What happens when allocation factors don't add up to 1?

Question

I am working with a database in which a multi-output process has outputs in several different units, some in kg, some in m3 (for some reason they are using m3*a, but I'll roll with it). Auto-calculating physical allocation factors doesn't work in this case, but the database author has manually entered values into the "Physical allocation" and "economic allocation" columns. Neither of these columns adds up to 1.

The user manual does not explain *how* allocation factors are applied during the LCI computation, so I ask here: is this approach viable?

(1) are the allocation factors normalized somehow so that the process impacts are fairly allocated according to the proportions specified?

(2) is it normal for "correct" allocation factors to not add up to 1 in the case that the reference flows have mixed units?

(3) regardless of that, shouldn't economic allocation factors *always* add up to 1?

More generally, can someone from GreenDelta explain how the allocation factors are applied to compute exchange values for the partitioned processes? The user manual describes how the factors are calculated, but not how they are implemented. I imagine it is straightforward, but it would be good for the algorithm to be stated explicitly, ideally in the comprehensive user manual.

Andreas Ciroth · Answer 1 · 2019-05-12T11:08:20+0000

Hi, physical and economic allocation factors in openLCA always add up to 1 since they are calculated automatically. For causal allocation factors, it is reponsibility of the modeler - or put it differently, in causal allocation, which means manual entry of allocation factors, the modeler has the power to set each of the factors for each exchange as he or she wants. The factors entered then are not further automatically normalised (to your point 2). Personally, I think it always makes sense to have allocation factors adding up to 1.

I am pasting some pictures from a simple multioutput process, setting calculating phyisical allocation factors gets 1/3 and 2/3, setting one of the factors manually to 0.2, for CO2 output, calculates then the respective amount for CO2 as 0.2*2kg.