Infinite Coke Demand because of division by 0 leading to infeasible solution

[Timon-R]

I used a forked version of the newest PyPSA code that has some additional features which, however, shouldn't be cause of the bug. (https://github.com/Timon-R/pypsa-eur/tree/model_testing version:7bad187)

Bug description

The coke demand for NL is inf leading to an infeasible solution.

The proximate cause is infinite values in the resource industry_sector_ratios_2040.

I could solve/avoid the error by replacing the line:

today_sector_ratios = demand.div(production, axis=1)

in the script build_industry_sector_ratios_intermediate.py
by

today_sector_ratios = demand.div(production, axis=1).replace([np.inf, -np.inf], 0)

(also importing numpy in the header)

I am not sure if this fix is elegant or if there are deeper issues leading to this error.

[lisazeyen]

Hi @Timon-R this is probably related to our update of the JRC-IDEES and could happen if you use another reference year then we tested for in the config (year=2019). If the production in one country is zero this lieads to infinity values for the today_sector_ratios. I will adapt the code as you suggested!