Good hot roof?
I have been doing research and reading for the past year on cathedral ceilings and retrofitting additional insulation.
I am in climate zone 5, between Hartford CT and Springfield MA. I have a house that was built in 79 post and beam style. At some point in its life the house was renovated and all the exposed beams and tongue and grove decking was covered up. Fiberglass bats were sandwiched between the t&g decking and a metal stud grid-work. This was all covered by 3/4″ sheetrock. The best that I can tell, the beams are 8″ thick, the bats are all that insulate the roof at the moment, roughly 8″ thick.
I would like to add rigid insulation this April during a re-roof. I want to make sure my plan of attack will work. The roof assembly will be as follow, working from the inside out:
1 1/2″ t&g
7/16″ osb( seams tapped with flashing tape, grace i&w where the roof plane meets the wall plane for air sealing)
2 3/4″ foil polyiso, staggered and seams tapped on last layer.
I realize that this type of assembly will have the shingles running hotter that with a vented assembly. I also realize that polyiso loses some of its r value in the winter months. My goal is to have a successful assembly on a budget. If I could I would add furring strips on top of the rigid foam before running the new OSB. Does my plan of action make sense? Do I have any potential failures in this type of assembly?
My second question is in regards to overhangs. House is built on an East West axis, with the eaves having no overhangs. Because of this, in conjunction with a lack of flashing on some large east facing windows, there was bulk water intrusion on the east wall. I will be addressing these windows but have stopped water intrusion by removing the wooden siding and wrapping the house in tyvek(covering the windows). The other windows have already been replaced and flashing integrated with the housewrap.
What is the best way to mount the overhangs on the eaves considering I will have 5.5″ of insulation on top of the original roof deck?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
One more specific detail. The Grace I&W will wrap around from the roof plane onto the wall plane and integrate with the tyvek.
In US climate zone 5 (= all of MA) you need at least 40% of the total R to be on the exterior of the roof deck, and you'd have to derate polyiso to about R5/inch for wintertime performance in that kind of stack up. Code min is R49.
Your stackup has about R14 in polyiso (derated), and about R30 in fiber, R44 total (which is sub-code) and only about 32% of the R-value to the exterior of the fiber layer, which is risky.
If you bumped the polyiso to 4" you'd have R20 out of R50, which just barely squeaks in.
You also don't specify whether there is an (necesssary) air barrier between the metal studs and the fiberglass(?).
Are you planning to just leave the pre-existing batts in place? If yes, take the time to measure it, and find out if there is any air gap between the fiberglass and the t & g.
If there is a rigid enough air barrier between the batts and the metal studs it's worth drilling and blowing in fiberglass or cellulose from the top before applying the Grace Ice & Water Shield, to make sure there are no thermal bypass channels, and fill in any compressions or voids.
In MA you can can get factory seconds & reclimated roofing polyiso for pretty cheap (1/4-1/3 of retail virgin stock) from places like Green Insulation Group in Worcester, or Nationwide Foam in Framingham. A couple layers or maybe even three of 2" reclaimed roofing iso seems like the "right" solution here, with a lower cost than 2-3/4" of virgin stock foam.
There are others, sometimes advertising in the local Craigslist and elsewhere:
You have to keep checking back to find them, some of the small-time foam reclaimers only advertise when they have an upcoming job that's bigger than their available storage space or don't have it pre-sold.
Download and save a copy of the Mass Save Deep Energy Retrofit Builder Guide:
Refer to the Attic/Roof retrofit section beginning on p.48, in particular, the unvented roof drawing on page 50. Note the "At least 40% of total R-value" prescriptive specified for the rigid foam layer. Take that seriously!
Dana, thank you for the response. I'm sorry that I omitted a detail. I will have 2 layers of 2 3/4" foam for a total of 5.5". I did purchase factory seconds from Ron at green insulation. 24 per 4x8 sheet. I think that the suggestion to blow in additional insulation is a great idea. Considering, however, I have a limited schedule and time frame to accomplish this work I may continue with what I have planned. I am doing the work myself, with support from family and friends. It won't hurt to get estimates on the insulation. Would dense packed be an ideal candidate?
In general, dense-packed cellulose performs better than fiberglass batts, because the cellulose is more likely to fill all the nooks and crannies of the rafter bays. Moreover, dense-packed cellulose is less air-permeable than fiberglass, so you are less likely to have problems with convection and air leaks.
As it stands, with the addition of polyiso, I will have almost 50% of my insulation(r value of 27.5) on the exterior of the sheathing. With the addition on dense packed, I break that ratio. Should I be concerned?
What will be the total thickness of the fiber (cellulose + fiberglass) layer(s)?
That is a difficult question to answer. The existing roof structure has some space between the batts and the t&g deck, however, I am unsure of the exact spacing.
Total depth of the assembly, from t&g to finished sheet rock is 11". Sheet rock is 3/4". I am guessing that the metal grid-work is 2" +/- thick with the first layer running vertically, attached to the beams, and the second layer running horizontal attached to the first layer. This should make the beams 8". Batts sit on the metal grid-work. dropping slightly below the bottom of the beams as they rest on the horizontal metal studs. I will need to open up the sheet rock to make measurements and find the exact spacing.
In addition, wouldn't the process of dense packing compress the batts somewhat? I would assome the best course of action would be to install the cellulose between the batts and the t&g deck.
I made a discovery today while trying to get a better idea of my roof assembly. Pulling a corner of my asphalt shingle off I discovered that on above the t&g I have 3" of foam insulation. Im unsure what this rigid insulation is. It has a approximately 1/2" of a fiberboard material with a foam layer of 2 1/2". This acts as the roof deck. The asphalt shingles were nailed directly to this surface.
My question is does anyone know what this foam board is?
Secondly, with 8.5" of foam, and potentially 8" of dense packed, do I have the potential of any moisture issues? Im calculating a rough r value of 66? I am derating the polyiso to r4 and including the 3" of foam board i found today. R4x8.5 is r34 for the foam. R4x8 for the dense packed is 32. total of r66. Martin or Dana, any input? Your professional opinion would be appreciated.
The product you describe -- 2.5 inches of rigid foam with a 1/2-inch-thick fiber-based sheathing material bonded to one side -- is called nailbase. These days, nailbase panels have OSB adhered to one side of the rigid foam. The sheathing material that you describe as "1/2 inch of a fiberboard material" doesn't sound like OSB, but the principle is the same. It's an early version of nailbase.
Concerning your questions about the ratio of rigid foam above the roof sheathing to fluffy insulation below the roof sheathing, I urge you to read this article: Combining Exterior Rigid Foam With Fluffy Insulation.
In that article, you'll read that in your climate zone (Zone 5), this type of roof needs at least 41% of the total R-value of the roof assembly to be in the form of rigid foam above the roof sheathing. So your proposed assembly is fine; you have a significant margin of safety.
By the way, I like to use an R-value of R-3.7 per inch for dense-packed cellulose, not R-4 per inch. However, either way that you calculate the R-value of your proposed cellulose layer, the assembly is safe.
Thank you for the prompt response.