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
Q&A Spotlight

Does This Roof Need More Foam?

Spray foam contractors recommend one thing, the building inspector says another. Who's right?

How much is enough? A spray polyurethane foam contractor is recommending less insulation than is required by code. That's a problem. Photo: mlwoodchuck / CC BY-NC-ND / Flickr

A GBA reader named 88Clayton is insulating the attic in his house with spray foam and has run into a problem: the level of insulation recommended by contractors is at odds with what the building inspector wants.

“I’m on the southern edge of [Climate Zone] 4,” 88Clayton writes in a  Q&A post. “My spray foam contractor recommended R-20 of open cell. That’s what is currently sprayed on my roof deck. The code inspector came by today and is insisting on R-38, as if it [were] a conventional vented attic with loose blown fibers.”

With an air leakage rate of 1.7 ach50, the house is tight when compared with typical construction. Even so, the R-20 recommendation 88Clayton got from all four foam contractors is well below code required insulation levels for his climate zone (R-49 in the International Residential Code) and less than what he has in the walls of his house—R-25.

“It’s a bit strange to have walls that are better insulated than [the] roof,” he says. “Where is the realistic point of diminishing returns? I know the encapsulation itself has benefits that go beyond just raw R value, even though the inspector doesn’t quite get this, but perhaps I do need more foam thickness before the diminishing returns kick in.”

That’s where we begin this Q&A Spotlight.

It’s closed-cell foam you really want

When foam is sprayed directly on the underside of the roof deck, it should be closed-cell, not open-cell, writes Zephyr7. “Open-cell has moisture issues here!” he writes.

Open-cell foam is an effective air barrier, but it’s vapor permeable. Closed-cell foam is both an air barrier and a vapor barrier, so moisture is not able to migrate through the foam and collect in the roof sheathing.

“I’ve seen spray foam contractors say R-20 is enough because ‘spray foam is so awesome’ or ‘it seals really well,’ ” Zephyr7 continues. “R-20 is not code minimum R-38. You don’t get a pass because you used spray foam.”

Zephyr7 isn’t the only GBA reader who points out that R-20 isn’t enough insulation in Climate Zone 4, but 88Clayton insists that open-cell foam is commonly used in attics where he lives. “Almost all the homes are getting open-cell,” he says.

And building scientist Joseph Lstiburek of Building Science Corp. agrees, says Jon R. He references an article on the topic, noting that Lstiburek finds open-cell foam in the attic is fine with one big caveat—the attic must be a conditioned space so moisture can be removed from the air.

The less is plenty argument

Yes, the attic could probably use a little more insulation, 88Clayton admits. But, he adds, “I’m willing to bet R-20 of open-cell on a conditioned attic roof deck is a better performer than the typical R-38 house with loose blown on vented attic floor.”

In fact, he adds, typical houses are getting building department approval with between 5 inches and 7 inches of foam, with “little if any” air sealing of penetrations and gaps on the attic floor.

That, says Russell Miller, is the “snake oil” that many spray-foam installers are trying to sell their customers. David Pilati writes that the pitch for less-than-code insulation from foam installers is “typical and tired.”

The argument that R-20 of spray foam is better than R-38 of blown in insulation also attracted the attention of Peter Engle.

“This is a false equivalent, and an old argument by the spray foam industry that has been thoroughly debunked,” Engle writes. “It counts on old testing done on ‘typical’ fiberglass installations from years ago. Those ‘typical’ installations didn’t have any air seals and did not meet current code, even at that time. They wouldn’t meet code now, as the code requires air sealing as well as insulation.

“Your argument essentially argues that your house with spray foam insulation that doesn’t meet code is better than an assumed ‘typical’ fiberglass house that doesn’t meet code,” Engle continues. “Since neither one meets code, there’s not much point in arguing, is there?”

Bottom line? The building inspector is right, he argues. The roof does not meet code and needs more insulation.

What about HVAC ducts in the attic?

88Clayton is having trouble accepting the fact that a conditioned attic insulated to R-38 will perform the same as a code-compliant R-38 vented attic insulated with loose fibers.

“In terms of the actual system as a whole in the real world,” he write, “I find it hard to believe these are equal. If for any reason, houses like mine have second floor ducts in the attic.”

Once mechanical systems are installed in the attic, says Dana Dorsett, it’s time to insulate the roof deck.

“Installing the mechanicals in the attic forces one to take the more expensive step of insulating at the roof deck,” he says. “It’s often cheaper to build a horizontal utility chase below the attic floor for the ducts, insulating the attic floor above the chase, than it is to insulate at the roof deck. But that’s an option that has to be decided as part of the framing design.”

Making R-38 work in the attic

Dorsett adds a few thoughts on how 88Clayton might approach building code requirements for more insulation. Although R-49 is the code requirement in Climate Zone 4, R-38 might be acceptable if whole-house performance is considered.

“R-38 of open-cell-foam-only will fit in 2×12 rafters,” Dorsett says, “but it would need an interior side vapor retarder to control moisture cycling, and not a true vapor barrier. To do it without the interior side vapor retarder, a minimum of 15% of the total R would have to be closed-cell foam at the roof deck with open-cell under that.

Dorsett suggests that 2 inches of closed-cell foam (a type that uses a hydrofluoro-olefin blowing agent to lower the global warming potential) with 6 1/2 inches of open-cell foam would fit in a 2×10 rafter bay.

Another option is 2 inches of HFO-blown foam with R-23 mineral wool batts—a combination that would fit in a 2×8 rafter bay.

Why installers recommend so little foam

If the code requires R-49, or even R-38, why do foam installers suggest R-20 worth of open-cell foam?

One reason, Dorsett says, is that R-20 is about the most that can be installed in a single pass of open-cell foam with a density of 1/2-pound per cubic foot. If installers put in more without waiting for the first lift to cure, they run the risk that the foam will catch on fire. Plus, there are issues with char, shrinkage, cracks, and separation between the foam and the framing.

Another reason is that in areas where energy costs are low, R-20 is the most that can be installed with a “reasonable” payback when considering the net present value of future energy costs.

“Foam guys will give the usual BS about how much more airtight it is, but a competent blown-fiber contractor spends as much or more time on the air sealing details than blowing insulation,” Dorsett says.

“It’s easier to make a rationale for R-49 cellulose at $1.50- $2 per square foot than R-49 of open-cell foam at $4- $5.50 per square foot,” Dorsett says. But the fact is that the current requirement in the International Residential Code is R-49, and 88Clayton’s attic isn’t even halfway there.

“To get compliance on a U-factor basis it has to be under U-0.026, which is R38.5 ‘whole assembly,'” he adds, “factoring in the thermal bridging of the framing, but also adding in the R-values of the roof deck, roofing, interior and exterior air films.”

In most houses, Dorsett says, that could be achieved with 6 to 7 inches  of continuous rigid polyiso foam board above the structural roof deck. That would be followed by a 5/8-inch plywood or OSB layer screwed to the structural deck with pancake-head timber screws.

“For less money,” he adds, “if the house has 2×6 rafters one could get there with R21-R23 batts between the rafters and 2-4 inches of polyiso up top, with plenty of dew point control at the roof deck.”

Our expert adds this

Closing thoughts come from Peter Yost, GBA’s technical director:

To be clear, the IECC 2015 Table R402.1.2 requires R-49 of insulation in ceiling and roof assemblies in Climate Zone 4. It says nothing about what type of insulation will achieve this performance value. We don’t know exactly what code applies in this Q&A, but I think it is safe to say that the local code for Clayton88 is in this ballpark and R-20 is just too far away for any code pathway compliance to offer a solution. Clayton88 needs more R value in his roof assembly.

