GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon
Musings of an Energy Nerd

Burying Ducts in Attic Insulation

Building codes provide guidance for contractors, but the details can still be tricky

If you bury attic ducts in cellulose or fiberglass insulation, you need to think about summertime condensation. In a hot, humid climate, moisture can condense on cool ducts and register boots when the air conditioning system is blowing cool air through the ducts.
Image Credit: Image #1: Home Innovation Research Labs

You shouldn’t install ducts in a vented unconditioned attic. As most GBA readers know, ducts belong inside a home’s thermal envelope, not in an environment that’s cold in the winter and hot in the summer.

Efforts by energy-efficiency experts to convince builders to move their ducts indoors haven’t been particularly successful. In states where slab-on-grade homes are common — in other words, in states where basements are rare — most builders still install ducts in vented unconditioned attics.

According to building scientist Joseph Lstiburek, the energy penalty associated with attic ducts is often in the range of 30% — that is, 30% of the home’s heating and cooling energy is lost to the attic. By performing two important steps — carefully air sealing all of the duct seams, and moving the air handler into the home’s conditioned space — the energy penalty associated with attic ducts can be lowered from 30% to about 10%.

Even better performance can be achieved, however, if the ducts are installed near the attic floor and buried in deep cellulose or blown-in fiberglass insulation. Historically, the main impediment to implementing this approach has been the concern that condensation can form on the exterior side of the ducts during the summer, making ceilings damp.

Why do attic ducts sweat?

If ducts aren’t buried with insulation, they usually don’t sweat. Most attic ducts are insulated flex ducts (that is, flex ducts with R-6 or R-8 fiberglass insulation, along with an exterior polyethylene jacket). When these ducts are exposed to exterior air, the plastic jacket never gets cold enough to permit condensation, because the hot attic air keeps the plastic jacket warm. But if the same ducts are buried in cellulose or fiberglass insulation, the added insulation lowers the temperature of…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. User avater
    Dana Dorsett | | #1

    Diffusion venting
    According to Lstiburek:

    The best way to do buried ducts, unless you live somewhere where you have to worry about ice dams, is to close up the soffit vents, and replace the ridge vent with a diffusion vent.”

    I see prescriptive practices regarding diffusion venting have been addressed in the IRC 2018 (at least for US climate zone s1, 2, & 3.) See section R806.5, subsection 5.2:

    Lstiburek has been advocating putting this into code for awhile now!

  2. User avater GBA Editor
    Martin Holladay | | #2

    Response to Dana Dorsett
    Thanks very much for your comment. I was unaware that Lstiburek's lobbying efforts on behalf of his vapor diffusion port idea had been accepted by code officials in the latest version of the IRC.

    Note to GBA readers: Most jurisdictions have not yet adopted the 2018 IRC, so it's premature to assume that your local code official will agree with an attempt to adopt the vapor diffusion port approach.

    I have edited my article to include information on the new code provisions for vapor diffusion ports.

    Below is an image showing the provisions of the 2018 IRC, section R806.5.5.2, cited by Dana. Click the image to enlarge it.


  3. User avater
    Peter Engle | | #3

    Science is cool.
    I was also not aware of the new section on diffusion ports. Very cool. And it shows that actual science can still make a difference. Joe understands the physics and proposed these as a potential solution for several different climates. Then, he did quite a bit of serious and careful experimental study to show that these work, and the code community took notice and adopted the techniques. It all actually happened relatively quickly, for the building code folks. Nice job. Can't wait until he's got rules for colder climates.

  4. User avater
    Dana Dorsett | | #4

    Code official signoff @ Martin
    Most code officials I've dealt with (OK, it's a small number, to be sure) usually go along with approaches & values that appear in newer versions of the code not found in the current state codes if you can point them to the relevant sections.

    I have a bit of concern about the minimum 20 perms requirement in subsection 5.2.3 if the house being built is in a hurricane zone. There is a hurricane resilience aspect to unvented roofs which is desirable, but materials north of 20 perms that are sufficiently rigid and can be fastened well enough to withstand hurricane force winds without ripping out (turning the unvented roof into a vented roof) are limited.

    Weather resistant & rigid vapor permeable 3/4" asphalted fiberboard runs about 15 perms, half-inch about 19 perms, and wouldn't quite meet code. MDF without the asphalt, overlaid with a high permeance housewrap probably makes it at 3/4", but the vapor permeance of untreated MDF isn't normally found in the specs for MDF, since in it's usual applications vapor permeance doesn't matter.

    Weather resistant fiberglass faced gypsum board such as 5/8" DensGlass is specified at 17 perms (minimum). Half-inch DensGlass is 23 perms (min), but is half-inch gypsum board really rigid enough with sufficient fastener retention for the uplift of hurricane force winds at the roof peak mounted to 24" o.c. trusses/rafters? (I doubt it, but don't know.) It has to be better than housewrap, but is it good enough?

    Until diffusion venting becomes more common it's doubtful that products to solve this issue will magically appear. Perhaps future revisions of the code would allow slightly lower permeance if combined with slightly larger vent ports at the ridge (currently limited to 12" max from the ridge, per subsection 5.2.1.)

  5. Malcolm Taylor | | #5

    Diffusion ridge vents
    Does anyone sell a manufactured ridge vent big enough for a diffusion vent at the peak? Joe L's detail seems to show a fairy large area was necessary.

  6. User avater
    Jon R | | #6

    diffusion vent
    It would be interesting to know how the required 50 cfm to the attic balances out in terms of diluting humidity in the attic vs providing a little bit of positive pressurization (reducing humidity gain via summer infiltration).

  7. Randy Williams | | #7

    Vapor diffusion port
    The way I'm reading the code is the vapor diffusion port, which is located at the peak of the roof, is the only part of the roof assembly that requires a minimum 20 perm. I'm in zone 7, so the code doesn't apply to me. I could be wrong.

    Peter Yost has an article talking about this code change on the JLC website.

  8. Mike Roberts | | #8

    I build in climate zone 4 (Bellingham WA) and the only ducts in the attic are ducts to the HRV. Do these ducts (both those going to outside and those going to indoor locations) have the same condensation concerns as those connected to an Heating/Cooling system?

  9. User avater GBA Editor
    Martin Holladay | | #9

    Read the installation instructions for the HRV. The fresh air duct connecting the fresh air intake to the HRV needs to be insulated (to minimize the chance of condensation). If the HRV is located in a vented unconditioned attic (usually a bad idea in a cold climate, but perhaps possible in Zone 4), then you also want to make sure that the ducts conveying fresh air from the HRV to the registers are insulated, and you will also want to insulate the ducts conveying stale air from your exhaust grilles to the HRV (so that you don't lose heat scavenged by your HRV in the winter).

    1. Mike Roberts | | #10

      Sorry, I didn't mean to imply that the ducts are un-insulated entirely. We insulate the ducts or use insulated flex duct. What I was asking is if we can bury the ducts in blown in attic insulation without the concerns about the condensation issues addressed in your article.

  10. User avater GBA Editor
    Martin Holladay | | #11

    The concerns about condensation addressed in this article arise with ducts connected to an air conditioner. These issues do not arise with ducts connected to an HRV, because ducts connected to an HRV never get cold enough to raise condensation concerns.

  11. user-7527360 | | #12

    "Buy large diameter R-6 flex duct, and pull the R-6 flex duct over R-8 flex duct, as if you were sliding a sleeping bag over the R-8 duct. You end up with “flex duct inside of flex duct.”"

    Does code allow this? Isn't this method of achieving R-13 flex duct creating multiple concentric vapor barriers?

    1. User avater
      Dana Dorsett | | #13

      I'm not sure about code (I doubt it's even addressed) , but all flex duct is concentric vapor barriers- not a problem since it's air-tight.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |