bug report

[aken]

hi,
I specified the mole fraction of benzene to be 0.9 in distillate and of toluene to be 0.98 in the bottoms, and the system did converge. But once l expose the reflux ratio as output variable in the cape open properties, it converges but with wrong results.

[jasper]

Quoting Ross Taylor:

I'll agree that there is a bug but this column appears to have been something different in the past.

I confirm that I can get the same results as the poster. But when I open the file and try to solve it as a stand-alone model in ChemSep itself (i.e. NOT using CO properties) I see it has a Liquid-liquid K-value model that had been selected earlier. This is only meaningful for extraction operations so I am unsure why that got set for this case. When I reverted the model to something sensible (DECHEMA, NRTL, Antoine) the column converged to the correct results. Then I switched it back to CO properties and it also converged correctly.

lodeccibmgcgiklh.png (3.82 KiB) Viewed 7743 times

I am unsure why a LL K-value model would lead to the specification being wrongly used, but it is not relevant when a correct model is set.