7 replies to this topic

[javaloyes]

Hello,
In the ethanol dehydration with benzene or cyclohexane, I have some problems with the simulation of the heterogeneous azeotropic distillation system using Hysys. I’m trying to design the typical system with two distillation columns (the azeotropic and entrainer recovery columns) as it is described, for example in “ Distillation design and control using aspen simulation” by W.L Luyben.
Following the Luyben explanation, I’m able to converge the system with Aspen, but the problem is that the simulation of the first column is very tricky, and I’m not able to converge it using Hysys instead of Aspen.
Maybe it is because of the fluid package, but I’m using the same in both simulators (aspen and hysys).
In other post people recommended to find the document called “advanced column modeling document”, but I can’t find it.

Please, somebody can help me to find this tutorial, or help me with the simulation?
Thanks for your time.

[Technical Bard]

What thermo package are you using? If you are using the Aspen Properties package in HYSYS it may be a little finicky because the way Aspen thermo works with the HYSYS solver isn't always stable.

[javaloyes]

hello!

I'm using the uniquac fluid package of HYSYS. I tried to use the Aspen Properties package in HYSYS, but it doesn’t work well.

I changed the solving method from "Modified HYSIM Inside-Out" to the "sparse continuatinuation solver" which is recommended for azeotropic distillation and supports two liquid phases on the trays of the column.

With this solving method I'm able to converge the system (although I can’t close the reflux ratio loop) the main problem is that this solver is for ideal stages, and I need to change the stage efficiencies using MATLAB in order to optimize the total number of stages (efficiency ~= 0 ==> by pass).

If you are interested, I can mail you the hysys file.

Thank you

[Chesps]

Just a suggestion may be you have tried it already, but in case not try checking in Azeotropic modified HYSIM INSIDE-OUT solver (In parameters Tab) >> SOLVER >> Dropdown>> Modified HYSIM INSIDE-OUT and check-in azeotropic option. Also keep two liquid phases prediction on.

[chemdoc]

for mixtures with strong non-ideal behaviour (which is not the case) the Russel's procedure may suffer, better to use a full Newton correction or a limiting factor (damped correction on Inside-Out parameters), however (usually) it works fine for these problems, you may review the inputs and change some parameters to help convergence.

[javaloyes]

Thanks you all for your help.

I tried to converge the azeotropic column with the HYSIM INSIDE-OUT with the two liquid phases prediction on and is no way to converge it.

I will attempt to change some parameters to help the convergence, as the damping factor.

Regards

[Chesps]

You are right if you keep two liquid phases prediction on you will have to use sparse solver, sorry for earlier post.
I have not done azeotropic distillation simulation in HYSYS but have done it in ASPEN PLUS, in ASPEN PLUS also it is troublesome somtimes to get convergence unless we have our hand calculations ready to give proper specs.

[smalawi]

my experience with Hysys solver for two phase and Azeo's is very bad. this applies the same to UniSim and guess similar products. i believe the solver is not stable and you need lots of tuning and manipulation to get a solution that still may not be reliable !

the column solver in Pro/II was better for two phase but not too sure about Azeo's

cheers,
SM