Montecarlo Errate Values

Question

Good morning,
I am using OpenLCA 2.4.0 with Ecoinvent 3.10 Cut-Off and the EF 3.1 impact assessment method (version 2.7.3). I am experiencing issues with the Monte Carlo analysis, specifically with the GWP100 indicator: it returns values that are 100 times higher than the actual indicator value. This occurs across multiple studies; in particular, I am conducting LCAs on lime and on glass, and for both, the Monte Carlo simulation yields extremely high values from the very first iterations.
I have often encountered issues with uncertainty analysis in OpenLCA — do you have any suggestions on what I could do to resolve this?
Thank you.

Answer 1

Hm and why do you think this is a wrong value? The result of the simulation depends on the system you simulate, evidently, and e.g. if you have structures that are numerically unstable (subtracting one big amount from another and using this result in further calculations, e.g., but this is an entire topic on its own) then you can have quite different results in the simulation compared to the normal calculation. Or also of course if you have very high uncertainties.