Dynamic Simulation And Optimization Of Grade Transition For Ethylene Slurry Polymerization Process Based On Simultaneous Approach

Due to the low production cost, versatile mechanical and rheological properties, high density polyethylene (HDPE) is one of the largest synthetic commodity polymers that is widely used in the film, pipe, and container industries. Slurry polymerization of ethylene is an effective process to produce HDPE. To meet different specifications of end-use properties for HDPE, a plant usually produces several grades of the product with different qualities through the same process but with different operating conditions. Fast grade transition has been the significant effort of both academy and industry during the past decades. In this paper, a rigorous equation-oriented (EO) model is first developed for the dynamic process based on polymerization mechanism with multiple-site-type Ziegler-Natta catalyst. The average molecular weight calculated based on moment method is used to represent the polymer quality. Perturbed-chain statistical associating fluid theory (PC-SAFT) is used to generate thermodynamic data of the polymer system, based on which, surrogate models are further developed as a part of the EO model to calculate the thermodynamic properties of each streams. Simultaneous method is used to solve the differential-algebraic equation (DAE) system. The orthogonal collocation finite elements (OCFE) method is used for the discretization. With Lagrange interpolation polynomials and Gauss collocation, the original model is reformulated to a large-scale nonlinear programming (NLP) problem. A general thermodynamic calculation package is established to overcoming the difficulty of solving large-scale problem, which can also improve the efficiency and accuracy of simulation. Industrial grade transition is established by changing operation conditions. Considering safety of production process and stability of polyethylene, the optimization problem of minimizing transition time is formulated. To solve the problem, the main research contents and contributions of this paper are as follows:1.A rigorous dynamic model is developed for the process based on polymerization mechanism with multiple-site-type Ziegler-Natta catalyst. The average molecular weight based on moment method is used to represent the polymer quality. PID controllers are used to control temperature, pressure and liquid level of reactor during the grade transition process to ensure the production stability.2. Surrogate models based on Kriging technique are proposed and developed to describe the relationship between the physical conditions and the material properties for the PC-SAFT equations. The good simulation results at several steady-states show the accuracy and robustness of the Kriging models. Through calculating first-order and second-order derivatives of Kriging function, a general thermodynamic calculation package is established, which can improve the efficiency and accuracy of simulation.3. Simultaneous approach based on orthogonal collocation finite elements is studied. The equation-oriented (EO) model is obtained by discretizing all state and control variables.4. A series of dynamic simulation of grade transition process is completed based on the simultaneous approach. The average molecular weight profile and other state variable profile are presented. The equation-oriented (EO) model is validated in comparison to that in Aspen Dynamics. Industrial grade transition process simulation is focused. Both results indicate the EO model can predict the quality of polymer well.5. In order to solve dynamic optimization problem, grade transition time serves as objective function while polymer quality and production safety is ensured. A two-layer solving strategy is presented. Base on the approach, optimal operation conditions profile of industrial grade transition process is obtained.