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

Insulating the underside of a concrete deck

jonscholl | Posted in Energy Efficiency and Durability on

I have a 2-story garage project where the lower level is built with ICF walls…R-25….and PEX radiant heated floors with an R-10 underslab insulation but the ceiling consists of the metal concrete pans and +/- 8″ of concrete and rebar. Assume that space above is unheated… might also get insulated and heated from time to time when the homeowner needs to do automotive work…..but in general it will be unheated.

The homeowner needs to maintain a 55 degree minimum heat in the lower level because it will be used for storing antique cars which do not play well in freezing temps. The dimensions of the space are 24′ x 34′ x 8′ high. Due to steel support beams the 34′ direction is broken into (2) 12′ sections and (1) 10′ section.

The question is how “best” to insulate the ceiling..both from an upfront cost and a long term energy performance cost? I had been thinking of attaching (3 rows) of PT 2×4’s flat to the steel pans using Tapcons and then installing regular 2×4’s…also in the flat….. in the perpendicular direction at 24″ OC to attach our drywall to. We would then run our electricals and piping as needed in this 2x space and then spray foam with closed cell to the bottom face of the regular 2×4’s. This would give us a minimum of 3″ of foam where we just have our 2x materials attached directly to the steel pan and as much as 8″ of foam where the steel pan goes up. But as I was reading various articles about spray foaming cathedral ceilings and roof decks, I became concerned about thermal bridging and water vapor getting into the 2x materials over time to cause rot and rusting of the screws and metal pan. We are trying to build “100 Year Buildings” so I am trying to be cautious without breaking the client’s budget. Maybe they are mutually exclusive?? Project is located outside of Philadelphia so we are a boarderline Zone 5. Martin had a table showing that IRC requires R-20 insulation to ensure water vapor does not condense on cold surfaces? I thought it was R-10??

Any comments are welcome! Thanks.
Jon Scholl

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

    I hope that this project is at the design stage. Construction hasn't started, right? Because GBA readers know that all insulation decisions need to be finalized before construction begins. (For more on this topic, see Plan Ahead For Insulation.)

    If you are still at the design stage, the most valuable change you can make is to switch from an 8-foot ceiling to a 9-foot ceiling. Unless you have lots of room to work with, it's going to be hard to solve this insulation problem.

    So -- can you design a higher ceiling?

    Otherwise, you have two choices:

    (a) Insulate from below with closed-cell spray foam. The spray foam will need to coat not only the steel pans but the beams that support the slab. Everything needs to be covered with the spray foam to avoid thermal bridging. In Zone 5, you're aiming for R-49 insulation -- so this is going to be expensive. (A 9-foot ceiling would be cheaper.)

    (b) Insulate from above, using several layers of rigid foam on top of the slab, followed by a new slab above the rigid foam. This will only work if you can address the thermal bridge at the perimeter of the slab.

  2. jonscholl | | #2

    Sorry to say Martin.... it is built and 8' is what we have to work with. AND the client wants to keep the bottom portion of the beams exposed so that he might be able to use a chain hoist attached to a wheeled carriage on the bottom flange of the I-beam. So I have shot holes in all of your concerns. Insulating above is also not an option....unless of course we take a closer look at the insulating of the whole garage above ...which we are planning on doing......and heating it on a consistent basis......which we were not planning on doing. I suppose then the underside of this concrete pan slab need not be insulated at all, eh?

  3. GBA Editor
    Martin Holladay | | #3

    The situation isn't good, so all you can do is compromise. Put as much spray foam as you can on the ceiling. If the beams stick through, you've got some big thermal bridges.

    It won't be a well insulated ceiling, but you've got a lot of limitations preventing better options.

  4. Dana1 | | #4

    Unless you are actively humidifying the garage, there are no real moisture sources to speak of and the dew point of the garage air will track that of the outdoor air, with minimal risk of moisture.

    After running the electrical, insulating and encapsulating the whole mess with open cell foam from the steel on down, fully filling your 2x4s is more appropriate and higher performance than 3" of closed cell foam, and would prevent convective transport of moisture to cold surfaces. If you are for some reason still concerned about wintertime vapor diffusion, install 2 mil nylon (Certaineed MemBrain) behind the ceiling gypsum as a "smart" vapor retarder, or use half perm "vapor barrier latex" primer on the gypsum itself.

    The beam is still a big thermal bridge, not great, but not a huge issue for zone 4A (Philly is not zone 5) location only heated to 55F. Insulating the space above (heated or not) would be a big improvement.

  5. jonscholl | | #5

    Thank you for your additional comments. Am I understanding correctly that you feel Open Cell foam is a better material in this situation than Closed Cell?....because I can "fill" the whole space with Open Cell for less money than the 3" of Closed Cell?

    The upper flange of the two steel beams will be encased in foam also which means thermal transfer is only happening thru the web but it is steel and cold loves steel....I get that.


  6. Expert Member
    Dana Dorsett | | #6

    At 10 cents per R per square foot you can sure get a lot more for the money with open cell foam than closed cell at 17-18 cent's per R-foot. You don't really need the lower vapor permeance of closed cell in your application, and it's higher R/inch buys you damned little when thermally bridged by 2x4s, let alone a steel I-beams.

    With 2x4s over 2x furring you have 6" of depth, which will give you R22 at center cavity if you fill it all up, at a cost of ~$2.25 per square foot, and with a longer path through the framing the thermal bridging is lower.

    With 3" of closed cell even R7/inch rated foam (usually exaggerated- figure R6 really), you'd have only R21, with half the distance (= twice the thermal bridging) through the wood framing & I-beam web, at a cost of ~$3 per square foot.

    That's more money for lower average performance. Even if you went the 2-mil nylon "just to be sure", that adds only ~12-15 cents per square foot in material cost, which would still cost less overall.

    From a total greenliness point of view, 6" of half pound foam runs 3 lbs of polymer per square foot compared to 6 lbs per foot for for 3" of 2 lbs. foam. That's half the polymer use, with higher net performance. And open cell foam is blown with water, which has a very low environmental impact (d'ya think? :-) ), where as almost all closed cell is blown with HFC245fa, a powerful greenhouse gas (~1000 x CO2 @ 100 years). It's better to save the closed cell foam for situations where it's lower vapor permeance and higher R per inch actually buy you something. Here it would be buying you less than nothing compared to a full-fill open cell solution.

    How are the walls & ceiling of the room above constructed? Any windows?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |