Effective approximations for concentration-polarization in Pd-membrane separators

Olga Nekhamkina*, Moshe Sheintuch

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

25 Scopus citations


Approximate models are proposed to simulate the transmembrane hydrogen flux in an empty membrane separator. The hydrodynamic field is constructed under the assumption of constant density with a single unknown parameter - the normal wall velocity (. vw). The 2-D concentration profiles are derived using a known velocity distribution. The problem is closed by definition of vw via the transmembrane flux, which admits Sievert's law. Such an approach allows to derive two approximate models that are governed by the set of ODE equations with respect to the average variables coupled with algebraic relations to describe the radial profiles. The first model accounts for analytical concentration profiles ci(. z,. r), the second one presents model reduction using the mass transfer coefficient (. kc) expressed as Sherwood (Sh) number. An analytical expression for local Sh is derived (Sh. =. 6 in a tube).We identified a parameter Γ which represents the ratio of diffusive to permeating flux and suggest that for Γ>. 6 the concentration polarization effect can be neglected. We address two axisymmetric geometries (i) a tube with transport at its wall, as is the case in a membrane of an integrated reactor; (ii) an annular cylinder with transport at an inner tube, as is the case in a separator. The proposed approximations are validated by comparison with experimental and CFD simulation data.

Original languageEnglish
Pages (from-to)835-845
Number of pages11
JournalChemical Engineering Journal
StatePublished - 5 Jan 2015
Externally publishedYes


  • Concentration polarization
  • Mass-transfer
  • Membrane reactor
  • Numerical simulations
  • Pd-membrane
  • Sherwood number


Dive into the research topics of 'Effective approximations for concentration-polarization in Pd-membrane separators'. Together they form a unique fingerprint.

Cite this