Monday, May 27, 2019

Construction - Window Installation

Most of the north and west exterior walls of the house are earth sheltered and the garage abuts the east wall.
Only 12 second windows were built instead of the 14
shown on the plans.  The first story windows stayed
true to the plans, making a total of 19 large and 1 small
 south-facing windows 
  Therefore, we would have been hard pressed to fit glazing into any of these walls.  But, no loss, we intentionally designed the house with all south-facing glazing (except for one small east window under the entry portico).  It took twelve 3' x 5' windows on the second story, four 3' x 5' and three 3' x 6' windows on the first story to meet minimum code for the ratio of glass-to-floor area.  The interior of the house is laid out so that all spaces except the bathrooms and the TV viewing area have windows.  The living room has no ground level windows but, since there are no second story rooms above it, it is generously lit by the second story clerestory windows.

 Green Features 
All the windows and the two exterior doors are designed for maximum sustainability.
     -  Energy efficient glass                -  Recessed profile               
     -  Energy efficient frames             -  Overhangs                
     -  Fixed vs.Operable                     -  Angle of incidence 
     -  Swing closure                            -  Peripheral seal            
     -  Translucent (frosted) glass        -  Glass-to-floor ratio

(Click on any photo to enlarge it for better viewing.)
    
Energy Efficient Glass
Our windows and doors have all of the bells and whistles,
like double glass and low-e coating, that make them first quality.  But, at +/-R-4, they are like holes in the wall compared to our fifteen inch, R-60 walls.  According to Daniiel Chiras in his book The Solar House; Passive Heating and Cooling, the problem is not just the glass -- about a fourth of heat gain or loss occurs through the frames, making it imperative that both the glass and the frames be energy efficient.

Energy Efficient Frames
A fiberglass frame has the same coefficient of thermal expansion as window glass unlike wood and vinyl frames (vinyl is by far the biggest mismatch).  When the frame and glass expand and contract at the same rate, air leakage between the frame and glass that might compromise the seal of the double panes or admit air into the living space becomes moot.  Unlike vinyl, fiberglass is resistant to UV rays, is much stiffer and accepts paint.  And it is made from sand, the most abundant substance on earth, rather than petroleum.

Fixed vs. Operable 
The top label at the right contains a map of the US with shading to indicate the area for which the window is designed.  Notice that it encompasses all but the southern reaches of the country.  The second label shows a reverse picture; only the southern reaches are shaded.  Fortunately, both labels cover St Louis.

Each label has a series of four numbers near the top.  The first number is the U-value which is an inverse expression of R-value in that the lower the number, the higher the R-value.  Notice that the U-value for the top label is lower than for the bottom label.

The two labels were photographed on side-by-side windows.  The top label is on a fixed unit and the bottom on an operable unit.  The fact that the maps and the U-values show an energy rating that is better for the fixed glass window validates our decision to limit the number of operable windows.  Since we will be depending on the energy recovery ventilator for most of our fresh air, having only 8 of 19 windows that are operable will be sufficient for additional ventilation when it is needed or desired.

Swing Closure
Windows and doors that are hinged prevent air infiltration better than sliding doors and windows.  To say it another way, casement, hopper and awning windows and hinged doors are more energy efficient than double- or single-hung windows or doors and windows that slide horizontally.  The reason is that swing doors and windows can be locked tight against a gasket-like seal while sliding units need a looser fit to be able to move up and down or side-to-side.  While the fit of modern sliding units are infinitely better than their ancestors, their air sealing still falls short of the gasket-type.

Frosted Glass

Greenhouses utilize the solar advantage of translucent  glass.  The glass in old-time greenhouses was painted with whitewash.  Newer greenhouses use "clear" sheet plastic that is, in reality, translucent in that you can't see clear images through it or they utilize translucent corrugated fiberglass panels.  Accordingly, we have 19 south-facing windows at least 3' x 5' in size of which more than half have translucent (frosted) glass instead of clear glass. 

The solar performance of the two types of glass differ in this way.  Sunshine through clear glass warms whatever it shines on.  If "it" is thermal mass, like concrete overlaid with ceramic or porcelain tile and of an appropriate color, the solar heat is absorbed into the mass without overheating its surface, which is a good thing.  If "it" lacks thermal mass (think wood or carpet floors, furniture and even drywall), "it" overheats and warms the air when it is not needed.  In a sense, precious heat is wasted.  

Frosted glass, on the other hand, diffuses (scatters) the sun's rays.  The heat rides the air currents within the dwelling until it either finds thermal mass or it conserves heat in the thermal mass by providing heat that would otherwise be coming from the thermal mass.  In our case, the thermal mass is the downstairs floor, the lower 10' of the long concrete back wall of the house and the soil under and behind them as well as the soil  beneath the insulation-watershed umbrella adjacent to the house.  Consequently, all of our first story windows are transparent because of their proximity to the thermal mass whereas all but two of the second story windows, where mass is missing, are frosted.

Recessed Profile
Today's new construction windows have nailing fins that position
the surface of the glass flush with the outside surface of the wall.  While this design eliminates air leakage between the window and the wall framing, it is not as energy efficient as it could be if the glass were not flush with the wall.  Here's why.

The rate at which heat is lost through the glass is dependent on the rate at which heat is removed from the outside surface of the glass.  If the heat can pass through the glass and linger for awhile on the outside surface of the glass, it slows the transfer of more heat.  What determines the amount of lingering?  Air movement.  Wind can remove the heat as fast as it forms and, since heat seeks cold, the rate at which heat is drawn through the glass accelerates, a phenomenon called "wind washing".

Our windows, by being positioned about 10" inside the plane of the wall and being south-facing, are sheltered from the prevailing north and west winter winds.  As such, the amount of wind-washing is about as minimal as it gets.  The red arrows in the nearby photo bracket the 15" thickness of the wall; the green arrows delineate the area on which the window rests; the blue arrows show the distance from the window to the outer surface of the wall -- about 10" after the trim boards are in place.
The distance between the top of the windows and the 2'
overhangs is greater for the first story than for the second
story but the difference in the amount of sunshine reaching
 the floors inside is negligible even for the taller windows
 in the foreground.  (The sheathing is protected by lumber
wraps and sheet plastic which will be replaced with proper
house-wrap and metal siding eventually.)

Overhangs
Sunshine through the windows is a blessing during the cooler months of the year but very unwelcome from May through September.  At our St Louis latitude, it takes a two-foot overhanging soffet to block the midday sun during the hot summer months.  

Early morning and late afternoon sun can shine under the overhang but, with the windows recessed, the adjacent wall offers shade in a way not possible with flush-mounted windows.  However, after the end of May and until the first of September, the sun angle increases enough that sunlight is completely blocked all day, even during the early and late hours. With the windows recessed, though, the sun that does reach the floor during late spring and early fall covers an area a third less than would be the case with flush-mounted windows.  As long as the overhang is reasonably close to the top of the windows, the height of the windows
Even with tall windows and an overhang a good
 distance above the tops of the windows, very little
 midday sunshine reaches the floor in early May;
 later in May none will.
is not critical. The nearby photos show the extent to which the sun reaches the floor during the first week of May through windows 6' tall and an overhang +/-2' above the tops of the windows.

Angle of Incidence
The same windows in late afternoon after the sun has
dropped low enough to shine under the overhang; the
amount of sunlight reaching the floor is restricted by
the recessed profile.
Another important concept in understanding passive solar is the angle of incidence, the angle at which sunlight strikes the glass in windows or the roof of a greenhouse.  The more perpendicular the sunshine is to the glass, the more energy absorbed by the glass.  The steeper the angle of the sunshine, the more energy reflected by the glass.  The practical take-away is that at least 90% of the sun's energy reaching a vertical window during the summer months does so at such a high angle that it is reflected rather than raising the temperature inside the house.  By the same token, the amount of energy absorbed through vertical windows at winter solstice time is at least 90% of what could be expected through glass that is tilted toward perpendicularity with the winter sun, as often seen in pictures of passive solar homes and greenhouses.  But tilted glass is not a good idea because it absorbs too much unwanted heat in summer unless it is covered with an exaggerated overhang, which is rare in the pictures I've seen.

In summary, our passive solar design utilizes three things to maximize solar gain when it is needed and minimize solar gain when it is not..  First, our windows are vertical rather than tilted for the reasons just discussed.  Second, we have overhangs designed for our latitude that provide adequate overhead shading.  And, third, our windows are recessed for lateral shading.

Peripheral Seal
The decision to recess the windows meant using "replacement" rather than "new-construction" windows.  No doubt about it, the latter with their nailing fins are better at blocking air leakage between the windows and the framing and sheathing that surrounds them unless the air sealing around replacement windows is done with extreme care.  

Sealing our windows was complicated by the fact that I built the housing for them as premade wall sections using custom jigs several years in advance of even ordering the windows. I deliberately made the window openings larger side-to-side and top-to-bottom than the window manufacturer specified because I knew that it would be extremely difficult to enlarge the openings once the units were in a wall.  As a result, the gap on either side of the windows averaged 1/2" and the gap at the top of the windows more like 3/4".  


All three sizes of backer rod --  3/8", 1/2" and 5/8" -- were
 needed in various combinations to close the gaps 

between the window frames and the rough openings on
 the exterior side 
At first glance, it would seem that all one would have to do for fool-proof air sealing is to fill the gaps with spray foam but even minimal expanding foam that is designed for this purpose is capable of distorting window frames if used in excess.  Rather the window manufacturer recommends that the foam be limited to a 1" strip located about 1" into the space between the window frame and the rough opening on the interior side.  On the exterior side, it recommends a course of properly-sized backer rod just inside the outer edge of the frame and, in doing so, leaving enough room for a bead of caulk.  Since our windows are inset, the caulk will be added when the jam extension between the window frame and the trim boards around the window opening are installed.  In this manner then, the spray foam provides the air sealing and the backer rod and caulk provide weather sealing.  The wide 3/4" gaps at the tops of windows were partially filled with plywood ahead of the foam and backer rod.

Glass-to-floor Ratio
In his definitive book, The Passive Solar Energy Book; A Complete Guide to Passive Solar Home, Greenhouse and Building Design, Edward Mazria recommends for our temperate climate (average winter temperatures of 35 to 45 degrees) a ratio of 0.11 to 0.25 sq ft of south-facing glass for 1.0 sq ft of floor area "in order to keep the space at an average temperature of 65 to 70 degrees during most of the winter".  Only our concrete first floor can be counted as floor area for calculating the ratio, so 2,800 sq ft of floor area to 354 sq ft of south-facing glass gives us ratio of 0.13, which falls at the lower end of Mazria's recommended range.  

Such a low ratio could be discouraging for us older folks who do better with temperatures in the mid-70s and the numbers get even worse because Mazria's figures for the temperate zone do not differentiate for cloud cover.  Here in the St Louis region, according to Daniel Chiras, only 47% of possible sunshine during the months of December, January and February actually reaches the glass.  Therefore, a conventional winter-centric passive solar house would definitely require supplemental (CO2-emitting) heat.  However, when heat coming through our windows during winter is added to the heat generated by our Annualize GeoSolar summer-centric system, it is almost certain that our thermal mass will maintain a comfortable year-round temperature hovering around our preferred 74 degrees without any supplemental heat.  Don Stephens, the father of AGS warns that too much heat could actually be a problem eventually that would necessitate mothballing part of the solar collector.  (For an understanding of AGS, click on "Featured Post" in the left column above.)

Installation
Not much needs to be said about the actual installation of the windows except for some considerations dictated by working alone.  Most of the windows were 3 x 5' and about, I would guess, 60-70 lb.  Knowing that I would be working alone, I requested that the double windows be shipped as individual windows rather than joined together with a factory-installed mullion.  Even then it was a stretch to hoist the second story windows to place without using a ledger on which to set them part way up.  By using temporary vertical stops against which to set the window, I could lift it into the opening without its falling outward and, by installing the stops exactly plumb and at the exact location of the exterior surface of the window in the rough opening, the window ended up automatically in the position it needed to be in except for plumbing in a left-right direction. 

The additional work required for the site-made mullions is at least partially compensated by the knowledge that they are more energy efficient than factory-made mullions would have been since they are filled with foam board insulation.

Monday, May 20, 2019

Construction - Standing Seam Steel Roof - Part II: Installation

The prior post covered the preliminary steps leading up to the actual installation of the steel.  This post details the installation of the garage roof that was actually the third section of roof that I had installed but I am using it as exhibit A because, by being a simple rectangle, it is easier to describe. In a subsequent post, I will deal with the hip-roofed porch which is far more challenging.  

Installing Moisture Barriers
After I identified the truest eave-rake corner, I proceeded with the installation in the following order
1.   Metal drip edge over the junction of the sheathing and the fascia (eave side only; the rake sides would be handled with a steel rake trim over the roof panels later); the drip edge is installed first so it can be overlapped by the ice shield in step two.
2.  Two courses of stick-down ice shield paralleling the eave; the ice shield covered the drip edge; code requires that the ice shield extend 24" beyond the interior face of the exterior wall and, since our eaves overhang 24" and the wall is 15" thick, two 36" courses of shield were necessary to meet code.  Unfortunately, the installation was being done on warm-ish winter days that were still too cool for proper adhesion of the shield to the sheathing so we used a torch to seal the edge at the eave then held our noses and used roofing nails to secure the remainder, hoping that the hot summer sun will activate the adhesive later; ordinarily, the ice shield would be unnecessary for an unheated garage but, since ours will be seriously insulated and warmed considerably by the AGS system, I decided to error on the side of caution.
3.  Roofing nails to fasten as many courses of 30# felt paper as it took to reach from the ice shield to the wall of the second story (first photo below (clicking on any of the photos will enlarge them for better viewing)); the junction between the future steel siding on the second story wall and the steel roofing panels would eventually be protected from moisture by a metal trim piece called "dormer flashing"; however, rather than depend on it alone, I continued the felt paper from the roof up the wall a ways as a secondary barrier, taking care not to bend it so abruptly as to compromise its integrity; most roofing contractors use staples in a slap stapler in a rather random fashion but I like the idea of using roofing nails along the overlapped edges of the felt because, when driven tightly, the heads help to seal the hole in the felt against water in a way that staples may not, especially when the paper is damaged by the head of the slap stapler should it be held at an angle to the surface of the paper.
4.  A mason line paralleling the eave to which the ends of the panels could be referenced for a straight installation (second picture).  In our case, the line was positioned for the 1 1/2" overhang specified for the  Rain Handler system that we are using on all sides of the house in lieu of conventional gutters.

It is important to point out that the virtually impervious 30# felt can be used in this situation only because the sheathing, which will inevitably collect moisture, will be able to dry from underneath even though the roof is of cathedral design.  The trusses supporting it are 12" thick and I plan to hold the insulation down from their top edges several inches to allow air to circulate between the insulation and sheathing.  (For a in-depth discussion of moisture abatement, check out the previous posts on Design - Vapor and Air Barriers and Design - Vapor and Air Barriers (Cont'd)). 

Major Problem
As described in the post preceding this one, the metal panels for the garage were the wrong length and had to be reordered.  Unfortunately, I did not catch the mistake until the moisture barrier was in place and, as luck would have it, a strong wind soon had its way with the 30# felt despite its toughness.  A small section blew away but most of it stayed in place even after pulling loose in many areas.  I re-nailed it with Grip-Rite Round Cap Roofing Nails which held it reasonably well until the replacement 
Green arrows = felt paper being unrolled just ahead of
the installation of the steel panels; red arrow = felt paper
extending up the wall a ways for better protection of the
wall-roof interface

panels were available but knowing that it would not be wise to leave the Grip-Rites under the metal panels because they would be proud enough possibly to dent the steel from below if it were stepped on.  Consequently, it was just easier at little additional cost to pull the Grip-Rite nails and lay down another layer of felt but doing both a little at a time just ahead of the next panel to be installed rather than laying it all down at once and risking another wind event.

Observation of Note
I bought 30# felt paper from our local (traditional) lumber yard as well as from one of the big box retailers and, despite only a small price spread, found a significant difference in quality.  Fortunately, I found the discrepancy early enough on the first roof to replace the big box paper with the better quality lumber yard paper before the first panel went down.  And, of course, I used the heavier paper for the rest of the roofs.

Installing the Steel Roof Panels
After the taut mason line has been installed along the eave as a reference for the ends of the steel panels, the panels can then be installed from left to right or from right to left, depending on the location of the truest corner
The mason line is tautly strung so that it designates the
amount of overhang for the steel panels but situated just
low enough that the panels do not touch and distort it
that was identified by using the 3-4-5 method (actually, 9-12-15 or 12-16-20 for more accuracy if the roof is large enough).  Knowing which corner forms the best right angle is a starting point but it is probably not accurate enough to install the first panel without further checking.  If the first panel is not aligned perfectly perpendicular to the mason line, the rake edges of the panels will not align with each other, forming an amateurish saw-tooth edge to the roof, the jaggedness of which is directly proportional to how much the first panel is off.  

To satisfy the 9-12-15 right triangle, the lower end of the
 first panel is anchored with one screw then the "0" end of the
 tape measure is held at the 12' mark on the mason line (red);
 the other end of the tape measure is held on the outside
 edge of the panel at the 9' mark (green) while the upper end of
 the panel is swung in or out until the 9' mark on the panel 
precisely underlies the 15' mark on the tape measure

I installed the first panel next to the rake with the standing seam side towards the rake.  In order to verify the absolutely critical right-angle-ness of the first panel to the mason line, I drove a screw through the predrilled hole in the panel closest 
to the eave then, with help from a neighbor, redid the 9-12-15 measurement.  Any right-angle discrepancy could easily be corrected by moving the top end of the panel slightly left or right to align it with 15' on the tape measure.  As long as care was taken to align succeeding panels with the mason line, the ends of the panels were nice and even.  By contrast, our first installation was the smallish rectangular roof on the west side of the house that we did before cold weather.  I did not think to double-check the right-angle-ness of the first panel as just described and ended up with a little sawtoothness.
Resulting smooth, versus 
saw-toothed, edge

The leading edge of each panel contains a row of holes to receive the screws that are furnished by the manufacturer (for a better understanding of the components, visit ProSnap Steel Roofing on Menards website).  The screws should be no further than 2' apart which meant one in every other hole.  The next panel is lined up with the mason line at the lower end and its standing seam ridge is aligned over the lessor ridge of the preceding panel.  It is then a simple matter of snapping the panels together by pressing on the ridge of the second panel, starting at the eave, until it snaps to place over the ridge of the first panel then repeating the process up the slope of the roof until the panel is fully seated on the prior panel for its full length.  The snapping can be done with the heel of the hand or with a rubber mallet.

Installing the Rake Trim
After the panels were in place, it was time to add the trim pieces that divert water from the rake edges and the junction between the house wall and the roof.  First, the rake trim.  On the side of the roof where the first panel was installed, the trim fits against the surface of the fascia in a vertical direction and overlaps the standing seam ridge in horizontal direction.  It is fastened to the flat part of the panel with screws having neoprene grommets under the heads that keep water from penetrating around the screws.  Since the trim overlaps the standing seam, any water that runs under the edge of the trim is caught by the standing seam so that it is unnecessary to use any sort of caulking under the edge of the trim.

The installation of the rake trim on the opposite
The panels nearly installed; notice the block on the
 fascia to hold the ladder off of the mason line and
out of the way of installing the panels behind it; notice
that the felt paper extends up the wall a ways to be
covered later by the house wrap and cladding
side of the roof is complicated by the fact that there is no standing seam under the trim to divert water.  Therefore,
 to serve as a water barrier, so-called "tape mastic" that has the consistency of very dense plumber's putty but looks like weather stripping is applied to the under-edge of the trim before the latter is screwed to place.

The rake trim came in 12' lengths so I started the first piece at the eave and overlapped it for a few inches with a second
The rake trim did not fit snugly against the fascia; it had
to be tightened down with screws after painting the
 fascia with the final coat of paint
12-footer.  It took still another shorter piece to reach the second story wall.  The trim did not fit the facia tightly in several areas, especially where two pieces overlapped, so I resorted to a few grommeted screws to snug up the vertical part of the trim after the fascia had been painted.


Installing the Dormer Flashing
The dormer flashing is used at the top of the roof to divert water from the junction of the second floor wall and the roof panels.  It slips up under the house wrap and the wall cladding and overlaps the ends of the steel panels.  In doing so, it rests on the tops of the standing seam ridges leaving gaps under it.  In order to keep water from blowing in through the gaps, the latter are filled with so-called "closure strips" that are die-cut to fit the contours of the flat part of the panels and coated with adhesive. They are pressed to place between the ridges just inside the lower edge of the dormer flashing to prevent, or at least minimize, water penetration.  Once the closure strips are in place the flashing is secured with grommeted screws long enough to penetrate through the tops of the standing seam ridges and screw into the sheathing.  The vertical flange of the flashing is also screwed or nailed to the wall sheathing before the house wrap and wall cladding are installed.