| The new generations of Ziegler-Natta catalysts and ethylene polymerization processes are always the core inpetus of polyethylene manufacturing technology. However, the multi-site Ziegler-Natta catalysts usually exhibit special complexity of its ethylene polymerization kinetics. Therefore, it is still hard to accurately quantify the minimum number of active site types and apparent kinetic behaviors for each of site type of multi-site Ziegler-Natta catalysts, morevoer, the effect of pre-polymerization on the polymerization kinetics of catalyst is also still unclear. Besides, the complex thermodymic properties, the series reactor system and complex microstructure of polyethylene product for the new type of Borstar bimodal polyethylene process also present a challenge for building a theoretical model of full process. Deeply understanding the ethylene polymerization kinetics of multi-site Ziegler-Natta catalysts with the method that combines experiment and modeling, then based on above, modeling of Borstar bimodal polyethylene process, and further accurately controling the microstructure of polyethylene in large-scale process, which have important theoretical significances and value of industrial applications.This thesis quantified the effect of pre-polymerization on kinetics for ethylene homo-polymerization and ethylene/1-hexene copolymerization based on the simplified single-site Ziegler-Natta catalyst polymerization model. Moreover, the surface morphology of the pre-polymer particles correlated to and explained these changes in polymerization kinetics. Then, the new methods for multi-site Ziegler-Natta catalysts were developed to estimate the minimum number of active site types and apparent kinetic parameters on each of site type needed to accurately model polymerization kinetics and polymer microstructural data. Next, the equation of state of perturbed-chain statistical associating fluid theory (PC-SAFT EOS) with updated parameters was applied to predict thermodynamic properties and phase equilibria for the complex system of containing supercritical fluid and polymer in ethylene supercritical coordination polymerization process. Finally, a robust steady-state model was built-up for an industrial successive supercritical slurry-phase and gas-phase catalytic polymerization reactors of Borstar bimodal polyethylene process. In addition, the feeding conditions of the reactors were also analyzed.The following results were achieved in this thesis:1. A simple single-site polymerization kinetic model was used to quantify the effects of several pre-polymerization conditions (temperature, pressure, time,1-hexene/ethylene molar ratio and hydrogen) on the apparent kinetic constants of ethylene homo-polymerization and ethylene/1-hexene copolymerization with a 4th generation commercial Ziegler-Natta catalyst. And the surface morphology of pre-polymer particles also correlated to and explained these changes in apparent polymerization kinetics. It was found that the apparent propagation rate constant (kP) was the highest when intermediate pre-polymerization temperatures and time (50 ℃ and 30 minutes), high ethylene pressures, and low 1-hexene/ethylene molar ratios were used. The apparent activation rate constants (Ka) increased for low pre-polymerization temperatures, long pre-polymerization time, high ethylene pre-polymerization pressures, and high 1-hexene/ethylene ratios. Finally, the apparent deactivation rate constants (kd) decreased merely with increasing pre-polymerization temperatures. When hydrogen was added to the polymerization, the value of kP decreased, while Ka and kd increased, but addition of hydrogen to the main polymerization does not affect the trends for polymerization kinetics without hydrogen. In addition, a rule of thumb for this catalyst was proposed with correlating surface morphology of the pre-polymer particles through a scanning electron microscopy (SEM) to polymerization kinetics that pre-polymer particles with high concentration of polymer fibrils have lower apparent kP and Ka.2. A method was proposed to estimate the minimum number of active site types for multi-site catalyst by deconvoluting the time-dependent molecular weight distributions and average comonomer fraction profiles (MWD/SCBD) of ethylene/1-hexene copolymers made with a 4th generation commercial Ziegler-Natta catalyst in a semi-batch slurry reactor. These distributions were measured by a high-temperature gel permeation chromatography equipped with an infrared detector (GPC-IR) at four different polymerization time and different 1-hexene contents. Then these information were used to infer how the fractions of polymer made on each site type varied with polymerization time. This model estimates the minimum number of active site types needed to describe these copolymers, the MWD of polymer populations made on each site type, and their average comonomer fractions. This model is useful to quantify the microstructure of olefin copolymers made with multi-site catalysts using the least number of adjustable parameters.3. A method was proposed to quantify the polymerization kinetics of ethylene and a-olefins with a 4th generation commercial Ziegler-Natta catalyst based on the approach of estimating the minimum number of active site types for multi-site catalyst. The method determines the leading apparent polymerization kinetic constants for each of active site in a Ziegler-Natta catalyst by simultaneously fitting the instantaneous polymerization rate, cumulative polymer yield, and polymer molecular weight distribution measured at different time during a series of ethylene homo-polymerization and ethylene/a-olefins copolymerization experiments in a semi-batch slurry reactor. For each site type this approach estimates the apparent rate constants for site activation (Ka,i), monomer/comonomer propagation (Kp,i or KP,i), and site deactivation (kd,i) needed to describe ethylene homo-polymerization and ethylene/a- olefins copolymerization kinetic behaviors, as well as average molecular weights (Mn and Mw) and molecular weight distributions (MWD) of each polymer population needed to describe the MWD of these polymers. This approach quantifies the behavior of olefin polymerization with multi- site catalysts using the least number of adjustable parameters needed to consistently model polymerization kinetics and polymer microstructural data.4. The PC-SAFT EOS with updated parameters was applied to describe the thermodynamic properties and phase equilibria for complex system of containing supercritical fluid and polymer in ethylene supercritical coordination polymerization process with a 4th generation commercial Ziegler-Natta catalyst. Based on the data of literatures, the pure components parameters and binary interaction parameters of PC-SAFT EOS were regressed. The re-parameterized PC-SAFT EOS can accurately predict the thermodynamic properties and phase equilibria for light components and complex system with polymer of supercritical ethylene coordination polymerization system in a wide range of temperature (340-380 K) and pressure (2.0-7.0 MPa). The re-parameterized PC- SAFT EOS was validated by real industrial data. The calculated H2/C2H4 molar ratio and slurry density in supercritical loop reactor and H2/C2H4 and 1-C4H8/C2H4 molar ratio and recycle gas density in fluidized bed reactor agree very well with industrial data with an average error less than 3%.5. A robust steady-state model was built up for an industrial successive supercritical slurry-phase and gas-phase catalytic polymerization reactors of Borstar bimodal polyethylene process by an advanced software tool, Polymer Plus. The full process model included the rigorous re-parameterized PC-SAFT EOS thermodynamic model, multi-site Ziegler-Natta catalyst polymerization model, and reasonable reactor model. A step iterative strategy was applied to finely tune the kinetic parameters on each of site type needed to describe the key operating parameters, average polymer properties and MWD of polymer in each reactor. This model is capable of predicting the process variables and polymer properties and validated by real plant data under four kinds of multi-steady-state operating conditions. Thereafter, the effects of feeding rate for the main reactors such as catalysts, ethylene, hydrogen and propane on the average polymer properties and bimodal MWD were also revealed. |
PDF Full Text Request