Benzene Column Delta T Conrol Using Reflux

If you have the opportunity, I'd strongly recommend you read Shinskey's
"Distillation Control" (McGraw-Hill) in particular the chapter on reboiler control.

the reason operators like flow control to the reboiler is that it is predictable and stable.

you can use temperature to modulate the steam flow setpoint but it is not a good idea to control the steam valve by the temperature controller only. The swings are too wild.


Given that you have old instrumentation, leave it alone, when you go to cascade controls you also have to build into your controls various constraints (pressure, boilup, L/V, etc.) to keep the column stable. That is easy for modern DCS type systems, though commissioning the controls is potentially difficult.

Shinskey's book is a good one to have, Page Buckley is also a good refeence-though mostly dealing with modeling.

The column may have an inherently unstable control scheme. Check with a reference by Henry Kister, the Perry's Chemical Engineers Handbook, or those mentioned by others. Then implement a control scheme that is stable. This may be the root of the problem the operators prefer to run it on manual and not automatic.

I have seen several unstable columns that are run in "manual" or have layers of DCS constraints placed on them to make up for this problem. My opinion on this last part is that controls engineers over emphasize programming over the basics.