Hydrodynamic Study of a Hybrid Electro-Flotation Column

Abstract

Bubble columns are used in the mining industry for mineral recovery but are also widely utilized in the chemical and petrochemical industry. The hydrodynamic characteristics of their performance is a field of interest with a number of points, which are nonetheless poorly understood, and a considerable amount of methods have aimed to shed light on the flow regimes that prevail in the columns. The study of the hydrodynamic part of a flotation process should consider characteristics such as air flow, volumetric gas fraction, flow field, and bubble size, along with the mechanical and design factors and pulp properties. The present work aims to elucidate the characteristics of the gas phase of a hybrid flotation system. For this purpose, a hybrid flotation column was designed and constructed and the bubbles size distributions at different radial positions in the flotation column were computed by analyzing high resolution digital images. A patented electrical impedance technique was employed to instantaneously measure the local volumetric gas fraction. Flow dispersion in the column was studied by residence time distributions using conductivity tracers. The experimental results are discussed to comprehend the variation in the gas fraction in the column. In particular, the study showed that the size of the bubbles changed from the center to the walls of the column, and this was observed both radically and vertically. Moreover, the size of the bubbles affected the volume fractions, and no coalescence of the bubbles was observed. Finally, the dispersion of the tracer in the working solution was distributed uniformly in the volume of the column, with a time difference for the four positions of the column.

1. Introduction

The most important parameters that affect a flotation process are the slurry chemistry and the hydrodynamics of the phenomenon [1]. The performance of a flotation column is affected by the characteristics of the flow field [2] and, moreover, importance is also paid to the “micro-hydrodynamics” observed due to the lack of homogeneity with the appearance of the local turbulent flow, which is responsible for the bubble−particle interactions and flotation kinetics [3].

Studies have shown that the presence of the gas phase changes the local characteristics of the flow [4,5,6], while suspended fine solid particles can prevent the generation of turbulent flow [5,7]. Therefore, the study of the hydrodynamic part of a flotation process should consider characteristics such as air flow, volumetric gas fraction, and bubble size, along with the mechanical and design factors and pulp properties.

Air flow (or superficial gas velocity) is an extremely important parameter in the flotation process [8]. In mechanical flotation devices, the apparent velocity is recommended to not exceed 3 cm/s, because at higher values, the efficiency of the process decreases and entrainment is observed [9]; meanwhile, values smaller than 1 cm/s reduce the kinetics of flotation. In general, air flow affects the air dispersion, bubble size, and volumetric gas fraction [10]. Higher air flow rate values result in a lower particle residence time in the froth zone, preventing bubble coalescence [11]. This leads to an increase in yield, as the coalescence of bubbles is one of the reasons for the detachment of the particles from the bubbles. In addition, increasing the air flow also increases the surface area of the bubbles, enhancing the probability of particle to bubble collision [12]. At large values, a reduction in the efficiency of flotation is caused, provoking entrainment. Finally, on the other hand, low air flow leads to the opposite results and, more specifically, to the rupture of the bubbles in the foam phase and thus to a reduction in recovery [12]. Values of 0.01–0.03 m/s are considered optimal gas velocity values for industrial scale columns. Increasing the air volume fraction up to a point can improve the flotation kinetics; however, at high values (>30%) it leads to a decrease in efficiency [13]. Furthermore, increasing the apparent velocity leads to entrainment of the solid particles, reducing the selectivity of the process [14].

Most early studies of the flow in the flotation columns assumed that laminar flow dominated. A significant amount of older research stated that the flow velocity in the center of a column was upward, whereas near the walls it faced downward [15]. Furthermore, Xia et al. established that the gas–liquid flow in the column could cause turbulent pulsation [16]. Recently, the prediction and evaluation of these parameters have been managed utilizing CFD simulations. This approach allows for the simulation of various experimental conditions in different flotation devices, such as the Denver [17], Jameson cell [18], or flotation column [19]. Flow fields have been studied throughout the years and developments in experimental fluid dynamics (EFDs) and computational fluid dynamics (CFDs) have been managed. Particle image velocimetry (PIV) is an advanced EFD-based velocity measurement method that was developed in the late 1970s. In general, the flotation process is accompanied by turbulent flow, independent of the flotation devices. In a turbulent flow field, mineral particles interrelate with the bubbles instantly after the injection of the gas phase in the system. Several models have been reported that describe the turbulence in a flotation system, such as k-ε, k-ω, Reynolds stress model, and large eddies simulations. Additionally, the simulation of the multiphase flow can be modeled with either the Eulerian–Langrangian or the Eulerian–Eulerian approach. The former considers that the fluid is a continuous medium and the bubbles and particles are discrete phases, whereas the latter considers continuous and dispersed phases as continuous media that can penetrate each other. It is important to consider that precise liquid–phase hydrodynamic properties can be achieved from the numerical simulation of the single-phase flow. Yet, the requirements for the prediction of the flotation efficiency cannot be assembled by the single-phase flow simulation.

The size of the bubbles is crucial for the successful recovery of mineral particles with regards to the technique of flotation. For instance, microbubbles have a lower relative velocity, which offers a number of advantages for the flotation of fine particles. For a constant volume of air in the cell, as the bubble size decreases, the number of bubbles increases noticeably and leads to an increase in collision probability. Furthermore, decreasing the bubble size increases the gas hold-up and increases the bubble concentration, which leads to an increase in the probability of bubble−particle collision. To that end, the bubble size and shape has been studied extensively over the past years and frameworks for a number of techniques have been developed. The most common methods to measure conventional-sized bubbles are optical and photographic techniques [20], image analysis [21,22], and electrical impedance [23,24]. There are a large body of experimental data that confirm that the use micro or nano bubbles enhance the recovery of fine mineral particles. To that end, recently, a number of techniques have been developed that can be utilized for measuring micro and nano bubble size distributions. Among these methods are laser pulse methods [25], nanoparticle tracking analysis [26,27], laser-diffraction-based technologies and dynamic light scattering (DLS) [28] X-ray techniques [29], and indirect measurements through dissolved oxygen reverse estimation [30].

Several techniques have been developed that aim to measure the volumetric gas fraction in a two-phase flow. Among them are capacitance sensors [31], magnetic resonance [32,33], ultrasonic methods [34,35], radiation attenuation techniques [36,37], and wire mesh sensors [38,39]. The void fraction in a two-phase flow can also be measured by applying acoustic emission technology for measuring acoustic emission signals. The most widely used method is the optical method, as it is non-intrusive and accurate. The measurements are made using high-speed cameras and tomography algorithms. In previous studies [40,41], it was found that the laser transmittance varies with the area of the gas-liquid surface when an infrared laser beam passes through the gas−liquid two-phase flow. Information about the gas void fraction of the two-phase flow can be derived from the laser transmittance. Another method is the electrical capacitance tomography technique, which images the two-phase flow in a pressurized pipeline and is used for industrial applications [42]. In addition, electrical impedance tomography is a method that was introduced in past decades it results in accurate estimations of the local gas phase dispersed in a liquid medium, taking into account the local axial, radial, and angular components of the dispersed phase [43].

Bubble columns are widely used in the chemical, mining, and petrochemical industry and the hydrodynamic characteristics of their performance is a field of interest with a number of aspects that are still poorly understood. Recently, a considerable amount of work has been devoted to developing methods that aim to understand the flow regimes that prevail in columns. Among these methods is the radioactive particle technique (RPT), which is used to obtain flow measurements [44]. Two non-intrusive methods for two-phase flow study are the laser doppler velocimetry (LDV) [37,38] and the particle image velocimetry (PIV) [45,46,47,48]. Experimental data derived from these methods show that the flow field in a system with bubble evolution could be transformed from laminar to turbulent.

The scope of the work is the construction of a hybrid flotation column that generates in situ microbubbles, by employing electrolysis of water in order to study and characterize the hydrodynamics of the combined microflotation. To this end, the structure of this work is as follows: first, the bubble size distributions are determined by capturing photos of both conventional and electrolytic bubbles, and their size is calculated using custom-made image analysis software. Then, the turbidity of the flotation medium is estimated by employing a scattered light method. The gas holdup is determined by utilizing an electrical technique that estimates the volumetric gas fraction in the current flotation system. Finally, an NaCl solution serves as a conductivity tracer for the estimation of the residence time distributions in an effort to study the mixing conditions inside the flotation column.

2. Materials and Methods

2.1. Experimental Set Up—The Hybrid Flotation Column

The laboratory device used for mineral flotation in the present study was a laboratory scale custom constructed column (Figure 1). The flotation column consisted of three plexiglas cylindrical sections, connected with two flanges, and it was completely sealed. Plexiglas is a preferential material used for laboratory-scale flotation columns mainly due to its stability and, moreover, it can also be used for three-phase flow observation.

The column’s height was 60 cm, its diameter was 7 cm, and the width of the walls was 3 mm. At the top of the flotation column, around the outer surface, a concentric cylindrical overflow weir was mounted where the recovered particles were collected. The column could be divided into two zones (Figure 1b): the collection zone, where particles collided with the air bubbles and the froth zone, where the recovered particles were collected.

Two boron-doped diamond (BDD) electrodes served as an electrolysis unit, as described in a previous study [49]. The unit was supported on the inner wall of the flotation column, 6 cm above the column’s bottom (Figure 2), so that the produced microbubbles could be dispersed homogeneously in the column’s volume. The electrodes were placed horizontally and parallel and connected to an external power supply. There are many studies indicating that combining conventional-sized bubbles with microbubbles leads to enhancing the fine particle recovery [50,51,52]. The hybrid column was capable of producing bubbles with an average bubble diameter less than 40 and over 400 µm, enhancing the flotation of fine mineral particles.

Given the fact that the gas phase of a flotation system is crucial for the successful attachment of mineral particles onto bubbles, experiments for measuring the size and number of the bubbles involved in the flotation slurry were conducted. The gas fraction involved in the flotation process was characterized using both optical and electrical methods. The optical methods were employed to determine the bubble size distributions and turbidity of the flotation medium (correlation of bubbles size and number), whereas the electrical diagnostic techniques were involved in the determination of the volumetric fraction of the gas bubbles.

Materials and Reagents

The frother used in this research was the anionic surfactant sodium oleate (NaOl, ≥82% fatty acids, Riedel-de Haen, Seelze, Germany). The pH was adjusted using 0.1 M NaOH/HCl (Pancreac); moreover, pine oil was employed as the frother to improve the stability of the froth [53]. Sodium chloride (NaCl; VWR Chemicals, Radnor, PA, USA) was utilized as a background electrolyte, and throughout the flotation experiments dionized water (~10 μS/cm) was used. The experiments were conducted on the aforementioned hybrid electroflotation column.

2.2. Diagnostic Techniques

2.2.1. Optical Measurements—Bubble Size Distributions

The size of the bubbles was determined by capturing the bubble images using a high resolution digital camera (a 20 MP Canon EOS 70, Tokyo, Japan) equipped with macro lenses and extension tubes for efficient image magnification (Figure 3). A custom-made image analysis software (BubbleSEdit software) was used to automatically detect the contour of the bubbles and to measure their size and consequently obtain the corresponding bubble size distributions [54].

It is worth mentioning that in order to calculate the size of the bubbles from the photos taken, it is necessary to convert the pixels to μm. The calibration of the optical measurements was achieved by capturing a standard 56 μm thick wire with the camera settings used per the experimental conditions. Moreover, the possible alteration of the dimensions of x,y inside the flotation column in the presence of water was taken into consideration.

2.2.2. Intensity of Bubbles-Induced Scattered Light Method

An optical system (Intensity of bubbles-Induced Scattered Light method) was able to estimate the turbidity of the flotation medium by measuring the intensity of the light that passed from one side to the opposite side of the flotation cell (non-intrusive measurements) [55]. The change in light intensity due to scattering was related to the formation of bubbles in the two-phase system. The factors determining this change were the number and size of the dispersed air bubbles. The optical measurement system consisted of a cold light source placed on one side of the flotation device, while a photodetector was placed on the other side of the device (S121C, Thorlabs Inc., Newton, NJ, USA), which was connected to a computer via USB and power meter software (PM101 Photodiode and Thermal Power Sensor Interface with USB, RS232 (Thorlabs GmbH (Lübeck), Lübeck, Germany)) (Figure 4). For the measurement, a frequency generator was used and the current was adjusted and remained constant, so that any change in the intensity of the light recorded by the sensor was only due to its diffusion.

