Negative shadow price for Global CO2 constraint (also in linopy model )

[martavp]

The CO2 shadow price associated with the 'GlobalConstraint-CO2limit' is negative in the postnetwork obtained from pypsa-eur. I checked the linopy model m.dual['GlobalConstraint-CO2limit'] and it is also negative there.

Actually, I realized that for the m.dual['GlobalConstraint-CO2limit'] a long array is saved. I was expecting to get only one value as I do when I ask for m.dual['GlobalConstraint-co2_sequestration_limit'].

You can access my linopy model here and reproduce this by

m.linopy_read_netcdf(filename)
m.dual['GlobalConstraint-CO2limit']

I'm using the (almost) latest versions, in particular
python 3.11.7
pandas 2.1.4
pypsa 0.26.2
linopy 0.3.2
xarray 2023.12.0
so the fix implemented in September is included. I also checked the release notes and this does not seem to be fixed in the version of the past few days.

@FabianHofmann Would you mind checking if you are also getting negative values for the CO2 shadow price? thanks!