10 replies to this topic

[Icehot1]

I need to pick a few base methods for different components in Aspen Plus, and I'm having a hard time finding information on it. It seems it should be possible in HYSYS. Is it not possible in Aspen Plus? I would need to use NRTL for a couple polar molecules and Wilson for oil compounds. Until now I've only needed to pick a single option for the base method when starting a new document.

[PaoloPemi]

do you mean to define different models (for fugacity) for different components in the same stream ?  Which is the reason for that ?

Can you explain a bit better your problem ?

I am not familiar with Hysis but I doubt you can mix models as explained (but I might have missed something in your post).

My usual approach (for mixtures including polar / non polar components) with another software (Prode Properties) is to select a EOS with complex mixing rules, for example Peng Robinson - NRTL with Huron Vidal mixing rules (just to mention a possible combination).

This is certainly possible with Aspen.

[Icehot1]

Well, I'm might very well be saying something that isn't possible. I'm doing a project and getting little help. It seems the supervisor wants me to find the correct base method(/s?), and since I'm using both polar and non-polar components, a certain base method doesn't work for all. NRTL would work well for the polar fluids, but Wilson (from what I've read) would work well for the oil. But Aspen only let's me pick one base method. I interpret the supervisors instructions as if I would need to use both, to account for both behaviors, but it's not possible (?). I'm a little confused by all this. Can these Huran Vidal mixing rules be used to "combine" NRTL and Wilson?

[breizh]

Hi,

To support your work consider to review the documents attached related to thermodynamic models .

Good luck

Breizh 

[Icehot1]

Thanks breizh, I'll look into this.

[PaoloPemi]

complex mixing rules allow to combine the concept of activity (see NRTL, Wilson, UNIQUAC etc.) with an equation of state such as Soave Redlich Kwong , Peng Robinson etc.
Simulators include different mixing rules (for example, my copy of Prode Properties has Huron Vidal, MHV2, LCVM plus several others and Aspen should include similar options) .
Depending from application,  you can select a complex mixing rule as Huron Vidal to obtain good results for both polar (for example water-methanol etc.) components and non-polar components such as hydrocarbons.
An advantage of this approach is that the procedure applies the classical quadratic rule for all non polar components thus obtaining the same results of base EOS.
I would suggest that you discuss this topic with your supervisor since there are also other models as, for example, CPA (with association) , SAFT (with association) etc.  which could be (or not) suitable for your application....

[Icehot1]

Thanks for the info, PaoloPemi.

[SilverShaded]

Its really not clear what your are trying to do.  Are these compounds on the same stream?  What are the polar compounds and how are they mixing with the 'oil' compounds (real or pseudo?).  Is this an oil stream with some polar compounds in it like H2S COS H2O etc?

[Icehot1]

For the most part the polar compound is methanol, and from what I've read NRTL or UNIQUAC would simulate that best. And regarding the non-polar, for the most part it's a certain fatty acid methyl ester (as biodiesel), which Wilson would simulate well. Other minor compounds would be water, triglyceride and glycerol. They would be in the same stream yes. And I would be seeing how it would be affected to run that stream through certain equipment. My problem is that I'm not sure if the base method I was choosing was good enough for all compounds, and I though you could pick what base method would work for each compound, but after having read some literature I think I was quite wrong. I now believe I can only choose one for this task, and possibly use some complex mixing rule with it, but I think I'm not equipped to handle that at this stage. The best I could do is pick the "best" method, and I believe NRTL is it, since the polar interactions outweigh the non-polar. Do you think I'm on the right path?

[SilverShaded]

Guess it depends what you are trying to to do.  I'm assuming this is all one phase, or is it two phase or do you need to check if its two phase?  Do you need to flash it?  Seems like the main thing to worry about is the methanol and ester and NRTL seems a logical choice.  Tuning anything to get the trace compounds correct will require real data to tune whatever method you pick and maybe not worth the effort if they really are trace compounds.  Are any of these azeotropic?  If not, then NRTL and SRK should be reasonable.  If you have real data you can tune the NRTL or SRK parameters to match the data.

 

[Icehot1]

Sorry for the late respond. Thanks, I've come to the conclusion of using just NRTL.