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

Spray foam against underside of roof deck? Or install vent baffles?

dfvellone | Posted in General Questions on

my location is zone 6.
my roof design is a cape: 5′ of rafters get insulated with sprayfoam from the plate to the intersection of the collar ties which form the “drop” ceiling and are also insulated.
My plan had been to install site built baffles in order to provide a 2″ airspace between the foam and underside of the roof deck. Rafter tails are exposed – no fascia and soffit, and the vents originate where the siding meet the roof deck and the vents dump into the  short attic space. The ridge would be vented.
I’ve just had a foam installer out who is encouraging me to spray the foam against the underside of the roof deck directly omitting the vents and baffles.
I could go this route and instead provide gable vents.
My initial preference for the vented rafters arose out of my apprehension to seal off the bottom of the roof deck making it impossible to ever detect any roof leaks until damage occurs. With the 2″ space I can sight down the channels. I’ve included a sketch of the roof.
Am I wasting energy installing the baffles and vents and would I be further ahead just installing the foam against the roof deck?
Thanks, Daniel

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

    You have forgotten to provide a few details. We don't know:

    1. If you are talking about open-cell spray foam or closed-cell spray foam.

    2. The intended thickness of the spray foam that will be installed.

    3. The depth of your rafters.

    In general, your approach is better (since your approach provides better ventilation of the third-floor attic, and since your approach makes it easier to perform future roof sheathing repairs). The main disadvantage of your approach is that providing a 2-inch-deep vent channel reduces the room available for insulation -- and in most cases, that means that your roof R-value will be less than optimum.

  2. dfvellone | | #2

    2lb closed cell foam sprayed to 8" depth.
    I generalized regarding the vent. My rafters are a full 10" deep so the vent channel is acutally 1 3/4" - I'm using 1/4" luan for the baffle material. The entire length of the vent channel is 65".
    And to further clarify, the structure is built: roof deck is on as well as roofing material, so I'm well beyond the point of considering applying insulation to the top of the roof deck.
    Thanks, Daniel

    1. GBA Editor
      Martin Holladay | | #3

      Thanks for filling in the missing details. My answer is unchanged: installing the baffles will provide a better installation than spraying directly against the underside of the roof sheathing.

      As I'm sure you know, 8 inches of closed-cell spray foam is neither cheap nor particularly green -- but in your circumstances, the spray foam may make sense.

      1. dfvellone | | #4

        Thanks for the input Martin.

    2. Expert Member
      Akos | | #5

      8" of close cell spf in between real 2x10, you end up with only R29 roof. The thermal bridging of the rafters conduct a lot of heat, short circuiting the insulation value of the spray foam.

      You are way better of to stick to your initial vented plan go with open cell foam. This give you an R22 roof, but would cost significantly less. The cost increase and R value difference will never pay in energy savings.

      If you can go down to 1" vent space, you get an R2 with open cell for very small extra installation charge.

  3. MattJF | | #6

    Are you open to furring down the rafters with strips of foam on the inside. This will increase depth of the rafters and significantly reduce heat loss through framing.

    8" of closed cell foam is a LOT. Adding 1" thick polyiso strips and using open cell foam, rockwool or high density batts in the bays will be much more effective both thermally and cost wise.

    Do you need to build to code or as just as good as you can?

    1. dfvellone | | #7

      Would foam furring make that much difference? I could still apply it since I've only installed the finished ceiling on the flat, but isn't it the thermal bridging from outside that compromises the integrity of the foam?

      As far as my seemingly having blown the benefit of foam at all (between the inevitable failing seal between the rafters and the foam, and the thermal bridging) I suppose that any foam invariably is a waste and for that matter a strong argument in favor of fiberglass batts. The current house I live in and constructed in exactly the same manner has fiberglass batts throughout the roof/ceiling assembly and though I did use polyethylene against the inside of the rafters and collar ties I heat the house with a small fraction of wood most others do with similar sized homes and never get ice buildup at all on the roof or eaves.

      So, I'll reword my original question: given the limitations of current construction and code requirements aside, is foam worth it at all in the slopes, or should I just stick with batts?

      1. Expert Member
        Akos | | #9

        Foam is worth it for air sealing, this is especially true for typical construction where the ceiling/floor joist are used as ties for the rafters, so they poke through your air tight layer. Trying to seal this intersection up without spray foam is next to impossible.

        A bit of closed cell is worth it when you can't vent the roof. Otherwise stick to open cell or batts.

        With older houses, there are a lot of times issues with non standard distance between rafters. In this case, to go with batts is a lot of labour, unless you are DIY it would not be worth it.

        For overall R value, in zone 6, going much above R25 full assembly is not really worth it in energy savings but can sometimes reduce the size of the HVAC load enough to be able to downsize. This might pay be enough to pay for a bit of extra insulation.

        Heat doesn't care which way the thermal bridging is reduced, works just as well on the inside as the outside.

