This projet is intended to download and preprocess the data for german power plants, especially renewable energy plants which is available to download from the official Markstammdatenregister part of Bundesnetzagentur for a possible import to OpenStreetMap. The data is downloaded using open-MaStR and is processed with pandas and geopandas e.g. like in the given examples using custom query strings. Note that most of the detailed data is not acutally used as it's not relevant in this context. Since 2023 the data can be used under the OSM licensing terms, see Permission_OSM_MaStR. As usual, no permission is granted to actually import any data using this tool without consulting the community (and me, the author) beforehand. OSM data is downloaded and filtered using pyrosm and osmium.

Note that the dataset for solar and storage is very large and takes a long time to download, because it also includes millions of small home installations. The full data can also be viewed after downloading using e.g. a regular sqlite viewer etc where all columns and possible values can be investigated. Also see documention of open-MaStR and of MaStR for reference.

This script mainly uses the german names, but these are translated at the end. Refer to OSM wiki and the documentation of Markstammdatenregister for matchingthe names/tags between them. Relevant data is start/end/opening date of a plant, location, power output and for wind turbines e.g. manufacturere, height, rotor diameter. Others can be added/removed as required. Most relevant is the 'MaStR-Nummer der Einheit' (OSM ref:mastr) in the form of SEExxxxx that can be used to cross-reference with the original data and therefore allows to update plants over their lifecycle. This is issued for every individual generator(!) and can serve as a unique id.

Note there's also 'EEG-Anlagenschlüssel' (OSM ref:EEG) in the form Exxxxxxxxxxxxxxx, but this is only issued for renewable energy plants and only after the validation of the grid operator which can take a few months even after the plant is already in operation, so this is not ideal to use for cross-referencing. Similar there's also 'MaStR-Nummer der EEG-Anlage' in the form EEGxxxxx which is used to identify the whole plant e.g. to group multiple generators at the same location together.

Some plants e.g biomass, waste and non-renewable have even more additional specific values e.g. 'KWKxxxx' for Kraft-Wärmekopplung.

So it is generally recommended to primarily use the 'SEExxx' value as this doesn't depend on the energy carrier and avoids confusion with different ids. But since ref:EEG was/is still used with OSM it can be included to potentially match existing tagging. For plants where only the whole plant is mapped because it is not obvious how to differentiate on the ground this might be included, too (e.g. biomass plants often have multiple generators in the same building). These plants have multiple "SEExxx" numbers attached to it in the dataset.

To execute install the locally using pip install -e . This installs the requirements and the provided utilities in editable mode locally.

The workflow usage is something like the following:

• Download data for one energy carrier using the provided Mastrdata class
• optionally define additional columns and their translations as dict depending on the energy carrier
• optionally define and apply custom query string filter depending on the energy carrier
• using the provided DataFilter to apply general filters not depending on the energy carrier
• using the provided PostProcessing to e.g. adjust names of manufactureres, format power values as needed and in the last step rename the columns to match OSM
• print data, plot map etc.
• compare to existing OSM data of some area using geopandas sjoin_nearest and matching on some existing osm key=value for confirmation.

Some filters can be used (see the help page of script) directly as options, others can be added via custom query strings "key1 = 'value1' and/or key2 = 'value2' and/or ...."

Optionally the queried data can plotted on a map, to compare the data with existing OSM data. Another option is to plot the distance of the MaStR-Data compared to existing OSM data. There are some weird cases where ways and not nodes were used to map plants. Specify column to match on, or if not given show only missmatches outside max_distance. Some example maps and a table overiew are included.

A utility to show cases where OSM tags aren't formatted in standard ways is also provided. Further a small utility for directly querying selected values from the downloaded MaStR database by known ref numbers (e.g. Exxxx, SEExxxx ,KWKxxx) is also provided.