Electricity Generation Through Enhanced Geothermal Systems

[LukasFrankenQ]

Addresses Issue PyPSA/pypsa-eur#541.

This PR adds potential for electricity generation through enhanced geothermal systems (EGS) as in this paper. The data's resolution is on the level of countries.

EGS draws heat from great depths (1-10 km) to achieve the temperatures of 100-150 C required for electricity generation through steam turbines or binary cycle power plants.

The EGS heat can also be used for (duh) heating. This is currently not implemented, as we are still in contact with the authors of the paper to obtain the relevant cost estimations. However, in terms of implementation both electricity and heat would be quite canonical, and in that sense I thought it sensible to move forward with the electricity part and create a new PR for heat at a later stage.

Open questions (that probably should find good answers before merging):

1. Currently, the egs-scripts draw from egs_costs.xlsx (20kb) and egs_global_potential.xlsx (500kb), which are simply stored in data/. Is this okay or should they be added to the databundle?
2. How EGS electricity changes the system is still under investigation. Its costs are assumed to be subject to a strong learning rate, and a naive (and temporally sparse) run with 2050 costs builds egs capacity up to the posed constraints in Germany, Austria, Switzerland and parts of Italy. I am currently setting up the infrastructure (i.e. the Edinburgh cluster) to test the implications more rigorously through experiments with multiple investment periods. Results will be summarized in a short report.
3. The implementation uses pycountry, which is a tremendously slim package, which I would love to add to the environment.yaml. I suppose I would make a PR at pypsa/pypsa-eur for this?

Possible Improvements:

1. Heat becomes available through drilling at various depths. Through the cost of drilling, potential becomes a function of LCOE, and the paper shares its data thus as: x MW available at LCOE <= 5 Euro/MWh, y MW available at LCOE of 5-10 Euro/MWh, etc.... The present implementation coarsens this binning, to the LCOE steps 0-50, 50-100, and 100-150 Euro/MWh. To accommodate the linear nature of the model, each of the bins is implemented as a separate generator sitting on the same bus. Our freedom with regards to choice of bins is limited, as it is dictated by the data provided through mail exchange with the paper authors. However, I am happy to discuss how we can obtain capital and marginal cost from LCOE, to make our own choice of binning, it should be possible while respecting the different discount rates etc..

2. The implementation of heat and electricity could potentially be coupled, since EGS can lend itself to generation through CHP. Landau is such an example, where hot water from the well at 160 C is used for electricity generation and the residual heat with temperature 70-80 C for district heating. See here, here, and here.

Big thanks to @lisazeyen for tips along the way!

[LukasFrankenQ]

Also, I do not know why github thinks I made a whole new Snakefile...

[lisazeyen]

@LukasFrankenQ thanks a lot for your PR!
We should add the additional data to the data bundle and not the the github repository. Can you convert the excel to an csv and change the code accordingly?

[LukasFrankenQ]

@LukasFrankenQ thanks a lot for your PR! We should add the additional data to the data bundle and not the the github repository. Can you convert the excel to an csv and change the code accordingly?

Absolutely! Should I send them over to be added on zenodo?