- Email: [email protected]
Flow patterns and transitions in a rectangular three-phase bubble column
Powder Technology 260 (2014) 27–35
Contents lists available at ScienceDirect
Powder Technology journal homepage: www.elsevier.com/locate/powtec
Flow patterns and transitions in a rectangular three-phase bubble column Weiling Li, Wenqi Zhong ⁎, Baosheng Jin, Yong Lu, Tingting He Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, School of Energy and Environment, Southeast University, Nanjing 210096, China
a r t i c l e
i n f o
Article history: Received 5 December 2013 Received in revised form 19 February 2014 Accepted 1 April 2014 Available online 12 April 2014 Keywords: Flow pattern Flow regime transition Gas–liquid–solid flow Particle effect Three-phase bubble column
a b s t r a c t The flow patterns and transitions in a rectangular gas–liquid–solid three-phase bubble column were studied. The influence of solid volume fraction, particle size and particle density on the flow regime transitions of the threephase bubble column was investigated experimentally. Experiments were carried out for solid volume fraction Vs = 0.03–0.3, average particle size dp = 48 μm–270 μm, particle density ρp = 2500 kg/m3–4800 kg/m3, and superficial gas velocity Ug = 0.007 m/s–0.7 m/s in a rectangular bubble column measured 0.8 m tall, 0.1 m long and 0.01 m wide. Four distinct flow patterns and three transition points were observed in this experimental system, and the four flow regimes were discrete bubble regime, transition regime, bubble coalescence regime and strong turbulent regime, which were determined on the basis of criteria as well as schematic diagrams and typical flow pattern images obtained from a high-resolution digital charge couple device (CCD) camera. Typical flow patterns maps were plotted for illustrating flow regime transitions under different particle conditions. It was found out that particle volume fraction and particle density had an effect on the flow pattern transitions; when increasing the values of particle volume fraction and the particle density respectively, the values of the flow regime transition points all decreased. Particle size had a little effect on flow regime transition points when the particle size shifted in the range of 48 μm to 150 μm; when particle size was larger than 150 μm and increased to 270 μm, the operation ranges of the transition regime and the bubble coalescence regime decreased. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Gas–liquid–solid three-phase bubble columns have been widely utilized in chemical, petrochemical, energy conversion, biochemical and environmental processes [1–5], such as direct coal liquefaction, Fischer–Tropsch (FT) synthesis, CO2 removal from flue gases in a bubble column [6], heavy oil hydrocracking using catalyst [7], and sodium bicarbonate production [8]. However, there are still some difficulties on the bubble column scale-up technology for industry utilization due to the lack of comprehensive knowledge on the flow characteristics. Flow patterns and transition are two of the important hydrodynamic characteristics of the gas–liquid–solid bubble column, and the mass and heat transfer, mixing efficiency, pressure gradient, momentum loss, pipe vibration and reactor volume productivity of these systems all vary greatly with different flow patterns. Hence, flow regime and transition study is an important respect on understanding the flow characteristics of three-phase bubble columns, and it provides
⁎ Corresponding author at: School of Energy and Environment, Southeast University, Sipailou 2#, Nanjing 210096, Jiangsu, China. Tel.: + 86 25 83794744; fax: + 86 25 83795508. E-mail addresses: [email protected] (W. Li), [email protected] (W. Zhong), [email protected] (B. Jin).
http://dx.doi.org/10.1016/j.powtec.2014.04.002 0032-5910/© 2014 Elsevier B.V. All rights reserved.
fundamental knowledge on bubble column design, optimization, operation and scale-ups. Flow regime transition is usually influenced by many factors, such as column geometry, gas distributor design, gas flow rate and operating conditions (temperature, pressure and the physical properties of gas, liquid and particle) [9,10]. Numerous studies have investigated flow patterns in bubble columns, some studies were focused on gas–liquid two-phase systems, and some investigated three-phase systems. The shapes of bubble columns mainly included cylindrical and rectangular two types [11]. When the liquid phase operated in batch mode, Shaikh et al. pointed out that four types of flow patterns had been observed, and they are homogeneous (bubbly), heterogeneous (churn-turbulent), slug and annular flow [4]. Three flow regimes with two flow regime transition points have been identified [12–14], and they were homogeneous regime, transition regime, and the heterogeneous regime. Angeles et al. [7] summarized three types of flow regimes in bubble column reactor, viz., homogeneous bubble flow, heterogeneous bubble flow and churn turbulent slug flow regime. Some scholars [10,15] studied two principal flow regimes, homogeneous and heterogeneous regimes in bubble column systems. Nedeltchev et al. [16] studied a bubble column and identified five flow regimes, and they were: a dispersed bubble regime, first and second transition regimes, a coalesced bubble regime consisting of four regions (called 4-region flow) and a coalesced bubble regime consisting of three regions (called 3-region flow). When
28
W. Li et al. / Powder Technology 260 (2014) 27–35
the liquid phase and the gas flow were operated in co-current mode, Shiea et al. [17] detected dispersed bubble flow, discrete bubble flow, coalesced bubble flow and slug flow regimes in their experimental bubble column. Most of these researches were conducted in cylindrical bubble columns, and therefore rectangular bubble columns needed more studies. A rectangular bubble column with a width of 200 mm and a depth of 15 mm was utilized to study the turbulence characterization under both homogeneous and heterogeneous regimes [18]. Particle properties influence the flow regime transitions, and some studies have been performed on the particle effect. Mena et al. [15] discussed the solid effect on homogeneous–heterogeneous flow pattern transition in a bubble column, and provided detailed suggestions on the explanation of the dual solid effect observed. Rabha et al. [19] studied the effect of particle size on the intrinsic flow behavior in a slurry bubble column in order to provide reliable guidelines for optimizing catalyst particle size and solid concentration. Gan [20] studied the solid circulation velocities and holdup profiles in a gas–liquid–solid bubble column, and analyzed the effect of the particle addition on the gas–liquid flow and the three-phase flow hydrodynamics in bubble columns [21]. There is still a lack of detailed understanding of the effect of particles on regime transition, such as the particle size and density effect on the flow structure at different operating conditions [22]. The objective of the study was to investigate the flow patterns in a rectangular bubble column through experiments, and to study the effect of the particle volume fraction, particle size and particle density on the flow regime transition points. This paper is structured as follows. Following the introduction, experimental work is described in the
second part. Subsequently, four flow patterns and three transition points are detailed in the third part; flow regime maps are analyzed, followed by concluding remarks.
2. Experimental section A gas–liquid–solid three-phase bubble column experimental system was built, and the schematic diagram of the experimental set-up was sketched in Fig. 1. The experimental system consists of four sections: a gas–liquid–solid three-phase bubble column, an air supply system, a differential pressure signal acquisition system and a digital image acquisition system. The bubble column measured 0.8 m tall, 0.1 m long and 0.01 m wide and was made of transparent, 6-mm-thick Plexiglas. The air supply system consists of an air compressor, pipelines and flowmeters. The differential pressure signal acquisition system includes the diffused silicon pressure transmitters (GB-3000E and GB-3000HK), a data acquisition card (CDAQ-9188), power supply, positive wire, negative wire and data collection software (Labview 2010 Signal Express). The measurement range and measurement accuracy of the diffused silicon pressure transmitters are 0–2.5 kPa and 0.05%, respectively. The digital image acquisition system consists of a high resolution digital CCD camera and a computer for photo collection. The high-speed digital camera used features a high-speed consecutive shooting mode and shoots 16 frames/s in JPEG mode and thus could be used to capture a series of gas–liquid–solid flow structure images. A 1000-W floodlight was used for lighting and to ensure high photograph resolution.
Fig. 1. A schematic diagram of the experimental set-up. 1. Computer; 2. Power supply; 3. Data collector; 4. Differential pressure sensor; 5. Gas outlet; 6. A three-phase bubble column; 7. Metal sintered plate distributor; 8. Flowmeters; 9. Light; 10. CCD camera.
