Sealing & Conditioning a TX Attic – no spray foam, no exterior insulation, no soffit vents?!
Hi folks-
I just moved into a house in Austin, TX; a late 70’s single story with a vented attic.
The HVAC and ducts are in the vented attic.
The attic is vented with ridge vents and gable vents, but NO soffit vents.
My long term plan is to re-roof with exterior insulation above properly air-sealed 5/8″ ZIP Sheathing, then condition the attic space. But I need to save a lot of money before I can do that. In the meantime, I’m wondering if I can take some half-measures to help.
Can I seal all the vents in the attic, then use Rockwool batts between the rafters, followed by 1/2″ drywall mounted to metal channel over the rafters to improve the situation? This will allow me to air-seal (drywall), add R-value (Rockwool), but still maintain the health of the roof sheathing by allowing it to dry to the inside, which will be dehumidified, conditioned space?
I understand this is far from an ideal solution, but it is one that I can undertake slowly, on my own time. I’m also not willing to consider spray foam at this time because of climate, health, cost, and installer risk.
Thanks!
damen
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Not sure I 100% follow your plan. You want to establish the air-barrier at the underside of the rafters (in the attic), then condition the attic (presumably you'll also clean the attic and have the insulation on the attic floor removed)? You'd add a supply and return in the attic?
So it's basically a vaulted ceiling situation, but with the attic as a big conditioned chase.
Another idea would be, as you save, to work on air-sealing the ceiling plane as the air-barrier, but getting loads of cellulose blown in to bury the ducts? That's allowed in Austin (Climate Zone 2) and is going to be _way_ less expensive than rockwool in the rafters. You might even have someone remove existing insulation up there, to help with the ceiling air-sealing. That' what Risinger did on his "Real re-model" house (before he, uh, tore it down to reframe it :)
https://www.youtube.com/watch?v=1BcqeQJqp08&list=PLDYh81z-Rhxj4mulCrScAArjx7Db14t78&index=4&t=0s
Hi James, thanks for your response.
Sounds like you do understand the plan. Yes, I currently have 5 tons of HVAC for a 2400sf house and would add supply, return, and a dehumidifier that drains outside.
Which rolls right into the next point: the HVAC duct situation; it's a total mess. The ducts are way undersized for the 5ton that the previous owners rammed in here. When the heat or AC kick on, the fan noise is super loud, like makes you raise your voice a whole lot to be heard.
But maybe you're on to something here. I have been considering moving to a ductless mini-split solution for the house that could eschew any mechanicals in the attic, although it's not clear to me yet how that would work with air distribution the 4 small bedrooms (anybody have thoughts on that?). Then I could proceed with the plan you laid out: remove all the mechanicals up there, have it vacuumed out, seal around all the penetrations, and blow in cellulose.
Ha, I follow Risinger's channel and know exactly what you're talking about :)
Thanks again for your thoughts- very helpful!
damen
In Austin you'd be able to get away with dense packed cellulose in an unvented assembly. It might work with rock wool if the metal roofing isn't a highly reflective "cool roof" type, but it's iffy.
See Table 3 of this document, in particular the "2A Houston" row, and the "Spray fiberglass" and "Cellulose" columns.
https://www.buildingscience.com/sites/default/files/migrate/pdf/BA-1001_Moisture_Safe_Unvented_Roofs.pdf
The 1.8lbs density fiberglass is pretty similar to rock wool batts from a hygric buffering point of view.
At the simulated R30 with fiberglass it's likely to have a higher than desirable moisture content in the north facing pitches for at least 4 weeks out of the year even with dark shingle roofing, whereas it's just fine with cellulose.
If batts is your only option, cotton (denim) batts have similar moisture handling characteristics to cellulose, but isn't quite as "green".
How deep are the rafters/truss chords?
[edited to add] The thermal diffusivity of cellulose compared to rock wool is also quite favorable, which will delay and lower the peak attic-ceiling temperature even at the same R-value.
Great reply, thank you Dana!
I read the paper in full, thanks so much for bringing that to my attention.
