- Email: [email protected]
Ultrasound treatment on tailings to enhance copper flotation recovery
Minerals Engineering 99 (2016) 89–95
Contents lists available at ScienceDirect
Minerals Engineering journal homepage: www.elsevier.com/locate/mineng
Ultrasound treatment on tailings to enhance copper flotation recovery A.R. Videla a,⇑, R. Morales a, T. Saint-Jean a, L. Gaete b, Y. Vargas b, J.D. Miller c a
Mining Engineering Department, Faculty of Engineering, Pontificia Universidad Catolica de Chile, Chile Physics Department, Faculty of Science, University of Santiago, Chile c Metallurgical Engineering Department, College of Mines and Earth Sciences, University of Utah, USA b
a r t i c l e
i n f o
Article history: Received 24 March 2016 Revised 5 September 2016 Accepted 24 September 2016
Keywords: Ultrasonic treatment Ultrasound Flotation Copper tailings Copper recovery
a b s t r a c t As ore grades fall, the amount of tailings production for the same copper production is expected to rise. Flotation recovery of copper sulfide from the El Teniente plant has deteriorated in recent years, in this regard ultrasound treatment of tailings for enhanced copper recovery was evaluated in laboratory experiments. The impact of the ultrasound wave was examined under different conditions, with the conclusion that improved recovery occurs when ultrasound is applied during conditioning and flotation. The results can be explained by the effect of acoustic cavitation that cleans particle surfaces and minimizes slime coatings, facilitating the action of the reagents. In this way, improvement in copper recovery of up to 3.5% were obtained. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction One of the biggest challenges facing the mining industry is the deterioration of mineral grades. This problem makes it necessary to improve the recovery of current operations to offset the steady decline of Cu grades. Such a decline of average saturation grade has meant a fall of 1.0% Cu to 0.86% of global average grade during the last decade (Chile, Ministerio de Minería, Corporación Chilena del Cobre, 2015). It is estimated that in Chile the amount of tailings generated only as products of copper sulfide concentration reaches 481 M ton annually, a figure that will continue to rise as mineral grades fall. That is why it is interesting to explore the use of new technologies to improve overall recovery and reduce tailings. Ultrasound treatment is used in various industries for surface cleaning (Farmer et al., 2000; Gallego-Juarez, 1994; Zhao et al., 2007). Ultrasound treatment has lead to interesting research in the mineral processing industry to improve coal processing (Ozkan and Kuyumcu, 2006, 2007; Ozkan, 2012; Ambedkar et al., 2011), fluid-solid particle separation (Riera-Franco de Sarabia et al., 2000) and flotation of oil shale cleaning (Altun et al., 2009). Moreover, various studies to optimize the flotation process by using ultrasound treatment have been conducted (Aldrich and Feng, 1999; Ishak and Rowson, 2009; Kang et al., 2009; Ozkan and Gungoren, 2012; Schlesinger and Muter, 1989; Vargas et al., 2006), but none of the studies contemplated the re-treatment of ⇑ Corresponding author. E-mail address: [email protected] (A.R. Videla). http://dx.doi.org/10.1016/j.mineng.2016.09.019 0892-6875/Ó 2016 Elsevier Ltd. All rights reserved.
the poorest tailings in order to recover bornite, chalcopyrite, cuprite or molybdenite. Following this line of research, Ozkan (2002) has shown that pre-treatment of magnesite by ultrasound increases magnetite recovery. In his analysis, he assumed a cleaning effect on the surface of the mineral. Similarly, Kang et al. (2009) have shown that the application of ultrasound radiation in coal flotation improves the efficiency of the process, reducing ash recovery. Also, sulfides were oxidized due to the ultrasound effect. Aldrich and Feng (1999) observed an improvement in the ultimate recovery of sulfur from Merensky Reef, Africa, using ultrasound treatment, despite the negative effects of the concentration of solids, temperature, conditioning time and gas flow. Meanwhile, Vargas et al. (2006) found that conditioning by means of ultrasonic radiation could preserve the concentration of copper sulfides, but decrease the recovery of iron sulfides, resulting in improved selectivity in the flotation, reducing the recovery of pyrite and increasing the concentrate grade. Previous work (Stoev et al., 1992) have shown the significant positive impact that may have the application of mechanical vibration and the acoustic wave excited by them in several mineral processing units, particularly in flotation. Their results were promising and show applications of vibroacoustic excitation at low frequencies to promote attachment of mineral particles to air bubbles, increment of recovery, increment of flotation rate, increment of selectivity, control bubbles size, emulsification of reagents, desorption of reagents, and froth control among others. The principle that justifies the use of ultrasound relates to the same phenomena
90
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
described in this seminal work, a process based on the wave spreading in the aqueous medium. Additionally, as the wave frequency increases reaching high intensities such in ultrasound, the vibroacoustic waves can lead to acoustic cavitation, a phenomenon that causes bubble generation due to progressive pressure drop in each wave cycle. The bubbles begin to oscillate along the sound field in the first stage; after a few cycles the bubbles reach a critical size at which they resonate in a nonlinear regime. This regime makes it difficult for the bubbles to expand to larger sizes, making the medium’s response increasingly important, due to the inertial forces that cause shrinkage of the bubble in the compression cycle (Gaete et al., 1997). Eventually the bubbles collapse under considerable elevated pressure emanating from the medium (Mitome, 2003). This implosion generates a supersonic microjet of fluid in the area where the bubble collapses. It has been shown that this microjet possesses great capacity to erode a surface which would remove contaminants on the surfaces of the particles, thus improving the efficiency of the concentration process. This study suggests that the cavitation