With thanks to Stephen Thwaites of Thermotech Fiberglass Fenestration.

Windows are one of the most important elements of a building envelope. They provide views to the outdoors, and bring in sunlight while protecting the building interior from the elements. Unfortunately, windows tend to do a poor job of resisting the movement of heat. Compared to an opaque wall, which will likely have three to six inches of insulation, windows are weak thermal insulators and usually make up the weakest link in the building envelope. Given that windows take up increasingly large fractions of walls in modern buildings, this presents some challenges to building designers who wish to provide a high performance building envelope. For example, a typical wall built in a commercial building in Ontario might have an effective thermal resistance (or R-value) of R-15 with four inches of insulation. A standard double-glazed window with argon between the panes of glass, and thermally broken aluminum framing might have a U-value of 0.4 or less (an R-value of R-2.5 or greater). If the windows take up 40% of the wall area (window-to-wall ratio or WWR of 40%), the effective thermal resistance of the wall is reduced from R-15 to R-5. This is not great. Naturally we want to improve the thermal performance of the wall. We might decide that the easiest thing to do is to double the amount of wall insulation, so that the wall goes from R-15 to R-30. This increases the total R-value from R-5 to R-5.6. What happened there? Remember that the windows are the weakest link in the envelope. Improving the envelope by improving the wall performance is like strengthening a chain by strengthening the strongest links. We haven’t changed the weak links, and the improvement in performance is marginal. Most of the heat is being lost through the windows and reducing the heat lost through the wall doesn’t change this fact. So how can we improve the window performance? One way to do this is to simply turn some of the window area back into wall area by halving the window area from 40% to 20% of the total wall area. This improves the total R-value from R-5 to R-7.5, an improvement which is more than four times greater than one gets from doubling the wall insulation. An even larger improvement from R-5 to R-8.3 is seen from doubling the window performance from R-2.5 to R-5 (U=0.2) by specifying triple-glazed windows. Only when we combine all three of these measures does the total R-value of the wall rise to R-15.

Case WWR Window R-Value Wall R-Value Total R-Value
Base 40% 2.5 15 5.0
Wall R-value x 2 40% 2.5 30 5.6
Window area x 1/2 20% 2.5 15 7.5
Window R-value x 2 40% 5.0 15 8.3
All of the above 20% 5.0 30 15.0

So is getting rid of windows the answer? Of course not. It is important to keep in mind that windows can be a source of heating energy from the sun, not just a source of heating losses. Heat gains in a window are characterized by the solar heat gain coefficient and in some cases can be large enough to offset the heat losses. South facing windows with high solar heat gain and with low heat losses can provide a net energy benefit to the building. In addition, properly placed windows in combination with a good daylighting strategy can save on lighting energy and provide high quality illumination.

So where does this leave us? Provide enough windows so that people in the building have good views. If you can choose between better windows and more wall insulation choose better windows every time. Design the windows for good daylighting to save on lighting energy. Put more windows on the south and fewer windows on the north to maximize solar gains and minimize heat losses. This strategy will maximize the benefits of windows will minimizing their weaknesses.