(With a riddle at the end: The first person to answer it correctly gets a copy of my new book.)
Storm windows save a lot of energy, especially in old buildings where the main window is single-pane. But they may not work the way we think they work.
In Energy Audits and Improvements for Commercial Buildings, I wrote the following paragraph, in which I make a strong statement. I say that storm windows, if left partially open, still save energy for the portion that is still closed and in place between the indoor window and the outdoors. Huh?? You mean if the storm window is only half closed, and outdoor air could be freely wafting up between the main window and the closed portion of the storm, it’s still saving energy? Here’s what I wrote:
Storm windows are vulnerable to being left open, and therefore typically require an active operation and maintenance program to ensure that they are closed in winter. Interestingly, a double-hung storm window, if left open (in place for the upper half of the window, open for the lower half), still delivers much of its intended insulating function. The air between the primary window and the upper half storm is warmer than the outdoor air below it, and so mostly does not move, despite being open to the outdoor air below. Therefore, we cannot assume that a half-open storm window is entirely ineffective; we can only assume it is not functioning for the area that is open.
The statement came from some temperature measurements I took on the surface of windows in our office building, on a January weekend a couple years ago. I measured the indoor surface temperature (the glass temperature) at the center of some windows, when it was a chilly 28 degrees outdoors. And I found that this indoor glass surface temperature, at the upper half of the window, where the storm window is still in place outside it, is about as warm as if the storm window is fully closed, even though the bottom half of the storm window is open. And even though the air trapped between the upper window and closed upper storm is in open contact with the outdoor air below.
Why is this? I can only guess that the upper half of the window is trapping relatively warm air (relative to the outdoor air below it), and since warm air is more buoyant than cold air, it stays where it is.
This past weekend I decided to test my speculation, by measuring the air temperature between the upper window and the storm. It happened to be 31 degrees outdoors, another cold day, very similar to the January day two years ago when I did the indoor window surface temperature measurements. I went back to our office building, and opened a storm window (raised the lower storm sash), went outdoors, and waited for temperatures to come to equilibrium now that the storm was open. It is a first floor window, so I could reach it from outdoors. Then I stuck a piece of black duct tape on the end of a stick, and pushed it up between the window and the storm. I let the tape come to temperature equilibrium, and then measured the trapped air temperature with an infrared thermometer, using the tape as a target, up through the opening below. My neck is just now starting to get over the crick I got from holding the stick and shooting the temperature while looking up into the air pocket from below, with my head stuck in through the open storm window and up against the main single-pane window.
The results? With 65 degrees indoors and 31 degrees outdoors (air temperatures), if my trapped-warm-air guess is correct, we would expect to measure a temperature warmer than 31 degrees up in that pocket of air trapped between the top half of the window and the storm, even though the bottom half is wide open to the outdoor air. The envelope please: I measured 43.5 degrees air temperature between the window and storm at the center of the upper half of the double hung window. WAY warmer than 31 degrees. That upper half of the storm window is still working, even though it is exposed to outdoor air right below it. The indoor room CANNOT lose as much heat through the main single-pane window to 43.5 degrees as it can to 31 degrees. Not physically possible.
How about as we get closer to the outdoor air, toward the bottom of the upper half of the window? I lowered the stick and tape and took another measurement. Still 43.5 degrees, even ¾ of the way down the upper half of the window, within 8” of the open outdoor air below. So, that upper storm is likely working well over its full height.
Recall that not only is the warm air pocket trapped, but the upper half of the window now has not only the main single-pane window, AND the original upper-half of the storm, it now ALSO has the raised lower storm sash. It’s basically triple-pane up there. But the key, I think, is still that the warm air is trapped.
Now, if that top half of the storm were lowered even slightly, leaving an air gap at the top, I would imagine that the warm air would be pumped right out by buoyancy (stack effect), and the storm window would lose its effectiveness.
What does this all mean? Well, before I start on conclusions, let me say that we still certainly don’t want to leave storm windows open. When open on the bottom half, heat is being lost at a high rate through an old-fashioned single-pane window. But what it means is that when we go to calculate how much heat is being lost (or how much heat is saved by closing the window), we cannot multiply the full area of the window by the heat loss through a single pane window. To be conservative, we should multiply it by one half of the window area.
Another takeaway: Many people LIKE to keep storm windows open in the winter, so they can occasionally crack open the main window for ventilation. But, we should be telling them, “Just open the storm window slightly, like a few inches. You will get plenty of ventilation, but the storm window will save almost as much energy as if it is fully closed!” Most triple-track aluminum storm windows have an open position (a notch on each side to hold the storm window lower sash) that is in fact just a few inches open. Perfect. We should use this position for ventilation, rather than raising the storm window all the way up.
Now, are open storm windows REALLY such a problem. In other words, are there really so many of them that are left open? Well, I did a survey of windows in an, ahem, engineering office in an old building. This particular office building HAPPENS to be the home of a fairly progressive firm, with a particular interest in energy conservation. We would expect a building like this, with a thoughtful tenant like this, to have all those storms nice and closed, right? Well, I found that out of 39 old double-hung windows with double-hung triple-track aluminum storm windows, there were 14 that were at least half-open, most of which (9 of ‘em) were in fact fully open (lower sash fully raised). That’s over one-third of the storm windows that are significantly open.
Wanna know something else interesting? ALL the open storm windows in this example building are on the second floor! Any guesses as to why? MY guess is that the stack effect, which draws cold air into the first floor, and pushes it out of the building on the second floor, causes drafts on the first floor, so people are more likely to close their storm windows. They are also more likely to be getting the ventilation they want, because of this infiltration on the first floor. On the second floor, the warm air is leaving the windows, so there are fewer drafts. And people may want more ventilation.
Closing thoughts: My main focus in this discussion has been on conductive heat losses. I’m not talking about infiltration rates with open or closed storm windows, and I presume that closed storm windows reduce infiltration, and that partially-open storm windows are as bad for infiltration as fully-open storm windows. And, to reiterate – I like storm windows to be closed. I just don’t want us to overestimate the savings from closing them. And if they need to be left partially open, for occasional ventilation, it is MUCH better to leave them open just a crack of a few inches than it is to leave them open to the max.
OK, now here’s the riddle. Above, I describe the warm pocket of air between a main window and open storm window, at the top half, and I even say “it stays where it is.” Now, I’m going to go out on a limb and speculate that this warm pocket of air is constantly moving. I bet you that inside the pocket of warm air, up against the warmer main window, air is rising as it is warmed. At the top, it turns around and starts falling as it is cooled by the colder storm window, up against the interior side of the storm window. The first good answer to the following question gets a copy of my new book: As the colder air in the pocket falls toward the bottom of the storm window, why does it not fall out of the air pocket into the outdoor air, to be replaced by rising outdoor air, pulled upward by stack effect?