House energy retrofit project 11
[Click any image for a larger version.]
With the sheathing and sill area clear all the way around, the guys
wanted to try and get the metal base flashing continuous around the house.
It was easy enough to send through the basement-bulkhead gap with only
a little creative bending.
A larger impediment was the side stair unit, still too close to the wall.
They declined my offer of the truck jack I already knew could lift the thing
in favor of just getting a bunch of guys to huff away at it with long
2x4s. A nice little challenge? "Give me a
and I shall move the world!" They broke a couple of boards in the process
but managed to inch both ends of the stoop far enough out to work behind.
Here was some actual rot needing replacement wood, not to mention
minor structural fixes. The stoop had sat right up against the old siding,
creating a channel against the house that would trap and hold water.
The ants, of course, loved this especially with a nice shady and damp
area around it under the stairwell, otherwise only ever used by a
passing chipmunk or two.
The dead wood and the door itself got pulled, exposing the whole area.
We realized that there had been *nothing* helping hold the old door up --
it had been delicately balanced across the kitchen subflooring and the
rotting sheathing. No wonder the metal sill had gotten to feeling
squishy whenever I stepped on it; it's amazing it never collapsed
into the hole under me.
The dark void at the lower right is that same convoluted return duct from
the kitchen, that had the weird flopped-over dust pack in it. Here I
could see the other side of how the returns were constructed, simply
using a stud bay void to bring air down around the edge of the subflooring
to the basement. Not exactly what you'd call "ducts in conditioned
space" as cold air from cracks in the original wall's sheathing would merrily
join the flow. No biggie now, this entire mess was about to become
In theory I could have tried to clean that duct out more from here but I
didn't want to get in the way.
The rest of the bad wood got torn off and new framing put down for the door,
which in this case would be a new one. More borate was applied in the
process. Even if some ants were left inside the structure, once everything
was sealed up they'd die because they wouldn't be able to get *out*
to find food.
Again, the mudsill itself was rock-solid despite the years of moisture
and neglect. The framing and remaining sheathing looked a little ratty
but was still solid enough to not bother worrying about.
The sheathing got filled back in with new plywood, and then the flashing
could go through on top of that.
The foundation-to-sill connections are weird in this house,
with the outer sheathing
surface a little shy of the block wall below on the front
and back of the house, but sitting rather *proud* of the wall along the
east and west gable ends. This left a bit of sill and sheathing still
exposed under the flashing metal, and we would have to eventually come
up with a fix to seal this up. I had tried to do a little spray-foam
work under the siding along here a few years ago, the tattered remnants
of which were still visible.
The related oddity is that the visible end joists in the basement aren't
right at the outer wall, there's a stud-width of inaccessible void behind
them and the lowest sheathing here at the outside, which could make
air-sealing along both those walls a little interesting. Here I was
primarily concerned with insect protection.
Foaming and window installation proceeded along the east wall, ignoring the
door for the time being.
It was almost like a race, with two more guys working the west wall the
same way, and whoever was on the ground frantically trying to prep
and pass them materials.
The west-side guys had one more window to install, but they gave it the
same attention to detail as the rest. Here they're in the computer room
prepping another unit, before passing it out through the very opening
it was about to get installed in.
As they worked up to the second story, supplying materials began to involve
more of that dynamic aerial art as they tossed stuff up -- such as whole
assembled buck pieces.
They'd swap between indoors and out so smoothly I hardly noticed -- of
course both first-floor doors were hanging wide open pretty much all day,
so there was no ceremony in them just coming in and wandering upstairs
to help shim and line up another window. I had done what I could to
assure them that they absolutely shouldn't worry about traipsing in
and out of my house to do what they needed to do, other than occasionally
winding up on candid camera.
A little vignette: tools of the trade when you're doing rigid foam. Caulk,
short and long screws with buttons ready to pop into the panels, Dow tape.
They were going through caulk
fast enough that they weren't worrying about trying to close off tube
nozzles between applications, and some of their guns were the type that
leave a little remaining pressure on the plunger after the trigger is
released. My own cheapo caulk gun had been one of those too, but it
eventually broke entirely and I went and got one of the expensive ones
that actually works right. As geeky as it sounds, that was one happy day.
Foaming had proceeded along to the side-door space, which was still open
because we didn't have said side door to install yet. Here's a nice look
at the new wall cross-section, though: Kitchen subfloor and interior
drywall, old and new
rough framing, old sheathing and new plywood to match, one 2-inch polyiso
layer, and finally the buck assembly. And at this location, all arranged
with a slight downhill slope as the door sill would eventually sit here.
Outside of that would go strapping and siding, for a total wall thickness
close to ten inches.
An impressive pile of foam cut-offs had built up in the easel area, but
they were trying to use small pieces where they could for minor fill-in.
I went around to check out the new windows a little, and discovered that
a couple of the latches felt loose. Not only that, the screws holding
them down turned out to be completely stripped in the fiberglass. Okay,
now what? I probably could have have come up with a fix but decided that
since these were brandy-new, this kind of quality hiccup should have never
left their shop and I called Serious for suggestions. More on this later.
Bottom line was that I simply tested them all gently with a screwdriver
and blue-taped all the ones that were stripped.
With everything so opened up all day, this was about the lowest I'd seen
the CO2 meter reading. It suposedly uses readings below our 400 ppm
"hockey stick baseline" to recalibrate its own low threshold automatically,
and to reach that it basically has to be in plenty of fresh air which
it certainly was this particular day.
The temp/humidity between the expensive CO2 meter and
the cheapo T/H thing from Amazon has never quite matched, but all these
measurement widgets I've accumulated are well enough in the same
ballpark that I'm not too worried about it.
Today had built up another collection of dead house parts. A couple of
the old windows had sort of fallen apart in the removal process, but none
of the glass had broken yet even when they were pitched into the dumpster.
They landed fairly flat and I guess the glass was well enough supported in
the muntins that it wasn't unduly stressed. I also rescued my Simplex
and latch parts off the old door, and even swiped the hinges off the
frame because ya never know when a couple of door hinges will be useful.
The old door was a ThermaTru, filled with closed-cell foam inside, and
actually had fairly decent thermal resistance and a good magnetic
weatherstrip. I had even painted it at some point. But it was rusting
out at the bottom, and the wood frame was basically toast. Time to
let that one go.
Very close examination of one of the old windowsills revealed that it
wasn't a single piece of wood, but an assembly finger-jointed together
so precisely it was almost impossible to discern. It's barely visible
in the big picture and that's even after a little attempted
contrast-boost in post.
Foaming had gotten just around the rear corner, and it was time to once
again do the roof-angle saw trimming.
We can't see it here due to his butt in the way, but the lighting
wire over this door had been accomodated and brought out through the
foam the same way as out front.
Corner seams in the foam were lapped in a way to prevent any straight air
channels to the outside, and each layer got a strip of flashing to seal
its seam against infiltration. See the big picture for how how I took
Note carefully that many of the foam boards are obviously not dead-flat;
their surfaces often have a little height variation. This could leave a
few small air channels between layers and what they're up against. I
would say that alone is reason enough to thoroughly tape all layers,
using notes from the field here to mildly disagree with what Joe
Lstiburek has recently said on the subject.
With the windows all installed east and west, the guys were getting near
the roof peak. Note how the foam is exactly flush with the roof sheathing,
continuing the plane of the wing-wall roofline.
A last triangle went onto the peak, and the Grace got flopped back over
everything. There was still a little work left to do here so they didn't
seal it down.
The new door they thought would show up didn't arrive that day, and they
had to do something about the opening. To their query I assured them I was
fine with using the front door for however long they needed me to, and they
simply boarded up the side and ran some tape over the edges to make it
nominally waterproof. They expected to get a door any day and were
kind of surprised it hadn't been delivered that morning.
As it turned out, I didn't have a side door for about another week. They
taped an extra piece of housewrap over the plywood the next day for a
little more water-protection in the meantime.
After one of their longest workdays on the project
the guys finally buttoned up and got rolling for home, and
I turned my attention back to the front door a little. I realized that
as careful as they'd been with the flashing, water streaming straight
down the front face of the house would probably hook underneath and
cling its way in toward the top of the doorframe. To try and head this
off, especially as rain was expected that night, I ran a strip of gaff
tape along the lower corner to serve as sort of a drip edge. The house
now had absolutely no overhangs, but I figured the fellas knew what they
were doing and had left things nominally able to deal with weather,
thought nothing more about it, and headed off to my class that evening.
It rained hard in the interim, and I returned home to a bit of a disaster.
The supposedly Graced-up roof *was* leaking, exactly through those holes
I had not only circled *and* mentioned to the crew, and had been assured
that they were fine and wouldn't let water in. Wrong: here it was.
Over near one end of the attic where there were two screw-holes close
together at the roof edge, enough water was
coming in to build up and do that typical run-down-the-rafter thing
like I would occasionally see from some of the old leaks I'd battled
in the past.
