Weighing insulation options for attic, roof, and gable ends
I’m seeking insulation advice to turn our 1950s cape cod attic into living space. Here is a description of the situation:
(1) Location is near Harrisonburg, VA (zone 4)–code level roof insulation is R38.
(2) There is zero insulation in the walls and roof (there is cellulose in the attic floor).
(3) The walls from the inside out appear to be 2×4 frame sheathed in 1x lumber wrapped in tar paper covered in rigid siding (transite, maybe?) covered by a plasticy, spongy WRB (I assume) clad in vinyl siding.
(4) There are no (intentional) soffit vents, just a fan at each gable end. However there is daylight at the eaves.
(5) Insulating the exterior roof deck is not feasible.
(6) I’d like to minimize encroachment of the roof into the room. The rafters are 5.5″ deep, 16″ o.c.
(7) Gable ends are north and south. There are two west-facing dormers.
I think the air-sealing qualities of spray foam would be greatly beneficial here, and the local contractors use Demilec closed cell which claims R7.4/inch. That would allow me to reach code-level insulation without deepening the rafter bays. Another option is for spray foam up to the collar ties (where a ceiling will be installed) and then blow in cellulose on that ceiling. That is significantly cheaper, but does it make sense to spray foam against the lower part of the roof and have no insulation against the upper part? What would I have to do to ensure proper venting of that space (preferably without the fans)? Also, how can I best address the daylight at the eaves? Backing rod & caulk? I assume it’s a bad idea to leave it up to the foam to take care of it since some would could end up on the exterior.
As another option, if I can source Kingspan Kooltherm, does it make sense to cut and cobble the first 2 or 4 inches of the roof insulation and encase it in closed cell spray foam (an idea derived from https://www.greenbuildingadvisor.com/article/high-humidity-in-unvented-conditioned-attics)? I am assuming this would be cheaper than a 100% spray foam fill (using myself for labor), but do not yet know the cost of Kooltherm.
Then again, I haven’t even addressed thermal bridging. Would 1″ polyiso be sufficient? This article seems to say “yes”: https://www.greenbuildingadvisor.com/article/walls-with-interior-rigid-foam.
As for the walls, there are gaps in the sheathing large enough to stick my fingers in, so they need a lot of air-sealing help. Does it make sense to use closed cell spray foam on them when the drying potential to the exterior is suspect (see point 3 above)? Would I be better off making the gable-end air-barrier at the drywall and using cellulose to provide a hygric buffer? The contractor dissuades me from using open cell on the walls because because it is a different installation crew from the closed cell (ergo high fixed cost).
Thanks for any and all help!
–Paul
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
HFO blown foam is crazy-expensive and it's performance is undercut by the thermal bridging of the framing. R38 with 5" of thermally bridging rafter is not the same as R38 of 10" rafter depth. The 5" of wood is only good for R6, the 10" is good for R12- half the amount of heat transfer through the framing fraction.
In zone 4A it only takes ~30% of the total R outside the first condensing surface to have sufficient dew point control. R38 x 0.3= R11.4. With 2" of Demilec you'd be at ~R15, and would only need an additional ~R23 to meet code, and it would have sufficiently low vapor permenance to protect the roof deck from winter moisture, but still sufficiently vapor open for a reasonable seasonal drying path (unlike 5".) That's do-able with 5.5" thick R23 rock wool batts, which would be a total depth of 7", about what you'd have with a 2x8 rafter.
If the rafters are 2x6, adding Bonfilioli strips of 1.5" polyiso + 1x furring would bring the total depth to 7.25", exactly the same as a 2x8 rafter, but it would raise the R-value of the framing to R12, which is better than a 2x10 rafter.
https://www.finehomebuilding.com/membership/pdf/9750/021250059.pdf
Even taking it above the collar ties all the way to the ridge (recommended) is going to come in cheaper and higher performance than a 5" Demilec solution, impinging only 2.25" at the cathedralized ceiling.
So, what's that 2.25" really worth to you?
On the gable ends a flash-inch of closed cell for air sealing and compressed 3" thick rock wool sound abatement batts or compressed R13 fiberglass (unfaced or kraft faced, just not foil faced) would work from a moisture point of view. That would be about R17 at center cavity, but no thermal break over the framing. If the goal is to bring it up to IRC code min, the flash-inch of close cell, and 1" polyiso edge strips lightly glued or cap-nailed (no furring) to the studs would allow for R15 rock wool or fiberglass batts. It would be over R20, and a higher R framing fraction than 2x6. The wallboard would need to be long-nailed/screwed through the 1" foam but that's not usually a problem.
It's very easy to cut 1.5" wide foil faced polyiso strips cleanly & accurately using a steel wallboard taping knife that has been sharpened on the edges and a straightedge (a 4' level is pretty handy). See this instructional vidi:
https://www.finehomebuilding.com/2009/04/15/theres-a-better-way-cutting-rigid-insulation-2
Thanks Dana, fantastic response! For the gable ends I may use both methods mentioned. The stairs come up to the south gable end with a landing of just 36" from nosing to framing, so I don't know if I can get away with deepening the wall in that area. But the north wall is ripe for added depth, as well as the south wall outside of the landing, and perhaps the dormer walls too.