- Email: [email protected]
The segregation behaviors of fine coal particles in a coal beneficiation fluidized bed
Fuel Processing Technology 124 (2014) 28–34
Contents lists available at ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
The segregation behaviors of fine coal particles in a coal beneficiation fluidized bed Qinggong Wang a, Weidi Yin a, Bin Zhao a,b, Hairui Yang a, Junfu Lu a,⁎, Lubin Wei c a b c
Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing, 100084, China College of Metallurgy and Energy, Hebei United University, Tangshan 063009, China School of Chemical and Environmental Engineering, China University of Mining and Technology, Beijing 100083, China
a r t i c l e
i n f o
Article history: Received 25 October 2013 Received in revised form 13 February 2014 Accepted 18 February 2014 Available online xxxx Keywords: Coal beneficiation Fluidized bed Fine coal particles Segregation
a b s t r a c t The segregation behaviors of fine coal particles in a coal beneficiation fluidized bed (CBFB) were investigated in this work. The size range of 1–8 mm was taken into account and three separate size fractions were studied comparatively in the experiments, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. Both a clean coal sample and a gangue sample were used as the processed material to study the segregation behaviors of both light particles and heavy particles. The dense bed was divided as seven layers from bottom to top and the particle distribution in each layer for each sample was fully demonstrated. The influences of the particle density, particle size and the fluidized air velocity were revealed, the segregation patterns under different conditions were compared and the segregation mechanism was carefully analyzed. The results showed that the flotation and sedimentation of the particles in CBFB were still largely influenced by the particle density for the fine size range particles, and density stratification occurred even within each size fraction sample. The weight fraction in each layer showed a quadratic increase along the bed height for the coal particles. For gangue particles, a large fraction deposited in the bottom while the mass proportions in the middle layers also showed an increased tendency. With a decrease of the particle size, both the particle segregation and the density stratification phenomena deteriorated seriously. It was proved that particle feed size should be above 3 mm as the separation effect was quite inefficient for finer particles. By increasing the fluidized air velocity, the bed density slightly decreased but the bed turbulence was largely strengthened by the increasing bubble boiling effect. The flotation and sedimentation of the particles in 5.5–8 mm were obviously affected while no clear influence occurred to the rest of the two size fractions. Moreover, the results in this work provide a group of data that are quite suitable for CBFB numerical modeling studies. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Coal beneficiation fluidized bed (CBFB) is a new dry technology for coal separation developed in recent decades for coal upgrading and/or size separation. In operation, a low velocity bubbling fluidization regime is formed in the bed; thus CBFB is one kind of typically dense gas–solid fluidized system characterized by micro-bubbles and high bed densities [1,2]. Under this condition, the feed coal can be segregated as flotsam (refers to the particles sank to the bed bottom) and jetsam (particles rise to the top half of the bed) by its physical properties, such as particle density, size and shapes. The basic concepts, developments as well as separation performances of CBFB have been widely reviewed in literatures [3–8]. The segregation behaviors of the coarse coal particles in CBFB are largely determined by the particle density, and the particle motion is controlled by the balance between the gravity and the effective ⁎ Corresponding author at: Department of Thermal Engineering, Tsinghua University, Qing Hua Yuan No.1, Haidian District, Beijing, 100084, China. Tel.: +8610-62792647. E-mail address: [email protected] (J. Lu).
http://dx.doi.org/10.1016/j.fuproc.2014.02.015 0378-3820/© 2014 Elsevier B.V. All rights reserved.
hydrostatic buoyancy, whereas the Archimedes' principle can be roughly used to explain the separation mechanism [9]. But for the fine particles, interactions between the coal particles and the bed medium as well as the interactions between the coal particles and the bubble phase become relatively significant, thus the separation mechanism is more complicated [10]. At present, the particle size range that can be effectively processed in industrial CBFB technology is reported to be 6– 50 mm while finer particles are difficult to deal with [11].Therefore, the influence of coal feed size on the performance of CBFB has gained interest in laboratory researches [12–14]. However, the segregation behaviors and particle distribution patterns in CBFB have not been fully revealed as the bed is traditionally divided as two layers or at most three layers to discuss the overall property of the product stream and the refused stream, which is very useful for the industrial production, but quite not enough to embody the particle behaviors in the academic respect. Besides, the available literatures conform to the fact that the separation efficiency in CBFB decreases with coal feeding size [12,15], but the detailed segregation characteristics for fine particles are not fully demonstrated and a clear proof of the lower feed size limit is not proposed and still remains a challenge.
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
2. Experimental setup The experimental setup of the CBFB used in this study is schematically shown in Fig. 1. The body of the rectangular fluidized bed was made of Plexiglas with a cross-sectional area of 0.5 × 0.05 m2, and a height of 1.0 m. The fluidized air was supplied by a fan blower and sent to the bottom of the fluidized bed through an air distributor. The air flow rate was measured with romameters and could be regulated by valves to form the minimum fluidization condition and the designed working conditions for coal beneficiation. Even with a porosity rate of 2% and a thickness of 20 mm, the pressure drop of the air distributor was not enough to ensure the fluidization uniformity in the bed as the length–breadth ratio of the bed cross-section was large. To make compensation, a sand layer of 15 mm with sands of 2–3 mm size naturally packed was set under the air distributor to increase the pressure drop.
3000
Plexiglass Air Distributor: porosity rate =2% hole diamter = 2mm thickness = 15mm Sand Layer: sand size = 2-3mm thickness = 15mm
2500
Pressure drop (Pa)
More important, the computational fluid dynamics (CFD) has become a useful and promising method to study the gas–solid flow in CBFB technology [16–18]. However, the widely existing experimental data in the references, which are the bases for validating the CFD model, mainly concentrate on the performance of the apparatus and the assessments of separation efficiency by the ècart probable moyen (Ep) value (calculated from the distribution curve/partition curve in which the percent of feed reporting to clean coal is plotted against specific gravity). Those expressions reflect the coal washability and separation quality by the dry CBFB methods, but the detailed particle distributions in CBFB are not fully provided and the information is not appropriate for CFD modeling [19–21]. Besides, confined by the mesh size and accuracy of the numerical modeling, the particle size modeled is only restricted to fine ranges. Therefore, in this study, the segregation behaviors of fine coal particles of 1–8 mm in CBFB were carefully studied to reveal the particle distribution characteristics, to demonstrate the lower feeding size limit and to provide some available data for numerical modeling. Three size ranges were taken into account separately in the experiments, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. Both a clean coal sample and a gangue sample were used as the processed objects to comparatively study the segregation behaviors of the light particles and the heavy particles. The influences of the particle density, particle size and the fluidized air velocity were mainly concerned. The dense bed was divided as seven layers from bottom to top and the distribution characteristics in each layer for each sample were clearly demonstrated.
29
2000
1500
1000
500
0 0.03
0.06
0.09
0.12
0.15
0.18
Air Velocity (m/s) Fig. 2. The pressure drop of the air distribution layers.
By this way, a satisfactorily uniform fluidization state was obtained in the working conditions. The whole pressure drop of the air distribution layers at different air velocity is shown in Fig. 2. To monitor the pressure distribution in the bed, a group of pressure taps were set along two columns both on the front and the back walls as the bed is horizontally wide. The pressure drops were measured by a group of U-tube water manometers. Based on the measured bed pressure drop between two measured points at different heights, the bed porosity at the referred average height was calculated and then the average bed density was obtained. One kind of Geldart B magnetite powder was used as the bed medium in the measurement. It had a real density of 4181 kg/m3 and a size distribution of 10–600 μm with a mean particle diameter of 204 μm. The detailed particle size distribution is shown in Fig. 3. The minimum fluidization velocity umf of the magnetite powder was 0.067 m/s, which was measured according to the pressure drop of the bed with a gradual increase of airflow rate. Both a clean coal sample and a refused gangue sample with apparent density difference were considered in the experiment to reflect particle motion and distributions of both the light product and the heavy product in CBFB beneficiation. Three fine size fractions were selected for each sample, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. The average particle density for each size fraction was tested and reported in Table 1. The gangue sample had a density twice of the coal sample and little difference was observed for different size fractions within the same sample group. An initial height of 0.3 m was filled with magnetite powder in each operation. The weighted feed coal and/or gangue sample was then 100
Cumulative (%)
80
60
d43 = 204 µ m d32 = 131 µ m
40
umf = 0.067m/s
20
0
0
100
200
300
400
500
600
Diameter (micron) Fig. 1. Schematic drawing of the experimental system of the coal beneficiation fluidized bed. (1) Fluidized bed; (2) air box; (3) air distributor; (4) sand layer; (5) fan blower; (6) rotameter; (7) U-tube water manometer group.
