Air Sealing Challenges for Small Buildings
Even the smallest, simplest homes are hard to air seal to Passive House standards.
In 2013 I embarked on my first Passive House (PH) project. Unlike many PH builders, I am neither an architect nor a licensed contractor. I am a longtime owner builder. As such, my perspective is a bit different from most (not necessarily better, just different, and probably less informed that it ought to be). As a longtime HERS Rater and PHIUS+ Rater, I knew that my biggest challenge was probably going to be keeping the building infiltration to no more than 0.6 ACH50. My approach to both designing and constructing the building focused on that goal.
To date, the shell has been completed, windows installed, and roof finished. With a PH net air volume of 1,316 cubic feet, I’m currently testing at 20 CFM50, without local exhaust, or 0.75 ACH50. This is tighter than most forced-air systems, but not quite good enough for PH certification. I’m guessing that at least some of that CFM is from an ill-fitting blower door frame. I’m hoping to track down enough remaining leaks to achieve 0.6 ACH50 before long.
Accessory Dwelling Unit—North-South Section
Figure 2. Once the shell was coated with the air barrier, the only vulnerable location would be the seam between the stem wall and the sill plate where the termite shield is located. I designed that seam to include an EDPM gasket and lots of mastic.
Note: All blower door numbers cited here are for single-point tests at 50 Pa. In my experience, the single-point tests match the multipoint tests in low-wind conditions in my location. However, an official PHIUS (Passive House Institute US) certification requires averaging multipoint pressurization and depressurization tests.
The biggest design challenge I faced was planning restrictions. The building is located in Berkeley, California. It’s what is called an Accessory Dwelling Unit, or ADU. Berkeley has strict size and height requirements. Those requirements constrained the building to a 300 ft2 external footprint, and a 12-foot average roof height. With 9-inch-thick walls, the interior conditioned floor area is only 239 square feet. Because of that small space, I worked hard to find a design that includes some loft space. See the north-south section (the air barrier is shown in red) in Figure 1.
The full height is just less than 16 feet. The average roof height is just less than 12 feet calculated from the wall top plates. I put the interior slab at grade so that I could gain a few more inches of headroom in the loft, resulting in an 8-inch concrete stem wall above grade for termite protection.
Once I had a basic design in place, I wanted to make the shell construction as tight as possible as easily as possible. My first choice for the shell assembly was Structural Insulated Panels (SIPs). They come in large sections, with minimal seams—what could be easier to make airtight? That plan was quickly scuttled once it became clear that my structural engineer, a so-called SIPs specialist, didn’t know anything about my seismic requirements. The panels he chose were not certified for use in my earthquake-prone zone. I found this out after submitting my first plan set to the building department. (In subsequent discussions with other builders, it became clear to me that SIPs aren’t necessarily all that easy to make airtight, because of their prerouted horizontal and vertical chases for wiring and plumbing and less-than-precise fabrication, so I’m glad that plan never materialized.)
My next choice included a new structural engineer, an SIP roof, and 24-inch OC 2 x 6 framing with additional exterior insulation. Window placements and sizes were optimized to minimize framing. I resisted all suggestions from my wife to include bay windows and window seats—any architectural enhancements that would make the shell more complicated. (In retrospect, I think my concerns were unfounded.) At this stage, two key ideas emerged:
- Construct a fully enclosed shell when framing, with minimal seams and penetrations, coat it completely with a liquid-applied air barrier, and test it for leakage as early as possible.
- Make the roof overhangs a separate structure, outside the air barrier (thanks to Allen Gilliland at One Sky Homes for the suggestion).
For the shell, I decided to make the edges of the SIP roof flush with the exterior walls and wrap the underinsulation vapor barrier up over the top of the above-grade concrete stem wall. Once the shell was coated with the air barrier, the only vulnerable location would be the seam between the stem wall and the sill plate where the termite shield is located. I designed that seam to include an EDPM gasket from Conservation Resources and lots of mastic. See Figure 2 for a look at the complete assembly.
My plan was to have the shell framed and sheathed, test the air leakage, and then start making changes in a controlled fashion, with regular blower door testing, so that I could test the effect of minor changes throughout the process. As an owner builder, I can take my time (in fact, probably too much time), a luxury that most construction projects don’t have.
My biggest concerns were the shell, the windows, and the ventilation equipment. I planned to tackle them in that order, one at a time, in such a way that I could (hopefully) determine the individual effect of each item on the air leakage.
I was out of the country on vacation when the slab was poured and the sill plates framed, so I had no control over the quality of that crucial detail. My backup plan, in case the sill sealing wasn’t done very well, was to tape the interior side of the seam (which I ended up doing).
My young framers weren’t the most experienced builders, but they were enthusiastic about working on a PH. The day the framing was done was a momentous occasion. The “tank,” as the lead carpenter called it, had only one opening—the front door.
None of the crew had ever seen a blower door test. Just looking at the framing from the dark, windowless interior, you could see seams with visible light shining through. The initial ACH50 number was 425 CFM50, about 25 times the 16 CFM50 target based on the volume as defined by PH criteria. The lead carpenter suggested that we pressurize the building and insert spray foam into all the visible cracks. Twenty minutes later we were down to 113 CFM50. That was encouraging progress.
The next step was to cover the exterior holes and seams with Prosoco’s Joint & Seam filler, and the whole exterior surface with Prosoco’s R-Guard, a water-resistive air barrier. The exterior sheathing is ½-inch CDX plywood. It contains various imperfections in and between layers that could add to the air pathways created by the framing process. (In retrospect, I should have used ACX plywood instead for a much higher-quality exterior surface). After we slimed the exterior with Prosoco’s products, the CFM50 dropped from 113 to 51.
To track down the remaining leaks, I tried using both an infrared (IR) camera with the blower door depressurizing the house, and a smoke machine with the blower door pressurizing the house. Neither technique worked well. The IR camera was useless, because there was no significant temperature difference between inside and outside the building at the time of the tests, despite running a heater for 12 hours the previous night. The smoke machine was equally useless, although a little bit more informative. As soon as a wisp of smoke escaped the building shell, it would dissipate. We did find a few seam issues that way, but nothing significant. I then tried the following steps:
- Foaming under the exterior edge of the termite shield where there were some imperfections. No improvement.
- Closing up an overlooked ⅛-inch gap around the incoming water supply. The CFM50 dropped to 45, demonstrating that a gap that small contributes significantly to air leakage.
- Installing backer rod at the interior of the sill plates. The CFM50 dropped to 40. (I later replaced the backer rod with SIGA tape.)
- Caulking the SIP screw penetrations in the roof. No improvement.
- Taping and foaming the edges of the door opening to minimize leakage around the blower door frame. Slight change.
Totally out of ideas, with dreams of PH certification fading fast, I desperately checked the interior slab penetrations once again. I discovered that the framers, who claimed they had sealed all the slab penetrations before the first blower door test, had missed the wall-mounted toilet waste line. I sealed that up. The ACH50 dropped to 20. Lesson learned—a 3-inch-diameter hole at 50 Pa is worth about 20 CFM50.
- At this stage, I had learned the following things: The framers had done a pretty good job on the sill plate connection. Perhaps the connection design was hard to mess up. Maybe it was the combination of a simple connection detail and good workmanship.
- Double- and even triple-check the planned penetrations. Things can change on a construction site from day to day, accidentally and inadvertently.
- Smoke and IR have limited usefulness. It depends on the test conditions.
- At these low leakage levels, the test equipment can be a significant source of errors. I suspect that less-than-perfect sealing around the blower door frame is contributing at least 2–3 CFM50 to the leakage numbers—potentially the difference between passing and failing. I have no easy way of proving that.
Now that I had a relatively airtight shell, it was time to install the windows. I had never really thought much about the airtightness of windows before working as a PHIUS+ rater. I’ve come to realize that they can make all the difference between greater or less than 0.6 ACH50. I decided to use Alpen dual-pane 725 fixed and awning windows (five fixed, six operable) for several reasons:
- The Alpen windows are NFRC-rated, so I can take advantage of California’s utility incentives.
- They are less expensive than PH-certified windows imported from Europe.
- They are durable, thermally stable fiberglass, which I like (except for the nasty manufacturing side effects).
- Their U-factor and SHGC specifications work well for a PH.
I chose awning instead of casement windows because the Alpen awning windows have two locking points instead of one, which I hoped would make them more airtight.
Installation was straightforward. I cut the openings, sealed them with Prosoco FastFlash, installed anchor straps on the frame interiors according to Alpen’s instructions, wrapped the full frame with Hannoband expanding foam tape, and installed them. The Hannoband tape is a miraculous product. Delivered in rolls, it’s about ⅛ inch thick. Once you unroll it, it slowly expands to as much as ¾ inch, depending on which version you use, over a period of about an hour. It’s supposedly a vapor-permeable air barrier, designed to seal cracks and seams, perfect for window installation.
After installation, my next blower door test came in at 95 CFM50. The Hannoband tape worked pretty well as an air barrier, but it wasn’t perfect. In addition, it was impossible to get it perfectly installed so that all window-edge cracks were sealed. The next step involved sealing all the exterior window edges with Prosoco Joint & Seam Filler.
The CFM50 dropped 39 CFM. Good, but not good enough. I then covered all the interior window seams with more Joint & Seal Filler. The final test came in at 20 CFM, indicating that the Alpen windows have no significant leakage. I plan on retesting the infiltration on this project at least a year after it’s completed. The windows have exterior gaskets. This may be one way to see if the seals degrade over time as the windows are repeatedly opened and closed.
After dialing in the shell and the windows, my next-biggest concern was the ventilation. A PH in California has to comply with two ventilation standards: Passive House and ASHRAE 62.2. In a larger house, it’s often possible to combine the two and satisfy both requirements with one well-designed heat recovery ventilator (HRV) system. In my small house, with a volume of only 2,260 cubic feet, those two standards present some interesting complications.
ASHRAE 62.2 requires 7.5 x (# bedrooms + 1) + .001 * CFA (18 CFM) continuous ventilation. PHPP (Passive House Planning Package) recommends 35 CFM per kitchen and 25 CFM per bathroom (60 CFM) using an HRV. PHPP also recommends an average ACH of about 0.35. Sixty CFM is almost three times that rate. Looking at various HRV alternatives, a standard ducted HRV seemed like serious overkill from both a cost and a complexity perspective, plus I just didn’t have the space for the unit. I decided to install a pair of Lunos e2 units to provide whole-house ventilation that would meet both ASHRAE 62.2 and PH requirements at an ACH50 rate of 0.31 running at 22 CFM50. Before doing that, however, I wanted to make sure I could meet the ASHRAE local exhaust requirements without jeopardizing my blower door numbers.
ASHRAE 62.2 requires local exhaust for the kitchen and the bath of 100 CFM and 50 CFM respectively, or continuous rates of 5 ACH and 20 CFM. I couldn’t use the Lunos to satisfy either of those requirements—they switch from supply to exhaust mode approximately every 75 seconds. I would need a pair in both the kitchen and the bathroom to guarantee continuous ventilation. Plus I wanted to make sure I had adequate kitchen local exhaust to the outdoors to mitigate cooking odors.
My initial plan was to install an Energy Star kitchen range hood, vented to the outdoors, and an Energy Star 50 CFM bath fan, also vented to the outdoors. This introduced two additional factors:
- Typical residential units have very leaky backdraft dampers, which would totally destroy my PH ACH50 test results (local exhaust openings must be in their natural operating condition, unsealed, when testing).
- With the tight construction and small space, supply air should be added to the structure whenever the locale exhaust units are turned on.
To solve both these problems, I decided to attach motor-activated backdraft dampers to the local exhaust outlets and add separate fresh-air inlets that also have inline motorized backdraft dampers. The success of this strategy depends entirely on how well the motorized backdraft dampers seal
Before buying all this ventilation equipment, I devised an experiment designed to find out just how leaky the motorized backdraft dampers were. Assuming I’d have one on both the supply and the exhaust side of the local exhaust equipment, I purchased a pair of 4-inch dampers. I then installed them directly on the exterior and sealed the perimeter of the openings with SIGA tape.
I then sealed the exterior duct openings with blue painter’s tape and ran my blower door tests again. The results were identical to my CFM50 numbers before I installed the dampers (19 CFM50 and 20 CFM50), indicating that the installation was tight.
I then uncapped the duct and ran the blower door tests again. The results were not unexpected: 24 CFM50 and 27 CFM50. Essentially, each 4-inch damper added about 2.5 CFM50 to the air leakage. With another set of 6-inch dampers for the range hood, I expect at least another 5 CFM50 leakage.
That’s the current status of the project. I’m still exploring ways to reduce the infiltration down to acceptable levels. Regarding the ventilation, I’m reconsidering using a ducted HRV, although I’m resisting the idea because of the sheer size of the equipment and the cost. I’m also investigating higher-quality dampers, techniques for sealing the dampers themselves, and combining the two supply dampers into one. I’m also going to spend some more time trying to track down another 5 CFM50 of building leakage. Although it’s a small building, when you’re looking for what amounts to a series of pinholes, it can feel like a very large building. I’ll be applying a coat of Prosoco Cat 5
after finishing the window flashing. That may make a difference. I’ll pull out the IR camera and smoke machine again. If all that fails, at least I will have conducted an interesting series of mystery science experiments.
Steve Mann, bitten by the PH bug a few years ago, is now building his first (hopefully) certified PH home in Berkeley, California. He is a certified Passive House Consultant and Builder, LEED AP+ Homes, both a HERS and a PHIUS+ Rater, and past president and current board member of the California Association of Building Energy Consultants. Contact Steve at firstname.lastname@example.org.
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.