Fluidised bed membrane reactor for ultrapure hydrogen production via methane steam reforming: Experimental demonstration and model validation

Abstract

Hydrogen is emerging as a future alternative for mobile and stationary energy carriers in addition to its use in chemical and petrochemical applications. A novel multifunctional reactor concept has been developed for the production of ultrapure hydrogen ( < 10 ppm CO ) from light hydrocarbons such as methane for online use in downstream polymer electrolyte membrane fuel cells. A high degree of process intensification can be achieved by integrating perm-selective hydrogen membranes for selective hydrogen removal to shift the methane steam reforming and water–gas-shift equilibriums in the favourable direction and perm-selective oxygen membranes for selective oxygen addition to supply the required reaction energy via partial oxidation of part of the methane feed and enable pure CO 2 capture without costly post-treatment. This can be achieved in a proposed novel multifunctional bi-membrane bi-section fluidised bed reactor [Patil, C.S., van Sint Annaland, M., Kuipers, J.A.M., 2005. Design of a novel autothermal membrane assisted fluidized bed reactor for the production of ultrapure hydrogen from methane. Industrial and Engineering Chemistry Research 44, 9502–9512]. In this paper, an experimental proof of principle for the steam reforming/water–gas-shift section of the proposed novel fluidised bed membrane reactor is presented. A fluidised bed membrane reactor for steam reforming of methane/water–gas-shift on a commercial noble metal-based catalyst has been designed and constructed using 10 H 2 perm-selective Pd membranes for a fuel cell power output in the range of 50–100 W. It has been experimentally demonstrated that by the insertion of the membranes in the fluidised bed, the thermodynamic equilibrium constraints can indeed be overcome, i.e., increased CH 4 conversion, decreased CO selectivity and higher product yield (H 2 produced/CH 4 reacted). Experiments at different superficial gas velocities and also at different temperatures and pressures (carried out in the regime without kinetic limitations) revealed enhanced reactor performance at higher temperatures ( 650 ∘ C ) and pressures (3–4 bar). With a phenomenological two-phase reactor model for the fluidised bed membrane reactor, incorporating a separately developed lumped flux expression for the H 2 permeation rate through the used Pd-based membranes, the measured data from the fluidised bed membrane reactor could be well described, provided that axial gas back-mixing in the membrane-assisted fluidised bed reactor is negligible. This indicates that the membrane reactor behaviour approached that of an ideal isothermal plug flow reactor with maximum H 2 permeation.