(540f) Study of Mass Transport Regimes in Membrane Steam Reformers

The success of membrane reactors for hydrogen separation depends significantly on the development of membranes having high permeability and selectivity with respect to hydrogen. Pd-based membranes have been proposed and their performance is being continuously improved. In conditions of high membrane flux, external mass transfer resistances, such as those due to the transport of hydrogen from the bulk of the catalytic bed to the membrane surface, may limit the performance of the reactor. In the present work a two-dimensional model has been developed and the influence of different operating conditions on the performance of the reactor has been investigated. An isothermal reactor has been considered as a first step for a more complete analysis. Methane steam reforming has been considered as a case study, but the conclusions reached may be extended to other reaction systems. The velocity profile has been evaluated through Darcy's law, in which gas velocity is correlated to pressure drops and bed permeability. Both axial and radial velocity components have been accounted for. Mass transport equations for all species present have been developed by considering consumption/production through the steam reforming reaction, and transport by convection and dispersion in the radial and axial directions. Variations of gas density due to pressure gradients and composition changes have been accounted for. The membrane has been considered to have infinite selectivity towards hydrogen and Sievert's law has been used to describe hydrogen flux through the membrane. Dimensionless parameters which govern the system's behavior have been identified and, according to the values of these parameters, simplified models have been proposed. The equations of the general model have been solved using the finite element solver COMSOL Multiphysics. Macroscopic quantities, such as hydrogen production and yield, as well as concentration and mas flux profiles within the reactor have been evaluated. The presence of an optimal operating pressure has been identified. The existence of a boundary layer, close to the membrane, in which the reaction occurs has been noticed and may serve as a basis for the optimization of the reactor's design.