I managed to barely squeak through October with a net monthly surplus of power, just 5.68 kWh (for reference that's about a day's worth of usage). November will almost certainly be chipping away at my current annual surplus. Artisan Electric stopped by towards the end of the month to do my first annual inspection and cleaning and found that everything is right on track compared to their pre-install estimates.
Solar production, grid purchase/sale, and use.
The heady days of summer surplus are clearly behind us now, with my percent of power from solar dropping to 102% and each panel only generating an average of just 12.67 kWh for the month. In comparison each panel generated about 34 - 37 kWh in July.
Percent of household energy use produced by solar.
Cumulative production by panel (DC).
Production was slightly above estimates for this month, which is always nice to see. I produced 164.43 kWh versus the estimate of 158 kWh. For this month I also updated the installer's estimates with more accurate numbers (I'd estimated the previous numbers from a chart) and also corrected them to account for the system degradation over time. There's not a lot of change this year, but it's expected that production will drop about 0.5% per year.
Comparison with installer's estimate.
The daily numbers show that production is much less reliable, particularly towards the end of the month. October 22nd was particularly gloomy; I only generated 10% of my own power that day (just 0.49 kWh).
Daily solar production, grid purchase/sale, and use.
Percent of household energy use produced by solar.
This month the panels showed a bit more variation in production on sunny days (vertical smears in the chart below). On cloudy days the production is pretty uniformly bad.
Energy by panel, broken out by day.
Looking at the monthly totals for each panel it's clear that A1 and A5 are under-performing the others. It's important to remember, though, that the 10 - 12% lag is only about 1.5 kWh over the course of the whole month. Total production is low enough that small absolute differences look much more significant than they might otherwise be. The lower numbered panels are on the western end of the roof and are more subject to shading from a nearby tree than the others.
Lower numbered panels are lagging a bit.
Significant difference emerge when comparing to the average.
It's interesting to see which days had inconsistent sunlight throughout the day. For example, October 8th seems to have been relatively sunny in the morning but cloudier in the afternoon while a few days later the opposite was true.
The hourly production chart is still one of my favorites. The effect of shorter days, lower sun angle, and worsening weather is very clear when comparing the curve for October to the sunnier months. The best producing hour this month is only as good as the mornings or evenings in the summer. It's interesting to see that the peak is already closer to noon than 1:00, even before the time change off of DST.
Production by hour for the past 6 months.
The 2014 financials continue to look good, with the state incentive payments for October coming in at about $89. My bill credit this month with Seattle City Light will be a measly 29 cents.
Conversion efficiency is starting to show the characteristic curve of lower efficiency on days with less total power. I've pulled the last 11 months of this data and will be writing a separate post about it soon.