9 replies to this topic

[daraj]

Hi all, need some inputs.. From a gax mixture containings organics like benzoid acid, maleic anhydride etc. along with water, light gases, when I condense this mixture in reality the water and light gases present in the stream should go along with the vapor and not condense and only the heavy organics must condense. But aspen plus predicts some water condensation and some light gases inn the liquid phase as well. I declared all the light gases as HENRY components. Using NRTL-RK. Still some water and gases are present in the liquid phase after the mixture is cooled. any idea how to rectfiy this? i cna always use some splitter and get rid of this issue. But i want the model to correctly predict the composition instead.

[breizh]

Hi,

most probably your thermo models are not the right ones ! 

Why did you select Henry ?

you should have an option where the software is proposing to you models .

Good luck

Breizh 

[daraj]

thanks Breizh, Asoen usually expoects you to declare the HENRY components. As to the choice of thermo package it is up to the user. NRTL is because of some polar compounds and water. declration of HENRY components for light gases is a common approach recommended by aspen itself. this will take care that the light gases dont stay dissolved in liquid phase. 

[breizh]

Hi,

You may take a look at this document.

Good luck

Breizh 

[latexman]

But aspen plus predicts some water condensation and some light gases inn [sic] the liquid phase as well. I declared all the light gases as HENRY components. Using NRTL-RK. Still some water and gases are present in the liquid phase after the mixture is cooled. any idea how to rectfiy this?

 

On what basis is Aspen Plus wrong?  How much is "some"?  Where's your real measured data and comparison?

 

A flash is only one theoretical stage, and quite frankly, I would expect "some water condensation and some light gases inn [sic] the liquid phase".  Benzoic acid and maleic anhydride are known to hydrogen bond with water, not to the point of ionization, but, there is a weak to medium attraction.

[daraj]

Hi latexman, the gas mixture is actually cooled ina  condenser to 50-60C. I dont have actual real data but i was told that industrially due to the low mole fractions of water (pressure is just above 1 bar) water does not condense or is very little. Aspen predicts almost 10% of incoming water condensing after the exchanger based on liquid and vapor fractions reported in the 2-phase mixture i get after the condenser.

[latexman]

1 barg and 40-50 C, yes, water in the liquid phase is quite reasonable.

[SilverShaded]

Henrys components are used so that it will predict the solubility of gases in liquids, thats the whole point of using them.  Like O2 solubility in water.  It's just more accuracte than using PR or SRK as the henrys constants are either from experimental data or estimated in some way.

Edited by SilverShaded, 28 March 2022 - 04:23 AM.

[daraj]

silver suppose if the condenser had some solid formation(due to desublimation of certain compounds) would NRTL-RK be able to handle that, if i define certain components as solids, and used a GIBBS to do the desublimation? or REQUIL?

[SilverShaded]

Thats a complicated issue.  Many substances that sublimate actually dissociate into two or more compounds, secondly the 'typical' thermodynamics packages like PR, SRK are tuned for Vapour liquid fugacity predictions.  It is possible to predict solid vapour pressures (although off the top of my head I forget exactly how).

Compounds that dissociate use more obscure packages like Hayden Oconnell, for solid vapour pressures you probbaly need to use a packages specifically develeoped for doing that, A+ would be your best bet, i think it has solids capability, or maybe other vendors.

Edited by SilverShaded, 29 March 2022 - 04:23 AM.