Missing impact factors in EF 3.1 or EN15804 3.1

[conradspindler]

I found some impact factors that might be missing during comparisons of the ecoinvent implementations for EF 3.1, EN15804 3.1, IPCC 2021.

1. "Hexafluoroethane" from 4 different air emission compartments does not exist in the climate change (GWP100) categories for "EN15804 3.1", but exists in "EF 3.1" and "IPCC 2021" (high impact factor of 12400 kgCO2eq/kg).

2. "Perfluoropentane" from 1 air emission compartment does not exist in the climate change (GWP100) categories for "EN15804 3.1" and "EF 3.1", but exists in "IPCC 2021" (high impact factor of 9220 kgCO2eq/kg). It is marked as "ecoinvent orphan" in the LCIA CSV files in this Github repository, but still used in the processes "methylene diphenyl diisocyanate production, ISOPA", "polyether polyols production, long chain, ISOPA", "polyether polyols production, short chain, ISOPA", "toluene diisocyanate production, ISOPA".

3. "Tetrafluoromethane" from 4 different air emission compartments does not exist in the climate change (GWP100) categories of "EN15804 3.1" for 3 compartments ("air/urban air close to ground", "air/non-urban air or from high stacks", "air/low population density (long-term)"), but fully exists in "IPCC 2021" and "EF 3.1" and partly exists for the 1 remaining compartment "air/unspecified" in "EN15804 3.1" (high impact factor of 7380 kgCO2eq/kg).

4. "Sulfuryl fluoride" is not characterized in the climate change (GWP100) categories for "EN15804 3.1" and "EF 3.1", but exists in "IPCC 2021". I guess that this is a mistake in the original data from the EF developers, since this substance is also not characterized for climate change in the original EF reference system. The EF system has the substance available (called "sulphuryl difluoride" in EF with CAS 2699-79-8) but in the LCIA tables they only characterize it in ecotoxicity and not in climate change. This must be a mistake, since "Sulfuryl fluoride" is also a high impact greenhouse gas with a factor of 4630 kgCO2eq/kg and it is present in AR5 and AR6.

5. Just a tiny difference that "Sulfur hexafluoride" for GWP100 has a factor of 25200 kgCO2eq/kg in the preliminary IPCC 2021 report and 24300 kgCO2eq/kg in the final IPCC 2021 report. EF 3.1 and EN15804 3.1 are based on the preliminary report (as far as I know), but the "IPCC 2021" method could be in principle updated to the final report. I could only find this one difference between the preliminary and final report. Not sure, if it should be changed in "IPCC 2021", since it would lead to differences between "IPCC 2021" and "EF 3.1". And not sure if it should be changed in "EF 3.1", since it would be different to the original EF factors based on the preliminary report.

This will hardly affect any impact results, since the substances will typically not have a large contribution in standard ecoinvent processes, but we can keep track in this repository for updates/improvements.

[thomassonderegger]

Very helpful as always, thank you, @conradspindler! Please double check once v3.12 is released, but we should have caught and corrected all issues.

[conradspindler]

@thomassonderegger

Julia found that SF5CF3 with a high impact factor (Trifluoromethylsulfur pentafluoride, CAS 373-80-8, 18500 kgCO2eq.) is not included in the IPCC 2021 methods if needed for modeling. It is not used in any process and I guess you are only mapping what is needed for your reference system?

There might be more substances from IPCC that are not used in any process / in the ecoinvent reference system and of course, one can also include them manually if needed. But because the IPCC method is used a lot and the IPCC report has not so many factors in total, maybe it would be useful to include all of them by default, even if not used in processes. This helps when needed for modeling, and it could help compliance when e.g. reporting all substances from the IPCC report (even if is the result is low or zero for example).