Where I can find the binary interaction parameters

[longcloud]

Hi CalebBell,

Where is your binary interaction parameters databank, I just found this in your code, are they generated by the code? I know Aspen and ProII have a lot binary parameters, so just confused about this.

if kijs is None: kijs = [[0]*self.N for i in range(self.N)] self.kijs = kijs

[CalebBell]

Hello longcloud,

Aspentech and SimSci are extremely large corporations with the resources to conduct comprehensive literature surveys or simply purchase their data. For example, I believe Aspentech licenses the complete VLE database from DDBST to perform regressions to obtain their binary interaction parameters. DDBST has 1024504 data sets currently, and charges on the order of $35 USD per set. The multiplication of those values results in a very large number.

You read the line correctly; kijs are assumed 0 unless specified by the user. For example, the H2S-CH4 system is known to have kij=0.08. To use that value, use:

>>> SRKMIX(T=190, P=40.53E5, Tcs=[373.2, 190.564], Pcs=[8936865.0, 4599000.0], omegas=[0.1, 0.008], zs=[ 0.07, 0.93], kijs=[[0, .08], [0.08, 0]])
<thermo.eos_mix.SRKMIX object at 0x7fe0e2f33590>

If you are aware of a large database of free VLE data, please let me know. The ones I am aware of are small and not critically evaluated:

• https://www.cheric.org/research/kdb/hcvle/hcvle.php
• http://vle-calc.com/index.html

Sincerely, Caleb

[longcloud]

Hi Caleb,

Thanks for your replay, You made a fantastic library for chemical engineer. I worked at AspenTech in China. It would be pleasure if I can offer help to this library.

Thanks, Cloud

[CalebBell]

Hello longcloud,

I appreciate the offer of help and the compliment. The following are all ways that you could contribute to thermo:

• Find places the results from thermo don't seem right to you and submit bugs
• Find inputs that raise exceptions to thermo where they shouldn't
• Request features (It is useful to know what people are interested in, even if I can't make any promises about adding them)
• Write examples; for example performing some type of analysis on a chemical or mixture
• Find new sources of digital and reasonably licensed chemical data
• Promote thermo to engineers who might be interested in using Python

I would be very thankful if you looked into any of them.

Sincerely, Caleb

[longcloud]

Hi Caleb,

I'm writing a web UI for your thermo, once it down, every engineer can use it without know python. And I assume they will ask more questions. I will show it to you.

Thanks, Cloud

[longcloud]

Hi Caleb,

We made a web for thermo, the component and binary is not much. But we will improve it. CalcPlayer

Thanks, Cloud

[sivarammurugan]

Hi Caleb, First of all kudos to you for this amazing package . Regarding the binary interaction parameters , Chemsep has a a data bank , which includes the interaction parameters for sizable number of binary systems. As I checked with Chemsep authors, these databanks themselves are released under following license terms http://www.perlfoundation.org/artistic_license_2_0. So we are free to use these data for this package. Could you consider adding these data to thermo, or if you are ok with it , I can create a pull request to upload these files. Thanks Sivaram

[rmoreira00]

Hello longcloud,

Aspentech and SimSci are extremely large corporations with the resources to conduct comprehensive literature surveys or simply purchase their data. For example, I believe Aspentech licenses the complete VLE database from DDBST to perform regressions to obtain their binary interaction parameters. DDBST has 1024504 data sets currently, and charges on the order of $35 USD per set. The multiplication of those values results in a very large number.

You read the line correctly; kijs are assumed 0 unless specified by the user. For example, the H2S-CH4 system is known to have kij=0.08. To use that value, use:

>>> SRKMIX(T=190, P=40.53E5, Tcs=[373.2, 190.564], Pcs=[8936865.0, 4599000.0], omegas=[0.1, 0.008], zs=[ 0.07, 0.93], kijs=[[0, .08], [0.08, 0]])
<thermo.eos_mix.SRKMIX object at 0x7fe0e2f33590>

If you are aware of a large database of free VLE data, please let me know. The ones I am aware of are small and not critically evaluated:

* https://www.cheric.org/research/kdb/hcvle/hcvle.php

* http://vle-calc.com/index.html

Sincerely, Caleb

Hi Caleb,

Thank you very much for your work. From now on, I intend to give thermo a shot on my development of HVAC systems. I have been coding Cubic EoS myself for computing fluid mixture thermo. properties, but I rather try help on somebody else code and finish my HVAC modeling.

Thus, I am posting some things that may be helpful to you.

Component database: with VLE and NRTL parameters: Chemsep: http://chemsep.org/downloads/source.html Another component database: https://www.chemeo.com/ Korean database: https://www.cheric.org/research/kdb/ for UNIFAC: http://www.ddbst.com/unifacga.html

MATLAB EoS routines including PC-SAFT, for instance: http://hpp.uva.es/open-source-software-eos/ Link for downloading the routines: https://drive.google.com/file/d/0B9uaKxwk5j7TTS1BR2lQcGZZOW8/view

Worth citing open source software almost as capable as Aspen Properties: DWSIM (http://dwsim.inforside.com.br/wiki/index.php?title=Main_Page)

Regards, Raul

[longcloud]

Hi, @CalebBell and @rmoreira00 ,

Thanks for your information. Do you have any paper about bk10, we find the original paper of Braun, but the calculation need search the nomogram when calculate logK=alfa log(K10) + log(beta), can you find the formula of alfa and beta?

Thanks a lot. Cloud

[CalebBell]

Hi all, There is a small sample database in the development version of thermo now, https://github.com/CalebBell/thermo/blob/master/thermo/interaction_parameters.py with data from ChemSep.

Interaction parameters won't be loaded by default; it is left to the user to load them, or use their own, or a combination of the two. You can see the variant.

I don't know much about bk10 but I suspect it is quite obsolete.

Sincerely, Caleb