2.2.3. Electrical Technique—Gas Fraction Measurements

The electrical technique was employed to determine the volumetric gas fraction in the current flotation system. The gas holdup parameter affected the flotation performance and defined the bubble-flow density, which was mostly related to the flotation kinetics [56]. The estimation of the gas holdup at different locations on the flotation column could be established with a unique patented method called I-VED, which is an electrical spectroscopy impedance technique with multiplexer technology (Figure 5) [57].

The technique measured the ohmic component of electrical impedance, which is equivalent to the electrical resistance or vice versa of the electrical conductivity of the medium. The principle of the present technique is based on the fact that the formation of a gas phase (non-conductive) in a liquid volume (conductive phase), increases the electrical resistance (or equivalently reduces the electrical impedance) of the medium. The change in electrical resistance was affected by the quality characteristics of the bubbles, such as their size and number. The current technique permitted the evolution of local volumetric gas fraction during the flotation process. This expanded the single-point information by showing, in real time, the dispersion of the gas phase along the column height. Electrical impedance of the studied flotation medium occurred by employing metallic electrodes that were placed in pairs along the column’s height (Ch1–4, Figure 5i(a)). The distance between the electrodes of each pair was 32 mm (Figure 5i(a)) and corresponded to D/2 of the column (where D is the inner diameter of the column). According to previous experimental data [58,59], an electrode distance equal to D/2 ensured the radial homogeneity of the impedance measurements, while maintaining the local behavior of the measurement for the corresponding column height. The impedance measured was the equivalent intervening between the opposed electrodes. The electrical impedance data obtained during measurement were transformed into gas fraction time series using Maxwell’s model. Measurements were conducted at a frequency of 25 kHz in order to electrically excite the flow inside the column, generating an input current passing through the medium. Experimental data indicated that an excitation frequency of 25 kHz provided almost pure resistive behavior.

The present diagnostic technique took advantage of the fact that the electrical impedance of the medium used in the flotation runs decreased in the presence of air (gas phase). Therefore, the difference in the measured impedance of the flotation medium, before and after the introduction of air, could be transformed to the gas phase evolution in time and height.

Τhe specialized electronic equipment used to perform measurements with the I-VED technique consisted of the following devices:

• Variable frequency signal generator (20 MHz, 33220A, Agilent, Santa Clara, CA, USA)
• Data acquisition card (NI USB-6363, X Series Multifunction DAQ, National Instruments, Austin, TX, USA, 96 kHz/16 bit).
• Digital multimeter 51/2-digit (34405A, Agilent)
• Laptop (ASUSPRO, 4 GB RAM) with Sigview 5ch I-VED FFT v4.1 software.

2.2.4. Flow Field Investigation

An NaCl solution was utilized as a conductivity tracer for the estimation of the residence time distributions in an effort to study the mixing conditions inside the flotation column when a liquid jet was discharged. In particular, in this method, a high conductivity solution was injected into a solution with a lower conductivity and at the same time the electrical resistance of the mixture was measured locally at four different heights of the column, while dispersed air (0.7 L/min) was also inserted. Essentially, the time evolution of the electrical resistance developed described the flow field along the height of the column. Certainly, the presence of air bubbles strongly affected the flow field. The electrical resistance was measured using the electrical impedance technique I-VED using multiplexing, the principle of operation and measurement processing. With the use of multiplexing, the time evolution of the dispersion of the injection current in the liquid of the device in the four measurement channels of the gas phase fraction was simultaneously studied. In each experiment, the column contained an NaCl aqueous solution with a low specific conductivity (150–300 μS/cm) and 3 mL of an NaCl aqueous solution of higher specific conductivity (90 mS/cm) was injected into it. The solution was injected close to the bottom and centrally so that the dispersion was as homogeneous as possible. The sampling frequency of the measurements in the four channels was 25 kHz [60].

3. Results and Discussion

3.1. Bubble Size Distribution

Figure 6 shows illustrative photographs of dispersed air bubbles in the flotation column in the presence of sodium oleate as the frother. The experiments were conducted at a concentration of 120 mg/L, which was the optimal amount during the flotation experiments realized using the same apparatus, when sodium oleate was used as the collector for magnesite particles [49]. The photographs were captured at a steady state for the four different heights of the column (y = 17, 26, 36.2, and 46.5 cm) and at four radial distances (z = 1, 5, 10, and 15 mm) from the inner wall of the column.

