Ph.D Thesis

Ph.D StudentSimakov David Stas
SubjectAutothermal Packed-Bed Catalytic Membrane Reformer for
Hydrogen Generation by Steam Reforming of Methane
DepartmentDepartment of Chemical Engineering
Supervisor PROFESSOR EMERITUS Moshe Sheintuch
Full Thesis textFull thesis text - English Version


A novel concept of a membrane autothermal packed-bed reformer for hydrogen production by the endothermic steam reforming of methane coupled with the exothermic methane combustion, for power generation in polymer electrolyte membrane (PEM) fuel cells, has been proposed and studied theoretically and experimentally. The steam reforming equilibrium is shifted toward formation of hydrogen by using Pd-Ag alloy-based hydrogen separation membranes, providing also extra-pure hydrogen that can be directly fueled to PEM fuel cells. The heat required for the endothermic steam reforming is provided by the catalytic combustion of methane over an adjacent packed-bed, using air as an oxygen source. The reformer is designed to be operated independently on any external source of heat.

Simulations of the adiabatic system exhibited dynamic thermal fronts and mapped the acceptable domain of operation bounded by stationary fronts, separating domains of upstream and downstream-moving fronts. It has been found that the front velocity and the operation domain boundary strongly depend on the exothermic-to-endothermic flow rates ratio and on the hydrogen separation rate. An analytical approximation for the thermal front velocity has been developed.

The proof-of-concept has been experimentally provided in a lab-scale (10-50 W) non-adiabatic unit, focusing first on the determination of the key operation parameters. It has been shown that the reactor can be independently operated at steady state with the enthalpy required for the steam reforming and for heat losses compensation provided by methane oxidation. The reactor performance was mainly defined by the dimensionless feed-to-membrane flow rates ratio and by the exothermic-to-endothermic flow rates ratio. Further experimental study focused on the experimental demonstration of two approaches for the optimization of hydrogen generation in terms of power output and process efficiency: increasing membrane flux and recycling reformer products to the combustion section.

Next, a comprehensive non-adiabatic model with no adjustable parameters has been defined, validated versus experimental data by numerical simulations, and then extended for the hydrogen generation optimization. A model-based analytical algorithm for the thermodynamic optimization in terms of the reformer efficiency and fuel cell power output has been developed, defining the reformer operation window for a required fuel cell power output. The optimized methane reformer is expected to be suitable for fueling kW-range PEM fuel cell stacks, providing power density of ~1 kW/L and efficiency of up to 0.8, which is expected to provide an overall natural gas-to-electric power conversion efficiency of ~0.5 in a combined reformer-PEM fuel cell unit.