Heat and Mass Transfer in Co-Current Gas-Liquid Packed Beds
Analysis, recommendations and new correlations

Faïçal Larachi, Lamia Belfares, Ion Iliuta and Bernard P.A. Grandjean
 Department of Chemical Engineering and CERPIC
Université Laval, Ste-Foy, Québec, Canada, G1K 7P4
Corresponding author: flarachi@gch.ulaval.ca

I&EC Res., 42, 222-242 (2003)

Abstract :
Meticulous inspection of the literature has unveiled the weakness of several empirical methods for predicting the macroscopic mass and heat transfer characteristics relevant to gas-liquid co-current downflow and upflow packed-bed reactors. In response, using a wide experimental database consisting of 5,279 measurements for trickle beds (down-flow), and 1,974 measurements for packed bubble columns (up-flow), a set of reliable correlations has been recommended for the prediction of the gas-liquid interfacial area (agL), the volumetric liquid- (kLa) and gas-side (kga) mass transfer coefficients, the wall  and bed liquid-solid mass transfer coefficients, the wall heat transfer coefficient (hw), the bed effective radial thermal conductivity , and the particle-to-fluid heat transfer coefficient (hp). Some of these correlations are from the literature and others have been developed by combining artificial neural networks and dimensional analysis. The accuracy of the proposed correlations surpasses by far the performances of the available methods sometimes by up to a ten-fold reduction in scatter. Notwithstanding the substantial reduction in scatter, these correlations have been thoroughly tested for phenomenological consistency and have been shown to restore the expected trends documented in the database.

Keywords fixed bed, up-flow, down-flow, mass and heat transfer parameters, correlation

ERRATUM   


You can get the tbr-pbc-2.zip  file that contains an Excel worksheet simulator to compute heat and mass transfer properties in TBR's and packed bubble columns.


You may also download our  Excel worksheets simulators for  Trickle-bed or Flooded Bed reactors.

The neural correlation was developped with the software NNFit