There are some pretty big differences between open-cell and closed-cell spray foam (see this article for more information on getting spray foam insulation right). Here are some things to keep in mind:

  • The expansion rate for open-cell foam is about 100:1; it’s about 30:1 for closed-cell. This means that it is difficult to spray in tight spaces and shallow cavities without a lot of wasteful overfill, that must then be shaved. It also means that if the lift is too deep, you can get large pockets of air called “rollover.” I have seen cavities as large as a softball in a 2×12 framing cavity.
  • All polyurethane spray foams involve an exothermic reaction, but I have never seen enough heat generated in open-cell to create the problems that can be experienced with closed-cell. This is related to density (and probably to water content of the two). Here is what the SPFA has to say about exothermic reactions in spray foam (from Spray Polyurethane Foam Alliance FAQ #9): “The chemical reaction used to create polyurethane and polyurethane foams give off heat, which is called an exothermic reaction. Closed-cell SPF products, when installed at pass thicknesses exceeding manufacturer’s recommendations, can generate excessive heat that can be trapped inside the foam. This heat can results in poorly formed foams that dramatically reduce coverage rates and diminish product performance, causing loss of R-value or shrinkage. In extreme cases, where closed-cell foam is applied at thicknesses several times the manufacturer’s limits, can generate enough heat to self-ignite the foam. The SPF contractor should always follow manufacturer’s installation instructions regarding pass thickness and times between passes to eliminate damage caused by exothermic reactions.”

It’s interesting to note that there’s a bit of a void in the building code when it comes to guidance on open-cell spray foam in roof assemblies in Climate Zone 4. Here is what the 2015 IRC says about unvented attic and unvented enclosed rafter assemblies:

“R806.5 Unvented attic and unvented enclosed rafter assemblies.

“In Climate Zones 5, 6, 7 and 8, any air-impermeable insulation shall be a Class II vapor retarder, or shall have a Class II vapor retarder coating or covering in direct contact with the underside of the insulation.”

And here is what the Joe Lstiburek article referenced in this Q&A says about “conditioning” the attic space in a cathedralized roof/attic.

Footnote 2:….”5.1.5. In climate zones 1, 2, and 3 air shall be supplied at a flow rate greater than or equal to 50 cfm per 1000 ftof ceiling.” Lstiburek, at the time, was working on this code change based on BSC/Building America research that has since been adopted in the 2018 IRC in the section on vapor diffusion ports, but only for CZ 1–3 to date.

So, do we need a Class II vapor retarder and dedicated attic space ventilation in CZ4? My cut: Neither the code nor Joe Lstiburek states that you should not include these in unvented roof assemblies with open-cell spray foam and both make sense from a building science perspective. I think Clayon88 should deepen his spray foam application to achieve the code minimum, apply a Class II vapor retarder, and include dedicated ventilation to the attic at the rate of 50 cfm per 1000 sf of floor area.

-Scott Gibson is a contributing writer at Green Building Advisor and Fine Homebuilding magazine.


  1. John Clark | | #1

    It's disappointing that after 15 yrs of being in the marketplace SPF continues to suffer from application related problems and continued cost pressures from less expensive competing products.

  2. Nola Sweats | | #2

    Regarding the supply of conditioned air at 50cfm per 1000 sq ft to the attic, that seems to be for a central air duct, which operates only when the a/c or heat is running. Given that a/c and heat don't run much here in the southern end of Zone 2A in the late fall, winter and early spring, it seems like continuous venting from the conditioned space into the attic would be preferable. Is there a reason why less-frequent-but-drier air directly from a duct would be preferable?

    Alternatively, does it make sense to use a small fan (20-40 cfm), perhaps humidity/temperature controlled, to vent attic air to the outside and suck in conditioned air from the living space when the humidity or temp gets too high in the attic?

    1. User avater
      Peter Engle | | #5

      Humidity issues tend to be worst in winter and summer. In winter, warm/humid air inside the house can condense against cold(er) sheathing and other exterior surfaces, causing mold and rot. In summer, the very hot and humid outside air can condense if it reaches cold(er) interior surfaces chilled by the A/C system. In spring and fall, most US locations have relatively moderate humidity, and the lack of large differences between indoor and outdoor temperatures tends to limit the risk of condensation within building cavities. So, conditioning the attic when the air handler operates is a pretty good approximation of dehumidifying during only the highest-risk times of year.

      An exhaust-only solution can work if the house is not super tight and not already depressurized by other exhaust fans. There is an additional cost in energy to condition the air that leaks in to replace the exhausted air. If going this route, exhaust air from the very top of the attic space.

