Replies

1. GBA Editor
   Martin Holladay | Mar 27, 2015 02:15pm | #1

   Glen,
   Talk to a spray foam contractor. I don't think your plan is a good one.

   Your 2x12s are 11.25 inches deep. If you fill the bays with 11 inches of batts, where will the spray foam go?

   If you try to install cellulose in open, sloped rafter bays without any roof sheathing, the insulation will be blown away by the wind, and it certainly won't stay put when the spray foam contractor starts spraying.

   The usual way to do this type of retrofit work from above is to empty the bays of existing insulation, and then get a spray foam contractor to install as much spray foam as you can afford.

2. 
   ironglen | Mar 27, 2015 02:36pm | #2

   Thanks for the prompt reply Martin. Looks like I was off on my measurements (I thought old 2x12 lumber was 'true' to measure unlike the newer lumber being 1.5x11.25). Still, if the plan is carried out with rockwool or glass batts to 10" or even 9", with closed cell sprayed on top, would this not follow your guidelines in the summary article, only spray applying the foam from the topside followed by sheathing, rather than interior access spraying the underside and loose filling to capacity? Inexpensive 1/2" rigid foam sheets could easily and inexpensively be fixed in the rafter bays prior to spraying as well, to hold the batts, assist in sealing, and maximizing spray foam coverage and sealing. I believe it would be an immense cost saving plan compared to contractor spray foam only, with good r value to cost performance. Your thoughts appreciated.

3. 
   ironglen | Mar 27, 2015 04:46pm | #3

   I explained my project to a spray foam contractor and he said the set-up cost would be cost prohibitive for this small job (only half a living room ceiling) and other contractors would agree. He suggested that I use foam boards and seal the edges where they meet the rafters using cans of spray foam. Doesn't the insulation need to be airtight as well as in contact with the roof sheathing to prevent any condensation? Any gap between the two is unacceptable, correct?

4. Expert Member
   Dana Dorsett | Mar 27, 2015 06:24pm | #4

   Most rock wool batts these days are R4.2/inch or better, if installed with a perfect fit.

   Installing 1/4" fan-fold XPS (sold as siding underlayment) with the seams taped over the rafters (and under the new OSB deck) as a vapor retardent air barrier would be sufficiently protective of the roof deck in your climate. The thin plastic facers on fan-fold XPS are typically 0.7-1.5 perms- about the same as an inch of closed cell spray foam, which allows the roof deck to dry at a reasonable rate, but isn't so vapor open that the roof deck takes on significant wintertime moisture accumulation. Key to making the fan-fold XPS perfectly air-tight is to locate the seams only over rafters caulking it under the edges, and taping over the seams with housewrap tape prior to putting down the new roof deck. It comes in 4' x 50' sections, so it should be pretty easy to make every edge & seam coincide with a framing element. (eg: http://www.foamular.com/assets/0/144/172/174/73641387-12a8-45af-99a9-bb84726ae053.pdf < this particular version is rated 1 perm max, which is about right.)

   There could still be wintertime condensation events at the foam/fiber interface, but unless you actively humidify the air in winter it won't amount to much. If you insulated with cellulose blown in mesh instead of rock wool batts you would get a more perfect fit, lower the air leakage from the interior, and the cellulose would safely buffer the wintertime moisture drives.

   Blown cellulose is DIY-able with a 1-stage rental blower from a box store, using landscaping fabric as the blowing mesh. You'd have to roll the pillowed-out mesh flat again before stapling down the fan-fold XPS to get it to sit flat, but this approach is pretty safe, even if it's a bit outside the strict letter of code for unvented insulated roof assemblies in climate zone 3B. The density you're likely to achieve would be in the 2.5-2.8 lbs range (3 at the outside) but that should be sufficient to prevent settling over time in a climate as dry as yours, if somewhat problematic in much colder wetter climates with much more wintertime moisture cycling.

   Note the roof deck performance of cellulose-only in different climate zones in table 3 of this docuement, as well as the performance with 1" of ccSPF (which is about 1.2 perms, comparable to 1/4" fan-fold XPS.)

   http://www.buildingscience.com/documents/bareports/ba-1001-moisture-safe-unvented-wood-roof-systems

   Unfortunately they didn't simulate a zone 3B climate, but the cellulose-only performance was still pretty safe in both zones 3A & 3C (Atlanta & San Fransisco) which are substantially more humid than zone 3B.

5. 
   ironglen | Mar 27, 2015 06:47pm | #5

   Thanks Dana, I was hoping you might add your expertise as well. Your solution sounds achievable and cost effective, more so than spraying closed-cell, though perhaps less moisture resistant and insulating. I take it that I would complete the fan-fold XPS and re-roofing prior to installing the blown cellulose from the interior, through holes in the sheet rock?

   I just spoke to a representative at foam it green, and he understands what I want to achieve, but is running it by other associates to see if one of his products would work best in my plan. They have a regular, fast cure spray as well as a slow-rise formula, which I think would be ideal for my plan to allow time to move the osb into position and secure before the product begins to skim/dry, ensuring excellent adhesion and sealing.

   I'll find out next week what they suggest, but what do you think regarding this plan? I like your idea, especially the cost, and I know my idea is a little unorthodox, but it is really not new, just achieved through a different method. It really is creating a single sided SIP panel with additional loose fill/batting insulation, though it has thermal bridging via rafters.

6. 
   ironglen | Mar 27, 2015 06:59pm | #6

   Dana, would this product be comparable to the foamular one?
   http://www.lowes.com/ProductDisplay?partNumber=304092-210-304092.0&langId=-1&storeId=10151&productId=3122451&catalogId=10051&cmRelshp=req&rel=nofollow&cId=PDIO1
   I believe it is Kingspan's:
   http://www.trustgreenguard.com/product-applications/xps-siding-underlayment/xp-series-fanfold-underlayment/

7. 
   Dana1 | Mar 27, 2015 09:34pm | #7

   The Pactiv box-store branded fan-fold shown in your box store link is perforated on a grid, and highly vapor permeable, not very protective at all, and not even an air-barrier. The unperforated Pactiv XP14 would do it, at 1.7 perms.

   It would be easier to verify the quality of the insulation job if you blow the cellulose from the exterior using mesh to contain it, prior to the XPS. (This of course requires a stretch of dry weather.)

   Using slow rise 2lb foam the OSB would be essentially be permanently glued to the rafters (you almost wouldn't need nails) but any gaps or voids would be hard to find & fix. (Some installers use infra-red cameras to track the spread of the foam as it expands, to ensure they are installing enough, but not too much.)

8. 
   ironglen | Mar 27, 2015 11:46pm | #8

   Yes, I see it listed on the lowes site as perforated, then on the kingspan site XP14 is listed as 1.7 perms compared to the foamular's 1.0 perms using the same standardized test. I am concerned how well it seals when shingles are nailed throughout the osb and protrude through the barrier skin. Is there an alternative shingle fastener for the shallow 3/12 pitch that won't penetrate the barrier?

   I see the benefits filling the cavity with cellulose, and we often have successive days of sunshine to facilitate the installation. Additionally, I was going to add significant cellulose to the minuscule amount of existing attic insulation in remaining non-cathedral areas.

   I had the same thoughts regarding the slow rise foam, and with its longer skim/dry time thought it might be more manageable to obtain an even coat prior to setting the osb deck atop it. I believe blowing cellulose, followed by slow rise foam will likely be the plan taken, due to my concern about the numerous nail holes in the barrier along with the reduced r value. I sure would prefer using an XPS type barrier for the cost savings alone, and I know spray foam can be difficult and very messy.

9. GBA Editor
   Martin Holladay | Mar 28, 2015 05:36am | #9

   Glen,
   You are describing the "cut-and-cobble" method. I'm not a fan of the cut-and-cobble method for unvented roof assemblies; there have been reports of failures. For more information, see Cut-and-Cobble Insulation.

   If your job is really too small for a spray foam contractor, you could consider buying one or more two-component spray foam kits and doing the job yourself.

10. 
    Dana1 | Mar 28, 2015 09:30am | #10

    Cut'n'cobble is indeed riskier than a full-sheet of XPS with polyolefin facers, since the long term air seal of that many panel seams is dubious at best. A 4' x 50' sheet with seams coincident with framing is far easier to guarantee, and unlike the slow-rise foam solution, is inspectable. That's not to say the risk is very high with any of them in a climate as dry as yours.

    Don't sweat the nail holes for the shingles. The hole itself is filled with vapor-impermeable steel, and it has a compression fit surrounded by XPS. The total cross sectional area of higher permeance in the annular ring around the nail is miniscule. On subsequent re-roofing there may be nail holes left open, but the total area of those perforations is still tiny. Vapor diffusion is a function of total area- it's a far cry from perforated fan-fold with much larger holes on a 1"-2" grid.

    But in a dry 3B / 4B climate cellulose alone is not very risky at all. When you get to dry climates as cool as 5B it is of some concern though. Note: There are cellulose vendors who claim that at 3.5lbs density cellulose alone is safe for the roof deck in unvented assemblies even in snowy zone 6A. Though I'm personally skeptical about those claims over a 50 year lifecycle, even in those colder -wetter climates it may still take decades for dense-packed cellulose to show up as an actual problem.

    In your case, with 1.7 perm (or even 5 perm) thin polyolifin-faced XPS between the fiber and roof deck, and standard latex paint on the finished interior ceiling (3-5 perms) the rate of vapor diffusion from the interior to the roof deck in winter is extremely slow. The key to keeping moisture transport to the roof deck low is sufficient air-tightness to eliminate significant air convection through or around the XPS layer. That's pretty easy to do with large fan-fold sheet goods, nearly impossible with cut'n'cobbled sheet foam, and (IMHO) less than certain with slow rise foam.

11. 
    ironglen | Mar 28, 2015 12:24pm | #11

    I agree with you both, that the cut and cobble appears quite risky: I was surprised when the contractor suggested it. I won't use it.

    I'm so thankful to have all of you on this forum advise me on this matter. I would have thought an inch or two of either regular or especially slow rise foam would have easily been the preferred barrier in this application. The XPS product appears to be the cost effective application to better insure against condensation problems. It will be cost effective as well, perhaps 20% of the two part spray foam cost.

    Dana, would the XPS seams be overlapped ~1.5" at the rafters, sealed with PL foam rated cault, housewrap taped as you previously stated, and nailed to the rafters with 1.25" shingle nails?

    While on this topic, the cathedral ceiling has large,unsealed, poorly insulated kick walls on three sides: any recommendation? Once again, I was planning to spray foam an inch or two, and reuse fiberglass batts in the remaining vertical cavity spaces, but perhaps there is another alternative method.

12. GBA Editor
    Martin Holladay | Mar 30, 2015 09:49am | #12

    Glen,
    I'm not sure what you mean by "kick walls." Maybe this is a same thing that most of us call kneewalls. If so, you may want to read this article: Two ways to insulate attic kneewalls.

13. 
    ironglen | Mar 30, 2015 11:05am | #13

    My mistake again, knee walls is the correct term for the areas needing insulation. Thank you Martin for the links to your detailed, informative articles as they are supplying exactly what I need.

    Now that I have a better understanding of the cathedral ceiling, I am looking at the lower porch area that has the blistered shingles and it has no ventilation. I will try to take a picture of this roof, to show the complex issue.

    Is there a way to change the font color or text style for hyperlinks in the posts? I almost missed one in an earlier post but I happened to scroll my cursor over it!

14. 
    ironglen | Mar 31, 2015 02:38pm | #14

    I found the answer to addressing the adjacent non-ventilated lower porch area in Martin's article: a slim, mid-roof vent. http://www.finehomebuilding.com/how-to/departments/energy-smart-details/attic-ventilation-strategies.aspx

    Thanks Martin!