 assembly. R20 is almsot 3.5" of closed cell spray foam, I'd try for 4" or a little more. For an R60 assembly, you need about R30 worth of closed cell spray foam, and I'd probably target a 6" layer in that case to be safe. The rest of the insulation can be open cell, mineral wool, pretty much whatever you want. The spray foam would be your air barrier, just make sure you haven't missed any spots, especially around the soffit area.
If you want to put in vent channels that should be fine, but then you have a vented roof and you need to go all the way -- vent channels of the full rafter bay width in every rafter bay (ideally, although you could spray foam only some unvented rafter bays, but I don't see a point). You don't want to do partial vent channels or other oddball stuff. Go all vented, or go all unvented. You can build the vent baffles with 1/4" waferboard and 1x2 furring strips as spacers. This is cheap and goes fast when installed with a finish nailer. You can then insulate with whatever insulation you want, but you WILL need an air barrier behind your T+G ceiling if you aren't using spray foam. I have in the past used 1/2" polyiso behind the T+G as a combination air barrier and thermal break, so you don't necassarily have to go with drywall.
I would try for a vented assembly if possible as they tend to be safer. You may be able to do loose fill insulation if you have enough clearance above the ceiling so it's worth checking into that.
Bill
Bob and Bill,
Thanks for your respective comments.
Bob, I'm not sure why you say that all mono pitch roofs/cathedral ceilings are by their nature, unvented. There are plenty of details for vented cathedral ceilings out there, including in Martin's article on GBA about how to build insulated cathedral ceilings (https://www.greenbuildingadvisor.com/article/how-to-build-an-insulated-cathedral-ceiling). Thanks for the recommendation on foaming over the transition from the wall, we will definitely be doing that too keep a continuous air and vapor barrier.
Bill, Thanks for your recommendations on foam thickness. I had thought about a lot of that - the foam we're using has a published R of 6.7 after 90 days cure, so 3" is R20. 1. That's the absolute minimum needed to meet the ratio, plus factoring in thermal bridging, imperfect application thickness, etc., I was also leaning towards switching to 4" of closed cell, which I will probably do it we don't vent (I think we're going to go with R49 total). If we decide to vent, I was thinking that meeting that ratio which is prescribed for unvented roofs is less critical so I might save a few bucks and just stick with 3" of cc and do the rest with oc. Although I think technically I could use all open cell or other air permeable insulation if we vent, we would still do a hybrid approach - definitely want an air/vapor barrier even with the air gap between the foam and sheathing and don't want to worry about detailing a drywall ceiling because we need fire sprinklers and have a bunch of recessed lights (flush not cans, but still) and some remote in line exhaust fans and ducts that we'll be running between the finished ceiling and the bottom of the foam insulation that I'd like to be able to access in the future by unscrewing a few T&G boards and not messing with a drywall ceiling - I'm assuming you think that works as long as we use the SPF? I like the polyiso suggestion but it would still limit my access to remote vent fans that I'm planning to locate in that void. Do you agree that the ratio is less critical if I'm not spraying directly onto the sheathing?
I am leaning toward venting, in which case we will attach baffles the full width of each bay leaving a 3" vent channel all the way from eave to ridge, except possibly for a few bays where the vent channel might not be able to be clear, in which case we would spray directly onto the sheathing in those bays (with 4" of cc) and make sure that the transition back to the vented bays is continuous. My insulation guy is recommending stapling up Solex insulation (without furring strips) for a quicker installation - he uses a high volume low pressure (HLVP) sprayer that goes on gently and shouldn't blow out the Solex even without furring strip backers (he'll just attach to the sides of the truss chords). He will make a very thin initial pass to give it a little more rigidity and then do the full spray.
I would normally consider loose fill but it's challenging with a cathedral ceiling and avoiding future settlement, plus again I'd lose that "conditioned" void above the ceiling where I will have some fans and ducts.
I'm glad to hear that you think venting is the best approach. I'd be very interested in thoughts about the 4 specific concerns I described in the next to last paragraph of my original post:
1) It is a warm-humid environment in summer (with cold, snowy winters), with high outdoor humidity levels which could unnecessarily introduce moist air to the underside of the sheathing in the vent channels.
2) We have a very tall detached garage very close to the house on the eave (low) side, which might block a lot of potential for wind exposure from that side which would limit the usefulness of the vents.
3) Potential for truss uplift caused by basically putting all or part of the top chord in the ventilation channel outside of the building envelope while the rest of the truss members are inside the envelope.
4) Possibility that thermal bridging through the trusses might be made much worse by exposing the top chord and bottom of the roof sheathing to the exterior air in the vent channel rather than keeping them inside the conditioned envelope (or at least inside the foam).
Any thoughts on these or other potential downsides/risks of venting?
Thanks again everybody.