    2. User avater
      Jon R | | #7

      There are various dehumidification options in 5.1.5. I'm sure they aren't exactly equivalent in effectiveness or energy use.

  3. Antonio Oliver | | #3

    Just want to point out that the use of the phrase "the code" implies one uniform code across localities. This is not the case as the version of the IECC in effect across the country varies from locality to locality.

    1. John Clark | | #4

      At GBA "the code" usually means current IECC. Now whether states and localities wish to adopt the current code is of course up to them.

      1. User avater
        Peter Engle | | #6

        I somewhat disagree. "The Code" for most of us is whatever version of the IRC has been adopted locally. That governs all aspects of residential construction, including energy efficiency. The IECC is a broader code that governs both residential and commercial construction. The residential aspects of the IECC are embedded in the IRC, and the IRC folks work hard to synchronize the two. The IECC is sometimes a bit more flexible if designers with to use it, and the IRC allows for that as well. Still, the IRC is the top level "code" for most of us.

        Of course, local exceptions and delays in adoption make everything more difficult. In NJ, we are on a sort-of 5 year cycle of code adoptions, which is totally out of synch with the I-code people's 3-year cycle. And we take exception to lots of good work from the I-code people because, of course, our politicians know far more than them.

        1. John Clark | | #9

          Well ya, but I was thinking within the sphere of GBA articles that current IRC/IECC code is almost always part of the context absent more detailed information. Especially in a Q&A scenario.

        2. User avater GBA Editor
          Brian Pontolilo | | #10

          When I reference the code here, I'm referencing the current 2018 International Residential Code (IRC), which as Peter pointed out is the current home building code book from the International Code Council and includes the most recent International Energy Conservation Codes where applicable to residential construction. I believe that Martin has always leaned on the most recent IRC for his reference too.

          To Peter's point, this is a challenge for home building journalists. When we published content in FHB, from an architect or builder in one area, the work may or may not have agreed with the local codes in any given reader's jurisdiction. We got a lot of mail about that.

          Readers were often appalled with what they saw in the magazine. If we published a horizontal deck railing, which codes in many areas do not allow, we'd be sure to get some feedback from upset readers. I can remember the subject line of one email that read, "Railings of Death."

          We were always happy to get and publish reader feedback on safety issues, but the fact was often that the architects and builders were building within the law of the building codes that applied to their projects.

          1. Malcolm Taylor | | #11

            Railings of Death have just been approved in our latest code revisions. Canny investors in BC are lining up to buy shares in funeral homes.

    2. Robert Hronek | | #14

      You are talking about the roof decking on the other side of the insulation. So if the outside air is in the 90s the roof deck is probably 140°. What about the thermal bridging?

      There is a product called aero barrier that could be used to get a similar air seal and then use dense pack to get the r value and eliminate the bridging.

      Spray foam is so expensive that the benefits are oversold and substandard r values used. Then there is the fact the every job is a science experiment. Mix part a and part b at the proper ratios at the proper temps and apply at the proper thickness. I have smelled spray foam that has had a lingering order.

  4. User avater
    Jeremiah Gammond | | #8

    Here in Northern Ontario, the R-value varies depending on whether it's a roof with attic space (R-60) or without attic space (nom. R-31)

  5. Patrick Stuart | | #12

    Looking at this from another angle: does it make monetary sense to spray foam all cavities? For the Ohio weatherization program, we’ve adopted the option of using rigid insulation (polystyrene) for spaces between joists, with spray foam only used as filler for leftover gaps. It’s used primarily in bandboard and basement/crawlspace applications (attics typically get blown with cellulose), but the reduction in expensive spray foam usually makes up for the increased labor cost. Just curious if anyone does this in attics.

    As an aside, in my past architectural life I was intimately familiar with railings of death. Next to politics, religion and OSU/Michigan rivalries, there were few other topics that caused such gnashing of teeth and rending of garments.

  6. User avater
    Greg Labbe | | #13

    I wonder what this foam will look like in a year once it's covered in drywall... I bet the foam will de-laminate as time marches on and wood works it's magic.

    Did anyone do a blower door test to see if and where this foam leaks air before it gets covered?

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |