Improvements to distillation unit model

[jghouse88]

Pending items:

• [x] Tray model replacing component temp and press ref
• [ ] Add support the newer version of the square flash formulation where we removed the need for bubble and dew point
• [ ] Initialization improvement to include a linear temperature profile before activating energy balance
• [ ] Add compatibility to scaling framework
• [x] Add compatibility to generic properties framework
• [x] Separate out make_phase_split and propagate_state methods as utility methods and use these in condenser and reboiler unit models
• [ ] Add report() method for column
• [ ] Check that flowsheet.report() works with distillation columns - there appears to be a bug somewhere.

[ksbeattie]

@jghouse88 is this still possible for Aug? Is there someone else who could take it over?

[adowling2]

@agarciadiego Are you using this branch?

[lbianchi-lbl]

@jghouse88 @adowling2 we're looking at open PRs in preparation for the August release. Is this work expected to be completed in the next few days? Otherwise, we can keep tracking this PR as part of the November release.

[dallan-keylogic]

Check that flowsheet.report() works with distillation columns - there appears to be a bug somewhere.

Found the cause of this while investigating #480. The UnitModelBlock.add_inlet_port() method currently adds an additional _state_block attribute to a port while generating it. That attaches the state_block to the feed tray's port, but because the attribute is not one added by Pyomo, it isn't copied when the port is extended to the level of the overall column. We are currently figuring out a better method to make stream information available, but the distillation column is not at fault for this.

The condenser.distillate and reboiler.bottoms ports should be extended to the column level as well. Otherwise Pyomo's SequentialDecomposition tool will get confused when trying to build a graph of the flowsheet and won't record that arcs attached to those ports are attached to the distillation column.

[ksbeattie]

@jghouse88, seems like this should get a new assignee. Can you pick one?

[ksbeattie]

@jghouse88 is this still relevant?

[andrewlee94]

@ksbeattie It is still very relevant, but we lack the resources to address it right now.

[adowling2]

@agarciadiego has been working on extensions of this. We have ideas on improving initialization. But our work has been exploratory. It is far from generalized.

[andrewlee94]

I was thinking @agarciadiego would be a good person to take this over, however he probably won;t have time in the next few months as I believe he is committed to other projects.