search

highRES-model / highRES-Europe-WF

https://highres-europe-wf.readthedocs.io/en/latest/
MIT License
0 stars 2 forks source link

Possibility to improve hydropower representation through better data #9

Closed OskarVagero closed 2 months ago

OskarVagero commented 9 months ago

Marianne recently mentioned The Energy and Industry Geography Lab and while exploring this I compared the hydropower database to ours. The installed capacities are different (not by a large extent), but could be worth comparing more in detail and especially for disaggregation, as the file contains some spatial information beyond country.

OskarVagero commented 9 months ago

Data documentation: https://joint-research-centre.ec.europa.eu/system/files/2023-11/Data%20documentation%20EIGL%20v1_6.pdf

OskarVagero commented 4 months ago

jrc_highRES_hydro_comparison

It is worth noting that there seems to be no reported capacities in Denmark, Estonia, Greece, Luxembourg and the Netherland.

OskarVagero commented 4 months ago

Looking at it specifically for Norway, where we have a very good comparison in terms of the Norwegian Water and Energy Directorate's (NVE) hydropower database: https://temakart.nve.no/tema/vannkraft

Units in MW

NUTS_ID NVE JRC abs_diff perc_diff
NO02 2577.13 2420.82 156.31 6.065274
NO06 1947.14 1834.07 113.07 5.806978
NO07 5385.44 5158.83 226.61 4.207827
NO08 4167.41 3405.40 762.01 18.284978
NO09 6348.22 6114.17 234.05 3.686860
NO0A 14969.39 14201.90 767.49 5.127063
OskarVagero commented 4 months ago

After discussing with @absolute-afloat, we think that the best play might be to use the JRC database (also available and maintained on GitHub, https://github.com/energy-modelling-toolkit/hydro-power-database). Unless we expect a large amount of decommissioned hydropower plants, the higher value seems more likely (easier to miss a power plant than to make one up out of thin air).

We can contribute to the hydropower database where we have better data (e.g. NO, SE and ES).

OskarVagero commented 2 months ago

This issue was resolved with PR #43. Closing issue.