【作者】 冯连芳；

【导师】 王凯； 胡国华；

【作者基本信息】 浙江大学 ， 化学工程与技术， 2006， 博士

【摘要】 聚合反应机理、聚合物体系以及聚合反应器操作的复杂性、耦合性和高度非线性，聚合反应过程状态和聚合物质量表征的关键参数（如聚合率、分子量等）在线测量的困难，聚合过程的系统设计、流程模拟与先进控制很大程度上依赖于以严格反应机理和热力学状态方程为基础的全流程数学建模，成为当前国际主流发展方向。乙烯工业状况是一个国家工业现代化程度的重要标志，而乙烯工业中最能体现科学技术水平的是聚烯烃技术。基于反应特性和过程特性的烯烃聚合过程模型化与优化命题涉及聚合物化学和反应动力学、多尺度聚合反应器模型化、聚合物结构性能和质量表征、大规模模型求解和优化计算、聚合反应过程控制等多方面的融合，交叉了高分子化学、聚合反应工程、高分子材料、过程系统工程多学科，无论在理论研究还是实际应用都是目前急需解决、极有挑战性的前沿研究课题。论文在对聚烯烃工程技术及其发展趋势系统了解和认真把握的基础上，以代表性的丙烯液相本体聚合和气相聚合串联的工艺流程，从工业化聚合装置的角度对丙烯液相本体聚合反应器、丙烯气相聚合反应器进行了剖析；研究了丙烯聚合复杂体系的热力学物性计算模型及其参数确定方法；从烯烃聚合反应过程的催化机理和反应动力学入手，基于Ziegler-Natta催化聚合的多活性中心反应动力学对液相和气相多个反应器串联的工业丙烯聚合装置全流程动态建模；进行了反应器操作条件分析、动态特性分析、牌号过渡模拟，以及流程优化模拟。论文取得了以下的创新性研究结果：1、通过再参数化PC-SAFT状态方程，建立了丙烯液相本体和丙烯气相聚合反应体系中丙烯—氢气—聚丙烯体系各组分物性的计算方法。以经典文献数据为基准，分别得到了丙烯、氢气、聚丙烯纯组分的PC-SAFT方程的各3个模型参数，包括链段数m、链段直径σ、能量参数ε／k_B以及丙烯—氢气、丙烯—聚丙烯相平衡的二元交互参数k_ij。采用本文修正参数的PC-SAFT状态方程计算氢气、丙烯、聚丙烯热力学性质的精度优于P-R方程、S-L方程和文献参数的PC-SAFT状态方程。2、综合分析Ziegler-Natta催化丙烯聚合反应机理，建立了适合工业装置建模的、简化的丙烯聚合基元反应组合，包括助催化剂活化、单体活化、链引发、链增长、向单体链转移、向氢气链转移、催化剂自失活反应，并确定了各基元反应的动力学常数及反应活化能的合理范围。依据高、低熔融指数两个不同牌号聚丙烯样品的分子量及分布分析，确认该催化体系的活性位为6个。动力学参数的敏感性分析表明，链增长速率常数和链失活速率常数对聚合率敏感、链转移常数和链增长速率对分子量敏感，为聚合过程建模指明了模型参数的修正策略。3、从流体混合的角度对搅拌釜式丙烯液相本体聚合反应器进行了工程剖析，明确反应器优化的原则是维持聚丙烯颗粒的悬浮、促进分子量调节剂氢气在聚合体系中的分散和传质、强化液态丙烯蒸发撤热和丙烯循环气在反应器内的分散以及夹套的传热。实验研究了Hypol工艺液相搅拌釜反应器的气液搅拌功率特性、气含率规律和气液搅拌传热特性，获得了一系列新型的实验关联式，可用于该类搅拌反应器的设计和优化。其中：平均气含率：ε_g=0.0393[P_G+P_A)／V]^0.53V_s^0.254通气状态搅拌功率特性：N_PG=7.0N_Q^-0.0574Fr^-0.233通气状态搅拌传热特性：Nu=0.675Hz^0.236Pr^0.33Vis^0.14Hz=Re^2.73+6.22×10⁹(N_Q-0.70Fr^1.85)Re_G4、以搅拌流化床丙烯气相聚合反应器为对象，对框式搅拌桨、双层锚式搅拌桨与多孔分布板、半锥帽分布板协同作用下的搅拌功率特性、流化特性、压力脉动进行研究，考察搅拌桨、气体分布板、粉粒直径、表观气速对流态化的影响规律，建立了临界流化速度、床层压降的关联式。U_mf=0.083×(（ρ_p-ρ_g）^0.69)╱(μ_g^0.38ρ_g^0.29)×d_p^1.07△P=3.55×10^-4L₀^1.15（1-ε₀）（ρ_p-ρ_g）搅拌转速提高，使床层压降△P与表观气速U_g特性曲线的转折点变平缓；完全流态化状态时，搅拌功率与粉粒直径、床层高度、分布板结构基本无关；床层压降与搅拌桨形式、搅拌转速无关。搅拌流化床中搅拌的作用在于防止颗粒的粘结、消除床层中的沟流、抑止床层中的节涌、稳定流态化、提高流化质量。5、搅拌流化床中，优化的搅拌桨叶应该有适当数目的水平叶片以抑制和破碎气泡、消除沟流，有一定数目的垂直叶片起刮壁作用，并且垂直叶片应避免一直从床层底部延伸到床层顶部防止诱导气泡短路。上下双锚式搅拌桨的流化质量优于框式搅拌桨，在搅拌的协同作用下导向型半锥帽分布器的流化质量优于多孔板分布器。明确了搅拌流化床反应器对不同粒径聚丙烯的适用性，提出粒径分布均匀化、增加床层高度、适当提高气速的优化策略不仅可以将Hypol气相流化床由粉末状聚丙烯转换成小球型颗粒聚丙烯的生产，而且能强化生产负荷、提高反应器的时空产率。6、分析工业化丙烯聚合流程，以各反应器聚丙烯生成量为考核标准建立了基于单活性位反应动力学的全流程稳态模型，以聚合物的分子尺度质量指标——分子量及其分布为目标建立了基于多活性位反应动力学的全流程稳态和动态模型，并采用工业装置操作数据进行了验证。在此基础上，模拟分析并阐明了反应器操作条件，如反应温度、反应器液位、气相氢气浓度、进料流量，对聚合率、聚合物分子量等过程状态量的影响规律；考察了脉冲扰动、过程阶跃、测量噪声作用下的丙烯聚合反应过程的动态特性，为过程的优化控制提供定量依据；提出并模拟验证了氢气的气液传质行为影响丙烯液相本体聚合反应过程的动态特性的新观点；对牌号过渡过程的模拟表明所建立的全流程动态模型有很好的适应性。7、应用所建立的全流程稳态和动态模型，对丙烯聚合流程组合的优化进行了模拟研究，提出了“液相+气相”和“液相+气相+气相”的优化流程，确定了相应的优化操作条件，可以大幅度提高聚合反应装置的时空产率。对生产宽分子量分布聚丙烯的可能性进行了模拟分析，表明液气两釜串联流程最适合生产宽的分子量分布的聚丙烯。更多还原

【Abstract】 Polymerization mechanisms, systems and reactors are often complex, coupled and highly non-linear. It is difficult to assess, on-line, the state of the polymerization process and the quality of the polymer product such as the degree of polymerization and polymer molecular weight. The future development trend in this area is that the design, simulation and control of polymerization processes reply more and more on thorough mathematical modeling based on rigorous polymerization mechanisms and thermodynamic state equations. The level of ethylene industry, especially that of the polyolefin industry, is an important indicator of a country’s level of industrialization. Modeling and optimization of olefin polymerization processes based on polymerization and process characteristics have to integrate polymer chemistry and polymerization kinetics, multi-scale polymerization reactor modeling, characterization of polymer structures and quality, large-scale model solution search and optimization calculations, process control, etc. It is an interdisciplinary field with many unsolved and challenging fundamental research topics and practical applications, relating to macromolecular chemistry, polymer reaction engineering, polymer material and process system engineering.The aim of this work is to develop a rigorous mathematical model for an industrial propylene polymerization process in order to significantly improve its performances. For that purpose, typical industrial slurry and gas-phase propylene polymerization processes composed of slurry and fluidized-bed reactors in series are analyzed based on a systematic evaluation of the current olefin polymerization technologies and its future development trend. A model is developed to calculate thermodynamic and physical properties of the propylene polymerization system. A method is proposed to validate the parameters of the model. A dynamic model is then built-up for an industrial Ziegler-Natta catalyzed propylene polymerization process composed of two slurry and two fluidized-bed reactors in series. Finally, the operating conditions of the reactors and dynamic characteristics of the process are analyzed; the grade change and process optimization are simulated. Most relevant results obtained in this work are summarized as follows.1. The equation of state of Perturbed-Chain Statistical Associating Fluid Theory is applied to predict thermodynamic properties and phase equilibria. The three pure-component parameters of PC-SAFT model, i.e., segment number m, segment diameter σ, and segment energy parameter ε/k_B, and binary interaction parameter k_ij are regressed based on the literature data for propylene-hydrogen-polypropylene system. The accuracy of PC-SAFT state equation using the regressed parameters obtained this work outperforms those of Peng-Robinson equation, S-L equation and PC-SAFT state equation using the literature parameters.2. The mechanism of Ziegler-Natta catalyzed propylene polymerization isanalyzed in a thorough manner. Simplified elementary reactions involved in the propylene polymerization that are suitable for subsequent modeling of industrial processes are then proposed. They include the activation of co-catalyst and monomer, initiation and propagation of chains, transfer of radicals to monomer and hydrogen, and deactivation of catalyst. The reasonable ranges of the kinetic constants of these elementary reactions and their activation energies are also determined. According to the molecular weights and molecular weights distribution of two different polypropylene grades, the number of active sites for the catalyst is found to be six. A sensitivity analysis shows that the monomer conversion is most sensitive to the chain propagation and deactivation constants and that the molecular weight is most sensitive to the chain transfer and propagation constants. This provides guidance for choosing appropriate model parameters for process modeling.3. The agitated slurry propylene polymerization reactor is analyzed from the point of view of fluid mixing. It is found that the principle of reactor optimization is to keep polypropylene particles suspended, enhance the dispersion and mass transfer of hydrogen, intensify the heat transfer by the vaporization of liquid propylene and in the jacket, and improve the dispersion of circulating propylene gas in the reactor. The power consumption, gas holdup and heat transfer in the gas-liquid agitated reactor are investigated experimentally. The results can be used to design and optimize this type of agitated reactor.Gas holdup: Power consumption under gassing