How to overcome linking issues in worldsteel LCI "steel plate" process?

Question

Hello!

I'm concluding an LCA for a machine that is mostly made out of steel parts. I am using worldsteel database for steel parts. There seems to be some linking issues with the "Steel plate Global 2020, production mix, at plant, blast furnace route, 1kg, typical thickness between 2 to 20 mm. The maximum width is 1860 mm". It gives for example when using ReCiPe2016 midpoint method a total of ~600kg of CO2 per 1kg of manufactured steel plate in category GWP. When examining the issue, there are many links for example from used process water, steel scrap etc. that link to different steel products and for some reason the quantity needed from them is many times higher than in reality. This does not show up in the model graph but the issue comes up when calculating the results.

I have tried to use different providers based on what seems to affect the result but this has not changed the situation.

Any ideas on this?

Answer 1

Hi Aleksi,

indeed, these are GaBi datasets originally, and not meant for the autoconnect feature. You should not link it to other processes automatically. The many products that go in and out of these aggregated datasets are probably due to a property of the GaBi calculation algorithm that struggles with looped system and needed to stop the iterations at one point, but this is my interpretation.

Best wishes,

Andreas