Researches On Modeling, Dynamical Behavior And Controllability Analysis Of A Novel Double-effect Distillation Process For Methanol Purification

Next to alkene and arene, methanol is the third massive organic product in chemical industry. Recently the uses of methanol for new sources of energy, especially for the fuel cell and the fuel of vehicles, have been in practice applications and are developing widely. Accompanied by this development, the market demand and the capacity of production of methanol are increasing rapidly. In the synthesized methanol industry, the purification of raw methanol is not only the important process determining the quality of product, but also a key factor influencing the energy cost of production. It is important for promoting the development of methanol production to study the new techniques for distillation of methanol with saving the energy and water consumptions. Along with the cost increase bringing by global energy source crisis, the 3-column double-effect distillation process has been widely adopted in China due to its ability of reducing energy consumption and annual cost.A novel double-effect distillation process (NLSF) reported by Wu and Chen (2005) has been studied in detail in this dissertation. Compared with the traditional 3-column double-effect distillation process widely used today, this novel process could reduce the heat consumption by 24.7% and soft-water consumption by 38.0%. However, this novel process enhances the coupling of different units in the system and the interactions between the two distillation columns, which results in the complexity of system responses to feed disturbances and increases significant difficulties in operation and control. Therefore, it is necessary to study the dynamical behavior and controllability of this novel double-effect distillation scheme before putting it into implementation of industrial application.In this dissertation, a high-fidelity model of NLSF process was developed first with the model parameters optimized. Then with this model, the open-loop dynamical behavior of NLSF was studied, the controllability analysis was carried out, a few of possible controlling schemes were proposed, the closed-loop dynamical behaviors of these schemes were studied with comparisons. At last, the difference of dynamical characteristics and controllability between NLSF and the other classical processes (LSF and FS) were also compared. The detailed work and main contributions of this dissertation are as follows:1. The rigorous steady-state and dynamic models of NLSF process were setup on chemical engineering principles with the coupling characteristics considered sufficiently and by use of the commercial software Aspen Plus and Aspen Dynamics. In order to make the dynamical model more reliable and rigorous, two distillation columns were simulated by RADFRAC models and their reboilers and condensers were simulated by independent HeatX models with rigorous hydraulics method and the equipment and feed stream parameters derived from the practice factory data.2. It is crucial to solve the optimization of steady-state models of the three processes for comparing their performances. The optimization of heat-integrated double-effect distillation system is a puzzle of optimization of multivariable nonlinear function with strong constraints. The method of distributed parallel genetic algorithm coupled with a sequential modular process simulator was employed to optimize the parameters of LSF, FS and NLSF processes. The energy consumption and product quality of the three optimized steady-state processes were analyzed comparatively.3. For the NLSF process of heat-integrated double-effect distillation, the coupling between variables is heavily, so it is necessary to study the dynamical behavior of system in order to setup proper control strategy. A detailed study on the open-loop dynamical response of the system to disturbances of feed and operation variable were carried out by observing the interrelations of system variables. On the summation of open-loop dynamical characteristics of NLSF and the combination of practice situations, several rational control strategies were proposed.4. The controllability analysis is necessary to compare the performances of different control strategies for NLSF. A few related control indexes were adopted in controllability analysis to predict and to distinguish the performances of different strategies. Six feed disturbances were chosen as the representatives of the real operating situations. The closed-loop dynamical performances of different control strategies under these disturbances were studied comparatively to check the results of controllability analysis and to find the best one.5. In order to understand the characteristics of the dynamical behavior of NLSF and to guide the choice of industrial methanol distillation process, the controllability analysis and the closed-loop dynamical behavior studies were carried out with different control strategies for LSF, FS and NLSF processes respectively. The optimal control strategy of each process was found. Then a comparative study on the dynamical response curves was carried out for the three processes with their own optimal control strategies.At the end of dissertation, the conclusions of present work and suggestions for further research were given.