How We Turned Our House Into a Giant Foam Box
My wife Wen Lee and I are in the midst of renovating our rather generic three-bed, two-bath, 1,400 ft2 suburban Pasadena area home to be net zero energy. At the outset I decided it would be a good and reasonable idea to do most of the work myself, not realizing how presumptuous and arrogant it was to think that I could quickly learn multiple design and building trades and do competent work. So inevitably here we are now two years into a one-year-long project. Things are progressing, albeit s-l-o-w-l-y.
The main bulk of the work has been in what we call the common area, which includes the kitchen, dining area, and living room. After vaulting the ceiling (and changing the roof structure), the next heavy lift was superinsulating and air sealing the walls and the roof/ceiling assembly.
I chose to insulate the walls and ceiling by installing 3 inches of polyisocyanurate rigid foam (polyiso for short) in between the structural members. This method is often derisively referred to as “cut-and-cobble,” and it is generally regarded by professionals as residing squarely in the realm of do-it-yourself homeowner strategies. Because cut-and-cobble is so labor intensive, it cannot be considered cost-effective by any traditional accounting method. My experience will do nothing to discredit this perception. (For more, see “Establishing Priorities.”)
The whys of the project—net zero energy, insourcing versus outsourcing labor, cost-effectiveness, and so on—are probably more important to us than the nitty-gritty hows. These priorities led us to do some unconventional, even inefficient, things. For example, the reasonable way to insulate the wall would be to remove the stucco and install sheathing and insulation on the outside of the house, rather than do everything from the inside the way we did it. Working from the outside would have had many benefits, including not having to cut and cobble the rigid foam, keeping the studs warm in the winter, and not losing any interior square footage. But it had one major downside from our perspective—namely, it doesn’t lend itself to being insourced. In order to keep costs down and most importantly, to take advantage of as many learning opportunities as possible, we wanted to do as much work ourselves as we could. It’s not a good idea to expose the exterior of a house for long periods of time, so the work would have had to be done quickly, and that would have meant lots of hired help. The stucco would almost certainly need to be contracted out—another expense, and another learning opportunity lost. Doing it from the inside took much longer, and it took an extreme amount of labor on our part. But we willingly chose to make this trade-off, at least for this learning project.
For more on the background and goals of the project, see “Establishing Priorities at the Outset.”
Our house was built in 1963, and the only insulation was 5 inches of blown cellulose on the floor of the vented attic. The walls were uninsulated, as was the subfloor above the vented crawl space. Figures 1a and 1b show the wall assembly before and after the renovation.
There is no sheathing on the exterior of the house—the studs touch building paper and then stucco. There were occasional diagonal braces set into the outboard side of the studs, but there wasn’t nearly enough lateral bracing. (Which keeps your nice rectangular wall from turning into a nasty parallelogram in the event of an earthquake.) This lack of sheathing creates two problems. First, there is no drainage plane to allow any moisture getting through the stucco to escape or to dry out. Second, the walls don’t meet today’s seismic code.
To address the first problem, I inserted ½-inch wood sticks to create an air gap in between the building paper and the 3-inch rigid foam. This should give any moisture that gets into the wall assembly a chance to dry out, rather than getting trapped and rotting the studs. Admittedly, ½ inch is not a lot, but I think it should suffice in our very dry climate. To address the second problem, after putting 3 inches of rigid foam in the stud bays, I installed plywood across the inboard side of the studs.
Fitting the rigid foam into the stud bays was no easy task. This is where the labor intensity part starts to kick in. The stud bays were generally spaced roughly 16 inches on center (OC), but there was some variation. And the studs were generally more or less parallel, but some deviated significantly. In practice this meant that each piece of rigid foam had to be custom-measured and cut for each stud bay.
After the stud bays were insulated, I air sealed with acoustical sealant caulk and one part canned spray foam.
After weeks of work, the walls were insulated and air sealed. To address the lack of shear strength due to the absence of external sheathing, my brother joined us in our project and we installed plywood across the inboard side of the studs. Earlier, per the structural engineer, brackets and extra studs were installed at regular intervals to help support the shear walls and to tie the house to the foundation. Adding the plywood would complete the shear wall. Because it was continuous, the plywood installation went much more quickly than the cut-and-cobble rigid-foam installation.
After the plywood was finished, the walls were insulated, air sealed, and structurally sound. However, heat could still travel through the uninsulated parts of the wall—the studs—and through the plywood. This circumvention of insulation is called thermal bridging.
To prevent this thermal bridging, inboard of the plywood I installed 1-inch continuous polyiso rigid foam.
Like the plywood, these foam sheets could be installed continuously, so it went relatively quickly.
Just for good measure, I air sealed the seams of the 1-inch foam.
Finally, on top of the 1-inch foam, I put 2 x 3 members horizontally, spaced 2 feet apart. The purpose of these 2 x 3 furring strips is to create a place to run utility lines—wiring, water supply lines, communication wires, and so forth—in the exterior walls, without compromising the performance of the wall. This service cavity keeps every part of the exterior wall fully insulated and air sealed while still hiding the utility lines behind the wall’s finish surface. Service cavities also make installing the plumbing and electrical wiring easier because you don’t have to constantly drill holes through studs and snake tubing through them. Service cavities add a little bit more carpentry labor and they make your wall thicker, which slightly reduces your interior area.
Once all the utilities are run, I’ll attach drywall to serve as the finish surface.
So that’s the wall assembly.
The ceiling was similar but even thicker and more complicated (see Figure 2a).
The original roof assembly was fairly typical: 2 x 6 rafters spaced 24 inches OC, covered with 1 x 6 skip sheathing, then continuous sheathing, building paper, and asphalt composite shingles. The attic below was 2 x 6 ceiling joists spaced 16 inches OC, with R-19 loose-fill cellulose that was installed in 1983.
We changed the roof structure in order to vault the ceiling. This converted the ceiling and the roof into a single assembly. To carry the now-unsupported 15-foot span, 2 x 12 rafters were sistered to the existing 2 x 6 rafters. This added depth allowed us to install a lot more insulation (see Figure 2b).
This 2-inch gap is intended to vent the underside of the roof assembly, both to keep it cooler in the summer and to allow any moisture that gets into the assembly to dry. One-inch-thick strips of corrugated plastic—like many straws bundled together—are tacked to the top of the blocking for each rafter bay. These openings serve as an inlet for cool air at the eave. Another vent opening is made at the ridge to allow the hot air from the assembly to escape. The air in the vent space will heat up—buoyant hot air exiting at the top, cooler air entering from the bottom to replace it—creating a convective loop that will passively cool the underside of the roof deck.
The rigid foam had to be cut for each rafter bay, but the rafters were more uniformly spaced than the studs, and much more likely to be parallel. This uniformity meant we could mass-produce the foam strips, which made the job go much more quickly than it did with the wall insulation. As with the walls, I air sealed the seams between the rigid-foam pieces and where the rigid foam touched the rafters.
Inboard of the 3-inch polyiso, I installed kraft-faced R-19 fiberglass batts, holding them in place awkwardly and stapling the flaps to the medial sides of the rafters.
To prevent thermal bridging, I then added 1-inch polyiso continuous insulation across the new 2 x 12 rafters. These sheets had the added benefit of supporting the fiberglass batts.
For more on the background and goals of the project, see “Establishing Priorities at the Outset."
Just as I did with the wall, I used 2 x 3 furring strips to create a service cavity for the ceiling.
Once the utilities are installed, the finish surface for the ceiling will be attached to the furring strips. It’s not going to be drywall, though. That would be too simple.
The final value for the wall assembly is roughly R-26, and the ceiling/roof is roughly R-46. The performance of the new building envelope cannot be assessed yet, but needless to say, I expect good things. Anyone contemplating this degree of intervention in their own home should be under no delusions as to just how much labor is involved. But I can attest that the satisfaction makes it worthwhile.
Enter your comments in the box below:
(Please note that all comments are subject to review prior to posting.)
While we will do our best to monitor all comments and blog posts for accuracy and relevancy, Home Energy is not responsible for content posted by our readers or third parties. Home Energy reserves the right to edit or remove comments or blog posts that do not meet our community guidelines.