13 replies to this topic

[SilverShaded]

Some Background:

 

Lockdown has given us all lots of spare time were we are bored out of our minds (at least some of the time).  To that end i started playing around with some old distillation column code i had and made some improvements and put together a simple process simulator.

 

The main improvements are that it is 'generally' not neccessary to enter estimates for any column conditions for the column to solve. (at least for the columns i've tried).

 

The code has been optimised to some extent and basic stream flashes are fast (~2ms for a crude flash with 80 components), columns generally solve in a few hundred ms.

 

Currently it has a basic Hysys/Petrosim/Unisim style interface and works i'm a roughly similar way (although a myriad of improvements are of course possible).

 

However, this is 'a work in progress' and doubtless has plenty of annoyances, bugs and things that are just plain daft.

 

What i'm wondering is if anyone would be willing to help test it?  Not right now but possibly in a few months time.

 

My current aim is to get a crude and vacuum column flowsheet running sucessfully and once that's done it would be nice to get feedback on things like bugs and column configurations that fail to solve etc...  After that some basic petyluk columns are on the waiting list for testing.

 

If anyone might be interested please send me a message

 

Thanks for reading this.

 

Regards,

Silver

 

 

 

[StealthProg]

Hi, I might be interested to help can you tell us a bit more about the software?

[SilverShaded]

Hi Stealth!

 

Sure, the program (or versions of it) actually has a fairly long history but has never been publically available.  My interest in simualtion started back in the mid 80's with the devleopment (as far as  know) of the worlds first object orented simulator.  However due to the IP and copyright onwership restrictions at the time it was never publically available.  The current program is a a total rewrite using a combination of standard thermodynamic methods, publically available data and some IP developed by myself over the previous decades.  In additon to that some of the solver methodology has been borrowed from the excellent Sim42 open source project that was around about 20 years ago.  The main part borrowed from those algorithms was the concept of a forget pass, orginally I had not used a forget pass but instead propogated empty values around the flowsheet with the idea that it would be less computation time not to have to re-calculate verything on every pass.  However, the forget pass simplifies the whole thing so I went that way for the moment.

 

The program is not open source, although it may be downloadable for free, eventually, when i think the quality and reliability is good enough.

 

Currently it has a substantial component database, a fully working assay manager, and most of the basic unit operations.  I am currently working on upgrading the interface around complex column configurations (e.g. Columns with side strippers/pumparounds and eventually petyluk columns) the solver is working via script input and the interface needs to be properly connected to it.

 

Some of the IP that will not be made public is around the column solving aglorithm which has been developed to be far more robust than most commercial simulators.

 

In the case of Peng-Robinson, results for common compounds, e.g. gases, light hydrocarbons are identical to Hysys.

 

Im looking for people to help test it, I guess the benefit is you will get a copy of the software and free use.  Testing is really important to iron out any remaining bugs with all the different column confiurtaions that are possible and different therm methods, Peng-Robinson is thoroughly tested but some of the others are so tested yet.

 

Some Screen shots below;

 

https://flic.kr/p/2kxWwu9

 

https://flic.kr/p/2kxVVH9

 

https://flic.kr/p/2kxWwuu

 

http://https://flic.kr/p/2kxVVHQ

Edited by SilverShaded, 02 February 2021 - 03:23 AM.

[SilverShaded]

Kinetic Reformer Model now added!  Kinetic Isomerisation Model to follow.

Attached Files

•  RefResults.png   45.08KB   1 downloads

Edited by SilverShaded, 03 September 2021 - 06:58 AM.

[horatorres]

I can help you to test the software

 

horacio

[SilverShaded]

I can help you to test the software

 

horacio

Yeah that would be great, theres probably a lot of very obvious bugs in it write now that i could easily fix if i had a list.

[ChE Matt]

I'd be happy to help test it out.  Let me know!

 

edit: I've been mucking about with just building a bit of a "simulator" in Excel, really just want to be able to put in stream info, get a flash calc, and be able to take the properties of the fluids and use them for equipment design.  So the equations behind process simulation and equipment design are of great interest to me.

 

I felt like I had to say something more because, oddly enough, this is my first post here?  I've been coming here for quite a while!

Edited by ChE Matt, 27 September 2021 - 09:34 PM.

[SilverShaded]

I'd be happy to help test it out.  Let me know!

 

edit: I've been mucking about with just building a bit of a "simulator" in Excel, really just want to be able to put in stream info, get a flash calc, and be able to take the properties of the fluids and use them for equipment design.  So the equations behind process simulation and equipment design are of great interest to me.

 

I felt like I had to say something more because, oddly enough, this is my first post here?  I've been coming here for quite a while!

Thanks for the offer.  At some point i'll make version downloadable or maybe just send anyone interested an install. 

 

I did a lot of stuff in excel, it's a great way to prototype and debug code, ended up a somplete CDU simulation in excel with pumparounds and strippers, used it a huge amount to try out various solving methods.  Also for things like Lee-Kesler enthlpy calcs.  Actualy found some of the commerical and at least one free simulator often get Lee-Kesler enthlapy wrong, although usually by a smallish amount, due to the mutliple roots you can get from the LK equation.

 

Have to admit, recently haven't made much progress due to workload but a keen tester would really help speed that up.

[ChE Matt]

Welp, let me know if I can help!

[ChE Matt]

One thing I am wondering.  I've come across some information but, nothing I can download easily.  Usually I have to grab physical property data out of a PDF and wrangle it into a form that I can use in excel.  Latest data set I'm trying to get my hands on is heat capacity factors.  I'll be tackling vapor pressure pretty soon.  Where do you get thermodynamic data?

[SilverShaded]

Simulators rely on ideal gas heat capacity correlations and use departure functions to get real gas and real liquid enthalpies/entropies (or heat capacities).  So usually what you can find is ideal gas heat capacity correlations.

YAWS is a good place for general information.

Edited by SilverShaded, 01 October 2021 - 03:42 AM.

[breizh]

Hi,

I like  this resource : Korean data base

화학공학소재연구정보센터(CHERIC) | 연구정보 | KDB | Pure Component Properties

 

Breizh

[ChE Matt]

Hi,

I like  this resource : Korean data base

화학공학소재연구정보센터(CHERIC) | 연구정보 | KDB | Pure Component Properties

 

Breizh

 

So maybe I haven't found it yet, but it looks like I have to go to each individual compound to view the data.  Sorry if it's staring me in the face, but is there somewhere on that website where there's a file download with the data?

 

Thanks!

 

edit: OK, I found this page, lists fortran code for calculating certain properties:

화학공학소재연구정보센터(CHERIC) | 연구정보 | KDB | Calculation Modules

 

But, I don't know if the thermodynamic data is stuffed in there.  I doubt it.

Edited by ChE Matt, 01 October 2021 - 12:45 PM.

[SilverShaded]

 

Hi,

I like  this resource : Korean data base

화학공학소재연구정보센터(CHERIC) | 연구정보 | KDB | Pure Component Properties

 

Breizh

 

So maybe I haven't found it yet, but it looks like I have to go to each individual compound to view the data.  Sorry if it's staring me in the face, but is there somewhere on that website where there's a file download with the data?

 

Thanks!

 

edit: OK, I found this page, lists fortran code for calculating certain properties:

화학공학소재연구정보센터(CHERIC) | 연구정보 | KDB | Calculation Modules

 

But, I don't know if the thermodynamic data is stuffed in there.  I doubt it.

 

If you have access to Hysys and know some VBA you can suck out the entire database.  Most of it is openly available info anyway so no copyright issues.