Producing Hydrogen by Variable Volume Membrane Batch Reactors with Modulated Liquid Fuel Introduction

A novel dynamically-controlled membrane batch reactor, which integrates the variable volume operation of CO₂/H₂ Active Membrane Piston (CHAMP) with the direct injection of liquid fuel of Direct Droplet Impingement Reactor (DDIR) for improved power density, is described and experimentally characterized in this article.

Experiment

A laboratory-scale CHAMP-DDIR, comprising of an actively-controlled micro injector and a variable volume piston-cylinder reactor chamber for liquid fuel atomization, is used with a Pd-Ag foil membrane and Cu/ZnO/Al₂O₃ catalyst to steam reform methanol for hydrogen generation as illustrated in Figures 1 and 2.

Figure 1. Schematic of the experimental CHAMP-DDIR reactor and piping/instrumentation diagram.

Figure 2. Photograph of the experimental test bed used in CHAMP-DDIR performance characterization.

Two modes of CHAMP-DDIR operation, batch reaction with dynamically-adjusted reactor volume and pulse-modulated fuel injection, were examined. Their performance was quantified using metrics such as volumetric power density, and hydrogen yield measured using a mass spectrometer (Hiden Analytical Quadrupole HPR-20 QIC), and compared with those for a standard operation (single fuel injection with fixed reactor volume). The experimental results revealed that the residence time for the same hydrogen yield can be decreased by compressing the reactor volume during the conversion cycle. The reduction in residence time was mainly due to higher hydrogen partial pressure in the reactor chamber and therefore higher rates of hydrogen permeation.

Furthermore, pulse-modulated fuel injection experiments showed that a major reduction in required reactor volume can be accomplished with multi-shot split fuel introduction, as shown in Figure 3 (a) and 2 (b). Both the reduction in required reactor volume and the reduction in required cycle time increase the volumetric power density of CHAMP-DDIR. An important improvement in volumetric power density is seen as a result of integrating time-modulated fuel introduction with reduced reactor volume. The confirmed power density enhancement realized via the dynamic compression of reactor volume was 17%, and the improvement accomplished via time-modulated fuel introduction was 38% (for 85% hydrogen yield efficiency under the restrictions of the same maximum operating pressure and total amount of fuel), as illustrated in Figure 3 (c).

Figure 3. Results for time-modulated fuel injection experiments with constant, but different reactor volumes (6 cm³ and 4.5 cm³) under constrained pressure: (a) H₂ permeation rate, (b) H₂ yield efficiency, and (c) volumetric power density for 4 shot injection (4.5 cm³) and single injection (6 cm³) fixed volume operations for the reactors operated under a constraint of the maximum peak pressure. Initially, the hydrogen generation rate is higher with the single shot/large volume operation as compared to the multi-shot/small volume operation. However, because the smaller volume reactor maintains a higher pressure over most of the cycle, total H₂ yield becomes similar around 70% yield efficiency. For 85% hydrogen yield efficiency, volumetric power densities were improved by 38% by splitting fuel injection into 4 shots and using a 25% smaller volume.

Conclusion

The improved power density and potential for hydrogen throughput control shows the value of CHAMP-DDIR for applications such as transportation, where the ability to meet time varying power demands and high power density are crucial.

Project Summary by:

Andrei G. Fedorov
George W. Woodruff School of Mechanical Engineering
Georgia Institute of Technology
Atlanta, GA 30332-0405
USA

Paper Reference: Yun, T. M., Kottke, P. A., Anderson, D. A., and Fedorov, A. G. (2015) “Experimental investigation of hydrogen production by variable volume membrane batch reactors with modulated liquid fuel introduction” International Journal of Hydrogen Energy, 40, (6) 2601-2612

This information has been sourced, reviewed and adapted from materials provided by Hiden Analytical.

For more information on this source, please visit Hiden Analytical.