Unvented 1920s attic insulation dilemma
Hello all. I have an old 1920s unvented roof in Minnesota. Before purchasing the home, there was no insulation at all. The walls and basement are fairly straightforward. I am having issues with the attic, which is unconditioned space used solely for storage. Do to several factors, just insulating the floor of the attic is not an option.
The rafters are 2×6 and the floor joists are 2×12. Currently I used mineral wool batts tight to the sheathing and then overlaid aluminum faced poly iso on the bottom side of the rafters. This is a radiant heated home with original single pane widows so it’s fairly leaky with a very low existing relative humidity.
Would the poly iso stop the vapor from condensing on the bottom edge of the sheathing?
Or to be safe, should I remove everything and install a site made baffle to allow a 1” air space between the mineral wool and the sheathing, even if it isn’t vented? Would these baffles, while reducing the insulation thickness help overall by preventing wind washing through the old sheathing (boards)?
Unfortunately I have already installed the mineral wool / poly iso method, but knowing if I should fix this now and save myself from rotten sheathing in the future would be good.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Hi William,
The foil faced polyiso rigid insulation is a class I vapor retarder and can be detailed as an air barrier. Both are helpful to reduce the likelihood of condensation on the sheathing, and the continuous insulation is helpful for mitigating thermal bridging. That said, fibrous insulation in the unvented rafter bays is not a great idea. If you are willing to take down the polyiso and start again, you have options for both vented and unvented assemblies. The good news is that you should be able to design an assembly that reuses the insulation you already have. I suggest that you read this: How to Build an Insulated Cathedral Ceiling
Thank you Brian. I read the article and am still a bit confused. It was my understanding that cut and cobble vents in an unvented roof would still not be a good solution. Is the only way to remove and use CC? Right now, this approach would be cost prohibitive as estimates ran $11-20k just for cc flash coat.
I mentioned this in another reply. Would adding venting above the existing sheathing alleviate this issue when we reroof (shingles are roughly 10 years old).
If you are decide to wait for your reroof, or to reroof a bit early, consider exterior rigid foam insulation. Then you can insulate the unvented rafter bays with any material you'd like, including keeping the fiberglass.
As Dana pointed out below, if you keep the polyiso on the inside, your roof cannot dry inward, which some people believe to be a problem (the moisture trap). Others see less of a problem as long as things are dry when the work is done and roofing and air sealing details are diligently executed. I tend to be cautious and conservative. I like wall and roof assemblies that can dry in one direction (or both directions).
If I were to address this situation now, would you recommend cut and cobbling poly iso panels 1" shy of the sheathing and then putting the mineral wool back in? This would require furring out the rafter that additional 2". If so, what would you put on the underside of the mineral wool, a smart vapor barrier like MemBrain?
I guess the other, easier solution is to not insulate the rafters and bring the insulation down to the level of the joists.
>"The rafters are 2×6 and the floor joists are 2×12. Currently I used mineral wool batts tight to the sheathing and then overlaid aluminum faced poly iso on the bottom side of the rafters. "
That's a moisture trap. Roof decks don't dry rapidly through layers of roofing even on hot dry summer conditions, let alone though rain/dew wetted roofing or seasonal layers of snow.
>"Or to be safe, should I remove everything and install a site made baffle to allow a 1” air space between the mineral wool and the sheathing, even if it isn’t vented?"
It has to be vented to reliably keep the roof deck dry.
In Climate zone 6 applying half the total R as closed cell foam to the underside of a DRY( test it first- it must be << 20% moisture content) roof deck with the rest of the cavity fill as fiber can work, if the interior side is at least semi-permeable to water vapor (latex paint on gypsum board, or "smart" vapor retarders.) In zone 7 (northern MN) it has to be at least 60%. Using strips of foil faced polyiso on the rafter edges to provide more cavity depth is fine, but not a full layer, which is a true vapor barrier creating the moisture trap.
Assuming full-dimension 2x6s and 1" polyiso edge strips there would be 7" of cavity depth. In zone 6 with 3" of HFO blown closed cell foam (R20-ish) it would be fine to compress R23s into the 4" remaining 4" would deliver another R17-R18 for R37 - R38 total which would be a 52-53% foam-R. The R6 thermal break on the rafters brings the framing fraction up to R13 - R14 (comparable to a milled 2 x 12 ).
In zone 7 it would take 4" of HFO blown foam, compressing some R15 batts into the 3" remaining depth (or 1.5" polyiso edge strips, to yield the full R15.)
Forgive my ignorance, but wouldn't drying to the inside be impossible on damp days with CC foam directly applied to the underside of the sheathing?
>"Forgive my ignorance, but wouldn't drying to the inside be impossible on damp days with CC foam directly applied to the underside of the sheathing?"
Whether a damp day or dry, ANY day of the year there would be more drying capacity toward the interior through a few inches of foam than toward the exterior through typical roofing.
At 3" most closed cell foam is still running about 0.3 perms, which is still ~3x as vapor permeable as a typical; #30 felt + asphalt shingle layup. At 4" it's closer to 0.25 perms, still more than 2x as permeable as the roofing.