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

Help to plan foam insulation with AC ductwork inside 4 1/2″ high rafter bay cavity in roof assembly.

berryjb | Posted in GBA Pro Help on

Our house was built around 1850 and has no insulation or air sealing in the roof. The finished attic has a partial cathedral ceiling, and the high velocity 4″ AC flex ductwork runs from the air handler which is located behind a kneewall in the attic, up one side of the cathedral slanted ceiling (7ft), across the flat part of the ceiling (about 10ft), and partially down the other side of the slanted ceiling, through rafter bays which are only 4 1/2″ deep, due to the very old house construction. Ceiling material is plaster lathe. I’d like to air seal and foam the roof assembly, but I’m concerned about spray or even pour foam crushing the flex duct, or with causing moisture issues with the flex duct not getting well insulated (part of the duct will have to remain in contact with the underside of the roof. I also get conflicting advice on open cell vs closed cell applied within the rafter bays, and overall on the underside of the roof assembly. I’m considering trying to replace the flex duct which goes through the slanted ceiling with PVC, if I can even access the ductwork enough to do so. I don’t have an obvious solution to moving the ductwork out of the rafter bays. Any ideas?? 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. GBA Editor
    Martin Holladay | | #1

    Existing flex duct can certainly be replaced with galvanized duct if you are worried about crushing the flex duct.

    However, that's a fairly trivial issue compared to other problems.

    First, ducts shouldn't be routed through insulated rafter bays.

    Second, I hope that the air handler is inside of the home's thermal envelope, not outside of it.

    Third, I hope you are coming up with an insulation plan that will result in an R-value that at least meets minimum code requirements (R-38 in Zones 2 and 3, or R-49 in Zones 4 through 8). You can't do that in 4.5-inch-deep rafter bays. For more information, see How to Build an Insulated Cathedral Ceiling.

  2. Dana1 | | #2

    Putting the duct in the rafter bay is a terrible idea from the get-go. Consider removing it and routing it between floor joists, even if it means the AC air would then be blown from the floor up rather than the preferred ceiling down. (It IS a high-velocity AC, after all, which won't have any problem overcoming stratification issues.) If the floor joists are somehow inaccessible or blocked, even have the air flow coming out laterally from the knee wall is better than running it up the rafter bay.

    Insulating the rafter bays have enough issues to pay attention to without the complication of the duct. What is you location/climate zone ? In climate zones 1-3 dense packed cellulose might be a better/safer option than a half-pound pour (open cell foam), or non-expanding injection foam. In climate zones 4 & higher a 2-lb pour of closed cell foam would be better, but at 4.5" of thickness it's vapor permeance is a mere ~0.25 perms it all but blocking the roof deck's ability to dry toward the interior, and the expense is pretty high.

    I you are planning to re-shingle in the next 5 years, cellulose now, and exterior rigid foam under the new roofing makes sense in any climate. The necessary amount of exterior foam required to protect the roof deck over the long term varies with climate. An inch would be plenty in zone 4, but by zone 6 the exterior R value needs to be at least as high as the cavity R-value (R16-R17 if cellulose).

  3. berryjb | | #3

    Followup - I agree that having any ductwork within the rafter bay is, politely, "suboptimal" - I unfortunately don't have a good option to relocate it. I'd hoped that using PVC might help reduce moisture problems within the bays. Although I will need to reroof at some point, costs estimates to add rigid foam on top of the roof decking have been shockingly high. We're in climate zone 4, so I'm struggling to even come close to recommended R value, and get any improvement I can, without causing moisture problems.
    I was hoping for a closed cell, low expansion foam I can add to the slanted part of the cathedral ceiling, without having to bore holes though the plaster - I do have access to the bays, both from the kneewall area, and from above the flat portion of the ceiling - so far all solutions I have involve cutting holes ...
    Will I cause harm if I don't insulate the rafter bays which contain the ductwork?? Thank you! Jim

  4. berryjb | | #4

    ... more - and yes, naturally the AC unit/air handler are currently in the unconditioned part of the attic - hence part of the motivation to change the thermal envelope to include the AC.

  5. Dana1 | | #5

    It's not always possible to cost effectively (on a lifecycle basis) retrofit an antique to IRC 2012 code minimums, but that's not to say you shouldn't insulate even if you can't hit that mark.

    Insulating the rafter bays with 4.5" deep celluloseo open cell foam (R16) with R15 of rigid foam (say 1.5" of polyiso + 1.5" EPS) over the exterior would outperform R38 batts between joists/rafters due to thermally breaking the rafters, and is probably do-able, and in some instances would be cost-effective.

    Getting to R49 by stacking 3" of polyiso + 2-3" EPS on the exterior may be difficult & expensive to implement on complex roof lines. With simple roof lines and using reclaimed foam rather than virgin stock it might be cost effective, but it may be more cost effective to install half the foam and use the savings on solar electric panels.

  6. berryjb | | #6

    Unfortunately 3 x dormers on the east side make this a complex roof ... followup please to the idea of insulating all the bays, except for those with duct work - will this cause moisture issues? How about if I air seal the top and bottoms of each of the rafter bays with the duct work? Another follow up please on open cell vs closed cell, especially in the rafter bays. I'd understood that adding in cellulose would cause moisture issues in the cathedral ceiling - correct?? Thank you! Jim

  7. Dana1 | | #7

    2lb pours are expensive and risky (blow-outs and voids are both common, and very difficult to fix) and as noted in my first post, nearly eliminates all drying potential for the roof decking. The thermal performance of the high R/inch foam is is also severely undercut by the thermal bridging of the rafters.

    In climate zone 4 on the unshaded east, south, or west facing pitches with typical dark (and not "cool-roof") shingles you'll be just fine blowing in cellulose. But on steeply pitched north facing pitches you may need to paint the interior side with "vapor barrier" latex to bring the vapor permeance down to 0.5 perms (which is not ideal, but it'll need something.)

    In zone 4A & 4B, as long as at least 30% of the total center-cavity R above the roof deck, you can use fiber insulation in the cavities with no risk. In marine zone 4C that drops to 20%. So if you installed cellulose (highly recommended over open cell foam here, for both blowout risk and hygric buffering reasons) you'd be no more than R17 in the cavities, and adding 2" of foam (any type) above the roof deck when you re-roof would make it moisture-safe from wintertime interior moisture drives, while maximizing the ability to dry toward the interior.

    If that plaster ceiling has multiple layers of leaded paint or foil/vinyl wall paper it's vapor retardency is quite low, and you would need to retain roof deck venting to avoid creating a moisture trap.

  8. Dana1 | | #8

    We're doing a lot of cross posting a the same time here...

    Not insulating the bays with the ducts doesn't create more moisture issues than it already has, but if the cavity has conditioned space air now that you're blocking any soffit-to-ridge venting the roof deck will experience somewhat higher moisture cycling. Blowing it full of cellulose would be better than leaving it exposed to the conditioned space air in winter without venting.

    Your understanding of cellulose is a bit off base. Cellulose doesn't cause moisture issues- it mitigates them, by sharing the moisture burden with the structural wood. It is highly vapor permeable, and fairly air-permeable, but when dense-packed it's air-permeablity is low enough to not much matter, as long as you have an air-tight interior. (Some installers claim dense-packed cellulose is OK for cathedral ceilings even in zone 6, but I find that assertion dubious. In zone 4 it's sorta-OK, totally fine for all pitches except highly shaded or north facing pitches. But even on the north facing issue it would take a decade or more for a real problem to occur, as long as the vapor retardency to the interior isn't lower than 0.5 perms.

    Open cell foam air-seals better, but is fairly vapor permeable, loses performance when wet (regaining it as it dries), and does not protect the roof deck nearly as well as cellulose or closed cell foam. Closed cell foam is protective at 1-2" at a vapor permeance between 1.2-0.6 perms when applied directly to the roof deck, but approaches moisture-trap territory at 4.5". It's also ridiculously expensive at $5 per square foot or more at 4.5" depth.

    Almost any roof configuration no matter how complex can be retrofitted with 2" of exterior foam. If you're going to re-roof in under a decade, cellulose cavity fill + exterior foam is the right way to go. With 2" of polyiso it's wintertime performance would be about R11 in your climate. With 2" of EPS you'd be at R8.5-R9 in winter. XPS would deliver about R10.5-R11 in winter for the first few decades, but it would eventually fall to that of EPS as it loses it's climate damaging HFC blowing agents.

  9. berryjb | | #9

    You've given me a lot more to think about - I'd thought that cellulose in a non vented rafter cavity was a definite no. The three dormers on the east side will keep at least some portion of that side of the roof in shade, except for perhaps mid summer. Thank you very much for the advice! Jim

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |