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

The gift that keeps on giving: the wafer of cold air between floors in my house

user941025 | Posted in General Questions on

Hi again, GBA Forum:

Two and a half years ago, I had the roof ripped off my house and a second story put on. I paid for the envelope including door/window installs, the plumbing, and additional ductwork runs, and did the electrical, insulating of the second floor, and finish work myself.

Since the floor of the original attic was 2×4 construction (1920s bungalow) and there has been some settling over the years, the new second floor sits on a new sturdier deck that’s been built to be level, spanning the width of the house on LVLs, about a foot or so higher than the previous one. Insulating in that area was in the contract, and indeed that was done: By the time the subfloor was screwed down and the job was “done” and ready for me, there was an insulating job performed that covers the newly-made space. So between the old attic floor and the bottom plates of the new second floor. 2″ (or so) thick pieces of polyiso were foamed into place between where the new LVL met the rim joists. Raise your hand if you already see a problem. I certainly didn’t know this was what was going to happen; I guess I used to expect professionals to be professional and do their jobs well, so that I can go do my job, instead of sitting and micromanaging people in an area where I don’t personally have a lifetime of experience.

Ductwork and plumbing were run in this ostensibly-conditioned new space between the first and second floors. The plumbing access on the second floor is a small doorway to this space. A wall in the first floor was opened to run ductwork up from the basement (which I finally started to close up yesterday).

The problem is, when the first winter came (I’m in Minnesota), it became clear that there was still cold air penetration –significant– coming in below where the polyiso had been foamed in (which ended at the top surface of that previous attic floor). I started hanging indoor/outdoor digital thermometer leads into the between-floor space that first winter and measured temperatures of (if I remember correctly) about 40F. I’ll be calling this my cold air wafer. If you like, you can think of it as an Oreo cookie (I don’t care) where the sandwich cookies are conditioned space and the filling is a layer of The Great Outdoors.

A team of insulators came in and dense-packed the perimeter of my cold air wafer to about 18″ in. Maybe it helped. I’m not completely convinced–seems like there’s still room for cold air penetration. But it’s two years later and I’ve had a couple of energy audits, the house is mostly comfortable, the house has long since been sided and the flooring has long been down on top of the subfloor and I live upstairs and I deal with it.

So that’s what it’s been up to this point. Fast forward to yesterday: I finally got around to starting to close up the first floor interior wall where the supply ductwork runs from the basement up to that between-floor space, to feed the second floor. This wall is in the center of the house. With a piece of sheetrock up I can now feel cold air streaming out through those smaller gaps around that sheetrock. (Previously, with an entire open 27 sf piece of wall, it wasn’t apparent.) I’d tried foaming up around the top plate where the ductwork penetrates into the wafer but there’s still cold air. So while that’s unfortunate, now it seems that I have a new issue: this uninsulated ductwork (since it was run in an area that was supposedly inside the envelope) was previously kept in a (relatively) warm space, in part thanks to the open wall. Once I finish sealing this wall up, I’m going to be sending this warm air inside the duct in this leaky interior wall up into the cold air wafer, through a space colder than it has been. I’ll pour far less heated air into this cold wafer –that’s great– but I’ll also be up against the unknowns of how my uninsulated ductwork and my plumbing are going to do, once we learn how cold it really turns in there, when I’m not sending hot air in.

I’m curious to hear what to expect and how I can mitigate any problems at this point. I could go in under the top plate here again and try to foam the bejeezus out of it, which helps with the air leakage between the first floor and the wafer, but I’m concerned about how cold it’s going to get in there. Short of dense-packing the entire wafer or removing the siding on the entire house in a poor attempt to get back in and extend the insulation down around the perimeter, neither of which is really feasible, I’m not sure how to help this problem, nor what problems to really expect.

And, finally, since I mentioned the plumbing access on the second story, above, I may as well come back to it: that’s an uninsulated piece of plywood hatch in the middle of the second floor layout, between my bedroom and the cold air wafer. My bathtub is the metal membrane between my second-floor bathroom and the cold air wafer. My second-floor floor is the 3/4″ ply partition –with an additional 1/4″ of flooring for luxurious insulating warmth– between my conditioned space and the cold air wafer.

I look forward to hearing your opinions on how I should deal with the weirdly invasive cold air problems stemming from a construction job that passed inspection in 2008, and what to expect with my HVAC and plumbing running through that space once I turn it even more outside the envelope by closing up my wall. Thank you.

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. davidmeiland | | #1

    In my experience, most inspectors are not looking for possible air leakage at all, so passing inspection means nothing relative to this type of problem.

    You say you had a couple of energy audits. Did either of them find the air leakage sites and tell you how to fix them?

  2. GBA Editor
    Martin Holladay | | #2

    You have described what happens when renovations occur in an older home without performing the preliminary step of defining and establishing an air barrier for your home. As you have discovered, old homes have many discontinuities that can contribute to air leakage.

    If you are now throwing up your hands -- and it sounds like you are -- it's time to hire a home performance contractor experienced with weatherization work.

    If air is rushing upward through interior partitions, that's a clue that it is also escaping your envelope somewhere near the top of your house. You have to start with air leakage basics:
    - Outdoor air is entering your home through the crawl space or basement, so concentrate your air sealing work there.
    - Interior air is exiting your home through cracks near the top of your house, so you need to identify these cracks and seal them.

    Needless to say, it may be necessary to demolish some interior finishes to accomplish this work. Good luck.

  3. user941025 | | #3

    Hi, David.

    Energy audits and sealing leaks: Yes--one of those audits was even recently--a month ago. Sure. Other leakage sites were found and fixed both times.
    Recent blower door test CFM50 1484. Compared to other housing stock in the area, that's "acceptable." Now about this space I'm about to turn totally unconditioned....

  4. user941025 | | #4

    Thanks, Martin.

    To agree with you and then bring it back to the specifics of my newer concern:

    Penetrations to the attic are located and sealed. And for sure, there is plenty of insulating that can be done in my basement, and I recognize its effect is not insignificant.

    But I'm not convinced that the primary issue I'm facing, in what I'm describing, is cold air rushing in through my basement and hot air escaping upward through my attic.

    For one thing, any interior or exterior walls *above* this cold air wafer sit on top of the new second floor subfloor, where the penetrations are sealed with foam (the notable exception is the plumbing access hatch in the center of the room). There's no continuous chute past the cold air wafer for hot air to continue upward. Penetrations to the attic have similarly been sealed by a weatherization team that came in after I did the initial insulating. Stacks, vent fan ductwork, electrical wire penetrations and light penetrations, etc.: all foamed where they penetrate, sealed, etc.

    That weatherization team that dense-packed the perimeter of the space between floors did so in an attempt to mitigate the problem of that cold air penetration between those floors, around the area of the old and new rim joists. (They sealed many other places, too.)

    Back to the new problem facing me: where I do undoubtedly have heat leakage is to the perimeter, somewhere, of that unconditioned space between floors that was supposed to be conditioned. And I'm uneducated about what will happen with the plumbing and the temperature of the air in the ductwork running through that unconditioned space, and as I prepare to seal up a section of wall just below that unconditioned space, and close up a heat leak to that cold space for good, which might have been inadvertently beneficial, I am asking the opinions of this forum of people who have much more experience than I do. Thanks again, in advance.

  5. user-869687 | | #5


    If your air leak is just from the perimeter of floor framing into the air volume within the floor, then it seems like the problem should mostly go away if you continued with the dense packing strategy. You say there was 18" of dense pack around the perimeter, but is there really enough support to maintain density, at the edge of a joist bay? More likely the insulation has slid out over the ceiling and is hardly packed at the edges. If you dense packed the whole floor and embedded the heating ducts, there would be real density and much more resistance to air flow.

  6. user941025 | | #6

    That is likely true. I asked the insulators who dense-packed the perimeter if they'd fill the entire cavity originally, and they gave me some sort of explanation for why they wouldn't, and didn't. Maybe there's a legitimate counterargument to your suggestion, but I don't know it if there is.

  7. davidmeiland | | #7

    You need someone with an IR camera to find the cold spots. I like Thomas' explanation, and it would be easy to verify.

  8. user-723121 | | #8

    I did a 2nd story addition for a dear friend about 2 years ago, the rim joist area got very special attention by me. The exterior sheathing was R-10 Thermax and I insulated on the inside of the rim with an additional R-20 Thermax. Expanding foam was used to fill any gaps so the entire assembly was airtight and well insulated. With 150 sf of south facing glass this new space is bright and super comfortable with some 12' ceiling to boot.

  9. user941025 | | #9

    I'm bumping my thread because I'm alarmed with what I just came home to.

    It's 97F at 6:30pm right now. It's been hot and humid here in Minneapolis for a few days and nights solid, and I've been running the air conditioning (lest you judge, there are tenants and a dog here during the day too).

    One duct upstairs backs up against a drywall skylight tunnel that's visible from the first floor. That ductwork, unsurprisingly, is uninsulated because it was supposedly being installed in a conditioned space (which is in fact the unconditioned wafer between the two floors, described extensively above).

    Anyway, there is heavy condensation on the face of the drywall (meaning, inside the conditioned space) that's up against the back of that ductwork (that's the ductwork to the second floor, which runs through that wafer of unconditioned space).

    I'm attempting to attach a picture. It was tough to get a good one. Anyway, the condensation seems to extend significantly farther than just where the ductwork runs. What you're looking at is the drywall behind the back of a register boot.

    Anyone care to describe what precisely is happening here? Should I expect heavy condensation in that unconditioned space, too? Should I drill in and see if water runs out at this spot? If so, since my second floor deck is built upon LVLs, am I gearing up for eventual failure of those LVLs due to moisture?

    What I'd like to see is the builder who added on the second floor go back and remove the siding on level two, cut into the stucco on level one, extend the sheathing and insulate well, and then replace the exterior. Maybe I'm wrong, but my sense is that the builder's negligence has caused some problems here which are continuing.

    If you have constructive recommendations or questions, then I invite you to talk to me. Thanks in advance.

  10. user941025 | | #10

    Here's that picture. You're looking up a skylight tunnel from the first floor. This is near the bottom of that tunnel. You're looking up a wall face. The register boot is behind that wall, blowing the other way.

  11. GBA Editor
    Martin Holladay | | #11

    Your drywall is the coldest surface in the room, so that is where the moisture is accumulating.

    The solution is to insulate the ductwork.

  12. user941025 | | #12

    At this point, the only way to insulate the ductwork is probably to pack that entire unconditioned wafer, the width and length of the house, with cellulose (as Thomas Jefferson suggested above). Does that pose any risks in terms of drawing moisture further inside the house? I can't remember why, exactly, the insulators who came in with the cellulose to try to solve the problem said they didn't recommend it. (If there really is no good reason not to fill it, and no problems that would arise from it, then I think that's going to have to happen.)

  13. homedesign | | #13

    The problem could be lack of insulation on the ducts AND poor/orNO air sealing of the ducts.

  14. user-869687 | | #14

    Yeah, what John Brooks said. When I saw your picture I immediately thought, the drywall is cold because all the cooling is leaking from the ducts and chilling that air space inside the floor. There could be a serious leak like a disconnected duct joint, which would explain most of the problems.

  15. user-869687 | | #15

    The thing that's so suspicious about condensation on a ceiling is that the temperature inside that interstitial space must be considerably cooler than the room. It suggests that the mechanical system is delivering more cooling into that void than to the room, at least in proportion to the air volume. That's why it seems to indicate a massive duct leak rather than just uninsulated ducts or moderately leaky ducts.

  16. user-788447 | | #16

    Hmmm . . . It's always a little difficult with words only to communicate potential weak points of assemblies. The dew points we have had recently in Minneapolis are record highs. If you are air conditioning your home you would expect with a proper thermal envelope that at the same time your air conditioner was sufficiently dehumidifying the air. As you already have determined enough outside air is infiltrating the home elevating the relative humidity leading to condensation.

    I've read about a method of dens packing floor joist spaces where they will first insert an inflatable diaphram (or something of the sort) some distance back from the rim to act as a stop for the dens packing.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |