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

Attic insulation question

Rob Moore | Posted in GBA Pro Help on

I have read many of the posts and articles about different ways to bring an unconditioned attic into the thermal envelope of one’s home and I’m currently weighing some options. I will likely use spray foam under the rafters of my attic (9.5 inches). I’ve read the various debates about using less spray foam than would be required to achieve the desired R-value (eg R of 49 in my zone 4a area), with the argument (from the spray foam manufacturers it seems) being that spray foam yields a more effective R-value so you don’t need as much as you would think based solely on R-value calculations. This has been countered by experts here who say that is just not correct.

I’ve been discussing this with an insulation contractor whose company uses spray foam as well as other forms of insulation and he has laid out various options. My inclination is to go with 3-4 inches of closed cell under the rafters and then supplement with fiberglass for the remainder of the depth (as Martin has shown in a recent FHB article), however, he is presenting the argument that there would be very limited benefit to adding that extra fiberglass insulation even though on paper it adds an R-value of say 20 to the 21-28 that the closed cell would provide.

Although I’ve read the back and forth about the “effective” R-value of spray foam, and I understand that R49 is better than R28, I haven’t actually seen any data that compares these two situations: 4 inches of closed cell vs 4 inches of closed cell plus fiberglass on top of that. I know this topic of attic insulation has been talked about a lot, but was hoping someone could point me to data that will help me decide what to do. Thanks.

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. Expert Member
    Dana Dorsett | | #1

    In zone 4A you would not need more than R15 out of a total R49 to be closed cell foam in that stackup to be able to use a class-III vapor retarder (standard latex paint on wallboard) at the interior side air-barrier for the fiber layer.

    So at most you would need 3" of close cell foam and R30 fiberglass below that, but the fiberglass needs an interior side air barrier to perform at R30, as well as a Class-II or Class-III vapor retarder.

    There's nothing magic about closed cell foam that would give a higher performance at a lower R value. I don't know what sort of "data" you are looking for. It sounds like you're looking for an excuse to stay at 4" / R28, and skip meeting code minimums?

    The R7/inch is an exaggerated number to begin with. Even if they were able to foam a sample in a lab that hit R7/inch, the fully depleted R value is more like R5/inch after the HFC245fa leaks out, which will happen within the lifecycle of a house.

    Closed cell polyurethane is one of the most expensive commonly used types of insulation, at about 15-18 cents per R per square foot it adds up. Installing 4" of EPS above the roof deck (R17 at about 10 cents per R foot) and R30 fiber under the roof deck would likely be cheaper than 4" of closed cell foam under the roof deck, and will surely outperform it. That would only be R47 at center cavity, but would meet code on a U-factor basis since there would be no rafters thermally bridging the exterior EPS. Unlike HFC blown foams, the R-value of EPS is stable over time. EPS is blown with pentane, which escapes very quickly, and it's R-value tested at an already fully-depleted condition.

    HFC245fa is a powerful greenhouse gas (~1000x CO2 @ 100 years), and is likely to be banned by international agreements soon. (Phasing out HFCs for all commercial uses, including use air conditioning refrigerants was a topic of discussion during Indian PM Modi's recent visit to the US.) At least one very low impact HFO1234yf blowing agent is commercially available in the US, but so far only LaPolla is using it. Demilec has announced they will begin phasing over to HFO1234yf "soon", but haven't given a date certain. Pretty much everyone else with an R6/inch or higher 2lb foam is using HFC245fa.

    Open cell foam air seals just as well as closed cell, and an all open cell R49 solution with an interior side Class-II vapor retarder works too. For 0.7lb open cell foam that takes about 12-13", for half pound foam 13-14". At 10-13 cents/R-foot it may be easier than monkeying around with closed cell + fiber solutions, and may even be cheaper.

    Like EPS, the R-value of open cell foam doesn't change over time, since it's blowing agent (=water) dissipates quickly and is not present in the test samples. A foot of half pound foam has the same amount of polyurethane as 3" of 2lb foam, and has more than 1.5x the thermal performance- it's even lower impact than EPS. From a green-greener-greenest point of view, it's the greenest foam insulation there is (except for reclaimed/reused foam from building demolition.)

  2. Rob Moore | | #2

    Thanks Dana, I appreciate the response. I have a few comments:

    I'm not looking for an excuse to use less insulation, I'm actually trying to find the data to counter the insulation contractor's claim that 4 inches of closed cell is sufficient. I was going to follow one of Martin Holladay's proposal in FHB of doing closed cell plus batt insulation to achieve close to R49.

    We have 10 inch rafters so unless we sister all of them, which will add expense, we have 9.5 inches to work with. I'm not wedded to closed cell (I've read the discussions about risks of open cell in the attic and think it is readily solved by moisture control by ventilating the space if needed). But it allows me to get to a sufficient R value for the foam part and still add more via fiberglass.

    Although I realize insulating above the roof with rigid foam is a good way to go, it requires replacing the entire roof and then redoing all the fascia, etc, a very expensive undertaking for a retrofit. Maybe it would be less costly than closed on the underside for a new construction, but the roof alone will cost over $30,000 to replace (it's about 5000 sq ft), and that doesn't count all the exterior trim work to accommodate the thicker roof, or the labor costs of installing all that rigid foam.

    Do you have a sense of how much extra benefit would be accrued by adding fiberglass to 3" of closed cell under the rafters? Also, if I did 3" of closed and then fiberglass why do I need a vapor retarder? I thought that was only needed with open cell in zone 5 or colder?

  3. GBA Editor
    Martin Holladay | | #3

    In your climate zone (Zone 4A), if you want to combine closed-cell spray foam with fluffy insulation, then a minimum of 31% of the total R-value of your insulation should come from the spray foam layer.

    You have 9.5 inches to play with. Here's what you can do:
    2 inches of closed-cell spray foam = about R-12.5
    7.5 inches of fluffy insulation = R-27.75
    R-value of both layers = R-40.25
    Ratio of spray foam R-value to total R-value: 31%

    That stack-up would work.

    R-value is R-value, no matter where it comes from (spray foam or cellulose or fiberglass).

    If you went with only 3 inches of spray foam, and nothing else, you'd have about R-18.75.
    If you went with only 4 inches of spray foam and nothing else, you'd have R-25.

    The assembly I'm suggesting is R-40.25. If you choose that assembly, the rate of heat flow through your roof will be only half what it would be if you stopped at 3 inches of spray foam. It's good to double the R-value. Doubling the R-value cuts the rate of heat flow in half.

  4. Rob Moore | | #4

    Thanks Martin. Some additional questions for you:

    1) The insulation contractor said something to the effect that I would need a minimum of 2" of closed cell to ensure that the dew point doesn't fall on the sheathing. But if R-value is R-value, and if I add fluffy under that closed cell then the dew point issue should be taken care of, correct?

    2) Where does the 31% come from? What principle is at play here?

    3) I understand the view that R-value is R-value and that doubling it will halve the heat flow out of the roof. But how can I counter the claim about "effective R-value" of foam vs fiberglass? I understand that it's my house, and I can tell the insulation contractor that I want 2 or 3" of foam plus the rest in fiberglass, but the scientist in me would like to be able to offer up some data to convince him, other than pointing him to the posts here.

    4) If I went with closed plus fluffy, but don't plan to cover with drywall (would be very pricey for 3,000 sq ft of roof surface) what else do I need? Do I need some sort of fire/combustion barrier? If so, will certain types of fluffy insulation work (eg rock wool or other)?

    5) Leaving aside for a moment the risks of open cell under the roof, what about 8" of open cell (R 28-30)? I'm guessing you would say it has overall less R value than 2-3" closed plus fluffy, and likely not less expensive, so why bother.


  5. GBA Editor
    Martin Holladay | | #5

    To understand the general principles behind building assemblies that combine a layer of foam insulation (either rigid foam or spray foam) facing the exterior with a layer of fluffy insulation facing the interior, see these three articles:

    Calculating the Minimum Thickness of Rigid Foam Sheathing

    How to Build an Insulated Cathedral Ceiling

    Combining Exterior Rigid Foam With Fluffy Insulation

    The 31% figure is a climate-specific figure that applies to roofs, not walls. You can find that 31% figure in one of the tables I created for the last of the three articles listed above (Combining Exterior Rigid Foam With Fluffy Insulation). While the article talks about rigid foam rather than spray foam, the same principles and ratios apply to assemblies with spray foam on the exterior side of the fluffy insulation.

  6. GBA Editor
    Martin Holladay | | #6

    You wrote, "The insulation contractor said something to the effect that I would need a minimum of 2 inches of closed cell to ensure that the dew point doesn't fall on the sheathing."

    On this point, your insulation contractor and I (mostly) agree. You'll note that my proposed assembly also included 2 inches of closed-cell spray foam. My only quibble with your contractor would be that the surface where you need to be concerned about condensation is not the sheathing; its the interior side of the cured spray foam. That's the "first condensing surface" -- the surface that has to be warm enough in winter to avoid condensation problems. If there is no fluffy insulation -- only spray foam -- you could get away with anything from 1/2 inch of spray foam to 1 foot of spray foam, because the interior side of the cured spray foam is always at room temperature. The only thing that cools down that surface is the fluffy insulation.

    You wrote, "But if R-value is R-value, and if I add fluffy under that closed cell then the dew point issue should be taken care of, correct?"

    I don't know what you mean by "taken care of." But if you really want to install just 2 inches of closed-cell spray foam with an R-value of R-12.5, and stop right there, you could. You've have a very low R-value, but you wouldn't have any moisture problems. The only thing that makes the 2 inches of spray foam risky (from a moisture perspective) is adding more than R-28 of fluffy insulation on the interior side of the spray foam. The thicker the fluffy insulation, the riskier the assembly -- because the fluffy insulation makes the cured spray foam cold.

    For more information on the idea the R-value is R-value, whether the R-value comes from cellulose or spray foam, see these three articles:

    Understanding R-Value

    Air Leakage Degrades the Thermal Performance of Walls

    It’s OK to Skimp On Insulation, Icynene Says

  7. Rob Moore | | #7

    Martin, thanks for the explanations and references. I will look over the material in those links.

    Any thoughts on my question of what else one would need in addition to the closed cell and fiberglass if not installing drywall in the attic?

  8. GBA Editor
    Martin Holladay | | #8

    When installed in an attic, the spray foam has to be protected on the interior side of the spray foam with either a thermal barrrier (usually 1/2-inch drywall) or an ignition barrier. If you hope to avoid installing 1/2-inch drywall, you will not be permitted to use your attic for storage.

    If your local code official agrees that your attic has only limited access, and will not be used for storage, you may be allowed to use a less stringent barrier to protect the spray foam: namely an ignition barrier. Blown-in fiberglass, blown-in cellulose, or mineral wool can be used as an ignition barrier, but not fiberglass batts (to the best of my knowledge).

    More information can be found in this article: Thermal Barriers and Ignition Barriers for Spray Foam.

  9. D Dorsett | | #9

    If installing fiber that's anywhere near the 29% - 71% split in total R the fiber also needs an interior side air barrier that is also a class-III or Class-II vapor retarder. If the fiber is left exposed to the air convection moves quite a bit of moisture to condense at the surface of the foam, If you're pushing the ratio limit, you would be in danger of collecting too much moisture there by vapor diffusion alone unless the interior surface is at least semi vapor retardent. (Standard latex primer on gypsum board is fine- it doesn't have to be pretty, just air tight.)

    BTW: " ensure that the dew point doesn't fall on the sheathing..." is a nonsensical statement. A dew point is a temperature, not a location within the stackup. The relevant temperature in this context is the dew point of the conditioned space air in winter, which in your location is between 40-45F. The sheathing will of course dwell at temperatures well below that temp, but 2" of closed cell foam is a class-II vapor retarder, and will limit the amount of vapor diffusion from the interior air to the roof deck to limit it's peak moisture content to acceptable low-risk levels. But that closed cell foam won't keep moisture from condensing on the foam itself if the fiber layer is too thick, lowering the average temp at the interior surface of the foam below the conditioned space air's dew point.

    Open cell foam is also an air barrier. With 2" of closed cell and the rest being open cell there is very little danger of moisture accumulation in the open cell foam, but to limit that possibility (there are reports of high moisture cycling in summer in open cell insulated attics), an interior side vapor retarder is still a good idea. If it's open cell foam the vapor retarder layer doesn't necessarily need to be ultra-air tight, since there is no convection mechanism to move moisture through the open cell foam, only vapor diffusion.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |