0 votes

Unfortunately, I cannot merge ecoinvent database.  When I import ecoinvent_391_cutoff_lci.zolca and ecoinvent_391_cutoff_upr.zolca as two standalone database, it works. However, when I merge ecoinvent_391_cutoff_upr.zolca to  ecoinvent_391_cutoff_lci.zolca, it failed and nothing happened. 

And when I export ecoinvent_391_cutoff_upr.zolca into JSON-LD, it also failed and got the error message 

Failed to export data sets

# Installation details

* openLCA version: 2.1.1

* operating system: Windows 10

* os architecture: amd64

* os version: 10.0

* native libraries loaded: false

* with sparse matrix support: false

# Full error stack trace

in openLCA by (140 points)

1 Answer

0 votes
by (5.6k points)
Yes, our very first Nexus release from ecoinvent 3.9.1 had the same UUID numbers for processes in the unit-process database and in the system-process database, so if you merge them, nothing will happen. When you download the subsequent Nexus release version 2 or newer ("n2" inside the file name or newer) we have changed the UUID numbers so that both databases can be merged into one. As a suggestion, I would import the smaller unit-process database into the larger system-process database and not vice versa.

But may I ask, what is the advantage of merging them, since it is two exactly equal databases, one with the full supply chain (unit-processes), and one fully aggregated (system-processes). There is no calculation advantage anymore, since the calculations for the unit-process database are very fast now in openLCA 2. It is a matter of seconds when using openLCA 2 and selecting the lazy/on-demand calculations. Maybe I'm missing other advantages of merging them. Merging also comes with some small drawbacks that, for example, search functions or finding linkings / usage might be slower, due to the large amounts of exchanges. My suggestion is to work with the unit-process database only and using the latest openLCA 2.2 release.