HVDC links being removed in simplify_network.py

[paulregg]

Checklist

• [ X ] I am using the current master branch or the latest release. Please indicate.
• [ X ] I am running on an up-to-date pypsa-eur environment. Update via conda env update -f envs/environment.yaml.

Describe the Bug

Hi Fabian,

I was reading your paper on hydrogen networks in PyPSA-Eur-Sec, and I noticed that my model runs install way less wind generation in Ireland than the models in the paper. I then noticed that no HVDC connection exists in my network between Ireland and Britain, which I think would be the reason for this, since all the electricity generated would have to go through Northern Ireland and Scotland.

I've been struggling to figure out why this HVDC connection is removed. I checked, and in pypsa-eur, it exists in base.nc and elec.nc, but is removed in elec_s.nc. Therefore, it must be happening in simplify_network.py, probably in the simplify_links() function, but I can't figure out why the connection would be stripped here.

I don't know if this is something you've seen before, and it's probably difficult without seeing what's happening on my machine, but I thought I'd check if you have any advice. I'm currently rerunning the pypsa-eur to recreate elec_s.nc, but it's taking a long time because it triggered a recreation of renewable resources.

I'm using wildcard simpl = ""

The link has index '14815' in simplify_network.py, node 5937 is found to be a stub and deleted, which results in the DC link being deleted.

I've also checked, and node 5937 is not a stub, so I don't know why this is occurring. It turns out node 5935 in elec.nc is a stub, with only one connection (to 5937, the GB side of the IE-GB link). Therefore, in the busmap_by_stubs() function, first 5935 then 5937 is added to the stub list.

I've checked, and 5935 is a stub in base.nc as well.

I haven't looked in detail into other links, but by comparing my results and the 2022 Hydrogen network paper, I notice several that might be removed as well: Scotland - Norway 3rd cross-channel UK-FR connection (near Portsmouth) UK-NL Denmark-NL UK-Denmark

from config.yaml :

clustering: simplify_network: to_substations: false # network is simplified to nodes with positive or negative power injection (i.e. substations or offwind connections) algorithm: kmeans # choose from: [hac, kmeans] feature: solar+onwind-time

links: p_max_pu: 1.0 p_nom_max: .inf include_tyndp: true under_construction:

Error Message

No error message, since network still functions normally without the missing links

[pz-max]

Hi @paulregg , sorry for the late response some of us were on holiday. Thank you for the bug report. Seems like something is going wrong. We will discuss the issue and provide you soon with more feedback.

[fneum]

closed via 27d4af8e6574d8b39f1a442caa37d896ff1ae876

was indeed a bug related to the new ENTSO-E grid extract, required a few manual corrections