Application of the rough set data analysis for the development of the control structure of an FCC unit
The rough set data analysis (RSDA) is widely used for the determination of non-linear relationships in many different areas. In this paper it is applied to investigate the dependencies among the variables in fluid catalytic cracking unit (FCC) consisting of reactor and regenerator parts. The analysis has been performed to understand FFC behaviour and to develop its control structure. About 140 operating points of an industrial FCC unit are considered and analysed using the rough set methodology. The ROSETTA toolkit is applied to discover the dependencies among the controlled and the manipulated variables. It is determined that the temperature of the regenerated catalyst has a significant influence on the product quality (motor octane number) and it should be controlled. A new control loop is designed and its controlled variable is the temperature in the regenerator unit. To avoid afterburning in the regenerator part of the FCC unit, several temperature measurements are recommended with the use of a special control unit, a high selector (HS).The manipulated variable of the new control loop is the feed flow of the bottom product of the main distillation column (BMC) in the reactor part of the FCC unit.