Posted by S. R. Naidu on November 22, 1999 at 01:19:56:
I am faced with a poorly performing DMC controller on a propylene / propane splitter. The operations group is currently running the unit with 30-40% less feed than original plant test condition. Due to the significant model mismatch, DMC controller has been turned off & the system is under regulatory control. Management wants to get back to DMC control with nontrivial fix / modeification that will assure reliable multivariable control over widely ranging operating conditions. Any ideas or suggestions will be welcome. Should I develop models befitting varying operating regions & then program the controller to choose the models pertinent to the active operating reion ?
Is it possible to configure the CCF to work with a collection of models - choose appropriate model & switch to another if necessary seamlessly ?