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

Is six inches of open-cell spray foam enough for Climate Zone 3A attic?

user-2310254 | Posted in Energy Efficiency and Durability on

I am (mildly) renovating a circa 2006 three-level townhouse and trying to be prudent with the budget. On Friday the HERS rater completed a blower door test, and we met with an HVAC/Energy Performance contractor we have both worked with on a previous house. So far we have a general plan for air sealing the home’s more significant air leaks.

I am less comfortable with our initial discussion on how to improve the townhouse’s attic insulation and would value the GBA community’s input.

The HVAC/Energy Performance contractor is suggesting converting the attic to a semi-conditioned space and installing six inches of open cell foam. (The conditioning will be done by a new five-speed heat pump that also has a dehumidification feature. Service to the third floor is provided by ducts in the attic. A second system services the first and second floors.)

I questioned the amount of insulation since that will be far short of current recommendations for Zone 3A. But both the rater and contractor advised that installing R-38 would not be cost-effective and anything beyond six inches would offer diminishing returns.


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

    Q. "But both the rater and contractor advised that installing R-38 would not be cost-effective and anything beyond six inches would offer diminishing returns. Thoughts?"

    A. Here at GBA, we always advise homeowners and builders to install at least as much insulation as is minimally required by building codes. Unfortunately, spray foam contractors have been giving homeowners bad advice -- advice which is often frankly illegal -- for years.

    For more information, see "It’s OK to Skimp On Insulation, Icynene Says."

  2. user-2310254 | | #2

    Thanks, Martin.

    I am glad you included the link so I could reread the Icynene article. After posting my question, I started to do a slow boil about how the contractor handled my insulation question. Similarly, I am disappointed that he has already decided that my project needs four tons of cooling (700 square feet per ton) even before we have completed the Manual J.

  3. Expert Member
    ARMANDO COBO | | #3

    I am on the side to condition an attic by insulating with 2" R13 closed cell foam against the roof decking and 7.5" R26 open cell foam minimum, total R39 assembly. I also design ventilation of 50 cfm per 1,000 sf of attic space.

    1. user-2310254 | | #4

      Thanks, Armando.

      I had a flash and fill approach in mind as well. The contractor also did not see any value in spraying the top plate, which I found a little puzzling.

  4. Expert Member
    Dana Dorsett | | #5

    Let me get the straight:

    The "...HVAC/Energy Performance contractor..." is trying to sell you on 700 square feet per ton with out benefit of a load calculation, R22- ish roof deck insulation instead of the code minimum R38 , and doesn't see the value in insulating the top plate?

    Are you still talking to them? This reads like hackery, not Energy Performance or HVAC design.

    A ratio of 700 square feet per ton would be more typical of an insulated 2x4 framed house with single pane windows.

    Most homes come in with ratios north of 1000' per ton, with many in the 1400' per ton range. With a 5-speed heat pump you MIGHT be able to get away with oversizing by 2x without a large efficiency & comfort hit, but it's hardly optimal.

    It may be true that more than 6" of open cell foam may not be "cost effective" on short to mid-term energy savings, but that's partly a function of being a premium priced product. On full lifecycle energy savings it's all but certainly going to be cost effective even at the high cost per R of open cell foam. (That is, unless energy costs take a precipitous dive and keep going down over time.) Get it quoted at the code-minimum R38 as well as R22 (or whatever the 6" was supposed to be.)

  5. user-2310254 | | #6

    Hi Dana,

    Yes. I was surprised by his recommendation and am scheduling other vendors to take a look at the project. I have already had a room-by-room Manual J completed and (if I am reading it correctly) the current infiltration rate is 5.9/ACH50. This is the pre-airseal rate, so things should improve significantly after that work is done.

    I intend to follow Armando's advice and install an adequate layer of foam in the attic. If R-22 was "good enough," I would leave things as-is.

