Abstract
Heat-integrated distillation has long been considered as one of the technologies to reduce energy consumption in distillation columns. Despite the economic advantage, it introduces great difficulties in process operation due to the nonlinear, multivariable, and interacting characteristics of an energy-integrated system. In this study, a high-purity heat-integrated distillation system with light split/reverse configuration is utilized as a test case to separate methanol from water. Controllability of this system using nonlinear control algorithms is addressed in this study. A modeling/control approach is studied in this paper to handle the complicated characteristics of the system. A representative low-order nonlinear model is established and utilized to construct nonlinear simplified decouplers, thereby diagonalizing the input-output mapping of the distillation model. Linearizing feedback transformations are then adopted to reduce process nonlinearity for the nearly diagonalized system. Finally, PI controllers are designed on the basis of this transformed linear system. Significant performance improvement in control tests of the utilized nonlinear control approach over linear control strategy is achieved.
