Advanced process control (APC) using straightforward design and deployment of model predictive control (MPC) with System 800xA DCS enables higher level of automation and optimization of crude oil separation
Distillation columns are widely deployed in the process industry. Their use is recommended when there is a need to separate components that have different boiling points. The main idea is to introduce the raw mix of components, usually in liquid form, into the middle section of the distillation column. Through successive vaporization and condensation steps the low boiling point components are concentrated at the top of the column and the high boiling point components are concentrated at the bottom. A typical example is the separation of crude oil into components such as gasoline, kerosene and diesel. Since each step influences the others, the problem is naturally multivariable.
However the process is highly repeatable and thus well suited to modeling via empirical or data-driven modeling. Typical process variables are temperatures, pressures and compositions at the different levels of the column, while the main actuators are feed, firing, reboilers for the column bottom, pump-around flows, cooling rates and overhead pressure control bypass.