Prediction of drop size in a pulsed and non-pulsed disc and doughnut solvent extraction column
Graphical abstract
Introduction
Liquid–liquid extraction is an important separation process. Various types of solvent extraction columns have been used for a range of applications in the chemical, petroleum, nuclear, hydrometallurgical industries and other areas for many years. Disc and doughnut solvent extraction columns (DDC) are attractive from both safety and economic stand points, in particular its simplicity of design, less space consumption, higher throughput and no internal moving parts (Movsowitz et al., 1997). In the column, pulsing may be introduced by compressed air (pulsed disc and doughnut column, PDD column) in order to enhance the efficiency of the column.
Knowledge of the hydrodynamic parameters including drop size is of fundamental importance in the design of liquid–liquid extraction columns. Drop size affects the dispersed phase holdup, the residence time of the dispersed phase, and the allowable throughputs. Furthermore, a combination of the Sauter mean diameter together with the dispersed phase holdup gives a measure of the interfacial area of contact between the phases, which is important for the prediction of mass transfer performance in the column. It is therefore important to understand the drop size distribution in the column under different operating conditions and for different systems and geometries. It is also important to be able to predict the drop size distribution as a function of the column geometry, operating conditions and solvent physical properties.
There have been many studies reporting the drop size in solvent extraction columns, including rotating disc (Mao and Slater, 1994), Kühni (Dongaonkar et al., 1991), Wirz-II (Rinconrubio et al., 1994), perforated-plate (Kumar and Hartland, 1986), Karr (Smith et al., 2008) and packed (Torab-Mostaedi et al., 2011b) columns. Jahya (2002) and Torab-Mostaedi et al. (2011a) studied the drop size in a PDD column at relatively high pulsation rates.
To date, most of these studies have been conducted at relatively high pulsation intensities as low pulsation intensity and non-pulsation were thought to be of little practical importance due to poor mass transfer performance. However recently a number of industrial scale PDD columns have been operating with no pulsation. As the correlations and models for drop size from the literature have been developed for pulsing conditions only it is important to investigate how the performance of these columns change when operated under non-pulsing conditions. It is also important to develop a new model that can predict the drop size over the full range of pulsation conditions seen in practice, including low pulsing intensity and non-pulsing conditions.
In this study, the drop size of the dispersed phase was measured over a wide range of operating conditions, including pulsing and non-pulsing conditions, using a 75 mm diameter disc and doughnut column. The experimental data was then compared to literature correlations. An empirical correlation was developed to predict the mean drop size based on the experimental data and the literature data for different liquid systems.
Section snippets
Previous studies
The equilibrium droplet size distribution in a solvent extraction column is usually represented as the Sauter mean diameter (d32) and is calculated from the following equation:
In the absence of pulsation or at low levels of pulsation, the breakup of drops is controlled by the ratio of buoyancy to interfacial tension forces. A limiting value of the drop size under such conditions may be predicted from (Chang-Kakoti et al., 1985):where the constant, C1, is a
Equipment
The pilot scale PDD column used in this study consists of a 1.0 m long QVF® precision bore glass column with an internal diameter of 72.5 mm, enclosing a stack of alternating stationary discs and doughnuts. A t-piece is located on top of the main column section to act as the organic phase outlet, and below the column an expanded glass hemispherical section provides an area for settling of the dispersed droplets before exiting from the bottom of the column. The fluid was pulsed by the motion of
Drop size distribution
Fig. 2 shows the typical droplet size distribution under different pulsation intensity. Broad and multimodal drop size distribution is observed at no pulsation condition (f = 0 Hz) which corresponds to the mixer-settler region. The selection of drop diameter range will affect the results of drop size distribution. In this work, the selected drop size ranges are 0.1 mm for f = 1.5 Hz and 0.25 mm for f = 0 Hz. With an increase of pulsation intensity, the drop size distribution tends to narrow and a single
Correlation of drop diameter, d32
In order to design industrial solvent extraction PDD columns, it is important to be able to predict the drop size via correlations that cover the range of operational parameters and physical properties relevant to the particular system. As described previously, there are no existing correlations for predicting d32 in a PDD column that can be used for both non-pulsing and pulsing conditions.
Based on the published research (Kumar and Hartland, 1996) and analysis above, the d32 is determined to be
Impurities of operation of pulsed disc and doughnut columns
This work clearly shows that as the pulsation energy input is decreased, for constant continuous and dispersed phase velocities, the drop size increases and the drop size distribution is broader. The inclusion of data and models to predict this is an important development and enables the prediction of an operating window for this type of column. Operations using these columns and trying to maximize throughput and recovery have found that they can do this by reducing pulsation but maintain
Conclusions
The conclusions from this study may be summarized as follows:
- (i)
The droplet size distribution in PDD column is broad and multimodal at low pulsation conditions (Mixer settler region) and it tends to narrow and become single modal with increasing pulsation intensity (Emulsion region).
- (ii)
The cumulative drop size distribution in PDD column can be predicted well using the Weibull function.
- (iii)
Under non-pulsing conditions, the d32 in a PDD column is not obviously affected by continuous phase velocity, and
Conflict of interest statement
The authors declare the following competing financial interest.
Funding
The authors would like to acknowledge the funding provide by the Australian Research Council through Linkage grant LP130100305 and BHP Billiton, Olympic Dam, for this project, and would also like to thank the Particulate Fluids Processing Centre for the resources provided for this project.
References (18)
- et al.
Flooding and drop size in a pulsed disc and doughnut extraction column
Chem. Eng. Res. Des.
(2011) - et al.
Drop sizes and distributions in rotating disc contactors used for liquid–liquid extraction
J. Sep. Process Technol.
(1985) - et al.
Mass transfer and axial dispersion in a Kühni extraction column
AIChE J.
(1991) - et al.
Statistical Distributions
(2000) - et al.
Statistical Distributions
(2011) Fundamentals of the hydrodynamic mechanism of splitting in dispersion processes
AIChE J.
(1955)Performance of the Pulsed Disc and Doghnut Solvent Extraction Column
(2002)- et al.
Gravity settling in liquid/liquid dispersions
Can. J. Chem. Eng.
(1985) - et al.
Prediction of drop size in pulsed perforated-plate extraction columns
Chem. Eng. Commun.
(1986)
Cited by (29)
A data-driven model of drop size prediction based on artificial neural networks using small-scale data sets
2024, Chinese Journal of Chemical EngineeringStudy on the dynamic interfacial tension with micro interface sensor
2023, Chemical Engineering ScienceCitation Excerpt :Microfluidic technique has also been applied for the DIFT determination. For example, some researchers (Steegmans et al., 2009; Wang et al., 2009; Wang et al., 2016b; Glawdel et al., 2012; Wang et al., 2015; Cui et al., 2020) have determined the transient interfacial tension at the droplet pinch-off moment by measuring the droplet size. In our previous studies (Lan et al., 2016; Lan et al., 2019), the DIFT throughout the whole dispersion process is determined by pressure measurements in a microfluidic device.
Performance prediction of disc and doughnut extraction columns using bayes optimization algorithm-based machine learning models
2023, Chemical Engineering and Processing - Process IntensificationEstimation of droplet size distribution by using maximum entropy programming and population balance equations in pulsed disc-doughnut column
2022, Chemical Engineering Research and DesignCitation Excerpt :It was observed that this column was performed under lower holdup and larger Sauter-mean drop size compared to the standard column (Li et al., 2018). Performance evaluation of this column in pulsed and non-pulsed conditions showed that the agitation intensity plays a vital role in achieving the desired level in the column (Wang et al., 2016). A chemical system containing nitric acid in the aqueous phase and a solvent of tributyl phosphate in the organic phase was used to investigate the behavior of the PDDC column.