6 replies to this topic

[yclee99]

Dear All,

 

As I known, the steady state simulator does not take piping volume or equipment volume (such as simple heat exchanger volume, compressor volume) into consideration. How about the process dynamic simulator? Does the process dynamic simulator normally take volume into consideration, besides tank, vessel, column, separator's volume ? 

 

 

[Bobby Strain]

Yes.

Bobby

[yclee99]

Hi Bobby,

 

Thanks for the reply.

 

Can I say that all the volume such as piping volume are taking into consideration? How you deal with the robustness (i.e. very small volume which might cause instability of the model) or performance (i.e. dynamic performance may be slow down due to increasing number of pressure/flow nodes) of the model? 

 

Actually, I am working in OTS (Operator Training Simulation). Normally, the process dynamic model for the OTS is quite big as we don't separate the model to smaller model. Otherwise, it does not fit the training purpose. 

Edited by yclee99, 20 July 2014 - 11:42 PM.

[bponcet]

yclee99,
for OTS you may consider simplified models for columns, piping etc.
I know several vendors doing that.
Simplified models can run very fast,
consider that in most cases for training you don't really need
accurate results but an idea of how flows / pressures / temperatures change
in order to react quickly and properly.

[yclee99]

Hi bponcet,

 

You are right. That is typical process model for OTS. However, some clients are asking to include volume to several places which I mentioned above. I don't really get it why volume to be considered for such places? I believe with and without volume the dynamic behavior are quite similar, except the rate of pressure change if believe. Or is there something that I miss out which client is looking for?

[bponcet]

adding volume shouldn't be that difficult if you consider, for example, gas law with constant (average) compressibility factor (as we do for segments in pipelines),
it should be not too difficult to integrate the resulting ODE set,
for VLE things are more difficult but, again, you can adopt simplified models (constant volatility or similar) and integrate these.

More information could be available if you provide some details about your project.

[kabtik]

If you are developing an OTS then fidelity is very important and if it is then the more accurately you capture the existing facility the better it is. If you are not going to include volumes then are you saying that there will be nothing like holdup in your system. If you cutback on volume for what ever reason then you may also cut down on other things in the same proportion so as to get a dynamic response that simulates the plant's dynamics. Perhaps your OTS is going to be just a steady state model, which may not be a good tool for training.