Sunday, 1 March 2015

(re)discovering the year

We've had our solar PV system for over a year now.  One of the many plots we make is its output, measured every five minutes through the day, for each day:

(The horizontal time axis runs from 3:00am to 9:00pm GMT, and the vertical axis runs from January 2014 to February 2015.  The colour indicates the average production in kW in the five minute interval.)

Hey, there's a pattern!

We started generating in January last year, as the days were getting longer.  The changing day length is clearly visible.  Where we live, sunrise to sunset is a mere 7 hours 41 minutes on 21 December, and a whopping 16 hours 48 minutes on June 21.  I got those times from a website, not our generation data.  The system doesn't switch on exactly at sunrise and switch off exactly at sunset: it depends on clouds and scattering and stuff.  But the sinusoidal trend over the year is clear from the chart.

We get more power in June not only because the sun is up for longer, but it also generates more at peak, because the sun is higher in the sky, shining down fully onto the panels, not at an oblique angle.  So last June we saw this average over the month:

(Here the horizontal time axis runs from 3:00am to 9:00pm GMT, and the vertical axis runs from zero to 8kW.  The orange regions indicate the minimum, lower quartile, median, upper quartile, and maximum generation at that time, over the month. The black line indicates the monthly mean generation at that time.  The number is the average kWh production per day.)

The statistics show how variable the generation is.  On the best June day we got 53.63 kWh, but on the worst June day generated only 18.35 kWh.

Come December the generation plot looked like this, narrower and lower, with only a third the average generation:

There is still massive variation, from 27.63 kWh to a minuscule 0.34 kWh (it barely seemed to get light at all that day!)

But even an average of 12 kWh/day is a lot better than nothing!

No comments:

Post a Comment