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

Community and Q&A

Insulate this!

FrankFulton | Posted in Green Building Techniques on

This one-room addition has no foundation. There is insulated in walls (likely R11) and ceiling (likely R19 – certainly no more), but not in the floor. As you can see, outdoor air can simply pass through the open space under the floor. When the flat EPDM roof needs to be replaced in 5-10 years, we will install exterior foam board outside the roof deck. Heating is primarily warmed air from rest of house or propane fireplace. Cooling is AC from rest of house.

I’m considering digging 2′ out under the floor, and sealing the underside with foam. What approach would you take?

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 avater GBA Editor
    Martin Holladay | | #1

    You don't describe the foundation, but I'm guess that this room is on piers. For more information on pier foundations, see Crawl Spaces vs. Skirts.

    Excavating the soil under this addition is obviously a good idea. So is adjusting the grade nearby, if possible. If you can afford to put in a real crawl space foundation (with concrete walls), all the better.

    If you keep the pier foundation, you might want to follow the advice in this article: How to Insulate a Cold Floor.

  2. FrankFulton | | #2

    Thank you. I found your article on crawl spaces vs. skirts excellent and exactly relevant. You are correct the room is on piers. Tentative plan will be 1) dig this out, 2) install bats properly between joists, 3) install 1-2" polyisio sheathing flush withe the joists, 4) cover with OSB or plywood, 5) leave the space vented.
    Unless I am convinced otherwise, I believe this approach will provide the best balance of cost and effectiveness. I don't believe there are any pipes or ducts beneath the addition, and I can't see any use for a retrofitted foundation or crawlspace. Am I missing something? Also, would spray foaming the entire underside be another option? If so, which approach would you recommend?

  3. User avater GBA Editor
    Martin Holladay | | #3

    One key element is digging out the dirt so that there is good access (and so the area stays dry in the future). Once you have done digging the area out, you have a new problem: drainage. Unless the grade around the outside of the building is properly adjusted, the crawl space might become a mosquito-breeding pond. So you have to think this through. (That's where the advantage of a traditional crawl space with concrete foundation walls starts to shine.)

    Which is better, batts plus continuous rigid foam or spray foam? The answer depends on the skill of the person doing the air sealing work. If the air sealing work is conscientious, either approach achieves good results. Sometimes, air sealing is easier with spray foam -- but if the rigid foam installer is conscientious, either approach works.

  4. FrankFulton | | #4

    Thanks Martin. I'll post more photos later today - there is a minor pre-existing drainage issue.

    Wouldn't I replace most of the dirt back from where it came, after performing the sealing?

  5. User avater GBA Editor
    Martin Holladay | | #5

    The higher the crawl space, the better. Good ventilation makes it less likely that the underside of your floor assembly is damp (assuming the wind can blow under there). If the soil is only a few inches below your floor assembly, you certainly would want to cover the dirt with polyethylene, held down by rocks or bricks.

    If you have puddles or a pond, however, you have a problem.

  6. Malcolm Taylor | | #6


    I've done what you are contemplating and it's a lot of work. To get adequate space to fasten sheet goods to the bottom of joists you need about two feet. The work is done on your back, it's very hard to get the good, tight joints needed to keep pests out, and the language used is generally unsuited for children. Think though the sequence of construction carefully. Add the perimeter blocking for the insulation layer early as a guide. Make some plywood jigs to hold one end of the sheets in place. Get a short 12 volt impact driver for fasteners - and good luck!

  7. Jon R | | #7

    You could install rigid foam on the interior side.

  8. FrankFulton | | #8

    Malcolm, Thanks for sharing your experience. Did you have an assistant? Would hiring a spray foam installer have changed results?

    Jon, Too destructive for a nice wood floor.

  9. Malcolm Taylor | | #9


    It's a tough go installing the sheet goods by yourself, that's where the jigs helps. I'm not sure spray foam has much benefit. Putting the batts in the bays and installing the rigid foam is pretty easy if you have a perimeter of 2"x material to square things up. The plywood soffit can then be nailed to it at the edges and screwed through the foam into the joists. Also a lot easier if you have marked the joist locations on the perimeter 2"x. I'd think about bedding the plywood in caulk at the edges, and caulking or taping the seams to keep carpenter ants and other pests out.

    Our code precludes untreated materials close to grade to avoid rot. The longevity of the soffit, as Martin stressed, will depend on keeping the area underneath well drained. That may be your biggest challenge. The work itself is just unpleasant. But then a lot of building tasks are.

  10. FrankFulton | | #10

    Malcolm, What benefits did you experience as a result of this project?

  11. Malcolm Taylor | | #11


    In terms of comfort the change was huge, and the seasonal movement of the hardwood floor was sharply reduced. Because the area was a fairly small part of the house, any energy savings weren't obvious.

  12. FrankFulton | | #12

    Thanks Malcolm, very helpful. This will be a "phase two" update for us, along with adding exterior foam board where we have siding (2nd floor). I'll still post a few more photos for feedback on water drainage, and perhaps Martin will then convince me to look at pouring concrete walls.

    Merry Christmas!

  13. Malcolm Taylor | | #13


  14. FrankFulton | | #14

    Martin, Malcolm, Jon R, et al

    After a week weather in the 20s: this room is the coldest room the house and very drafty, cold, and uncomfortable in the mornings. So, I'm motivated to address sooner rather than later.

    1. Above Jon R mentions installing floor insulation from the inside. Has anyone tried this approach? Lifting the wood floor seems easier than digging a foundation.
    2. Martin, the ceiling is a flat EPDM roof. Can rafter bays beneath EPDM roof be dense-packed with cellulose? Or, do the cautions you advise here still apply w/EPDM (ie, full vent or spray foam)? The current roof seems vented through the rafters, with no good exhaust, and this venting extends through the floor joists where the addition was added! There is soffit access to the rafter bays.

    Maybe we just need a great thick rug!

    Thank you.

  15. Malcolm Taylor | | #15

    Installing rigid insulation on the existing floor may well be easier. Remember though that apart from the insulation, new subfloor and flooring you also have to deal with new trim, cutting down doors and the transition in height between the old and new floors.

  16. FrankFulton | | #16

    Malcolm, thank you.

    Because this room has a low ceiling already, I was actually thinking of cut and cobble between the floor joists (perhaps with one sheet polyiso above the joists to reduce thermal bridging). Depending on the depth of the joist, this would get us to R20 or so. Given the current room for improvement, I think we are seeking a "good enough" solution, and bang for our buck. We can also reduce ceiling penetrations (8 recessed lights in a 12x12 room), and perhaps increase ceiling/attic insulation - see my query to Martin above.

    Any thoughts?

  17. Malcolm Taylor | | #17

    I'd probably start digging.

  18. User avater GBA Editor
    Martin Holladay | | #18

    If you have a flat EPDM roof, filling the rafter bays with cellulose is not an option. Here is a link to an article that will tell you what you need to know: Insulating Low-Slope Residential Roofs.

    You can cut-and-cobble between the floor joists if you want, but the result (in terms of air sealing and thermal performance) won't be as good as a job that includes a continuous layer of rigid foam on the underside of the floor joists.

    Remember: one possible answer to a "cold room" problem is simply to improve the heating system or the heating distribution system. A heating contractor can design a heating system for any room. If this room is cold, just improve the level of heat. That may mean a longer hydronic baseboard unit or a bigger forced-air duct -- depending on the type of heating system you have.

  19. FrankFulton | | #19


    To close the loop here: in the addition room, I like the "improved heating plan" best. There is only one small supply register here, at the very entrance to the addition room. Low hanging fruit. Thank you.

    Regarding the flat EPDM roof: I met with the roofer who installed the flat EPDM (as well as a section of slate) after the roof sagged under a 36" snow 10 years ago. He proceeded to tell me that the flat roof did not need to be vented (which is incorrect) and that houses like this need to breathe (oh boy)... his visit led me to discovering the "decorative soffits" and also some fungus on the ceiling rafters under the EPDM roof. The fungus was white in color on the wood, with some black in color on the aluminum soffits. There are no leaks, and I am told the EPDM will be OK for 15 more years.

    What should we do about any fungus on the rafters, eg how to inspect, assess, repair, prevent further damage?

    The entire ceiling/roof assembly is 10-12" tall, so it is also underinsulated. See attached photo. I do not want to lift the roof until absolutely necessary, when we will add external foam.

    Last detail: the EPDM roof rafters run all the way across the house, into the currently wide open floor joists that we plan to seal next week. I suppose we could leave a vent, so the EPDM roof could vent all the way to the top attic. Again, trying to get the envelope right.

    (How do these people stay in business for 30 years?? Gosh I hope he didn't also screw up the slate installation.)

    Thank you.

  20. User avater GBA Editor
    Martin Holladay | | #20

    You have described "fungus on the ceiling rafters under the EPDM roof."

    Where can you see the fungus? Are the rafters exposed on the interior of the room (meaning that there is no ceiling)?

    Or are you looking at exposed rafter tails on the exterior of the building?

    Fungus (for example, species of mushrooms) grow on damp wood. The fungus is probably harmless, but it is a sign that the wood may be punky. That raises the question: Did you assess the soundness of the wood? (I usually use an awl for this task.)

    Is the wood dense and firm? (That's a good sign.) Or is it spongy? (Bad sign.)

    If the rafters are rotten, they clearly need to be replaced.

    If the rafters are damp, a smart investigator needs to determine the source of the moisture.

  21. FrankFulton | | #21

    Thank you, Martin.

    1. I saw this fungus while looking at rafter tails on the exterior of the building, which were hidden behind the decorative aluminum soffits in a 2' overhang. (I peeled back one soffit.) I will get an investigator here to confirm my observations, and identify the source of dampness. (Perhaps it could be humid air through the leaky canned lights?)

    2. Given that the low-sloping roof will need to be vented for now, where should I airseal?
    The EPDM roof rafters run all the way across the house, into the currently wide open floor joists that we plan to seal next week. I suppose we could leave a vent, so the EPDM roof could technically vent all the way to the top attic. Or, just entirely seal the joint between the flat roof and sloped roof? Again, trying to get the envelope right. So glad I found GBA.

  22. User avater GBA Editor
    Martin Holladay | | #22

    Once again, I'm going to direct your attention to a relevant article: Insulating Low-Slope Residential Roofs.

    Please take the time to read it.

    As the article makes clear, if you have a low-slope EPDM roof with typical rafters -- you have told us that the rafters measure 10 inches to 12 inches -- with a ceiling on the underside of the rafters and roof sheathing above the rafters, you have no attic. This type of roof assembly can't be vented, because you need at least 6 inches of air between the top of your insulation and the underside of your roof sheathing if you want a vented assembly with a low-slope roof.

    The only type of roof assembly that will work for you is an unvented roof assembly.

  23. FrankFulton | | #23


    That is a very helpful clarification, thank you. I'd read the article multiple times but because of our decorative sofits had believed the low-sloped roof to be vented.

    Is there ever an advantage to creating a partition within an unvented roof assembly?

    I had planned on blocking the joist bays where the low-sloped rafters intersect with the main house (ie, extend into second floor joists - currently wide open on both sides). This would block the colder air from entering the second-story joist bays. This solution is not as elegant as adding external foam over the low-sloped roof, but seems a 10y fix until we replace the EPDM.

    I don't need a GC (yet) - am just enjoying a very steep learning curve as a first-time homeowner. GBA is an impressive community - I would be glad to engage you, Dana, or another expert if you offer remote advisory services.

  24. User avater GBA Editor
    Martin Holladay | | #24

    If the rafters of your flat roof are continuous with your home's floor joists, you definitely need airtight blocking between the rafters where the insulation ends.

    I'm guessing that you may not understand the definition of "soffit." (I apologize if I am stating the obvious.) A soffit is a horizontal board underneath a roof overhang. A soffit vent is an opening in a soffit to allow air to flow into the space above the soffit.

  25. FrankFulton | | #25

    Awesome, thank you.

    And, you are correct re my misunderstanding. We have aluminum siding w/holes (designed to look like soffit vents) covering the horizontal wooden boards (soffits). But, in spite of the vented-looking aluminum, no vents are present based on my inspection. I'm learning.

  26. FrankFulton | | #26

    Martin, Dana, et al,

    Updates on this space (see photos):

    1. These 2x8 rafter bays have no soffit - they are completely wide open under 14' length of overhang. (!) Wind can wash right between the ceiling and flat roof. Have you seen this kind of "vented flat roof" before? Anyone want to guess on the cfm?

    My plan is to seal these bays with rigid foam from the inside - we are removing the recessed lights and patching the drywall ceiling anyway. We'll also test for mold and moisture above the ceiling - I understand your counsel re: insulating a flat roof, and we are open to insulating the roof deck from beneath if needed.


    Q1. How should I inspect roof deck for moisture from the inside?

    Q2. What are risks of insulating underside dry roof deck w/SPF and cellulose, as Dana suggests elsewhere? eg, what happens if there is a water leak in the future, or if the plywood somehow is damaged?

    Q3. Why is the "foam sandwich" ill-advised? ie, If we insulated from underside now and got to R30, why not add more exterior insulation in 10y when EPDM needs to be replaced?

    Thank you.

  27. User avater
    Dana Dorsett | | #27

    In order:

    A1: A 2-pronged moisture content meter can be used to spot-check the roof deck. Ideally it would be under 15% everywhere. If it's over 20% anywhere it needs to be allowed to dry before installing closed cell foam on the underside.

    A2: As long as the R-value of the SPF is sufficient for dew point control on the amount of cellulose-R the cellulose stays dry. If the roof leaks, the foam keeps it from dripping elsewhere. At 3" close cell foam is even structural, and would behave pretty much as the structural roof deck, though peeling up and replacing any rotted plywood would still be in order when it's time to re-roof. Even 1" of closed cell foam(R6-R17, about 1-perm) is sufficient protection for the roof deck when there is 6.25" (R23-ish) cellulose under it in most climate zones, but not sufficient for dew point control. But at only 20% of the total R isn't adequate for dew point control north of US climate zone 3. But 2" SPF (R12-R14, 0.5 perms) and 5.25" (R19.5) the foam would be at least 38% of the total, and pretty safe even in zone 5, relying on the hygric buffering capacity of cellulose. A smart vapor retarder on the interior side would be cheap insurance if it's marginal.

    What's your climate zone?

    A3: A foam sandwich isn't a problem unless the wood is over 20% moisture content when the top side foam gets added. With an inch or two of closed cell foam and no polyethylene on the interior side the roof deck can still dry reasonably (if slowly) toward the interior.

  28. FrankFulton | | #28

    We're CZ4, near Baltimore, MD. Of course, we'd prefer to put off the expense. But if our air sealing is placing the roof at risk for moisture problems, we're willing to bring down the ceilings and spray 2" of cc foam and cellulose (or just 4" of cc foam?).
    Please confirm: if there is rotten plywood when we reroof, the foam will be "lost?" What to do at that point, as we won't bring down the ceilings again (but would consider pricing the addition of external foam to increase R value)?
    Thank you.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |