Windows-Water and Air Sealing

I’m going to make a statement that not everyone will agree with, windows leak water.  Not all are leaking now, some will develop the leak later, as the window and the sealing products used to install the window age.  The leak may be an error or failure in the installation, or it could be the window itself that is leaking.  It could be a design error, water funneling off a roof over a window, such as in the photo.  The leak may only be an incidental amount of water, or it could be a major bulk water concern.  The leak may also only be active during the once every ten-year major rain fall event.  The point is windows will leak water.

Now the question becomes, which windows will leak to the point where damage to the structure occurs?  If you can answer that, I have some lottery ticket numbers I am curious about.  We don’t know, so we need to build all window assemblies so that they will handle water leaks, just in case.

Most of the window installation details dealing with water are not new, use a sealant between the sides and top nailing flange and sheeting (when installing “outie” windows with nailing flanges), shingle lap the water resistive barrier (WRB), tape all but the bottom flange and use a head flashing.  During the design of the building, we should try to keep windows from seeing water to begin with by using large roof overhangs and kickout flashings on roof to wall intersects.

All of those details are to help keep water out, but we also need to address the water when it does get in so that it exits as quickly as possible.  We do this by sloping the rough opening windowsill, and then protecting the sill using some sort of water barrier.  We don’t add sealant or tape the bottom nailing flange, which would prevent the water from leaving the opening.

We’ve gone through the effort to do our best at keeping water out of our window assemblies and then given the any water that does find a pathway in a way to exit at the sill.  What if the water leak is at or near the window head, does that water have a clear pathway to the sill?  There is a common installation method, at least in my market, that I believe is affecting how the space between the rough opening framing and the window is drying after a wetting event, it’s a common method used to air seal and insulate this space, canned minimal expanding foam.  The photos are of a test jig I set up using clear plexiglass and a window rough opening mockup.  When using canned expanding foam around window openings, you cannot see how much foam is ending up in the opening.  The foam may be expanding all the way against the window flange, or maybe it is leaving uneven surfaces between the nailing flange and inside of the framing, much like the photos below show.  Do we want the foam to expand to fill the entire void?

My thought is no, and here’s why.  We want any water, either incidental or in bulk form, that gets into the system to freely drain out.  If the canned expanding foam fills the space, or leaves areas where water can become “stuck”, draining and drying will be slowed and damage to water sensitive materials may occur.

A better method

I recently sat through a training given by a major window manufacturer on how they suggest insulating and air sealing the window opening after the window is properly installed.  They want to see a backer rod installed tight to the nailing flange, then add any insulation, which could be either a loose fiberglass, mineral wool or minimal expanding canned spray foam to fill the remaining space (they also mentioned still air makes a good insulation).  Their thought is, with the backer rod being round and the space at the nailing flange being mostly square, any water can find its way out of the system by flowing down the small gap between the window nailing flange and round backer rod.  The illustration shows their detail at the window head, the same strategy would be used on the sides.  (Number 6 shows the backer rod detail tight to the nailing flange.) They did propose something different at the sill, the backer rod should be pulled back towards the inside to allow better drainage out the sill.  The design also includes sealant (number 3) on the interior side of the space, the sealant could be changed to one of the newer window tapes on the market.  For the most part, I think this is an improvement on the detail, but my argument is I don’t think the space between the nailing flange and backer rod is big enough to promote both drainage and drying.

The following photos show what I think detail should be, back to my clear plexiglass and window rough opening mockup.  I think we need a slightly larger space between the backer rod and nailing flange to equalize pressures inside the cavity, the backer rod will also provide a smoother surface (compared to the foam insulation alone) to allow easier drainage in the assembly.  The foam backer rod should be placed close to the same plane as the window glass to align the thermal boundary, but shy of contacting the window nailing flange.

Projects I am involved in, I prefer to not have the window extension jamb factory installed.  This simplifies the backer rod and air seal installations.  The photos below show the detail I included on the barndominium project.

After the window extension jambs are installed, more insulation could be included between the rough opening framing and extension jamb, though that’s less important than getting a good air seal around the window.

Several of the window manufacturers are making changes to the way they suggest air sealing around the window, usually changes aren’t made to systems that are working perfectly.  I believe they are moving in the right direction, but this small alteration to the manufacturer’s design can add more durability to the home.