Researches On Modeling, Optimization And Control Of A Heat-Integrated Distillation Column System

Distillation is an important and energy-intensive separation technique for fluid mixtures in chemical and petrochemical industry. With the increasing competition in the chemical industry and an energy crisis in the world, lots of improved approaches for decreasing the operation cost in distillation process have been developed, e.g. optimal and advanced control technology, new techniques including the heat pump distillation, and heat-integrated distillation column system.A two-column heat-integrated distillation system is studied in detail in this dissertation. 50% energy savings can be achieved in this distillation system by running one of the columns at a higher pressure and integrating the condenser of the high-pressure (HP) column with the reboiler of the low-pressure (LP) column. Because the heat integration leads to significant difficulties in operation and control due to the strong coupling in material and energy of the two columns, it is a challenge to operate this system correctly so that the plant is operational and the energy savings are achieved.The main research wok and contributions of this dissertation are as follows:(1) The dynamic rigorous model of the heat-integrated distillation column system is introduced which is based on vapor-liquid equilibrium, component material balances and energy balance for each tray. The non-ideal liquid phase is computed by Wilson model. The compositions of vapor and liquid phases, vapor and liquid flow rates, temperature, pressure, and liquid holdups are the state variables in the model. The rigorous system model is a set of nonlinear differential and algebraic equations (DAEs) which are discretized by the orthogonal collocation on finite elements with three points. The model equations are solved by conventional Newton-Raphson algorithm.(2) It is crucial to estimate the tray efficiencies and feed composition for improving the accuracy of model and implementing of advanced control strategy for the heat-integrated distillation column system. A new parameter estimation approach is proposed which is addressed by turning weight factors in the least-squares dynamicoptimization, hi this work, through analysis of the solution procedure of the resulting optimization problem, it is found that parameter estimation with a weighted least-squares method is equivalent to a multi-variable closed-loop control system. The weighting factors in the objective function are equivalent to the controller parameters. Like any control system synthesis, the tuning can be made based on the output-input sensitivity analysis. The effectiveness of the proposed approach is illustrated by comparing the results of this method with the results of the general least-squares method.(3) To implement the dynamic optimization and control strategy, it is necessary to estimate the unmeasurable state variables in every time interval. A new approach, which is called optimization-predictor method, is proposed to estimate the unmeasurable state variables in such a large-scale system. The results show that the proposed approach is effective for real-time state estimation of the nonlinear large-scale heat-integrated distillation column system.(4) Two-level control strategy integrated dynamic optimization and control of a heat-integrated distillation column system is proposed to minimize the heat supply to HP column when the feed rate and composition are changed, and the product qualities are satisfied. The optimization variables are heat supply to HP column reboiler, HP column reflux rate, LP column reflux rate, and feed rate of LP column.Finally some conclusions and future researches are drawn in this dissertation.