Despite my letter to the Journal of Light Construction last year, I think energy audits are tremendously valuable. We’ve recently started using blower door tests and thermal imaging on some of our work, and so far, the results are excellent. Heat loss in buildings is weirdly unpredictable, so if you have access to a company with these tools, I recommend using them. Fine Homebuilding recently ran a great article on audits, so I won’t elaborate too much on this.

Nevertheless, there are a few crude tricks you can use to tell if you’re pumping heat into the outdoors. One of the easiest involves donning winter attire and craning your neck skyward.

In principle, your roof shingles should be approximately the same temperature as the outside air. Why? Because your shingles are on the outside of the house, specifically outside of the insulation. In winter, if the shingles are significantly warmer than the outside air, your home is losing heat. So what diagnostic tool can we use to determine if we have heat loss? Do you really have to carry a thermometer around your icy roof in January? There has to be an easier way!

After a snow-fall, most shallow-pitch roofs should be covered with an even layer of snow. In below freezing temperatures there should be very little snow melting. Pockets of thawed snow are generally indicative of heat loss.

What about roofs where snow doesn’t uniformly accumulate such as those with steep pitches, or metal? In this case, you can check your eaves and gutters for large icicles. See wiki’s ice-damming article here.

To complicate matters a bit, keep in mind that roof ventilation is designed to lower roof shingle temperature. So while hot patches and ice-damming almost certainly indicate heat loss, the reverse is not always true. A well ventilated roof may mask heat loss problems.

Earlier this week, with the temperatures hovering around freezing, we took a few photos of local roofs. Keep in mind that all of the photos below were taken in the same hour.
In this one you can see a ton of heat loss through the eaves. It’s tough to air-seal and insulate above exterior walls, so this is a common leakage point. Notice the garage (which is unconditioned) has fairly uniform snow coverage except directly adjacent the house.

Who knows what the heck is going on here? This is roof should have a lot of snow coverage – it has a really shallow pitch.

This is a commercial building, with a slightly steeper pitch. Here you see how roof penetrations can cause heat loss.


We’ve had a chilly winter, and partly as a result, our clients are requesting a lot of insulation improvements. Proper insulation upgrades require two strategies. The first is air-sealing. The second is the addition of insulation.

You knew it was going to get self-serving at some point. Here’s a place we air-sealed a few weeks ago, and photographed in the same hour as the other homes above. The house is within a few blocks of the *bad* examples. This was only a couple of hours after we added insulation to the attic – strategy 2 for those paying close attention. In other words, these results are largely from the air-sealing part of the job, not the addition of insulation.

Another view from the front of the house.

Rear of the same house. Chimneys are often a major source of heat loss, so this is kind of cool.

So could our results be from proper attic ventilation instead of proper insulation? In my opinion, probably not. I wouldn’t say that the house is particularly well ventilated, there are 3 small gable end vents, one of which is pictured above. That’s probably less than the attic square footage:vent area ratio of the ‘bad’ houses pictured above.

Every job is different – different size, different scope, caveat, caveat, caveat! Nevertheless, on this job, the total cost for air-sealing and adding R20 cellulose – essentially doubling the existing R value – was less than $5000.00 for over 1600 square feet of attic space.

Thanks for reading,


Leave a Comment