On Tuesday, March 26, 2013 1:13:37 PM UTC-7, wrote:
On Mar 26, 4:03*pm, Metspitzer wrote:

I say it's total BS. The best study I saw on roof color was
done in FL. They took a bunch of identical houses, put varying
types and color of roofs on them, fully instrumented the houses,
and measured during AC season.

The conclusion was that if you had a shingle roof, it made
about a 10% difference in energy usage if you had a black
rood or a white roof. And that was the energy usage with
the houses unoccupied. They did another test when they
were occupied and the energy usage difference dropped to
just a few percent. Which makes sense, because when
occupied the energy usage is going to be higher, because
people are opening door, turning on TVs, cooking, etc.

Was that based on stock white shingles or shingles painted white?
Because surfaces painted white reflect roughly 90% of the sunlight,
while regular white shingles reflect only 10%.