Changes for external coupling of must-run interconnectors

[mabijkerk]

While reviewing documentation of must-run interconnectors in quintel/documentation#203, I started testing the functionality and ran into some issues. The changes I made include:

• Renamed all relevant inputs for clarity
• Moved inputs for FLH to the relevant external coupling input folder
• Fixed coupling groups for FLH inputs
• Added step value and priority to inputs
• Removed disabled by from regular capacity inputs
• Normalized base curves sum to 1/3600 instead of 8760
• Added base curves to other relevant datasets

Notifying both @kaskranenburgQ and @louispt1. Would be good to have a discussion about this and see whether we can align on how the functionality should work, perhaps I've missed something. There are a couple of to-do's still left before we can merge:

• Must: fix aliases to the nl2019 curves that don't work yet (see Semaphore error)
• Must: discuss whether using a capacity profile is correct
• Must: discuss (with client) wether to migrate inputs
• Nice: discuss whether to incorporate the must-run input in other inputs

Goes with quintel/etmodel#4372

[kaskranenburgQ]

After some testing I found that the profile that is uploaded functions as an availability profile and not as a capacity profile.
Furthermore, with the implemented modelling I can set a capacity of 1 GW, 8760 flh for export, and a profile that only has an export availability in the first week of the year.
This could lead to the following result:
The exported electricity on a yearly base is in this case 8760h * 1GW = 8760 GWh
However, this is all exported in the first week of the year, as you can see in the graph below:

Due to these settings, the actual output capacity of the interconnector is around 52 GW in the first week of the year.
Is this behavior adequate?
In a way it is, and if it is an expert function it might be okay.
If we leave this behavior in, we must explicitly document it.
@mabijkerk What do you think?

[mabijkerk]

Due to these settings, the actual output capacity of the interconnector is around 52 GW in the first week of the year.

Well-spotted, I think this is probably why we prefer using capacity profiles. I'll discuss this with the client.

[mabijkerk]

After discussing with the client we drew the following conclusions:

• There's no need for migration
• We should remove the FLH inputs: the goal should be to retrieve the FLH from the profiles. This would also prevent the issue noted by @kaskranenburgQ that states that the resulting capacity can exceed the specified capacity. This will however probably require us to update the export_must_run and import_must_run subtypes in January.
  • We should add this to the list of issues in Fix minor front-end issues with coupled must-run interconnectors etmodel#4345
• We can then merge this PR without the ETModel PR, since that only contains locales for the FLH inputs