Online Identification and Control of a Fluidized Catalytic Cracking Unit

M. Khalilian and R. Dhib (Canada)


Identification, Catalytic Cracking, Process Control, Optimum Predictive Control.


Model-based controllers are known to be efficient for maximizing process throughputs in presence of intense plant upsets and hard constraints. This work deals with an online implementation of an Optimum Predictive Controller (OPC) onto a reactor/regenerator in a Fluidized Catalytic Cracking Unit (FCCU) for a large refinery plant. The OPC package developed by Treiber Control Inc. was used to generate dynamic models using Pseudo-random Binary Sequence (PRBS) tests. Process data collected in time intervals, for the reactor/regenerator and gas plant, were recorded in a Process History Data (PHD), a refinery database system. A set of model-based multivariable controllers was designed and successfully implemented onto the FCCU process in the refinery. The controller runs on Honeywell's TDC3000 Distributed Control System Application Module.

Important Links:

Go Back