[Solved] negative CML impact factors correct?

[gebbissimo1]

Hi everybody,

I was calculating some CML indicators for the ecoinvent v2.1 process "pig iron, at plant". However, some of the indicators were negative (human toxicity, freswhater and marine sediment toxicity). Please find below a screenshot of the sankey diagram.

To me, these results look rather suspicious. Since however the supply chain is rather complex I was wondering whether anybody could tell me, if...

• ... those results are indeed correct. I already found a few more processes with negative CML impact factors. If so, could anybody give a brief explanation why the creation of pig iron or similar processes should have negative impact factors at all?
• ... it is due to some bug in openLCA, e.g. using ecoinvent v2.1 instead of ecoinvent v2.2 data (according to this post: http://openlca.org/forum/viewtopic.php?f=18&t=15699)

Clarifications are very much appreciated, thanks in advance
Christian

[gebbissimo1]

//update

It was indeed an error related with the interface between ecoinvent v2.1 database and openLCA. Ecoinvent v2.2 yields positive CML impact factors for all categories