GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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

Insulating I-joists in a garage ceiling

Tony2slow | Posted in General Questions on

Hello. I have been doing some research to insulate a 2 car attached garage with a room on top. Size is approximately 27×27, new construction. This is in zone-6, Augusta Maine region, and code is R30 for floors.

The ceiling/floor is built with 16″ wood I-Joists set at 12″ on center. There are no plumbing or ducts in the space, only electrical wiring. I have air sealed the rim joist with caulking, then 2″ of Tuff-R using Great-stuff foam. With I-joists there is 16″ between the Advantech subfloor and the future sheetrock ceiling, and 13″ in the web area as chords are 2×4’s, with slightly under 12″ between the plywood webs.

Current plan is to use Roxul batts cut in half to fit between the joists. Options range from min R30 at 7.25″ thick to max R53 at 12.75″ thick. I am thinking that R46 will be acceptable, which is two layers of R23 at 11″ thick. The insulation will be installed so that it sits on the bottom chord, so there will be 3.5″ of dead air space above the insulation below the subfloor.

Also planning to put a vapor barrier below the insulation before the ceiling sheetrock, mostly to keep garage moisture from getting to the upstairs.

Not sure if I missed any information that would be important. Are there any issues with this plan, or anything that I have missed? Thank you for any hints and tips.

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. user-2310254 | | #1

    Tony. I think you will be advised to follow a cold-floor strategy for insulating the garage ceiling. This Fine Homebuilding article outlines the key points (, such as air sealing, completely filling the joist bay, and holding the insulation tight to the subfloor.

    A continuous layer of rigid foam, properly detailed and covered with 5/8 drywall, would help with thermal bridging and air sealing.

    It sounds like the rim joist might be a little under-insulated. Is the joist space between the garage and the rest of the house completely sealed?

  2. Dana1 | | #2

    Sculpting batts to fit I-joists perfectly is nearly impossible, made worse by the non-batt-standard spaceing. Blown insulation is really the "right" solution. This can be a DIY with a single stage box store rental blower. You can use landscaping fabric as a blowing mesh if you can't find any locally, or you can pre-install the ceiling gypsum and cut 3" holes every 10' along the joist bays. (or one hole in the middle of a 25- 30' span.) If you insert the blowing hose to the end of the bay and draw it back a foot or so, then blow until the blower stalls, back it out another foot, blow until it stalls, repeat you can hit 2.5lbs+ density with a bit of practice. with an intial blow of 16", with settling over time it'll stablize at about R50 with a 1-2" air gap between the cellulose & subfloor after a couple of decades. The air gap is fine, as long as you've adequately air sealed & insulated the band joist, which you have.

    With 16" I-joists the temperature striping along the floor won't be very significant as compared to dimensioned lumber joists. The web is only about 1/3 the width of 2x lumber, and with 15-16" of joist it'll be about R15, which isn't a terrible thermal bridge. (It would take 3" of continuous foam to barely double that!). You might feel the striping on a -20F day if you leave the garage doors & windows open, but most of the time it won't be noticeable.

  3. GBA Editor
    Martin Holladay | | #3

    I agree with Dana Dorsett. It's best to fill the entire joist bay volume with a blown-in insulation. Even though you end up with a higher R-value than the code minimum requirement, it's money well spent -- and you'll avoid potential problems with convective air currents or air movement (always possible when an insulation cavity includes "dead air space").

    If you ignore this advice and use R-46 batts with 3.5 inches of air space above the batts, it's essential that your air sealing work (especially at the rim joist area, but also at the ceiling) be impeccable. Air leaks will undermine all of the batts' R-value.

  4. Jon_R | | #4

    Omit the cold side vapor barrier. If the garage is too humid, ventilate it more. Do air seal very well.

  5. GBA Editor
    Martin Holladay | | #5

    Jon is right, of course. You never want to install polyethylene on the cold side of an assembly -- that's a disaster waiting to happen.

    Of course, your garage ceiling should be as airtight as you can make it.

  6. Tony2slow | | #6

    Wow, this is all good information. Let me hit up some of these points.

    Steve - Not sure I have enough head room for a layer of foam under the drywall. The rails for the garage doors and opener bar were installed for future 5/8 drywall. We cut the height tolerance pretty tight for the doors. Rim joist has caulking around perimeter of each bay, 2” Tuff-R board (R13) with Great-Stuff-Pro foam around the perimeter of the foam board, and Roxul will but up against the foam board. Do I need to add additional foam board? If so - what R value am I looking for? The garage is an add-on, and the house and garage + new room are being kept completely separate. Well, yes there will be a door between the two structures, but the rest is and will be completely sealed.

    Dana - I was not planning on sculpting the Roxul, just slice down the center and install so it rests on the bottom chord of the joist. Easy-peasy.!? I bought a bag and tried it; just to be sure it would work before I spent any real money. I am not sure that renting a blower and doing the blown insulation as a diy one-off job attempting to dense pack would have the best outcome. How many bays would I get through before I get the hang of it and think I know what I’m doing… possibly all of them?
    If the blown in will settle over time anyway, then I will still have the dead air gap under the floor anyway. The polyiso sheets that I have state that there is an additional R-value with a small ‘ideal air space’. Plus there is this,

    Martin – filling with blown in is probably around R40, per one insulation estimate, (although my calculation is >R50 for 16" thick), the Roxul batts are R46, which is similar, but has dead air space. From my descriptions, do I need additional air sealing at the rim joist? I think that the next step is to get all of the wiring holes plugged up. Not too concerned about going over code, but estimates have been very expensive for just getting to code, and none were going to air seal.

    Jon – no vapor barrier on garage side – got it. Air sealing well is planned, I don’t want smelly exhaust in the big room.

    What if I use 1.5” Tuff-R (R9.6) in between the top chords of the I-Joists. Then install R23 or R30 Roxul up against the Tuff-R and top chord plane. This would put the insulation at the hot side at R32.6 or R39.6. The dead air space would be larger and on the cold side, I would think that this method would have a higher degree of thermal bridging than leaving the air space at the top. Cost would only be one or two hundred more than the original R46 plan.

    By the way, the garage will be insulated, we will have to ignore that gaping 9x7 hole for a few minutes twice a day, but not heated, so I am thinking that it will not be as bad as an open crawl space.

    Thanks for all of the data to ponder.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |