GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Guest Blogs

Frugal Happy: Giant Foam Box, Part 2

It’s “man battles the heat” as three layers of insulation go into the roof

Step 1: A layer of rigid foam insulation. The first step in a process that ultimately took four months was to cut pieces of 3-inch-thick rigid foam insulation and install them between the rafters.
Image Credit: Chris Stratton and Wen Lee
View Gallery 10 images
Step 1: A layer of rigid foam insulation. The first step in a process that ultimately took four months was to cut pieces of 3-inch-thick rigid foam insulation and install them between the rafters.
Image Credit: Chris Stratton and Wen Lee
Not much to start with except undersized rafters and a cavity without any insulation. After the retrofit, the roof is 17 inches thick and has three distinct layers of insulation. The R-value of the assembly is 46. The pieces of 3-inch-thick polyiso insulation are installed between rafters against strips of foam measuring 2 inches square. This creates a 2-inch gap between the sheathing and insulation for ventilation. Cutting back the sheathing at the ridge creates an opening for air to exit the roof assembly. The slot is covered by a ridge vent, which keeps water out but allows air to exit. Fiberglass batts went in after the first layer of rigid foam. The batts are nominally 6 inches thick. The roof now has three layers of insulation, plus a ventilation gap. Chris used acoustical sealant and canned foam to seal the seams of the second layer of rigid foam insulation. Chris used 2x3s to create a service cavity below the insulation.

Editor’s Note: This post is one of a series by Chris Stratton and Wen Lee, a husband-and-wife team living in the Los Angeles area who are turning their 1963 suburban house into an all-electric, zero-net energy home. They chronicle their attempts at a low-carbon, low-cost, and joyful lifestyle on their blog Frugal Happy. This post was written by Wen.

After spending months painstakingly insulating our walls, Chris turned his attention to insulating the vaulted ceiling.

You may be wondering, “Is this guy a glutton for punishment?” Valid question.

In many ways, insulating the ceiling was even harder than insulating the walls. The assembly is thicker and more complex. Chris had to install the insulation while standing atop a ladder, which made the job awkward and dangerous. Moreover, much of the insulation work took place during the hot summer months, when the air temperature at ceiling height often neared 100°F. It was laborious and slow-going. In the end, the process lasted more than four months.

So yes, maybe Chris is a bit of a glutton for punishment. But for the end product of a foam-filled, air-sealed, attractive cathedral ceiling with an insulation value of R-46, there was no question in Chris’s mind he wanted to do it. The process helped create a comfortable home with steady indoor temperatures and low energy bills year-’round.

Since we converted our flat ceiling to a vaulted ceiling, the ceiling and roof became one and the same. So our roof went from zero insulation to superinsulation. The two drawings at the bottom of this post show the roof assembly before and after the work (see Images #2 and #3 below).

That’s a big change! By the time Chris has completed the retrofit, our ceiling/roof will have 10 distinct layers and be 17 inches thick. Wowza. Let’s walk through the insulation phases one by one.

Phase 1: Rigid foam and ventilation

Chris wanted to ensure that the finished ceiling/roof assembly was properly ventilated so that any moisture that gets into the assembly (e.g. from rain) is able to evaporate. He did this by constructing long ventilation baffles, or chutes, from rigid insulation. The chutes allow air to flow from soffit vents at the eaves to the ridge vent at the top of the roof. (For more information on this step, see “Site-Built Ventilation Baffles for Roofs.”)

Often people install plastic vent baffles for this purpose, but since Chris was planning to insulate using polyisocyanurate rigid foam, he decided to use the foam itself to serve as the chutes (see this fun video from Fine Homebuilding). Two birds with one stone!

Each panel of foam is pushed up against 2-inch-square “sticks” of foam glued into the upper corners of each rafter bay with Loctite PL 300 foamboard latex construction adhesive. This creates a 2-inch gap above the insulation for air to flow (see Image #4 below). As you probably guessed, what you see in the photo is just half a chute — when it’s complete it will extend all the way up to the ridge.

