5 replies to this topic

[marchese85]

hello

Edited by marchese85, 07 September 2010 - 04:34 PM.

[saeed.s]

It s difficult troubleshooting a simulation without seeing it, but you can also use Aspen batchsep software that is a powerful & handy software in batch separation

[Padmakar Katre]

Dear,
First of all let me know about your steady-sate simulation results. Are these in line with the standard operating data you have based on which you are going to validate your model. As far as you are using a 3-phase distillation template and NRTL as activity coefficient model I don't find any problem in handling the simulation. (Try with other activity coefficient models available.

[Chellani]

To answer few of your questions
1. Solving method - Only sparse continuation solver can detect phase splitting on intermediate stages. Legacy and modified I/O would always detect phase splitting for condenser/receiver
2. Damping factor - try not to use it initially. If you are really interested in trying it, try adaptive damping which would tune your damping factor. Hope this option is available in HYSYS, I am using Honeywell's UniSim Design.
3. Radio button in 2/3 phase of column - Try 'Auto Water Draws', again; not sure about its availability in HYSYS.

Solving method really doesn't play any role in dynamic simulation. Not sure that you've got solid phase in your model, if yes; you need to put some extra efforts and intelligence. Not sure about how more than 3 phases are handled in dynamic simulation in HYSYS, Honeywell's UniSim Design provides few good options for doing that.

Edited by Chellani, 28 June 2010 - 03:13 AM.

[joerd]

You'll have to define two fluid packages, sharing the same component list. Both would use NRTL, but you will specify different BIPs. In the flowsheet, you can pick the fluid package for each unit operation. Hysys will automatically insert stream cutters where applicable; in case of a subflowsheet(column)-to-flowsheet interface, this is handled on the subflowsheet transition page instead of by a stream cutter operation. These serve to help you specify how the transition from one property package to the other should be handled, i.e. by PH-flash, PT-flash, etc.
This will not help convergence of your column, by the way.

[joerd]

The easiest way to accomplish what you want is to model your column+reboiler separately from the accumulator. Make the column a "reboiled absorber" then connect the overhead stream to a condenser and a separator. Liquids from the separator should be fed back to the column as reflux via a recycle operation.