How was that possible?

A couple of years back, after all major air sealing, we had a blower door test result of 2.1 ACH @ 50 pascal. This February the results dropped to 0.62 ACH @ 50 pascal in a follow up test! How was that possible?

From what I have read and heard, once the major air sealing is completed, it becomes increasingly difficult to tease out additional improvements in the blower door test results. And a drop from 2.1 to 0.62 ACH @ 50 pascal is – well – a massive improvement.

Let’s start by looking at what I mean by…

Major air sealing

In one sentence: We took good care of the building envelope.

On the 1st floor we had an application of closed cell foam followed by a layer of open cell foam. The 2nd floor just got one layer of closed cell foam. The closed cell foam on both floors was acting as our air barrier across the masonry wall plane.

The closed cell foam only provides a functioning air barrier if it is diligently installed. On a sloppy spray job, you may have to contend with leaks in your air barrier. I point to some examples here: on the open cell job on our 1st floor.

We air sealed the top of the attic with drywall, which was taped and mudded. The gaps around the attic edges were sealed with another bead of foam.

All penetrations through the building envelope were diligently sealed, typically with foam. Examples are:

  • The ERV fresh air intake and exhaust
  • The range hood exhaust
  • Any electrical conduits leading to the exterior
  • Any low voltage conduits leading to the exterior
  • The supply and return lines to our minisplits
  • Etc.

You can seal around electrical conduits, but that still leaves with a big hole – the conduit itself. To plug the conduit, we diligently used duct putty.

I made sure we had decent weather stripping on our exterior doors. And then there were the windows. We had good quality replacement windows on the 1st floor, while the 2nd floor still had the old double hung vinyl windows. However, I made sure the perimeter of each window was properly sealed.

This was our baseline that gave us the blower door test result of 2.1 ACH @ 50 pascal. Now let’s take a look at what may have caused the drop to 0.62 ACH @ 50 pascal.

The 1st floor

I took a mental walk through all improvements on the 1st floor, major or minor, since we completed the major air sealing.

There was some additional air sealing on the kitchen back door. This is a fancy way of saying, “I installed some additional weather stripping.” The back door was pretty air tight to begin with, though, so I am not sure that this effort contributed that much.

There was the transom window over the kitchen back door. It had a temporary window that most likely was not properly air sealed. I didn’t replace it with a properly sized window until I ordered the replacement windows for the 2nd floor last fall. This should have contributed to the reduction in air leakage.


I suspect that the biggest impact lies with the air sealing work I did on the 1st floor casement windows. Regular readers may recall my air leakage problems in the corners of the casements. The manufacturer was not able to resolve the issue, but was gracious enough to refund me the money for the leaky windows.

As described in a past post, the problem came down to 1/16 inch and I was determined to plug that gap. I invested a little research time online and found a single coated, low density, PVC foam tape, 1/16 inch thick. I installed it on the operable part of the casement, right across from the gaskets on the casement frame. The foam tape pushed against the gaskets, closed the gaps, and eliminated the air leaks. I could tell it did, because the cold drafts on windy winter days disappeared.



2nd floor

Like on the 1st floor, I did some additional air sealing around the 2nd floor back door. The most significant improvement may have come from my work on the door threshold, which was not properly sealed.

But the biggest reduction in air leaks must have come from the 2nd floor replacement windows, which I installed last fall. I thought I had done a good job air sealing the old double hung windows after we finished with the spray foam installation, but the blower door test numbers tell me otherwise.


I clearly underestimated the amount of air leakage from the 1st floor casement windows and the effect of good quality replacement windows on the 2nd floor. Or, I overestimated the effect of my air sealing efforts on those old double hung windows.

The more I think about it, the more I am convinced that the windows were the major contributors to the drop from 2.1 to 0.62 ACH @ 50 pascal.

So, maybe it is not that difficult to improve your blower door results after all major air sealing after all! Or maybe I have a unique definition of what “major air sealing” means.

Related posts:

A magic 0.6 – as airtight as it gets

Reaching a major milestone is a reason to celebrate. But what do you do when you exceed all expectations and it takes you by surprise?

Old buildings like ours are notoriously leaky. Even a lot of newer buildings are. That’s something you will notice when you try to escape the cold drafts in your own home on a windy winter day. We know how leaky our building was before we started with the deep energy retrofit, because we conducted a blower door test on the 2nd floor unit. It leaked a whopping 4,763 cfm @ 50 pascal.

The blower door basically depressurizes the building or unit by 50 pascal. At that point you take a reading of the airflow. In our case, the 4,763 cfm @ 50 pascal is the air volume that is leaking through the building or unit. To get to a metric that has a common denominator, the airflow is converted into air exchanges per hour or ACH @ 50 pascal. We ended up with 13.9 ACH @ 50 pascal.

Our results were not that unusual. Blower door test results often range from 7 – 15 ACH @ 50 pascal, depending on the building.

What are the air leakage standards?

Air leakage standards in the U.S. are a moving target. There is no national standard; instead, a builder or homeowner can pick from a list of requirements. Here are some examples:

The PHIUS standard discontinued to use the metric of air exchanges per hour (ACH), and instead replaced it with cubic feet per minute (cfm) of air infiltration per square foot of gross building envelope area, measured at 50 and 75 pascal. The rationale behind this metric is that it can be scaled, unlike ACH. Because ACH is based on the building volume, it allows larger buildings to have a greater rate of air leakage compared to smaller buildings.

How did we do?

We ran a blower door test on the 2nd floor unit after we finished with the spray foam insulation and major air sealing. The results were 720 cfm @ 50 pascal or 2.1 ACH @ 50 pascal. That was a remarkable improvement over the 13.9 ACH @ 50 pascal and caused some excitement.

About a year later we conducted a blower door test on the finished 1st floor unit during a project tour organized by the Chicago Community Loan Fund.That test gave us a reading of 630 cfm @ 50 pascal which also equated 2.1 ACH @ 50 pascal.


I have learned over the past years through reading and talking with green builders that once the major air sealing is done, shaving the last few 1/10th off an ACH @ 50 pascal is painstaking work. It involves a lot of chasing, tracking and sealing small air leaks.

All our major air sealing work was completed. My hope was that I would be able to get a final result of just under 2 ACH @ 50 pascal by plugging the last few leaks I could find. That would have made me very happy!

In for a major surprise…

This past week, I hosted another tour of our deep energy retrofit. It was organized by Eco Achievers. Part of the tour was another blower door test demonstration. This time we tested the 1st and 2nd floor unit together.


The reading of 385 cfm @ 50 pascal seemed unreal. So much so that I asked to start the test over again including another recalibration of the blower door equipment. Yet we got the same reading for a second time.

385 cfm @ 50 pascal for the 1st and 2nd floor unit equates 0.62 ACH @ 50 pascal! If I use the PHIUS metric, I come up with 0.05 cfm50 per square foot of gross envelope area.

Do I need to say that it took a while for this to sink in? Never in my wildest dreams would I have thought a less than or equal to 1 ACH @ 50 pascals would be possible on our deep energy retrofit. I guess it’s time to celebrate! And I have to find out how I managed to drop from 2.1 to 0.62 ACH @ 50 pascal.

Related posts:

Fussing over fence details

My nicely aligned posts alone won’t make a fence. I had to make up my mind about the fence panels, so let’s get back to basics for a minute:

Using metal in the fence panels is a risky proposition as it might get snatched by scavengers overnight. Instead, we opted to use pressure treated lumber. We installed the four by four fence posts, and I planned to use two by fours for the fence rails.

The rails alone won’t suffice. I needed a somewhat solid fence panel to keep trash from blowing into the rain garden vegetation. I considered a whole gamut of ideas, but let me make this short.

The concept of woven fence panels persisted. Woven, like wicker furniture or an old fashioned willow basket. This would add a level of surprise or contrast, as it would not be expected in an urban context. Yet I’d have to make it sufficiently robust to persist through the urban pressures.

And this is where I had to rely on metal after all: half inch or number four rebar. While I still wanted to use a wooden material like willow for the horizontal weft, the vertical warp had to be rebar.

But how could I prevent the rebar from growing legs at night? By slipping it through a pre-drilled top rail and into a three quarter inch hole in the bottom rail. A handrail attached to the top rail would lock the rebar in place.

One problem was solved, but another one was created. I ended up with two very different fences in close proximity: the woven panel knee fence around the parkway and our typical Chicago style black metal yard fence.

Patchworks of different style can generate something visually stimulating. But in this case, creating a connection between these two different enclosures and as such weaving the parkway landscape into the remaining landscape on our property would be more inviting. Rather than passing a semi-public landscape on the parkway side and a private landscape behind the property fence, we would prefer to invite observers to pass through an extension of our private landscape, which reaches all the way to the curb.

How could I begin to weave those two landscape together? Literally by weaving. I could use the same willow material that I plan on using in the knee fence panels, and weave a solid panel into the bottom our our yard fence. The added benefit would be that even more blowing trash close to the ground will be blocked by semi-solid paneling and kept out of our plantings.

Related posts:

Pounding parkway paver

We set the knee fence posts along the street a good 12 inches back from the curb. That gave us enough room to install an paver strip that should measure at least 18 inches from the front of the curb to the face of the fence.

This would be enough room for passengers in parked cars on the street to open the door and step out. The fence in turn would prevent passengers from stepping into the parkway landscape or their car doors from swinging into the parkway vegetation.

We will transform the parkway into a rain garden. That means the elevation of the rain garden will be around six inches lower than the current elevation. To help with the transition between the two different elevations, we added a small hard edge along the east and west end, which serves as a curb for the rain garden.

A lot of things can go wrong in an urban environment. Someone could end up driving into the parkway. There is little I can do about that. But I can prepare for other things. For instance, what if someone drives onto the paver strip?

To prepare for this eventuality, we put down a solid six inch aggregate base with recycled material. On top went a setting layer of coarse sand into which I laid the pavers.


The pavers themselves need to have a certain mass to be cut out for the job. Some of the salvaged limestone pieces that I had stored in the yard were perfect for the job. At 12 inches wide and four inches thick they had the weight and soundness I was looking for.


I wouldn’t be surprised if the paver strip ends up being better built than the adjacent road itself.

Related posts:

Getting my posts in a row

Setting fence posts starts with the corners. Once the corners are set, get a mason line and a level. That makes it fairly straightforward to get all other posts perfectly lined up, plumb and at the same elevation.

When a post was positioned, we braced it and temporarily connected it with two by fours to the adjacent posts so that it stayed in place. After a full line of posts was set, we carefully poured the concrete footing around each of the post bases. We knew that if we didn’t do it carefully, we might have moved post base and would have had to start re-aligning. And that is a time suck, believe me!

There is a simple trick on how to make pressure treated posts last. Don’t set them on concrete, set them in concrete.

If you set them on concrete, water has difficulty getting out and the post base becomes soggy which accelerates rot, even if they are pressure treated. If you make sure that the bottom one inch is sitting in soil – or even better, on pea gravel – the post base has a chance to drain and dry out, which should make it last longer.

Related posts: