Custom Unit Stream Normalization Error

[nrgeng]

The Unit Operation based on COFE Flowsheet (Custom Unit) was used to configure a sub-flow sheet. An error occurred. The same error has occurred in two different simulations.

• A material steam (15x) was reported as non-normalized prohibiting the simulation from solving.
  • The subject stream functions in its respective flow sheet, but not when its solved flow sheet is loaded in a Custom Unit sub-flow sheet.
    The mole fractions add to 1.00000 (so streams are actually normalized).

Since I cannot complete my simulations with this error, I hope that you can help.

Non-normalized Stream Error Message

Sub-Flow Sheet v1.0 (Forum).jpg (206.21 KiB) Viewed 33287 times

[jasper]

Would it be possible to see the fsd?

[nrgeng]

I can provide more information but not the fsd file. I hope this is sufficient for a resolution.

The overall flow sheet uses two sub-flow sheets. The first sub-flow sheet compresses and partially dehydrates saturated ambient air, then passes the partially dried air to the second sub-flow sheet. The second sub-flow sheet should accept the output (Stream 15x) of the first sub-flow sheet because the stream was created by the first sub-flow sheet from saturated ambient air. The non-normalized composition error makes no sense to me.

The saturated ambient air stream (Stream 4 (2)) and two possible second sub-flow sheet input stream analyses are provided (see attached file). Stream 15x is created by the first sub-flow sheet.

Stream 15x as input elicits the error message from the second sub-flow sheet, but Stream 15x org does not. Neither Streams change when [Normalize Compositions] is selected with the right mouse button in Document Explorer. Hopefully you can unravel this mystery.

Stream 15x Compositions.pdf

Stream Analyses

(6.38 KiB) Downloaded 1527 times

[jasper]

Please contact me via my private mail, and I will send you an alpha of the new COFE. You can check whether this problem is still present, and validate that your other two problems have been resolved.

[nrgeng]

I installed the alpha software as instructed. Thanks. The Disappearing Stream Types and Toolbar Redundancy problems were not encountered during a brief use of the alpha s/w.

The normalization problem has been determined to be a partial change in stream component compositions when streams are changed. Switching from Stream 15x to Stream 15x_Test (water removed) changed all component compositions except the last two which are Argon and Nitrogen. The MW was recalculated but with two wrong compositions.

The attached file shows that not all of the component compositions were changed when Stream 15x_Test was connected to the sub-flow sheet Custom Unit after disconnecting Stream 15x. No other action was undertaken. Just disconnecting Stream 15x then connecting Stream 15x_Test in its place.

Does this reporting provide enough information to repair the problem?

Stream 15x Compositions 2.pdf

Partial Composition Data Overlay

(6.53 KiB) Downloaded 1551 times

[jasper]

The description by itself is not sufficient to find the problem. I cannot reproduce it after a few tries.

Can you construct an example with this behaviour that you can share?

[nrgeng]

Further testing revealed that the overall flow sheet solved when switched from dry to wet by beginning with the inner sub-flow sheet and working successively outward. Cannot switch from dry to wet or vice versa just by changing the input stream to the overall flow sheet due to non-normalized composition error. Unable to solve wet even with a slight composition change. The last two component compositions are not replaced with the new stream data as reported before. The error is still present in the alpha s/w.

I will attempt to construct an example with this behavior that can be shared. Thanks for your help.

[jasper]

I suspect that it has to do with thermo selections in outer and inner flowsheets. When reconstructing, try copying your problem, keep thermo setup and remove all relevant data. Then any new embedded unit may reproduce the issue.

[nrgeng]

I created a new, simple sub-flow sheet example to share, but the problem did not present itself because switching streams solved. I decided to re-create my original sub-flow sheet in the alpha version and hope for the best. Your comment seems to suggest the same. Thanks for the advice.

[nrgeng]

After re-creating the sub-flow sheets and the overall flow sheet in the alpha s/w, with "incomplete specification" of each stream (wiped clean), and connecting an input stream containing only dry air (four compounds), the overall flow sheet failed to solve with the following message.

error: calculate failed for unit Sub Flow System: Failed to flash internal material: CalcEquilibrium failed: Failed to wrap v1.0 CalcEquilibrium of type "TP" to v.1.1: CalcEquilibrium failed: CalcEquilibrium failed: TP flash failed: composition not set for one or more compounds

The last phrase TP flash failed: composition not set for one or more compounds may be connected to the non-normalized composition error messages. Stream data may not be exchanged for all stream composition array elements between the outer and inner flow sheets.

Does this help to identify the problem? I am running out of ideas. As I stated in a previous post, the simulation solves, i.e., no configuration logic errors, wet or dry if I start with the innermost sub-flow sheet and work outward.