基于多尺度CFD耦合PBM的甲醇制烯烃反应器模拟及放大效应研究

发布时间：2018-05-31 14:45

本文选题：MTO + 放大　；参考：《中国科学院大学(中国科学院过程工程研究所)》2017年硕士论文

【摘要】：新兴的甲醇制烯烃(MTO)工艺可实现甲醇到低碳烯烃的转化过程。因为甲醇可方便地从煤和天然气中获得,因此MTO工艺架起了煤化工和石油化工之间的桥梁,有望在不久的将来成为生成低碳烯烃的主要路线。目前,MTO工艺已实现商业化运行,而相关的化学反应工程的基础研究很少报道,特别是反应器的放大仍然依赖于经验和逐级放大实验,缺乏可靠的理论指导。传统的反应器放大准则主要关注放大过程中的流动参数的相似性,而忽略了流动和反应的双向耦合,因此无法有效指导实际放大过程。近年来,随着计算流体力学(CFD)技术和多相流理论的迅速发展,特别是介尺度理论的兴起,采用CFD模拟研究反应器的放大过程将有助于深刻理解反应器中流动和反应及其耦合行为随着反应器放大的变化规律,有望缩短传统的基于实验的反应器放大进程。然而,反应器的放大势必涉及操作流域的一系列改变(如从鼓泡流域到湍动流域的变化)以及随之带来的对停留时间、反应行为等的影响,这给CFD模拟带来新的挑战:1)反应器放大过程中涉及流域转变,不同流域存在不同的流动结构,如鼓泡床中的"气泡"和快速床中的"颗粒团聚物",在不同流域中选择合适的曳力本构关系是个非常关键的问题;2)受限于计算量,在CFD模拟中,一般采用集总模型描述反应动力学。而集总反应动力学模型通常基于小型流化床实验,过滤了外部流场变化带来的影响,它是否适用于操作在不同流域的大规模反应器的模拟需要深入探索和验证;3)工业MTO反应器尺寸大且颗粒停留时间长,Lu等[1]人提出的全混流(CSTR)模型用作CFD模拟的初值预测从而加速反应模拟的方式是否也适用于工业反应器的加速模拟,还需进一步研究。针对上述挑战,本文围绕MTO反应器的放大过程开展了一系列研究。论文第一章介绍MTO工艺的发展现状、工艺特点、各种模拟方法、以及已开展的模拟工作。在此基础上,引出本文的研究内容。论文第二章以大连化学物理研究所开发的MTO工艺(DMTO)的放大过程为切入点,开展不同尺度的DMTO反应器的放大模拟研究。其中,流动模型采用双流体模型(TFM),曳力模型根据不同反应器的操作状态(如鼓泡和湍动状态),分别采用了 EMMS/bubbling模型和最新开发的二步法模型,反应动力学采用了平行反应的七集总模型。在此基础上,分析了流动和反应行为随着反应器放大的变化情况,进一步考察了反应动力学模型对模拟结果的影响。研究表明,上述模拟方法可准确预测不同尺度反应器内的流场分布和甲醇转化率,但对主要产物(乙烯和丙烯)的预测,则随着反应器的放大而逐渐偏离实验。考虑产物转化反应的交叉动力学模型代替平行反应模型,依然无法改善对反应产物的预测。由于产物的选择性与催化剂的焦炭含量密切相关,而基于TFM的模拟无法区分同一固相中的颗粒差异,从而不足以准确预测催化剂颗粒的焦炭含量分布。为在模拟中考虑焦炭含量分布的影响,论文第三章提出采用群体平衡模型(PBM)来描述由焦炭含量确定的催化剂颗粒分布情况。先由全混流反应器模型估算颗粒停留时间分布,结合焦炭的生成速率方程,计算得到焦炭含量分布,作为TFM耦合PBM模拟的初始值,采用离散法求解PBM中的焦炭含量的分布密度函数。在此基础上,对DMTO示范反应器进行了二维模拟。研究表明,上述方法能够有效地预测焦炭含量分布,显著提高模拟对主要产物选择性的预测。论文最后对全文进行了总结,并对未来如何完善流化床反应器的放大模拟工作进行了展望。
[Abstract]:The new methanol to olefin (MTO) process can realize the conversion process of methanol to low carbon olefin. Because methanol can be easily obtained from coal and natural gas, the MTO process has erected a bridge between coal chemical and petrochemical industry. It is expected to become the main route of producing low carbon olefin in the near future. At present, the MTO process has been commercialized. The basic research of the related chemical reaction engineering is rarely reported. Especially, the amplification of the reactor is still dependent on the experience and the step by step amplification experiment and lack of reliable theoretical guidance. The traditional reactor amplification criterion mainly focuses on the similarity of the flow parameters in the amplification process, but neglects the two-way coupling of the flow and reaction, so there is no one. In recent years, with the rapid development of the computational fluid dynamics (CFD) and multiphase flow theory, especially the rise of the mesoscale theory, the CFD simulation of the amplification process of the reactor will help to understand the flow and reaction in the reactor and its coupling behavior with the amplification of the reactor in recent years. It is expected to shorten the traditional experiment based reactor amplification process. However, the amplification of the reactor is bound to involve a series of changes in the operation of the basin, such as the change from the bubbling basin to the turbulent River Basin, and the consequent effects on the residence time, reaction behavior, etc. this brings new challenges to the CFD model: 1) the amplification process of the reactor involves the amplification process of the reactor. There are different flow structures in different basins, such as "bubbles" in bubbling beds and "particle aggregation" in fast beds. Choosing appropriate drag constitutive relations in different basins is a key problem; 2) limited to the amount of calculation, in CFD simulation, the lumped model is generally used to describe the reaction dynamics. The model is usually based on the small fluidized bed experiment, filtering the influence of the change of the external flow field. Whether it is suitable for the simulation of large-scale reactor operating in different basins needs deep exploration and verification; 3) the size of the industrial MTO reactor and the long time of the particle residence, the Lu and other [1] people's total mixed flow (CSTR) model are used as the initial of the CFD simulation. For the above challenge, a series of studies have been carried out on the amplification process of the MTO reactor. The first chapter introduces the development of the MTO process, the process characteristics, the various simulation methods, and the simulation that has been carried out. On the basis of this, the second chapter in the second chapter takes the amplification process of the MTO process (DMTO) developed by Dalian Institute of Chemical Physics as the breakthrough point to carry out the magnification simulation study of the different scales of the DMTO reactor. Such as bubble and turbulent state, the EMMS/bubbling model and the newly developed two step model were used respectively, and the reaction kinetics adopted the seven lumped model of parallel reaction. On this basis, the effect of the flow and reaction behavior with the change of reactor magnification was analyzed, and the effect of the reaction kinetic model on the simulation results was further examined. The results show that the simulation method can accurately predict the flow field distribution and the conversion rate of methanol in different scale reactors, but the prediction of the main products (ethylene and propylene) is gradually deviated from the experiment with the amplification of the reactor. Because the selectivity of the product is closely related to the coke content of the catalyst, the TFM based simulation can not distinguish the particle difference in the same solid phase, so that the coke content distribution of the catalyst particles is not accurately predicted. In order to consider the influence of the coke content distribution in the simulation, the third chapter of the paper proposes the use of the group equilibrium model. Type (PBM) is used to describe the distribution of catalyst particles determined by coke content. The distribution of particle residence time is estimated by the model of total mixed flow reactor, and the distribution of coke content is calculated with the formation rate equation of coke. As the initial value of TFM coupling PBM simulation, the distribution density function of coke content in PBM is solved by dispersion method. On this basis, the two dimensional simulation of the DMTO model reactor has been carried out. The study shows that the above method can effectively predict the distribution of coke content and significantly improve the prediction of the selectivity of the main products. Finally, the paper is summarized and the future improvement of the amplification and Simulation of the fluidized bed reactor is prospected.
【学位授予单位】：中国科学院大学(中国科学院过程工程研究所)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ221.21;TQ018

【参考文献】