phenomenon caused by ultrasound could improve the flotation tailings of copper sulfides. The tailings consist of fine and ultra-fine particles of gangue, and a percentage of particles containing copper that were not floated due to lack of release and surface contamination. Copper sulfides inside the tailings might have been oxidized and ultrasound treatment may have removed the oxidized layer from the particles surfaces as well as helping the desliming. This would mean an improvement in the overall recovery of the system before sending the tailings to their final deposition. Subsequently, the methodology of the development and evaluation of laboratory flotation tests for processing the tailings of copper sulfide is presented. For this evaluation a ‘Denver’ flotation cell has been changed to be able to control the ultrasound field applied to the particles. The results are analyzed via obtaining the kinetic curve flotation, its parameters and the resulting Cu grade. 2. Materials and methods 2.1. Materials To develop this study, copper sulfide tailings, provided by Valle Central Mine, were used. The tailings have an average copper grade
of 0.12%, with most in the form of chalcopyrite, and with a specific gravity of 2.7 [g/cm3]. The tailings consist of fine and ultra-fine particles under 70 mesh, which have previously passed through a stage of grinding and flotation in the mine of ‘El Teniente’, a branch of the national copper company CODELCO Chile, a mill plant located in central Chile. Three tailings samples of 500 ml out of the 100 L sample were analysed to characterize the particle size of the supplied material. After tailings were homogenized, increments of 50 ml were taken from different parts of the container to complete one sample. The samples were extracted and analysed via laser diffraction technology with a Mastersizer 2000. Table 1 and Fig. 1 display the results of the analyses and distributions of the size for each sample, respectively. As can be seen, the results show that the tailings is mostly composed of particles of about 10 lm, with 50% of the particles under 14 lm and 90% of the particles under 220 lm. Sample one is coarser than sample two and three. This difference may due to the fact that the upstream classification process is not operating perfectly and few coarse particles may have end in the final sample. In average though, the product size is below the limit. No agglomeration was observed in the sample. Samples were sonicated 20 min before the particle size analysis. 2.2. Ultrasound treatment In this study we evaluated the use of ultrasound to improve the recovery of copper sulfides by applying three different conditions, derived from experience gained from evaluating heavy minerals in previous research (Celik, 1989; Cilek and Ozgen, 2009; Djendova and Mehandjiski, 1992; Aldrich and Feng, 1999; Ozkan, 2002; Vargas et al., 2006). The three ultrasound conditions applied are: Ultrasound Conditioning (UC) where ultrasound is applied only during conditioning; Ultrasound Flotation (UF) which applies ultrasound only in flotation; Flotation and Conditioning with Ultrasound (FCU), in which ultrasound is applied throughout the process of conditioning and during flotation. These tests were compared with the results of a basic test called Standard Flotation (SF) that considers the same operating conditions but without applying ultrasound during any of the processess. All tests were run in random order and in duplicate to control the variability of
Table 1 Characterization of collected samples. Sample 2
Sample 3
Average
Standard deviation
7,48 0,86 7,00 158,02 2,76 20,12 402,70
3,32 1,15 5,20 40,76 2,12 10,82 118,53
5,07 1,11 5,43 59,73 2,27 10,79 130,01
5,29 1,04 5,88 86,17 2,39 13,91 217,08
2,09 0,16 0,98 62,94 0,34 5,38 160,86
Volume [%]
Volume [%]
Sample 1
Uniformity Specific surface [m2/g] Average particle size D[3,2] [lm] Average particle size D[4,3] [lm] 10% Under diameter [lm] 50% Under diameter [lm] 90% Under diameter [lm]
Volume [%]
Characteristic
Parcle Size (Diameter) [μm]
Parcle Size (Diameter) [μm]
Parcle Size (Diameter) [μm]
Fig. 1. Particle size distribution in three samples collected from Minera Valle Central feed. Sample 1, 2 and 3 are shown from left to right.
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95 Table 2 Experimental design for flotation with ultrasound. Identification
Conditioning with ultrasound [yes/no]
Flotation with ultrasound [yes/no]
SF UC UF FCU
No Yes No Yes
No No Yes Yes
results. Table 2 summarizes the conditions of application of ultrasound in the four types of tests. The flotation tests were conducted in a ‘Denver’ cell with a stainless steel reactor of 8 L effective capacity, fitted with 9 ultrasonic transducers (Clangsonic 2045-68LB P8), as shown in Fig. 2. The transducers were connected to a Clangsonic SONOCLG XR 2000G generator of 2000 W. Each transducer emits 100 W of power at a frequency of 20 kHz, obtaining an equivalent power of 112.5 W/L. The ultrasonic frequency was kept fixed in one of Eigen mode of the transducers during testing. A calibrated probe, designed especially to measure cavitating acoustic fields (Gaete et al., 1993), has been used to determine the effective intensity of the acoustic field present in the flotation cell. The probe’s design was based on stainless steel and brass, which transforms external mechanical disturbances into electrical signals that are sent to Arduino hardware, where the signals are processed with a code written in Python. The probe measures 35 cm in length and 1.3 mm in diameter. With these dimensions it is possible to make three-dimensional scans inside the Denver cell without excessively modifying the internal acoustic field. The experimental system was mounted within a steel structure that allows three-dimensional scans controlled by a computer with an accuracy of 0.1 mm in all three axes. Using that setup it was possible to measure the acoustic pressure at three different planes inside the cell as shown in Fig. 3. The experimental set-up allow to measured the acoustic pressure on the scanned planes gener-
(a)
Denver Flotaon Machine
(b)
Cell
Transducers Fig. 2. Schematic design of the Denver equipment and the flotation cell: (a) Outline of the Denver flotation equipment and the cell, side view; (b) Outline of the flotation cell, top view.