I THOUGHT I WAS FUCKING DONE WITH THIS.
It wasn't just the attic, either. Downstairs, in that very first example
opening, water was seeping in somewhere around the sheathing from above
and *dripping* far in enough in to miss the backdam pan below
and land on more framing. Had something gone wrong with all that exacting
flashing work?? I was really hoping there was some other cause and all
that wouldn't have to be torn apart and redone.
Even the front door wasn't exempt, as some of that pan flashing
under it had suffered a couple of rips and I found that a little water
was visible from the basement, dribbling down through there onto the
mudsill. My little tape drip-edge probably
prevented a total deluge onto the living room floor, but the whole bottom
of the doorframe was completely soaked. At that point I put down the
camera and grabbed the absorbent rags.
I was *so blindingly pissed*. Leaks around the shed dormer with its
decrepit shingles were one thing,
but I had *never* had any leak issues below second-floor level before.
This was about the closest I came to freaking out about this project, but
at the same time the other part of my brain was calm and confident that all
of this would be worked out and diagnosed and fixed. That didn't do me
any good right *now*, though, in the middle of the goddamn night with water
coming in a bunch of places. Fortunately, the rain had slacked off and
wasn't predicted to be heavy for the rest of the night. And technically,
if the builders had screwed up somehow they'd ultimately be on the hook
for making it right.
I managed to avoid ripping out a, shall we say, very strongly-worded
email message to the Synergy guys, but did send something asking
about our future roofing-material planning and mentioned the leaks as
Morning was still quite damp, but at least it was clearing a little. I
greeted the PM as he got out of his truck with the leak news
and proceeded to go through everything I'd found. He was
totally baffled as to why any of it was happening, other than that
assumption that *removed-screw* penetrations in Grace were self-sealing
maybe wasn't universally
true after all. The front door stuff was pretty obvious,
as the flashing around it had definitely sustained a little damage and
there wasn't much above yet to prevent water from flowing right down the
door and frame.
He assured me he'd figure it all out and make good on it before end of day.
What he didn't actually do was poke his head up in the attic and look at
the locations of those holes I was trying to tell him about.
They resumed work on the roof edge lines they'd reached the previous day,
and wanted to do a more convoluted air-seal in that area. They
cut back the foam layers to different levels and did the same sort of
staggered flashing thing as on the vertical corners, except that here they
had to cut some fairly thin patch pieces to interlace with the flashing.
And of course they didn't have the roof layers on yet to really tie this
It looked a little complex from the ground, and I wasn't really following
what they were doing here. I had already delivered my little rant for the
day, I wasn't going to bug them for a while.
I did supply them with several crap-towels to wipe down the foam with, as
it still had a lot of clinging water. Even the super-sticky Dow tape
won't adhere to wet surfaces.
The result of this detailing was a fairly intricate maze of interleaved
foam and flashing layers. These corners are apparently one of the tricky
areas that tend to have air leaks if not executed right.
The Grace was finally brought down and stuck over all this, making the
whole corner of the house look "gift-wrapped" as a friend referred to it.
This is why the Grace was left long and overhanging all the edges, so it
could tie in all the foam up the sides and flash *over* it to the outside.
And here, one of the guys was finding and patching the screw holes left
behind when they'd pulled some of the staging brackets off before. A
squirt of "black jack" or similar roof-patching goop, and a little
shaping so it was down flush with the surface. The theory on the
front-window leaks was that water had come in right here and gone straight
down inside the original wall. Seemed plausible to me given where it
had been appearing in the opening.
With a little more amusing cutting and patching, they finished up the
peak. I was shooting against a very bright but still overcast sky,
which is why the pix came out a little weird as I tried to stop down
far enough to get the detail.
I'm not completely clear on why they went through all this top-edge
complexity, because if you think about it, if air tries to leak
out through the old roof or wall sheathing and comes up underneath the Grace,
no amount of flash interlacing below that is going to close the potentially
straight path out over the top edges of the wall foam. Sure, there's an
extra layer of the blue Carlisle peel-n-stick under there but ultimately the
air seal at this point comes down to how the Grace is brought over the foam
and secured down.
As the foaming had worked its way around to the rear of the house, it was
time to deal with the basement bulkhead. There was plenty of room for the
first layer, but not *quite* enough for a full 2 inches of second layer.
They cut down a piece of foam to fit, and squished it in there to just
match the dimensions of the bulkhead frame piece.
That way, they could flash it back to the first layer of foam like with
the windows and doors, and make sure it was as water-managed as
everything else. The bulkhead frame itself is designed to shed water
off to the sides.
This left the remaining space open at the sides, but that was all under the
sill flashing metal and now less likely to get water intrusion. We'd come
up with something to fill these later.
With the bulkhead now built out to the right level, foaming and window
installation could proceed upward.
Not all adhesive building materials make perfect connections, and it's
sometimes really hard to get them to stick onto non-favorable surfaces.
Most of the connections between dissimilar materials were backed up with
runs of the thinner Dow tape to make the best "water wedge" possible at
those points. The top edges of those applications are critical and need to
be stuck down really well without any water or dust in the way, because
technically every horizontal run of tape creates a little tiny reverse
flash if it peels away even a little.
Where the underlying materials often didn't lay flat
it was very easy for the tape to develop a "fishmouth" on the upper edge.
We did find some spots where the tape itself had accumulated water
pockets inside, such as here. I was more worried about the potential
effects of this than the crew; they were looking ahead to when it would
all be covered under siding.
But water *does* get under siding sometimes...
If that water pocket collects at a horizontal seam between two foam panels
and can't fall out, where's it going to go? Potentially inward between
the foam edges, and now it's running down in between the two applied
layers. Hit another fishmouth in the tape on that inner layer, and ...
you can see where I'm going with this. While it seems like a small
thing and a long shot for bringing bulk water inward especially with
the Tyvek underneath all that, we need to consider
the effects of many years over any particular erroneous path.
In a truly inspiring outburst of anal-ness, in his own words, the PM also
took note of this and the presence of a lot of completely superfluous
tape on the *lower* edge of other materials, where all it does is
potentially catch and hold more water. There's no need for any
horizontal tape along the bottom of other adhered flashing layers, and he
was now all fired up about that and wanted to fix a bunch of these
unneeded lower-edge details. I managed to catch
a portion of his lecture
about "planning for failure". The whole point of having
layers lapped over each other is to let water drain outward and downward,
not trap it. Same reason the bottom flange of a window is never caulked or
taped over; if water gets behind there it needs to seep *out* easily. So
now he was asking for all that horizontal tape to get removed or at least
cut away to free any lower edges of other materials, such as the Grace used
as pan flashing here and the bottom edges of the Carlisle over it, and for
everyone to stop applying tape that way.
On-the-job training in action.
They went around and took care of most of the areas they could easily reach,
but not every possible point and it was sort of too late to revisit all of
the horizontal tape seams without some undesireable effect on the
surrounding materials. You can't just rip this off the foam without
taking the foil-face with it.
At least one case study I had read recommended using little handheld rubber
rollers on construction tape to stick it down really well, but that would
have increased the crew's workload even more. We acknowledge that the
outer face of the
foam *is* the supposedly weatherproof cladding for some number of days
during the project, and that due care is needed on these details. Some
from Building Science even suggest shifting all horizontal tape applications
upward to get more adherent surface area above any breaks in the external
this had everyone much more attuned to water issues and they knew the
front door had some problems, so they put a temporary water-kicker over
the front door to deflect most of it off to the side. Simple L of
aluminum, taped down at the upstream edge.
An hour after they left for the day more storms rolled through and it got
a good stress-test, too. Here we see it working: most of the water that
would have come down over the door got sent off to the side in a stream
like a tall cow on a flat rock. It definitely helped keep the door drier.
The windows were getting their own brute-test, too. Normally water shouldn't
ever be streaming straight down a window like this if a house has any
overhang, but here I didn't yet so the window was also the drainage plane.
They shed the water okay, but there was certainly a lot of water left in
the bottom of the sash frame that hadn't made it into the fairly generous
weep holes in the bottom. While it's all fiberglass and thus rotproof,
these would probably have to be cleaned out once the windows were
But the leaks weren't totally gone. With the kicker and some additional
splash-blocking stuff I had wedged into the open crevice surrounding the
front door it didn't seem to leak bulk water anymore, but it
still certainly got wet. The various patching up top had taken care of
some of the leaks *but not all*; a little water was still coming in some
of my circled holes in the attic. WTF?? At this point I was too tired
of this to even be angry anymore, and realized that the most poignant
and shocking feedback on the work I could tell the guys in the morning
was "it's still leaking".
There seemed to be many more screw buckets in the tool dump that night,
or maybe they just organized the stacks better. They were definitely
going through great numbers of different sizes while working along the
walls. Plenty of rolls of flashing, too. The crap-towels were also down
here drying as they'd invariably have to be used again the next day.
Next section (12)