TL;DR: No, please do not do this.
Introduction
When conducting life cycle assessments, practitioners often work with multiple LCA databases and wonder whether combining them would streamline their workflow. Whilst database merging seems logical, it typically leads to significant methodological errors that compromise assessment accuracy.
In some cases, yes, you can merge databases, but ONLY if the database instructions explicitly say so. For example, Agribalyse and Ecoinvent can be merged, or the various databases within LCA Commons are designed to connect. However, this is permissible with the caveat of ensuring you merge the exact versions stated.
Why shouldn’t I merge databases?
The problem arises from inconsistent flow names. For example, a flow of N2O might be called “nitrous oxide” in one database, and in another “dinitrogen monoxide”, or CO2 might be classified as “CO2“, “carbon dioxide”, or “carbon dioxide, fossil”. When applying impact categories, the characterisation factors will only be for one of these names for each flow. Hence, you end up ignoring perhaps half of the emissions, because half of them had a different name.
Is this a common mistake?
Yes, and by breaking the foundation of the modelling, the database used, there is some really good work out there that gives the wrong results.
Need LCA, EPD, or CBAM consultancy?
Or have a research proposal to collaborate on?
Global commercial consultancy • Horizon Europe, UKRI & Innovate UK research partner