        Provided you can seal up your floor joists, a budget higher performance DIY roof assembly might be:
        -vent baffle
        -R28 fluffy
        -2" polyiso with taped seams over the rafters
        -drywall installed directly on the foam with long screws.

        This would give you roughly an R35 roof.

        For the attic area, just blow in as much insulation as you can. Extra R value there is almost free.

  4. Jon_R | | #8

    > isn't it the thermal bridging from outside that compromises...

    Thermal bridging from either side is a negative. Inside (low R wallboard) is slightly more harmful than outside (higher R OSB or plywood + something covering it) - because the exposed wallboard conducts more heat to the thermal bridge (the joist). Like a cooling fin.

    > any foam invariably is a waste

    I expect that the air impermeability and edge sealing of open cell spray foam (vs fiberglass) has a significant and durable advantage. And don't heed any warnings about open cell risks that are based on un-vented, poorly conditioned or no-vapor-barrier assemblies. Anyone have data?

    1. Deleted | | #12


  5. MattJF | | #10

    Closed cell only makes sense if you need to control moisture in an unvented roof deck. You won't be able to put foam against the deck only in the sloped areas, it's all or nothing with venting.

    What is your plan above the flat ceiling? Do you have ducting up there?

    If you don't, I would plan to keep the entire roof vented. Air seal the ceiling drywall and pile R60 cellulose on the flat ceiling. Air seal your 1.5" or 2" baffles and have them extend well past the fluffy insulation. I believe there may be some benefit to running them all the way to the ridge and using them as a second air barrier. You may use rigid foam for your baffles, but you need to use unfaced 1" or 1.5" foam with enough permeability to not get wet or use thick foam making up 50% of your R value (I think that is right for zone 6).

    Here is an example of foam furring:

    This article highlights the effects of thermal loss through framing:

    One option: Fur the rafters out 1.5", 2" baffle spacing, 4" reclaimed or seconds polyiso R24-26, R23 Rockwool.

    If you want to keep the existing sloped ceiling height, 8" of polyiso foam is cheaper and more environmentally friendly.

    These approaches are probably only viable as a DIY due to labor involved.

    1. dfvellone | | #11

      There is no ducting in the attic and with baffled and vented rafters a continuos ridge vent will be included. The flat is getting cut and fill polyiso and skimmed with 2" of closed cell.

    2. Jon_R | | #13

      > it's all or nothing with venting.

      You are free to vent some areas of a roof and not others. In some cases, it may require under shingle vents.

  6. MattJF | | #14

    Do you have reclaimed or seconds polyiso? It is hard to justify using rigid foam where blown insulation can work so well. Air sealing a ceiling is pretty straight forward. You can use open cell foam to seal and then fill with fluffy stuff.

    Move the cut and cobble polyiso to the slopes.

    1. dfvellone | | #15

      The slopes comprise 300 sq' of ceiling area while the flat is 731. I'm installing cut and fill polyiso in the flats because it's the larger space as well as easier to fill: slopes have the obstacle of rafter tails every other bay from the rafter extensions. I'm not interested in loose fill. I need to access the space and I've seen too much rodent damage even in all borate cellulose.
      My concern regarding open cell is its water absorbing quality. Should the roof spring a future leak it would be a small disaster.

      1. MattJF | | #16

        Cut and cobble is a lot of work. I have no interest in doing it 8" deep. You will either need to leave half inch gaps around the foam and can foam the edges or fit it tight and learn how accurate your framing was. Spraying small deep gaps 6-8" deep afterward with commercial spray foam will likely leave a bunch of voids.

        It's your house and money, you do what you want. If you are committed to ccSPF, please use a HFO blown foam. Just know it is not generally justified in your application from an environmental, cost, or building science perspective. I don't know if you have pricing, but 8" of ccSPF over 300SF is $2400-$3600. Know too that they can't actually put 8" in as the stuff is terrible to shave down and they will put 7-7.5".

        Roof leaks are not typically large volumes of water and open cell dries okay from a single event. It would just soak a small area before dripping though, not absorb through the entire ceiling. I am not sure I would want something really waterproof in the ceiling without a drainage path, I have this attic pool image in my head, but again leaks are usually small volumes.

  7. dfvellone | | #17

    My original question having been derailed, I'd like to clarify some confusing advice.

    The integrity of the adhesion of the foam is so poor and the thermal bridging of the 2 x 10 rafters so severe that 8" of closed cell foam will only yield r29, but I should instead use open cell foam (along with 1" foam furring strips on the inside edge of the rafters) which, setting aside the above mentioned issues with closed cell foam in the same scenario, at best will offer me somewhere in the neighborhood of r30?

    And lastly, "A bit of closed cell is worth it when you can't vent the roof. Otherwise stick to open cell or batts."
    The conclusion from this seems to be that the ventilation of the roof will allow the heat that escapes from open cell and batts to escape, and closed cell doesn't require ventilation... because it seals better? Its r value works better?
    Maybe I'm missing something and it wouldn't be the first time.

