Dear Qazi,

building phase envelops is not only a matter of optimization, there are many other theoretical difficulties that you should solve them first. For example building the retrograde region of the envelope is difficult and has special considerations. Finding Cricondotherm and Cricondobar points is not easy and it is not only a matter of calculating Kxy. Also finding the true critical point of a mixture with EoS is rigorous and sometimes very complex. Just by using PR-EoS and without considering these limitations you can never truly build or even understand the phase envelope. I do strongly suggest you to study Mr. Michelsen papers published in Fluid Phase Equillibria Magazine first and broaden your knowledge before going any further. One tip, if you have Methane in your mixture, phase behavior of your mixture is more complex than common mixtures. Try to use another EoS, even not a cubic one is preferred.

I think it all depends on what software resources you currently have available to you. If you have access to Matlab or Mathematica you might luck out and have someone provide the code for you to use. But it is never easy to use code that was developed by someone else in a research environment, as it tends to be very specific. As an alternative you might want to look at the following:

http://www.jmcampbell.com/tip-of-the-month/2009/09/how-to-tune-the-eos-in-your-process-simulation-software/

This article succinctly discusses how to optimize BIP in the context of using the ProMax Process Simulation Software package(PSSP). But whatever PSSP you use you will be working with a thermodynamic package that has been developed by experts in the field. and you are not trying to reinvent your own. Your task is then to fine-tune the system to optimize your own BIP which now is a simple regression problem, easily done with software such as MatLab or Mathematica.

Also check out the NIST website and their thermodynamic software resources.

Alternative, if you search the web you will find many PHD theses where this generic problem has been addressed from scratch using different approaches ... You may luck out, but from my experience, I favor working with a sound thermodynamic package rather than try to muddle through code (often not well documented) in a PhD thesis.

Qazi,

If you have access to the Aspen Plus process simulation package then you may want to consider using the Data Regression feature of this tool. Aspen Plus provides a robust tool for regressing Binary Interaction Parameters for Equations of State like Peng Robinson and Redlich-Kwong-Soave, along with activity coefficient models, like NRTL, and UNIQUAC. There are definitely other tools as mentioned above that you may want to consider for this work, but I have some familiarity with the Aspen Plus tool and recommend it as an option.

Thanks Brian Higgins for the good discussion, actually i have develop my own Matlab codes to predict Pxy diagram , but they way most of the authors used is to predict the equilibrium constant (Kcal=y/x) and using experimental data to calculate Kexp and then using optimization toolbox of matlab to minimize AAD(K) . They problem I am facing in my program the global composition , I used the fixed global composition let say (CH4=0.2 and CO2=0.8) by mol fraction and then calculate the bubble point pressure and the composition of CH4 & CO2 in both the liquid and vapour fraction, from where I calculate the equilibrium constant (K). then to extend this further I am using the previous mole fraction of CH4 and CO2 as a guess to calculate another bubble point pressure and so on , the problem I am facing is when I try to used the experimental feed as a global composition in my model to get the bubble point pressure , I couldn't get the result I was expected , I tried Matlab optimization tool box to optimize BIP but the result are not improved at all. That is the reason why i am seeking for someone codes to understand what have they done to achieve the required result which is lacking in my model.

Dave Modeste , I have aspen plus in my University and I have little hands on experience on that , would you guide me how can I optimize the BIP using aspen plus. I have seen Datafit option but don't know how to proceed further.

Dear Qazi,

building phase envelops is not only a matter of optimization, there are many other theoretical difficulties that you should solve them first. For example building the retrograde region of the envelope is difficult and has special considerations. Finding Cricondotherm and Cricondobar points is not easy and it is not only a matter of calculating Kxy. Also finding the true critical point of a mixture with EoS is rigorous and sometimes very complex. Just by using PR-EoS and without considering these limitations you can never truly build or even understand the phase envelope. I do strongly suggest you to study Mr. Michelsen papers published in Fluid Phase Equillibria Magazine first and broaden your knowledge before going any further. One tip, if you have Methane in your mixture, phase behavior of your mixture is more complex than common mixtures. Try to use another EoS, even not a cubic one is preferred.

Qazi,

That is very good that you have access to Aspen Plus. I am not sure which version your university is using, but my comments will be related to version 8.1 of aspen. It should also apply to recent versions like 7.1, 7.2 or 7.3, with slight modification. Before doing that my suggestion is that you go to the aspen plus support site and complete the registration process. (this is free and will allow you to access examples, and documentation for many aspen plus tools. there may be some video webinars available also that are relevant to your area of interest).

http://support.aspentech.com/supportpublic/AllFrameFile.asp?AllNews.asp

here you can then do some browsing to find the info you need. I suggest you try the "Knowledge base" and "Product Video tutorials" See Menu on the left side of the page.

The Datafit tool is different from the Data Regression tool and you may want to consider using the Data Regression tool first, since this tool is well suited to regress data to the parameters of both Equation of state and Activity coefficient parameters. Of course this depends on the quality and quantity of data you have. There is also a thermodynamic consistency test for VLE data that can be used to assess the quality of the data, and this may be helpful as you get going.

Once you open aspen plus in version 8, you can choose the Properties Option from the

selection pane at the bottom left corner of the screen, and look for the data regression menu at the top selection pane (Run Mode) of the screen. In the menu pane on the left side of the screen, there is another menu that will have an option for Regression that you can select to get started.

Anyhow, this is quite a bit at one time, and doesn't go into details of regression but should help you to get to the Regression modules in Aspen Plus. Please give this a try and you can ask additional questions to this thread, along with contacting the Aspen Tech Hotline directly by phone or email.

Babak Maghbooli I know that other factor also influence in building phase envelope , can you please give me a link of the article you were talking about .

Dave Modeste thanks alot for the wonderful explanation , I will surely gona do that and will let you know if I found any problem , hope that is okay with you.