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

How to Insulate the Attic in a 1910 Remodel

A homeowner contemplating an energy overhaul of his North Dakota home weighs the options for insulating the attic

An attic in need of a helping hand: The attic of Tim Lange's North Dakota home offers several energy-related challenges, including inadequate insulation and framing that doesn't allow enough insulation between the rafters.
Image Credit: Tim Lange

Tim Lange is taking on a major renovation of his 1910 home in North Dakota that will include a new roof, exterior spray-foam insulation, and new doors and windows. His quandary is what to do in the attic.

“I think I’ve got a good handle on the exterior insulation process — using window bucks to create an ‘outie’ style window is the current plan,” Lange writes in a Q&A post at GBA. “The third floor and attic are where I need some help.”

In the attic, Lange is dealing with three distinct zones: the area behind the kneewalls, the sloped ceiling in the living space, and the flat ceiling in the living space. He assumes the rafters are 2x6s on 2-foot centers. Judging from the photos he provides, none of it appears adequately insulated.

Lange is considering a number of possible approaches. “For the area behind the kneewall,” he writes, “one option would be to remove the fiberglass, add a layer of spray foam against the roof sheathing and finish with dense-packed cellulose to get the desired R-value,” he writes.

“I don’t see a way to effectively insulate the sloped section of the ceiling,” he continues. “There is just not enough room and it can’t be accessed easily. I suppose we could remove this part of the ceiling, fill the cavity with spray foam, then Sheetrock back over the foam. UGH. One suggestion was to dense-pack the area behind the kneewall and in the sloped ceiling area and pile lots of loose-fill in the main attic area. Not code-compliant, I know. But they do it out East.”

He could gut the third-floor attic to remove all the lathe and plaster, then spray-foam the underside of the sheathing to R-49 — that’s what’s recommended for his Climate Zone 7 locale — or spray 1 inch of foam and fill the rest of the cavity with dense-packed cellulose. Finally, he could ventilate the underside of the sheathing with a 1-inch air gap using rigid foam.

There are plenty of possibilities. What’s his best plan of attack? That’s the topic for this Q&A Spotlight.

Foam on top of the roof sheathing is the best option

“By far the best solution is to install one or more layers of rigid foam above the roof sheathing, followed by an additional layer of OSB or plywood and new roofing,” GBA senior editor Martin Holladay tells Lange.

Spraying 1 inch of foam on the bottom side of the roof sheathing and then using cellulose won’t work, Holladay says, because the minimum R-value for the foam layer in this type of construction is R-30. To get that kind of performance, Lane would need to install 5 inches of closed-cell foam, followed by enough cellulose to get to R-49.

Of the three possible approaches Lange has outlined, Holladay likes the third — installing ventilation baffles — assuming Lange can manage to install soffit vents and a ridge vent.

Addressing the moisture problem

Using 2 inches of closed-cell foam on the underside of the roof deck would protect the sheathing, writes Dana Dorsett, but Lange would still need a smart vapor retarder on the interior or some foam on the exterior if he planned to augment the foam with cellulose.

“In Zone 7, a minimum of 60% of the total center-cavity R-value needs to be low-perm foam in that type of stackup,” Dorsett says. “The 2 inches of foam (R-14) + 3.5 inches of cellulose (R-13) adds up to about R-27. For minimal dew point control at the foam/fiber boundary it would take another R-5.5 on the exterior, but to get to a code-minimum R-49 it would take R-22, or 4 inches of rigid polyiso.

“The polyiso would underperform its labeled R-value in winter, but it would also be thermally breaking the rafters, so the performance loss isn’t too terrible,” Dorsett continues. “For minimal dew point control, a minimum of 2 inches exterior + 2 inches of closed-cell on the underside would be good insurance.”

Holladay says that using 2 inches of closed-cell foam instead of 5 inches would violate a section of the 2012 International Residential Code dealing with unvented attic assemblies.

“It’s not technically a violation and will be pretty safe as long as there is also exterior foam. It’s more of a gray area,” Dorsett replies. “The 2 inches of closed-cell foam is a Class-II vapor retarder, which is sufficient to protect the roof deck from interior moisture drives, and with a couple of inches of rigid foam (of any type) above the roof deck, there would still be dew point control at the first condensing surface, which is the interior side of the spray foam.”

Although this method would not be the best option in new construction, Dorsett continues, it’s about “as good as it gets in a retrofit with 5 1/2 inches of rafter depth.”

Dorsett points to an article published by the Building Science Corporation suggesting that his plan would work.

“This does not provide control at the foam/fiber boundary without interior vapor retarders, though, but adding sufficient exterior foam does,” he says. “Putting a bit more R-value on the exterior would also raise the average temperature at the roof deck, further lowering its (already vapor-retardancy-limited) wintertime moisture accumulation.”

Limiting the amount of interior spray foam to 2 inches gives the roof deck some capacity to dry toward the interior, Dorsett says.

Will foam insulation mask roof leaks?

One thing that’s been nagging Lange is that using spray foam on the underside of the roof deck would make it difficult to spot roof leaks.

“Will leaking shingles eventually allow water to drip through the spray foam and onto the ceiling below — giving notification that the shingles had a leak?,” he asks. “I’m concerned that the spray foam will hide a bigger issue and instead of replacing shingles, would then need to do shingles, sheathing and possible rafter fixes.”

The same concern would apply if Lange were to add rigid foam insulation on top of the roof deck, followed by a new layer of sheathing. “If the shingles are leaking,” he asks, “how does the owner notice this?”

That, says Holladay, is the risk he runs with that type of assembly. “If you want early detection of roof leaks, you want a house with a vented unconditioned attic,” he says. “If you decide to insulate along your roof slope, to bring your attic inside your home’s thermal envelope, you lose the ability to detect roof leaks early.

“It’s your choice — take your pick.”

Insulating the attic from the inside alone

Screwing through 10 inches of foam placed on top of the roof deck and hitting framing members accurately seems a little dicey, so Lange has come up with an alternative that calls for a combination of spray foam and rigid insulation that can be installed entirely below the roof deck.

After gutting the interior, he would create a ventilation channel 1 inch deep below the roof deck with rigid foam, and fill the rest of the cavity with closed-cell spray polyurethane, for a total R-value of 27. Next he would install 4 inches of continuous extruded polystyrene (XPS) applied below the rafters in the cavity behind the kneewalls and on the sloped section of ceiling on the other side of the kneewall.

In the center section of the attic ceiling, where Lange doesn’t want to give up any headroom, he would install 4 inches of foam board on top of the horizontal ceiling joists. To keep the R-value of the various sections of the roof assembly fairly uniform, Lange would add 3 1/2 inches of spray foam in the spaces between joists in this center section. Using foam board instead of blown-in cellulose would retain a narrow crawl space for inspecting the roof.

But, after talking with his spray-foam installer, Lange has decided to use baffles in the rafter bays from the soffit to the open space above the ceiling joists — an easier ventilation option than making baffles on site with 1-inch XPS board.

Lange also thinks that introducing a thermal break on the underside of the rafters is worth considering. That would be in the form a furring strips, spaced out from the rafters by 1 inch, and the gap filled with spray foam. It would cost him $1,275 for this 1-inch-thick thermal break, in addition to the 4 inches of spray polyurethane foam already planned.

“It’s hard to calculate the return on investment for this added insulation,” he adds, “My gut feel is do it — we have come this far. I’ll be installing the furring strips, so the labor is basically free.”

Even with the added 1 inch of foam, the total thickness would be 5 1/2 inches, or R-35.75 at R-6.5 per inch. That’s less than the R-49 required by code, but the plan apparently has the blessing of the local building inspector.

Our expert’s opinion

Here’s what GBA technical director Peter Yost has to say:

There certainly are a lot of moving parts in this roof/attic insulation discussion, so I finally decided to make a quick drawing — crude as it may be — of the “final” configuration (see below).

Here’s the color key (marker color selection is no more meaningful than what I could find on or near my desk):

1. Yellow = spray foam insulation, presumably all closed-cell.

2. Orange = existing sprayed urea formaldehyde (SPUF) wall insulation.

3. Pink = rigid insulation board.

And I wanted to focus on these issues:

Venting/ventilating the roof and attic. I like where this landed finally, with soffit-to-ridge ventilation on the underside of the roof deck. In general, the difference between venting and ventilating is the latter conveys actual air movement (soffit to ridge, driven by stack effect convection) with the former only supporting gable-to-gable diffusive drying potential. (I know that gable-to-gable air movement can be increased by wind forces, but in my experience this happens precious little, especially in truncated attics like this one).

This roof/attic assembly has good-to-excellent drying potential for the roof framing and sheathing to the exterior with this ventilation of the underside of the structural roof assembly.

Air barrier continuity. This was not really covered in the Q&A exchange, but again, in the final configuration there is good continuity throughout the roof/attic assembly. Note that particular attention is needed to connect the spray foam air barrier at the spots on my drawing marked by stars.

One way to deal with the middle starred point is to redo the kneewall framing after the rigid insulation is in place, screwing the angled top plate of the kneewall through the foam into the rafters. This completes the thermal break of the rigid foam on the sloped portions of the assembly as well.

Thermal bridging. This final configuration has this covered well, too, even if it flip-flops or jumps at the sloped/flat ceiling transition.

I don’t have the sloped ceiling furring strips and subsequent added depth of spray foam insulation as a thermal break in my sketch because I did not understand its need, given the continuous rigid insulation on the underside of the sloped ceiling rafters.

Drying potential of flat ceiling. Depending on the rigid insulation chosen to be on top of the ceiling overlays, keep the interior sheathing and finish vapor-open to accommodate drying to the interior of this portion of the attic/roof assembly.

Caution on the depth of closed-cell spray foam lifts. How deep closed-cell spray foam can be installed per pass is dependent on the spray foam manufacturer and its particular recommendations or requirements. The depths in this roof/attic assembly are just deep enough for this to be a concern, almost regardless of the manufacturer or product. I hope this goes without saying, but always use a certified company and installer (complying with ABAA or SFPA certification programs) for any spray foam product (since this type of insulation is site-manufactured).

Sprayed urea formaldehyde foam (SPUF). This is really nasty stuff. In my remodeling days in the Northeast, I never encountered SPUF that was not seriously deteriorated in composition and volume, with its insulation value (and even more important, its airtightness) also trashed. I’m not sure how much of its original R-value you can count on to augment the exterior spray foam in the exterior walls. But I am sure that no one looks forward to having to gut this building’s exterior walls in addition to the second story…


  1. TIM LANGE | | #1

    Project Update
    Nice write up Scott; illustrates the complexities of the project and the trade-offs. The gutting of plaster and lathe and insulation is almost done. The roof boards are in very good shape, a few areas show moisture presence likely due to condensation.

    The hand-drawing showed a layer of rigid foam on the underside of the sloped section. This was considered but current plan is to use the foam as furring strips only...

    For the thermal break on the underside of the rafters of the sloped section - planning to use strips of 1" thick XPS, 1.5" wide. Going to choose an XPS with high compressive strength (60 psi) since drywall will be attached through these strips and don't want the rigid foam to compress. The SPF will lap 1/2" over these 1" thick rigid foam furring strips.

    Calculated R for sloped roof assembly: Without any thermal break on rafters, the assembly has 4” of SPF (using R6.5 per inch), the overall R for 16” oc is R20.7. Adding 1” foam furring strips to bottom of rafters and increasing to 5” of SPF increases to R27.8. Increasing to 1.5” furring foam strips: R31.21. The framing is 24" oc, not 16" oc. Will soon have to decide how much headroom to give up in order to get more R.

    Interesting idea on the kneewall removal and re-install after thermal break installed. Too much effort.

    Added a picture showing the settling and shrinkage of the '70s foam.

  2. Bo Jespersen | | #2

    Great points
    Thanks as usual for a good article. We run into this in Maine all the time and I applaud the owner being able to do this right. My post is geared towards other ideas on this issue without removing the interior finish.

    I ask because typically, a client cannot afford to remove the interior finish in a 2nd floor cape & we are left trying to insulate as is. This usually means working around batting/loose insulation that is already in the slope that cannot be easily removed and sometimes eliminating all ventilation. Side attics are usually exposed like this one and are more easily upgraded with spray foam and are either vented (preferable) or not. The real trick, and my greatest bane, lies in how to handle those damn finished slopes.

    We have been dense packing these for ages with cellulose with apparently good results but based on BSC evidence and the one job I investigated, this is risky. I was able to pull apart a nicely dense packed slope once in winter and was heartbroken to see it was full of frost 3" interior from the sheathing. We surmised that the cavity had absorbed moisture during summer and did not have time to dry out before freezing temps set in.

    Here are some options I am bouncing around and could use feedback on for finished slopes with no venting:

    1.) If we have to insulate a slope with no venting, we start using blown rock wool or the new Spider fiberglass insulation instead of cellulose. This will still choke off venting, but will not be hygroscopic like cellulose and should therefore dry faster if it didn't absorb summer time humidty. This is my theory.

    2.) We insulate side attics, but unless we can remove interior finish as the owner did above, we don't insulate finished slopes at all.

    3.) We have looked into pourable foam which seems like a good idea, but with the potential of old insulation still in the cavitiy, I have concerns of its effectivness.

    4.) That's all I've got!

    This has kept me up nights so I could use some fresh advice:)

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

    Response to Bo Jespersen
    Thanks for sharing the anecdote about the 3 inches of frost on the exterior side of the cellulose-insulated slope. Anecdotes like yours reinforce my opinion that you don't want to use an air-permeable insulation -- even Spider insulation or mineral wool insulation -- in an unvented roof cavity.

  4. George Hawirko | | #4

    How to Insulate the Attic in a 1910 Remodel
    Why Insulate an Old Wood Building it will only be a Total Wast of Time & Money.

  5. Bo Jespersen | | #5

    Best to leave it alone then?
    Thanks for the quick response, Martin.

    3 Questions;

    1.) If you cannot strip the finished slope, would you then leave the cavity as is (empty or poorly insulated) rather than take the risk?

    2.) Have you seen or heard of pour-in foam used in these cases?

    3.) Do you see any potentially ill effects of insulating the side attics with spray foam directly to the roof deck but then leaving the finished slopes untouched? I don't see a hazard here, but would love another opinion.


  6. Kohta Ueno | | #6

    Injected Polyurethane Foam (IPF) Method
    As a point of information, a *good* spray foam contractor might be able to insulate existing closed cavities using closed cell polyurethane foam and an injection method, through holes drilled from the interior. I note "good," because this is a somewhat finicky method--the installer is working "blind," and needs to do things like "timed shots" relative to the dimensions of the cavity in order to avoid underfill problems. Overfilling, of course, runs the risk of breaking or bulging interior or exterior finishes. Also, I believe a slow-rise formula is required; an infrared camera is often used as quality-control tool during installation.

    The main reason I know about this is conversations with Henri Fennell, who is currently a spray foam consultant. This web page talks more about the method.

    Injected, poured, blown-in, or foamed-in-place (FIP)

    And apologies if this is pushing the edge of commercialism, but Henri is teaching an online course on Heatspring on this method:


    Obviously, existing closed cavity installations will be have R-values limited by the depth of the cavity.

  7. User avater GBA Editor
    Martin Holladay | | #7

    Response to Bo Jespersen (Comment #5)
    To give you enough courage to handle your next interaction with a homeowner, we're going to do a role-playing exercise.

    Bo [speaking to homeowner]: "In order to insulate this sloped ceiling, we're going to have to remove the drywall so we can install some closed-cell spray polyurethane foam."

    Homeowner: "My goodness! That's going to be disruptive and expensive!"

    Bo: "Yes."

    Homeowner: "I don't want to do that because it's disruptive and expensive."

    Bo: "OK. Here's my card. Give me a call if you change your mind."

    Here's my point: It's not your job to come up with a cheap way to insulate your customer's house. It's your job to recommend an insulation method that complies with building science principles and best practices. That's how you avoid callbacks and litigation.

    If a homeowner wants you to insulate the small triangular attic behind a kneewall, you can do that. If they can't afford to insulate the sloped ceiling section, that's the homeowner's choice. If the sloped section gets a buildup of frost and starts dripping, it won't be your problem as long as you didn't insulate that part of the house.

  8. Bo Jespersen | | #8

    Much apperciated
    Two excellent responses gentlemen and I thank you. I will look into the pourable foam a little more and think HeatSpring is a great resource- perfect timing, Kohta.

    This is a simple enough dialogue, Martin (My 9 year old and I have been reading my old Calvin and Hobbes books lately and I can think of many ways to lighten this converstation up:) and it's time to make a change. Perhaps my willingness to walk away will be enough for them to understand the consequences of filling a cavity with air-permeable insulation. As Straube says, not every roof will fail using this method, but some will.

    I have attached a great BSC report that really hit home for me.

    Thanks again and we will keep rockin'!

  9. Patrick Stuart | | #9

    WAP standards
    Just to add to the ventilation issue, the DOE Weatherization Assistance Program (WAP) requires adequate ventilation prior to any attic insulation. Granted, most of what we do is loose-fill cellulose (knee walls and slopes are dense packed), but ventilation is key. Using blown fiberglass where you think there’s moisture isn’t a bad idea, e.g., we require it in mobile home floors for the same reason. It’s a little more expensive, but wet cellulose is basically ruined and then promotes mold. But the bigger issue is providing for adequate ventilation; we won’t install unless that’s addressed first.

    BTW: I’m all for a Bill Watterson and Gary Larson reunion tour. :)

  10. Bo Jespersen | | #10

    Good info, Pat
    Thanks, Pat, and sounds like we have the same sense of humor:)

    So if you are asked to dense pack a finished slope how do you maintain or create an air space? Do you remove the interior finish and start over or have you come up with a way to thread a baffle up the slope?

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |