Deal with 3-hourly-resolution in pypsa-eur

[ClaraBuettner]

The currently running pypsa-eur results created by @ulfmueller consider every 3rd hour of the year. That is fine for things like installed capacities.
But we also use some time series data from pypsa-eur (e.g. for load time series of foreign countries) which we need in an hourly resolution. To be able to do that, two prenetworks are build in pypsa-eur, one in a 3-hourly resolution (that is solved afterwards) and one in an hourly resolution (that is not solved and was only created for being able to solve this issue).

So the needed data from pypsa-eur is available. However, it still needs to be considered in the implementation of the interface from pypsa-eur to powerd-data.
From what I know, at least this function needs an update:

powerd-data/src/egon/data/datasets/pypsaeur/__init__.py

Lines 419 to 444 in 4ca597c

	def neighbor_reduction():
	network = read_network()
	# network.links.drop("pipe_retrofit", axis="columns", inplace=True)
	wanted_countries = [
	"DE",
	"AT",
	"CH",
	"CZ",
	"PL",
	"SE",
	"NO",
	"DK",
	"GB",
	"NL",
	"BE",
	"FR",
	"LU",
	]
	foreign_buses = network.buses[
	~network.buses.index.str.contains("|".join(wanted_countries))
	]
	network.buses = network.buses.drop(
	network.buses.loc[foreign_buses.index].index
	)

Currently, only one, solved network (with every 3rd hour) is used. But we would actually need to read in two and use some data from the solved network (e.g. installed capacities) and other from the prenetwork in an hourly resolution (e.g. load time series).

There might be other datasets that need adjustments like this. This one is only the one that directly came to my mind. As I am neither creating the pypsa-eur results nor running powerd-data, I think it would be great if you could have a look at this @CarlosEpia and @ulfmueller. In case I should support you, jest let me know :)

[ClaraBuettner]

The preapred_network should be also used here:

powerd-data/src/egon/data/datasets/industrial_gas_demand.py

Line 490 in 4ca597c

n = read_network()

(So far, static data is used to it would also work link this. But I think it is better to consistently use the prepared_network for loads)

Edit: Done in 0fbb926

[ClaraBuettner]

Not related to the 3-hourly resolution but the path should be also updated here (or better, use read_network)

powerd-data/src/egon/data/datasets/chp/__init__.py

Lines 595 to 604 in 4ca597c

	target_file = (
	Path(".")
	/ "data_bundle_egon_data"
	/ "pypsa_eur_sec"
	/ "2022-07-26-egondata-integration"
	/ "postnetworks"
	/ "elec_s_37_lv2.0__Co2L0-1H-T-H-B-I-dist1_2050.nc"
	)
	network = pypsa.Network(str(target_file))

Edit: Done in f100620

[ClaraBuettner]

Only used for eGon2035 but to be sure, we should also use the prepared_network function here:

powerd-data/src/egon/data/datasets/gas_neighbours/eGon2035.py

Lines 642 to 651 in 4ca597c

	target_file = (
	cwd
	/ "data_bundle_egon_data"
	/ "pypsa_eur_sec"
	/ "2022-07-26-egondata-integration"
	/ "postnetworks"
	/ "elec_s_37_lv2.0__Co2L0-1H-T-H-B-I-dist1_2050.nc"
	)
	network = pypsa.Network(str(target_file))

Edit: Done in fcc026e

[ClaraBuettner]

Solved in the run-status2019-and-egon100 branch, can be closed when that branch was merged to dev