Synthesis Of Two-Dimensional Reactor Network Based On Transfer Fundamentals

A reaction system, which is the core of process in typical refining and petrochemical industry, often depends critically the economic and environmental performance of process. And it provides essential conditions for design and control on separation and heat exchange processes.The essence of reactor network synthesis is to find optimal reactor types, sizes, controls and interconnections which will optimize a specific objective function. Three significant strategies for reactor network synthesis are superstructure optimization, attainable region and targeting. These researches are based on one dimensional ideal reactor models, which are different from practical reactors in plant, also the objective function is minimizing reactor volume or maximizing yield. There are still many questions need to be answered, for example, length, diameter, reaction distribution in a reactor and investment of a reactor system. What’s more, there are still some necessary constraints need to be considered for non-isothermal reactor network synthesis, such as heat transfer rate equation.In this paper, the two-dimensional reactor network model is accomplished through two steps, in order to establish a practical reactor network which is approximate to the industrial production of tubular reactor network:(1) Unit simulation of a tubular reactor:firstly, analyzed reaction system characteristics through simulating the one dimensional pseudo-homogeneous tubular reactor, then optimizing a singular variable control problem about temperature profile; Secondly, a two-dimensional reactor model is established by considering heat transfer, mass transfer, fluid flow and the reaction kinetics, in which model partial differential equations are used describing temperature, concentration distribution on axial and radial directions.(2) Synthesis of tubular reactor network:a state space framework composed by series two-dimensional reactors and distribution network is constructed for the synthesis of reactor networks. Considering the heat transfer rate equations in the non isothermal reactor system, as well as the yield of demands, the total annual cost (TAC) as an objective function is optimized and evaluated on economic performance. A mathematical model is established according to the super structure mentioned above, which constitutes a mixed-integer dynamic optimization (MIDO) problem with partial differential constraints. Through the decomposition strategy as well as the discretization algorithm, the MIDO problem is simplified into a nonlinear programming (NLP) problem. Simulation results include concentration distributions and nominal operating conditions are used as initial values for improving the efficiency of targeting. Consequently, the optimal solution is solved in the AMPL and GAMS environment.Finally, two industrial production cases, which are epoxidation of propylene and alkylation of toluene, are illustrated with the approach proposed above. And optimal reactor networks are obtained simultaneously as well as optimal decision variables, which demonstrate industrial feasibility and efficiency of the proposed approach.