    One last thing, I've had 6 different insulation contractors out here to give me quotes and not a one of them even sprays open cell foam. Now, as a lifelong professional builder I'll be the first to admit that construction trends don't necessarily follow the best practices when it comes to the environment or even intelligent building, but these have all been reputable installers.

    1. Expert Member
      Akos | | #18

      " integrity of the adhesion of the foam is so poor "

      I don't know where this came from, spray foam sticks like crazy to most building materials.

      "thermal bridging of the 2 x 10 rafters so severe "

      That is correct. This is true no matter what insulation you put between the studs. This is why cut and cobble is a HUGE waste of time (the high R value of the rigid insulation is short circuited by the wood).

      "ventilation of the roof will allow the heat that escapes "

      Ventilation is not about allowing heat to escape (it helps a bit with that also) but allowing moisture to escape. With enough closed cell spray foam under the deck or enough rigid above the deck, interior moist air can't condense, if the there is no condensation, you don't need drying, thus no need to vent.

      The amount of closed cell spray foam or rigid insulation is somewhere between 20% to 50% of the R value of the overall cavity insulation depending on your climate zone. This percentage is the "bit of closed cell" that is worth it.

      Open cell foam is still vapour permeable, there is a risk of condensation, thus you need to vent.

      Overall I would follow these for a roof:

      -if you can air seal the assembly and can vent it, use batts
      -if you can't/hard to air seal but CAN vent go with open cell
      -if you can't air seal can can't vent go with closed cell+batts

      Avoid more than 3" of cc SPF between dimensional lumber, the extra bit of SPF does little for insulation, just costs more to install.

      Additionally a bit of continuous rigid insulation on either the inside or the outside will significantly increase the overall R value, in colder climates it is worth it.

    2. GBA Editor
      Martin Holladay | | #19

      Q. "The integrity of the adhesion of the foam is so poor and the thermal bridging of the 2 x 10 rafters so severe that 8 inches of closed cell foam will only yield R-29."

      A. Who said that the integrity of the adhesion of the closed-cell spray foam is poor? I was under the impression that the spray foam hasn't even been installed yet.

      You might want to read this article: "Installing Closed-Cell Spray Foam Between Studs is a Waste."

      Eight inches of softwood lumber (the rafters) have an R-value of about R-10, which is less than the R-value of 8 inches of closed-cell spray foam (about R-52), but which is not nothing. I think that the whole-roof R-value of this assembly is closer to R-32 than R-29.

    3. Expert Member
      BILL WICHERS | | #20

      >"The integrity of the adhesion of the foam is so poor and the thermal bridging of the 2 x 10 rafters so severe that 8" of closed cell foam will only yield r29,"

      Spray foam sticks like crazy, it's like spray adhesive. Spray foam even sticks to wet surfaces, the moisture just helps cure the spray foam. There is some truth behind the thinking that closed cell spray foam helps make a structure more rigid -- it's durable, and adheres VERY strongly to wood surfaces. You pretty much have to destroy a roof assembly that has been insulated with spray foam if you want to remove any of the sheathing, for example.

      By simple averages and 24" rafter spacing, 8" of closed cell spray foam and 2x10 rafters you get about R33 for the assembly, about R32 with 16" rafter spacing. The thermal bridging does make a big difference in the performance of the overall assembly.

      >"The conclusion from this seems to be that the ventilation of the roof will allow the heat that escapes from open cell and batts to escape"

      No, escaping heat is energy loss, but it won't damage your roof. What you need to be concerned with is MOISTURE migrating through the open cell foam or batts and then condensing on the cold underside of the roof sheathing. This is where mold and rot come from. Closed cell foam acts as both an air and vapor barrier so there is no moisture migration, and if the foam is thick enough, the inside surface of the foam will be warm enough that any indoor moisture won't condense.

      Open cell foam has also had some issues holding on to moisture over time and causing other kinds of problems when used in roof assemblies.


  8. dfvellone | | #21

    I had assumed that the estimated r29 value was a combination of thermal bridging and the effect of expansion/contraction of the lumber that would compromise the seal. I didn't expect thermal bridging to be the sole factor.
    No, no insulation has been installed yet. I was referencing the ideas that I'd read here and making my own mistaken conclusions. I told you it wouldn't be the first time.

    I can air seal the assembly as well as vent it - I prefer to vent it and this had been my plan from the start - and I'm firmly set on using foam in this location (my walls are all getting dense pack cellulose). Am I no further ahead from a performance standpoint (moisture and insulation value) using closed cell? Open cell will perform as well for me in this same situation? Would it pay to use baffle material that would act as a vapor barrier when using open cell or should the open cell be able to dry both ways?
    Thanks and especially for your patience.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |