Long-term contexts

[bnjmnmorelli]

Hello, I am trying to better understand the implementation of long-term emissions throughout ecoinvent compatible LCIA methods. Note: i am referencing the openLCA implementation of ecoinvent 3.10. Information in this post is helpful. The logic in this post makes sense to me for LCIA methods without a defined time horizon. But I am trying to understand why long-term emissions are characterized in an indicator like GWP100 (this is true in the IPCC 2021 method and ReCiPe 2016 v1.03, midpoint(H), for example). Impact world's v2.2 indicator 'Climate change, human health, short term' includes null characterization factors for long-term contexts, as a counter example.

Also, in looking through ecoinvent harmonized LCIA methods, I noticed that the EF v3.1 method does not include ecoinvent's long-term contexts for freshwater ecotoxicity or human toxicity carcinogenic (to name two). I expected them to be there given that there is also an explicit 'no LT' version.

[thomassonderegger]

Dear Ben, I am not very familiar with IMPACT World+ and how they treat short- and long-term. The ecoinvent approach might be simplistic: no matter when an emission happens, it is treated as if it appeared now. This is true whether the long-term emissions, so those only happening in the very distant future, are included or excluded. Not sure this helps :-)
We change methods as little as needed (we might remove duplicates, for example), and this also should give an answer to the EF question: if you look into the original EF files, you find that some CFs for long-term flows are set to 0. We take these 0 as is, but this undermines the approach of implementing CFs given for a flow also for long-term compartments and then having a no LT version of the method to compare (I wasn't aware of this, to be honest).