Pulsed solvent extraction column performance: study and comparison of standard disc and doughnut internals and Tenova kinetics internals
AffiliationChemical and Biomedical Engineering
Document TypePhD thesis
Access StatusOpen Access
© 2018 Dr. Wen (April) Li
Pulsed columns were used as high efficiency solvent extraction contactors for a range of applications including uranium recovery (Olympic Dam Operations, Australia) and for cobalt and nickel extraction (Goro Nickel, New Caledonia). Traditionally, disc and doughnut-shaped internals were used in these systems and their operating performance had been studied extensively and found effective for processes with normal and fast kinetics. For processes with slow kinetics and high feed phase ratios, the new Tenova kinetics internals which included novel contacting elements on the contours of the “disc and doughnut” plates and optimized spacing between the plates may provide enhanced performance. The developers of the new internals expected these internals to result in less back-mixing, higher column holdup and improved mass transfer at the same flux during operations. However independent experimental data and pilot testing was required to verify these improvements. This research aimed to study the effect of operating characteristics (flowrates, pulse frequency and pulse amplitude) on dispersed phase holdup, flood point, drop size distribution, mass transfer and axial dispersion performance for Tenova kinetics internals and quantify the potential improvements over the traditional standard disc and doughnut internals using two different liquid-liquid systems with slow and fast kinetics. It was observed that the solvent extraction column efficiency, especially for the relatively slow kinetics systems, had been optimized to receive higher extraction efficiency and hereby making it more cost-effective. Dispersed phase holdup and drop size, as the parameters to estimate the hydrodynamic performance, were measured in terms of pulsation intensity, dispersed and continuous phase velocities. Holdup decreased with increasing pulsation intensity to a minimum value at (Af)min and then increased with higher pulsation intensity for both types of internals and both liquid-liquid systems. And holdup increased with increasing dispersed phase velocity but neglected change with the continuous phase velocity. Dispersed phase droplets drop size decreased as pulsation intensity increases up to 0.3 m/s for both standard disc and doughnut internals and Tenova kinetics internals, and no noticeable effect of dispersed and continuous phase velocities was observed. At low Af, the Sauter mean diameter for Tenova kinetics internals was higher than the standard disc and doughnut internals. The correlations were refitted with the pilot scale experimental data in this study for both holdup, Sauter mean diameter and the drop size volume distribution, and the overall absolute average relative errors for holdup and Sauter mean diameter were 16% and 11%, respectively. Axial dispersion performance of the pulsed column with two types of internals was estimated in terms of pulsation intensity, dispersed and continuous phase velocities for the concentration jump of the continuous phase. The axial dispersion coefficients for both internals decreased with increasing continuous phase velocity, and little impact of dispersed phase velocity on the axial dispersion coefficients was found for both liquid-liquid systems and both types of internals. The axial dispersion coefficients with Tenova kinetics internals were lower compared to the standard disc and doughnut internals using water-Alamine 336 system and high pulsation intensity condition for water-LIX 84 system. Correlations were selected and refitted to the 100 experimental axial dispersion data points and resulted in the overall absolute average relative errors of 54% and 41%. Mass transfer experiments were conducted with two liquid-liquid systems with different kinetics for both types of column internals under the effects of pulsation intensity, dispersed and continuous phase velocities. The height of a mass transfer unit Hoc decreased with increasing pulsation intensity and increased with increasing dispersed and continuous phase velocities for CuSO4-LIX 84, relatively slow kinetics system. For H2SO4-Alamine 336, fast kinetics system, no change of Hoc was observed due to the system reaching the equilibrium state inside the 2 m high solvent extraction column. The height of a mass transfer unit for CuSO4-LIX 84 system was much larger than H2SO4-Alamine 336 system. The correlation of Hoc was regressed to the experimental data with the absolute average relative error of 16%, and the overall mass transfer coefficient of continuous phase could be predicted with correlated values of Hoc, xd and d32. And an approximate estimated value of reaction rate for CuSO4-LIX 84 system was reported.
- Click on "Export Reference in RIS Format" and choose "open with... Endnote".
- Click on "Export Reference in RIS Format". Login to Refworks, go to References => Import References