Application Of Dense Oxygen-permeable Membrane Reactor In Oxidative Steam Reforming Of Ethanol To Produce Hydrogen

One of the challenges for the practical application of ethanol steam reforming (ESR) is the poor stability of the catalyst due to coking. Ni-based catalyst is promising for ESR reaction due to its high catalystic ativity and low cost. However, the carbon deposition on the Ni-based catalyst needs to be further suppressed. The ethanol oxidative steam reforming (EOSR) can be operated at an autothermal condition. Furthermore, the high potential of reaction of oxygen with the carbon formed on the catalyst surface may facilitate carbon elimination.A reactor with a dense oxygen-permeable membrane can feature a gradual supply of oxygen to the reactor through the permeation of oxygen across a dense perovskite layer which is a mixed conductor. In this work, a perovskite-type Ba0.5Sr0.5Co0.8Fe0.2O3-δ(BSCF) dense membrane reactor (DMR) was applied to the EOSR reaction.In chapter 2, a thermodynamic analysis of EOSR was carried out with a Gibbs free energy minimization method. The addition of oxygen lowers the enthalpy of the system and favors the heat recycle. The heat released from the exothermic reactions makes up exactly for the requirement of the endothermic reactions. Moreover, the addition of oxygen can eliminate the carbon deposition. Then, EOSR in a fixed-bed reactor (FBR) and thermodynamic analysis were compared. The result shows that in the temperature range 750-900 oC and oxygen flow rate range in 0.684-1.267 ml/min, with H2O/EtOH=1 and feed rate (H2O & EtOH) at 40 ml/min, the H2 and CO yields increases with the increase of the temperature and the oxygen flow. The formation of graphitic carbon is thermodynamically unfavorable for all these conditions. However, in the experiment, filament carbon with a rough surface was formed and the Ni particles were encapsulated by the deposited carbon.In chapter 3, the performance of a Ni/α-Al2O3 catalyst both in DMR and FBR was compared for hydrogen production in EOSR reaction at high temperature. The catalyst in the FBR showed a perfect selectivity of H2 and CO while that in the DMR was lower due to oxygen permeation. The catalyst in the DMR showed a stable performance for 188 h at 850 oC, while that in the FBR the catalyst was stable only for 72 h. Eventually, the stability of the perovskite BSCF membrane materials was measured under the reaction conditions, the results exhibits that the BSCF membrane material sustains longer than the stable time of the catalyst used. Nevertheless, the surface of the membrane exposed to the reaction atmosphere was partially decomposed and formed carbonates.In chapter 4, the catalytic membrane reactors (CMRs) were prepared by combinating a porous catalyst with the dense oxygen-permrable membrane. The permeability of the CMRs was investigated in detail. It was found that the oxygen permeation was dependent on the thickness of dense membrane and controlled by the bulk diffusion when the temperature was higher 750 oC. The CMRs were applied for EOSR. It was found that the limiting step for oxygen permeation was shifted from the bulk diffusion controlling step to the surface exchange controlling step. After the steady reaction for 20 hours, it was found that there was no carbon deposition on the CMRs. The result was attributed to oxygen permeation and lattice oxygen (Oo*) of CMRs.In chapter 5, the hydrogen production via EOSR in a tubular DMR was sequentially simulated with the aide of an ASPEN PLUS software. The tubular DMR was divided into multi-sub-reactors, and the Gibbs free energy minimization sub-model in ASPEN PLUS was employed to simulate the EOSR process in the sub-reactors. A FORTRAN sub-routine was integrated into Aspen Plus to simulate the oxygen permeation through membranes in the sub-separators. The simulation results indicate that there is an optimal length of the tubular DMR at the operating temperature and steam-to-ethanol ratio, under which hydrogen and carbon monoxide formation reach their maxima.