On-site pure hydrogen production by methane steam reforming in high flux membrane reactor: Experimental validation, model predictions and membrane inhibition

Michael Patrascu*, Moshe Sheintuch

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

87 Scopus citations

Abstract

Measurements in a membrane methane reformer packed with catalytic (Pt(3)Ni(10)/CeO2) foams and equipped with a 175cm2 Pd membrane showed that high conversion and high hydrogen recovery can be achieved with sweep flow and high pressures (over 90% conversion and over 80% H2 recovery at 525°C for 0.25NL/min CH4 feed flow). Increasing pressure above 10bar does not lead to higher fluxes due to stronger permeance inhibition. A mathematical model predicts the reactor's performance well in terms of axial temperature profile, exit compositions and permeate flow, when membrane permeance is calibrated with experimental results. However this value is significantly lower (by ~80%) than values measured in pure H2 in the absence of reaction. This apparent permeance inhibition is attributed to coadsorbates, although the only strong inhibitor in these conditions is thought to be CO, and its concentration (<1%) cannot justify this strong inhibition. We suggest a mechanistic explanation for this. Concentration polarization effects are not negligible but a 1-D model solution, using an approximation of this effect, shows it cannot account for this large inhibition.

Original languageEnglish
Pages (from-to)862-874
Number of pages13
JournalChemical Engineering Journal
Volume262
DOIs
StatePublished - 5 Feb 2015
Externally publishedYes

Keywords

  • Hydrogen production
  • Membrane reactor
  • Methane steam reforming
  • Modeling
  • Pd membrane
  • Permeance inhibition

Fingerprint

Dive into the research topics of 'On-site pure hydrogen production by methane steam reforming in high flux membrane reactor: Experimental validation, model predictions and membrane inhibition'. Together they form a unique fingerprint.

Cite this