7 replies to this topic

[panoskagiou]

What is the difference between using UNIQUAC and UNIQUAC-HOC model for the simulation of Vapor Liquid Equilibrium for the mixture water-acetic acid?

I attach you these two plots of VLE. In the one I use only the UNIQUAC model to describe the Vapor Liquid Equilibrium and in the other one I use the combination of UNIQUAC-HOC to describe the Vapor-Liquid Equilibrium in ASPEN PLUS. And I can notice that the second model is the most suitable because it takes into account the dimerization of acetic acid in the vapor phase. But I cannot really understand this dimerization how affects the composition in the vapor phase.

​HOC: Hayden O' Connel

 

Thank you in advance!!!

Attached Files

•  Vapor Liquid Equilibrium.pptx   727.92KB   22 downloads

[MrShorty]

I don't know how much detail you need. As a basic overview to fix ideas, I like to look at it as one of those "multi-step reaction equilibrium" problems you should recall from beginning chemistry. In this case, we have:

 

ROOH(aq) ↔ ROOH(g) described by our activity coefficient equation

2ROOH(g) ↔ (ROOH)2(g) described by Hayden-O'Connel

 

Recall from LeChatelier's principle that, as ROOH is consumed by the second reaction, more ROOH will be drawn out of the aqueous phase to maintain the phase equilibrium between ROOH(aq) and ROOH(g). As more ROOH(g) is in the vapor space, more (ROOH)2 will be formed to maintain that equilibrium. These "reactions" continue until equilibrium is established for both "reactions". The overall effect is that you will have more total ROOH [ROOH+(ROOH)2] in the vapor than if you ignored the second reaction.

 

It is similar to the dissolution of CO2, where we have CO2(g) ↔ CO2(aq) [+H2O] ↔ H2CO3 ↔ H+ + HCO3- that we use to explain why CO2 is so much more soluble in water (especially if you have a base/buffer in that water) than other gases.

[panoskagiou]

Dear Sir,

 

Thank you so much for your reply. I send you a personal message to ask you more. Your answer is really helpful.

[serra]

you may find useful this paper discussing last version of Hayden-O'Connel method

Aqueous  cross second virial coefficients with the Hayden-O'Connel method (Kyle Byshop and J.P.O'Connel)

not sure if Aspen adopts original or revised correlation,

the software which I use (Prode) has a model for vapor fugacity based on Hayden-O'Connel and

in principle you should be able to combine the virial method (vapor phase) with different models for liquid phase

as discussed by MrShorty...

[panoskagiou]

Dear Serra,

 

Thank you so much for your reply and the paper. I think that ASPEN PLUS uses the proper correlation. I will check the paper out. This binary system is really interesting. 

[MrShorty]

In your PM, you asked about understanding more about the thermodynamics of this mixture. We like to say in our lab that acetic acid-water is one of the least ideal (most thermodynamically complex) systems we've looked at. Can you be more specific? What exactly about the thermodynamics of this system are you looking to understand?

 

In your PM, you mentioned that we like to talk about ideal/non-ideal in terms of intermolecular forces, and that is certainly true. At the same time, I also think that these models also have a simply empirical element to them. I have often noticed that we do not use electrolyte models with acetic acid, even though we know that acetic acid is a weak acid and dissociates to some extent in water.

 

I guess my point is that these mathematical models we use are not always more than empirical equations that we use because they work to relate the properties we are interested in (T, P, x, y, H, etc.). In a case like this, I am not convinced that a simple gamma-HOC model describes the behavior of water-acetic acid in great detail down to the molecular level.

[panoskagiou]

Generally I am interested in mostly from a process engineer perspective. So I think that the UNIQUAC-HOC is good enough to design the separation for example of this mixture. But with my question I would like to get as answer how a fresh graduate MSc process engineer will deal with different mixtures in order to find the best model to describe the thermodynamics of the mixture. To find the best model depends mostly on experience and less on knowledge about thermodynamics.

 

Thank you for your answer!

[MrShorty]

I agree that experience has an important part to play in this sort of thing. My mentor (an old school engineer from before modern simulation programs) would often worry that new engineers come out of school with a misplaced trust in the "black box" simulator. I put stuff into the black box, results come out, and they must be The Truth ™ because the computer did it. I hope that new engineers have the ability to use the black box, with a healthy mistrust of its results. In a case like this, a new engineer may use the tool he knows (a basic gamma-phi model), and calculates the results in the simulator. Then, because he knows that the computer can be wrong, he immediately seeks out experimental work to check against the simulator's results. In this case, he should quickly come up with your left picture where he can see that his simulation results do not replicate experiment. At this point, hopefully he knows enough about finding answers to questions to seek further information (from experienced mentors/colleagues, literature research, contacting simulator tech support, etc.). My old high school geometry teacher once said that education is more about learning how to learn and find answers than it is about having all of the answers at graduation. Hopefully our new engineers are entering the field not only with good engineering knowledge, but also knowing how to find and learn what they don't yet know.