Transient numerical modeling and model predictive control of an industrial-scale steam methane reforming reactor

Abstract

A steam methane reforming reactor is a key equipment in hydrogen production, and numerical analysis and process control can provide a critical insight into its reforming mechanisms and flexible operation in real engineering applications. The present paper firstly studies the transport phenomena in an industrial-scale steam methane reforming reactor by transient numerical simulations. Wall effect and local non thermal equilibrium is considered in the simulations. A temperature profile of the tube outer wall is given by user defined functions integrated into the ANSYS FLUENT software. Dynamic simulations show that the species distribution is closely related to the temperature distribution which makes the temperature of the reactor tube wall an important factor for the hydrogen production of the reformer and the thermal conductivity of the catalyst network is crucial in the heat transfer in the reactor. Besides, there exists a delay of the reformer's hydrogen production when the temperature profile of the tube wall changes. Among inlet temperature, inlet mass flow rate and inlet steam-to-carbon (S/C) ratio, the mass flow rate is the most influencing factor for the hydrogen production. The dynamic matrix control (DMC) scheme is subsequently designed to manipulate the mole fraction of hydrogen of the outlet to the target value by setting the temperature profile trajectory of the reforming tube with time. The proportional-integral control strategy is also studied for comparison. The closed-loop simulation results show that the proposed DMC control strategy can reduce the overshoot and have a small change of the input variable. In addition, the disturbances of feed disturbance can also be well rejected to assure the tracking performance, indicating the superiority of the DMC controller. All the results give insight to the theoretical analysis and controller design of a steam methane reformer and demonstrate the potential of the CFD modeling in study the transport mechanism and the idea of combining CFD modeling with controller design for the real application.