Negative exchanges in Vanadium datasets

[dbantje]

I've noticed that I got negative scenario-wide extraction of Titanium (using pathways):

I've traced back the issue to redox-flow Vanadium batteries (which mostly make up the orange bars on the top). When looking into lci-batteries-vanadium.xlsx I find the following negative exchange:

In a resulting premise-generated database, I also find some negative biosphere flows for Iron and Titanium - but I guess those are the result of post-allocation correction?

In any case, maybe there is just that one error in the imported database? I'll check what happens if I change that one and will report back.

[dbantje]

Based on aggregate pathways results, just changing the sign seems to give reasonable results:

I don't know whether any of the code or data in metals.py is based on the inventories.

[romainsacchi]

@alvarojhahn, do you remember anything about this negative sign? Can we assume it is an error? How should it be corrected?