Cathedral ceiling in octagon-shaped home
I live in an octagon shaped house with a cathedral ceiling over the second floor. 3/4 of the interior space on the second floor is open and the ceiling is a little over 1,000 square feet. Four sides of the roof/ceiling are rectangular, while the other four sides come to points. The house has a small third story room at the top with four walls where the four pointed corners come together. This room has a flat ceiling but the roof for it is pointed so I assume it is an insulated attic but it has no access or ventilation. The house has a metal roof (about 10 years old) with exposed fasteners on top of a layer of shingles on top of a deck of plywood. The cathedral roof assembly on the second floor is failing. I bought the house three years ago thinking a metal roof was great.
The rafters are 2 x 8 so that is only 7 1/4″ of space for insulation. The insulation is fiberglass, most likely from when the home was built in 1980 and has evidently seen years of condensation under the deck. The ceiling had cheap plywood planks that we ripped down, and it was hiding drywall with cracks in the joints for warm moist air to do its damage. We noticed discolored leaks coming through the ceiling one winter but found out later it wasn’t a roof leak or ice dam but rather condensation from a new humidifier we installed. The soffit is vented but the roof assembly has no air channel to bring the air up and out. I believe the roof geometry and lack of space is too complicated to vent. The second floor bathroom vented directly into a crawl space under another portion of the cathedral ceiling, with leaky drywall installation. Having opened up some of the ceiling to take a peek, the deck is literally crumbling from years of condensation. Much of the roof is spongy if you walk around on it.
Even though the roof doesn’t seem to be leaking from rain or ice, I know the decking is rotten and I don’t believe I can fix the insulation, air, and drywall problems without ripping off the whole roof.
After reading a lot about insulating an unvented assembly, my plan is to put a white colored standing seam metal roof on, with ice and water recommended by the contractor for the whole roof. I’m assuming all or most of the decking will need to be replaced and am planning on using 1/2″ plywood. I don’t have enough room to put R-20 rigid foam on top due to some low windows on the third-floor room so I need to insulate underneath. I also do not want to lower the ceiling to increase the area for insulation. The vented soffit will need to be replaced or closed off in some manner. Thus my original plan was to install closed cell spray foam. I live in Zone 5, so I need at least R20 of foam on the underside of the deck and the rest could be more spray foam or insulation batt to get to R38. Because my ceiling is fairly large, I don’t think I can follow the R30 rules for cathedral ceilings. The building inspector said this was a good plan and that I would need to cover it with a thermal barrier. I’m now questioning this approach due to the lack of drying direction as well as needing drywall instead of tongue and groove boards.
If the metal on top and the underlayment doesn’t allow water that might eventually get on the plywood to dry outward, and the closed cell spray foam acts as an air barrier and possibly a vapor retarder, am I not dooming this deck to rot since it doesn’t seem like it could dry to the interior either? You might be able to argue that the rafters could allow for some drying direction just as you have thermal bridging issues with the rafters, but one likely has nothing to do with the other.
I know that open cell foam is considered risky for this situation, so I haven’t considered it. After reading about “smart” air barriers with vapor control, such as Intello and Majrex, it seems like it would be safer to install something like Rockwool batt with a smart air barrier that lets moisture escape inward. I know this wouldn’t allow me to bring the insulation up to code (R-30 Rockwool is 7.25″ thick) but having enough insulation seems pointless if the deck traps moisture and could rot out again.
All that said, would it be best to use all closed cell spray foam, mixture of closed cell + air-permeable insulation + drywall, or air-permeable insulation + smart air barrier with vapor control + drywall? If drywall is used, do I still need a vapor retarder paint with the option suggested?
The building inspector may not require spray foam due to the high cost but regardless, I don’t want to cause the same problems for me or someone else down the road. I also realize that there is the additional cost of putting up all new drywall, if using tongue and groove planks do not exempt a thermal barrier over closed cell spray foam.
There seems to be a lot of confusion about whether or not solid tongue and groove lumber is allowed. Michigan code R316.5.2 Roofing says a thermal barrier is not required, over foam plastic, if it is covered by tongue and groove planks, but the wording is muddy and makes me think the tongue and groove planks need to be engineered with glue, like plywood or engineered flooring, which doesn’t make sense.
If I only use closed cell spray foam to get to R38, which acts as an air barrier, can I leave out the gypsum board and just hang solid tongue and groove boards? I’ve read that tongue and groove ceilings are leaky, but if the foam is the air barrier, does it matter?
Lastly, would it be helpful, if I could afford it, to add a radiant foil layer on top of the underlayment on the deck, with 1x purlins opening up an air space under the metal roof? The roofer said he has done this for others to help cool the roof, but I’m not sure it would be worth it, or help with drying outward. The second and third floors are hard to keep cool in the summer, so that may be a good reason to do it unless the spray foam solves that problem.
I just don’t want to make a huge 25K plus mistake and would appreciate any help the community could provide. Thanks for your time.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Without a picture of the roof, it is hard to make specific recommendations. By the sounds of you can vent the rectangular sections with ridge vent like this at the base of the 3rd story popup:
https://www.pac-clad.com/wordpress2017/wp-content/uploads/2017/02/SS-Ridge-Vent-Slope-to-High-Wall.svg
If you strap out the whole roof, it should create enough air path for the triangular sections to vent into the square section to do a top vented roof assembly. This would mean your roof underlayment would need to be vapour open, definitely something other than ice and water.
Unless your roof is low slope, there is very little benefit for peel and stick over the whole surface, a standard synthetic underpayment would work.
Close cell spray foam might be simplest but more costly option, section of my roof is with closed cell SPF under the deck with ice and water on the top with metal roofing, no issues with (zone 5). The one thing to watch is that the roof deck is dry before the spray foam goes on.
With T&G and no drywall, you need either closed cell SPF or lot of closed cell SPF with a bit of permeable insulation. With your rafters something like 3.5" cc SPF+3.5" mineral wool for thermal barrier and T&G over that. The cost of this might be similar to going with open cell foam + Membrain+ drywall though.
Unless your inspector requires R38, there is not much energy savings to go up to that R value. No matter what you put between the rafters, the thermal bridging of the wood limits the R value of the overall roof assembly, the R value of the SPF would be wasted. With 7" SPF you would end up with an overall roof R value of ~R26 VS R20 with mineral wool batts/open cell SPF.
If you don't mind loosing a bit of ceiling height, your best value and performance would be to go with a top vented roof, full height mineral wool batts between the rafter bays with 1" of foil faced polyiso under the rafters. You can tape the seams of the polyiso to air seal the ceiling and install the T&G directly over the foam (2.5" finish nails should hold through the foam). This would give you an overall R26 roof. If you don't mind strapping out the foam to install the T&G, going up to 2" thick foam would be even better.
Thanks for your input Akos. If I could vent the top, what you say makes a lot of sense to me regarding drying outward. I've attached a picture of the house. I think my main problem is going to be those windows being framed so close to the roof, for the system you referenced. It doesn't mean there isn't another way to do it however.
I didn't even think about putting a small amount of foam under the rafters, acting as insulation and an air barrier, but I don't really want drywall as my finished ceiling anyway. If the building inspector allows it, do the nails holding the T&G go through the foam and into the rafters? I'm assuming the foam can't hold the T&G by itself. The previous homeowner nailed plywood sheet to the drywall, with glue, and it sagged and started failing. If I go with closed cell spray foam, the contractor says they can get it to 3 1/2" thick at R24.5 or 5.42" thick at R38. It sounds like you are saying that might not be accurate.
For low windows and metal roofs, what I've done is stop the metal roof a 8 to 10 inch bellow the window. The slope of the roof over this distance adds enough height for metal work. Looking at your roof, this might not still not be enough to be able to double strap. If your roof deck is not totally rotten everywhere, it might be worth the cost of the closed cell spray foam to not have to touch the metal roof.
The problem isn't the R value of the spray foam, the numbers your installer quoted are for the foam itself. The problem becomes when you put that foam between wood studs. The wood rafters conduct a fair bit of heat (a 2x8 rafter is only R8.5), so the overall R value of the whole assembly (studs+foam) is much lower than the foam itself. Generally, it is not worth it to have more than an inch or two of close cell spray foam between studs, more than that, the R value of the foam is wasted.
