Frictional Pressure Drop in Three-phase Fixed-bed Columns with Upward Flow of Gas and Liquid: Critical Revision of Experimental Data

Abstract

The technological relevance of packed-bed columns with simultaneous flow of gas and liquid is stressed in the literature, e. g. [ 1 ]. Packed-bed columns that process gas and liquid streams can operate in upward co-current flow, in downward co-current flow or in counter-current flow [ 1 ]. One major drawback of the co-current upward flow is that it develops a much higher pressure drop than that corresponding to the other two modes of operation [ 1 ]. Hence, reliable tools for predicting pressure drop are a must to size the column and to estimate the cost of impulsion. However, the disparities among the equations published for the prediction of pressure drop are remarkable, e. g. [ 2 ]. The observed disparity is derived from the experimental data that led to the formulation of those equations. A conclusive comparison of correlations is not possible without testing the data itself. However, information about experimental errors is almost absent in the revised literature. Reports of reproducibility levels, of test-runs performed under experimental conditions explored by other authors, of experimental variances or of relative error are lacking. To address this issue, experimental data on frictional pressure drop measured in packed columns with upward co-current of gas and liquid, published in twenty different sources, were collected and analyzed. The purpose was to build a databank devoid of untrendy data. Following the main experimental trends and with the aid of diagnostic plots along with “ high-order polynomials ” [ 3 ], data subsets from eleven sources were retained to ensemble a database comprising two hundred and thirty seven experimental observations. The selected experimental observations, together their respective references, are attached to be downloaded. The files contain all the experimental information necessary for the development of tools for predicting frictional pressure drop. Certain experimental works were not considered because they do not inform whether the data correspond to the frictional contribution or the total pressure drop. Most of the rejected data correspond to the lower values of the pressure drop. In most of the rejected data, the influence of the frictional pressure drop over the total pressure drop is significantly attenuated by the fluid static head. A brief comparison with the largest data bank [ 2 ] is pertinent. A comprehensive study of experimental information from various sources should involve data handling. However, Larachi et al. [ 2 ] do not report any data analysis. As mention above, the experimental database attached to this summary was thoroughly analyzed for identifying inconsistent data. In addition, the data bank available in this link contains data with packings not involved in [ 2 ] as Intallox saddles and Berl saddles. Columns packed with Intallox saddles are characterized by truly high bed porosities of about nine tenths. Therefore, Intallox saddles allowed to collect data that evidence frictional pressure drop values much lower than the lowest ones in [ 2 ].

[ 1 ] “ Multiphase Catalytic Reactors: A Perspective on Current Knowledge and Future Trends ”; M. P. Duduković, F. Larachi and P. Mills. Catalysis Reviews, volume 44, number 1, pages 123 – 246 ( 2002 ). [ 2 ] “ Two-phase Frictional Pressure Drop in Flooded Bed Reactors: A State-of-the-art Correlation ”; F. Larachi, Z. Bensetiti, B. P. A. Grandjean and G. Wild. Chemical Engineering & Technology, volume 21, number 11, pages 887 – 893 ( 1988 ). [ 3 ] “ External Liquid Saturation in Fixed Beds with Co-current upward Flow of Gas and Liquid: Revisiting Experimental Data and Data Representation ”; S. P. Bressa. Trends in Chemical Engineering, volume 18, pages 29 – 45 ( 2020 ).