(137b) Autothermal Reforming of Ethanol in a Fluidized Bed Membrane Reactor for Ultra-Pure Hydrogen Production

ABSTRACT In recent years, many efforts have been made to develop processes and reactor technologies for the production of cheap, ultra-pure hydrogen that can be used in efficient PEM fuel cells. Nowadays, hydrogen is mainly produced by steam reforming of natural gas in multi-tubular reactors. The main drawback of natural gas reforming is that this reaction leads to a H2 rich gas mixture also containing carbon oxides and other by-products. Consequently, in order to produce pure H2, chemical processes are carried out in a number of reaction units (typically high temperature reformer, high and low temperature shift reactors) followed by separation units (mostly pressure swing adsorption units are used). An attractive way to produce hydrogen is the reforming of renewable fuels (e.g. bio-ethanol) inside membrane reactors, where hydrogen production and hydrogen separation through selective membranes are integrated in one apparatus. In this paper the production of ultra-pure hydrogen via autothermal reforming of ethanol in a fluidized bed membrane reactor has been studied. The steam reforming of ethanol is an endothermic reaction requiring a great amount of heat. The needed energy for the steam reforming is obtained by burning part of the hydrogen recovered via the hydrogen perm-selective membrane. In this configuration, the air used for the combustion is never in contact with the reacting mixture, which makes thus autothermal reforming with integrated CO2 capture feasible. Simulation results based on a phenomenological model show that it is possible to obtain overall autothermal reforming of ethanol while 100% of hydrogen can in principle be recovered at relatively high temperatures and at high reaction pressures. At the same operating conditions, ethanol is completely converted, while the methane produced by the reaction is completely reformed to CO, CO2 and H2.

Keywords: Membrane Fluidized Bed, Ethanol Reforming, Autothermal Reforming, Hydrogen, Membrane Reactor.

* Corresponding author

FLUIDIZED BED MEMBRANE REACTOR MODEL In our fluidized bed model ethanol and steam enter the reactor where the catalyst is mildly fluidized. The hydrogen produced is directly recovered as an ultra-pure stream via perm-selective dead-end Pd-based membranes. Part of the hydrogen is recovered in different (U-shaped) Pd-based membranes where it reacts with oxygen (fed as air) producing heat. The heat is transferred to the bed, supporting the reforming reaction. A typical phenomenological two-phase model for a membrane assisted fluidised bed reactor is considered here, where the number of CSTRs in the cascade and the sizes of the CSTRs are directly related to the extent of gas back-mixing in each phase. The overall (bubble and emulsion phase) component mass conservation equations have been formulated, accounting for chemical transformations in the emulsion phase and a net gas production due to the chemical reactions and gas withdrawal (hydrogen) via membranes. The overall energy balances are also accounting for possible energy exchange via the sweep gas. The exact number of CSTR in both emulsion phase and bubble phase need to be determined by comparing the model results with experimental results. However, no experimental data on fluidized bed reforming of ethanol are available at the moment, so the number of CSTR in both phases were kept equal to the number of CSTRs determined for the methane reforming (Nb =5 and Ne =35) [1]. Empirical correlations for the mass transfer coefficients and fluidization properties of the fluidized beds have considered. The kinetic rate expressions for ethanol steam reforming have been taken from Mas et al. [2]. Selective removal of H2 using Pd-based membranes has been modelled with a lumped type flux expression, using experimental data from [3]. Pure component physical data have been taken from Daubert and Danner [4], while mixture properties have been computed following Reid et al. [5]. A direct validation of the fluidized bed membrane reactor for ethanol reforming is not possible at the moment since no experimental results are available for this reactor type. However, the model can simulate a typical plug flow reactor, if the number of CSTRs considered is sufficiently high and permeation through membranes is ignored. By simulating a reactor with 200 CSTRs and no membranes, the results in terms of ethanol reforming agree fairly well the experimental results presented by Mas et al. [2].

RESULTS AND CONCLUSIONS The theoretical feasibility of ultra-pure hydrogen production via autothermal reforming of ethanol in a fluidized bed membrane reactor has been demonstrated through a theoretical investigation. The simulation results with a phenomenological two-phase model showed that pure hydrogen can be recovered through Pd-based membranes, realizing complete ethanol conversion while the heat needed for the reforming reaction can be easily supplied by burning approximately 15% of the hydrogen recovered, obtaining overall autothermal reforming. Mixing between air and the reaction mixture is circumvented so that the CO2 separation after the reactor is easier, since the exhaust gas of the reactor consists of a mixture of CO, CO2 and water. The simulations also show that a critical point to be experimentally studied in the fluidized bed membrane reactor is the extent of bubble-to-emulsion phase mass transfer limitations. In the worst case (limitations calculated as in a fluidized bed without membranes) the mass transfer limitations inside the reactor result in an increase of the membrane area of 88% with respect the case in which no mass transfer limitations prevail.

REFERENCES [1] Gallucci F., van Sint Annaland M., Kuipers J.A.M., Autothermal Steam Reforming of Methane in a novel Fluidized Membrane Reactor. Part 2 Comparison of reactor configuration, Topics in Catalysis, 2008; 51(1-4), 133-45 [2] Mas V., Bergamini M.L., Baronetti G., Amadeo N., Laborde M., A kinetic study of ethanol steam reforming using a nickel based catalyst, Topics in Catalysis, 2008;51(1-4):39-48 [3] Gallucci F., van Sint Annaland M., Kuipers J.A.M., Autothermal Steam Reforming of Methane in a novel Fluidized Membrane Reactor. Part 1 Experimental Demonstration, Topics in Catalysis, 2008; 51(1-4), 133-45 [4] Daubert T.E., Danner R.P., Physical and Thermodynamic Properties of Pure Chemicals; Core Edition, Taylor & Francis: London (1998) [5] Reid R.C., Prausnitz J.M., Poling B.E., The Properties of Gases and Liquids. McGraw-Hill Book Company: New York (1988)