Not that bad -- just look at the generation after 4pm (when it is worth ~ 2x). You will not like the results for the 20% penalty.
Ok. It appears that I am completely wrong. Ran the hourly calculations.
180 degrees (South) Peak production: 0.26MWh Non Peak: 6.2MWh
300 degrees (WNW) Peak production: 0.38MWh Non Peak 4.9MWh
Another factor that I also thought would make a big difference is morning cloud cover. It only makes a slight difference, 190 degrees is optimal for total production.
My MATLAB code:
f = fopen('pvwatts_hourly-180.csv')
for n=1:19
fgetl(f);
end
C = textscan(f,'"%f","%f","%f","%f","%f","%f","%f","%f","%f","%f","%f"');
month = C{1};
hour = C{3};
wh = C{11};
peak_wh=0;
for n=1:length(month)
if month(n) >= 6 && month(n) <= 10
if hour(n) >= 15 && hour(n) < 20 %peak hour 4-9pm June 1st - October 31st + daylight savings time!
peak_wh = peak_wh+wh(n);
end
end
end
peak_kwh=peak_wh*5/7*1/1e3 % peak kwh is only on week days
non_peak_kwh = sum(wh)/1e3-peak_kwh