0 votes


I have a question about uncertainty factors used in Ecoinvent’s Pedigree matrix, to convert data quality indicators into additional uncertainties.
As specified on Ecoinvent website, these factors are based on expert judgment, what is a bit vague. I haven’t succeeded to find more information about these expert judgment, so I would enjoy if you have some literature on that topic.
I’ve also read this article from (Ciroth et al, 2016), with interesting research about using empirical data rather than expert judgment. Was this way just dropped, and why ?

Best regards,

in openLCA by (150 points)

1 Answer

0 votes
by (93.1k points)
selected by
Best answer
Hi Emilien,

interesting question. The full report is available on our website: https://www.greendelta.com/wp-content/uploads/2017/03/Pedigree_report_final_May2012.pdf.

Why these factors are not offcially used I cannot say, but we certainly had a somewhat unfortunate timing. Our project was linked to the ecoinvent 3 release. This release brought many new ideas (the different system models, a new data format, a new IT environment for ecoinvent, market processes, and so on). The entire ecoinvent 3 release took much much longer than initially planned, with many draft releases that were not really stable. In the course of these delays, also a new management was established, and it is understandable that the idea of the new management was to rather consolidate and not to implement many additonal innovations. We are providing "our" empirical uncertainty factors as an additonal data quality system with the ecoinvent databases on Nexus, which means that you can use them. A downside is that they do not distinguish between different sectors, and arguably, technological and also geographical change varies across sectors, but also the official ecoinvent factors are not differentiated by sector.

So everything is currently a bit stuck, and using the old, still official uncertainty factors from ecoinvent for an uncertainty analysis is probably not more than an interesting exercise - but, I can tell that we are currently working on a refined version, thus hopefully the situation will improve.

Best wishes and thanks again for raising the point,

by (100 points)
Hi Andreas, I'm curious to know, beyond this publication/research are there any follow-up projects being undertaken and when are they expected to be completed?