Numerical studies of the influences of bypass on hydrogen separation in a multichannel Pd membrane system

Highlights

A multichannel Pd membrane system with flow bypass is designed.

H2 recovery can be improved by flow bypass significantly and up to 219%.

H2 recovery by the first membrane is independent of the bypass ratio.

A contour map and a correlation from regression analysis are established.

This study provides a practical insight into the flexible operation for H2 separation.

Abstract

A multichannel palladium (Pd) membrane system in association with flow bypass is designed for hydrogen separation with high recovery, and the mass transfer phenomena in the system are simulated by developing a computational fluid dynamics (CFD) model. Two Pd membranes are installed in the system. The predictions suggest that the H2 recovery (HR) can be substantially improved by the bypass. The higher the feed gas Reynolds number, the more pronounced the improvement of H2 recovery by the bypass. The HR by the first membrane is independent of the bypass ratio (BR), revealing that the enhancement of HR is completely contributed by the second membrane. An increase in H2/CO2 molar ratio in the feed gas reduces HR, but raises the H2 permeation rate. The maximum HR by the second membrane always develops at the feed gas Reynolds number (Rer,M1) of 500, regardless of bypass ratio. This reveals that the aforementioned Reynolds number is an appropriate condition for H2 separation in the designed membrane system. Based on the HR in the absence of flow bypass (i.e., BR = 0), the higher the Rer,M1, the larger the intensification of H2 permeation. A contour map and a correlation from regression analysis in terms of Rer,M1 and BR are established. Under a desired H2 recovery, the combination of Rer,M1 and BR can be suggested to provide flexible operation for H2 separation in the membrane system.

Graphical abstract

Keywords

  • Palladium (Pd) membrane;
  • Multichannel membrane system;
  • Bypass mechanism;
  • Hydrogen separation and recovery;
  • Regression analysis;
  • Numerical simulation

Be the first to comment

Leave a Reply

Your email address will not be published.


*