For air to flow successfully through each chute, there needs to be a way for air to 1) enter, and 2) exit. In the photo, you can see that Chris created openings at the base of the rafters (where daylight is shining through along the top of the wall). Later, he installed plastic soffit vents.

But where will the air exit? At the ridge, the highest point of the roof. Chris cut out openings along the ridge too (see Image #5 below). Then he climbed on top of the roof and installed a ridge vent, carefully covering it with ridge shingles (see Image #6 below). That way air from inside can flow outside, but when it rains, water will not leak into the house.

Another benefit of ventilating the roof in this manner is that on very warm days, hot air (which rises) will flow out the top vents, drawing cooler air into the chutes from the bottom — a passive way of cooling the house down.

Working by yourself can be exhausting and lonely (even when there are unlimited podcasts to listen to), not to mention s-l-o-w! One weekend, Chris organized a work party (a.k.a. bribed people with free pizza), and luckily, my brother Bin and our adventurous friends Merc and Dan showed up to help. With four people, they were able to install a whole bunch of chutes in one afternoon.

As with the wall insulation, every piece of rigid foam had to be carefully measured and custom cut for each bay, and then gently pounded into place with a block of wood and a 4-pound mini-sledge. The pieces were snug enough to stay in place with friction alone, but Chris sealed the edges with acoustical sealant.

Over the next few weeks, Chris continued to slowly but steadily install the foam on his own. As he progressed, he had to climb ever higher on the ladder. Handling an 8-foot-long piece of foam while balancing on top of a ladder (in sweltering heat) is no small feat, people.

Of course, insulation doesn’t work at its best if there are leaky seams. So Chris made sure to air seal (with acoustical sealant and canned spray foam) along all the seams of the rigid foam, creating an insulation layer that is completely airtight.

Phase 2: Fiberglass

After the rigid foam, the next layer to go in was 6-inch-thick fiberglass batts (see Image #7 below). That’s the cotton candy-looking stuff. It may look soft and fluffy, but you don’t want to roll around in it because millions of teeny tiny spiky things will become embedded in your skin, and you will be thoroughly itchy.

Thankfully these are kraft-faced batts, which means they are covered on one side by paper. This is designed to keep out moisture, but it also makes the batts easier (and less itchy) to handle and install.

Chris installed the fiberglass batts by stapling the paper facing to the rafters. This isn’t a particularly labor-intensive process, but because the batts were so long and floppy, and Chris was trying to install these things by himself so high up, it was rather awkward.

It also got challenging when he had to get into tight corners. Even so, the fiberglass batts went in much faster than the rigid foam.

Phase 3: More rigid foam

Similar to what he did with our walls, Chris next installed yet another layer of rigid foam, this time 1 inch thick, continuously across the rafters rather than between them. Why more rigid foam? To reduce thermal bridging through the rafters.

The foam sheets are not very heavy, but at 4 feet by 8 feet they are rather unwieldy for a single person to handle. Chris got creative with ways to hold up the large sheets while on a ladder (such as using another ladder).

Two layers of rigid foam, filled with fluffy pink fiberglass. That’s right, it’s an insulation sandwich (see Image #8 below)! Fast forward a few weeks, and behold! The fully insulated ceiling: 10 inches of insulation (and months of Chris’ blood, sweat, and tears) hidden behind a dull gray facade.

It looks nice and tidy in the end, but getting to this point was not. It’s all out of sight now, but we’ll always remember that it’s there. Once more, Chris meticulously air sealed the final layer of rigid foam with acoustical sealant (the white stuff) or canned spray foam (the orange stuff). Because, darn it, you can never air seal enough (see Image #9 below). The air sealing ensures that there are no air leaks in the ceiling of our giant foam box house.

Phase 4: Furring strips and a service cavity

Just as we did with the walls, we created a service cavity on the inside of the insulation for wiring before putting on the finished ceiling. To do this, Chris installed 2×3 furring strips. He used a laser level to ensure that the strips were perfectly level and co-planar (see Image #10 below). Chris can use the furring strips to anchor electrical junction boxes for lighting fixtures later on.

At long last — the furring strips are in, and the insulation is complete! We are ready for wiring, lighting, and the finished ceiling surface (which will be wood paneling).

It was long, it was laborious, it was sweaty… but it was worth it. The best thing about insulation is that if it’s done right, it only needs to be put in once. Once it’s in, it pretty much works forever (or at least for the rest of the building’s functional lifetime).

With superinsulation installed, our house now needs less than half the air conditioning it used to, and an even smaller fraction of heating. If all homes had this level of insulation, imagine how much energy (and greenhouse gas emissions) could be saved. I’ve said it before and I’ll say it again: Insulation is unsexy, but it makes a difference!


  1. truthbox | | #1

    roof insulation
    Great write-up on this process. I've really enjoyed reading your series of posts. This is a very robust assembly, but, clearly, very labor intensive. Can you comment on why you decided to locate the insulation on the inside rather than stack up rigid foam on the exterior side of the roof (or a combination)?

    Also, is there a risk here in having fiberglass sandwiched between two impermeable layers?


  2. Joshua_Elliott | | #2

    Jealous of your ceiling height
    Considering the same thing on my home but the ceiling height in the attic/second floor is low enough that I'm not sure I can stand to lose the extra inches. Only other options seems to be waiting until the roof needs replacement and then putting additional insulation on the roof deck.

  3. JC72 | | #3

    Back in 2000 I so wanted to do this when we
    opted to finish the 3rd floor of our townhome. Unfortunately in my case I wasn't aware that I could buy reclaimed foam and my truss spacing was all over the place.

    Question for the science folk: Would the ends (i.e. soffit & ridge) of the fiber insulation have to be capped in order to prevent convective looping between the layers of foam?

  4. GBA Editor
    Martin Holladay | | #4

    Response to Chris Stratton (Comment #3)
    Q. "I don't have a great intuitive sense of permeability, so I'd love to hear thoughts from knowledgeable others about whether the fiberglass is a big screw up, or not so much."

    A. Not so much, in my opinion. Two factors are usually present in cases of damp cathedral ceilings: (1) lots of air leakage that pulls indoor air through the assembly -- the air usually exits the assembly through cracks near the ridge -- and (2) a cold climate that encourages condensation or moisture accumulation.

    You don't have either of these risk factors.

  5. GBA Editor
    Martin Holladay | | #5

    Response to John Clark (Comment #4)
    Q. "Would the ends (i.e. soffit & ridge) of the fiber insulation have to be capped in order to prevent convective looping between the layers of foam?"

    A. You always need attention to airtightness whenever you are building this type of assembly. Chris and Wen planned their ventilation channel carefully -- so the air that enters the soffit vents passes above their site-built ventilation baffles (made of polyiso), and exits at the ridge vent.

    There should be no opportunities for outdoor air to enter the cavity where the fiberglass batts are installed.

    Convective looping sometimes occurs even in the absence of any outdoor air leaks -- for example, when one side of a fiberglass-insulated assembly is very cold, and the other side of the fiberglass-insulated assembly is very warm. That won't occur here, however, because the rigid foam on both sides of the fiberglass-insulated assembly ensure that these surfaces are neither very cold nor very hot.

  6. jcstratton | | #6

    @ Joshua
    Yeah the depth of the ceiling from the underside of the original 2x6 rafters to the finish ceiling surface is significant -- something close to 9 inches (1.5 of that is the service cavity). To us, because we went from flat to vaulted, it still feels like a huge gain in space. But if you're used to a vaulted space, it may feel like a somewhat of a loss.

  7. jcstratton | | #7

    @ Martin (comment #5)
    Thanks Martin. This is great to hear.

  8. jcstratton | | #8

    @ user-6816482
    Thanks for reading. I decided to locate the insulation inside because I didn't want to have to replace the roof, which is in pretty decent shape. Yes in retrospect the fiberglass sandwich with polyiso bread may not have been a great choice for vapor diffusion. I used fiber faced polyiso on the interior because it has a much higher permeance than foil faced (26.0 vs 0.05, according to this). Also we have a very dry climate, so that should help. I don't have a great intuitive sense of permeability, so I'd love to hear thoughts from knowledgeable others about whether the fiberglass is a big screw up, or not so much.

  9. wsuman | | #9

    The rigid foam you show is a
    The rigid foam you show is a class B. Doesn’t it require a minimum 15 minute thermal barrier, like 1/2” gwb, between occupied space?

  10. jcstratton | | #10

    @ wsuman (comment #9)
    Hmm, good question. I'll let others answer this more definitively. This article discusses the "thermal barrier" fire requirements for spray foam, I don't know if the requirements are the same for rigid foam? In any case, a relevant passage from the list of exceptions to the 1/2" drywall requirement for ceilings with spray foam:

    When the spray foam is installed in a roof assembly and is protected on the interior side with a layer of tongue-and-groove boards or plywood measuring at least 15/32 inch thick (2009 IRC R316.5.2);

    Our ceiling will have 1/2" tongue-and-groove hardwood and there's a 1.5 inch air gap between the back of the tongue-and-groove and the rigid foam, if that matters.

    Also I've showed the plans for the ceiling assembly to my city's plan reviewers, and they signed off on it, for what it's worth. I realize that's a different question than whether or not this is sufficiently safe or advisable.

  11. Mark_Pollard | | #11

    Polyiso perm rating

    Who is the manufacturer of the paper faced polyiso? It is my understanding that all polyiso is closed cell foam, thereby making it a Class I vapor barrier. The paper facing is a way to reduce cost and provide a substrate to adhere different products to the panels (I.e. membrane roofing). Aluminum facing is more expensive for obvious reasons but makes taping panels easier for air sealing purposes, and if an air gap is provided between itself and the inner layer (i.e. drywall), it will see a marginal gain in R-value due to long wave radiation being reflected back.

    Like you and Martin discussed above in the comments, your assembly is very low risk for moisture accumulation due the climate you live in. Lucky dog! Great job with your house, your neighbors should be envious and maybe inspired to "Save the BTU's" as well.

  12. brendanalbano | | #12

    re: polyiso perm rating
    Mark, in general my understanding is that foil faced polyiso is a class I vapor retarder, and fibergalss faced polyiso is a class II vapor retarder, and for both, the facing is providing most of the vapor retarding properties. Here's an article with a chart that touches on generic perm ratings for polyiso:

  13. user-6970808 | | #13

    Another Foam Building Advisor article
    What could possibly be construed as "green' about building another tox-box out of foam, fiberglass and construction adhesive? Come on now, let's get some advice on actually building green.

  14. GBA Editor
    Martin Holladay | | #14

    Response to User-6970808
    This is a guest blog written by two homeowners who are striving to reduce their carbon footprint by minimizing the number of miles they drive and reducing the amount of energy needed to run their home. Their goals and methods are of interest to Green Building Advisor readers, in my opinion.

    GBA welcomes guest blog authors who have a variety of opinions and use a variety methods, and we don't require all guest bloggers to agree to any list of opinions or rules before we publish their blogs.

    I'm not sure what a "tox-box" is, but I'm confident that the home remodeled by Chris and Wen isn't poisoning the home's occupants.

    If you are interested in building a house that has no foam insulation, GBA has many relevant articles for you to read, including this one: "Building a Foam-Free House."

  15. JC72 | | #15

    I believe the OP is using re-purposed rigid foam in a pre-existing structure. That's green to me.

  16. wsuman | | #16

    John Clark is

    John Clark is correct, energy savings with a net reduction in footprint is unquestionably “green”.
    What is your go-to insulation? The readily available repurposed cotton fabrics, design flexible straw bales, cellulose..?
    I trust you wouldn’t consider mineral wool to be superior in that context. Do you know how much energy it takes to turn stone into fibers? Or how long for a carbon payback?
    There’s merit to many mainstream solutions, when rigid is reused, so much the better.

  17. user-6970808 | | #17

    No Perfect Insulation
    All insulating materials have their pluses and minuses. I do not see where in this article it is mentioned that this is re-purposed foam. I guess I could dig back through a lengthy blog to find that info but this article for all practical purposes is promoting foam, fiberglass and adhesives. Re-purposed foam is most likely done off-gassing and shrinking and any salvaged materials have much lower embodied energy for sure but I would like to see numbers comparing fiberglass/foam to mineral wool/cellulose in that department. wsuman? Any numbers for us? I do not believe "energy savings" are the only criteria for "unquestionably green" status, there are just so many other things to consider.

    Mineral wool is most certainly superior in terms of fire resistance, insect resistance, water repellency, not off-gassing toxic chemicals into your living space, longevity, dimensional stability and last but not least thermal resistance at temperatures where it matters.

    I also do not see mentioned in this article if the adhesives are no/low toxicity. If your house is poisoning you it is not green. In so many articles in GBA I see a high reliance on toxic adhesives and tapes to attain air tightness which then has to be offset by expensive ventilation systems which do no good if you live in a place that has wildfire smoke every summer. Sometimes it seems like complexity for complexities sake. Simplify, use local materials when possible, if you need gloves, mask and protective clothing to install a material it is not good for the builder, the planet, nor the people who will inhabit the structure.

    Martin, I'm not sure what "minimizing the number of miles they drive" has to do with green building unless of course you mean number of trips to the hardware store? Part of an energy aware lifestyle for sure but not really related to building. I do appreciate the wide range of tactics expressed here at GBA and even if we don't all agree we can still learn from each other. There is no "one right way". Thanks for all your cat herding.

  18. GBA Editor
    Martin Holladay | | #18

    Response to User-6970808
    If you want to engage in a dialog on this issue, it would be helpful if you would tell us your name. (I'm Martin.)

    I agree with you that there is "no perfect insulation." That's a much better starting point for dialog than accusing Chris and Wen of building a "tox box."

    Of course we all agree that "If your house is poisoning you it is not green," but there is no evidence whatsoever that Chris and Wen's home is poisoning them. There are many people who have suffered poisoning -- I'm thinking of the families in Flint, Michigan affected by high lead levels in the water -- and it is disrespectful to these families to toss off the word "poisoning" so casually. Poisoning happens, but green builders need to be precise and compassionate when discussing poisoning.

    The use of a mask and gloves in building is common sense, and these items of clothing are no indication that the builder is making a mistake. If you told the average builder that he or she would have to give up the use of gloves because using gloves isn't "green," the builder would laugh.

    You wrote, "I'm not sure what 'minimizing the number of miles they drive' has to do with green building unless of course you mean number of trips to the hardware store?"

    I have two reactions to this comment:

    1. Yes, minimizing trips to the hardware store is, indeed, part of green building. To read more about Chris and Wen's approach to this issue, see "Frugal Happy: A Car-Free Experiment."

    2. I urge you to read John Abrams's article, "Tracking Our Company’s Carbon Footprint." In that article, Abrams describes his company's efforts to determine their carbon footprint. After a series of calculations, Abrams concluded that "by far the largest source of energy use in our company at present is employees getting to and from work and driving around (hopefully not aimlessly) doing errands during the day."

  19. jcstratton | | #19

    @ user-6970808 (comments #13, 17)
    It's probably inadvisable for me to weigh in on this. But nevertheless...

    Anonymous user-6970808, I think you have a point. We (collectively) should be using non-petroleum derived building materials, adhesives, insulation, etc. But I hope that you can understand that high performance, non-petrol based materials are just beginning to become available in the US -- they are very unusual, very hard to get, and very expensive.

    I looked into sheep's wool insulation. Into gutex, rock wool. Because of some combination of expense, unavailability, as well as physical constraints, (i.e., the wall/roof would have to be too thick for the desired R value, and hence not compatible with a retrofit application), I decided not to go with these and went the rigid foam route instead.

    Also, as far as air quality and "poisoning", I don't know what your claims are based on. Can you point to peer-reviewed studies that have found that polyiso, formaldehyde-free fiberglass batts, acoustical sealant, one-part canned spray foam -- after everything has cured -- have negative chronic IAQ effects? I know of no such studies.

    As for the embodied energy and lifetime global warming potential of different insulating materials, I would point you Alex Wilson's
    analysis on this question from a few years back.

    As you say, "all insulating materials have their pluses and minuses." Given that this was my first renovation project and that I would be designing, paying for, and constructing everything myself, I opted not to ratchet up the already high level of difficulty/expense yet further by utilizing very exotic and hard-to-find building materials, and instead used slightly less exotic and only moderately-difficult-to-find (in my area), yet petroleum-derived, materials.

    It is fairly easy -- and dare I say, cowardly -- to offer an anonymous armchair critique of another's work. I would invite you to put forth an example of another extant high-performance residential project that utilizes the kind of materials and assemblies that you have envisioned. I really do think we all want the same thing: high-performance buildings that have good indoor environmental quality that have readily constructable assemblies and utilize affordable, readily available materials. Inevitably, in actual built projects (as opposed to conceptual ideal ones), not all of these aspects can be simultaneously realized and tradeoffs must be made.

    I personally find inspiration and hope (and maybe a path forward?) in the wood and cellulose-based passive house assemblies that are being developed in Europe, like this one shown by Bronwyn Barry from this year's passive house conference in Munich:

  20. Expert Member
    MALCOLM TAYLOR | | #20

    User 697 etc,
    You wrote:
    "In so many articles in GBA I see a high reliance on toxic adhesives and tapes to attain air tightness which then has to be offset by expensive ventilation systems which do no good if you live in a place that has wildfire smoke every summer."

    i think you need to un-pack this a bit. The ventilation systems in tight houses aren't primarily to reduce the residual toxins form building materials, their use is usually governed by reducing CO2, and moisture, both of which build up in every well-sealed house no matter what it is made of.

    Every house needs fresh air. Poorly-sealed ones get it inadvertently though leaks, well-sealed ones decide where it will be admitted from. In both cases the air needs to come from outside, and is reliant on the quality of that air to refresh the interior. i can't think of a house that wouldn't be affected by wildfires or similar problems - although perhaps those with ventilation systems incorporating filters might fare better.

  21. AntonioB | | #21

    Great article! I like reading other people's practical experience.

    On a side note, so often when we solve one problem we create another. We need to be conscious of this. I see a potential structural problem with this solution. In this particular house it may or may not be a problem. I can't judge, not having adequate information. But it's something to keep in mind.

    The continuous slice through the roof sheathing to create a ridge vent ... In high seismic areas a plywood roof diaphragm serves a very important role in the lateral load resisting system, the system that you rely on during earthquakes. Cutting continuously through the plywood effectively eliminates its ability to transfer lateral loads through the roof diaphragm to shear walls.

    So before using a solution like this I encourage you to consult with a structural engineer.

    One solution I've used in the SF Bay Area is to cut a series of 2" diameter holes on either side of the ridge, usually 2 on each side, in each rafter bay. This leaves enough of the plywood and the nailing pattern intact for the diaphragm to still function properly.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |