As mentioned on the home page, I have a photovoltaic installation on my roof and I keep a close eye on it. The entire installation was done by a electric utility company. But even though the system came with a monitor system and logs data to an online portal, I have written a few small programs in Python to log data into a database and display some of it. Since installation, I have been logging 10 different parameters every minute. The documentation for the inverter actually has the log file format explained so I could extract more parameters if I wanted. I also wanted to use a portal that the software in the inverter didn’t support so I wrote another script that exports daily numbers from the database and uploads them to the portal a couple of times each day.
This is a live graph of the output for the last 24 hours. Updated every 10 minutes between 05 and 22. I tried different graph modules for Python but ending up just feeding the kilowatt output data into RRDtool.
Click here for a larger more detailed version with one pixel pr. minute.
One of the programs generates a webpage for viewing on my smartphone so I can keep an eye on it from anywhere in the world where you can get a data connection.
I also has this status monitor in the hallway / corridor. It’s a picture frame with a special USB key inserted. The key generates 5 images with data it gets wireless from a monitor box connected to the ethernet and the picture frame switches between these images. The monitor box can collect data from up to 10 inverters.
Pictures from the installation
The inverter with the front open. The electrician are installing the 3-phase 400 Volt output, connecting the DC input from the photovoltaic generator(the solar panels) and the ethernet connection for monitoring.
Only 3 more panels to go
The two upper panels to the right of the chimney gets a shadow creeping across them in the afternoon around 16:00,17:30 18:45 and can be seen as a loss on the graph, marked with #1. Usually before it gets cloudy, the wind picks up and cools the panels better(#2) and when the sun breaks through, the panels produce better for a little while until they heat up again(#3). I am not sure what happened at #4, I am guessing it must be wind still cooling the panels.
It could be fun to get a temperature probe on the panels, measure ambient temperature and wind speed and add those to the graph. Kostal has a add-on kit for the inverter that measures the temperature as well as a reference measurement of radiation from the sun. Maybe I can find that at a reasonable price somewhere.