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

Rate my knee wall insulation plan

jmetrail | Posted in General Questions on

Hi all,

I want to re-insulate the upper floor of my 1920 1.5 story.

Here are pics showing current configuring and my planned new configuration.

My main questions are:

1. On the short exterior walls, is it weird that the shiplap and weather barrier are installed on the inside face of the studs, which prevents me from accessing those stud bays to add insulation? Is my proposal to create a small 2×6 wall immediately inside this exterior wall a reasonable one? Should there be some sort of channel and air barrier between the insulation and the shiplap?

2. I have seen it suggested that such an exterior wall would be designed to dry to the inside, and to not put a interior vapour barrier. Is that because the rigid foam use to create the air channel is not vapour permeable? What if I use some kind of vapour permeable fibre board to create the air channel instead? Could I then install rigid foam/vapour barrier on the interior have the wall dry to the outside as planned?

2. Should I be adding vapour barrier to the flat part of the ceiling between drywall and joists, taped to the rigid foam on the sloped parts?

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. jmetrail | | #1

    And a quick follow-up - under this plan, the attic will be vapour sealed against the second floor.

    However, there is no insulation or vapour barrier in the main floor exterior walls. Will this be a problem? Will moisture from the main floor travel through the balloon framing, get into the attic, and defeat the purpose of my vapour sealing the second floor?

    Would caulking the new fire blocks achieve a vapour sealed attic? Or is it important that the air in the lower exterior walls be permitted to move up to the attics?

  2. Expert Member
    Dana Dorsett | | #2

    It doesn't need to be a 2x6 assembly on the interior side since it isn't structural, and it's better if it isn't, since the thermal bridging of the framing timbers cuts in to the performance of the insulation. How tall is that stubby exterior wall section?

    You can probably just cap-nail or prop otherwise fasten 5.5" deep R23 rock wool or R21 fiberglass F batts in place, snugged up to each other in a continuous layer with a sheet of plywood as an interior air barrier cleated and caulked to your new plywood floor decking and ceiling from the interior side, not cutting into the insulation layer. That would yield something like R25 "clear wall R", but if the same batts were thermally bridged by 2x6 16" o.c.with both top and bottom plates it would be come in less than R15 clear-wall. That's a big difference in performance for about the same material cost.

    With the 5.5" air gap to the brick veneer and the roof above it has plenty of exterior drying capacity and zero chance of direct wetting from wind blown rain since it's also protected by roof overhangs.

    Do you even need a vapor barrier? Where are you located? Dry plywood has a vapor permeance low enough to meet the Canadian NBC definition of "vapour barrier". If the code inspector won't buy that argument you can paint it with half-perm "vapor barrier latex" and be done with it, but you shouldn't have to. A plywood vapor barrier is more resilient than 6 mil poly or vapor barrier paint, since it's vapor permeance varies with the humidity of the proximate air. If there is sufficient moisture in the rock wool to be a mold problem the vapor permeance of the plywood goes up, allowing a faster drying rate toward the interior. But half perm paint is about a half perm no matter what- it meets code where vapor barriers are required, but it restricts drying.

    True vapor barriers such as foil or polyethylene sheeting creates as many problems as it solves, and can usually be designed out of the assembly. But in extremely cold climates it's still useful.

    Using fiberboard for the vent channels the right way to go in a cooler climate, using plywood or OSB (also variable permeance) as the interior surface.

  3. jmetrail | | #3

    Dana, those are some fabulous suggestions. The end wall is about 5' high. I think I can get away with not having full studs.

    Would this be consistent with what you are suggesting for the end wall?
    1. Add 3/4" strapping to the shiplap, and add fiberboard over the strapping - so there is an air channel behind the insulation
    2. Add 5.5" R22 rock wool
    3. Potentially add some 2x2s vertically, spanning from floor to rafters, for nothing else than to hold in the rock wool and support the OSB - they would not extent to the shiplap, so no thermal bridging.
    4. Add OSB to the interior facing side

    I believe R24 is now the minimum for exterior above grade walls in Ontario. Will 5.5" Roxul get me to that? I know the 16" batts are rated R24, while the 24" batts are only rated R22. I know in this application there will be no difference, but an inspector may someday want to see a package showing nominal R24.

    If I need more R, instead of OSB, could I use 1" foam to hold in the Roxul? That would get me to R27. I think at 1", it would be semi-permeable and I would not add any poly. The only downside, I suppose, is that for flammability reasons it would need to be covered with drywall.

    As for vapour barrier, I am in Southern Ontario. I know it is usually recommended here, but as you note, the attic roof and exterior walls will be well ventilated and should dry well. That raises 2 questions.

    1. Do you think for the flat ceiling I should bother with poly vapour barrier under the rock wool batts? Based on your comments I am leaning towards no. Drywall with regular latest paint, combined with air sealing, I think should keep enough vapour out. Again, the attic will be well ventilated.

    2. For the sloped ceiling, I need 4" of XPS foam. Does that create a vapour barrier whether I like it or not? Would a solution to that be to use a vapour-permeable, air impermeable, tape to seal the edges of the XPS, instead of the usual tuck tape? Does such a tape even exist?

    Very much appreciate your thoughts.

  4. Expert Member
    Dana Dorsett | | #4

    The strapping creating the air channel for the fiberboard is redundant, if there is 5.5" air channels between the ship-lap and brick veneer. If you want to use the fiberboard as an air barrier you can just mount it to the ship lap and seal the edges & seams with housewrap tape covered over with fiber reinforced duct mastic to ensure long term adhesion.

    The R24 minimum in the codes ipresumes thermally bridging framing. A 2x6/R20 + R4 continuous insulation wall comes in at about R17 "whole assembly" due to the thermal bridging. A continuous R22 layer outperforms that. If you're going to add sheet foam to bring the number up to R24 or larger for the inspectors, install 3/4" or 1" EPS on the interior side of the rock wool. Sheet foam is vapor retardent and if installed on the exterior side it would have to be sufficient for wintertime dew point control at the foam/fiber boundary in order for the stackup to be configured for drying toward the interior. You have a huge exterior drying channel already, so use it!

    Depending on where you are in southern Ontario it's either comparable to a US zone 5 or zone 6 climate (mostly zone 6), neither of which would demand a vapor barrier on the ceiling. This map isn't super accurate, indicating that it's all zone 6, not that it matters a whole lot here:

    If your location averages under 4000 celcius heating degree days (base 18C) it's really zone 5.

    With a vented cathedralized or flat ceiling with at least 1" of clear vent channel from soffit to ridge, the roof can dry into the vent channel. It won't hurt to use poly sheeting behind ceiling gypsum, or vapor barrier paint on the gypsum, even though it's not necessary. As with the walls, an OSB or plywood ceiling would also be sufficiently low perm to meet NBC code definitions, but you might have to provide some documentation to convince the inspectors.

    XPS is one of the least green commonly used insulation materials due to the extremely high global warming potential of the HFC blowing agents used (>> 1000x CO2). As it loses it's blowing agent over a few decades it's performance drops to that of EPS of similar density, about R4.2/inch. EPS is blown with penatane (~ 7x CO2) most of which escapes the foam and is recovered at the factory. EPS is not dependent on the blowing agent for it's performance, which remains fairly stable over a 50-100 year lifecycle. It's also usually cheaper per R than XPS. Polyisocyanruate is also blown primarily with pentane, but may need to be derated for temperature in your climate if it's on the exterior side of other insulation. If it's the entire insulation layer it's good for ~R5-R5.5 per inch, if it's on the interior side of mostly fiber insulation it's good for R6/inch or more. Low density polyiso usually comes with foil facers. A foil facer is a true vapor barrier, and is easy to detail as an air barrier with any high quality aluminum tape. (Something with a temperature rated adhesive such as aluminum tapes used on ductwork will surely stand the test of time in a behind wall or ceiling gypsum application.)

  5. jmetrail | | #5

    Thanks for more great advice. The downside to EPS on the pitched ceiling is the R value - in order to get R-31, I am going to need 5" of EPS under the rafters.

    ISO is a great suggestion. I avoided it due to cost, but here is seems it is actually cheaper than XPS: - XPS at $43.99 for 32 sqft. - ISO for $29.99 for 32 sqft.

    Am I missing something? I was sure ISO was generally more expensive than XPS.

    At R-6 per inch, I only need 3.5" of ISO which is great.

    Here's my next problem though - how do I fasten 1x2 strapping for the drywall into the rafters, through 3.5" of foam? The only option I see is 5" structural screws - which are expensive and which have hex heads that would not even sit flush with the strapping. Do you have any ideas?

    I could cut off half an inch of foam, and maybe be able to use 4.5" wood screws, but then I will be down to R-30 - not quite enough to meet code.

    If it's not possible to fasten 3.5" foam + strapping under rafters, I am back to the idea of scabbing/sistering more wood onto the rafters - which I really don't want to do.

  6. Expert Member
    Dana Dorsett | | #6

    JMETRAIL: As a rule polyiso is cheaper per R than XPS, with just enough retail exceptions to prove the rule. (EPS pricing is usually running neck-in-neck with polyiso.)

    Use 1x4, not 1x2s, which will tend to be shaped like corkscrews and have a high potential for splitting. And there's more:

    With the 1x4 strapping & 3.5" foam you'd be at 4.25", which means there would be ~3/4" of rafter penetration using 40d nails. Running the 1x4s perpendicular to the rafters would be wide enough for 2 nails at every rafter crossing, and there would be quite a bit of fastener retention capacity if using RING-SHANK nails. I'm not sure what code requires in your area, but it's pretty common for builders in New England to shim 1x4 strapping and ring shanks to level the ceiling under trussed roof, supporting the weight of both the ceiling gypsum + 18" or more of cellulose.

    Jon R: Just a guess, but for a residential buildings in southern Ontario the code R-value requirements are probably ALL about heating season performance, not it's cooling season performance.

    And since it's on the interior side of fiber insulation between the rafters, the mean temp through the foam is never going to be that warm, nowhere near as warm as it gets when polyiso is the only insulation, under a black hot-mopped flat roof in a commercial or industrial building where temp derating would actually amount to something that matters. (EPS and XPS have high-temp derating issues too.)

  7. jmetrail | | #7

    Even if the ISO derates to R 5.6/inch, it's still better than XPS and I am still getting my needed R-31. It'll be below R12 of Roxul, so it shouldn't get to hot or too cold (I heard ISO performs badly when cold too).

    The 5" ring shank nails are a great idea. I would have thought the retention on nails wouldn't be good enough, but using ring shanks at 2 per rafter I think would work.

    I am quite certain in my jurisdiction I don't even need a permit to add or replace insulation, and because I am not moving walls or doing anything structural, I probably will not be pulling one. Even though an inspector likely won't be seeing it, I still want to do it to code though. Advice is much appreciated.

  8. Jon_R | | #8

    > At R-6 per inch, I only need 3.5" of ISO which is great.

    Until you read their fine print about aging or the test report, then it's ~R5.6/inch when warm (75F). Maybe R5 in heating applications (insulation < 75F).

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |