Six Years of Solar PV

[jebatty]

Six full years of solar PV are now history. The graph shows the results. 2019 set a record annual low, ending up slightly lower than 2016, the prior low year. 2019 year was plagued by smoke from Canadian and western US forest fires and day upon day of cloudy/rainy weather.

I'm leaving the average line at the five year period 2014-18 rather than have a rolling forward average. This is just an intuitive sense that the first five years should be years when the panels are producing the highest with minimal degradation. Panel specs that I have seen assume a 0.5% annual degradation. At this point I don't have the skill or measuring ability to actually plot degradation impact, but over time and as more years pass it may be possible to see degradation by the data I do collect. Day to day weather impacts annual production significantly. For example, seeing degradation is made more difficult by the impact of as little a single day with clear skies, with the maximum daily production I have seen being just over 90kWh, and another day with dense clouds producing just a few kWh. That single "HI" day vs "LO" day alone can impact annual production by about 0.5%.

 

[peakbagger]

Seasonal "noise" is far going to exceed degradation. Only way to do degradation would be to pull panels and use a standardized light source. Manufacturers know this and that is why they extend such long guarantees on degradation Highly unlikely someone will be able to prove it even if the manufacturer is still in the solar business. Most of the time I hear claims of underperforming systems its usually external like shading caused by landscaping. I know I have one 17 year old array that suffering from that problem, the small trees off to the sides have grown a lot taller in 17 years. Generally the actual silicon chip electrical conversion isnt the issue, its encapsulant deterioration. I have a 20 year ARCO/Shell/Siemens/Solar World) panel on my SHW system pump where the encapsulant is cooked, not as bad as the old ARCO crispy critters quadlams that gave a lot of folks their start in solar long ago, but still quite cloudy. It still does what it needs to so I am just going to run it until it does not.

Not that hard to fire up excel and generate some best fit curves for your past production and use that to compare current production over a discrete period compared to the historical curve.

 

[Brian26]

Solaredge monitoring with optimizers lets you look at very detailed individual panel performance. You can graph out each panels output and compare it to other panels. Here are my total outputs for each panel. The east panels get less sun in the morning so have slighty less output. It would be very easy to see if a panel is losing its performance.

The 2nd graph is an example of someone using the software to graph and compare panels. The blue line is a supposedly a bad panel with significantly less output visible.

 

[jebatty]

I have similar ability with my ABB microinverters to monitor/compare/graph each of my 46 panels. With some regularity on days of full sun I compare the output of each of the panels to verify each is producing about the same watts as the others. Variation is very minor.