Process design and control structure screening based on economic and static controllability criteria

• Published in: Computers & chemical engineering 2001, Vol.25 (1), p.177-188
• Main Authors: Seferlis, Panagiotis, Grievink, Johan
• Format: Article
• Language: English
• Subjects: Continuation method; Control structure selection; Parametric optimization; Process design; Sensitivity analysis; Static controllability
• ISSN: 0098-1354; 1873-4375
• DOI: 10.1016/S0098-1354(00)00641-4

A method is developed that screens process flowsheet and control system configurations, defined in terms of selected controlled and manipulated variables that exhibit poor steady-state behavior under the presence of multiple simultaneous disturbances and model parameter variations. The steady-state disturbance rejection characteristics of the nonlinear process models are considered in conjunction with economic criteria for the elimination from further consideration of poor designs early in the design process. A wide variety of static control objectives for the system is incorporated within an optimization framework that evaluates the required contribution of the selected input variables to alleviate the effects of the imposed disturbances on the control objectives specified in a hierarchical order. Sensitivity information of the optimal solution to the control design problem is utilized to identify the most influential design parameters on static controllability performance, hence providing the guideline for improving the disturbance rejection properties of the flowsheet. The proposed approach allows the study of plants with unequal number of manipulated and controlled variables and the investigation of cases with saturation of manipulated variables and hard bounds on the controlled variables. The method has been applied to a number of different designs for a complex flowsheet that involves multiple reactive and separation steps with recycle. Reactive distillation configuration proved to exhibit superior economic and static controllability performance compared with the conventional reactor–separator scheme for combined reaction kinetic parameter variations.