    The HERS rater and I are leaning toward using an UltraAire ventilator/dehumidifier in the attic instead of relying on the HVAC for dehumidification. I had a similar setup in my last home, and it did a decent job of introducing fresh air while controlling humidity.

  6. user-2310254 | | #7

    I just spoke with another foaming contractor, and I am starting to worry that my project may go off the rails. Like the other guy, he wanted to sell me R-20 (but will spray whatever I want, of course).

    He also did not want to spray closed cell and raised the issue of trapped moisture rotting the roof sheathing. My gut tells me this warning is (again) a reflection of how foam jobs are priced around here, but my gut has let me down on other occasions.

    He also said his company would need to install house wrap in front of the adjoining firewalls because the code does not allow foam to be applied directly to these surfaces. (The walls appear to be GP DenseGlass.) I was thinking of just having him blow in cellulose.

    As for the roof, I was hoping to get around the ignition barrier requirement by not using combustion appliances or storing anything in that space. If pressed for an ignition barrier, I was considering intumescent paint.

    These contractor interview are making wonder what is going to happen when Georgia eventually updated its code. If R-20 foam "delivers 90% of the benefit of R-38" what will they say when gov'ment requires more insulation?

  7. Expert Member
    Dana Dorsett | | #8

    If the moisture content of the roof deck is low when the closed cell foam goes on, it stays low. Checking it with 2-pronged moisture content meter in several places is a good idea (days or weeks before the foam crew arrives.)

    If/when the roof leaks it's going to be an issue for that section roof deck regardless whether there is closed cell instead of open cell foam on the interior. Bulk water leaks won't dry toward the interior through open cell foam quickly enough to matter.

    For the firewalls, cellulose is often cheaper than open cell foam, but not always. Rock wool (blown, or even batts if installed well) is by far the best material for fire walls, since it doesn't burn or melt at relevant temperatures found in a house fire. The inspectors would likely see that as an upgrade.

    Be sure the code inspectors are on-board with the intumescent paint approach before taking that route, or you'll end up paying for both the paint AND the thermal barrier assembly.

    Insulation contractors who speak in terms of percentages drive me nuts, since the paradigm starting point is what, R-ZERO, open air? A tent with just interior & exterior air-films with effectively no insulation value in the fabric? An empty framing bay?

    It's the latter, and here's his math:

    An empty framing bay with plywood on one side and half inch gypsum on the other is good for about R4 (U0.25)at center cavity when adding up all the air films & layers. Installing R20 brings that up to about R22 (U0.045), at center cavity if it's a full cavity fill. Installing R38 brings it up to about R40 (U0.025).

    So R20 compared to an empty bay reduces the heat transfer rate to about 18% of the empty stud bay (an 82% reduction) R38 reduces it to about 10% of the empty stud bay (a 90% reduction). So 82% is 90% of the 90% you get with R38 but so what? Where is it written that an empty stud bay is a good starting point, and that an 82% reduction from that is a "reasonable" end point?

    Why isn't R13 in a framing cavity a better starting point, since most homes insulated their attics to about R13 or more even before 1960, and nobody in their right mind would install less than that since the first energy price shocks of the 1970s.

    R13 delivers about R15 (U0.067) at center cavity. That would make the R22 (U0.045) delivered by using R20 a 33% reduction from the starting point, and the R40(U0.025) a 63% reduction. 33 is only 52% of the 63, a much lower percentage improvement than the alleged "...90% of the benefit of R-38."

    Stating it in terms of "...percentage of the benefit..." is a numbers game, and deceptive unless the starting point is stated. Zero insulation for a conditioned attic was really never an option (was it?), but that's the starting point necessary to make the contractor's statement a near-truth approximation of fact.

    He may or may not have calculated the math for himself, but it's probably just being passed around as insulation-contractor lore without thinking about it too deeply. Since the maximum amount of half-pound foam that can be sprayed safely at high quality in a single pass is about R20, that "90% of the benefit of R38 " soon became the foam-guy mantra. But without context it's crap.

    Radiant barrier trolls have a habit of applying even crazier numbers in the numbers game than foam guys, declaring the ~97% reflectivity of bright aluminum means radiant barrier delivers "...a 97% reduction in heat transfer..." ( as if we all lived in an unenclosed vacuum, without air or walls to effect that heat transfer. ) I once read a guy on a building forum bragging that his house was now "...97% efficient..." after having radiant barrier installed under his siding and in his attic. (I hate to think about how much they charged him for that "service"!)

    1. user-2310254 | | #11


      I love reading your answers. Thanks for taking the time to do the analysis.

      I agree some of these vendors sound like they are delivering talking points from a brownbag session. I've tried telling the foam guys in particular that I have a different agenda, and I am not interested in creating an underperforming attic.

      I do need to ask some questions about pulling permits, however.

  8. GBA Editor
    Martin Holladay | | #9

    Open-cell spray foam is the type of spray foam that is most associated with damp roof sheathing -- not closed-cell spray foam. So the contractor you spoke with has it backwards.

    More information here: "High Humidity in Unvented Conditioned Attics."

    1. user-2310254 | | #12

      Thanks, Martin.

      Honestly, I don't think any of these guys are thinking about vapor, solar drive, excessive humidity, or any those issues. They just figure that roofs leak, which they do. And you want to know when you have a leak.

      In a weird way, Dana's observation that a leak is a localized issue is quite comforting.

      On the plus side, the electrician I met with today really seemed to have a handle on his part of the project. I guess installing a subpanel is much easier than spraying foam.

  9. GBA Editor
    Martin Holladay | | #10

    Dana's analysis of the "R-20 foam delivers 90% of the benefit of R-38" nonsense is a good analysis.

    Here's another way to look at it: R-38 foam cuts heat flow in half compared to R-19 foam. Or, if you prefer: the rate of heat flow through R-19 foam is twice the rate of heat flow through R-38 foam. That's why you want R-38 insulation.

  10. Expert Member
    ARMANDO COBO | | #13

    I second Martin's math. All you have to do is perform a dew-point analysis (Wuffi or ASHRAE Fundamentals) to see the difference. R-values are linear, not a percentage of percentages. The contractors you are talking to are no different than the guys here in Dallas-Fort Worth area, CZ3A. The biggest majority are ignorant about building science, and a few that know a bit, try to "get" the job with lies and BS. I add that ATL is close to CZ4, which means that every so often, your winters will be like a CZ4, so be prepared for it.
    To me, you have two choices: 1. the #1 priority is to do it right, so you must follow R806.5 & Table N1102.1.2 or 2. the #1 priority is to save money and and make a risk management decision to accept the possibilities of future problems.

    1. Expert Member
      Dana Dorsett | | #24

      I find Martin's math better too- I was just mapping out the tortured path it took to come up with the foam-guy's percentage numbers.

      It's not rocket science, but it takes at least the 5th grade math to figure out the source of the percentage numbers. I suppose it's a good thing that very few rocket scientists are wasting their careers installing insulation, but I sincerely hope most insulation contractors made it past the 5th grade.

  11. user-2310254 | | #14

    A quick update. I spoke with my third foam contractor today, and he also could not see the value of going beyond six inches of open-cell foam--or air sealing the leaks between the attic and the rest of the home. He advised me that I needed the air leakage to avoid having a moisture issue. He said there was also no reason to seal the top plates since foam at the roof line would eliminate the wintertime stack effect.

    This was a particularly disappointing conversation since the firm's website includes a lot of detail on how they followed best practices for air sealing and insulating homes.

    Maybe it is time to rethink this project. Perhaps I should just air seal the attic, elevate the new HVAC equipment, and install R-60 of new cellulose. It would likely be cheaper as well.

