Active Constraint Regions for Optimal Operation of Chemical Processes

Whendesigning the control structure ofa chemical plant,withoptimaloperationand control inmind,it is important to know how the active set of constraints changes with disturbances.The generation of optimal active constraints regions using optimization generally is very time-consuming, and this paper discusses how to use process knowledge to minimize the need for numerical calculations—that is, minimize the number of optimization problems that must be solved. We consider the case of two disturbances, as this can be nicely represented graphically. In this paper, we study economically optimal operation of a reactorseparatorrecycle process and show how the set of active constraints changes with feed flow rate and energy cost as disturbances.Wealsoidentifytheeconomicalandphysicalbottleneckoftheprocess.Forthiscasestudyprocess,we find fivedistinct regions with different active sets. This illustrates that, even for simple cases, the map of active constraint regions can be quite complex.