Fig. 3. The particle size distribution of the bed medium (magnetite powder) used in the experiment.
30
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
a) 50%
Table 1 The average density of the samples.
1–2 mm
1371 2676
1384 2653
1397 2654
added into the bed. To minimize the interactions between fed particles themselves, while also ensuring the statistical reliability in data collection, 1 kg of feed sample in total (about 5% in weight fraction) was used in each operation. The fluidization was started and lasted for 15 min in steady state. The supply air was then cut off suddenly and the bed fell down immediately to be static. After the feed sample was added in the bed, the final bed height was usually higher than 0.3 m and the bed was then divided into seven layers from the top to the bottom. The bottom six ones were 0.05 m for each and the top one referred to the material above the level of 0.3 m. The lower part of the bed body was in layered structure, so the material in each layer was easily collected by spades from the top to the bottom, by removing each upper layer box after the material was strictly extracted. The feeding sample in each layer was sieved out and the mass proportion was measured. The particle size distribution was obtained by classifying the particles with a group of standard screens and the particle density was measured by a true volume and density measurement instrument. The testing of particle density was repeated three times for each sample to obtain the average value accurately. All the samples were in air dried state beforehand; thus the influence of moist content can be neglected and the experiments were conducted at room temperature. 3. Results and discussion 3.1. Influence of particle density Generally, the utilization of CBFB for coal preparation is gravitybased and the distribution of the feed coal particles is largely influenced by the particle density and its deviation to the mean bed density. For fine coal particles, the particle motion in CBFB becomes complicated; thus the influence of particle density should be studied separately by controlling the other factors the same. Under the fluidized air velocity of 1.5umf for the bed medium, the coal particles and the gangue particles were beneficiated in the CBFB separately first, and each size fraction was considered individually in each operation. Fig. 4 shows the distributions of the weight fraction and the density variation for each coal sample along the bed height. It is clear that the light clean coal particles mainly flow to the surface layers of the bed and the weight proportion decreases along the bed height. The fitting curves of the weight proportion are in quadratic functions. As the top layer (the 7th layer) is usually shallow than the next normal layers, the top fraction shows a decrease, and it has been plotted separately from the normal layers in the figures. At the same time, the slope of the particle distribution curve is obviously larger for the coarse particle sample than the fine one, which means that the floating/beneficiation effect of the pseudofluid bed is stronger for the coarse particles. As seen in Fig. 4b, the density stratification occurs even within the same size fraction sample. For example, for the coal of 5.5–8 mm, the average particle density in the bottom layers is larger than the top layers and an apparent decease tendency along the bed can be observed. Even for the finest sample of 1–2 mm, the density stratification can also be detected. For the gangue particles as shown in Fig. 5, the distribution of the heavy gangue shows an interesting pattern. For the size fractions of 5.5–8 mm and 3–5.5 mm, most gangue particles sank to the bed bottom and the deepest layer (the 1st layer) has a very high mass proportion. The distribution of the rest of the particles from layer 2 to layer 6 shows an increased tendency overall. But for the size fraction of 1–2 mm, the particle's sedimentation effect is not obvious. The 1st
40%
Proportion (wt%)
3–5.5 mm
30%
20%
10%
0%
1
2
3
4
5
6
7
Height (Layer)
b) Particle Density (g/cm3)
Coal Gangue
5.5–8 mm
5.5-8 mm 3-5.5 mm 1-2 mm
Single Sample 100% Coal
True density (kg/m3)
1.5
Single Sample 100% Coal
5.5-8 mm 3-5.5 mm 1-2 mm
1.45
1.4
1.35
1.3
1
2
3
4
5
6
7
Height (Layer) Fig. 4. The distribution of coal particles and density variations along the bed height. (a) Particle distribution; (b) density variation.
layer shows a much lower proportion and a whole tendency can be derived from the 1st layer to the 6th layer. Due to the same reason as above, the 7th layer has low proportions. The bed expansion is diminished as a less volume of the same weight was added for the heavier gangue particles. The particle density stratification is more obvious for gangue samples. The density varies from 2.5 g/cm3 to 2. 75 g/cm3 for the coarse samples while still not significant for the finest one. In addition, a binary mixture of half coal sample and half gangue was beneficiated in the CBFB and the integral density effect was demonstrated. Each size fraction was testified separately and the results are shown in Fig. 6. As seen from the particle mass distribution, a similar tendency but a combined effect of the single sample results as shown in Figs. 4 and 5 can be observed. The proportion in the bottom layer decreases when compared with the pure gangue sample while the increase tendency in the middle layers is flattened when compared with the pure coal sample. However, for the binary mixtures, the effect of particle density distribution along the bed height can be more clearly observed. The average particle density at the very bottom reaches to 2.7 g/cm3 where gangue particles mainly exist while the top layers are most coal particles. The fraction of coal or gangue sample in each layer can be estimated from the average particle density and the pure sample densities. Again, the finest size fraction of 1–2 mm is not obviously stratified both in the mass distribution and the density distribution even for two samples with a quite large density difference. As proved above, the distribution of fine coal particles was still obviously influenced by the particle density. The density ratio between the particle density and the mean bed density is an important factor to determine the particle segregation [22–24]. For the coal samples, the density ratio is about 0.7, the large particles are inclined to be swept up to the bed surface but still could occasionally descend in the bed as
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
a) 50%
50%
Single Sample 100% Gangue
Proportion (wt%)
40%
5.5-8 mm 3-5.5 mm 1-2 mm
30%
20%
10%
0%
5.5-8 mm 3-5.5 mm 1-2 mm
Binary Mixture 50% Coal + 50% Gangue
40%
Proportion (wt%)
a)
31
30%
20%
10%
1
2
3
4
5
6
7
0%
Height (Layer)
1
2
3
4
5
6
7
Height (Layer)
b) 2.8
5.5-8 mm 3-5.5 mm 1-2 mm
2.7
2.6
Single Sample 100% Gangue
2.5
5.5-8 mm 3-5.5 mm 1-2 mm
Binary Mixture 50%Coal+50%Gangue
2.6
Particle Density (g/cm3)
Particle Density (g/cm3)
b) 2.8
2.4 2.2 2 1.8 1.6
2.4
1
2
3
4
5
6
7
1.4
Height (Layer)
1
2
3
4
5
6
7
Height (Layer) Fig. 5. The distribution of gangue particles and density variations along the bed height. (a) Particle distribution; (b) density variation.
the rising bubbles are pushing it back. For the gangue particles, the ratio ranges from 1.23 to 1.36. When the ratio is not large enough, the swarms of bubbles exert a considerable lifting effort to draw the particles up; as a result, the particle proportions from layer 2 to layer 6 show an increased tendency as in Figs. 4 to 6. For a density ratio great enough (more than 1.28 as reported by Kunii and Levenspiel [25] for coarse particles), the particles fall down to the bed bottom and stay there permanently. As there is a general correlation between the ash content and particle density [26], the density stratification can be effectively used to remove the high ash content gangue in the raw coal to upgrade the low rank coal samples in the fine particle size range. However, the segregation behavior is also largely affected by the particle size. When the particles are as fine as 1–2 mm, the density stratification in the dry CBFB is negligible and the separation can only resort to some other methods such as electric technology [27] or froth flotation [28].
3.2. Influence of particle size As partly demonstrated in the former section, the particle segregation behavior in CBFB is a strong function of the particle size. As seen from Figs. 4 to 6, the beneficiation efficiency deteriorates seriously as the particle size decreases. To further study the influence of particle size and the combination effects of different sizes, a ternary mixture with one third of each size fraction of coal samples (or gangue samples) was beneficiated in the CBFB at the fluidized air velocity of 1.5umf for the bed medium. The particle density variations were neglected and the allocations of size fraction in each layer were mainly concerned.
Fig. 6. The distribution of the binary mixture and average density variations along the bed height. (a) Particle distribution; (b) density variation.
As seen in Fig. 7, the coal particles of 5.5–8 mm and 3–5.5 mm show quite similar mass distributions along the bed height while the fraction of 1–2 mm shows a weaker flotation effect. For the gangue particles, the size fraction of 1–2 mm shows a weaker sedimentation effect as the weight proportion in the bottom layer is quite small and the increase tendency along the bed height is the most significant. When particles are small enough, their movements are mainly dominated by surface drag and less dependent on the gravity force and hydrostatic buoyancy. Consequently, the misplacing effect is substantial [29–31]. The back mixing effect in the bubbling CBFB has such a strong influence to the 1–2 mm particles that the light coal and heavy gangue show convergent tendencies, which brings the difficulty to distinguish and separate them apart in the raw coal. According to Tanaka et al. [32], the possibility of the particles segregation/mixing in a fluidized bed is dependent on the difference in size and density between the beneficiated particles and the medium particles. The segregation condition can be roughly formulated by Eqs. (1) and (2) as follows, where ρ and d refer to the particle density and diameter, respectively, and εmf is the void fraction at the minimum fluidization state. When the particle properties fall into the region, the two species are mixable and hard to be segregated in fluidization. ð1−ε mf Þ b
0:5 b
ρbeneficated‐particle 1 b ρmedium‐particle ð1−εmf Þ
dbeneficiated‐particle dmedium‐particle
!
ρbeneficiated‐particle ρmedium‐particle
ð1Þ !2 b2
ð2Þ
32
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
forces such as the vibrating force or the magnetic field are suggested to be introduced to improve the beneficiation quality [10,27]. Due to these reasons, the feed particle size is suggested to be more than 3 mm under the working conditions and with the magnetite particles used in this study.
a) 50%
Ternary Mixture Coal Samples 33 wt% for each size
Proportion (wt%)
40%
5.5-8 mm 3-5.5 mm 1-2 mm
3.3. Influence of fluidized air velocity
30%
20%
10%
0%
1
2
3
4
5
6
7
6
7
Height (Layer)
b)
50%
5.5-8 mm 3-5.5 mm 1-2 mm
Proportion (wt%)
40%
Ternary Mixture Gangue Samples 33 wt% for each size
30%
20%
10%
0% 1
2
3
4
5
Height (Layer) Fig. 7. The distribution of the samples' proportion along the bed height (ternary mixture: 33 wt.% for each size fraction). (a) Coal samples; (b) gangue samples.
The segregation possibility for the particles and conditions used in this work is shown in Fig. 8. For the coarse fractions of 5.5–8 mm and 3–5.5 mm, the coal particles and gangue particles are extractive from the bed medium, whereas the segregation phenomenon is obvious as proved from Figs. 4 to 7. For the finest size fraction of 1–2 mm, the particles in this work fall into the mixable region by a large section, which indicates that the beneficiated particles and the medium particles are hard to segregate in the fluidized bed normally. Some external
Although the characteristics of the pseudofluid state in CBFB are used to beneficiate coal particles, the separation mechanism is not exactly the same as the conventional wet dense medium separator. The stability of the CBFB and the beneficiation quality are directly influenced by the fluidized air velocity, which will determine the average bed density on the one hand, and alter the behaviors of the bubble phase on the other. The bed density can reach the similar values required as in the wet process by adjusting the property of the bed material and the air velocity, but the viscosity of the pseudofluid phase is quite lower [2,33]. In this work, a group of different fluidized air velocities were tested, ranging from 1.2umf to 1.8umf of the bed medium, and particles of different samples and size fractions were beneficiated singly. With an increase of the fluidized air velocity, the averaging bed density decreases slightly in linear as seen in Fig. 9. Even when the average bed density is not obviously changing, the bubble size is greatly increasing when the air velocity is enhancing and the fluidization is in more violently boiling. The particle distributions under different air velocities are comparatively shown in Fig. 10 for different samples and size fractions. It can be seen that the fluidized air velocity has an obvious influence on the particles of 5.5–8 mm both for coal samples and gangue samples while it has little effect to the size fractions of 3–5.5 mm and 1–2 mm. When the fluidization number increases, the weight proportions in the top layers decrease while more fractions move to the lower layers for the light coal samples of 5.5–8 mm, which means that the flotation effect deteriorates to some extent. The intensive backing mixing effect is mainly attributed to the strengthened bubble boiling. As the bubbles rising in the emulsion phase are not spherical but always have a concave base, the lower pressure in this wake region would draw up the solid particles, while solids in the emulsion phase will move downward to fill the caused vacancies [12,34]. The light coal particles are more easily affected by the circulation mixing effect, and the big particles are more sensitive to the bubble size and the turbulence intensity according to the findings in Fig. 10. For the gangue particles, the fraction in the bottom increases apparently when the fluidized air velocity increases. As the particle density of gangue is larger than the averaging bed density, the bubbling mixing strengthens the bed disturbance and accelerates 2.1
3-5.5 mm
Density Ratio
5.5-8 mm
1-2 mm Eq. (1) Eq. (2)
10
1
mixable
Averagiong bed density (g/cm3)
100
2.05
2
1.95
1.9 0.1 0.1
1
10
100
umf =0.067m/s
1
1.2
1.4
1.6
1.8
2
2.2
Fluidization number
Size Ratio Fig. 8. Segregation possibility of the particles in the CBFB.
Fig. 9. The averaging bed density under different fluidization number. Fluidization number = uoperation/umf.
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
b) 60%
50%
40%
Proportion (wt%)
1.2 umf 1.5 umf 1.8 umf
Single Sample Coal 5.5-8mm
30%
20%
10%
40% 30% 20% 10% 0%
0% 1
2
3
4
5
6
1
7
2
3
Height (Layer)
1.2 umf 1.5 umf 1.8 umf
Single Sample Coal 3-5.5mm
6
7
30%
20%
10%
6
7
6
7
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 3-5.5mm
50%
Proportion (wt%)
Proportion (wt%)
5
60%
40%
40% 30% 20% 10%
1
2
3
4
5
6
0%
7
1
2
3
Height (Layer)
4
5
Height (Layer)
50%
60%
1.2 umf 1.5 umf 1.8 umf
30%
20%
10%
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 1-2mm
50%
Proportion (wt%)
Single Sample Coal 1-2mm
40%
Proportion (wt%)
4
Height (Layer)
50%
0%
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 5.5-8mm
50%
Proportion (wt%)
a)
33
40% 30% 20% 10%
0% 1
2
3
4
5
6
7
0%
1
Height (Layer)
2
3
4
5
Height (Layer)
Fig. 10. The distribution of the samples' proportion along the bed height under different fluidized velocities. (a) Coal samples; (b) gangue samples.
the sedimentation of heavy particles. Still, the effect is more obvious for the 5.5–8 mm size fraction and not significant for the rest of the fine ones. The fluidization of the beneficiated particles is facilitated by the viscous drag of the pseudofluid medium. The segregation behaviors of particles with different size and/or density are directly affected by the changing of the bed density and viscosity. As the fine particles have a lower minimum fluidization velocity in the dense medium, they are better mixed and their flotation/sedimentation is less influenced by the variation of the air velocity. Due to this reason, the aggregative fluidization regime is preferred in CBFB to ensure the stability of the bed for higher separation efficiencies. However, in industrial operation,
the air velocity is always set as high as possible to make the bubbling fluidization more uniform on one hand, and to speed up the stratification on the other.
4. Conclusions The particle distribution characteristics and the segregation behaviors of fine coal/gangue particles in a coal beneficiation fluidized bed (CBFB), which is still not fully clear to date, were revealed in detail in this work. Both clean coal samples and gangue samples of 1–8 mm in three separate size fractions were beneficiated purposely. The dense
34
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
bed was divided into seven layers and the segregation results were carefully analyzed. The results obtained can be summarized as follows: (1) The flotation and sedimentation of the particles in CBFB were still largely influenced by the particle density for the fine size range particles, and density stratification occurred even within each size fraction sample. The mass distribution showed a quadratic increase along the bed height for the light coal particles. For gangue particles, the mass proportions in the middle layers also showed an increased tendency surprisingly with a large fraction deposited in the bottom. (2) Both the particle segregation and the density stratification deteriorated seriously as the particle size decreased. The lower limit of the particle feed size was proved to be more than 3 mm and the beneficiation effect for finer particles was inefficient. (3) The increase of fluidized air velocity slightly decreased the averaging bed density while apparently strengthened the bubble boiling, in which the segregation efficiency of the 5.5–8 mm particles was obviously affected by the back mixing effect. Moreover, among the massive experimental data available, there is still not such a group that can be effectively used for numerical modeling for an academic purpose. The narrow ranges of particle density and size and the full information of particle distributions in CBFB of this work bring the convenience for CFD modeling validation, which, with no doubt, has arisen as a new effective tool in the coal beneficiation field. Acknowledgements Financial support of this work by the National Program on Key Basic Research Project (973 Program) of China (No. 2012CB214900) is gratefully acknowledged. The authors also gratefully acknowledge Dr. Xuemin Liu at the Department of Thermal Engineering, Tsinghua University for the preparation of the coal samples, and Xinhua Yang for the help in particle density measurements. References [1] Q. Chen, L. Wei, Development of coal dry beneficiation with air-dense medium fluidized bed in China, China Particuology 3 (2005) 42. [2] Y.M. Zhao, L.B. Wei, Rheology of gas–solid fluidized bed, Fuel Processing Technology 68 (2000) 153–160. [3] N.C. Lockhart, Dry beneficiation of coal, Powder Technology 40 (1984) 17–42. [4] J.A. van Houwelingen, T.P.R. de Jong, Dry cleaning of coal: review, fundamentals and opportunities, Geologica Belgica 7 (2004) 335–343. [5] R.K. Dwari, K.H. Rao, Dry beneficiation of coal—a review, Mineral Processing and Extractive Metallurgy Review 28 (2007) 177–234. [6] H. Katalambula, R. Gupta, Low-grade coals: a review of some prospective upgrading technologies, Energy and Fuels 23 (2009) 3392–3405. [7] A.K. Sahu, S.K. Biswal, A. Parida, Development of air dense medium fluidized bed technology for dry beneficiation of coal—a review, International Journal of Coal Preparation and Utilization 29 (2009) 216–241.
[8] S. Mohanta, C.S. Rao, A.B. Daram, S. Chakraborty, B.C. Meikap, Air dense medium fluidized bed for dry beneficiation of coal: technological challenges for future, Particulate Science and Technology 31 (2013) 16–27. [9] M.G. Rasul, V. Rudolph, F.Y. Wang, Particles separation using fluidization techniques, International Journal of Mineral Processing 60 (2000) 163–179. [10] X. Yang, Y. Zhao, Z. Luo, S. Song, C. Duan, L. Dong, Fine coal dry cleaning using a vibrated gas-fluidized bed, Fuel Processing Technology 106 (2013) 338–343. [11] Z. Luo, Q. Chen, Dry beneficiation technology of coal with an air dense-medium fluidized bed, International Journal of Mineral Processing 63 (2001) 167–175. [12] S. Mohanta, S. Chakraborty, B.C. Meikap, Influence of coal feed size on the performance of air dense medium fluidized bed separator used for coal beneficiation, Industrial and Engineering Chemistry Research 50 (2011) 10865–10871. [13] O. Prakash, D.D. Misra, Correlation of size and effective separation density in coal washing using heavy media, Journal of Mines, Metals and Fuels 27 (1979) 138–147. [14] J. Choung, C. Mak, Z. Xu, Fine coal beneficiation using an air dense medium fluidized bed, Coal Preparation 26 (2006) 1–15. [15] Z. Luo, Y. Zhao, Q. Chen, X. Tao, M. Fan, Separation lower limit in a magnetically gas–solid two-phase fluidized bed, Fuel Processing Technology 85 (2003) 173–178. [16] S. Azizi, S.H. Hosseini, G. Ahmadi, M. Moraveji, Numerical simulation of particle segregation in bubbling gas-fluidized beds, Chemical Engineering & Technology 33 (2010) 421–432. [17] Y. Zhao, L. Tang, Z. Luo, C. Liang, H. Xing, W. Wu, C. Duan, Experimental and numerical simulation studies of the fluidization characteristics of a separating gas–solid fluidized bed, Fuel Processing Technology 91 (2010) 1819–1825. [18] Q. Wang, J. Lu, W. Yin, H. Yang, L. Wei, Numerical study of gas–solid flow in a coal beneficiation fluidized bed using kinetic theory of granular flow, Fuel Processing Technology 111 (2013) 29–41. [19] Y.Q. Feng, A.B. Yu, Microdynamic modelling and analysis of the mixing and segregation of binary mixtures of particles in gas fluidization, Chemical Engineering Science 62 (2007) 256–268. [20] H. Lu, Y. He, D. Gidaspow, L. Yang, Size segregation of binary mixture of solids in bubbling fluidized beds, Powder Technology 134 (2003) 86–97. [21] P.K. Naik, P.S.R. Reddy, V.N. Misra, Optimization of coal flotation using statistical technique, Fuel Processing Technology 85 (2004) 1473–1485. [22] F. Garcia, A. Romero, J. Villar, A. Bello, A study of segregation in a gas solid fluidized bed: particles of different density, Powder Technology 58 (1989) 169–174. [23] A. Nienow, P. Rowe, L. Cheung, A quantitative analysis of the mixing of two segregating powders of different density in a gas-fluidised bed, Powder Technology 20 (1978) 89–97. [24] A. Rao, J. Curtis, B. Hancock, C. Wassgren, Classifying the fluidization and segregation behavior of binary mixtures using particle size and density ratios, AIChE Journal 57 (2011) 1446–1458. [25] D. Kunii, O. Levenspiel, Fluidization Engineering, 2nd ed. Butterworth-Heinemann, Newton, MA, 2005. [26] C. Zhang, G. Chen, T. Yang, G. Lu, C. Mak, D. Kelly, Z. Xu, An investigation on mercury association in an Alberta sub-bituminous coal, Energy and Fuels 21 (2007) 485–490. [27] Z. Luo, Y. Zhao, Q. Chen, M. Fan, X. Tao, Separation characteristics for fine coal of the magnetically fluidized bed, Fuel Processing Technology 79 (2002) 63–69. [28] A. Guney, G. Onal, O. Ergut, Beneficiation of fine coal by using the free jet flotation system, Fuel Processing Technology 75 (2002) 141–150. [29] Y. He, H. Chen, Y. Ding, B. Lickiss, Solids motion and segregation of binary mixtures in a rotating drum mixer, Chemical Engineering Research and Design 85 (2007) 963–973. [30] A. Nienow, N. Naimer, T. Chiba, Studies of segregation/mixing in fluidised beds of different size particles, Chemical Engineering Communications 62 (1987) 53–66. [31] Y. Soong, T.A. Link, M.R. Schoffstall, M.L. Gray, D.J. Fauth, J.P. Knoer, J.R. Jones, I.K. Gamwo, Dry beneficiation of Slovakian coal, Fuel Processing Technology 72 (2001) 185–198. [32] Z. Tanaka, X. Song, Continuous separation of particles by fluidized beds, Advanced Powder Technology 7 (1996) 29–40. [33] R. Sahan, B. Kozanoglu, Use of an air fluidized bed separator in a dry coal cleaning process, Energy Conversion and Management 38 (1997) 269–286. [34] S. Mohanta, S. Chakraborty, B.C. Meikap, Optimization process of an air dense medium fluidized bed separator for treating high-ash non-coking Indian coal, Mineral Processing and Extractive Metallurgy Review 34 (2013) 240–248.
Contents lists available at ScienceDirect
Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc
The segregation behaviors of fine coal particles in a coal beneficiation fluidized bed Qinggong Wang a, Weidi Yin a, Bin Zhao a,b, Hairui Yang a, Junfu Lu a,⁎, Lubin Wei c a b c
Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing, 100084, China College of Metallurgy and Energy, Hebei United University, Tangshan 063009, China School of Chemical and Environmental Engineering, China University of Mining and Technology, Beijing 100083, China
a r t i c l e
i n f o
Article history: Received 25 October 2013 Received in revised form 13 February 2014 Accepted 18 February 2014 Available online xxxx Keywords: Coal beneficiation Fluidized bed Fine coal particles Segregation
a b s t r a c t The segregation behaviors of fine coal particles in a coal beneficiation fluidized bed (CBFB) were investigated in this work. The size range of 1–8 mm was taken into account and three separate size fractions were studied comparatively in the experiments, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. Both a clean coal sample and a gangue sample were used as the processed material to study the segregation behaviors of both light particles and heavy particles. The dense bed was divided as seven layers from bottom to top and the particle distribution in each layer for each sample was fully demonstrated. The influences of the particle density, particle size and the fluidized air velocity were revealed, the segregation patterns under different conditions were compared and the segregation mechanism was carefully analyzed. The results showed that the flotation and sedimentation of the particles in CBFB were still largely influenced by the particle density for the fine size range particles, and density stratification occurred even within each size fraction sample. The weight fraction in each layer showed a quadratic increase along the bed height for the coal particles. For gangue particles, a large fraction deposited in the bottom while the mass proportions in the middle layers also showed an increased tendency. With a decrease of the particle size, both the particle segregation and the density stratification phenomena deteriorated seriously. It was proved that particle feed size should be above 3 mm as the separation effect was quite inefficient for finer particles. By increasing the fluidized air velocity, the bed density slightly decreased but the bed turbulence was largely strengthened by the increasing bubble boiling effect. The flotation and sedimentation of the particles in 5.5–8 mm were obviously affected while no clear influence occurred to the rest of the two size fractions. Moreover, the results in this work provide a group of data that are quite suitable for CBFB numerical modeling studies. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Coal beneficiation fluidized bed (CBFB) is a new dry technology for coal separation developed in recent decades for coal upgrading and/or size separation. In operation, a low velocity bubbling fluidization regime is formed in the bed; thus CBFB is one kind of typically dense gas–solid fluidized system characterized by micro-bubbles and high bed densities [1,2]. Under this condition, the feed coal can be segregated as flotsam (refers to the particles sank to the bed bottom) and jetsam (particles rise to the top half of the bed) by its physical properties, such as particle density, size and shapes. The basic concepts, developments as well as separation performances of CBFB have been widely reviewed in literatures [3–8]. The segregation behaviors of the coarse coal particles in CBFB are largely determined by the particle density, and the particle motion is controlled by the balance between the gravity and the effective ⁎ Corresponding author at: Department of Thermal Engineering, Tsinghua University, Qing Hua Yuan No.1, Haidian District, Beijing, 100084, China. Tel.: +8610-62792647. E-mail address: [email protected] (J. Lu).
http://dx.doi.org/10.1016/j.fuproc.2014.02.015 0378-3820/© 2014 Elsevier B.V. All rights reserved.
hydrostatic buoyancy, whereas the Archimedes' principle can be roughly used to explain the separation mechanism [9]. But for the fine particles, interactions between the coal particles and the bed medium as well as the interactions between the coal particles and the bubble phase become relatively significant, thus the separation mechanism is more complicated [10]. At present, the particle size range that can be effectively processed in industrial CBFB technology is reported to be 6– 50 mm while finer particles are difficult to deal with [11].Therefore, the influence of coal feed size on the performance of CBFB has gained interest in laboratory researches [12–14]. However, the segregation behaviors and particle distribution patterns in CBFB have not been fully revealed as the bed is traditionally divided as two layers or at most three layers to discuss the overall property of the product stream and the refused stream, which is very useful for the industrial production, but quite not enough to embody the particle behaviors in the academic respect. Besides, the available literatures conform to the fact that the separation efficiency in CBFB decreases with coal feeding size [12,15], but the detailed segregation characteristics for fine particles are not fully demonstrated and a clear proof of the lower feed size limit is not proposed and still remains a challenge.
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
2. Experimental setup The experimental setup of the CBFB used in this study is schematically shown in Fig. 1. The body of the rectangular fluidized bed was made of Plexiglas with a cross-sectional area of 0.5 × 0.05 m2, and a height of 1.0 m. The fluidized air was supplied by a fan blower and sent to the bottom of the fluidized bed through an air distributor. The air flow rate was measured with romameters and could be regulated by valves to form the minimum fluidization condition and the designed working conditions for coal beneficiation. Even with a porosity rate of 2% and a thickness of 20 mm, the pressure drop of the air distributor was not enough to ensure the fluidization uniformity in the bed as the length–breadth ratio of the bed cross-section was large. To make compensation, a sand layer of 15 mm with sands of 2–3 mm size naturally packed was set under the air distributor to increase the pressure drop.
3000
Plexiglass Air Distributor: porosity rate =2% hole diamter = 2mm thickness = 15mm Sand Layer: sand size = 2-3mm thickness = 15mm
2500
Pressure drop (Pa)
More important, the computational fluid dynamics (CFD) has become a useful and promising method to study the gas–solid flow in CBFB technology [16–18]. However, the widely existing experimental data in the references, which are the bases for validating the CFD model, mainly concentrate on the performance of the apparatus and the assessments of separation efficiency by the ècart probable moyen (Ep) value (calculated from the distribution curve/partition curve in which the percent of feed reporting to clean coal is plotted against specific gravity). Those expressions reflect the coal washability and separation quality by the dry CBFB methods, but the detailed particle distributions in CBFB are not fully provided and the information is not appropriate for CFD modeling [19–21]. Besides, confined by the mesh size and accuracy of the numerical modeling, the particle size modeled is only restricted to fine ranges. Therefore, in this study, the segregation behaviors of fine coal particles of 1–8 mm in CBFB were carefully studied to reveal the particle distribution characteristics, to demonstrate the lower feeding size limit and to provide some available data for numerical modeling. Three size ranges were taken into account separately in the experiments, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. Both a clean coal sample and a gangue sample were used as the processed objects to comparatively study the segregation behaviors of the light particles and the heavy particles. The influences of the particle density, particle size and the fluidized air velocity were mainly concerned. The dense bed was divided as seven layers from bottom to top and the distribution characteristics in each layer for each sample were clearly demonstrated.
29
2000
1500
1000
500
0 0.03
0.06
0.09
0.12
0.15
0.18
Air Velocity (m/s) Fig. 2. The pressure drop of the air distribution layers.
By this way, a satisfactorily uniform fluidization state was obtained in the working conditions. The whole pressure drop of the air distribution layers at different air velocity is shown in Fig. 2. To monitor the pressure distribution in the bed, a group of pressure taps were set along two columns both on the front and the back walls as the bed is horizontally wide. The pressure drops were measured by a group of U-tube water manometers. Based on the measured bed pressure drop between two measured points at different heights, the bed porosity at the referred average height was calculated and then the average bed density was obtained. One kind of Geldart B magnetite powder was used as the bed medium in the measurement. It had a real density of 4181 kg/m3 and a size distribution of 10–600 μm with a mean particle diameter of 204 μm. The detailed particle size distribution is shown in Fig. 3. The minimum fluidization velocity umf of the magnetite powder was 0.067 m/s, which was measured according to the pressure drop of the bed with a gradual increase of airflow rate. Both a clean coal sample and a refused gangue sample with apparent density difference were considered in the experiment to reflect particle motion and distributions of both the light product and the heavy product in CBFB beneficiation. Three fine size fractions were selected for each sample, e.g. 1–2 mm, 3–5.5 mm and 5.5–8 mm. The average particle density for each size fraction was tested and reported in Table 1. The gangue sample had a density twice of the coal sample and little difference was observed for different size fractions within the same sample group. An initial height of 0.3 m was filled with magnetite powder in each operation. The weighted feed coal and/or gangue sample was then 100
Cumulative (%)
80
60
d43 = 204 µ m d32 = 131 µ m
40
umf = 0.067m/s
20
0
0
100
200
300
400
500
600
Diameter (micron) Fig. 1. Schematic drawing of the experimental system of the coal beneficiation fluidized bed. (1) Fluidized bed; (2) air box; (3) air distributor; (4) sand layer; (5) fan blower; (6) rotameter; (7) U-tube water manometer group.
Fig. 3. The particle size distribution of the bed medium (magnetite powder) used in the experiment.
30
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
a) 50%
Table 1 The average density of the samples.
1–2 mm
1371 2676
1384 2653
1397 2654
added into the bed. To minimize the interactions between fed particles themselves, while also ensuring the statistical reliability in data collection, 1 kg of feed sample in total (about 5% in weight fraction) was used in each operation. The fluidization was started and lasted for 15 min in steady state. The supply air was then cut off suddenly and the bed fell down immediately to be static. After the feed sample was added in the bed, the final bed height was usually higher than 0.3 m and the bed was then divided into seven layers from the top to the bottom. The bottom six ones were 0.05 m for each and the top one referred to the material above the level of 0.3 m. The lower part of the bed body was in layered structure, so the material in each layer was easily collected by spades from the top to the bottom, by removing each upper layer box after the material was strictly extracted. The feeding sample in each layer was sieved out and the mass proportion was measured. The particle size distribution was obtained by classifying the particles with a group of standard screens and the particle density was measured by a true volume and density measurement instrument. The testing of particle density was repeated three times for each sample to obtain the average value accurately. All the samples were in air dried state beforehand; thus the influence of moist content can be neglected and the experiments were conducted at room temperature. 3. Results and discussion 3.1. Influence of particle density Generally, the utilization of CBFB for coal preparation is gravitybased and the distribution of the feed coal particles is largely influenced by the particle density and its deviation to the mean bed density. For fine coal particles, the particle motion in CBFB becomes complicated; thus the influence of particle density should be studied separately by controlling the other factors the same. Under the fluidized air velocity of 1.5umf for the bed medium, the coal particles and the gangue particles were beneficiated in the CBFB separately first, and each size fraction was considered individually in each operation. Fig. 4 shows the distributions of the weight fraction and the density variation for each coal sample along the bed height. It is clear that the light clean coal particles mainly flow to the surface layers of the bed and the weight proportion decreases along the bed height. The fitting curves of the weight proportion are in quadratic functions. As the top layer (the 7th layer) is usually shallow than the next normal layers, the top fraction shows a decrease, and it has been plotted separately from the normal layers in the figures. At the same time, the slope of the particle distribution curve is obviously larger for the coarse particle sample than the fine one, which means that the floating/beneficiation effect of the pseudofluid bed is stronger for the coarse particles. As seen in Fig. 4b, the density stratification occurs even within the same size fraction sample. For example, for the coal of 5.5–8 mm, the average particle density in the bottom layers is larger than the top layers and an apparent decease tendency along the bed can be observed. Even for the finest sample of 1–2 mm, the density stratification can also be detected. For the gangue particles as shown in Fig. 5, the distribution of the heavy gangue shows an interesting pattern. For the size fractions of 5.5–8 mm and 3–5.5 mm, most gangue particles sank to the bed bottom and the deepest layer (the 1st layer) has a very high mass proportion. The distribution of the rest of the particles from layer 2 to layer 6 shows an increased tendency overall. But for the size fraction of 1–2 mm, the particle's sedimentation effect is not obvious. The 1st
40%
Proportion (wt%)
3–5.5 mm
30%
20%
10%
0%
1
2
3
4
5
6
7
Height (Layer)
b) Particle Density (g/cm3)
Coal Gangue
5.5–8 mm
5.5-8 mm 3-5.5 mm 1-2 mm
Single Sample 100% Coal
True density (kg/m3)
1.5
Single Sample 100% Coal
5.5-8 mm 3-5.5 mm 1-2 mm
1.45
1.4
1.35
1.3
1
2
3
4
5
6
7
Height (Layer) Fig. 4. The distribution of coal particles and density variations along the bed height. (a) Particle distribution; (b) density variation.
layer shows a much lower proportion and a whole tendency can be derived from the 1st layer to the 6th layer. Due to the same reason as above, the 7th layer has low proportions. The bed expansion is diminished as a less volume of the same weight was added for the heavier gangue particles. The particle density stratification is more obvious for gangue samples. The density varies from 2.5 g/cm3 to 2. 75 g/cm3 for the coarse samples while still not significant for the finest one. In addition, a binary mixture of half coal sample and half gangue was beneficiated in the CBFB and the integral density effect was demonstrated. Each size fraction was testified separately and the results are shown in Fig. 6. As seen from the particle mass distribution, a similar tendency but a combined effect of the single sample results as shown in Figs. 4 and 5 can be observed. The proportion in the bottom layer decreases when compared with the pure gangue sample while the increase tendency in the middle layers is flattened when compared with the pure coal sample. However, for the binary mixtures, the effect of particle density distribution along the bed height can be more clearly observed. The average particle density at the very bottom reaches to 2.7 g/cm3 where gangue particles mainly exist while the top layers are most coal particles. The fraction of coal or gangue sample in each layer can be estimated from the average particle density and the pure sample densities. Again, the finest size fraction of 1–2 mm is not obviously stratified both in the mass distribution and the density distribution even for two samples with a quite large density difference. As proved above, the distribution of fine coal particles was still obviously influenced by the particle density. The density ratio between the particle density and the mean bed density is an important factor to determine the particle segregation [22–24]. For the coal samples, the density ratio is about 0.7, the large particles are inclined to be swept up to the bed surface but still could occasionally descend in the bed as
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
a) 50%
50%
Single Sample 100% Gangue
Proportion (wt%)
40%
5.5-8 mm 3-5.5 mm 1-2 mm
30%
20%
10%
0%
5.5-8 mm 3-5.5 mm 1-2 mm
Binary Mixture 50% Coal + 50% Gangue
40%
Proportion (wt%)
a)
31
30%
20%
10%
1
2
3
4
5
6
7
0%
Height (Layer)
1
2
3
4
5
6
7
Height (Layer)
b) 2.8
5.5-8 mm 3-5.5 mm 1-2 mm
2.7
2.6
Single Sample 100% Gangue
2.5
5.5-8 mm 3-5.5 mm 1-2 mm
Binary Mixture 50%Coal+50%Gangue
2.6
Particle Density (g/cm3)
Particle Density (g/cm3)
b) 2.8
2.4 2.2 2 1.8 1.6
2.4
1
2
3
4
5
6
7
1.4
Height (Layer)
1
2
3
4
5
6
7
Height (Layer) Fig. 5. The distribution of gangue particles and density variations along the bed height. (a) Particle distribution; (b) density variation.
the rising bubbles are pushing it back. For the gangue particles, the ratio ranges from 1.23 to 1.36. When the ratio is not large enough, the swarms of bubbles exert a considerable lifting effort to draw the particles up; as a result, the particle proportions from layer 2 to layer 6 show an increased tendency as in Figs. 4 to 6. For a density ratio great enough (more than 1.28 as reported by Kunii and Levenspiel [25] for coarse particles), the particles fall down to the bed bottom and stay there permanently. As there is a general correlation between the ash content and particle density [26], the density stratification can be effectively used to remove the high ash content gangue in the raw coal to upgrade the low rank coal samples in the fine particle size range. However, the segregation behavior is also largely affected by the particle size. When the particles are as fine as 1–2 mm, the density stratification in the dry CBFB is negligible and the separation can only resort to some other methods such as electric technology [27] or froth flotation [28].
3.2. Influence of particle size As partly demonstrated in the former section, the particle segregation behavior in CBFB is a strong function of the particle size. As seen from Figs. 4 to 6, the beneficiation efficiency deteriorates seriously as the particle size decreases. To further study the influence of particle size and the combination effects of different sizes, a ternary mixture with one third of each size fraction of coal samples (or gangue samples) was beneficiated in the CBFB at the fluidized air velocity of 1.5umf for the bed medium. The particle density variations were neglected and the allocations of size fraction in each layer were mainly concerned.
