ReCiPe 2016 endpoint: the method is probably errors-affected

Question

Hello, I suppose to have found some errors in ReCiPe 2016 Endpoint. I have tested the method with ecoinvent 3.6 cutoff in OLCA with method package 2.0.4 v2 and ecoinvent 3.5 in SimaPro.
The process analysed is:
market for electricity, high voltage | electricity, high voltage | Cutoff, S - IT

The characterization step reports higher values for water consumption:

And moreover the normalization is wrong, in detail, the method uses the same NFs than SimaPro (which perform: Normalized result=Characterized result*NF) despite OpenLCA perform the inverse operation (Normalized result=Characterized result / NF). In fact, by using the inverse factors to obtain the normalized values, the results are comparable.

I would like to have clarification about the different results obtained for water consumption and the mistake on the NFs.

Thank you,
Nicola

Answer 1

Hi, 

Thank you for pointing out the error with the normalization factors. we will fix it in the new version. In the meantime, please use the inverse of the values, to get the correct results.

As for the higher values for water consumption, the results may be different due to the difference in ecoinvent inventory between ecoinvent 3.6 and ecoinvent 3.5, and also due to regionalization of the water-based flows for SimaPro, which can be found in ecoinvent 3.5 regionalized databases (for e.g. ecoinvent 3.5 Cut-Off Unit Processes Regionalised) for openLCA.

Comments

Hi,
I'd like to follow up on this question. The normalization factors for Recipe2016 are similar to the ones found here: https://www.rivm.nl/documenten/normalization-scores-recipe-2016
Nonetheless, the normalisation factors for the endpoint method are completely different.
Should I use the characterisation results from Open LCA and then do the calculation with the RIVM values instead?
Thanks for the information.

---

Hi pwiche,

The normalization factors for endpoints will be fixed in the next version update. Please use the factors from RIVM for calculating the normalization factors. Thank you