  12. Expert Member
    ARMANDO COBO | | #15

    Steve - What do you mean by "elevate the new HVAC equipment"? Are you talking about creating a ventilated attic with the HVAC system in the attic?

    1. user-2310254 | | #16


      I am starting to worry about the competency of the foam installers--especially since I am asking them to do something that is clearly outside their comfort zone. The last "technician" I spoke with did not know the names of any of the products his firm uses or what a blowing agent was.

      I don't want a ventilated attic, but I also don't want a bad foam job. If I used cellulose, the new attic HVAC system would need to be suspended or at least placed on a platform to allow a full depth of insulation.

      Please let me know your thoughts.

  13. GBA Editor
    Martin Holladay | | #17

    If you want an unvented conditioned attic, you don't have to use spray foam for insulation. Other approaches are possible.

    1. If you have a simple roof configuration without valleys, hips, skylights, and dormers, you can install vent baffles under the roof sheathing and then insulate your rafter bays with fiberglass, mineral wool, or cellulose.

    2. Since you are in Climate Zone 3, you can consider creating an unvented attic equipped with vapor diffusion ports -- an approach newly permitted by the 2018 IRC. If you do this, you can insulate with fiberglass, mineral wool, or cellulose. For more information on the new code, see Image #2 on this GBA page: "Burying Ducts in Attic Insulation."

  14. Expert Member
    ARMANDO COBO | | #18

    If you still want to do a foamed attic, I would think that you could ask Carl Seville or the Southface Energy Institute in ATL some recommendations on a good foam installer in your area; I'm sure they know of at least one guy.
    Martin's #1 option is a good option, with lots of labor. #2 option, I don't agree with unless you install the air handler inside the conditioned space, which means to create a closet on the third floor, install an upflow handler so you can encapsulate or double-wrap (R8+R6) the ducts, or to encapsulate the air handler, you would have to build a attic room just for the air handler. I did that once.
    Joe Lstiburek has come up with another solution... Creating a vapor diffusion vent at the ridge and with no soffit or lower vents. See: Read all the way to the bottom, Figure 5.

  15. GBA Editor
    Martin Holladay | | #19

    You advised, "Joe Lstiburek has come up with another solution... Creating a vapor diffusion vent at the ridge and with no soffit or lower vents." That's exactly the same advice I gave as Option #2 in my comment (Comment #17).

    1. Expert Member
      ARMANDO COBO | | #21

      I was thinking of burying ducts (per your link) and not on Joe's method to use a vapor diffusion vents, which the way I understand, is still an unvented assembly that happens to have a vapor diffusion and a supply ventilation port. Joe gave a presentation here in DFW a month a go on this topic, among others. This is a rare time where I've disagreed with Joe (with all due respect) for two reasons, based on issues we had in the Southwest when we installed dense pack cellulose or batts (held by net or wire) against the roof decking over years past. 1. With time, gravity worked its magic and created an 1"-2" void between the insulation and the bottom of the decking, thus diminishing the insulation value. 2. We found way too many deficiencies in the installation of dense packed insulation between the truss webs. Its about the same reason why I specify to spray open cell foam on the rim joist of a truss system, because is almost impossible to install any other form of insulation there. For the record, I expressed my thoughts to Joe during his presentation, but his explanation didn't change my mind over our past experiences. I guess I agree to disagree.
      I don't have issues with double-wrapping dusts in a ventilated attic, as we've now done it a few times, however, I don't know of a way to double-wrap or encapsulate an air handlers in a ventilated attic, and probably not allow by manufacturers or code.

      1. user-2310254 | | #22


        So you think the netted cellulose will drop down and render the insulation useless. What about using batt insulation, adhesive-applied fiberglass, or open-cell in place of cellulose (assuming I could convince someone to spray 12 inches of foam)? Or do you think the vapor diffusion approach is still too risky?

        1. Expert Member
          ARMANDO COBO | | #23

          Steve - Let me start by saying that I'm not a researcher, nor I have the experience and knowledge that Joe and BSC have, my conclusion is based only on our work in the SW and a pretty good gut feeling.
          I probably would not have a problem using rigid foam or CC foam AND netted dense pack insulation against the roof decking on a roof assembly using dimensional lumber rafters or TJIs, and I may even try the vapor diffusion and a supply ventilation port with rafters or TJIs roof assembly, as I think one could do a good job of protecting against gravity. Having said that, I would never install any permeable insulation against the roof decking with a truss assembly other than open cell foam, obviously with rigid foam on top of the roof decking.

  16. user-2310254 | | #20

    Thank you, Martin and Armando.

    I wanted to take another look at the attic before replying. It has a truss framing system and is essentially a gable roof with a couple of ornamental dormers on the front.

    Lstibureks article indicates it is possible to net the roof and pump in cellulose. This approach appeals to me the most out of the various options we've discussed. Let me see if I can get an insulation contractor on-board.

  17. user-2310254 | | #25

    I would like additional advice on insulating my new (old) townhouse. Budget is getting to be an issue, so I am looking for possible savings.

    My original plan was to demo the most of the garage drywall and have the ceiling and interior wall spray foamed. Removing and replacing the drywall is pricier than I expected. So I am considering a Plan B, which would involve the following steps:

    - Leave existing fiberglass batts in place (They have largely dropped away from the subfloor.)
    - Blow cellulose into the garage ceiling (all web joists)
    - Block the area where the conditioned and semi-conditioned spaces meet (the garage is getting a heat pump water heater)
    - Foam the blocking to air seal as much as possible
    - Foam or caulk remaining penetrations such as around breaker box

    Will this work? Any other suggestions. The HERS rater will do another blower door test after the insulation is installed to look for problem areas.

  18. Expert Member
    Dana Dorsett | | #26

    How deep are the garage joists? And what spacing?

    From a moisture management point of view it's fine to fill the joist bays above the garage. If they're 16" TJIs it could be quite a lot of cellulose though. The fastener spacing on the ceiling gypsum may need to be tightened up, even if you don't dense-pack it.

  19. Expert Member
    ARMANDO COBO | | #27

    Steve - The way I treat floors above a garage is applying 2" R13 CC sprayed foam under the floor decking and then install the rest with 5" R18 OC foam, if floor trusses, or cavity fill dense pack cellulose or BIBs, if using TJIs or dimentional lumber. I know the code in CZ3 requires R19, however, I always call for R30 total. The reason I spec 2" R13 CC sprayed foam either way, is to avoid moisture, chemicals and CO2.
    For budget, I would still apply the 2" R13 CC sprayed foam and reuse your batts, if installed correctly. I do have to say that I've never seen batt insulation installed correctly in a web truss system, but since you are doing the work, you may have a bigger incentive to do it right.

  20. user-2310254 | | #28

    Thanks you both for the quick replies.

    Dana. The townhouse has floor trusses (probably 16 inches although I haven't measured them). So yeah, that would be a lot of shredded paper.

    Armando. I was hoping to avoid having to demo the garage drywall. It is not a lot of area, but the insulation contractor wants a pretty steep bump to install new material. My spend is seriously cutting into my wife's furniture budget.

  21. Expert Member
    Dana Dorsett | | #29

    16" of 2-2.5lb (not dense packed) cellulose weighs 3.3lb per square foot, and delivers R55-R60. Blown fiberglass would weigh half that, but wouldn't be as air retardent, nor would it buffer the summertime moisture when the subfloor temp in an air conditioned room drops below the outdoor dew point. If the ceiling is air tight and painted with vapor barrier latex the cooling season dew point issue isn't likely to come up.

    Armando's solution is very robust, but requires demoing the gypsum. An inch of closed cell would theoretically be enough, but most installers aren't adept at installing layers that thin.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |