Tony Wesley wrote:
wrote:
Windsun wrote:
... Basically, 100% uptime will cost you a lot more than 99% uptime.
If cloudy days were coinflips, storing enough energy for 1 day would
make 50% uptime possible, 2 would make 75%, with 3 for 88%, 4 for 94%,
5 for 97%, 6 for 98%, and 7 for 99%... 100% would be impossible.
Just an FYI, in the Detroit area, we just went from Dec 20 through Jan
5 (inclusive) with 14 minutes total sunshine.
I think what that means is that the clouds parted and exposed the solar
disk for 14 minutes total over that time period (and having spent my
holidays there, I cannot doubt its accuracy). It's not representative of
the total amount of solar energy that penetrated the clouds and reached
the surface of the earth. A little bit of math provides a quick sanity
check:
14 minutes of full sun = 233 Wh/m^2
233 Wh/m^2 = 13.7 Wh/m^2/day over 17 days
13.7 Wh/m^2/day = approximately 2 W/m^2 during daylight hours
That's a tad darker than Barrow, Alaska (in the Arctic Circle) in late
January, and no way was Detroit that dark. Indeed, if you check NREL's
data on the subject, Detroit averages 1300 Wh/m^2/day in December
despite the fact that long cloudy spells like the one you described are
the norm for the area at that time of year. You can also check NREL's 30
years of hourly data to see that peak horizontal insolation in Detroit
is usually greater than 200 W/m^2 and nearly always exceeds 100 W/m^2,
even on Detroit's cloudiest days.