Capturing pictures at different radial distances aimed to study the dispersion of the bubbles in the cross-section of the two-phase flow inside the flotation column. Visual observation of the indicative photos showed that the dispersion of the bubbles was not radially homogeneous, as the density of the bubble population appeared to be lower near the walls compared with the number of bubbles at the center of the column. Figure 7 presents the corresponding bubble size distributions for all positions and radial distances and, more specifically, shows the average diameter (dB,av) and the standard deviation (StDev (dB,av)) for each distribution. It is worth mentioning that for each distribution, at least 500 bubbles were processed in order for the average calculated size to be statistically correct.

The results showed that at the lowest height of y₁ = 17 cm and at a distance of 1 mm from the inner wall, the average size of the bubbles was 148 μm (±54), while at a radial distance of 15 mm, it was 246 μm (±128). Regarding the position y₂ = 26 cm near the wall of the column, the average bubble size was 191 μm (±69), while closer to the center (radial distance 15 mm) it was 230 μm (±91). The trend of increasing in size while moving radially towards the center of the column followed from the data derived from the pictures captured in position y₃ = 36.2 cm: close to the inner wall, the average size of the bubbles was 176 μm (±62), while at a radial distance of 15 mm, it was 210 µm (±90). In addition, the bubble distributions of the highest position (y₄ = 46.5 cm) appeared to follow the same trend. Near the wall, the average size was 148 µm (±54), while closer to the center it was 246 µm (±128). Generally, it was observed that in all four vertical positions of the column, the largest bubbles were located close to the center of the column, while the smaller bubbles were closer to the walls. This was attributed to the fact that the coarser bubbles rose rapidly in the center of the column, driving the smaller ones through recirculation to settle near the walls. The bubbles in the center line drew the liquid upwards. This created an imbalance and thus caused a kind of recirculation, and the liquid closer to the center flowed upwards while that closer to the wall flowed downwards. If the velocity of the downflow was greater than the relative velocity u_b of the bubbles, the bubbles appeared to move downwards [61]. Finally, no aggregation of the bubbles was observed along the height of the column, as the size of the bubbles did not seem to increase significantly at the highest locations. When the bubbles “met”, they bounced off each other without agglomerating because of the presence of the adsorbed frother on their surface [62].

Figure 8 outlines the effect of the frother concentration on the average diameter of the dispersed air bubbles for the four heights, and in particular for the radial position z₄ = 15 mm, which was judged to be the most representative position of the two-phase flow. The experimental data depict that the average size of the bubbles decreased when the frother concentration increased and did not change significantly in height [63]. It was generally observed that the same trend was followed for the bubble size, both vertically and radially, for the different positions in the column.

The presence of the surfactant led the solution surface tension decrease, resulting in a noticeable reduction in the bubbles size compared with the bubbles in the absence of a frother.

Indicative photographs of the bubbles produced with electrolysis of water in the flotation column in the presence of NaCl electrolyte (0.1 M) are presented in Figure 9. The photos were captured at the four aforementioned heights and radial distances of the column at steady state to study the dispersion of the electrolytic bubbles in the two-phase flow.

Figure 10 shows the bubble size distributions. The observation of the photos led to the assumption that the distribution of electrolytic bubbles was homogeneous, both vertically and radially.

In summary, Figure 11 indicates the results of the effect of the concentration of the sodium oleate frother on the average size of the electrolytic air bubbles in the flotation column for the four positions along the height of the column for the radial position z₄ = 15 mm, which was considered the most representative of the flow. The average bubble size decreased slightly with increasing the frother concentration and did not change appreciably at the different vertical positions.

3.2. Volumetric Gas Fraction

Under the framework of studying the hydrodynamic characteristics of the hybrid electroflotation column, the effect of the frother concentration on the volumetric gas fraction of the dispersed air was examined. Figure 12 presents the time series of the gas fraction for the four different heights of the flotation column (y = 17, 26, 36.2, and 46.5 cm) in the presence of sodium oleate (120 mg/L). In the first minute, the recorded fraction was zero and did not change, as no air was introduced to the column. In the sixth second, a sharp increase in the volume fraction was observed, while the two-phase system reached a steady state 50 s after the introduction of air. Air entrainment took place for 5 min, and the mean value of the volumetric gas fraction from the measurements recorded in the steady state region of each time series for locations y₁ = 17 cm, y₂ = 26 cm, y₃ = 36.2 cm, and y₄ = 46.5 cm was 5.4%, 5.6%, 5.8%, and 8%, respectively. The signals obtained in the presence of the highest frother concentration differed regarding the frequency of the peaks, compared with the signals with a lower oleate concentration.

The time that each height reached steady state varied by 1 s and the positions near the porous sparger achieved a steady state quicker. Because of buoyancy, the bubbles moved with a relative velocity to the liquid and thus had their own dynamics. Because of this, the gas phase fraction reached a steady state faster than the liquid flow.

Figure 13 shows the effect of the frother concentration (0, 60, 120, and 240 mg/L) on the variations in the electrical signal of the volumetric gas fraction during 10 s of injection (steady state). The peak signals in the presence of 60 mg/L sodium oleate depicted a higher intensity and this was attributed to the larger sized individual bubbles, which increased at a higher speed. Furthermore, when the frother concentration was 120 mg/L, the volumetric fraction was higher as the smaller bubbles that formed had a longer residence time in the column. Combining the data derived from the bubble size distributions at the same concentrations with the electrical signal measurements to calculate the volume fraction of the gas phase, it was concluded that the presence of smaller bubbles led to larger fractions as smaller bubbles had a longer residence time in the column due to a lower buoyancy speed, as already mentioned.

Moreover, the difference observed in the values of the gas fractions could also be attributed to the limited movement of the bubbles due to the adsorbed layer of the frother. More specifically, the coefficient of the drag force exerted on the bubbles increased, and as a result, the buoyancy speed decreased, which led to recording higher volumetric fractions [64]. The fact that the electrical signals were in line with the optical measurements proves that although a photograph represented the instantaneous state of the two-phase flow, the information it contained was sufficient to describe the phenomenon overall.

Figure 14 presents the effect of the frother concentration on the evolution of the gas fraction in the height of the flotation column. The final volume fraction of the gas phase (ε_f) was calculated from the average values of the last measurements of the steady state region from each time series in Figure 12. Apart from indicating the values of the gas fraction, the electrical signals could furthermore be utilized to qualitatively characterize the two-phase flow. Initially, Kostoglou et al. theoretically studied the interactions between adjacent bubbles in two-phase flow as a first attempt to characterize their size through electrical signal fluctuations [65]. Then, Evgenidis and Karapantsios demonstrated experimentally that the intensity and frequency of fluctuations around the mean value of the air volume fraction, in a two-phase flow with constant liquid and gas phase supply, depended on the mean bubble size [66]. More specifically, they developed an empirical relationship according to which the ratio of the standard deviation to the mean value of the air volumetric fraction (StDev (ε_av)/ε_av) increased with increasing the bubble size of the two-phase medium. The above possibility will be explored in the future.

3.3. Intensity of Bubble-Induced Scattered Light Method

To characterize the gas phase of the flotation system in the hybrid flotation column, experiments were carried out using the scattered light optical method. Figure 15 shows the time series of the light energy intensity (W(t)) for the four vertical positions of the flotation column in the presence of dispersed air in the absence of a frother and in the presence of three different concentrations, 60, 120 and 240 mg/L.

In the first 60 s of the measurement, the intensity of light that passed from one side of the device to the other before the introduction of air (baseline) was recorded. The light intensity of the single-phase medium was constant with time as the composition of the slurry did not change, and thus no diffusion of the energy occurred. The air intake (0.7 L/min) after 60 s was accompanied by a sharp drop in light intensity. The light diffused over the dispersed air bubbles it encountered along its path and the energy of the light beam exiting the other wall of the device became less than that of the light source. Air was introduced for 5 min and a steady state was achieved after about 10 s. Moreover, it was observed that at the end of the air intake and, in particular, regarding the higher position, y₄ did not return to its initial state (single-phase medium) as foam had formed, resulting in the impossible gradual restoration of the energy to the levels of the single-phase liquid medium.

We conducted a comparative study to manage the effect of the frother concentration on the change in the energy of the light source for the four vertical positions in the presence of dispersed air. As mentioned, the concentrations we studied were 60, 120, and 240 mg/L. The results depict that a smaller change in energy took place in the presence of water (20–40%) with the lowest frother concentration (40–60%), while as the concentration of oleate increased, the change reached 90%. It was obvious that the change in energy depended on the size of the bubbles of the gas phase and, more specifically, the larger the size of the bubbles, the smaller the change in outward energy, as the light diffused less. In addition, the smaller bubbles moved in the bulk and covered the entire volume of the column, causing the light to diffuse intensely; therefore, a greater change in the recorded energy was observed [67]. The decrease in the light intensity appeared to be uniform for all four vertical positions of the column. This shows that the gas phase was homogeneously distributed along the height of the column and, furthermore, the average size of the bubbles and their distribution did not differ significantly in the different positions.

Figure 16 shows the transmitted energy (%) of the light source due to scattering in the flotation column for various column heights and frother concentrations in presence of dispersed air. The greatest energy reduction was observed at the highest position of the column (y₄ = 46.5 cm). It was observed that at the highest positions and for sodium oleate concentrations of 120 and 240 mg/L, a small percentage of energy that reached the opposite side of the column was recorded, as the average bubble size was smaller due to a reduction in the surface tension.

Figure 17 summarizes the results of the hydrodynamic study on the flotation column regarding the effect of the frother concentration (red line) on the average size of the dispersed air bubbles, (black line) on the time series of the volumetric gas fraction, and (blue line) the energy of the light source that passes through the bulk due to diffusion, for the four different positions of the column in height. It can be concluded that at the highest positions, y₃ and y₄, the increase in the concentration of the frother led to an increase in the volumetric gas fraction, while the recorded energy of the light that reached the detector was lower due its diffusion at the bubble surfaces. The average bubble size decreased with increasing the concentration; however, it did not change significantly at the different vertical positions of the column. The measurements were conducted simultaneously.

3.4. Flow Dispersion

Figure 18 presents the evolution of the electrical resistance of the dispersion as countered by the electrode pairs. The evolution of the electrical resistance was due to the conducting tracer injected in the flow field and it was representative of the residence time distributions of the injection stream in the flotation column at the four different height positions in the presence of sodium oleate (120 mg/L) and dispersed air flow (0.7 L/min). Once a steady state was accomplished, the tracer was injected close to the ceramic sparger in order to ensure a presentative “picture” of the two-phase flow inside the column. Figure 18a shows that the dispersion of the tracer in the working solution was distributed uniformly in the volume of the column, while Figure 18b shows that the injection current was dispersed along the height of the column and reached the four positions with a time difference equal to 1 s, while a steady state (plateau) occurred after 7 s. The presence of bubbles in the flow affected the residence time distribution of the liquid in the column in two ways; initially, by forcing the dispersion of the tracer and, additionally, by controlling the time required to restore equilibrium at different heights. As the bubbles rose to the free surface of the liquid, a velocity profile was created around them, and thus the dispersion of the injection stream in the column liquid was enhanced.

4. Conclusions

In the present work, an experimental study of some of the hydrodynamic characteristics of a hybrid electroflotation column was conducted. For this purpose, optical and electrical methods were utilized to investigate the bubble size distributions, the volumetric gas fraction, and the flow field evolution in the hybrid column.

The experimental results can be concluded as follows:

• The study of the size distribution of the bubbles in the flotation column showed that at the four vertical positions of the column, coarser bubbles appeared in the center of the column, while smaller bubbles appeared closer to the walls through the recirculation zone near the walls. The average bubble size decreased when the frother concentration increased and did not change considerably at the different vertical positions. In general, that trend was observed both vertically and radially for the different positions in the column. Moreover, no coalescence of the bubbles was observed.
• The measurement of the volume fraction showed that the presence of smaller bubbles led to larger air fractions due to a lower buoyancy velocity and an increase in frother concentration as the volume fraction increased.
• The experiments realized with the scattered light optical method for correlating the bubble size and number showed that an increase in frother concentration caused a greater change in the energy of the light source passing through the flotation devices, due to diffusion over smaller and numerous bubbles.
• The study of the residence time distribution in the flotation column depicted that the dispersion led to a uniform trace concentration along the column and the steady state occurred 3 s after the dispersed air was inserted.