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
Musings of an Energy Nerd

Insulating a Cape Cod House

Capes are notoriously difficult to air seal and insulate

This photo of a Cape was taken in Levittown, New York, in 1947. When these homes were first offered for sale, they cost between $7,000 and $9,000 each.
Image Credit: Image #1: University of Illinois at Chicago library

If you own an older Cape Cod home, you have my sympathy. If you’re the type of homeowner who regularly tackles DIY projects, you’ve probably spent weeks chasing air leaks with a foam gun, lying on your back in a cramped attic. And there’s a good chance that, in spite of your efforts, your house still suffers from ice dams.

I’m sorry for your troubles. You deserve better.

If you are thinking of building a new Cape, it’s not too late to get the details right — as long as you’re still at the planning stage.

A Cape is a story-and-a-half design. Most Capes have second-floor bedrooms featuring 4-foot kneewalls and two sections of sloped ceiling. The sloped ceiling sections usually rise to meet a narrow horizontal ceiling in the center of the house. Most Capes have triangular attics behind the second-floor kneewalls and a tiny third-floor attic that is too cramped to stand up in.

Cape-style homes are associated with a famous peninsula in Massachusetts as well as a well-known 17,000-home residential development built in the late 1940s in Levittown, New York. Most of the original Levittown Capes were sold with an unfinished second floor; it was up to the homeowners to figure out how to install insulation up there. You can imagine the results.

What’s wrong with a Cape?

The basic problem with a Cape is that most examples have a poorly defined thermal boundary.

Should the triangular attics behind the kneewalls be considered indoor space or outdoor space? From a building science perspective, the answer is clear — these attics should be inside the home’s thermal boundary. However, most builders don’t have the foggiest idea where the thermal boundary in a Cape belongs. That’s why builders and homeowners are always scratching…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. user-5174690 | | #1

    Cape Cod with SIP roof
    I don't have a ton of experience with this style of home, but I was wondering if the use of SIP's at the roof in new construction would have any merit. You would still need to provide structural support, but it would solve some of the difficult air sealing challenges. Any thoughts?

  2. GBA Editor
    Martin Holladay | | #2

    Response to Brian Wolfgang
    I addressed your question in my article, which notes, "There are many ways to build an insulated sloped roof assembly. This type of roof can be vented or unvented. It can be insulated with fiberglass, cellulose, mineral wool, spray foam, rigid foam, SIPs, or nailbase."

    Depending on the span, a SIP roof many not need any structural support. Roof SIPs are engineered and manufactured to handle a variety of spans. Some SIPs include integral structural members. If you are uncertain about the structural needs for a planned SIP roof, talk to your SIP manufacturer.

    Of course, SIP roofs still have air sealing challenges. In a cold climate, it's a good idea (a) to seal SIP seams with interior tape as well as spray foam, and (b) to include ventilation channels between the exterior OSB face and the roofing. For more information on sealing SIP seams, see Air sealing SIP seams.

  3. JIM BAERG | | #3

    Been there, done that.
    As one who has pushed through the difficulties of insulating the attic of my own story-and-a-half home, I very much appreciate the clarity of this article. I was fortunate having a house that was 26 feet wide. That gave me pretty good head room in the kneewall and upper attics. Many of the post war Cape Cod houses are narrower and are much tougher to work in.
    In my case, I thought that insulating the knee wall attic floor (R-60) and knee wall itself (2" of taped ISO) was preferable because it allowed for higher levels of insulation than just filling the rafter cavity. The most difficult area to insulate was the sloped rafters between the two attics. I was able to slip cardboard baffles into those cavities, leave a 1" air space up against the roof sheathing, and still get 5" of cellulose blown in when the upper attic was insulated. Later I checked those sloped roofs with an IR camera and found some of the cavities with insulation voids. We drilled 2" holes through the plaster and dense packed. I'm pretty satisfied with the result and think that I've got a continuous, fairly air tight, insulation package.
    On a more technical note, I notice that the drawings in the article show new construction, rather than what typically exists in older houses. Still very useful, but I would caution readers to think through each building junction as it really exists when they plan their insulation and air sealing project. I've always found that drawing up detailed sections of existing conditions is very helpful before work begins.

  4. GBA Editor
    Martin Holladay | | #4

    Response to Jim Baerg
    Thanks for sharing your experience.

    I agree completely with your point that older homes are rarely framed the way that new homes are framed. Almost everyone who owns an older home learns this lesson quickly. I've seen log joists, rough-sawn joists, and planed joists. I've seen 16-inch-on-center framing, 24-inch-on-center framing, and lots of other types of framing -- like 26 inches on center, or 30 inches on center, or absolutely random spacing.

    I've seen balloon framing and platform framing. I've seen too much blocking and not enough blocking. I've seen double top plates, single top plates, and no top plates. I've seen oversized window headers and no window headers.

    So if you've got an older wood-framed home, you can't assume that the pretty pictures in Fine Homebuilding or GBA represent your house. They probably don't.

  5. woodyfingers | | #5

    proper roof ventilation
    I would like some feedback, please.

    I live in a story and a half cape. The main roof to the side of the dormer was looking lumpy, so I set up the roof jacks and opened things up. The 1/2" sheathing had delaminated and fiberglass insulation was against the underside of the sheathing. My solution was to rip and install 1 x 2's to the sides of the rafters flush with the top of the rafter. I had some 1" polyiso scraps waiting for place to spend eternity, so I loose fit them under the 1x2's and foamed the gaps. Installing the polyiso this way did compress the fiberglass about 2", but I created an 1 1/2" airspace below new sheathing. Is there a better way?

  6. GBA Editor
    Martin Holladay | | #6

    Response to Ed McGuire
    While the preferred solution -- adding rigid foam to the exterior side of the roof sheathing -- would have provided a higher R-value for your roof assembly, your approach is a good lower-cost Plan B.

    Of course, your approach only makes sense if you have soffit vents at the bottom and a ridge vent on top.

  7. Expert Member
    Dana Dorsett | | #7

    Potential trouble ahead.
    How much fiberglass is between the 1" polyiso and the interior ceiling?

    The roof deck will be fine now that it's vented, but unless the ratio of the polyiso R to total R is sufficient for dew point control in your climate zone, you may have a moldy moisture trap on your hands.

    In zones 1-3 you would be fine with 1" polyiso.

    In zone 4C it will be OK as long as there isn't more than R20 of fiber (2x6 rafters with compressed batts are fine, 2x8s maybe not, depends on the final thickness of the fiber.)

    In zones 4A & 4B it'll be fine as long as there isn't more than R16 in the fiber layer, which is fine if it's R19s in 2x6 raftes, could be a problem with high density batts or deeper rafters.

    For zone 5 & colder you have to derate the 1" polyiso to ~R5 for wintertime performance. For dew point control that R5 needs to be at least 40% of the total R in zone 5, 50% zone 6, and 60% in zone 7.

    So, in zone 5 it's good for at most R7.5 of fiber, which could be a problem even with compressed R11s in 2x4 framing, since an R11 compressed to 2.5" is about R9- you compressed it a bit more, so it's probably OK with 1.5-2" of fiber, but not more. With any 2x6 or deeper rafters it's definitely a problem.

    And in zones 6 & higher it's a problem.

    If the interior has an air tight polyethylene vapor barrier it might do OK even without sufficient foam-R for dew point control, as long as the rafter bays were pretty dry when you sealed it all in. If there are air leaks to the interior side or no vapor retarders other than latex, the risk in zone 6 or higher (or zone 5 with 2x6 or deeper rafters) is pretty real.

    If there's any salvation at the margins here, it appears to be a south facing pitch means that the average winter temps at the roof deck are a bit higher than on the north side, which means you can cheat the minimums a bit. If it's not even close, there's a good chance you'll have wet insulation at the end of the winter in any rafter bays that leak air or have more than 1-perm of vapor retardency on the interior.

  8. GBA Editor
    Martin Holladay | | #8

    I disagree with Dana Dorsett
    I wouldn't panic yet. I disagree with Dana's analysis.

    Dana is right that if you are building an unvented roof assembly, you need to pay attention to the ratio between the rigid foam layer and the layer of fluffy insulation (in this case, fiberglass).

    However, yours is a vented assembly (assuming that you have soffit vents and a ridge vent), so there will be at least some airflow through the ventilation space. This airflow will keep your roof sheathing safe.

    Concerning the question of whether site-built ventilation baffles need to be made from a vapor-permeable material, I answered that question in my article called Site-Built Ventilation Baffles for Roofs.

    The paragraphs below are quotes from my article.

    * * * *

    One of the reasons that builders install ventilation channels is to help damp roof sheathing dry out. Researchers now realize that ventilation channels can help a little bit at this task, but not as much as some people think. (In fact, during the summer, ventilation channels can actually add moisture to roof sheathing rather than remove it; for an example of this phenomenon, see Image #4, below.)

    The most important way to prevent the migration of moisture from the interior of a house to the roof sheathing is have a good air barrier at the ceiling. The reason is simple: the usual transport method for this moisture is air leakage, not vapor diffusion. Because of this fact, installing ventilation baffles that are airtight makes this type of roof assembly more, not less, robust.

    What if interior moisture is able to reach the underside of a ventilation baffle — isn’t it possible that the moisture might condense against the baffle (especially if the baffle is cold)? If so, isn’t this a good argument in favor of using vapor-permeable materials (for example, fiberboard, cardboard, or thin EPS) for ventilation baffles?

    The answers to both questions is a qualified yes. Anyone worried about this possibility should probably make their ventilation baffles out of a vapor-permeable material.

    That said, there really aren’t any reports of failures or problems resulting from the use of vapor-impermeable materials — for example, polypropylene, vinyl, or foil-faced polyiso— to make ventilation baffles. The main reasons:
    Not much moisture manages to make its way to the ventilation baffles (especially in homes that pay attention to airtightness);
    The air in the ventilation channels is often warmer than outdoor air, a fact which limits condensation; and
    Any moisture that does make its way there seems to be incorporated into the rafters via sorption. The ventilation channels are able to remove a limited amount of moisture from the rafters, and it appears that the rate of drying exceeds the rate of wetting.

  9. Expert Member
    Dana Dorsett | | #9

    The problem isn't at the roof sheathing.
    As I stated: "The roof deck will be fine now that it's vented..."

    The problem is that the first condensing surface (the interior side of the foam) is colder than the wintertime interior air's dew point. As long as it's air tight to the interior and low-perm, it's not a problem. But with air leaks it's a potential problem. With high vapor permeance at the interior it's a potential problem, but mitigated by solar gains on the south facing pitches.

    Counting on sorptive transfer of moisture around the vapor barrier via the rafters into the vent channel might be OK if it's just a thin sheet, with less than 200 mils of rafter to traverse, but with 1" polyiso there is 1" of wood that it has to get through. An inch of wood will run 1 perm or less unless it has a high moisture content (high enough to support mold), a class-II vapor retarder, which isn't providing a high drying rate for the cavity.

    This isn't time to panic, but it's not best practice by any means.

  10. charlie_sullivan | | #10

    A rare disagreement
    If Martin and Dana disagree, it must be an interesting and challenging question! So I can't resist adding my $0.02.

    It seems to me that the opportunities for moisture to escape when an impermeable baffle is used are much greater with an unconditioned attic than with a cathedral ceiling. With the unconditioned attic, the insulation in question fills a triangular area, and moisture can escape by traveling laterally through the insulation to the vented attic space, even if it can't escape through the baffle, and even if the drying through the rafters is minimal. I wonder how much of the field experience Martin's article refers to is with impermeable baffles installed in unconditioned attics rather than cathedral ceilings.

    Several other factors that may help explain why there aren't more reports of problems are 1) Many baffles are installed with without the edges air sealed, 2) Many are installed without any real protection against wind washing coming from the eaves, so there is outside air flowing across the bottom surface as well as the top, and 3) Unlike OSB, the vapor impermeable materials themselves are pretty tolerant of moisture, so if there is some condensation on that surface, it might never got noticed and might never cause a real problem.

    In any case, this case seems like reasonably low risk--with some foam, the condensation potential will be reduced relative to a simple uninsulated impermeable plastic baffle, and everyone agrees that the roof deck will be fine.

    More generally, it might be that the supposedly best commercial baffle product, DCI Smart Baffle, is also the most dangerous. It has 2" plastic flaps on each side, to provide spacing from the roof deck. Thus, the vapor path through the wood goes up from the 1/5 inch that Dana think might work to two inches, but without any of the benefit of the R-value of foam. Perhaps "Smart Baffle" refers to the fact that smart people are baffled by the 3D hygrothermal analysis that would tell us when, whether, and how it actually works well. I'm hoping that it works OK in an unconditioned vented attic, because I am in the process of installing a similar baffle configuration in my attic.

  11. bill02888 | | #11

    2x6 rafters, a roof leak, PV, and a raccoon
    The roof of my 40's Cape in central Rhode Island has sprung a leak -- under PV panels, of course. The asphalt shingles seem to be at the end of their life too. One end of the 2nd floor is a bedroom. The leak has already badly effected the original plastered and calcimine-coated rock lathe. The other end is unfinished attic space which had 1" of fiberglass insulation face-stapled to the rafters (seriously!) which was used for storage. Yes, it had 1" of fiberglass until a raccoon turned the attic into a B&B/rave last winter. The shredded insulation is all over the place. I can now see light coming in through the ridge vent when I look up at the attic ceiling. That attic wasn't conditioned at all, but the minimal insulation did help a little in moderating temperatures somewhat. And the air leakage from the living space below ... well, you know! ;-)

    I'm at a crossroads. Now is the best time for me to take action. Either I find somebody who can design and install R-49 rigid foam above the rafters, and some sort of roofing on top of that which will "solve all of my problems". Or I find a decent roofer who can just fix the leak and replace the existing shingles, then I can do the best air sealing and insulating that I can given the confines/restrictions of 2x6 20" o.c. rafter construction with no soffit vents. There's a gable vent for each "behind the kneewall" area, and a ridge vent. It seems impossible to do anything about the 5 1/2" space between the sloped part of the bedroom ceiling and the underside of the roof deck (currently T&G board sheathing). Does code allow for "sorry but this can't be brought up to R-49 without complete demolition" in cases like this?

    How would I go about finding a builder/roofer who "does" rigid foam insulation on top of rafters? I'm sure they exist. Surely there's a better way than calling up every builder/roofer within 50 miles?

    ... --- ...

  12. GBA Editor
    Martin Holladay | | #12

    Response to William Sherman
    Q. "Does code allow for 'sorry but this can't be brought up to R-49 without complete demolition' in cases like this?"

    A. Yes. In most municipalities, homeowners who engage in a renovation project are not required to meet all modern building codes. That said, it's always best to check with your local code inspector to learn about local ordinances. One other point: Just because it is legal to ignore insulation requirements doesn't mean it's a good idea.

    Q. "How would I go about finding a builder/roofer who does rigid foam insulation on top of rafters? I'm sure they exist. Surely there's a better way than calling up every builder/roofer within 50 miles?"

    A. You might want to call up a local energy rater -- one who is certified by RESNET or BPI -- and ask the rater if they know any local contractors who understand building science issues and green building methods. Lists of certified raters can be found by visiting the RESNET and BPI websites.

    You might want to follow the approach used by Pauline Guntlow, the author of a relevant GBA blog called "Solving an Ice Dam Problem With Exterior Rigid Foam." Guntlow wrote, "I called a number of general contractors to hear about their experiences using continuous rigid foam insulation above the roof deck on a sloped roof. (Commercial roofers have successfully used this approach for years on low-slope roofs.) Of the people I called, only two — a G.C. from Long Island and a building inspector from Massachusetts — had used this method."

    Good luck.

  13. Expert Member
    RICHARD EVANS | | #13

    Raccoon Rave
    Bill, I know your situation is unfortunate but I enjoyed the description of your roof issue. Too Funny!

    Good luck on finding a solution

  14. chimewind | | #14

    If I were to insulate cape cod style house by insulating over roof deck how do I compensate for two dormers on front of house with 12/12 pitch? Since I live in zone 7 I'd need to add about 12" of insulation to roof deck which means I'd have to compensate in some way the windows in the dormers.

    1. GBA Editor
      Martin Holladay | | #15

      If you have existing windows with window sills that are close to the level of the existing roof sheathing, you can't easily add lots of insulation on the exterior side of the roof sheathing -- unless you are willing to remove the existing windows and install smaller windows.

      1. chimewind | | #17

        I wonder if it makes sense to enlarge the dormer down the roof another foot. That way the bottom of the window would be same height again. Not sure if it would still look okay.

    2. anonymoususer | | #16

      Hi. This may be intolerable to some but we opted to remove the one dormer on our Cape and install a skylight in the resulting opening. I know many gba nerds object to both dormers and skylights but we went for what we hope is the lesser of two evils so we could add rigid insulation outside roof sheathing.

      1. chimewind | | #18

        Thanks for the suggestion. That's actually an interesting idea.

  15. rcarmichael | | #19

    Hello Everyone,

    I’m hoping for roof insulation advice for my 1940s 1-½ story cape in climate zone 6a with dark gray asphalt shingles. When we moved in about 10 months ago, I removed the 70’s wall and ceiling paneling in the northwest upstairs bedroom, partially to investigate a sporadic chimney leak and partially for aesthetic purposes. I would like to determine what (if any) insulation work I should do before closing the room back in with drywall.

    House Orientation and General Details:
    -The house has ridge and gable vents in the upper attic but no soffit vent as the house lacks overhangs.
    -Long sides of the house face southwest and northeast. The northeast side (rear of the house) has a small first floor addition on the northwest end of the house with a cross gable roof.
    -The roof is in decent condition with an estimated 5-10 years of life remaining.
    -The house gets moderate ice dams if I don’t rake the roof after it snows, especially on the northeast side that has a valley and gets less sunlight.
    -There are forced hot air ducts in the southwest crawl spaces of both bedrooms.

    Current Condition, Unfinished Bedroom (northwest):
    -2x6 rafters with R-11 foil-faced fiberglass in each of the 16 oc bays, leaving an air gap between the underside of the 1x sheathing and the back of the fiberglass.
    -At the bottom of the ceiling joist plane, the rafter insulation turns horizontal and terminates by overlapping the insulation from the opposite side in the middle of the 4’ ceiling joists.
    -Room is quite cold in the winter as there is no air barrier to stop warm interior air from being sucked out the ridge vent
    -2x4 gable wall has the same R-11 foil-faced insulation.
    -No insulation in the knee wall
    -Old and thin fiberglass in between the floor joists.
    -A few bays along the gable end wall have 2x4 rockwool instead of the foil-face R-11.
    -Northeast crawl space has a 4 foot piece of 2x4 rockwool blocking a passthrough to the attic of the addition.This attic has blown in cellulose and a gable vent.

    Current Condition, Finished Bedroom (southeast):
    -Knee walls insulated with unlabeled 2x4 paperfaced fiberglass
    -Crawlspaces are insulated with R-19 paper faced fiberglass batts in the floor joist bays, rafters bays, and gable walls. No blocking between bays.
    -Unknown insulation inside enclosed rafter bays, upper attic, and gable wall.

    Discussion of Possible Solutions and Questions
    -From this article as well as several others on GBA, it seems clear that it is best practice to include the crawl spaces in the thermal envelope of the house and to add continuous exterior insulation to the roof. I’m willing to consider that the next time I reroof, but don’t plan on doing so for several years. My goal for now is to make the unfinished bedroom livable, without doing anything that will result in moisture accumulation or damage the integrity of the house. If I need to rake the roof for now to prevent ice dams, I’m ok with that. With that said:
    -Should I add more insulation to the top attic and/or add thicker insulation to the rafter bays or will it potentially cause moisture issues?
    -Should I try to install ventilation chutes and some type of fascia vent (or extend the rafter tails and add soffit vents)? I’m not particularly willing to demo the plaster in the finished bedroom and am not confident I would be able to successfully install the chutes from either above or below. If I don’t have ventilation chutes in the whole house, does it make any sense to add them to only the unfinished section?

    Thanks for sticking with me through this long post. I appreciate any feedback you can provide.


  16. SSHabal | | #20

    @Martin, everyone can you please share your opinion about the new cape, where the attic is indie the thermal envelope as you show it but there is a ceiling as well. Would you insulate the roof along the attic and then along the attic ceiling? Or the whole roof assembly? Attaching a drawing for reference. Would appreciate your advice.

    1. MartinHolladay | | #21

      S.S. Habal,
      The sketch shows HVAC equipment located in an unconditioned attic. That is never a good idea. For more information on why this is so, see these two articles:

      "Keeping Ducts Indoors"

      "Solutions to the Attic Duct Problem"

      The sketch shows insulated kneewalls. This approach is not a good approach, in general. This article explains why:

      "Two Ways to Insulate Attic Kneewalls"

      1. SSHabal | | #22

        Thanks. I appreciate your advice.

        Defenetely the attic will be fully in the insulated envelope. I will continue my research as it appears my designer setting me up for issue you describing.

        1. Expert Member
          MALCOLM TAYLOR | | #23


          Definitely ditch the insulated knee walls, but I'd keep the small vented attic if you have no mechanicals up there.

  17. maine_tyler | | #24

    All these years later, do you still think the over-the-roof foam is 'the green' option here? The price to perform the exterior retrofit is extreme.

    Insulating contractors regularly improve capes from the inside despite the work being 'difficult'. If it was truly more difficult than foam over the roof, it would be more expensive, no? But it's no where near the price. Hardly any contractor wants to install foam on a roof. Maybe the argument is that it's cheaper but doesn't perform as well... OK sure, but by how much? I doubt there's a human lifetime ROI for doing the exterior foam vs standard interior work.

    I like the performance it offers, but it's hard for me to see this as the 'best' option in terms of 'green.' How many people have been steered this direction only to find it untenable?

    1. MartinHolladay | | #25

      Excellent questions.

      Q. "Do you still think the over-the-roof foam is 'the green' option here? ... Maybe the argument is that it [insulating from the interior] is cheaper but doesn't perform as well."

      A. As far as I know, I never said that installing rigid foam above the roof sheathing is the green option. (That doesn't sound like me. I've never liked the term "green.") As you correctly guessed, installing a continuous layer of exterior rigid foam results in better roof assembly performance than insulating between the rafters.

      Recent research into the embodied carbon emissions problem convinces me that the best "green" option is to perform almost no construction or remodeling work until our current climate crisis is under control. For more information, see "A Proposed Solution to the Embodied Carbon Problem."

      Q. "I doubt there's a human lifetime return on investment [ROI] for doing exterior foam vs. standard interior work."

      A. You may be right, but ROI calculations depend on the cost of the work. The most affordable approach involves the use of recycled (reclaimed) polyisocyanurate and volunteer (homeowner) labor.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |