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

Wisdom of closed cell foam insulation in an unheated attic

GEDeGr | Posted in Green Building Techniques on

I own a house in upstate NY due east of Albany.  The plant hardiness zone is 5b.
The house is a mid-19th century 2-story federal style farm house, with a dug basement (dry with concrete floor), and an attic space that was finished by a prior owner into two bedrooms for some of the 8 kids in the family.  (The house also has 1.5 story mid-19th century wing without a basement, and a mid-1980s addition consisting of two rooms off the back of the house and an attached 2 car garage, over which is a large bedroom and bath.)
My question relates to the attic in 2-story farmhouse.  To save on energy consumption, I stopped heating the finished attic space and closed it off from the 2nd floor, but not tightly sealing it.  The unheated old attic space was far from air tight, in fact had a lot of air leakage from both outside and inside the house.  Not surprisingly, the sheetrock in this cold unheated space grew a nice mold crop from condensation from warmer, moist air rising from the second floor.  
At this time, I am having all of the sheetrock removed, and the fiberglass insulation that was behind it removed.  The exposed structural elements will be treated for mold.
When that is done, I have been contemplating bringing the attic back into the heated envelope of the house with spray foam insulation.  My reading is leading me to the conclusion that my best bet is closed cell spray foam.  The aim will be to eliminate the migration of cold air from outside into the space.  
This is being done in conjunction with a larger project of tightening up the house and installing a geothermal heating system to be the primary heat source for the house (but keeping my current oil burner in reserve status).  The geothermal will be primarily water-to-water for heating, with a water-to-air auxiliary unit in the (to be foamed) attic, which will also provide cooling for part of the house in the summer.
Getting to the point(s):
1.  Is the overall attic plan generally sound?
2.  Is closed cell the way to go in the unvented, within the envelope attic-space?
3.  Is there any need for ventilation between the underside of the roof sheathing and the spray foam insulation? Would it be best practice to have ventilation between the sheathing and the foam?  If so, what are the concerns?  FYI, the roof sheathing to be foamed consists of the original 150+ year old planks, on top of which was laid pressed chip board (in sure that’s not the right name) panels when the house was re-roofed in the mid-1980s.  The current roof was installed in 2008, so it’s 12 years old.  
4.  Am I correct that with a properly air tight (vis-a-vis the outside), insulated within-the-envelope 3rd floor space, the only heating required is that which migrates up from the second floor, which is closed off, but not sealed off, from the second floor?  
5.  Am I also correct that the third floor, if properly sealed-from-outside air intrusion, should not have humidity/condensation issues?   
6.  A leaky roof insulated with closed cell spray insulation is reported by some to be a concern. Is this a real risk?  What can be done to mitigate the risk?
7.  A roofer has mentioned the risk of spray foam insulation lessening the useful life of roof shingles because they can’t radiate summer heat inward.  Is that a real risk?  I’ve read that was a concern that has been overblown.
Thanks.  And sorry for the verbosity.


  1. user-2310254 | | #1


    I will give your post a bump. In terms of priorities, it might make more sense to start with insulating and airsealing the basement to reduce the stack effect. Check out this link ( and the articles in the sidebar.

    Can you post a couple of pictures of your roof?


    1. GEDeGr | | #4

      Thank you. To keep my post short (?) I didn't mention that I am have spray foam insulation installed in the basement where the floor joists meet the sill, after some old fiber glass insulation is removed. I am also trying to see if I can figure a way to get into the shallow crawl space under the old part of the house without a basement. There is no way in I have found, other than pulling up the floor, which I'm very reluctant to do. At this point, I don't plan to insulate the laid-up stone basement walls.

  2. GBA Editor
    Brian Pontolilo | | #2

    Hi Garrett.

    The short answer is that you can create an unvented roof assembly with spray foam insulation between the rafters, and if you have the rafter depth required (you didn't say what the rafter dimensions are), create a insulated roof assembly that will create a tighter, more thermally efficient, and vapor controlled assembly. An unvented assembly, designed right, is often the best path in retrofits because you don't have to give up rafter depth to the vent channel.

    That said, filling the rafter cavities with expensive, and environmentally-harmful closed-cell spray foam is not necessarily the only option and not always the best option. As long as you install a thick enough layer of closed-cell spray foam for condensation control, you can then fill the rest of the cavity with a more cost-effective and environmentally-friendly insulation like cellulose. I suggest you read this article, which will offer options and explain the details needed to insulate your roof well: How to Build an Insulated Cathedral Ceiling.

    When it comes to heating the third floor space and controlling humidity, tightening up and insulating the roof assembly will help, but this depends on other factors as well, including the rest of the house's envelope details and your mechanical equipment and delivery systems.

    1. GEDeGr | | #5

      The rafters are 3x5s, so I have basically 5 inches depth in the rafter cavities, and 20-21 inch spacing between the rafters. The attic walls at the ends of the house as 4 by 4 framing, giving me 4 inches of depth between the structural elements.
      I have a closed cell spray foam proposal for 4" of closed cell foam (3.5 nominal depth) on the roof deck, which will work with the roof rafter system, and 3" of spray foam on the end walls, again, depths that will work with the construction. The depths of foam are stated to deliver R-25 on the roof deck and R-21 on the end walls. Does that sound right?
      Considering the shallow depths I'm dealing with, do you see a path to using insulation other than spray foam?

  3. Expert Member
    Dana Dorsett | | #3

    Closed cell foam (even HFO blown closed cell foam) is one of the least-green options:

    If the rafter can be vented from soffit to ridge, there are cheaper/better/greener options. If it can't be vented, a hybrid foam/fiber insulation approach

    You are in DOE climate zone 5A. Per the IRC chapter 8 & 11 prescriptives in zone 5 as long as R20 out of R49 (or a ratio of a bit more than 40% of the total) is closed cell foam tight to the roof deck the risk of moisture accumulation at the foam/fiber boundary over a winter is small. In an 11.25" 2x12 rafter bay that could be 3" of HFO blown closed cell foam (R21) at about $4/square foot), plus 8.25" of blown cellulose (R30-R31) for a heluva lot less per square foot, adding up to more than R50. The sequestered carbon of R30 cellulose pretty much cancels out the CO2e footprint of R20-ish HFO blown foam.

    If your rafters aren't 2x12s there are other options, so tell us how much depth you have to work with.

    The drying rate through 3" of foam is slow & seasonal, but relatively safe if the roof deck is under 20% moisture content (check it) when applied. It's more than twice the drying rate of an R49 all-closed cell foam solution.

    See also:

    (skip forward to Tables 3 & 4 if you don't have time to read the whole thing, but it's pretty short.)

    The mold problem in the attic is as much an infiltration problem as it is an insulation problem. With air leaking out of the attic it draws humid air up to the cold attic space from the conditioned space below. Spending some of the closed cell foam budget for air sealing and insulating the band joist & foundation sills in the basement (where accessible) is better than going higher-R with closed cell in the attic. The stack effect driving the infiltration is a function of the vertical distance between the lowest & highest elevation leaks in they building- sealing the bottom of the stack is almost as important as sealing the top.

    If you're not in Dandelion Energy's service area ( )a ground source heat pump solution isn't going to fare favorably against cold climate mini-split heat pumps. With any solution it's important to know the whole-house heat load as well as the room by room loads before diving in. Do NOT leave this up to HVAC contractors. Since you have a heating history on the "before" picture, it's possible to get a good handle on the whole-house load using these methods:

    (I can walk you through that if you're willing to share a ZIP code and oil fill-up dates & amounts.)

    The room by room loads can be inferred somewhat from the radiation sizing in each room, assuming the temperatures are reasonably balanced and comfortable. Otherwise, a Manual-J (performed by an engineer or RESNET rater, not an HVAC contractor) or I=B=R method done in a spreadsheet too can get you there:

    The Manual-J or IBR is likely to have numbers somewhat higher than a fuel-use derived load number, but as the whole house load number is not more than 25% higher than the fuel-use derived number it's going to be reasonable enough to use for estimating the room loads.

  4. GEDeGr | | #6

    I have 5 inches of depth with the old rafters, which are spaced about 21 inches apart.
    To keep my post short, I didn't mention that I am also planning on closed cell foam insulation at the foundation of the house (in all areas that are accessible [some areas are not]).
    I am in Dandelion's service area and started my exploration of the feasibility of GT with an analysis of my heating load (based on data collected by various sensors over two weeks, multiyear of fuel oil consumption data, etc).

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |