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

Insulation gap in or out?

ddubx6 | Posted in General Questions on

Installing a 5 1/2 batt of Roxul in a 6″ wall. Does it matter where the batt ends up? Up against the sheathing or up against the drywall?

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. GBA Editor
    Martin Holladay | | #1

    First of all, most 2x6s measure 5 1/2 inches.

    But if your wall is framed with full-dimension 2x6s, it's best to use thicker insulation (and compress it somewhat) rather than have an air gap. The air gap can allow convection currents which lower the thermal performance of the wall.

  2. Expert Member
    Dana Dorsett | | #2

    What-he said- leaving a gap allows free convection between the gap and the entrained air in the batt, reducing it's R value. But worse yet, a channel where air can bypass the batt completely can take an even greater toll. Even the best batt installation had some gaps and compressions cutting into performance, but leaving a half-inch gap would be a waste.

    Batts are designed for a compression-fit. When installing they need to first be tucked into the edges and corners of the stud bay to not leave a long skinny air gap where the sheathing meets the framing, then gently tugged back to where the surface of the batt is just proud of the stud edge plane, so that it will contact the wallboard fully when installed. It's far better to compress a batt a half-inch than to leave a quarter-inch full-surface gap.

    If you have a nominal half-inch of excess depth a cut'n'cobble job using unfaced 1/2-3/4" thick rigid EPS foam to fill that space may be a reasonable DIY solution.

    1. ArayaHomes | | #8

      Hi Dana and Martin, I am a performance home designer for some background on myself. I am just curious about the recommendations here not leaving a gap. If there is a gap of air on the interior side between insulation and interior finish, the air space will become warmer, so long as there is enough insulation depth in between the studs I'm not sure why the warmer air space would make much of a difference other than the potential for thermal bridging possibly as convective flow would be in more direct contact with the studs.. There is a building science article by Joe on ( and it seems to be in line with what I'm referencing here. Wondering what is actually the answer here. I have a feeling that this is obviously about a specific question by the person who first asked on this thread their question, however could you be more specific and general about the air gap debate. Thanks

      1. ArayaHomes | | #9

        I'll be more specific as to why I ask as well. This is an assembly I am evaluating and although I am not a big fan of this assembly, this is what we've got in my case. From exterior - interior ( Asphalt shingles, tar paper, 1/2" osb, 2x4 cathedral roof rafters exposed currently, filled with ccspf directly against underside of roof sheathing filled at +/- 2.5" thickness, air space of +/- 1", proposing continuous insulation of 1 layer or possibly 2 layers of 1" rigid rockwool to give the assembly a fighting chance to dry towards the interior, plywood veneer mechanically fastened to rafters. In this assembly the veneer so long as it is fairly air tight and the rockwool comfortbatt at 1 layer of 2" or 2 layers of 1" although permeable would help fight the heat flow, by the time the air space is hit, it's probably got little to no moisture inside of it and would be as warm as interior air since ccspf is the vapour barrier in this case outboard of the airgap inside of the rafter cavity. I would guess that the air space in this assembly will be of no harm and I don't understand how warm convective air could lower the r value of insulation, wouldn't it be cold air rather which would lower r value performance?

  3. Kamboji | | #3

    I'm in exactly the same situation as Darren with full 6-inch rafters. Isn't there always an air gap anyhow between drywall and insulation, due to the thickness of the strapping? Because of this, I was planning to press the Roxul all the way in.

  4. GBA Editor
    Martin Holladay | | #4

    Darren was talking about wall studs, not roof rafters. Walls usually don't have strapping.

    Rafter bays can be insulated various ways, and some builders install strapping under rafters before installing drywall. However, energy experts advise that it's best if the interior air barrier (the drywall) is in direct contact with the insulation, with no intervening air gap -- especially if the insulation is air-permeable.

    If the insulation in question is a layer of foil-faced polyiso, then the installation of strapping (and the air gap that comes with the strapping) makes sense. In that case, however, the polyiso seams need to be taped -- so that the polyiso becomes the air barrier.

  5. Kamboji | | #5

    Martin- thanks for the clarification. I am planning to put carefully-taped foil-faced polyiso below the Roxul under the rafters, so maybe the small air gaps on both sides of the polyiso are ok (fingers crossed). I have an uneasy feeling about it though because I'm planning to put foam board on top of the roof as well. (And code seems to prohibit having a class 1 vapor barrier underneath an unvented roof assembly with permeable insulation.) Maybe ok if for the above-roof foam I use EPS rather than foil-faced to allow vapor to escape?

    Sorry to hijack the wall discussion with a rafter discussion!

  6. Expert Member
    Dana Dorsett | | #6

    Unless you have at least a 1" gap between the roof deck and Roxul with both ridge & soffit venting, putting putting foil facers on the interior side is going to create a moisture trap. Shingles & roofing felt stackups are on the boundary between class-I & class-II vapor retardency when dry at about 0.1 perms, and when the roof is wet there is exactly zero drying toward the exterior.

    While 0.1 perms could be a "drys before November" vapor retardency in a cool but arid climate, the 0.05 perms of the foil facers means that the drying toward the interior is a "drys before November of year 2499" situation.

    If you put sufficient foam board above the roof deck to meet the IRC 2012 prescriptive R for your climate you'll be fine with Roxul fulling filling the rafters, and standard latex painted gypsum or unpainted OSB on the interior side as the vapor retarder.

    The prescriptive exterior-R is predicated on R49 total R (center cavity), and its the ratio of exterior R to cavity R that determine the temperature of the roof deck, which needs to average above the dew point of the conditioned space air to remain dry. (Typically 38-40F in US climate zone 5 & higher.) So if you're in climate zone 6, with R25-ish Roxul in the rafters you'd also need R25 foam above, but in climate zone 4A the minimum exterior foam for R25 cavity fill would be about R11, even though that would be below the IRC 2012 prescriptive for both the total R (R36 instead of R49) and the exterior foam R (R15.) But for the extra inch of foam you'd be closer to code on both.

    This also means that the more R you put under the roof deck, the more you need above the deck. Since it's both easier to install and more beneficial from a dew point control point of view to put the interior-side EPS you're considering on the exterior, that's where it belong. Having it on the exterior places the roof deck in a warmer (=drier) layer in the stack up.

  7. Kamboji | | #7

    Thanks so much for your detailed and thoughtful comments. I'm in climate zone 6 (near Bangor ME), so I was planning for R25 of foam on top as you said. Your suggestion makes a lot of sense, and also will save me money since as you noted, putting the foam under the rafters is more difficult labor-wise. Unfortunately this arrangement will leave me with slightly less than R-49, because Roxul is R-23. I could add more EPS to the roof than planned, but I'm not sure how far I can push it with the roofers, as I understand that it's challenging to install a very thick layer of foam. With polyiso, one only needs 4 inches to get about to R-25. But EPS is only R-4 per inch, right? So I'd need 6 inches of it? (I'm aware that XPS and EPS perform better on the roof in my climate than polyiso, thanks to Martin's blog post about this.) Perhaps an initial layer of polyiso, with EPS above it will get me to the right spot, while mitigating the negative effects of the colder temps on the polyiso. Sorry again for going off topic.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |