Hello!
I'm adding an issue to summarize the ideas on the architecture to facilitate coordination.
This issue is meant to collect feedback on the (a) inputs, (b) outputs, (c) methodology and (d) configuration options.
Feel free to comment and integrate relevant points into this message: the list can ideally be updated to track all relevant ideas that appear.
Motivation: validated grid data for power-system models using open data
Inputs
- Open Street Map data [main requirement for minimum viable product]
- the architecture shall support additional custom data with standard data format
Outputs
- Substations
- Lines
- Cables
- Converters/Transformers
- Their graph representation (basically substation ids for lines/cables)
- Electrical parameters as appropriate (Line types / impedance values)
- Rated power capacity
- Optionally visual representations of the resulting network
- While we start with HV lines, we can consider a modular architecture to support LV/MV in the future
Methodology
We agree on the conceptual steps of the model
- download data
- cleaning/harmonizing the raw data
- building the graph-representation of the network
- validation of the result: estimating error-metrics against known datasets and track changes in results over time.
pypsa-earth-status (name could be revised) can support on the validation
Options
- Regional scope: it shall be global as well as regional
- Line Types shall optionally change by region
- The cleaning steps must contain multiple heuristics, that can be optionally used for cleaning
- The building step shall require heuristics on how to merge the entries (e.g. using substation geometries, spatial tolerance, relations, etc)
Looking forward to hear your opinion!
I give permission to all maintainers to edit this message to track all relevant information that may follow in the discussion.
Hello!
I'm adding an issue to summarize the ideas on the architecture to facilitate coordination.
This issue is meant to collect feedback on the (a) inputs, (b) outputs, (c) methodology and (d) configuration options.
Feel free to comment and integrate relevant points into this message: the list can ideally be updated to track all relevant ideas that appear.
Motivation: validated grid data for power-system models using open data
Inputs
Outputs
Methodology
We agree on the conceptual steps of the model
pypsa-earth-status (name could be revised) can support on the validation
Options
Looking forward to hear your opinion!
I give permission to all maintainers to edit this message to track all relevant information that may follow in the discussion.