Possible overestimation of co2 emission of oil loads

[energyLS]

Checklist

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

Describe the Bug

At low or zero emissions scenarios without carbon sequestration, the model produces oil via Fischer-Tropsch and does not use it.
A possible explaination for this might be:

• The implementation of oil demands and the linked co2 emissions are not always correct, probably the co2 emissions are overestimated at some point so a perfect carbon cycle of oil generation (Fischer-Tropsch) and oil demand does not exist.

Possible solution

Check the co2 calculation of these oil consumers:

• Aviation
• Shipping
• Industry (Coal/oil)
• services
• agriculture
• residential oil

by searching (strg+f) for oil emission and checking wether the oil demand and emission is linked correctly by checking the emission factor costs.at["oil", "CO2 intensity"], temporal resolution (8760 division?), spatial resolution, ...
It should always be like this: The annual oil demand in p_set of oil load * costs.at["oil", "CO2 intensity"] should equal the p_set of emission load.

One example is e.g.

pypsa-earth-sec/scripts/prepare_sector_network.py

Line 744 in 0b58848

co2 = airports["p_set"].sum() * costs.at["oil", "CO2 intensity"]

[energyLS]

Aviation: 1437 MWh per hour / 373 t per h
Shipping: 219 MWh per hour / -> (`shipping_oil_share` should be removed)
Industry (Coal/oil): demand unclear / 0 t emissions 

land transport: 7 TWh per year / 210 t per hour
services: 0.42 TWh per year / 12 t per hour
agriculture: 973 MWh per hour / 253 t per hour
residential oil: 15 TWh per year / 456 t per hour

Other emissions somewhere else?

• Other generators?
• biogas upgrading?
  Next:
• check solved network

[energyLS]

Checking the co2 balance of a net-zero solved network reveals: There are biomass emissions which should not be there. Biomass is implemented as "neutral" carrier, not deducting co2 from the atmosphere nor emitting into the atmosphere.

Code investigation reveals:
These biomass emissions are added here, but this seems to be an error.
https://github.com/pypsa-meets-earth/pypsa-earth-sec/blame/6b6f04f1984b6c1293f00acf64e0b42bd79636a0/scripts/prepare_sector_network.py#L2217-L2223

[energyLS]

All issues tackled here: #245