(188l) Steady State Real Time Optimization of a Reactor-Separator-Recycle Process

Kaistha, N., IIT Kanpur
Real time optimization (RTO) for economic (near) optimum steady operation of a reactor-separator-recycle process module is evaluated. The RTO obtains the optimum value of the dominant unconstrained regulatory layer setpoint, namely, the reactor limiting reactant composition, by adjusting unknown parameters of a simple reaction kinetic model to best fit the available overall plant material balance data, and also the separator tray efficiency to best fit the measured column tray temperature profile. Rigorous non-linear dynamic simulation results demonstrate that for slow long term changes in the plant characteristics such as catalyst deactivation, column fouling and throughput / fresh feed composition changes, RTO delivers 4-5% higher energy savings and maximum achievable throughput compared to process operation at constant setpoint. Considerations of model adequacy are highlighted.