Simultaneous Simulation And Optimization Of Air Separation Process

Air separation unit (ASU) is an important chemical process that produces high-purity industrial gases (oxygen, nitrogen, argon and so on) through separation of air. It plays an significant role in the national economy as the products are widely applied in chemical industry, petrochemical industry, metallurgy, electric engineering, energy engineering, medical treatment, aeronautic engineering and so on. However, it is a huge technical challenge to meet the demand of high purity and strong connection among equipment during the simulation and optimization of the large-scale complex process, thus choosing a proper modeling method becomes a major issue. Equation-oriented approach (EO) has gained great recognition in the field of chemical simulation and optimization owning to the good convergence properties, high calculation efficiency and the convenience of switching between simulation and optimization. Focusing on the tight coupling variables, strong nonlinearity and dimension difficulty, some studies on modeling, simulation and optimization of ASU based on EO approach are carried out. The contributions of the thesis include the following four aspects.1. Thermodynamic property calculation methods, PR and RKS-BM are chosen for the cryogenic system. These methods, which are based on cubic equation of state (CEOS), are successfully solved with the EO approach by adding derivative constraints. The equations in the thermodynamic property model can be integrated with the MESH equations as an EO model instead of embedded in the sequential model.2. Physical parameter estimation based on air distillation tower are carried out to get precise binary interaction factors of PR and RKS-BM. Besides the difference of search path between sequential method and EO method are discussed, demonstrating the good performance of convergence of the EO approach.3. Considering the bad convergence of Aspen Plus, ASU are simulated and optimized based on the EO approach in this thesis. Property calculation equations and MESH equations are solved simultaneously under several operation conditions with the same initial values. The results shows the accuracy and high efficiency of the EO model.4. The influence of the operation variables and constraints of ASU are discussed, optimization problems aimed at energy-consumption and products output are proposed and implemented under load variation. The good results can provide guidence in industrial operation. Besides, feasible region and thermal-coupling variation trend are also analyzed based on the EO model.