Modeling and Demonstration of a Sub-Watt Scale Methanol Reformer

Jamelyn D. Holladay

A micro-scale methanol fuel processor was designed, modeled, built and tested for the production of hydrogen for a small fuel cell (<1W). Methanol was selected due to the relatively low reforming temperature. The fuel processor included the methanol-water vaporizer, preheater, catalytic methanol steam reformer, catalytic combustor, a vaporizer/preheater for the combustion fuel and oxidant, and in one embodiment a CO selective methanation reactor. The fuel processor had a novel radial flow design with the methanol reformer and catalytic combustor separated by a thin shim. The inlet flow of the steam reformer section and the combustor sections were impinging. The fuel processor, including all unit operations, had a volume less than 0.5 cm^3. While most methanol fuel processors in the literature used a pyrophoric Cu/ZnO catalyst, we used a new Pd/ZnO/Al2O3 catalyst which is non-pyrophoric. A Pd/Al2O3 based catalyst was also used for the catalytic combustor. In the CO methanation reactor, a Ru/Al2O3 catalyst was used. A rate equation for the Pd/ZnO/Al2O3 catalyst was developed as: (-r)(mmol / kgcat/ s) = 2.9047*10^10 exp(-94800/R/T)*P(meoh)^0.715*P (H2O)^0.088. Three generations of the fuel processor were built and tested. The first reformer had the smallest reformer, but was able to produce up to approximately 1.1 sccm of reformate with 73-74% hydrogen and a thermal efficiency of up to 9 wt%. The second generation expanded the methanol reformer reactor and was able to increase the production to 4.9 sccm reformate (72-73% hydrogen) and the thermal efficiency increased to ~33%. The third variation included the selective methanation reactor. The methanation reactor reduced the CO content from approximately 1% to less than 100ppm. The higher CO content reformate stream was suitable for higher temperature (~150°C) polybenzimidazole fuel cell. While the low CO content reformate stream may be suitable for PEM type fuel cells. A three dimensional model of the reactor was developed to explore design variations to improve the reactor performance by modifying the internal reactor gas flow. Two variations were examined, the first included multiple outlets rather than the single outlet in the baseline designs on the methanol steam reformer side. The second explored the use of fins to increase performance. The multiple outlet version did not improve performance. In the finned reactor version, three fins were included in the methanol steam reformer section. This increased the gas residence time in the hottest portion of the reactor. This version allowed a temperature reduction of almost 30°C at the equivalent inlet flow rate to achieve >99.5% conversion. In addition, the inlet flow rate could be increased by 2.7x at approximately equivalent temperatures. Finally, the model clearly showed the reactor was heat transfer limited. Therefore a new design which had the combustor exhaust envelope the steam reformer. This would not only better insulate the methanol reformer, but the exhaust gases could heat the sides of the reactor.

Modeling and Demonstration of a Sub-Watt Scale Methanol Reformer

Files and links (1)

Abstract

Metrics

Details