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

Yet another cathedral ceiling question

manglesio | Posted in Green Building Techniques on


I have a bit of a twist on the standard cathedral ceiling question. For reference, I’m in the Lehigh Valley, NE Pennsylvania, climate zone 5, the house was built between 1890 and 1900.

My house has a roof with two ridges, one gable-ended at the front of the house, which includes a finished loft on the 3rd floor but which had no air supply/return, one gable-fronted (gable-backed?) at the back of the house, which was totally uninsulated, to the point of having snow blow in during winter, and which contains the forced air furnace/ac unit, which was retrofitted during a gut reno and flip in 2010. There are soffit vents and gable vents at each gable, and the two are connected through a service door (an unsealed service door, naturally).

Anyhow, back to the roof. The loft room is insulated, as far as I can see from peeking in, with paper-faced fiberglass batts, facing towards the warm space. This is, I believe, standard, if totally inadequate. There are recessed can lights, because of course there are recessed can lights.

After seeing our first year’s worth of heating/cooling bills, having to repair the gutters due to ice, and having the loft unusable for 6 months a year (too cold in winter, too hot in summer), we decided to bring the furnace inside the building envelope, first sticking batts of Roxul (2xR15) between the rafters, then putting 3-inch rigid foam overtop (which is to say, underneath – towards the inside) and sealing with spray foam, which brings us to a theoretical R48 in the furnace area. We can see the positive change in that snow now just sits on the roof, and it’s obvious which section is insulated and which is not.

Last but not least, we ran a (small) air supply duct and a (large) air return duct from the furnace to the loft, so there’s a conditioned air supply now and the loft is usable 12 months a year. We have an ecobee sensor up there and it’s usually the same temperature (+/- 2 degrees) as sensors on the 2nd floor.

And after reading your article about vented and unvented cathedral ceilings we wonder if we messed up badly, possibly fatally, by a) not putting an air-impermeable but vapor-permeable layer on the warm side of the air gap, between the rock wool and roof sheathing,  and b) by not leaving enough space under the roof sheathing for the roof to dry. Also c), did we score the trifecta, and what we did is going to mess up both parts of the roof.

Fortunately, it’s a small house, a small roof, and it wouldn’t be the end of the world to pull it all out and redo. It would just be a pain.

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

    Did you remove the recessed lights? Did you tape and/or caulk the rigid foam to create an air barrier?

  2. Expert Member
    Peter Yost | | #2

    Hi Manglesio -

    A continuous air control layer is essential to all roof assemblies but extra important for unvented roof assemblies.

    You can't vent your way out of an air leakage problem.

    More important to have continuous control layer than it is to have venting.

    Venting is always a good thing, but it can't make up for a lousy or discontinuous air control layer.


  3. manglesio | | #3

    The rigid foam is taped or spray-foamed (or taped and spray-foamed) within an inch of its life, then checked over for leaks with an IR camera, so I'd say that this area is as close to air-sealed as it's possible for me to get it.

    I'll replace the can lights at some point in the near future (suggestions?), although they're above the loft as opposed to above the newly enclosed furnace room. The rest of the house is, well, 120 years old. There are air leaks.


    1. Expert Member
      Akos | | #5


      The newer low profile LED pot lights (ones with spring clips and a gasket on the back) are much tighter than the older can lights. When I was doing a blower test on a house, I was pleasantly surprised to not see leaks around these. If your existing cutouts are a standard size, these might be a simple retrofit.

  4. manglesio | | #4

    Update on this: I went in this evening to add a damper to a start collar without one. And, since it's cold out, I felt an air leak (and not a small one) from under the ductopus, where there's one run of flex duct tucked underneath another run of flex duct, which hid it from the IR cam.

    It looks like I have a few long evenings ahead of me, and this time I'm probably going to have to disconnect and straighten all the flex duct running along that side of the furnace to get at it.

    Thanks for your advice on sealing the attic tight - super tight, to the maximum extent possible - before worrying about the roxul. You guys are awesome. (And make sense, which is also useful.)

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |