+2 votes
644 views
Hello,

when using Monte Carlo (MC) simulation, the allocation factors defined in a process in openLCA remain fixed, even if uncertainty distributions are defined.
A simple example:
A process yields two products, A and B, and only emits CO2. Based on the output of products, allocation factors are calculated, e.g. 0.5, 0.5. Additionally, uncertainty distributions are defined for the output quantities of A and B.
When running MC analysis, the relative share of outputs varies according to the defined uncertainty distributions. Consequently, allocation factors should be adapted accordingly for each MC run. However, openLCA keeps the pre-calculated/-defined allocation factors fixed.

The same applies for product outputs defined by parameters.
If the output of product A in the example is C, and the output of product B is defined as C*D (C and D are parameters to be varied in the sensitivity analysis using the project function in open LCA), openLCA keeps the allocation factors fixed, even if several variants of C and D are defined in the project function of openLCA.
E.g. In the process inventory, we define C=1 and D=9, the allocaton factor of A and B are 0.1 and 0.9, respectively. When using the project function in openLCA other variants for C and D can be defined. However, whatever is defined, the allocation factors remain 0.1 and 0.9.

I posted this problem 2 years ago (!) in the former forum, but the problem was not fixed since then in any update of openLCA: https://archive.openlca.org/viewtopic.php?f=24&t=25729&p=29595&hilit=Monte+Carlo#p29595

Is there any way to solve this issue?

Thank you for your help
in openLCA by (150 points)

1 Answer

+2 votes
by (13.5k points)

This is correct. Allocation factors are currently static numeric values in openLCA. My favorite solution to make them dynamic would be to also allow formulas (with parameters that represent product amounts) for them.

To track this, I also created this issue on our Github repository.

...