GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

3 layers of Roxul vs. 1 and foam.

canadianexpy | Posted in Energy Efficiency and Durability on

So I’m trying to come up with a wall insulation that will work with my budget and prefer DYI.
I’ve priced Roxul as I like its qualities and come to the conculsion that putting 3 layers of R16 2×4 16o.c. is cheaper than 1 layer of Roxul R22 or R24 2X6 16 or 24o.c. and 4″ of Pink ridged foam on the exterior. I know the lumber and labor of a double stud wall will cost more. (pricing that currently)

I would like to know if the 3 layers of R16 Roxul will perform well in this stack up compared to the 1 R22 plus exterior foam idea. Will it actually get R48 or will whole wall R-value be way down
Also would putting 1.5″ Comfortboard R6 on the outside help in anyway?

House located in Toronto ,ON Zone 6
Thanks for any input

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. charlie_sullivan | | #1

    If you make the right thickness cavity for it, so it's not compressed any more or less than it was designed for, you'll get the rated R-value, or extremely close. And you'll avoid the high climate impact of pink foam. (EPS is better)

    1.5" mineral wool board on the outside would help keep the sheathing warm, reducing the potential for moisture in it. Unlike foam, there's no minimum thickness. Any amount will help.

  2. Expert Member
    Dana Dorsett | | #2

    Exterior insulation thermally breaks the subfloor, band joists and foundation sill. A double studwall does not, so it's apples to pears rather than apples to apples.

    The whole-wall R (average performance after calculating the thermal bridging) of a 2x6/R23 (rock wool) 16" o.c. with 1/2" CDX & 1/2" gypsum with 1.5" rigid rock wool on the exterior is about R22. Swapping out the 1.5" rock wool for 4" of pink XPS (use EPS or polyiso instead- much greener), delivers ~R35.

    The whole wall-R of the 2x4/R15 + stacked R15 batts + 2x4/R15 is about R32, but could be higher if you do some careful thinking & planning around thermally breaking the subfloor, band joist & foundation sill. Adding 1.5" of rigid rock wool to the exterior would bring that up to about R38.

    You may want to study the details of Case 2a, 2b, & 4 in this document:

    These are not identical to the assemblies you're talking about, but they are similar enough that the discussions are relevant.

  3. user-4524083 | | #3

    Dave - It sounds like you're considering a double wall ( or you're calling it triple) vs. 2x6 with foam on the exterior. I think a double wall of the Roxul ( two 2x4 walls) with a layer of Roxul in between,laid horizontally, would work nicely. The central layer of insulation, with no studs, would serve as the thermal break. If the air barrier were placed on the inside of the interior wall, minimal to no condensation would take place at that plane. The air barrier/ sheathing could also be placed on the inside of the exterior wall. You would get condensation there, but with excellent drying potential both to the inside and outside. I am considering such a stack up for an upcoming project, and am doing so because the batts of roxul are so much less expensive than the "boards"and I don't want to use foam. The framing will "water down" the R-value. If you're careful with the framing( placing windows where the studs would fall in their 16"o.c. plan, headers no bigger than what is necessary,etc.), you'd get something like R10 for the exterior wall, R15 (R-14 in Canada for some reason) for the middle section with no studs, and R-11 or so for the interior wall, which could be 24" o.c. as the outer wall would be the bearing wall, for R-36 or so. John Straube wrote an article for Roxul - " Residential Insulated Sheathing Design Guide- Roxul Comfortboard IS". There are some great tables in there for R-values, condensation,etc. for different stack ups. They are all for using the "boards" of roxul, but still very useful even if you use the batts. Best of luck to you on your project.

  4. canadianexpy | | #4

    Thanks for the quick replies and links I'm currently reading them .
    Wow I realized the whole wall r-value was lower but didn't know that it dropped that much.
    So a double stud wall with gap so there is enough room for 3 layers of Roxul R16 and 1 exterior layer of 1.5" Roxul R6 is R54 but whole wall is about R38???
    OK will the 16" to 24" O.C. make much difference? I'm hanging vinyl siding as the finisher and it recommends 16" minimum spacing for attachment, so I'll need to attach 16"OC furring strips but not sure attaching them to the sheathing will cause issues(more holes) .
    One other thing then, is the R-value that people talk about in say the 5-10-20-40-60 for a Net-Zero House or pretty good house based on whole wall value?

  5. GBA Editor
    Martin Holladay | | #5

    Q. "Is the R-value that people talk about in say the 5-10-20-40-60 for a Net-Zero House or pretty good house based on whole wall value?"

    A. Usually. But lots of people talk about R-values in an uninformed or casual way. When in doubt, it's good to get out a pencil and paper and double-check someone's claims.

  6. STEPHEN SHEEHY | | #6

    Reply to Kevin:
    One reason not to put the air barrier on the inside of the inside wall is that you then need to carefully air-seal all the receptacles and switches as well as all the plumbing penetrations. That's doable of course, but can be tricky and time-consuming.
    I used an air barrier membrane on the outside of the interior wall which didn't require any air sealing of receptacles, etc.

  7. Expert Member
    Dana Dorsett | | #7

    Bumping the spacing of a 2x4 load bearing wall to 24" doesn't usually meet code on structural capacity. (But 24" 2x6 is fine.) Even if you did the full advanced framing on it to get the framing fractions down to the 15% range (which may or may not be possibe) it's only the rough equivalent of adding a half-inch to the exterior rigid insulation. Take a look at Case 1b (standard 2x4) and Case 1bii in Table 3, p 13 (pdf pagination) of that BSC document I linked to.

    Seriously- read the first chapter or two of that document, and study the performance tables, and read the discussion details on difference wall cases. For convience, here's the link again:

    The easiest way to get more performance out of a double studwall stack up is to add more space between the studs and fill it with something cheap. R23 rock wool instead of R16 (is R16 really a standard product in Canada?) doesn't add the full R7 difference to teh whole-wall R, since it doesn't thermally break the floor joists, band joists or foundation sills, but it would be more than an R4 boost.

    And yes, most people when discussing building performance put it in terms of U-factor or whole-wall R of the assemblies, not the center cavity R value of just the insulation component, and that would be the case with the Pretty Good House concept.

    BTW: Toronto's climate averages are a bit less than 3750HDD- C (= 6750 HDD- F) annualy, which makes it a zone 5 rather than zone 6 climate per ICC definitions.

    In a reasonably designed air tight house a whole-wall R of R30 is sufficient wall-R to be able to hit Net Zero Energy with a PV array that fits on the roof. See Table 2, p10 of BA-1005:

    As solar panel & inverter efficiencies have edged up since that document was drafted in 2009, (as the installed cost dropped by more than half), you can probably get there pretty reasonably at R27 or R28 whole-wall type construction now, but not R22.

    [edit to add]

    If the batts are truly R16 @ 3.5" thickness, the double studwall whole-wall-R is really coming in about R34, not R32 (and about R40 if R6 Comforboard is added.)

  8. canadianexpy | | #8

    OK so in rushing to get answers I see Dana is right (again) thanks for the catch and correcting me!! R14 not R16 in my head thinking 16 OC just went to R16?!?!
    So 3 layers is R42 +R6 ridged rock=R48 makes whole wall?
    Ok Zone 5 not 6?? Someone told me that, thanks again Dana. The house is being built less than a 1800' from Lake Ontario so I'm right at the bottom of Toronto.
    With the double stud wall why not put the air barrier on the interior of the outside wall before the second wall goes up.? IS it not tricky to install on the outside of the interior wall.
    Thanks again for the links great reads =, so far..

  9. Expert Member
    Dana Dorsett | | #9

    No, that's R48 center-cavity, without adding in the R-values of the wood sheathing, gypsum or siding, and calculating the performance hit of the combined thermal bridging. With 2x4 framing about 25% of the face area of the wall has framing behind it. The R-value of most framing species used in this region is about R1.2 per inch, so at 3.5" it's about R4.2.

    When 25% of the wall layer is R4.2 and the other 75% of the layer is R14, the 25% that is R4.2 conducts a bit more heat that the other 75% of wall area- it's a thermal bridge. When modeling the parallel paths to account for this bridging, and adding up the R-values of the siding, gypsum and structural sheathing, add in the performance of the interior & exterior air films you can come up with the whole-wall R. (I run a spreadsheet model for 2D parallel path heat conductivity estimates, but it can be done long-hand as well.) Since the framing fraction is much more thermally conductive, the whole-wall R will always be lower than the center-cavity R.

    In the US R15 rock wool and fiberglass are standard products for 2x4 framing (as are R11 and R13, low & mid-density fiberglass.) Breaking it down for when using R14 at a 25% framing fraction (typ. for 16" o.c. framing) your double wall stackup with continuous R14 between the studwalls comes in at about R8.8 for each 2x4 layer for R17.6, then ~R12 for the middle R14 layer (derated to R12 for the thermal bridging of the subfloor, joists and band joist, window & door framing etc.), now we're up to R29.6, then add ~R1 for the combined ply wood sheathing + gypsum & siding, and another ~R0.85 for interior + exterior air films and you're in the R31-32 range.

    Add R6 for the 1.5" Comfortboard and you're at ~ R37 whole-wall, or about U0.027

    These are all estimates, but at any typical real assembly it'll come in about R1-R1.5 less than than whatever it would be with R15s. With greater detail on the other materials and the framing plan it's possible to come up with a more precise number, but that precision is wiped out by the variations of real-world construction & materials, but it's the right range.

  10. canadianexpy | | #10

    So it was Roxul that lead me to believe I was in Zone 6
    13A08 ROXUL COMFORTBOARD IS Application Guide - 04-11-2014 mentions on page 24 that Toronto
    is in Zone 6 under Recommendations for vapor control by climate zone.
    But looking at other links I see Toronto slips just in Zone 5.

  11. canadianexpy | | #11

    Thanks Dana I feel I little more comfort (pardon the pun) in that I'm heading the right direction. I really want to build a Pretty Good House or Net-Zero. and use a Mini-split to heat/cool the house. I was surprised that they recommend so much Attic insulation in one of the papers you linked. (R65).
    Once I get pricing back on double 2x4 wall construction compared 2X6 . I can decide on full wall details.
    Then weather a mini-split will work in this house. 2600sq'

  12. user-1137156 | | #12

    Installing the air barrier on the outer face of the inner wall is really straight forward if it is done while the wall lays on the 'deck' before being "erected". Putting the air barrier on the inner face of the outer wall increases the risk of condensation in the insulation. What is your foundation? Have you considered PWF? Is your house 1 or 2 story? Dana has pointed out that rim joist and floor framing are areas that need attention thermally..

  13. Expert Member
    Dana Dorsett | | #13

    Looking at the actual weather data over the past 25 years Toronto is SOLIDLY in zone 5, not just slipping in at the hairy edge.

    And when you are only 1/2 km from the lake front, there is even more moderation of the local temperature averages compared to 5 km inland (though it doesn't always seem like it.) A half degree temperature warmer over 200+ days of heating season peels another 100+ HDD off the total.

    Zone 6 starts at 4000 HDD-C, a number that has been exceeded (barely & rarely) in Toronto in just 5 years out of the past 25, the worst being 1996 at 4176 HDD. The 25 year running average is less than 3750 HDD, maybe even less than 3700 HDD.

    Take a peek at the bar graphs at the bottom of this page: (Hover the cursor over the top of the bar and it'll display the number.)

  14. STEPHEN SHEEHY | | #14

    Dave- it was pretty easy to put the air barrier on the outside of the inside wall. The outer wall was structural. After the roof trusses were installed, the inner wall was built on the floor and the membrane stapled to the studs, leaving a foot or so overhanging what would be the top. Then it was stood up and the overhanging section was stapled to the underside of the trusses. Then the ceiling/roof air barrier membrane was applied to the trusses. Then we strapped over that, hung drywall and blew cellulose from an access panel on the exterior wall under the peak.

    Putting the air barrier on the inside of the outer wall would have complicated the installation of the blown-in cellulose in the walls. We filled the cavity from the air barrier out to the sheathing on the outside of the outer wall. Then we insulated the inner wall with fiberglass batts after all the wiring and plumbing was done.

    Using Roxul batts as you suggested sounds like it would work, though.

  15. canadianexpy | | #15

    Thanks again guys great link, It doesn't surprise me as its 12C yesterday in December?!!
    I'll post questions when I finalize the wall assembly to see if it makes sense.
    It will be a 2 story house with full height basement block walls. Total floor space will be roughly 3800sq'

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |