PPC-SAFT Database Butene Isomer Parameters

[MichaelGadaloff]

Hello,

I think there may be an issue with the parameters for cis/trans-2-butene in the database for PPC-SAFT. Looking at the csv file from Esper et al., I think there are data for cis-2-butene, trans-2-butene, and a mixture of the two. But the names of cis-2-butene and the mixture have both been changed to 2-butene in the Clapeyron file.

I hope that makes sense, but please let me know if you would like me to clarify.

In the Clapeyron file:

In the Esper et al. CSV:

[longemen3000]

Oh, I see, they got mixed, same thing happened with decalin before, this just slipped out of my sight

[MichaelGadaloff]

Ah ok fair enough - thank you for replying so quickly. I think I also found another one: cis-2-pentene (line 630 in the Clapeyron CSV) is labelled as 2-pentene

[prehner]

Hi, to avoid confusion, can I suggest using the acronym PCP-SAFT for the version based on the dipolar term by Gross and Vrabec (and used by Esper et al.)? There is also a PPC-SAFT out there that is mainly developed by the groups of de Hemptinne and NguyenHuynh but that uses a different term for polar contributions.

[pw0908]

Hi, I knew there was a difference between PCP and PPC SAFT but I always assumed the Gross version was PPC since it just sounds more "correct"? I think there's also quite a few different implementations of "Polar PC-SAFT" beyond just Gross, de Hemptinne and NguyenHuynh (I think Chapman has his own?).

As an aside: have you benchmarked the dipole-quadrupole term in feos? The coefficients don't quite look the same as in the paper and there's an additional pi^2 in front of the a_3 term. The tests don't seem to cover this term and the original paper doesn't provide the parameters needed to reproduce the results.

[prehner]

The dipole-quadrupole term is a bit of an issue in feos as well, because we are not sure whether the only reference implementation that we have is correct. The issue with the pi^2 should be clear though: It can either be absorbed in the correlation integral parameters (which would lead to a deviation close to factor 10) or left there as prefactor in a_3.

[longemen3000]

On the PPCSAFT vs PCPSAFT naming issue, we can just add an alias that makes those equivalent in the meanwhile.