十七集总催化重整反应器的稳态模拟

Steady state simulation of a catalytic reforming reactor using a 17-lumped kinetics

文中按照集总理论的指导原则 ,从催化重整反应机理出发 ,提出了包含 17个集总组份的催化重整反应网络 ,建立了 17集总催化重整反应器的稳态数学模型。该反应网络将重整物料按分子大小集总为C6、C7、C8和C+ 9,每一个碳数的化合物又划分为正构烷烃、异构烷烃、环烷烃和芳香烃 4个集总 ,裂化产物C-5作为一个集总。集总组份与组份之间主要发生烷烃脱氢环化、环烷烃脱氢芳构化、烷烃异构化和加氢裂化反应 ,简化了一些发生几率小或对过程影响小的反应。在该 17集总催化重整反应网络的基础上 ,针对一半再生式催化重整反应装置 ,建立了 17集总的轴向和径向重整反应器的数学模型 ,并以工业装置实际生产操作数据 ,采用区域约束法和Marquardt法相结合的数值方法 ,分步优化估计了模型动力学参数。模拟结果表明 ,该模型能够预测计算装置生产工况 ,得到碳六至碳九的芳烃产率分布和产品烃组成 ,并预测重整汽油辛烷值和反应器温降 ,计算结果的偏差分别小于 1 5 (RON)和5℃。因此 ,所建立模型正确地反映了反应的特性 ,对于炼油厂铂重整装置的模拟效果良好。该模型对于重整装置的实时动态模拟也有一定的指导意义。

According to the guiding principles of the lumping theory and based on the study of the reaction mechanisms, a new complex reaction network with seventeen lumped components was proposed for catalytic reforming. In the network the reaction species were firstly lumped into C 6, C 7, C 8 and C + 9 according to the carbon number of hydrocarbon compounds, each carbon number components were lumped into normal-paraffin, iso-paraffin, naphthene and aromatics respectively, and cracking products with less carbon number totally were taken as another lumped component C - 5. Four main type reactions among these lumped components were considered in the reaction network, i. e., paraffin dehydrogenation and cyclization, naphthene dehydrogenation and aromatization, paraffin isomerization and cracking reactions to form C - 5 lighter products. For the purpose of simplification, some reactions seldom taken place and reactions of less importance were eliminated from the network. Furthermore, based on the formerly proposed seventeen lumps reaction network, the estimated results of parameters and industrial process simulation were reported. The mathematic models of axial and redial reformer reactors were established for a semi-regenerated commercial catalytic reformer. The kinetic parameters were estimated and optimized stepwise by a combined procedure of the restrained range and Marquardt's methods with a set of the plant operation data. And this model can predict C 6 to C 9 aromatic yield distribution and hydrocarbon composition of reformatted gasoline under varied operating conditions, and calculate the product gasoline octane and temperature profiles in the reactors with the maximum deviations between calculated results and practical value less than 1.5 RON and 5?℃ respectively. Therefore, it is concluded that the model can describes the reactor correctly and help to guide the dynamic simulation for this process.