For nailing up the T&G, you need to hit the rafters. With 1" foam and 5/8" T&G, a 2.5" finish nail should get around 1" embed into the rafter, it should hold well. I've nailed up fair bit of T&G with finish nailer and never had issues with sagging even with 24 OC rafters.
The most breathable roof assembly that fits your plans is this unvented compliant design:
metal roofing
some venting, even if not to vented roof specs. Consider mesh.
a permeable (preferably fully adhered) underlayment
plywood
strips of polyiso foam on bottom edge of rafters (or cross members)
open cell foam filling the extended rafter bays
Intello or MemBrain
T&G wood
Even non-vented metal roofing isn't airtight and so this will dry upwards at the equivalent of about 1+ perm. And it will dry well to the interior.
Is this superior to closed cell foam and impermeable underlayment - I believe so, but don't know of data matching this exact case.
Also consider less than typical EPS foam above the sheathing (continuous or between purlins). This is fine if your assembly meets specs without the EPS.
Thanks for the follow-up Akos, and for your help too Jon R. You've both given me more to consider. I was starting to feel a little paralyzed by all the options. I knew I didn't want to add more lumber to effectively drop the rafters but an inch of rigid foam might be just fine. I have to decide on the rafter insulation before I lean towards the taped sheets of polyiso, or strips of polyiso with Intello/MemBrain ideas. I didn't even know what top side double strapping was but when I read about it, I also read why you don't want to do single strapping under metal roofs due to condensation and rotting. Ugh...
I'm reluctant to use open cell because if there is a leak in the assembly, it's my understanding that the open cell will swell up. I believe there have been other mentions on this site about open cell being risky in roof assemblies too but it is tempting because it would breath more. I'm pretty sure my building inspector will also require gypsum if I do tongue and groove over foam unless there is a thermal barrier. I'm already reluctant to do spray foam because of cost and worry about curing correctly but even more so due to what Akos said about the thermal bridging. That said, I really like the idea of using full fill mineral wool batts with the polyiso + T & G if I can vent the roof with strapping or use mesh to help dry the deck.
Jon R: What is the mesh you are talking about specifically? I've read about cedar breather and some really expensive underlayment with mesh in it.
Akos and Jon R: Do either of you have a recommendation for underlayment that is permeable (with full adhesion or not)? Are we thinking of something like Vaprosheild or just plain old 30lb felt?
This is Martin's list from another post:
"In that case, you want to choose a vapor-permeable underlayment like Perma R Products PermaFelt, GAF Materials Deck-Armor, Cosella-Dörken Delta-Maxx Titan, VaproShield SlopeShield, Cosella-Dörken Vent-S, Nemco Industries Roof AquaGuard BREA, or Cosella-Dörken Delta-Foxx."
I've used the GAF product and felt. Felt is fine provided your roof is strapped, the heat from the metal roof tends to melt the tar causing to stick and tear as the roof panel expands. Felt is also easy to rip, so roofers prefer the synthetic ones.
My gut feel is the breather type venting will not allow enough air movement from the triangular section. Maybe if you go with roof panels with larger ribs, this might be less of an issue. See if you can get the roof double strapped first though.
When it comes to roof durability, the most important part is fixing your ceiling air leaks and disconnecting your humidifier. Getting that bath fan vented outside is also a must. An air tight ceiling with a bit of roof ventilation will last, no amount of ventilation can save a leaky ceiling.
The humidifier bringing up the winter time humidity is definitely not helping. If your winter humidity is low, generally means the house has air leaks. Best to seal those up first. Once you fix the ceiling, blower door directed air sealing is cheap and will easily pay for itself in energy savings.
See this video for a Matt Risinger recommendation:
https://www.youtube.com/watch?v=v0w21Bg1HJI
The bathroom fan was one of the first things we got fixed. The Mrs. won't like the lack of a humidifier but we know from experience how much of a problem it can be if cranked up. Thanks again for taking the time to read through the postings and sharing what you know Akos and Jon.