It confirms (as has been referenced in many places I've found, Lstiburek, this site, etc) the importance of the air control to keep moisture off the condensing plane (in my case, the underside of the roof sheathing). I assume that properly installed drywall (all seams taped and sealed) will provide this air control?
Most of the rafters are 2x6, so I don't necessarily like my odds of getting the full R30.
My updated solution then seems to be drywall over metal channel, filled behind with blown-in cellulose. Is that right?
Thanks again, everyone!
damen
I'm not clear what is meant by "...metal channel..." in the proposed stackup. Are you talking about
steel hat furring? Why metal? How deep?
Metal is 100s of times more thermally conductive than wood, severely undercutting the performance of any insulation installed between the metal elements.
By metal channel I mean resilient channel, like is used for acoustic performance. This creates around a 1/2" gap between the drywall and the studs.
I have a hard time believing the thermal performance of resilient channel will be worse than the thermal bridging of screwing the studs to the drywall, no?
Here's a link to the product:
https://www.amazon.com/dp/B0002D05OQ
Thanks!
damen
>"I have a hard time believing the thermal performance of resilient channel will be worse than the thermal bridging of screwing the studs to the drywall, no?"
Carbon steel at about 500 x more thermally conductive than Hemlock or pine framing. The 25 gauge steel in that is about 0.022" thick, and each edge of the channel has the same conductivity of a 11" wide rafter, and there are 2 edges per channel, the equivalent to a 22" wide rafter (instead of the 1.5" of milled 2x lumber.) So no matter what insulation is filled in between the strips of furring, the performance of that furring is severely undercut for that thin depth layer by the steel. Even steel studs (with only one penetrating sheet of steel) cuts the thermal performance of the insulation by at least half.
The cumulative cross sectional area of screws or nails is pretty tiny compared to the continuous edges resilient channel furring. Yes, the fasteners are small thermal bridges, but pretty negligible overall.
Using 1x or 2x furring perpendicular to the rafters as the nailer to gain more insulation depth has far less thermal bridging issues. A layer of 1.5' of cellulose between the 2x2 furring puts an ~R5.5 thermal break over most of the rafter, and the 1.5" x 1.5" thermal bridge at the intersection where the furring attaches to the rafter is good for about R2.
I know it's a bit pedantic, but just for the sake of completeness, resilient channel only has one edge of contact with the drywall and one edge on the studs, so I think the equivalent would be the 11" wide rafter, if I properly understand what you're saying there.
And I think my wording was a bit confusing; I'm not concerned about the screws. I was trying to refer to the amount of contact area between the rafter and the drywall being the concern. I've spent a lot of time on both houses I've lived in with a FLIR E8 Thermal Imaging camera, and it always amazes me how clearly I can "see" the studs behind the drywall.
Anyhow, it sounds like I should be going with some 1x furring either way, so I'll do that.
Thanks again!
damen
>"I know it's a bit pedantic, but just for the sake of completeness, resilient channel only has one edge of contact with the drywall and one edge on the studs, so I think the equivalent would be the 11" wide rafter, if I properly understand what you're saying there."
It has two sides of the channel that conduct from the drywall side of the insulation to the rafter side of the of the insulation (assuming it there is insulation filling in between the furring.) In a 2-D parallel path model, each of those thin steel webs penetrating the insulation conducts as much heat as an 11" wide plank would if the plank were as thick as the furring is deep.
But it's actually a 3-D problem. The surface area in contact with the rafter and with the dry serves to somewhat increase the amount of heat being conducted to/from the steel, but that's really a secondary aspect. The most relevant part is the conductivity of the part that fully penetrates the insulation layer. I didn't see dimensions in the description- a lot hat-furring is 7/8" deep. A continuous 7/8" layer of cellulose is good for about R3.2, but with steel hat furring 16" o.c. the combined performance after thermal bridging is less than R0.2. With 3/4" 1x4 furring 16" o.c. the 3/4 layer of cellulose is only worth R2.8, the furring is about R0.9, but the combined performance is still over R2, which in your stackup is a non trivial fraction of the total.