Cannot combine System and Unit Ecoinvent databases

Question

Hello,

I have a significant problem. I have been working with combined databases of Ecoinvent (System and Unit), unfortunately from 3.11 version i am not able to combine them. I saw that there were a problem with 3.9.1 version with overlapping UUIDs. I checked 3.11 version and the same problem exists here. What can i do in this case? Is there any tool to change UUIDs globally for entire database?

Answer 1

Which system model and which Nexus release version of ecoinvent 3.11 are you talking about?

The latest Nexus release from 31.01.2025 should have random and different UUIDs numbers for processes in the system process databases so that it can be merged with the unit process database. However, I would still suggest to only use the unit process database, since copying, backups or searches will be much quicker and there is very few arguments to work with the system process version alone or combined.

Comments

I  am talking about 3.11 version with EN 15804 addon from April 2025

Thank you for suggestion, but from my perspective it is very valuable to use both of those versions combined :)

---

Thank you for reporting it. In the default ecoinvent 3.11 database we had already new random UUIDs for the system databases so that users can merge system and unit if they want to. For the EN15804 add-on system database, you can find now a new uploaded database with the version date from today (2025-06-13) that contains completely new and random UUIDs for the processes so that you can merge it with the unit process database.

There is only very few customers merging unit and system, and sometimes it is a relict from the past (business as usual) where the calculation for unit processes was slower.  Can you tell me some of the advantages that you see by mixing the system and unit process versions? Is it for quicker clicking / quicker screening through some impact results?

---

Thank you very much for the support, really appreciate that. There are 3 key aspects:
- i have a customers where i have tens of process steps along with tens of different flows - that leads to the longer calculation time for unit "approach"
- when i calculate emissions with unit data i want to see which flows generates highest emissions, but when i use unit flow i don't see the flows i introduced but their components (such as heat instead of brick).
- on the other hand i use unit processes mostly for investigating emission sources (verification of dataset completeness, source of impacts)

That is why i prefer to use those two approaches simultaneously