manfredherrmann / awattar-statistics

simple visualization of aWATTar HOURLY dynamic prices

Statistics for aWATTar hourly tariff

There is a new, very interesting offer in Germany for a dynamic power contract. The company is called aWATTar and offers an HOURLY tariff which is based on the EPEX Spot Day-Ahead price. I don't have a smart meter yet and only know my yearly power consumption. So I wasn't able to use the Lohnt sich aWATTar? project.

I wanted to be able to get some rough idea about the expected costs of aWATTar HOURLY and therefore I created this visualization script.

Data Source

The source of this data is the official (German) aWATTar API. Please note that this price is without VAT. So you will need to add 19% in order to get the actual prices for customers.

All data is only collected once and then cached. For convenience, this repository comes with the pre-downloaded data from 2013-12-22 until 2021-09-24.

Usage

In order to get the latest daily data, please run ./awattar-statistics.py update first. This will download and append the missing data to the file historical-data.txt.

Then run ./awattar-statistics.py calculate which will create all the files and visualizations. This will take a few minutes of generation.

Output

This script outputs a few CSV files which you can use for further processing, for instance Libreoffice Calc or Excel. The following three outputs are created:

• daily prices (data/daily/2020-04-22.txt): prices per hour, format is {hour};{price}. Please note that there might be 25 or 23 hours on days where daylight saving changes, i.e. there might be no hour 02 or two hours 02a and 02b.
• monthly minimum, average, median and maximum prices (data/monthly/2020-08-*.txt): Those files contain the hourly minimum, average, median or maximum prices for the whole month.
• yearly minimum, average, median and maximum prices (data/yearly/2020-*.txt): Those files contain the hourly minimum, average, median or maximum prices for the whole year.

There are three types of visualizations:

• daily charts (plot/daily/2020-04-22.png): bar chart for a single day
• monthly percentile charts (out/plot/monthly/2020-08.png): percentile chart for a month
• yearly percentile charts (out/plot/yearly/2020.png): percentile chart for a complete year

I find the monthly percentile charts the most interesting. Those look like the following: This graph shows the percentiles for each hour:

• The minimum 12:00-13:00 price in July 2020 was -6ct per kWh.
• 10% of the days, the 12:00-13:00 price was below zero, 90% it was above.
• The median price for 12:00-13:00 was a little below 3ct per kWh. 50% of the days, the price between 12:00 and 13:00 was therefore below 3ct per kWh and the other half of the days, the price was above.
• The maximum price between 12:00 and 13:00 in July 2020 was around 5.5ct per kWh.

Contributing

If you want to have another visualization, please feel free to open an issue or - if you can - a pull request.