Model advice

Discusses use of COCO, the process simulation and modelling software suite from AmsterCHEM, downloadable from http://www.cocosimulator.org

Moderator: jasper

Model advice

Postby HenkF » 06 April 2015, 13:25

Hi Jasper,
attached are two model files.

The COFE_BIO flowsheet is converging within 6 cycles (from reset), the COFE_BIO.reaction doesn't initialize well.
Problem seems to be within the Superheater model, a sub-flowsheet containing 3 heat-exhanger models.

The difference between both flowsheets files:
* added the furnace sub-flowsheet (reaction fs),
* changed definitions of a TEA package for the flowsheet and sub-flowsheets (reaction fs)
* removed steam temperature controller

Could you please advice me on how to continue with the reaction flowsheet?

Thanks!
Henk
Last edited by HenkF on 10 April 2015, 18:59, edited 1 time in total.
HenkF
 
Posts: 87
Joined: 24 February 2013, 14:50
Location: Hengelo,NL

Re: Model advice

Postby jasper » 07 April 2015, 20:08

In unit Super Heaters, edit the material types. For the material type "water", change the property package to "water". I reckon this is your intent, but TEA is selected and it fails the water flash (as the water is super-critical).

I am not sure whether this solves your problem. Let me know of the problems remain. But this was at least the cause for the initial guess failure.
User avatar
jasper
 
Posts: 1128
Joined: 24 October 2012, 15:33
Location: Spain

Re: Model advice

Postby HenkF » 09 April 2015, 15:08

That simple... ;)

I will try how the convergence is doing. Regarding this: what actually is the influence of the pertubation parameters?
I currently define the mass balance relative tolerance to be 1e-5, and the recycle tolerance at a 1e-6 level.
In your opinion: are these levels set tight enough or maybe too tight?

Maybe for a future release: could it be made possible to copy a Flowsheet Option set to an embedded Flowsheet?

thanks!
kind regards,
Henk
HenkF
 
Posts: 87
Joined: 24 February 2013, 14:50
Location: Hengelo,NL

Re: Model advice

Postby jasper » 10 April 2015, 07:12

The perturbation should not be too large, as you will linearize too much non-linear behaviour and increase the chance to step over discontinuities in what is being perturbed. On the other hand, they should not be too small, or you will be bothered by finite precision of the function evaluation introducing noice in the Jacobian. Typically you will want it to be a few orders larger than the precision on the function you are evaluating.

So if thermo tolerance is set to 1e-8, unit operations that use thermo could be around 1e-6, and the flowsheet around 1e-5. Of course if you have embedded flowsheets it gets a bit more complex.

Regarding thermo: embedded flowsheets can take thermo from feed or product material objects. This always makes sure consistent thermo is used.
User avatar
jasper
 
Posts: 1128
Joined: 24 October 2012, 15:33
Location: Spain


Return to COCO (AmsterCHEM)

Who is online

Users browsing this forum: No registered users and 8 guests