(64d) Real-Time Optimization Strategies for the Delayed Coking Unit | AIChE

(64d) Real-Time Optimization Strategies for the Delayed Coking Unit

Authors 

Richmond, P. - Presenter, Lamar University
Soni, C. K. - Presenter, Lamar University


The Delayed Coker Unit (DCU) stands as one of the most profitable units in the refinery due to high profit margins on heavy crude oil refining. However, because of the semi-continuous nature of this upgrading process, the DCU separation process is perturbed frequently. Thus, this unit has a very complex relationship between the controlled variables and the manipulated variables, and it can be very difficult to control and optimize. 

In this paper we study the control and optimization of a typical DCU obtained from the literature. The unit is modeled using a dynamic simulation package and it is optimized using an equation oriented Real-time Optimization (RTO) package.  The DCU dynamic simulation is configured to run under the control of an industrial Distributed Control System (DCS) in order to provide a realistic platform to study the associated optimization problem.

There are a number of recycle streams including a small part of Coke Drum vapors that get recycled into the fresh feed. This internal (hidden) recycle depends on the effect of the other recycle streams and greatly affects the plant economics.  In a high capacity DCU plant, small changes in the recycle rates can cause a huge financial penalty, so it is important to both maintain optimum recycle rates and keep the products within specifications.