convergence issue

[aken]

hi Jasper,
This problem solves in aspen but not in COCO. where can l see the output report, so l can try to troubleshoot the non convergence issue. It says l should see the full output report, but not sure where. the example is taken from this paper under distillation: https://www.academia.edu/keypass/YnBraV ... n_Students

[jasper]

The full report is the output report:

ChemSep (TM) - COL2 8.14
Copyright (c) 1988-2019 Harry Kooijman and Ross Taylor
License: 2100- 6 LITE free license
Build: 6741aaa23cca
Checking data file C:\Users\jasper\AppData\Local\Temp\CS_2_1A98_1D594BC_51346B50.sep
Cape Open unit name: Column_1
Cape-Open unit id: CS_2_1A98_1D594BC_51346B50
Cape-Open run level: normal
Cape Open Initialized
Cape Open Thermodynamic-properties enabled
Reading configuration
Reading component property libraries
1 74-84-0
2 74-98-6
3 75-28-5
4 106-97-8
5 78-78-4
Reading thermodynamic property options
Reading reaction data
Reading specifications
Checking Component Data
74-84-0 Ethane
74-98-6 Propane
75-28-5 Isobutane
106-97-8 N-butane
78-78-4 Isopentane
Determining feed conditions
Run level: Initialization
Generating initial flow profiles
Generating initial composition profiles
Init 461 milliseconds
Run level: Complete model

Iteration log(Err/Tol)
Warning! Feed pressure < column pressure (Feed 1)
0 4.2211
1 5.6029
2 5.4828
3 5.8369
4 5.4045
5 6.1241
6 6.3122
7 6.1140
8 6.3097
9 6.4650
10 6.3675
11 6.0398
12 5.9676
13 6.0710
14 6.3645
15 5.9570
16 6.1321
17 6.2340
18 6.2392
19 6.2977
20 6.1682
21 6.3591
22 6.4565
23 6.3254
24 6.0348
25 5.9545
26 5.8462
27 6.1231
28 5.9488
29 6.1172
30 6.0107
Run level: Report
Convergence not obtained in 30 iterations
Time 674 milliseconds
Cleanup called with F
FixMem driver done

run time: 3.1 seconds

[jasper]

Quoting Ross Taylor: you could switch the solution method to 2-pass Ideal K and constant H first. Also, he notices that the column is too tall.

[aken]

Hi. As l read down the paper l noticed their feed tray position was positioned differently as the optimized it. I then took for 30 stages, and put feed on tray number 9 and boom it converged. For 30 stages, and tray 15 (middle) used as a feed tray and with the same specifications, it doesn't converge. My question is, is there a way to tell that non convergence of a distillation column is due insufficient number of trays, or lower reflux ratio than the minimum or its just infeasible separation. if one knows then it will be easy to troubleshoot nonconvergence.

[jasper]

Again borrowing on Ross Taylor's wisdom: the shortcut method (FUG, Fenske - Underwood - Gilliland), in the shortcut panel in the Chemsep GUI should give you quick insight in what may work well. This is not the final answer, as the shortcut method is inexact.

[aken]

Hi. I noticed the FUG only appears after the simulation has converged.

[jasper]

That is correct. The fug panel uses results from a simulation.

Attached is a document that contains tips to solve ChemSep columns if it does not work automatically.