MULTIVARIATE CONTROL STRUCTURE SYNTHESIS FOR FLUID CATALYTIC CRACKING UNIT USING RELATIVE GAIN ARRAY

Fluid catalytic cracking (FCC) is one of the most important process in the petrochemical industries. The process is multivariable, strongly nonlinear, highly interactive and is subject to many operational, safety and environmental constraints, posing challenging control problems. A novel mode has been developed that captures the complex dynamics of the process, which combines together models of the reactor with a detailed five lumped kinetic model for the riser, as well as regenerator, and fractionator. Control studies of closed-loop simulations were carried out using an advanced proportional-integral-derivative (PID) algorithm which implemented the discrete position algorithm with backward approximation and anti-windup. The method of relative gain array (RGA) was used for the evaluation of potential control pairings for the multivariate system for which the PID controller were implemented. The PID controllers were tuning using a genetic algorithm based optimization algorithm. The results indicate strong couplings among some of the control loops therefore there is a strong incentive for a more advanced multivariable controller such as nonlinear model predictive control (NMPC) based on economic objective.