(586h) Accounting for Anisotropy of the Packing to Improve CFD Simulation of Liquid Maldistribution in Structured Packing in an Inclined Column and Validation

Accounting for anisotropy of the packing to improve CFD simulation of liquid maldistribution in structured packing in an inclined column and validation

P. Béard, M. Fourati, P. Alix, IFP Energies nouvelles, Solaize, France

T. Maubert, C. Weiss, TOTAL, X. Courtial, G. Perdu, Prosernat, France

As natural gas always contains contaminants such as acid gases, in particular CO2 and H2S, acid gas removal is required to meet LNG (Liquefied Natural Gas) specifications. This process is performed in packed liquid-gas contacting columns operating under counter- current liquid-gas flow. In offshore conditions, the performance may be altered because of ship movement resulting in liquid maldistribution within packed beds.
In order to improve the design of such columns, this work investigates the liquid distribution in a pilot-scale column filled with a structured packing for different operating conditions. The column tilt angle ( ), diameter (D) and packed bed height (H) as well as the liquid mass flow rate (QL) were varied.
Unsteady 3D CFD (Computational Fluid Dynamics) simulations were carried out using ANSYS® FLUENT® with a two-phase liquid-gas Eulerian approach. The structured packing was modeled as a porous medium wherein gas-liquid momentum exchanges and liquid dispersion were taken into account. The previously developed CFD model dedicated to the studied packing [1] allowed to reproduce properly the sensitivity to the operating conditions [2]. Nevertheless, the liquid maldistribution was overestimated when the height of the column is increased. Thus, the model was improved to account for the packing geometry by blocking the flow in the direction perpendicular to the walls (Figure 1) leading to an anisotropic dispersion of the liquid phase. This led to a modified distribution of the liquid in the column (Figure 2) compared to the results obtained with the isotropic model.
Experimental measurements were carried out in a test rig installed on a hexapod moving platform allowing ship motion simulation (Figure 3) at Heriot Watt (HW) University. These data are used to validate the liquid maldistribution predicted by the numerical simulations.
Comparisons exhibit an enhanced agreement (Figure 4) with the more realistic description of the packing geometry developed in this study compared to previous results [2]. The predictivity of the new anisotropic CFD model remains very good for high values of the column tilt angle and the liquid mass flow rate (Figure 5) or if the column diameter is increased (Figure 6).

References

[1] Fourati et al., "Liquid dispersion in packed columns: Experiments and numerical modeling", Chem. Eng. Sci. 100, 2013
[2] Béard et al., "Liquid maldistribution in structured packing in an inclined column: CFD
simulation and validation”, 9th Int. Conf. on Multiphase Flow, 2016

Authors

Philippe Béard, IFP Energies nouvelles, philippe.beard@ifpen.fr
Manel Fourati, IFP Energies nouvelles, manel.fourati@ifpen.fr
Pascal Alix, IFP Energies nouvelles, pascal.alix@ifpen.fr
Thomas Maubert, TOTAL, thomas.maubert@total.com
Claire Weiss, TOTAL, claire.weiss@total.com
Xavier Courtial, Prosernat, xcourtial@prosernat.com
Gauthier Perdu, Prosernat, gperdu@prosernat.com

Figures :

Figure 1 : Principle of anisotropic flow within the packing.

-

a) Isotropic dispersion	b) Anisotropic dispersion

Figure 2 : Comparison of the liquid volume fraction distributions between the isotropic and the anisotropic CFD models for D = 0.6 m, H = 1.2 m, = 3°, Q L.

Figure 3 : Experimental test rig allowing ship motion simulation.

Figure 4 : Comparison between the isotropic and the anisotropic CFD models and validation for D = 0.6 m, H = 1.2 m, = 3°, Q L

Figure 5 : Validation of the anisotropic CFD model for D = 0.6 m, H = 2.4 m, = 5°, 2 Q L

Figure 6 : Validation of the anisotropic CFD model for D = 1 m, H = 1 m, = 3°, Q L