13 replies to this topic

[Catherine Kwan]

Hi all,

 

Does anyone know how to determine solubility of styrene and ethylbenzene in water at different temperatures?

Thanks for the help!

[PaoloPemi]

you do not specify the conditions, at some standard conditions

for many potential inquinants it's easy to find solubility in water,

 

http://www.inchem.or...sc/eics0268.htm

http://www.inchem.or...sc/eics0073.htm

 

for temperature dependent values you may consider databases as DDB

(there is some ternary) or add a temperature dependence based, for example on properties of

each pure substance but accuracy of that methodology will depend from operating point.

Edited by PaoloPemi, 05 February 2014 - 04:29 AM.

[ankur2061]

Catherine,

 

Solubility in water can be obtained from Henry's law constants for both Ethylbenzene and Styrene. Constants can be found at

the NIST chemistry webbook at:

 

http://webbook.nist....k=10#Solubility

 

http://webbook.nist....k=10#Solubility

 

Regards,

Ankur.

[PaoloPemi]

thanks Ankur,

perhaps the IUPAC-NIST solubility database should provide some additional data,

 

http://srdata.nist.g...mp2ChemNms.aspx

 

however, as said in previous post applicability depend from operating conditions which are not specified,

Paolo

[Catherine Kwan]

Thank you very much for the prompt reply to my question. Actually I am working on a school project involving the Hysys simulation of a styrene plant. And I am working on the 3-phase separator, trying to figure out what is the solubility of styrene and ethylbenzene in water at 40degC and 120kPa as simulated in our project. 

[PaoloPemi]

styrene and ethylbenzene have low solubility in water and a multiphase equilibria at some specified conditions

can produce unrealistic results if model for fugacity isn't accurate,

with google I have found several papers discussing this separation with experiments see for example

 

http://ijche.com/iss...2010-7-4/28.pdf

 

you may start by regress data provided in above links and tune the model, then compare results.

[PingPong]

And I am working on the 3-phase separator, trying to figure out what is the solubility of styrene and ethylbenzene in water at 40degC and 120kPa as simulated in our project.

It is an illusion to think that in a 3-phase separator the separation between the styrene phase and the water phase is perfect.

The amount of the many very small styrene droplets dispersed in the water phase coming out of the separator will be very very very very much more than the tiny amount of styrene that is really dissolved in the water. (The same story applies to the many very small water droplets dispersed in the styrene phase from the separator).

Moreover there are often small amounts of impurities present that act as surfactants and increase the solubility of hydrocarbons in water (and vice versa).

 

That is why in the real world we usually do not care much about the theoretical solubility of hydrocarbons in (pure) water.

 

Of course this should not stop your learning efforts, but is merely intended to show you that the accuracy of any solubility data that you find, and decide to use, is hardly relevant, so don't worry too much about it.

Edited by PingPong, 05 February 2014 - 01:58 PM.

[Catherine Kwan]

Thank you very much for your reply!  Now I have basic idea of what is happening.

Currently, we are using Peng Robinson as the fluid package (thermodynamic model) for our HYSYS simulation. It does not seem to be accurate in simulating the 3-phase separator unit. Is there any alternative fluid package that can more accurately model the simulation of a 3-phase separator? Thank you very much for youur help!

[ankur2061]

Catherine,

 

The density difference between EB and Styrene is very less (867 kg/m3 and 906 kg/m3 @20°C). So what are you trying to achieve by a physical separation in a 3-phase separator. 

 

Distillation might be a better idea. Since the atmospheric BP of EB and Styrene are close (136 & 145°C respectively) distillation may require to be done under vacuum conditions to enhance separation. Refer some of the links:

 

http://www.sciencedi...263876212002420

 

http://www.calvin.ed...FinalReport.pdf

 

Regards,

Ankur.

[breizh]

http://researchrepos...ries._81_P6.pdf

 

This resource from Internet may help you

Breizh

[PaoloPemi]

I think you should reconsider process design,

first, base your analysis, on experimental data ( best, from experiments on real separations) not on the output from a simulator feeded with generic parameters

(unfortunatelly it seems this is the preferred way to proceed nowadays, if the simulator converges one assumes the solution is feasible),

to simulate multiple liquid phases with a EOS such as Peng Robinson you would need complex mixing rules (with suitable BIPS) or at least (with vDW mixing rules) temperature dependent BIPS, differently, better to use models based on liquid activity,

in both cases for all critical applications make sure about VLE/LLE consistence of models, otherwise regress BIPs from experimental data points (the best option when experimental data available)

finally, for true multiphase use a suitable solver (I know Aspen Plus and Prode) , I do not presume such capability in your simulator.

[Catherine Kwan]

Thanks everyone for the valuable guide and input! Now, we have some idea of how to go ahead with our project. Really appreciate all the help! 

[JakeVincent]

Dear all,

 

I understand that the flowsheet of a styrene production plant using Ethylbenzene feed involves a three phase separator to remove the excess water and lights (CO, CO2, etc) prior to the distillation towers. With this three phase separator, water will be removed from the bottom, light gases from the top, and the organic phase in the middle will be sent to the distillation towers.

 

From what I know, the peng robinson (PR) package has weak correlations for water/ethylbenzene and water/styrene systems, ie water has a certain degree of solubility in the organic phase if we were to employ the PR fluid package on this 3 phase separator. Consequently, this will cause problems in hysys conversion for the downstream distillation towers (2 liquid phases will be encountered at the tower stages).

 

After some reading, I found that 2 options to study which is the appropriate fluid package to employ for this three phase separator:

1. Prode Properties (regression from experimental data to identify models that give the best data fitting)

2. Dechema chemistry data series

 

Prode Properties is basically some software available from www.prode.com. Any ideas on how we can correlate theoretical data (from hysys simulation) to empirical data (from Dechema chemistry data series) to identify the most accurate thermodynamic model for this particular 3 phase separator?

 

Cheers,

Jake

[PaoloPemi]

you may consider to open a different thread for this question,
anyway, I do not think you should have problems to regress experimental data from (for example) Dortmund Data Bank (DDB) or another source,
in my opinion the problem there was the proposed model itself,
i.e. if you use a EOS & vDW mixing rules (as Extended Peng Robinson in Hysys) for these mixtures you should expect quite large errors (in a range of temperatures, pressures, compositions),
depending from operating conditions a solution could be a model based on liquid activity or (better, for extended ranges such as relatively high pressures) a EOS with complex mixing rules,
probably a EOS with vDW mixing rules and temperature dependent BIPs can do the work, too.
In almost every process as first step one has to investigate about the reliability (meaning the capability to reproduce accurately real data) of thermo packages,

I believe this is a most important concept for students.

Edited by PaoloPemi, 10 February 2014 - 12:11 PM.