91
ated by the acoustic field: the displacement of the probe was driven for a 3D system computer-controlled with a device similar to the CNC machines. The acoustic pressures within the cell at the three levels indicated are presented in Fig. 4. It can be observed how the acoustic field concentrates in certain areas of the cell, raising the pressure up to almost 75 kPa. The lowest pressure level measured was around 15 kPa, therefore the ratio between maximum and minimum acoustic pressure inside the Denver Cell is 5. In Fig. 4 the root mean square (RMS) intensity of the acoustic field is shown on a color scale, where dark red represents the greatest intensity of the acoustic field and blue the lower intensity. In the areas of high energy concentration a greater effect of acoustic cavitation should occur. Fig. 4 reveals that exposure of the tailings slurry to the ultrasonic field is not uniform throughout the cell. The transducers have a fundamental excitation frequency of 20.804 kHz amplitude for a sinusoidal signal of 36.36 kHz frequency, which is maintained with small variations of approx. 1 kHz. This type of excitation results in widening the band of the power signal, allowing compensation for small differences in the resonance frequency of each transducer. Fig. 5 shows the excitation wavelength used for the modulation of the transducers. 2.3. Flotation experiments The operating conditions for the flotation experiment were kept constant for all tests and are summarized in Table 3. Reagent OrePrep X-133 was used as frother and Aero 343 as collector, both supplied by CYTEC. A similar procedure was follows for each flotation test. At the start of each test, the percentage of solids of the material was measured through a Marcy pulp density scale, adjusting the concentration with water until the required value in the cell was reached. Originally, time counting initiates the conditioning stage under stirring only (without aeration) by adding the respective doses of each reagent. Regarding reagent addition, pH is adjusted by adding CaO until pH 10 is reached, then 30 g per ton of collector and frother is added. Initiation of aeration ended the conditioning stage and started the flotation stage. During the flotation process CaO was added to maintain pH 10 throughout the test. The concentrate froth was collected manually via over-flow and paddling a spatula at regular time intervals of approximately 15 s, collecting samples at 0.5, 1, 2, 4, 7 and 12 min. To calculate the copper content, the products were dried and analyzed by atomic absorption spectrophotometry on a Perkin Elmer Analyst 100 instrument. The test was considered valid if the difference between the measurements of each Cu content between replicates was less than 2.5%. Regarding the calculation of the recovery, the results were adjusted to an equation of first order recovery (Wills and Napier-Munn, 2005) by adjusting the data according to the model proposed by García-Zúñiga:
Rt ðtÞ ¼ R1 ð1 eðktÞ Þ
ð1Þ
where Rt is recovery (%) of a valuable metal in an instant t (min); with R1 representing the maximum estimated recovery and k the kinetic constant of the flotation process, respectively. 3. Results
Fig. 3. Lateral views of plane under study inside the DENVER cell.
As Fig. 6 shows, the García-Zúñiga model (Eq. (1)) fits wells to the data under analysis. Fig. 6(a) shows that the recovery in the standard flotation (SF) test with no ultrasound applied. Fig. 6(b) shows the results for ultrasound during conditioning only (UC). Fig. 6(c) shows results for ultrasound during flotation only (UF) and finally Fig. 6(d) shows when ultrasound is applied during
92
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
Fig. 4. Distribution of the absolute critical pressure inside the Denver flotation cell. The first picture (upper) corresponds to plane 1 inFig. 3, the second one to plane 2 and the last one to plane 3.
Fig. 5. Excitation signal for the transducer system. The top signal is for a period of the wave that modulates in amplitude below. It can be seen how this variation is repeated over time.
93
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95 Table 3 Conditions of the flotation process. Parameters
Ranges
Conditioning time Frother (OrePrep X–133) Collector (Aero 343) Air impeller speed Solids percentage Pulp pH (CaO added)
6 min 30 g/ton 30 g/ton 1200 rpm 30% 10
Table 4 Average parameters adjusted to the flotation results using the García - Zúñiga model. Maximum recovery (R1), kinetic constant (k) and Coefficient of Variation (COV) are shown.
conditioning and flotation time as well (FCU). From Figures can be observed that after twelve minutes of flotation each condition SF, UC, UF, and FCU reaches 27.89%, 28.94%, 29.23% and 31.40% recovery respectively. Therefore, the application of ultrasound equals or exceeds the results obtained by standard flotation without ultrasound. The first order flotation model also helps to evaluate the kinetic response of the flotation process. In this regard, once the recovery limit (R1) was determined, we proceeded to calculate the kinetic flotation constant (k) according to Eq. (1). In Table 3 adjusted values and the variation coefficients are shown, obtained by the least squares method. All adjustments conform to a very low variation level as shown in Table 4. The FCU test shows recovery being 4 percentage points higher than standard flotation without ultrasound. This improvement was significantly above measurement error. In the best case, the FCU treatment presents a statistically significant difference with the Cu recovery obtained by SF with a 95% confidence. In the other two cases the statistically confidence is reduced but still is over 80%. From the results we can assure there is an experimental difference in the Cu recovery.
R1 [%] Average
C.O.V [%] R1
k [min1] Average
C.O.V [%] k
SF UC UF FCU
27,89 28,94 29,23 31,40
4,07 0,93 2,63 2,95
0,55 0,43 0,37 0,43
4,11 2,33 4,66 6,98
Fig. 7 shows a superposition of the mean values of the resulting flotation tests. It is graphically observed that the FCU test with ultrasound during conditioning and flotation improved the recovery presenting a faster kinetic response than standard flotation. Regarding the concentrate grade (% Cu), it was found that when ultrasound was applied the concentrate grade improved in relation to standard flotation. As shown in Fig. 8, the best result was obtained by applying ultrasound during the flotation stage only, improving from 0.6% Cu to a final grade of 0.77% Cu. One explanation for these improvements is that fine and ultrafine particles are superficially cleaned by the microjets from the collapse of cavitation bubbles. Cleaning the surface facilitates the action of the collector on the cleaned surface, increasing the percentage of available surface for adhesion of the collector, and so fomenting the action of the reagents. Another hypothesis is that the detected beneficial effect is due to an agglomeration action of ultrasound on the ultrafine particles (6–50 lm) in the areas of highest sound pressure (Fig. 4), which would allow them to acquire enough kinetic energy to emerge together with the bubbles
35%
35%
(a)
(b)
30%
30%
25%
25%
Recovery (%)
Recovery (%)
Identification
20% 15% 10%
15% 10%
Test 1
Test 1 Test 2
5%
Test 2
5%
20%
Mean
Mean
0%
0% 0
5
10
15
0
5
Time (min) 35%
35%
(c)
30% 25% 20% 15% 10%
(d)
25% 20% 15% 10%
Test 1 Test 2
5%
Test 1 Test 2
5%
Mean
0%
0
5
10
Time (min)
15
30%
Recovery (%)
Recovery (%)
10
Time (min)
Mean
15
0%
0
5
10
Time (min)
Fig. 6. García Zúñiga setting for the quadruples of (a) SF, (b), UC (c), UF and (d) FCU.
15
94
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
35%
Recovery (%)
30% 25% 20% 15% SF
10%
UC UF
5%
FCU
0%
0
2
4
6
8
10
12
14
(Vargas et al., 2006). Additionally, the effect of the ultrasound could stem from the generation of sufficiently small bubbles leading to a better particle bubble and particle size relation than the standard flotation without ultrasound, improving the flotation of fine particles. An analysis of some particle tailings has been conducted to determine if any surface cleaning ocurred. This analysis was undertaken via electronic scanning by SEM (Scanning Electron Microscopy), obtaining images presented in Fig. 9. In most cases a decrease of ultrafine particles is observed on the surface of the particles subjected to ultrasound, which would corroborate the hypothesis of the cleaning effect of ultrasound and would improve the efficiency of the reagents.
Time (min) Fig. 7. Average recovery kinetics according to type of ultrasound treatment.
0.90%
Cumulave Grade (%)
0.85% 0.80% 0.75% 0.70% 0.65% 0.60% 0.55%
SF UC UF FCU
0.50% 0.45% 0.40%
0
2
4
6
8
10
12
14
Time (min) Fig. 8. Concentrate grade as a function of flotation time for different ultrasound treatments.
4. Conclusions Based on the experimental results, ultrasound could be a complementary technology to process tailings in classical flotation processes given its potential to improve the recovery of the fine and ultrafine particles. For a standard copper concentrator overall recovery is around 90%, therefore 10% of the feed Cu is lost in the tailings and could be potentially recovered. As shown from the current results, if a 30% recovery is achieved from this 10% copper lost in the tailings, it could mean an increase of about 3.1 percentage points in the overall recovery, achieving a total of 93.1% Cu recovery. Therefore the use of ultrasound during conditioning and flotation of fines and ultrafines deserves further consideration although the economical incentive to apply this technology will depend on the balance between the marginal benefit obtained with the extra Copper production and the investment cost need to apply an equivalent ultrasound treatment to high volumes. Given the current cost of the high frequency ultrasound
Fig. 9. Images of concentrate chalcopyrite particles resulting in the concentrate after flotation from SF and FCU experiments (magnification 6.000). Above: SF concentrated particles. Below: FCU concentrated particles.
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
equipment there is still significant work to be done to make the technology attractive. It is believed that improvements are mainly explained by the phenomenon of surface cleaning via fluid cavitation produced by ultrasonic waves and bubble jets, resulting in cleaning the particle surface and improving the functioning of the collector. However, more research is required to understand and properly control the applied pressure field and the effects on the flotation of ultrafine particles. It would be particularly convenient to make further progress in understanding the potential action of agglomeration and deagglomeration. of the ultrafine particles via ultrasound. Other possible effects to be considered is the ultrasound effect on the liquid. It was observed that the energy density heats up the slurry in some degrees, which potentially leads to changes in the chemical equilibrium in solution, which may have an effect in pH, redox potential, z–potential and contact angle. According to the results shown, ultrasound applied at any stage of the process improves final recovery and kinetics of the flotation process of fine and ultrafine copper sulfide particles under evaluation. This improvement is above measurement errors of the tests and represents significant increase in recovery compared to the recovery obtained by standard flotation. Importantly, the latter feature may allow reduction in the volume of reactors needed to reach a given throughput level for the same Cu recovery, allowing reduction of facilities lay-out and/or increasing production capacity with the same equipment. Fig. 4 demonstrates that exposure of the tailings slurry to the ultrasonic field is not uniform inside the cell. Design should be optimized and controlled with an optimal distribution of the transducers in the reactor’s geometry. Another alternative to improve acoustic field uniformity would be the redesign of the cell so that the ultrasound effects are maximized, searching resonance conditions. In any case, transducer interference should be avoided as it is destructive and lowers the efficiency of the process. This aspect of the research and development program will be addressed in a subsequent investigation. The presented research work has been limited to evaluation of ultrasound impact on the particle solid surface. However ultrasound irradiation impacts on solid-liquid interfaces and liquid-air interfaces changing for instance the contact angle. Some of our preliminary studies have shown that a significant change in contact angle can be achieved by the solid surface vibration lead by an ultrasound field. These impacts require further attention and should be addressed in future research.
References Aldrich, C., Feng, D., 1999. Effect of ultrasonic preconditioning of pulp on the flotation of sulphide ores. Miner. Eng. 12 (6), 701–707. http://dx.doi.org/ 10.1016/S0892-6875(99)00053-9.
95
Altun, N.E., Hwang, J.-Y., Hicyilmaz, C., 2009. Enhancement of flotation performance of oil shale cleaning by ultrasonic treatment. Int. J. Miner. Process. 91 (1–2), 1–13. Ambedkar, B., Nagarajan, R., Jayanti, S., 2011. Investigation of high-frequency, highintensity ultrasonics for size reduction and washing of coal in aqueous medium. Ind. Eng. Chem. Res. 50 (23), 13210–13219. Celik, M.S., 1989. Effect of ultrasonic treatment on the floatability of coal and galena. Sep. Sci. Technol. 24, 1159–1166. Chile, Ministerio de Minería, Corporación Chilena del Cobre, 2015. Anuario de Estadísticas del Cobre y otros Minerales 1995–2014. Recuperado de.
Contents lists available at ScienceDirect
Minerals Engineering journal homepage: www.elsevier.com/locate/mineng
Ultrasound treatment on tailings to enhance copper flotation recovery A.R. Videla a,⇑, R. Morales a, T. Saint-Jean a, L. Gaete b, Y. Vargas b, J.D. Miller c a
Mining Engineering Department, Faculty of Engineering, Pontificia Universidad Catolica de Chile, Chile Physics Department, Faculty of Science, University of Santiago, Chile c Metallurgical Engineering Department, College of Mines and Earth Sciences, University of Utah, USA b
a r t i c l e
i n f o
Article history: Received 24 March 2016 Revised 5 September 2016 Accepted 24 September 2016
Keywords: Ultrasonic treatment Ultrasound Flotation Copper tailings Copper recovery
a b s t r a c t As ore grades fall, the amount of tailings production for the same copper production is expected to rise. Flotation recovery of copper sulfide from the El Teniente plant has deteriorated in recent years, in this regard ultrasound treatment of tailings for enhanced copper recovery was evaluated in laboratory experiments. The impact of the ultrasound wave was examined under different conditions, with the conclusion that improved recovery occurs when ultrasound is applied during conditioning and flotation. The results can be explained by the effect of acoustic cavitation that cleans particle surfaces and minimizes slime coatings, facilitating the action of the reagents. In this way, improvement in copper recovery of up to 3.5% were obtained. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction One of the biggest challenges facing the mining industry is the deterioration of mineral grades. This problem makes it necessary to improve the recovery of current operations to offset the steady decline of Cu grades. Such a decline of average saturation grade has meant a fall of 1.0% Cu to 0.86% of global average grade during the last decade (Chile, Ministerio de Minería, Corporación Chilena del Cobre, 2015). It is estimated that in Chile the amount of tailings generated only as products of copper sulfide concentration reaches 481 M ton annually, a figure that will continue to rise as mineral grades fall. That is why it is interesting to explore the use of new technologies to improve overall recovery and reduce tailings. Ultrasound treatment is used in various industries for surface cleaning (Farmer et al., 2000; Gallego-Juarez, 1994; Zhao et al., 2007). Ultrasound treatment has lead to interesting research in the mineral processing industry to improve coal processing (Ozkan and Kuyumcu, 2006, 2007; Ozkan, 2012; Ambedkar et al., 2011), fluid-solid particle separation (Riera-Franco de Sarabia et al., 2000) and flotation of oil shale cleaning (Altun et al., 2009). Moreover, various studies to optimize the flotation process by using ultrasound treatment have been conducted (Aldrich and Feng, 1999; Ishak and Rowson, 2009; Kang et al., 2009; Ozkan and Gungoren, 2012; Schlesinger and Muter, 1989; Vargas et al., 2006), but none of the studies contemplated the re-treatment of ⇑ Corresponding author. E-mail address: [email protected] (A.R. Videla). http://dx.doi.org/10.1016/j.mineng.2016.09.019 0892-6875/Ó 2016 Elsevier Ltd. All rights reserved.
the poorest tailings in order to recover bornite, chalcopyrite, cuprite or molybdenite. Following this line of research, Ozkan (2002) has shown that pre-treatment of magnesite by ultrasound increases magnetite recovery. In his analysis, he assumed a cleaning effect on the surface of the mineral. Similarly, Kang et al. (2009) have shown that the application of ultrasound radiation in coal flotation improves the efficiency of the process, reducing ash recovery. Also, sulfides were oxidized due to the ultrasound effect. Aldrich and Feng (1999) observed an improvement in the ultimate recovery of sulfur from Merensky Reef, Africa, using ultrasound treatment, despite the negative effects of the concentration of solids, temperature, conditioning time and gas flow. Meanwhile, Vargas et al. (2006) found that conditioning by means of ultrasonic radiation could preserve the concentration of copper sulfides, but decrease the recovery of iron sulfides, resulting in improved selectivity in the flotation, reducing the recovery of pyrite and increasing the concentrate grade. Previous work (Stoev et al., 1992) have shown the significant positive impact that may have the application of mechanical vibration and the acoustic wave excited by them in several mineral processing units, particularly in flotation. Their results were promising and show applications of vibroacoustic excitation at low frequencies to promote attachment of mineral particles to air bubbles, increment of recovery, increment of flotation rate, increment of selectivity, control bubbles size, emulsification of reagents, desorption of reagents, and froth control among others. The principle that justifies the use of ultrasound relates to the same phenomena
90
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
described in this seminal work, a process based on the wave spreading in the aqueous medium. Additionally, as the wave frequency increases reaching high intensities such in ultrasound, the vibroacoustic waves can lead to acoustic cavitation, a phenomenon that causes bubble generation due to progressive pressure drop in each wave cycle. The bubbles begin to oscillate along the sound field in the first stage; after a few cycles the bubbles reach a critical size at which they resonate in a nonlinear regime. This regime makes it difficult for the bubbles to expand to larger sizes, making the medium’s response increasingly important, due to the inertial forces that cause shrinkage of the bubble in the compression cycle (Gaete et al., 1997). Eventually the bubbles collapse under considerable elevated pressure emanating from the medium (Mitome, 2003). This implosion generates a supersonic microjet of fluid in the area where the bubble collapses. It has been shown that this microjet possesses great capacity to erode a surface which would remove contaminants on the surfaces of the particles, thus improving the efficiency of the concentration process. This study suggests that the cavitation phenomenon caused by ultrasound could improve the flotation tailings of copper sulfides. The tailings consist of fine and ultra-fine particles of gangue, and a percentage of particles containing copper that were not floated due to lack of release and surface contamination. Copper sulfides inside the tailings might have been oxidized and ultrasound treatment may have removed the oxidized layer from the particles surfaces as well as helping the desliming. This would mean an improvement in the overall recovery of the system before sending the tailings to their final deposition. Subsequently, the methodology of the development and evaluation of laboratory flotation tests for processing the tailings of copper sulfide is presented. For this evaluation a ‘Denver’ flotation cell has been changed to be able to control the ultrasound field applied to the particles. The results are analyzed via obtaining the kinetic curve flotation, its parameters and the resulting Cu grade. 