Fig. 6. The distribution of the binary mixture and average density variations along the bed height. (a) Particle distribution; (b) density variation.
As seen in Fig. 7, the coal particles of 5.5–8 mm and 3–5.5 mm show quite similar mass distributions along the bed height while the fraction of 1–2 mm shows a weaker flotation effect. For the gangue particles, the size fraction of 1–2 mm shows a weaker sedimentation effect as the weight proportion in the bottom layer is quite small and the increase tendency along the bed height is the most significant. When particles are small enough, their movements are mainly dominated by surface drag and less dependent on the gravity force and hydrostatic buoyancy. Consequently, the misplacing effect is substantial [29–31]. The back mixing effect in the bubbling CBFB has such a strong influence to the 1–2 mm particles that the light coal and heavy gangue show convergent tendencies, which brings the difficulty to distinguish and separate them apart in the raw coal. According to Tanaka et al. [32], the possibility of the particles segregation/mixing in a fluidized bed is dependent on the difference in size and density between the beneficiated particles and the medium particles. The segregation condition can be roughly formulated by Eqs. (1) and (2) as follows, where ρ and d refer to the particle density and diameter, respectively, and εmf is the void fraction at the minimum fluidization state. When the particle properties fall into the region, the two species are mixable and hard to be segregated in fluidization. ð1−ε mf Þ b
0:5 b
ρbeneficated‐particle 1 b ρmedium‐particle ð1−εmf Þ
dbeneficiated‐particle dmedium‐particle
!
ρbeneficiated‐particle ρmedium‐particle
ð1Þ !2 b2
ð2Þ
32
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
forces such as the vibrating force or the magnetic field are suggested to be introduced to improve the beneficiation quality [10,27]. Due to these reasons, the feed particle size is suggested to be more than 3 mm under the working conditions and with the magnetite particles used in this study.
a) 50%
Ternary Mixture Coal Samples 33 wt% for each size
Proportion (wt%)
40%
5.5-8 mm 3-5.5 mm 1-2 mm
3.3. Influence of fluidized air velocity
30%
20%
10%
0%
1
2
3
4
5
6
7
6
7
Height (Layer)
b)
50%
5.5-8 mm 3-5.5 mm 1-2 mm
Proportion (wt%)
40%
Ternary Mixture Gangue Samples 33 wt% for each size
30%
20%
10%
0% 1
2
3
4
5
Height (Layer) Fig. 7. The distribution of the samples' proportion along the bed height (ternary mixture: 33 wt.% for each size fraction). (a) Coal samples; (b) gangue samples.
The segregation possibility for the particles and conditions used in this work is shown in Fig. 8. For the coarse fractions of 5.5–8 mm and 3–5.5 mm, the coal particles and gangue particles are extractive from the bed medium, whereas the segregation phenomenon is obvious as proved from Figs. 4 to 7. For the finest size fraction of 1–2 mm, the particles in this work fall into the mixable region by a large section, which indicates that the beneficiated particles and the medium particles are hard to segregate in the fluidized bed normally. Some external
Although the characteristics of the pseudofluid state in CBFB are used to beneficiate coal particles, the separation mechanism is not exactly the same as the conventional wet dense medium separator. The stability of the CBFB and the beneficiation quality are directly influenced by the fluidized air velocity, which will determine the average bed density on the one hand, and alter the behaviors of the bubble phase on the other. The bed density can reach the similar values required as in the wet process by adjusting the property of the bed material and the air velocity, but the viscosity of the pseudofluid phase is quite lower [2,33]. In this work, a group of different fluidized air velocities were tested, ranging from 1.2umf to 1.8umf of the bed medium, and particles of different samples and size fractions were beneficiated singly. With an increase of the fluidized air velocity, the averaging bed density decreases slightly in linear as seen in Fig. 9. Even when the average bed density is not obviously changing, the bubble size is greatly increasing when the air velocity is enhancing and the fluidization is in more violently boiling. The particle distributions under different air velocities are comparatively shown in Fig. 10 for different samples and size fractions. It can be seen that the fluidized air velocity has an obvious influence on the particles of 5.5–8 mm both for coal samples and gangue samples while it has little effect to the size fractions of 3–5.5 mm and 1–2 mm. When the fluidization number increases, the weight proportions in the top layers decrease while more fractions move to the lower layers for the light coal samples of 5.5–8 mm, which means that the flotation effect deteriorates to some extent. The intensive backing mixing effect is mainly attributed to the strengthened bubble boiling. As the bubbles rising in the emulsion phase are not spherical but always have a concave base, the lower pressure in this wake region would draw up the solid particles, while solids in the emulsion phase will move downward to fill the caused vacancies [12,34]. The light coal particles are more easily affected by the circulation mixing effect, and the big particles are more sensitive to the bubble size and the turbulence intensity according to the findings in Fig. 10. For the gangue particles, the fraction in the bottom increases apparently when the fluidized air velocity increases. As the particle density of gangue is larger than the averaging bed density, the bubbling mixing strengthens the bed disturbance and accelerates 2.1
3-5.5 mm
Density Ratio
5.5-8 mm
1-2 mm Eq. (1) Eq. (2)
10
1
mixable
Averagiong bed density (g/cm3)
100
2.05
2
1.95
1.9 0.1 0.1
1
10
100
umf =0.067m/s
1
1.2
1.4
1.6
1.8
2
2.2
Fluidization number
Size Ratio Fig. 8. Segregation possibility of the particles in the CBFB.
Fig. 9. The averaging bed density under different fluidization number. Fluidization number = uoperation/umf.
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
b) 60%
50%
40%
Proportion (wt%)
1.2 umf 1.5 umf 1.8 umf
Single Sample Coal 5.5-8mm
30%
20%
10%
40% 30% 20% 10% 0%
0% 1
2
3
4
5
6
1
7
2
3
Height (Layer)
1.2 umf 1.5 umf 1.8 umf
Single Sample Coal 3-5.5mm
6
7
30%
20%
10%
6
7
6
7
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 3-5.5mm
50%
Proportion (wt%)
Proportion (wt%)
5
60%
40%
40% 30% 20% 10%
1
2
3
4
5
6
0%
7
1
2
3
Height (Layer)
4
5
Height (Layer)
50%
60%
1.2 umf 1.5 umf 1.8 umf
30%
20%
10%
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 1-2mm
50%
Proportion (wt%)
Single Sample Coal 1-2mm
40%
Proportion (wt%)
4
Height (Layer)
50%
0%
1.2 umf 1.5 umf 1.8 umf
Single Sample Gangue 5.5-8mm
50%
Proportion (wt%)
a)
33
40% 30% 20% 10%
0% 1
2
3
4
5
6
7
0%
1
Height (Layer)
2
3
4
5
Height (Layer)
Fig. 10. The distribution of the samples' proportion along the bed height under different fluidized velocities. (a) Coal samples; (b) gangue samples.
the sedimentation of heavy particles. Still, the effect is more obvious for the 5.5–8 mm size fraction and not significant for the rest of the fine ones. The fluidization of the beneficiated particles is facilitated by the viscous drag of the pseudofluid medium. The segregation behaviors of particles with different size and/or density are directly affected by the changing of the bed density and viscosity. As the fine particles have a lower minimum fluidization velocity in the dense medium, they are better mixed and their flotation/sedimentation is less influenced by the variation of the air velocity. Due to this reason, the aggregative fluidization regime is preferred in CBFB to ensure the stability of the bed for higher separation efficiencies. However, in industrial operation,
the air velocity is always set as high as possible to make the bubbling fluidization more uniform on one hand, and to speed up the stratification on the other.
4. Conclusions The particle distribution characteristics and the segregation behaviors of fine coal/gangue particles in a coal beneficiation fluidized bed (CBFB), which is still not fully clear to date, were revealed in detail in this work. Both clean coal samples and gangue samples of 1–8 mm in three separate size fractions were beneficiated purposely. The dense
34
Q. Wang et al. / Fuel Processing Technology 124 (2014) 28–34
bed was divided into seven layers and the segregation results were carefully analyzed. The results obtained can be summarized as follows: (1) The flotation and sedimentation of the particles in CBFB were still largely influenced by the particle density for the fine size range particles, and density stratification occurred even within each size fraction sample. The mass distribution showed a quadratic increase along the bed height for the light coal particles. For gangue particles, the mass proportions in the middle layers also showed an increased tendency surprisingly with a large fraction deposited in the bottom. (2) Both the particle segregation and the density stratification deteriorated seriously as the particle size decreased. The lower limit of the particle feed size was proved to be more than 3 mm and the beneficiation effect for finer particles was inefficient. (3) The increase of fluidized air velocity slightly decreased the averaging bed density while apparently strengthened the bubble boiling, in which the segregation efficiency of the 5.5–8 mm particles was obviously affected by the back mixing effect. Moreover, among the massive experimental data available, there is still not such a group that can be effectively used for numerical modeling for an academic purpose. The narrow ranges of particle density and size and the full information of particle distributions in CBFB of this work bring the convenience for CFD modeling validation, which, with no doubt, has arisen as a new effective tool in the coal beneficiation field. Acknowledgements Financial support of this work by the National Program on Key Basic Research Project (973 Program) of China (No. 2012CB214900) is gratefully acknowledged. The authors also gratefully acknowledge Dr. Xuemin Liu at the Department of Thermal Engineering, Tsinghua University for the preparation of the coal samples, and Xinhua Yang for the help in particle density measurements. References [1] Q. Chen, L. Wei, Development of coal dry beneficiation with air-dense medium fluidized bed in China, China Particuology 3 (2005) 42. [2] Y.M. Zhao, L.B. Wei, Rheology of gas–solid fluidized bed, Fuel Processing Technology 68 (2000) 153–160. [3] N.C. Lockhart, Dry beneficiation of coal, Powder Technology 40 (1984) 17–42. [4] J.A. van Houwelingen, T.P.R. de Jong, Dry cleaning of coal: review, fundamentals and opportunities, Geologica Belgica 7 (2004) 335–343. [5] R.K. Dwari, K.H. Rao, Dry beneficiation of coal—a review, Mineral Processing and Extractive Metallurgy Review 28 (2007) 177–234. [6] H. Katalambula, R. Gupta, Low-grade coals: a review of some prospective upgrading technologies, Energy and Fuels 23 (2009) 3392–3405. [7] A.K. Sahu, S.K. Biswal, A. Parida, Development of air dense medium fluidized bed technology for dry beneficiation of coal—a review, International Journal of Coal Preparation and Utilization 29 (2009) 216–241.
[8] S. Mohanta, C.S. Rao, A.B. Daram, S. Chakraborty, B.C. Meikap, Air dense medium fluidized bed for dry beneficiation of coal: technological challenges for future, Particulate Science and Technology 31 (2013) 16–27. [9] M.G. Rasul, V. Rudolph, F.Y. Wang, Particles separation using fluidization techniques, International Journal of Mineral Processing 60 (2000) 163–179. [10] X. Yang, Y. Zhao, Z. Luo, S. Song, C. Duan, L. Dong, Fine coal dry cleaning using a vibrated gas-fluidized bed, Fuel Processing Technology 106 (2013) 338–343. [11] Z. Luo, Q. Chen, Dry beneficiation technology of coal with an air dense-medium fluidized bed, International Journal of Mineral Processing 63 (2001) 167–175. [12] S. Mohanta, S. Chakraborty, B.C. Meikap, Influence of coal feed size on the performance of air dense medium fluidized bed separator used for coal beneficiation, Industrial and Engineering Chemistry Research 50 (2011) 10865–10871. [13] O. Prakash, D.D. Misra, Correlation of size and effective separation density in coal washing using heavy media, Journal of Mines, Metals and Fuels 27 (1979) 138–147. [14] J. Choung, C. Mak, Z. Xu, Fine coal beneficiation using an air dense medium fluidized bed, Coal Preparation 26 (2006) 1–15. [15] Z. Luo, Y. Zhao, Q. Chen, X. Tao, M. Fan, Separation lower limit in a magnetically gas–solid two-phase fluidized bed, Fuel Processing Technology 85 (2003) 173–178. [16] S. Azizi, S.H. Hosseini, G. Ahmadi, M. Moraveji, Numerical simulation of particle segregation in bubbling gas-fluidized beds, Chemical Engineering & Technology 33 (2010) 421–432. [17] Y. Zhao, L. Tang, Z. Luo, C. Liang, H. Xing, W. Wu, C. Duan, Experimental and numerical simulation studies of the fluidization characteristics of a separating gas–solid fluidized bed, Fuel Processing Technology 91 (2010) 1819–1825. [18] Q. Wang, J. Lu, W. Yin, H. Yang, L. Wei, Numerical study of gas–solid flow in a coal beneficiation fluidized bed using kinetic theory of granular flow, Fuel Processing Technology 111 (2013) 29–41. [19] Y.Q. Feng, A.B. Yu, Microdynamic modelling and analysis of the mixing and segregation of binary mixtures of particles in gas fluidization, Chemical Engineering Science 62 (2007) 256–268. [20] H. Lu, Y. He, D. Gidaspow, L. Yang, Size segregation of binary mixture of solids in bubbling fluidized beds, Powder Technology 134 (2003) 86–97. [21] P.K. Naik, P.S.R. Reddy, V.N. Misra, Optimization of coal flotation using statistical technique, Fuel Processing Technology 85 (2004) 1473–1485. [22] F. Garcia, A. Romero, J. Villar, A. Bello, A study of segregation in a gas solid fluidized bed: particles of different density, Powder Technology 58 (1989) 169–174. [23] A. Nienow, P. Rowe, L. Cheung, A quantitative analysis of the mixing of two segregating powders of different density in a gas-fluidised bed, Powder Technology 20 (1978) 89–97. [24] A. Rao, J. Curtis, B. Hancock, C. Wassgren, Classifying the fluidization and segregation behavior of binary mixtures using particle size and density ratios, AIChE Journal 57 (2011) 1446–1458. [25] D. Kunii, O. Levenspiel, Fluidization Engineering, 2nd ed. Butterworth-Heinemann, Newton, MA, 2005. [26] C. Zhang, G. Chen, T. Yang, G. Lu, C. Mak, D. Kelly, Z. Xu, An investigation on mercury association in an Alberta sub-bituminous coal, Energy and Fuels 21 (2007) 485–490. [27] Z. Luo, Y. Zhao, Q. Chen, M. Fan, X. Tao, Separation characteristics for fine coal of the magnetically fluidized bed, Fuel Processing Technology 79 (2002) 63–69. [28] A. Guney, G. Onal, O. Ergut, Beneficiation of fine coal by using the free jet flotation system, Fuel Processing Technology 75 (2002) 141–150. [29] Y. He, H. Chen, Y. Ding, B. Lickiss, Solids motion and segregation of binary mixtures in a rotating drum mixer, Chemical Engineering Research and Design 85 (2007) 963–973. [30] A. Nienow, N. Naimer, T. Chiba, Studies of segregation/mixing in fluidised beds of different size particles, Chemical Engineering Communications 62 (1987) 53–66. [31] Y. Soong, T.A. Link, M.R. Schoffstall, M.L. Gray, D.J. Fauth, J.P. Knoer, J.R. Jones, I.K. Gamwo, Dry beneficiation of Slovakian coal, Fuel Processing Technology 72 (2001) 185–198. [32] Z. Tanaka, X. Song, Continuous separation of particles by fluidized beds, Advanced Powder Technology 7 (1996) 29–40. [33] R. Sahan, B. Kozanoglu, Use of an air fluidized bed separator in a dry coal cleaning process, Energy Conversion and Management 38 (1997) 269–286. [34] S. Mohanta, S. Chakraborty, B.C. Meikap, Optimization process of an air dense medium fluidized bed separator for treating high-ash non-coking Indian coal, Mineral Processing and Extractive Metallurgy Review 34 (2013) 240–248.