condition: Heat transfer under gassing condition: 4. The power consumption, fluidization process and pressure fluctuation in an agitated fluidization bed with a frame impeller or a dual anchor impeller, cooperating with orifice gas distributor or half-cone-shaped cap gas distributor, are investigated. The effects of the impeller, gas distributor, powder diameter and superficial gas velocity on the fluidization process are studied and correlations between the critical fluidization gas velocity and pressure drop of the bed are obtained.As the rotation speed of the impeller increases, the transition point of the pressure drop vs. superficial gas velocity curve becomes smoother. There is no obvious relationship between power consumption and powder diameter, bed heightand gas distributor in the fluidization regime. At the same time, the pressure drop of the bed is independent of the type and the rotation speed of the impeller. The impeller in the agitated fluidized bed prevents the powder from agglomeration and slugging, stabilizes the fluidization and improves the fluidization quality.5. It is shown that for an agitated fluidization bed reactor, an optimized impeller should possess a suitable number of horizontal blades to restrain and break down air bubbles and several vertical blades to scrape the wall and prevent air bubbles from short circuit. The fluidization quality of the dual anchor impeller excels that of the frame impeller, and that of the half-cone-shaped cap gas distributor is better than that of the orifice gas distributor. The optimization policy, i.e., decreasing the polydispersity of the powder diameters and increasing the bed height and gas velocity, is proposed. It allows to convert the production of powder type polypropylene to that of small-sphere type polypropylene in the Hypol gas fluidized bed and to intensify the time-space yield load.6. A steady model of the Hypol polymerization process based on single active site reaction kinetics is developed with the yield of polypropylene of each of the reactors as a benchmark. Meanwhile, steady and dynamic models of the process based on multiple active sites reaction kinetics are developed with the molecular weight and its distribution as the targets. They are validated using plant data. Thereafter, the effects of the operating conditions of the process such as temperature, slurry level in the reactor, hydrogen molar fraction in the gas phase and its inlet flow rate on the process state variables such as polymerization yield and molecular weight, are investigated. The dynamic characteristics of the propylene polymerization process under pulse disturbance, process step action and measurement noise have offered quantitative support to the process optimization and control. The dynamic model simulates well grade transition processes. In addition, it is shown for the first time that the mass transfer performance of hydrogen influences the dynamic characteristics of the propylene polymerization process.7. Finally the steady and dynamic models are used to simulate new processes based on the existing one. The new processes are "a slurry reactor and a fluidized bed reactor in series" and "a slurry reactor and two fluidized bed reactors in series". Optimized operating conditions are then proposed. Under these conditions, the new processes can markedly produce much higher polymer yields than the existing process. Simulated results show that the "a slurry reactor & a fluidized bed reactor in series" process is most suitable for producing polypropylene with broad molecular weight distributions.更多还原