2. Materials and methods 2.1. Materials To develop this study, copper sulfide tailings, provided by Valle Central Mine, were used. The tailings have an average copper grade
of 0.12%, with most in the form of chalcopyrite, and with a specific gravity of 2.7 [g/cm3]. The tailings consist of fine and ultra-fine particles under 70 mesh, which have previously passed through a stage of grinding and flotation in the mine of ‘El Teniente’, a branch of the national copper company CODELCO Chile, a mill plant located in central Chile. Three tailings samples of 500 ml out of the 100 L sample were analysed to characterize the particle size of the supplied material. After tailings were homogenized, increments of 50 ml were taken from different parts of the container to complete one sample. The samples were extracted and analysed via laser diffraction technology with a Mastersizer 2000. Table 1 and Fig. 1 display the results of the analyses and distributions of the size for each sample, respectively. As can be seen, the results show that the tailings is mostly composed of particles of about 10 lm, with 50% of the particles under 14 lm and 90% of the particles under 220 lm. Sample one is coarser than sample two and three. This difference may due to the fact that the upstream classification process is not operating perfectly and few coarse particles may have end in the final sample. In average though, the product size is below the limit. No agglomeration was observed in the sample. Samples were sonicated 20 min before the particle size analysis. 2.2. Ultrasound treatment In this study we evaluated the use of ultrasound to improve the recovery of copper sulfides by applying three different conditions, derived from experience gained from evaluating heavy minerals in previous research (Celik, 1989; Cilek and Ozgen, 2009; Djendova and Mehandjiski, 1992; Aldrich and Feng, 1999; Ozkan, 2002; Vargas et al., 2006). The three ultrasound conditions applied are: Ultrasound Conditioning (UC) where ultrasound is applied only during conditioning; Ultrasound Flotation (UF) which applies ultrasound only in flotation; Flotation and Conditioning with Ultrasound (FCU), in which ultrasound is applied throughout the process of conditioning and during flotation. These tests were compared with the results of a basic test called Standard Flotation (SF) that considers the same operating conditions but without applying ultrasound during any of the processess. All tests were run in random order and in duplicate to control the variability of
Table 1 Characterization of collected samples. Sample 2
Sample 3
Average
Standard deviation
7,48 0,86 7,00 158,02 2,76 20,12 402,70
3,32 1,15 5,20 40,76 2,12 10,82 118,53
5,07 1,11 5,43 59,73 2,27 10,79 130,01
5,29 1,04 5,88 86,17 2,39 13,91 217,08
2,09 0,16 0,98 62,94 0,34 5,38 160,86
Volume [%]
Volume [%]
Sample 1
Uniformity Specific surface [m2/g] Average particle size D[3,2] [lm] Average particle size D[4,3] [lm] 10% Under diameter [lm] 50% Under diameter [lm] 90% Under diameter [lm]
Volume [%]
Characteristic
Parcle Size (Diameter) [μm]
Parcle Size (Diameter) [μm]
Parcle Size (Diameter) [μm]
Fig. 1. Particle size distribution in three samples collected from Minera Valle Central feed. Sample 1, 2 and 3 are shown from left to right.
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95 Table 2 Experimental design for flotation with ultrasound. Identification
Conditioning with ultrasound [yes/no]
Flotation with ultrasound [yes/no]
SF UC UF FCU
No Yes No Yes
No No Yes Yes
results. Table 2 summarizes the conditions of application of ultrasound in the four types of tests. The flotation tests were conducted in a ‘Denver’ cell with a stainless steel reactor of 8 L effective capacity, fitted with 9 ultrasonic transducers (Clangsonic 2045-68LB P8), as shown in Fig. 2. The transducers were connected to a Clangsonic SONOCLG XR 2000G generator of 2000 W. Each transducer emits 100 W of power at a frequency of 20 kHz, obtaining an equivalent power of 112.5 W/L. The ultrasonic frequency was kept fixed in one of Eigen mode of the transducers during testing. A calibrated probe, designed especially to measure cavitating acoustic fields (Gaete et al., 1993), has been used to determine the effective intensity of the acoustic field present in the flotation cell. The probe’s design was based on stainless steel and brass, which transforms external mechanical disturbances into electrical signals that are sent to Arduino hardware, where the signals are processed with a code written in Python. The probe measures 35 cm in length and 1.3 mm in diameter. With these dimensions it is possible to make three-dimensional scans inside the Denver cell without excessively modifying the internal acoustic field. The experimental system was mounted within a steel structure that allows three-dimensional scans controlled by a computer with an accuracy of 0.1 mm in all three axes. Using that setup it was possible to measure the acoustic pressure at three different planes inside the cell as shown in Fig. 3. The experimental set-up allow to measured the acoustic pressure on the scanned planes gener-
(a)
Denver Flotaon Machine
(b)
Cell
Transducers Fig. 2. Schematic design of the Denver equipment and the flotation cell: (a) Outline of the Denver flotation equipment and the cell, side view; (b) Outline of the flotation cell, top view.
91
ated by the acoustic field: the displacement of the probe was driven for a 3D system computer-controlled with a device similar to the CNC machines. The acoustic pressures within the cell at the three levels indicated are presented in Fig. 4. It can be observed how the acoustic field concentrates in certain areas of the cell, raising the pressure up to almost 75 kPa. The lowest pressure level measured was around 15 kPa, therefore the ratio between maximum and minimum acoustic pressure inside the Denver Cell is 5. In Fig. 4 the root mean square (RMS) intensity of the acoustic field is shown on a color scale, where dark red represents the greatest intensity of the acoustic field and blue the lower intensity. In the areas of high energy concentration a greater effect of acoustic cavitation should occur. Fig. 4 reveals that exposure of the tailings slurry to the ultrasonic field is not uniform throughout the cell. The transducers have a fundamental excitation frequency of 20.804 kHz amplitude for a sinusoidal signal of 36.36 kHz frequency, which is maintained with small variations of approx. 1 kHz. This type of excitation results in widening the band of the power signal, allowing compensation for small differences in the resonance frequency of each transducer. Fig. 5 shows the excitation wavelength used for the modulation of the transducers. 2.3. Flotation experiments The operating conditions for the flotation experiment were kept constant for all tests and are summarized in Table 3. Reagent OrePrep X-133 was used as frother and Aero 343 as collector, both supplied by CYTEC. A similar procedure was follows for each flotation test. At the start of each test, the percentage of solids of the material was measured through a Marcy pulp density scale, adjusting the concentration with water until the required value in the cell was reached. Originally, time counting initiates the conditioning stage under stirring only (without aeration) by adding the respective doses of each reagent. Regarding reagent addition, pH is adjusted by adding CaO until pH 10 is reached, then 30 g per ton of collector and frother is added. Initiation of aeration ended the conditioning stage and started the flotation stage. During the flotation process CaO was added to maintain pH 10 throughout the test. The concentrate froth was collected manually via over-flow and paddling a spatula at regular time intervals of approximately 15 s, collecting samples at 0.5, 1, 2, 4, 7 and 12 min. To calculate the copper content, the products were dried and analyzed by atomic absorption spectrophotometry on a Perkin Elmer Analyst 100 instrument. The test was considered valid if the difference between the measurements of each Cu content between replicates was less than 2.5%. Regarding the calculation of the recovery, the results were adjusted to an equation of first order recovery (Wills and Napier-Munn, 2005) by adjusting the data according to the model proposed by García-Zúñiga:
Rt ðtÞ ¼ R1 ð1 eðktÞ Þ
ð1Þ
where Rt is recovery (%) of a valuable metal in an instant t (min); with R1 representing the maximum estimated recovery and k the kinetic constant of the flotation process, respectively. 3. Results
Fig. 3. Lateral views of plane under study inside the DENVER cell.
As Fig. 6 shows, the García-Zúñiga model (Eq. (1)) fits wells to the data under analysis. Fig. 6(a) shows that the recovery in the standard flotation (SF) test with no ultrasound applied. Fig. 6(b) shows the results for ultrasound during conditioning only (UC). Fig. 6(c) shows results for ultrasound during flotation only (UF) and finally Fig. 6(d) shows when ultrasound is applied during
92
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
Fig. 4. Distribution of the absolute critical pressure inside the Denver flotation cell. The first picture (upper) corresponds to plane 1 inFig. 3, the second one to plane 2 and the last one to plane 3.
Fig. 5. Excitation signal for the transducer system. The top signal is for a period of the wave that modulates in amplitude below. It can be seen how this variation is repeated over time.
93
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95 Table 3 Conditions of the flotation process. Parameters
Ranges
Conditioning time Frother (OrePrep X–133) Collector (Aero 343) Air impeller speed Solids percentage Pulp pH (CaO added)
6 min 30 g/ton 30 g/ton 1200 rpm 30% 10
Table 4 Average parameters adjusted to the flotation results using the García - Zúñiga model. Maximum recovery (R1), kinetic constant (k) and Coefficient of Variation (COV) are shown.
conditioning and flotation time as well (FCU). From Figures can be observed that after twelve minutes of flotation each condition SF, UC, UF, and FCU reaches 27.89%, 28.94%, 29.23% and 31.40% recovery respectively. Therefore, the application of ultrasound equals or exceeds the results obtained by standard flotation without ultrasound. The first order flotation model also helps to evaluate the kinetic response of the flotation process. In this regard, once the recovery limit (R1) was determined, we proceeded to calculate the kinetic flotation constant (k) according to Eq. (1). In Table 3 adjusted values and the variation coefficients are shown, obtained by the least squares method. All adjustments conform to a very low variation level as shown in Table 4. The FCU test shows recovery being 4 percentage points higher than standard flotation without ultrasound. This improvement was significantly above measurement error. In the best case, the FCU treatment presents a statistically significant difference with the Cu recovery obtained by SF with a 95% confidence. In the other two cases the statistically confidence is reduced but still is over 80%. From the results we can assure there is an experimental difference in the Cu recovery.
R1 [%] Average
C.O.V [%] R1
k [min1] Average
C.O.V [%] k
SF UC UF FCU
27,89 28,94 29,23 31,40
4,07 0,93 2,63 2,95
0,55 0,43 0,37 0,43
4,11 2,33 4,66 6,98
Fig. 7 shows a superposition of the mean values of the resulting flotation tests. It is graphically observed that the FCU test with ultrasound during conditioning and flotation improved the recovery presenting a faster kinetic response than standard flotation. Regarding the concentrate grade (% Cu), it was found that when ultrasound was applied the concentrate grade improved in relation to standard flotation. As shown in Fig. 8, the best result was obtained by applying ultrasound during the flotation stage only, improving from 0.6% Cu to a final grade of 0.77% Cu. One explanation for these improvements is that fine and ultrafine particles are superficially cleaned by the microjets from the collapse of cavitation bubbles. Cleaning the surface facilitates the action of the collector on the cleaned surface, increasing the percentage of available surface for adhesion of the collector, and so fomenting the action of the reagents. Another hypothesis is that the detected beneficial effect is due to an agglomeration action of ultrasound on the ultrafine particles (6–50 lm) in the areas of highest sound pressure (Fig. 4), which would allow them to acquire enough kinetic energy to emerge together with the bubbles
35%
35%
(a)
(b)
30%
30%
25%
25%
Recovery (%)
Recovery (%)
Identification
20% 15% 10%
15% 10%
Test 1
Test 1 Test 2
5%
Test 2
5%
20%
Mean
Mean
0%
0% 0
5
10
15
0
5
Time (min) 35%
35%
(c)
30% 25% 20% 15% 10%
(d)
25% 20% 15% 10%
Test 1 Test 2
5%
Test 1 Test 2
5%
Mean
0%
0
5
10
Time (min)
15
30%
Recovery (%)
Recovery (%)
10
Time (min)
Mean
15
0%
0
5
10
Time (min)
Fig. 6. García Zúñiga setting for the quadruples of (a) SF, (b), UC (c), UF and (d) FCU.
15
94
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
35%
Recovery (%)
30% 25% 20% 15% SF
10%
UC UF
5%
FCU
0%
0
2
4
6
8
10
12
14
(Vargas et al., 2006). Additionally, the effect of the ultrasound could stem from the generation of sufficiently small bubbles leading to a better particle bubble and particle size relation than the standard flotation without ultrasound, improving the flotation of fine particles. An analysis of some particle tailings has been conducted to determine if any surface cleaning ocurred. This analysis was undertaken via electronic scanning by SEM (Scanning Electron Microscopy), obtaining images presented in Fig. 9. In most cases a decrease of ultrafine particles is observed on the surface of the particles subjected to ultrasound, which would corroborate the hypothesis of the cleaning effect of ultrasound and would improve the efficiency of the reagents.
Time (min) Fig. 7. Average recovery kinetics according to type of ultrasound treatment.
0.90%
Cumulave Grade (%)
0.85% 0.80% 0.75% 0.70% 0.65% 0.60% 0.55%
SF UC UF FCU
0.50% 0.45% 0.40%
0
2
4
6
8
10
12
14
Time (min) Fig. 8. Concentrate grade as a function of flotation time for different ultrasound treatments.
4. Conclusions Based on the experimental results, ultrasound could be a complementary technology to process tailings in classical flotation processes given its potential to improve the recovery of the fine and ultrafine particles. For a standard copper concentrator overall recovery is around 90%, therefore 10% of the feed Cu is lost in the tailings and could be potentially recovered. As shown from the current results, if a 30% recovery is achieved from this 10% copper lost in the tailings, it could mean an increase of about 3.1 percentage points in the overall recovery, achieving a total of 93.1% Cu recovery. Therefore the use of ultrasound during conditioning and flotation of fines and ultrafines deserves further consideration although the economical incentive to apply this technology will depend on the balance between the marginal benefit obtained with the extra Copper production and the investment cost need to apply an equivalent ultrasound treatment to high volumes. Given the current cost of the high frequency ultrasound
Fig. 9. Images of concentrate chalcopyrite particles resulting in the concentrate after flotation from SF and FCU experiments (magnification 6.000). Above: SF concentrated particles. Below: FCU concentrated particles.
A.R. Videla et al. / Minerals Engineering 99 (2016) 89–95
equipment there is still significant work to be done to make the technology attractive. It is believed that improvements are mainly explained by the phenomenon of surface cleaning via fluid cavitation produced by ultrasonic waves and bubble jets, resulting in cleaning the particle surface and improving the functioning of the collector. However, more research is required to understand and properly control the applied pressure field and the effects on the flotation of ultrafine particles. It would be particularly convenient to make further progress in understanding the potential action of agglomeration and deagglomeration. of the ultrafine particles via ultrasound. Other possible effects to be considered is the ultrasound effect on the liquid. It was observed that the energy density heats up the slurry in some degrees, which potentially leads to changes in the chemical equilibrium in solution, which may have an effect in pH, redox potential, z–potential and contact angle. According to the results shown, ultrasound applied at any stage of the process improves final recovery and kinetics of the flotation process of fine and ultrafine copper sulfide particles under evaluation. This improvement is above measurement errors of the tests and represents significant increase in recovery compared to the recovery obtained by standard flotation. Importantly, the latter feature may allow reduction in the volume of reactors needed to reach a given throughput level for the same Cu recovery, allowing reduction of facilities lay-out and/or increasing production capacity with the same equipment. Fig. 4 demonstrates that exposure of the tailings slurry to the ultrasonic field is not uniform inside the cell. Design should be optimized and controlled with an optimal distribution of the transducers in the reactor’s geometry. Another alternative to improve acoustic field uniformity would be the redesign of the cell so that the ultrasound effects are maximized, searching resonance conditions. In any case, transducer interference should be avoided as it is destructive and lowers the efficiency of the process. This aspect of the research and development program will be addressed in a subsequent investigation. The presented research work has been limited to evaluation of ultrasound impact on the particle solid surface. However ultrasound irradiation impacts on solid-liquid interfaces and liquid-air interfaces changing for instance the contact angle. Some of our preliminary studies have shown that a significant change in contact angle can be achieved by the solid surface vibration lead by an ultrasound field. These impacts require further attention and should be addressed in future research.
References Aldrich, C., Feng, D., 1999. Effect of ultrasonic preconditioning of pulp on the flotation of sulphide ores. Miner. Eng. 12 (6), 701–707. http://dx.doi.org/ 10.1016/S0892-6875(99)00053-9.
95
Altun, N.E., Hwang, J.-Y., Hicyilmaz, C., 2009. Enhancement of flotation performance of oil shale cleaning by ultrasonic treatment. Int. J. Miner. Process. 91 (1–2), 1–13. Ambedkar, B., Nagarajan, R., Jayanti, S., 2011. Investigation of high-frequency, highintensity ultrasonics for size reduction and washing of coal in aqueous medium. Ind. Eng. Chem. Res. 50 (23), 13210–13219. Celik, M.S., 1989. Effect of ultrasonic treatment on the floatability of coal and galena. Sep. Sci. Technol. 24, 1159–1166. Chile, Ministerio de Minería, Corporación Chilena del Cobre, 2015. Anuario de Estadísticas del Cobre y otros Minerales 1995–2014. Recuperado de.