Note from Conrad:This is a guest post by fellow Edmontonian Alice Harkness. Thanks for your excellent work Alice!

I experimented last winter with a very simple way of upgrading the insulation value of the beautiful original windows of my house: I added a third, Low-E glaze between the double hung window and the storm window.

Method

There is a 3/4 inch thick space inside the storm window created by the stop (the strip of wood that serves as the outside edge of the channel for the upper double hung window, and backing for the storm). Here’s what I did:

1. I bought a custom cut piece of Low-E glass from Can-Am Glass Products
2. I bought 1 inch Foam Backer Rod, slit it lengthwise but just halfway through, and cupped it around the edge of the new piece of glass. This saved me making a wooden frame, and added insulation value over using wood. 
3. I inserted the third glaze from the outside (taking off the storm window), and held it in place by sandwiching it between the storm window and the upper double hung window.
4. A backing needs to be made for the lower half of the window, and I did this using extruded wall insulation, cut to fit snugly into the channels and across the sill. I wrapped the foam board in aluminum foil as a temporary solution (I was thinking about whether a reflective surface would bring more light into the house, and whether it would be jarring to see a reflective surface in that spot. The obvious material to use in a permanent solution would be wood, but the foam board does lend itself to being easily removed for seasonal purposes.) The Low-E window requires no caulking because the 1 inch Foam Backer Rod, with the added thickness of 1/8 inch of glass, is pressed into the 3/4 inch width of the stop.
5. Finally, I filled the storm window’s vent holes and removed the wooden vent flap before putting the storm back on.

In this close-up photo the slit down the center of the gray Foam Backer Rod (to the right of the aluminum foil) is where the Low-E glaze is sitting.

I carried out my experiment on two windows. In the photo of my dining room two-window set (top), the left window has the addition of a third glaze while the right window does not. I did the same thing on my kitchen two-window set, upgraded one window and left the other with just its original wooden storm.

Results

Using an infrared gun to measure the temperature of the different surfaces, I was able to measure the degree of improvement that the third glaze provided. I used black electrical tape, which has a minimal emissivity, for taking the readings. With outside temperatures ranging from -7 C to -25 C, I found that the windows with the third glaze were about 3-5 degrees warmer than the standard windows. In general, this translated into an improvement of about 130 - 180%.

I also took the readings of a variety of double-glazed window units at friends’ houses, from older double-glazed picture windows to new Ener-Guide Star-rated windows, and found that the performance of my three-glazed windows was equal to (to within 2%) or considerably better than the windows in my survey (see note below).

As well as preserving the beautiful craftsmanship and authenticity of the original windows, the money saved by this method is huge. It cost my friends over $1,000 per window (labor and materials) to replace their original windows with Ener-Guide Star-rated windows. My experiment on windows of about the same size cost me about $100 per window ($85 for the glass plus the other materials).

This is a solution which works best for windows that remain shut year round. However, it is possible to use this method on windows that open in the summer. It just means the spring/fall ritual of putting on and taking off storms, and having to re-caulk the double hung and storm windows every fall.

Condensation

In regards to the problem of condensation, when the humidity in the kitchen went up to 45-50 because of a lot of cooking, the benefit of the third glaze became very clear: the unimproved window would get quite fogged up (on the side facing the room) while the improved one remained clear of any condensation. In regards to the potential problem of condensation between the glazes, I was pleased to see that I did not have this problem. The humidity in my house normally sits at around 30 in the winter, but during very cold snaps can drop as low as about 17. If a good level of humidity were maintained, would I run into the problem of condensation between the glazes? I don’t know, but I do know that, in contrast to a manufactured unit, I could remedy the problem myself by redoing the caulking on the double hung window.

Finally

To my mind the great beauty and value in these older homes lies in the fact that, because they were hand built originally, they can also be fixed “by hand”. If a window cracks I can go out and chip out the old putty, put in a new piece of glass, and spackle in some new putty. But if a window cracks in a manufactured unit the whole unit has to be taken out and replaced. So every time a person replaces their original windows with bought units, they are sacrificing their ability to maintain that part of their house.

This is one more, and essential, reason that I like this third glaze solution to the problem of cold windows: because its simplicity and hand built nature is integral to the nature of the house itself.

A note on calculations

To ascertain the relative heat loss of a window I divided the difference between the outside and inside ambient temperatures by the heat lost through the window (ie. the inside ambient temperature minus the window temperature).

In regards to the data itself, some of it was collected with an infrared gun made by NAPA (used for automotive diagnoses) and some was collected using a more sophisticated gauge made by Nexus (capable of a number of functions including infrared readings). I noticed that the Nexus gun gave me consistently higher readings than the NAPA gun, a difference which became significant when I ran the numbers through the equation above. (On testing the two guns against each other, I found the Nexus gave readings about two degrees higher than the NAPA.) Because of this I realized I could not pool all of the data together. So my conclusions have been drawn out of two sets of data.

On my own windows I took readings on four separate occasions, twice with the Nexus gun and twice with the NAPA gun. Within each category (ie. NAPA gun on the two unimproved windows; NAPA gun on the two improved windows...) the relative heat loss calculations came within a range of 5% of each other. As to the testing of double-glazed units in the houses of my friends, I have readings on two windows using the Nexus gun, and six readings on six different windows made with the NAPA gun. Comparing the average relative heat loss of my improved windows with the appropriate category of these double-glazed windows (ie. NAPA or Nexus), mine showed to be over 30% more efficient than the worst of the windows tested, and within 4% of the best (the brand new Ener-Guide window).

I used the hard Low-E glass (Sungate 500) which, according to manufacturer’s specs, improves insulation by one R-value. The soft Low-E glass (Solar Ban) could also be used. The Solar Ban is rated to improve R-value by two factors.