W. Li et al. / Powder Technology 260 (2014) 27–35
Air, water and particles were used as the gas, liquid and solid phases respectively. Air was supplied from an air compressor, and then it was measured by three flowmeters and passed through a metal sintered plate distributor into the bubble column, and the average diameter ds of holes in the sintered plate was 50 μm. Tap water was operated in batch mode. The properties of particles were shown in Table 1. Five kinds of particles were used in experiments. Glass beads A–D were particles with different particle sizes, but had the same density; iron powder had the largest particle density and had the same average particle size with the glass beads B. The solid phase volume fractions tested in these experiments were 3%, 9%, 20% and 30%. In Table 2, experimental and sampling conditions were listed. The superficial gas velocity used in the study was over the range of 0.007 m/s–0.7 m/s. When the liquid and solid phases were prepared in the bubble column, it was a static bed. Then air as the gas phase entered into the bed, and gas–liquid–solid flow was formed. Experiments were carried out at ambient temperature and atmospheric pressure. Three-phase flow patterns operated at different superficial gas velocities in the rectangular bubble column were recorded and photographs were used to be analyzed. The experiments were repeated under different conditions, operating at different solid phase volume fractions, particle size and densities. 3. Results and discussion
29
Table 2 Experimental and sampling conditions. H0 (mm)
Ug (m/s)
f (Hz)
t (s)
120
0–0.7
1000
60
(a1)
(a2)
(a3)
(a4)
(b)
(c)
(d)
(e)
3.1. Identification of flow patterns As illustrated in the experimental section, the gas flow rate was adjusted and increased stepwise by step, gas–liquid–solid three-phase flow structures formed and evolved gradually. Four flow regimes and three transition points were observed in this experiment system. A schematic representation of these flow regimes were shown in Fig. 2, which depicted the main characteristics of these flow patterns. The flow patterns and transition points included: (a) discrete bubble regime, DBR; (b) the first transition point, FTP; (c) transition regime, TR; (d) the second transition point, STP; (e) bubble coalescence regime, BCR; (f) the third transition point, TTP and (g) strong turbulent regime, STR. The criteria as well as typical images for the description of every flow pattern are presented as follows. 3.1.1. Discrete bubble regime In this flow pattern, bubbles disperse in the liquid–solid flow, which is operated at low superficial gas velocity. Bubbles have a spherical shape, all bubbles have the similar size, and this flow regime was also demonstrated by Mota et al. [10] and Mena et al. [15]. When the superficial gas velocity increases to 0.025 m/s, the bubble size becomes larger, bubbles have ellipsoidal shape, and more numbers of bubbles appeared, whereas all bubbles still distribute uniformly in the liquid–solid flow in this flow regime. Liquid phase has continuous distribution, and the distribution of solid phase depends on particle size and density which is illustrated in Fig. 2a1–a4. The bubble size was only influenced by the gas sparger in this regime [16]. Fig. 3 depicts the typical flow pattern images of discrete bubble regime recorded by digital CCD camera. All the images in Fig. 3 were collected from different gas–liquid–solid systems, such as, glass beads A was used as the solid phase in Fig. 3(a1) and Vs = 0.03; glass beads
Table 1 Particle properties. Particles
dp (μm)
ρs (kg/m3)
Glass beads A Glass beads B Glass beads C Glass beads D Iron powder
48 75 150 270 75
2500 2500 2500 2500 4800
(f)
(g1)
(g2)
Fig. 2. Schematic representation of various flow patterns and transition points: (a1–a4) discrete bubble regime, DBR; (b) the first transition point, FTP; (c) transition regime, TR; (d) the second transition point, STP; (e) bubble coalescence regime, BCR; (f) the third transition point, TTP; (g1–g2) strong turbulent regime, STR.
30
W. Li et al. / Powder Technology 260 (2014) 27–35
(a1) (a1)
(a2)
(a2)
(a3)
(a3)
Fig. 3. Typical flow pattern images of discrete bubble regime (DBR): (a1) Glass beads A, particle volume fraction, V s = 3%, U g = 0.036 m/s; (a2) Glass beads B, V s = 20%, U g =0.047 m/s; (a3) Glass beads C, Vs = 20%, U g = 0.014 m/s.
B and Vs = 0.2 was demonstrated in Fig. 3(a2). Although they had different particle size and volume fraction, they showed similar flow structures and were in the same flow regime. It can be observed that both of particle sizes and particle volume fractions will affect the three-phase flow characteristics, such as bed expansion heights and bubble sizes. 3.1.2. The first transition point In this case, deformed bubbles appear firstly; some small ellipsoidal bubbles lose their stability, coalesce and form bubble blocks and larger bubbles in the local region. Bubbles near the sidewall keep an ellipsoidal shape. Particles move with the liquid phase. When the gas–liquid–solid three-phase flow shows these characteristics, it means that the flow pattern changes from the discrete bubble regime to the transition regime. Fig. 4(a1–a3) depicts the typical flow pattern images of the first transition point in this experimental system. The bubble blocks are defined as the bubbles that have irregular shapes and have larger sizes than that surrounded mainly caused by the coalescence of small size bubbles which distribute uniformly in the homogeneous flow regime. Bubble blocks move upward near the central axis. The close-up of the bubble blocks photos was shown in Fig. 4(a4), which was illustrated by the two blue circles. 3.1.3. Transition regime In this flow pattern, bubble block behaviors become more obvious. The volume of the bubble blocks becomes even larger. Bubbles have coalescence phenomena and move upward in the bed region. The liquid phase distributes uniformly together with bubbles and moves upward. Particles, such as glass beads A, glass beads B and C all distribute uniformly in the bubble column, which can be elucidated in Fig. 5. This flow regime was also described in previous studies [14]. Ruthiya et al. [12] gave the schematic representation of this regime. Compared with the transition regime characteristics described by
(a4) Fig. 4. Typical flow pattern images of the first transition point (FTP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.058 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.058 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.047 m/s; (a4) the close-up of bubble blocks photos.
Ruthiya et al. [12], the central bubble plume appeared in the bubble column, but the liquid eddy was not observed obviously in this regime. 3.1.4. The second transition point In this point, liquid circulation appears firstly in local regions as shown in Fig. 2(d); larger bubbles form and eddies appear caused by the onset of bubble coalescence, liquid circulation and larger liquid momentum transferred by the gas phase. The quantity of small bubbles reduces. The system has stronger coalescence phenomena and from this point the gas–liquid–solid three-phase flow shifts from the transition regime to the heterogeneous flow regime, which was illustrated in Ref. [14]. The typical flow pattern images of the second transition point were highlighted in Fig. 6. 3.1.5. Bubble coalescence regime In this case, bubbles have a relatively regular movement, small bubbles forming at the distributor region, and some of them rise into the central region of the bubble column, coalesce and form large bubbles which then rise up in the bed. These large bubbles have an irregular shape. The numbers of small bubbles reduce in this flow pattern, and bubbles have large size distributions. Fig. 7 presents the typical flow pattern images of the bubble coalescence regime. This flow regime could be
W. Li et al. / Powder Technology 260 (2014) 27–35
(a1)
(a2)
(a3)
Fig. 5. Typical flow pattern images of transition regime (TR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.089 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.11 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.13 m/s.
observed more obviously when glass beads B was used as the solid phase, which was depicted in the Fig. 7(a2). The liquid phase moved upward in the middle portion of the column to the surface and then descended adjacent to the sidewalls building the gross liquid
(a1)
31
(a2)
(a3)
Fig. 7. Typical flow pattern images of bubble coalescence regime (BCR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.16 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.18 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.17 m/s.
circulation. The existence of a solid phase and the volume ratio of gas and liquid phases have an effect on shaping this flow regime. Nedeltchev et al. [16] described that the flow pattern shifted to coalesced bubble (3-region flow) when the superficial gas velocity was larger than 0.13 m/s at atmospheric pressure in a gas–liquid two-phase bubble column. The bubble coalescence regime from this experimental results and the coalesced bubble regime (3-region flow) investigated by Nedeltchev et al. [16] show similar flow structure characteristics. The coalesced bubble regime (3-region flow) consists of the central fast bubble region, the vortical and the descending regimes [23]. Bubble coalesced and broke up in the central fast bubble region, and bubble clusters and coalesced bubbles moved upward in this region. The vortical flow was surrounding the central fast bubble region, and vortices existed in gas–liquid–solid three-phases. The descending regime was located adjacent to the column wall; three-phases streams moved downward. These characteristic can be observed in Fig. 7. 3.1.6. The third transition point In this case, bubbles have separation behaviors, and they have several ways to rise upward and do not coalesce in the middle portion which is distinguished from the bubble coalescence regime. Bubbles have more than two routes rising to the surface. Typical flow structure images of the third transition point were depicted in Fig. 8. Analyzing the pressure time-series collected from this flow regime with the time–frequency domain analysis methods, the time–frequency distributions were distinguished from that of previous flow regimes. Therefore, it was necessary to distinguish the flow pattern which was operated at higher superficial gas velocity from the previous flow regimes.
(a1)
(a2)
(a3)
Fig. 6. Typical flow pattern images of the second transition point (STP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.13 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.16 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.16 m/s.
3.1.7. Strong turbulent regime In this case, the bubble column usually operates above 0.25 m/s. At high gas velocity, liquid has large momentum, and the gas–liquid– solid three-phase flow shows strong turbulent flow structures. Bubbles take a large volume of the bed as the volume ratio of gas and liquid phases increases further. Bubbles coalesce and break up strongly in
32
W. Li et al. / Powder Technology 260 (2014) 27–35
0.6
Particle volume fraction
Particle:Glass beads A dp= 48µm 0.4
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
0.2 DBR
BCR
TR
STR
0.0 0.0
0.2
0.1
0.6
0.7
Superficial gas velocity (m/s) Fig. 10. Flow regime map of gas–liquid–solid bubble column systems at different particle volume fractions (dp = 48 μm).
studies on three-phase bubble columns [10,12], the authors did not mention this flow pattern due to their gas velocities studied varied in the range of 0–0.3 m/s and the gas velocities studied were not larger than 1 m/s [24].
(a1)
(a2)
(a3)
Fig. 8. Typical flow pattern images of the third transition point (TTP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.2 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.22 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.22 m/s.
this regime. The liquid phase has high energy, discontinuous distribution and is in a very turbulent pattern. The liquid flow pattern is much more chaotic than that in previous regimes. Liquid phase and gas phase have spatial and temporal stochastic distributions. Fig. 9 illustrated the flow pattern images of a strong turbulent regime. In the previous
3.2. Flow regime map and flow pattern transition In gas–liquid–solid three-phase bubble columns, the system has different flow patterns when changing the gas flow rate. Four flow regimes and three transition points were observed at different superficial gas velocities as discussed in the previous part. Several factors may influence the transitions of different flow patterns, such as solid phase volume fractions, particle size and particle density. Studies were carried out to find out how the flow regime transition points shifted when all these factors changed. Hence, flow regime maps were plotted with all these factors, and the operation ranges of these flow patterns were illustrated and identified. 3.2.1. The effect of particle volume fraction Glass beads were utilized as the solid phase, and the particle volume fractions were 3%, 9%, 20% and 30%. Figs. 10 and 11 depicted the flow regime map under different particle volume fraction. The superficial gas velocity Ug was plotted on the abscissa axis and the particle volume fraction was plotted on the ordinate axis. As shown in Fig. 10, the particle size was 48 μm and Vs equaled to 0.03; the superficial gas velocity of 0.6 Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
Particle volume fraction
Particle:Glass beads C dp=150µm 0.4
0.2 DBR TR
BCR
STR
0.0
(a1)
(a2)
(a3)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Superficial gas velocity (m/s) Fig. 9. Typical flow pattern images of strong turbulent regime (STR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.24 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.24 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.83 m/s.
Fig. 11. Flow regime map of gas–liquid–solid bubble column systems at different particle volume fractions (dp = 150 μm).
W. Li et al. / Powder Technology 260 (2014) 27–35
33
Table 3 Summary of various Ug,c values. Ug,c (m/s), dp = 48 μm
Transition points
The first transition point The second transition point The third transition point
Ug,c (m/s), dp = 150 μm
3%
9%
20%
30%
3%
9%
20%
30%
0.058 0.13 0.2
0.047 0.13 0.18
0.047 0.13 0.18
0.047 0.11 0.18
0.058 0.18 0.24
0.047 0.16 0.22
0.036 0.11 0.18
0.036 0.089 0.16
Table 4 Comparison with literature data. Investigators
System
Bubble column type
The transition points (m/s)
Ruthiya et al. [12]
Nitrogen gas/silica, carbon/demineralized water, dp = 44 μm, 30 μm Air/Therminol LT
Rectangular bubble column
Ug,c1 = 0.06 m/s, Ug,c2 = 0.12 m/s.
Cylindrical bubble column
Air/water, studies of Bach and Pilhofer Air/water, studies of Oels et al. Air/tap water/glass beads, dp = 35 μm Air/tap water/glass beads, dp = 48 μm
– – Cylindrical bubble column Rectangular bubble column
Ug,c1 = 0.02 m/s, Ug,c2 = 0.08 m/s, Uc3 = 0.10 m/s, Uc4 = 0.13 m/s. Ug,c1 = 0.046 m/s. Ug,c1 = 0.039 m/s. Ug,c1 = 0.05 m/s, Ug,c2 = 0.125 m/s. Ug,c1 = 0.047 m/s, Ug,c2 = 0.13 m/s, Ug,c3 = 0.18 m/s.
Nedeltchev et al. [16] Kantarci et al. [24] Barghi et al. [25] This studies
Subscripts: g,c1—the first transition point; g,c2—the second transition point; g,c3— the third transition point; c3—the point shifting to coalesced bubble (4-region flow); c4—the point shifting to coalesced bubble (3-region flow).
the first transition point was 0.058 m/s. The second transition point was at 0.13 m/s and the third transition point was at 0.20 m/s. All the transitional gas velocity values were summarized in Table 3. When the particle volume fraction shifted from 0.03 to 0.09, the first and the third transitional gas velocities decreased; when the particle volume fraction increased further, shifting from 0.09 to 0.3, these two transition points kept the same values. The second transition point stayed at 0.13 m/s with the particle volume fraction of 0.03, 0.09 and 0.2, but the value decreased when that shifted from 0.2 to 0.3. Fig. 11 presented the flow regime map of the three-phase system with a particle size of 150 μm. When the particle volume fraction increased, shifting from 0.03 to 0.2, all the three transition points decreased, such as the second transition point decreased to 0.156 m/s at Vs = 0.09, which was identified in reference [14]. The reasons may be due to the high particle volume fraction which would increase the instability of the system, and therefore make the system shift into a turbulent pattern at lower gas velocities. It can be concluded that increasing the particle volume fraction can reduce three transitional gas velocity values, and this trend was obviously observed in Fig. 11 with dp = 150 μm. This conclusion was in agreement with the results of Mota et al. [10]. The experimental results of transitional gas velocities have been compared with literature data, which was shown in Table 4. The
transition point results of the rectangular and cylindrical two types have also been quantitatively compared. Comparing the results of this study with that of Ruthiya et al. [12] and Barghi et al. [25] in Table 4, the first and second transitional gas velocity values had good agreement, and the bubble column type did not have a large influence on the transition gas velocities.
3.2.2. The effect of particle size Glass beads were utilized as the solid phase, and the average particle sizes were 48 μm, 75 μm, 150 μm and 270 μm. Fig. 12 demonstrated the effect of particle size on the flow regime map of gas–liquid–solid bubble column systems. The superficial gas velocity Ug was plotted on the abscissa axis and the particle size was plotted on the ordinate axis. When the particle size was 75 μm and the particle volume fraction was 0.09, the first transition point was at 0.058 m/s; the second transition point was at 0.16 m/s and the third transition point was at 0.22 m/s. As observed from Fig. 12, the first transition point was not affected by the particle size, and the second transition point kept the same value when the particle size shifted from 48 μm to 150 μm. When the particle size increased from 150 μm to 270 μm, the values of the second and the third transition points decreased, system shifting into the next
500
9000
400 350 300 250 200
DBR
TR
BCR
STR
150
Particle:Glass beads C Iron powder dp=75µm
8000
Particle density (kg/m3)
Particle diameter (µm)
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
Particle:Glass beads Vs=9%
450
7000 6000
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
5000 4000
DBR
TR
BCR
STR
100 3000 50 0.0
0.1
0.2
0.6
0.7
Superficial gas velocity (m/s) Fig. 12. Flow regime map of gas–liquid–solid bubble column systems at different particle diameter (Vs = 9%).
0.0
0.1
0.2
0.6
0.7
Superficial gas velocity (m/s) Fig. 13. Flow regime map of gas–liquid–solid bubble column systems at different particle density (dp = 75 μm, Vs = 9%).
34
W. Li et al. / Powder Technology 260 (2014) 27–35
flow regime at a lower superficial gas velocity. It can be concluded that the particle size had a little effect on the operation range of flow patterns at the particle size of 48 μm–150 μm. However, the particle size can narrow down the operation ranges of transition regime and bubble coalescence regime when the particle size was larger than 150 μm. The three-phase system had the lowest gas velocity for shifting into the strong turbulent regime when the particle size was 270 μm.
Static bed
BDR
3.2.3. The effect of particle density Glass bead and iron powder were utilized as the solid phase, and the particle densities were 2500 kg/m3 and 4800 kg/m3, respectively. Fig. 13 depicted the effect of particle density on the flow regime map of gas– liquid–solid bubble column systems. As shown in Fig. 13, the superficial gas velocity Ug was plotted on the abscissa axis and particle density was plotted on the ordinate axis. Particle size and the particle volume
TR
BCR
STR
BCR
STR
(a)
Static bed
BDR
TR
(b) Fig. 14. Flow regime images of gas–liquid–solid bubble column systems at different particle density: (a) Glass beads B, Ug = 0, 0.025 m/s, 0.11 m/s, 0.18 m/s, 0.33 m/s; (b) Iron powder, Ug = 0, 0.025 m/s, 0.058 m/s, 0.18 m/s, 0.42 m/s. (dp = 75 μm, Vs = 9%).
W. Li et al. / Powder Technology 260 (2014) 27–35
fraction kept the same values which were equaled to 75 μm and 0.09 respectively, whereas the particle density changed. It can be observed that all the three transition points decreased when particle density increased. Fig. 14 showed the static bed and four flow regime images of two systems, which utilizes glass bead and iron powder respectively as the particle phase. Observing the images of the transition regime in Fig. 14, particle density had large effect on the flow structure, bed expansion height, bubble size distributions and the turbulent structures which were operated at high gas velocity. Bed pressure drop of the column also changed when increasing particle density. 4. Conclusions Experimental studies on the flow pattern and transitions were carried out in a rectangular gas–liquid–solid three-phase bubble column. Flow pattern and transitions under different particle conditions were investigated. Typical flow patterns by certain criteria as well as schematic diagrams and flow pattern images photographed by a high-resolution digital CCD camera were presented. Flow pattern maps were plotted for describing the flow regime transitions under different particle conditions. The key conclusions are as follows: (1) Four distinct flow patterns and three transition points were identified in the experimental system, and they were discrete bubble regime, the first transition point, transition regime, the second transition point, bubble coalescence regime, the third transition point and strong turbulent regime. (2) When increasing the solid volume fraction in a three-phase bubble column, shifting from 0.03 to 0.3, the superficial gas velocities of the three transition points decreased. (3) When the particle size shifted from 48 μm to 150 μm, it had a little effect on the operation ranges of flow patterns. When the particle size increased from 150 μm to 270 μm, the values of the second and the third transitional gas velocities decreased. (4) When increasing the particle density, shifting from 2500 kg/m3 to 4800 kg/m3, the superficial gas velocities of the three transition points decreased. The three-phase system had different bed expansion height, bubble size distributions and turbulent structures shown at a high gas flow rate when changing particle density. Nomenclature BCR bubble coalescence regime dp average particle size, μm ds the average diameter of holes in sparger, μm DBR discrete bubble regime FTP the first transition point Qg volumetric gas flow rate, m3/h STP the second transition point STR strong turbulent regime TR transition regime TTP the third transition point Ug superficial gas velocity, m/s U superficial velocity, m/s Vs particle volume fraction
Greek letter ρ density of phases, kg/m3
subscripts g gas g,c critical value of gas phase
l p s
35
liquid particle solid
Acknowledgments The authors would like to thank the National Natural Science Fund for Distinguished Young Scholars (No. 51325601), the Major State Basic Research Development Program of China (973 Program, No. 2010CB732206) and the Scientific Research Foundation of Graduate School of Southeast University (YBJJ1225) for the financial support. References [1] H. Li, A. Prakash, Influence of slurry concentrations on bubble population and their rise velocities in a three-phase slurry bubble column, Powder Technol. 113 (2000) 158–167. [2] B. Gandhi, A. Prakash, M.A. Bergougnou, Hydrodynamic behavior of slurry bubble column at high solids concentrations, Powder Technol. 103 (1999) 80–94. [3] Y.C. Li, Fluidization Processing Engineering Theory, Science Press, Beijing, 2008. [4] A. Shaikh, M.H. Al-Dahhan, A review on flow regime transition in bubble columns, Int. J. Chem. React. Eng. 5 (2007). [5] L.-S. Fan, Gas–Liquid–Solid Fluidization Engineering, Butterworths, Boston, 1989. [6] L. Cheng, T. Li, T.C. Keener, J.Y. Lee, A mass transfer model of absorption of carbon dioxide in a bubble column reactor by using magnesium hydroxide slurry, Int. J. Greenhouse Gas Control 17 (2013) 240–249. [7] M.J. Angeles, C. Leyva, J. Ancheyta, S. Ramírez, A review of experimental procedures for heavy oil hydrocracking with dispersed catalyst, Catal. Today 220–222 (2014) 274–294. [8] B.t. Haut, V. Halloin, T. Cartage, A. Cockx, Production of sodium bicarbonate in industrial bubble columns, Chem. Eng. Sci. 59 (2004) 5687–5694. [9] S. Nedeltchev, A. Shaikh, M. Al-Dahhan, Flow regime identification in a bubble column via nuclear gauge densitometry and chaos analysis, Chem. Eng. Technol. 34 (2011) 225–233. [10] A. Mota, A.A. Vicente, J. Teixeira, Effect of spent grains on flow regime transition in bubble column, Chem. Eng. Sci. 66 (2011) 3350–3357. [11] M. Pourtousi, J.N. Sahu, P. Ganesan, Effect of interfacial forces and turbulence models on predicting flow pattern inside the bubble column, Chem. Eng. Process. Process Intensif. 75 (2014) 38–47. [12] K.C. Ruthiya, V.P. Chilekar, M.J.F. Warnier, J. van der Schaaf, B.F.M. Kuster, J.C. Schouten, J.R. van Ommen, Detecting regime transitions in slurry bubble columns using pressure time series, AIChE J. 51 (2005) 1951–1965. [13] C. Vial, S. Poncin, G. Wild, N. Midoux, A simple method for regime identification and flow characterisation in bubble columns and airlift reactors, Chem. Eng. Process. Process Intensif. 40 (2001) 135–151. [14] W.L. Li, W.Q. Zhong, B.S. Jin, R. Xiao, T.T. He, Flow regime identification in a threephase bubble column based on statistical, Hurst, Hilbert–Huang transform and Shannon entropy analysis, Chem. Eng. Sci. 102 (2013) 474–485. [15] P.C. Mena, M.C. Ruzicka, F.A. Rocha, J.A. Teixeira, J. Drahoš, Effect of solids on homogeneous–heterogeneous flow regime transition in bubble columns, Chem. Eng. Sci. 60 (2005) 6013–6026. [16] S. Nedeltchev, A. Shaikh, M. Al-Dahhan, Flow regime identification in a bubble column based on both statistical and chaotic parameters applied to computed tomography data, Chem. Eng. Technol. 29 (2006) 1054–1060. [17] M. Shiea, N. Mostoufi, R. Sotudeh-Gharebagh, Comprehensive study of regime transitions throughout a bubble column using resistivity probe, Chem. Eng. Sci. 100 (2013) 15–22. [18] M. Sathe, J. Joshi, G. Evans, Characterization of turbulence in rectangular bubble column, Chem. Eng. Sci. 100 (2013) 52–68. [19] S. Rabha, M. Schubert, U. Hampel, Intrinsic flow behavior in a slurry bubble column: a study on the effect of particle size, Chem. Eng. Sci. 93 (2013) 401–411. [20] Z.W. Gan, Holdup and velocity profiles of monosized spherical solids in a threephase bubble column, Chem. Eng. Sci. 94 (2013) 291–301. [21] Z.W. Gan, S.C.M. Yu, A.W.K. Law, Experimental study of three-phase bubble column using phase-Doppler anemometry, Advances in fluid mechanics & heat & mass transfer, 2012, pp. 349–352. [22] M.K. Thet, C.-H. Wang, R.B.H. Tan, Experimental studies of hydrodynamics and regime transition in bubble columns, Can. J. Chem. Eng. 84 (2006) 63–72. [23] T.J. Lin, J. Reese, T. Hong, L.S. Fan, Quantitative analysis and computation of two-dimensional bubble columns, AIChE J. 42 (1996) 301–318. [24] N. Kantarci, F. Borak, K.O. Ulgen, Bubble column reactors, Process Biochem. 40 (2005) 2263–2283. [25] S. Barghi, A. Prakash, A. Margaritis, M.A. Bergougnou, Flow regime identification in a slurry bubble column from gas holdup and pressure fluctuations analysis, Can. J. Chem. Eng. 82 (2004) 865–870.
Contents lists available at ScienceDirect
Powder Technology journal homepage: www.elsevier.com/locate/powtec
Flow patterns and transitions in a rectangular three-phase bubble column Weiling Li, Wenqi Zhong ⁎, Baosheng Jin, Yong Lu, Tingting He Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, School of Energy and Environment, Southeast University, Nanjing 210096, China
a r t i c l e
i n f o
Article history: Received 5 December 2013 Received in revised form 19 February 2014 Accepted 1 April 2014 Available online 12 April 2014 Keywords: Flow pattern Flow regime transition Gas–liquid–solid flow Particle effect Three-phase bubble column
a b s t r a c t The flow patterns and transitions in a rectangular gas–liquid–solid three-phase bubble column were studied. The influence of solid volume fraction, particle size and particle density on the flow regime transitions of the threephase bubble column was investigated experimentally. Experiments were carried out for solid volume fraction Vs = 0.03–0.3, average particle size dp = 48 μm–270 μm, particle density ρp = 2500 kg/m3–4800 kg/m3, and superficial gas velocity Ug = 0.007 m/s–0.7 m/s in a rectangular bubble column measured 0.8 m tall, 0.1 m long and 0.01 m wide. Four distinct flow patterns and three transition points were observed in this experimental system, and the four flow regimes were discrete bubble regime, transition regime, bubble coalescence regime and strong turbulent regime, which were determined on the basis of criteria as well as schematic diagrams and typical flow pattern images obtained from a high-resolution digital charge couple device (CCD) camera. Typical flow patterns maps were plotted for illustrating flow regime transitions under different particle conditions. It was found out that particle volume fraction and particle density had an effect on the flow pattern transitions; when increasing the values of particle volume fraction and the particle density respectively, the values of the flow regime transition points all decreased. Particle size had a little effect on flow regime transition points when the particle size shifted in the range of 48 μm to 150 μm; when particle size was larger than 150 μm and increased to 270 μm, the operation ranges of the transition regime and the bubble coalescence regime decreased. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Gas–liquid–solid three-phase bubble columns have been widely utilized in chemical, petrochemical, energy conversion, biochemical and environmental processes [1–5], such as direct coal liquefaction, Fischer–Tropsch (FT) synthesis, CO2 removal from flue gases in a bubble column [6], heavy oil hydrocracking using catalyst [7], and sodium bicarbonate production [8]. However, there are still some difficulties on the bubble column scale-up technology for industry utilization due to the lack of comprehensive knowledge on the flow characteristics. Flow patterns and transition are two of the important hydrodynamic characteristics of the gas–liquid–solid bubble column, and the mass and heat transfer, mixing efficiency, pressure gradient, momentum loss, pipe vibration and reactor volume productivity of these systems all vary greatly with different flow patterns. Hence, flow regime and transition study is an important respect on understanding the flow characteristics of three-phase bubble columns, and it provides
⁎ Corresponding author at: School of Energy and Environment, Southeast University, Sipailou 2#, Nanjing 210096, Jiangsu, China. Tel.: + 86 25 83794744; fax: + 86 25 83795508. E-mail addresses: [email protected] (W. Li), [email protected] (W. Zhong), [email protected] (B. Jin).
http://dx.doi.org/10.1016/j.powtec.2014.04.002 0032-5910/© 2014 Elsevier B.V. All rights reserved.
fundamental knowledge on bubble column design, optimization, operation and scale-ups. Flow regime transition is usually influenced by many factors, such as column geometry, gas distributor design, gas flow rate and operating conditions (temperature, pressure and the physical properties of gas, liquid and particle) [9,10]. Numerous studies have investigated flow patterns in bubble columns, some studies were focused on gas–liquid two-phase systems, and some investigated three-phase systems. The shapes of bubble columns mainly included cylindrical and rectangular two types [11]. When the liquid phase operated in batch mode, Shaikh et al. pointed out that four types of flow patterns had been observed, and they are homogeneous (bubbly), heterogeneous (churn-turbulent), slug and annular flow [4]. Three flow regimes with two flow regime transition points have been identified [12–14], and they were homogeneous regime, transition regime, and the heterogeneous regime. Angeles et al. [7] summarized three types of flow regimes in bubble column reactor, viz., homogeneous bubble flow, heterogeneous bubble flow and churn turbulent slug flow regime. Some scholars [10,15] studied two principal flow regimes, homogeneous and heterogeneous regimes in bubble column systems. Nedeltchev et al. [16] studied a bubble column and identified five flow regimes, and they were: a dispersed bubble regime, first and second transition regimes, a coalesced bubble regime consisting of four regions (called 4-region flow) and a coalesced bubble regime consisting of three regions (called 3-region flow). When
28
W. Li et al. / Powder Technology 260 (2014) 27–35
the liquid phase and the gas flow were operated in co-current mode, Shiea et al. [17] detected dispersed bubble flow, discrete bubble flow, coalesced bubble flow and slug flow regimes in their experimental bubble column. Most of these researches were conducted in cylindrical bubble columns, and therefore rectangular bubble columns needed more studies. A rectangular bubble column with a width of 200 mm and a depth of 15 mm was utilized to study the turbulence characterization under both homogeneous and heterogeneous regimes [18]. Particle properties influence the flow regime transitions, and some studies have been performed on the particle effect. Mena et al. [15] discussed the solid effect on homogeneous–heterogeneous flow pattern transition in a bubble column, and provided detailed suggestions on the explanation of the dual solid effect observed. Rabha et al. [19] studied the effect of particle size on the intrinsic flow behavior in a slurry bubble column in order to provide reliable guidelines for optimizing catalyst particle size and solid concentration. Gan [20] studied the solid circulation velocities and holdup profiles in a gas–liquid–solid bubble column, and analyzed the effect of the particle addition on the gas–liquid flow and the three-phase flow hydrodynamics in bubble columns [21]. There is still a lack of detailed understanding of the effect of particles on regime transition, such as the particle size and density effect on the flow structure at different operating conditions [22]. The objective of the study was to investigate the flow patterns in a rectangular bubble column through experiments, and to study the effect of the particle volume fraction, particle size and particle density on the flow regime transition points. This paper is structured as follows. Following the introduction, experimental work is described in the
second part. Subsequently, four flow patterns and three transition points are detailed in the third part; flow regime maps are analyzed, followed by concluding remarks.
2. Experimental section A gas–liquid–solid three-phase bubble column experimental system was built, and the schematic diagram of the experimental set-up was sketched in Fig. 1. The experimental system consists of four sections: a gas–liquid–solid three-phase bubble column, an air supply system, a differential pressure signal acquisition system and a digital image acquisition system. The bubble column measured 0.8 m tall, 0.1 m long and 0.01 m wide and was made of transparent, 6-mm-thick Plexiglas. The air supply system consists of an air compressor, pipelines and flowmeters. The differential pressure signal acquisition system includes the diffused silicon pressure transmitters (GB-3000E and GB-3000HK), a data acquisition card (CDAQ-9188), power supply, positive wire, negative wire and data collection software (Labview 2010 Signal Express). The measurement range and measurement accuracy of the diffused silicon pressure transmitters are 0–2.5 kPa and 0.05%, respectively. The digital image acquisition system consists of a high resolution digital CCD camera and a computer for photo collection. The high-speed digital camera used features a high-speed consecutive shooting mode and shoots 16 frames/s in JPEG mode and thus could be used to capture a series of gas–liquid–solid flow structure images. A 1000-W floodlight was used for lighting and to ensure high photograph resolution.
Fig. 1. A schematic diagram of the experimental set-up. 1. Computer; 2. Power supply; 3. Data collector; 4. Differential pressure sensor; 5. Gas outlet; 6. A three-phase bubble column; 7. Metal sintered plate distributor; 8. Flowmeters; 9. Light; 10. CCD camera.
W. Li et al. / Powder Technology 260 (2014) 27–35
Air, water and particles were used as the gas, liquid and solid phases respectively. Air was supplied from an air compressor, and then it was measured by three flowmeters and passed through a metal sintered plate distributor into the bubble column, and the average diameter ds of holes in the sintered plate was 50 μm. Tap water was operated in batch mode. The properties of particles were shown in Table 1. Five kinds of particles were used in experiments. Glass beads A–D were particles with different particle sizes, but had the same density; iron powder had the largest particle density and had the same average particle size with the glass beads B. The solid phase volume fractions tested in these experiments were 3%, 9%, 20% and 30%. In Table 2, experimental and sampling conditions were listed. The superficial gas velocity used in the study was over the range of 0.007 m/s–0.7 m/s. When the liquid and solid phases were prepared in the bubble column, it was a static bed. Then air as the gas phase entered into the bed, and gas–liquid–solid flow was formed. Experiments were carried out at ambient temperature and atmospheric pressure. Three-phase flow patterns operated at different superficial gas velocities in the rectangular bubble column were recorded and photographs were used to be analyzed. The experiments were repeated under different conditions, operating at different solid phase volume fractions, particle size and densities. 3. Results and discussion
29
Table 2 Experimental and sampling conditions. H0 (mm)
Ug (m/s)
f (Hz)
t (s)
120
0–0.7
1000
60
(a1)
(a2)
(a3)
(a4)
(b)
(c)
(d)
(e)
3.1. Identification of flow patterns As illustrated in the experimental section, the gas flow rate was adjusted and increased stepwise by step, gas–liquid–solid three-phase flow structures formed and evolved gradually. Four flow regimes and three transition points were observed in this experiment system. A schematic representation of these flow regimes were shown in Fig. 2, which depicted the main characteristics of these flow patterns. The flow patterns and transition points included: (a) discrete bubble regime, DBR; (b) the first transition point, FTP; (c) transition regime, TR; (d) the second transition point, STP; (e) bubble coalescence regime, BCR; (f) the third transition point, TTP and (g) strong turbulent regime, STR. The criteria as well as typical images for the description of every flow pattern are presented as follows. 3.1.1. Discrete bubble regime In this flow pattern, bubbles disperse in the liquid–solid flow, which is operated at low superficial gas velocity. Bubbles have a spherical shape, all bubbles have the similar size, and this flow regime was also demonstrated by Mota et al. [10] and Mena et al. [15]. When the superficial gas velocity increases to 0.025 m/s, the bubble size becomes larger, bubbles have ellipsoidal shape, and more numbers of bubbles appeared, whereas all bubbles still distribute uniformly in the liquid–solid flow in this flow regime. Liquid phase has continuous distribution, and the distribution of solid phase depends on particle size and density which is illustrated in Fig. 2a1–a4. The bubble size was only influenced by the gas sparger in this regime [16]. Fig. 3 depicts the typical flow pattern images of discrete bubble regime recorded by digital CCD camera. All the images in Fig. 3 were collected from different gas–liquid–solid systems, such as, glass beads A was used as the solid phase in Fig. 3(a1) and Vs = 0.03; glass beads
Table 1 Particle properties. Particles
dp (μm)
ρs (kg/m3)
Glass beads A Glass beads B Glass beads C Glass beads D Iron powder
48 75 150 270 75
2500 2500 2500 2500 4800
(f)
(g1)
(g2)
Fig. 2. Schematic representation of various flow patterns and transition points: (a1–a4) discrete bubble regime, DBR; (b) the first transition point, FTP; (c) transition regime, TR; (d) the second transition point, STP; (e) bubble coalescence regime, BCR; (f) the third transition point, TTP; (g1–g2) strong turbulent regime, STR.
30
W. Li et al. / Powder Technology 260 (2014) 27–35
(a1) (a1)
(a2)
(a2)
(a3)
(a3)
Fig. 3. Typical flow pattern images of discrete bubble regime (DBR): (a1) Glass beads A, particle volume fraction, V s = 3%, U g = 0.036 m/s; (a2) Glass beads B, V s = 20%, U g =0.047 m/s; (a3) Glass beads C, Vs = 20%, U g = 0.014 m/s.
B and Vs = 0.2 was demonstrated in Fig. 3(a2). Although they had different particle size and volume fraction, they showed similar flow structures and were in the same flow regime. It can be observed that both of particle sizes and particle volume fractions will affect the three-phase flow characteristics, such as bed expansion heights and bubble sizes. 3.1.2. The first transition point In this case, deformed bubbles appear firstly; some small ellipsoidal bubbles lose their stability, coalesce and form bubble blocks and larger bubbles in the local region. Bubbles near the sidewall keep an ellipsoidal shape. Particles move with the liquid phase. When the gas–liquid–solid three-phase flow shows these characteristics, it means that the flow pattern changes from the discrete bubble regime to the transition regime. Fig. 4(a1–a3) depicts the typical flow pattern images of the first transition point in this experimental system. The bubble blocks are defined as the bubbles that have irregular shapes and have larger sizes than that surrounded mainly caused by the coalescence of small size bubbles which distribute uniformly in the homogeneous flow regime. Bubble blocks move upward near the central axis. The close-up of the bubble blocks photos was shown in Fig. 4(a4), which was illustrated by the two blue circles. 3.1.3. Transition regime In this flow pattern, bubble block behaviors become more obvious. The volume of the bubble blocks becomes even larger. Bubbles have coalescence phenomena and move upward in the bed region. The liquid phase distributes uniformly together with bubbles and moves upward. Particles, such as glass beads A, glass beads B and C all distribute uniformly in the bubble column, which can be elucidated in Fig. 5. This flow regime was also described in previous studies [14]. Ruthiya et al. [12] gave the schematic representation of this regime. Compared with the transition regime characteristics described by
(a4) Fig. 4. Typical flow pattern images of the first transition point (FTP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.058 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.058 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.047 m/s; (a4) the close-up of bubble blocks photos.
Ruthiya et al. [12], the central bubble plume appeared in the bubble column, but the liquid eddy was not observed obviously in this regime. 3.1.4. The second transition point In this point, liquid circulation appears firstly in local regions as shown in Fig. 2(d); larger bubbles form and eddies appear caused by the onset of bubble coalescence, liquid circulation and larger liquid momentum transferred by the gas phase. The quantity of small bubbles reduces. The system has stronger coalescence phenomena and from this point the gas–liquid–solid three-phase flow shifts from the transition regime to the heterogeneous flow regime, which was illustrated in Ref. [14]. The typical flow pattern images of the second transition point were highlighted in Fig. 6. 3.1.5. Bubble coalescence regime In this case, bubbles have a relatively regular movement, small bubbles forming at the distributor region, and some of them rise into the central region of the bubble column, coalesce and form large bubbles which then rise up in the bed. These large bubbles have an irregular shape. The numbers of small bubbles reduce in this flow pattern, and bubbles have large size distributions. Fig. 7 presents the typical flow pattern images of the bubble coalescence regime. This flow regime could be
W. Li et al. / Powder Technology 260 (2014) 27–35
(a1)
(a2)
(a3)
Fig. 5. Typical flow pattern images of transition regime (TR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.089 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.11 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.13 m/s.
observed more obviously when glass beads B was used as the solid phase, which was depicted in the Fig. 7(a2). The liquid phase moved upward in the middle portion of the column to the surface and then descended adjacent to the sidewalls building the gross liquid
(a1)
31
(a2)
(a3)
Fig. 7. Typical flow pattern images of bubble coalescence regime (BCR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.16 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.18 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.17 m/s.
circulation. The existence of a solid phase and the volume ratio of gas and liquid phases have an effect on shaping this flow regime. Nedeltchev et al. [16] described that the flow pattern shifted to coalesced bubble (3-region flow) when the superficial gas velocity was larger than 0.13 m/s at atmospheric pressure in a gas–liquid two-phase bubble column. The bubble coalescence regime from this experimental results and the coalesced bubble regime (3-region flow) investigated by Nedeltchev et al. [16] show similar flow structure characteristics. The coalesced bubble regime (3-region flow) consists of the central fast bubble region, the vortical and the descending regimes [23]. Bubble coalesced and broke up in the central fast bubble region, and bubble clusters and coalesced bubbles moved upward in this region. The vortical flow was surrounding the central fast bubble region, and vortices existed in gas–liquid–solid three-phases. The descending regime was located adjacent to the column wall; three-phases streams moved downward. These characteristic can be observed in Fig. 7. 3.1.6. The third transition point In this case, bubbles have separation behaviors, and they have several ways to rise upward and do not coalesce in the middle portion which is distinguished from the bubble coalescence regime. Bubbles have more than two routes rising to the surface. Typical flow structure images of the third transition point were depicted in Fig. 8. Analyzing the pressure time-series collected from this flow regime with the time–frequency domain analysis methods, the time–frequency distributions were distinguished from that of previous flow regimes. Therefore, it was necessary to distinguish the flow pattern which was operated at higher superficial gas velocity from the previous flow regimes.
(a1)
(a2)
(a3)
Fig. 6. Typical flow pattern images of the second transition point (STP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.13 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.16 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.16 m/s.
3.1.7. Strong turbulent regime In this case, the bubble column usually operates above 0.25 m/s. At high gas velocity, liquid has large momentum, and the gas–liquid– solid three-phase flow shows strong turbulent flow structures. Bubbles take a large volume of the bed as the volume ratio of gas and liquid phases increases further. Bubbles coalesce and break up strongly in
32
W. Li et al. / Powder Technology 260 (2014) 27–35
0.6
Particle volume fraction
Particle:Glass beads A dp= 48µm 0.4
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
0.2 DBR
BCR
TR
STR
0.0 0.0
0.2
0.1
0.6
0.7
Superficial gas velocity (m/s) Fig. 10. Flow regime map of gas–liquid–solid bubble column systems at different particle volume fractions (dp = 48 μm).
studies on three-phase bubble columns [10,12], the authors did not mention this flow pattern due to their gas velocities studied varied in the range of 0–0.3 m/s and the gas velocities studied were not larger than 1 m/s [24].
(a1)
(a2)
(a3)
Fig. 8. Typical flow pattern images of the third transition point (TTP): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.2 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.22 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.22 m/s.
this regime. The liquid phase has high energy, discontinuous distribution and is in a very turbulent pattern. The liquid flow pattern is much more chaotic than that in previous regimes. Liquid phase and gas phase have spatial and temporal stochastic distributions. Fig. 9 illustrated the flow pattern images of a strong turbulent regime. In the previous
3.2. Flow regime map and flow pattern transition In gas–liquid–solid three-phase bubble columns, the system has different flow patterns when changing the gas flow rate. Four flow regimes and three transition points were observed at different superficial gas velocities as discussed in the previous part. Several factors may influence the transitions of different flow patterns, such as solid phase volume fractions, particle size and particle density. Studies were carried out to find out how the flow regime transition points shifted when all these factors changed. Hence, flow regime maps were plotted with all these factors, and the operation ranges of these flow patterns were illustrated and identified. 3.2.1. The effect of particle volume fraction Glass beads were utilized as the solid phase, and the particle volume fractions were 3%, 9%, 20% and 30%. Figs. 10 and 11 depicted the flow regime map under different particle volume fraction. The superficial gas velocity Ug was plotted on the abscissa axis and the particle volume fraction was plotted on the ordinate axis. As shown in Fig. 10, the particle size was 48 μm and Vs equaled to 0.03; the superficial gas velocity of 0.6 Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
Particle volume fraction
Particle:Glass beads C dp=150µm 0.4
0.2 DBR TR
BCR
STR
0.0
(a1)
(a2)
(a3)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Superficial gas velocity (m/s) Fig. 9. Typical flow pattern images of strong turbulent regime (STR): (a1) Glass beads A, particle volume fraction, Vs = 3%, Ug = 0.24 m/s; (a2) Glass beads B, Vs = 9%, Ug = 0.24 m/s; (a3) Glass beads C, Vs = 9%, Ug = 0.83 m/s.
Fig. 11. Flow regime map of gas–liquid–solid bubble column systems at different particle volume fractions (dp = 150 μm).
W. Li et al. / Powder Technology 260 (2014) 27–35
33
Table 3 Summary of various Ug,c values. Ug,c (m/s), dp = 48 μm
Transition points
The first transition point The second transition point The third transition point
Ug,c (m/s), dp = 150 μm
3%
9%
20%
30%
3%
9%
20%
30%
0.058 0.13 0.2
0.047 0.13 0.18
0.047 0.13 0.18
0.047 0.11 0.18
0.058 0.18 0.24
0.047 0.16 0.22
0.036 0.11 0.18
0.036 0.089 0.16
Table 4 Comparison with literature data. Investigators
System
Bubble column type
The transition points (m/s)
Ruthiya et al. [12]
Nitrogen gas/silica, carbon/demineralized water, dp = 44 μm, 30 μm Air/Therminol LT
Rectangular bubble column
Ug,c1 = 0.06 m/s, Ug,c2 = 0.12 m/s.
Cylindrical bubble column
Air/water, studies of Bach and Pilhofer Air/water, studies of Oels et al. Air/tap water/glass beads, dp = 35 μm Air/tap water/glass beads, dp = 48 μm
– – Cylindrical bubble column Rectangular bubble column
Ug,c1 = 0.02 m/s, Ug,c2 = 0.08 m/s, Uc3 = 0.10 m/s, Uc4 = 0.13 m/s. Ug,c1 = 0.046 m/s. Ug,c1 = 0.039 m/s. Ug,c1 = 0.05 m/s, Ug,c2 = 0.125 m/s. Ug,c1 = 0.047 m/s, Ug,c2 = 0.13 m/s, Ug,c3 = 0.18 m/s.
Nedeltchev et al. [16] Kantarci et al. [24] Barghi et al. [25] This studies
Subscripts: g,c1—the first transition point; g,c2—the second transition point; g,c3— the third transition point; c3—the point shifting to coalesced bubble (4-region flow); c4—the point shifting to coalesced bubble (3-region flow).
the first transition point was 0.058 m/s. The second transition point was at 0.13 m/s and the third transition point was at 0.20 m/s. All the transitional gas velocity values were summarized in Table 3. When the particle volume fraction shifted from 0.03 to 0.09, the first and the third transitional gas velocities decreased; when the particle volume fraction increased further, shifting from 0.09 to 0.3, these two transition points kept the same values. The second transition point stayed at 0.13 m/s with the particle volume fraction of 0.03, 0.09 and 0.2, but the value decreased when that shifted from 0.2 to 0.3. Fig. 11 presented the flow regime map of the three-phase system with a particle size of 150 μm. When the particle volume fraction increased, shifting from 0.03 to 0.2, all the three transition points decreased, such as the second transition point decreased to 0.156 m/s at Vs = 0.09, which was identified in reference [14]. The reasons may be due to the high particle volume fraction which would increase the instability of the system, and therefore make the system shift into a turbulent pattern at lower gas velocities. It can be concluded that increasing the particle volume fraction can reduce three transitional gas velocity values, and this trend was obviously observed in Fig. 11 with dp = 150 μm. This conclusion was in agreement with the results of Mota et al. [10]. The experimental results of transitional gas velocities have been compared with literature data, which was shown in Table 4. The
transition point results of the rectangular and cylindrical two types have also been quantitatively compared. Comparing the results of this study with that of Ruthiya et al. [12] and Barghi et al. [25] in Table 4, the first and second transitional gas velocity values had good agreement, and the bubble column type did not have a large influence on the transition gas velocities.
3.2.2. The effect of particle size Glass beads were utilized as the solid phase, and the average particle sizes were 48 μm, 75 μm, 150 μm and 270 μm. Fig. 12 demonstrated the effect of particle size on the flow regime map of gas–liquid–solid bubble column systems. The superficial gas velocity Ug was plotted on the abscissa axis and the particle size was plotted on the ordinate axis. When the particle size was 75 μm and the particle volume fraction was 0.09, the first transition point was at 0.058 m/s; the second transition point was at 0.16 m/s and the third transition point was at 0.22 m/s. As observed from Fig. 12, the first transition point was not affected by the particle size, and the second transition point kept the same value when the particle size shifted from 48 μm to 150 μm. When the particle size increased from 150 μm to 270 μm, the values of the second and the third transition points decreased, system shifting into the next
500
9000
400 350 300 250 200
DBR
TR
BCR
STR
150
Particle:Glass beads C Iron powder dp=75µm
8000
Particle density (kg/m3)
Particle diameter (µm)
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
Particle:Glass beads Vs=9%
450
7000 6000
Discrete bubble regime First transition point regime Transition regime Second transition point regime Bubble coalescence regime Third transition point regime Strong turbulent regime
5000 4000
DBR
TR
BCR
STR
100 3000 50 0.0
0.1
0.2
0.6
0.7
Superficial gas velocity (m/s) Fig. 12. Flow regime map of gas–liquid–solid bubble column systems at different particle diameter (Vs = 9%).
0.0
0.1
0.2
0.6
0.7
Superficial gas velocity (m/s) Fig. 13. Flow regime map of gas–liquid–solid bubble column systems at different particle density (dp = 75 μm, Vs = 9%).
34
W. Li et al. / Powder Technology 260 (2014) 27–35
flow regime at a lower superficial gas velocity. It can be concluded that the particle size had a little effect on the operation range of flow patterns at the particle size of 48 μm–150 μm. However, the particle size can narrow down the operation ranges of transition regime and bubble coalescence regime when the particle size was larger than 150 μm. The three-phase system had the lowest gas velocity for shifting into the strong turbulent regime when the particle size was 270 μm.
Static bed
BDR
3.2.3. The effect of particle density Glass bead and iron powder were utilized as the solid phase, and the particle densities were 2500 kg/m3 and 4800 kg/m3, respectively. Fig. 13 depicted the effect of particle density on the flow regime map of gas– liquid–solid bubble column systems. As shown in Fig. 13, the superficial gas velocity Ug was plotted on the abscissa axis and particle density was plotted on the ordinate axis. Particle size and the particle volume
TR
BCR
STR
BCR
STR
(a)
Static bed
BDR
TR
(b) Fig. 14. Flow regime images of gas–liquid–solid bubble column systems at different particle density: (a) Glass beads B, Ug = 0, 0.025 m/s, 0.11 m/s, 0.18 m/s, 0.33 m/s; (b) Iron powder, Ug = 0, 0.025 m/s, 0.058 m/s, 0.18 m/s, 0.42 m/s. (dp = 75 μm, Vs = 9%).
W. Li et al. / Powder Technology 260 (2014) 27–35
fraction kept the same values which were equaled to 75 μm and 0.09 respectively, whereas the particle density changed. It can be observed that all the three transition points decreased when particle density increased. Fig. 14 showed the static bed and four flow regime images of two systems, which utilizes glass bead and iron powder respectively as the particle phase. Observing the images of the transition regime in Fig. 14, particle density had large effect on the flow structure, bed expansion height, bubble size distributions and the turbulent structures which were operated at high gas velocity. Bed pressure drop of the column also changed when increasing particle density. 4. Conclusions Experimental studies on the flow pattern and transitions were carried out in a rectangular gas–liquid–solid three-phase bubble column. Flow pattern and transitions under different particle conditions were investigated. Typical flow patterns by certain criteria as well as schematic diagrams and flow pattern images photographed by a high-resolution digital CCD camera were presented. Flow pattern maps were plotted for describing the flow regime transitions under different particle conditions. The key conclusions are as follows: (1) Four distinct flow patterns and three transition points were identified in the experimental system, and they were discrete bubble regime, the first transition point, transition regime, the second transition point, bubble coalescence regime, the third transition point and strong turbulent regime. (2) When increasing the solid volume fraction in a three-phase bubble column, shifting from 0.03 to 0.3, the superficial gas velocities of the three transition points decreased. (3) When the particle size shifted from 48 μm to 150 μm, it had a little effect on the operation ranges of flow patterns. When the particle size increased from 150 μm to 270 μm, the values of the second and the third transitional gas velocities decreased. (4) When increasing the particle density, shifting from 2500 kg/m3 to 4800 kg/m3, the superficial gas velocities of the three transition points decreased. The three-phase system had different bed expansion height, bubble size distributions and turbulent structures shown at a high gas flow rate when changing particle density. Nomenclature BCR bubble coalescence regime dp average particle size, μm ds the average diameter of holes in sparger, μm DBR discrete bubble regime FTP the first transition point Qg volumetric gas flow rate, m3/h STP the second transition point STR strong turbulent regime TR transition regime TTP the third transition point Ug superficial gas velocity, m/s U superficial velocity, m/s Vs particle volume fraction
Greek letter ρ density of phases, kg/m3
subscripts g gas g,c critical value of gas phase
l p s
35
liquid particle solid
Acknowledgments The authors would like to thank the National Natural Science Fund for Distinguished Young Scholars (No. 51325601), the Major State Basic Research Development Program of China (973 Program, No. 2010CB732206) and the Scientific Research Foundation of Graduate School of Southeast University (YBJJ1225) for the financial support. References [1] H. Li, A. Prakash, Influence of slurry concentrations on bubble population and their rise velocities in a three-phase slurry bubble column, Powder Technol. 113 (2000) 158–167. [2] B. Gandhi, A. Prakash, M.A. Bergougnou, Hydrodynamic behavior of slurry bubble column at high solids concentrations, Powder Technol. 103 (1999) 80–94. [3] Y.C. Li, Fluidization Processing Engineering Theory, Science Press, Beijing, 2008. [4] A. Shaikh, M.H. Al-Dahhan, A review on flow regime transition in bubble columns, Int. J. Chem. React. Eng. 5 (2007). [5] L.-S. Fan, Gas–Liquid–Solid Fluidization Engineering, Butterworths, Boston, 1989. [6] L. Cheng, T. Li, T.C. Keener, J.Y. Lee, A mass transfer model of absorption of carbon dioxide in a bubble column reactor by using magnesium hydroxide slurry, Int. J. Greenhouse Gas Control 17 (2013) 240–249. [7] M.J. Angeles, C. Leyva, J. Ancheyta, S. Ramírez, A review of experimental procedures for heavy oil hydrocracking with dispersed catalyst, Catal. Today 220–222 (2014) 274–294. [8] B.t. Haut, V. Halloin, T. Cartage, A. Cockx, Production of sodium bicarbonate in industrial bubble columns, Chem. Eng. Sci. 59 (2004) 5687–5694. [9] S. Nedeltchev, A. Shaikh, M. Al-Dahhan, Flow regime identification in a bubble column via nuclear gauge densitometry and chaos analysis, Chem. Eng. Technol. 34 (2011) 225–233. [10] A. Mota, A.A. Vicente, J. Teixeira, Effect of spent grains on flow regime transition in bubble column, Chem. Eng. Sci. 66 (2011) 3350–3357. [11] M. Pourtousi, J.N. Sahu, P. Ganesan, Effect of interfacial forces and turbulence models on predicting flow pattern inside the bubble column, Chem. Eng. Process. Process Intensif. 75 (2014) 38–47. [12] K.C. Ruthiya, V.P. Chilekar, M.J.F. Warnier, J. van der Schaaf, B.F.M. Kuster, J.C. Schouten, J.R. van Ommen, Detecting regime transitions in slurry bubble columns using pressure time series, AIChE J. 51 (2005) 1951–1965. [13] C. Vial, S. Poncin, G. Wild, N. Midoux, A simple method for regime identification and flow characterisation in bubble columns and airlift reactors, Chem. Eng. Process. Process Intensif. 40 (2001) 135–151. [14] W.L. Li, W.Q. Zhong, B.S. Jin, R. Xiao, T.T. He, Flow regime identification in a threephase bubble column based on statistical, Hurst, Hilbert–Huang transform and Shannon entropy analysis, Chem. Eng. Sci. 102 (2013) 474–485. [15] P.C. Mena, M.C. Ruzicka, F.A. Rocha, J.A. Teixeira, J. Drahoš, Effect of solids on homogeneous–heterogeneous flow regime transition in bubble columns, Chem. Eng. Sci. 60 (2005) 6013–6026. [16] S. Nedeltchev, A. Shaikh, M. Al-Dahhan, Flow regime identification in a bubble column based on both statistical and chaotic parameters applied to computed tomography data, Chem. Eng. Technol. 29 (2006) 1054–1060. [17] M. Shiea, N. Mostoufi, R. Sotudeh-Gharebagh, Comprehensive study of regime transitions throughout a bubble column using resistivity probe, Chem. Eng. Sci. 100 (2013) 15–22. [18] M. Sathe, J. Joshi, G. Evans, Characterization of turbulence in rectangular bubble column, Chem. Eng. Sci. 100 (2013) 52–68. [19] S. Rabha, M. Schubert, U. Hampel, Intrinsic flow behavior in a slurry bubble column: a study on the effect of particle size, Chem. Eng. Sci. 93 (2013) 401–411. [20] Z.W. Gan, Holdup and velocity profiles of monosized spherical solids in a threephase bubble column, Chem. Eng. Sci. 94 (2013) 291–301. [21] Z.W. Gan, S.C.M. Yu, A.W.K. Law, Experimental study of three-phase bubble column using phase-Doppler anemometry, Advances in fluid mechanics & heat & mass transfer, 2012, pp. 349–352. [22] M.K. Thet, C.-H. Wang, R.B.H. Tan, Experimental studies of hydrodynamics and regime transition in bubble columns, Can. J. Chem. Eng. 84 (2006) 63–72. [23] T.J. Lin, J. Reese, T. Hong, L.S. Fan, Quantitative analysis and computation of two-dimensional bubble columns, AIChE J. 42 (1996) 301–318. [24] N. Kantarci, F. Borak, K.O. Ulgen, Bubble column reactors, Process Biochem. 40 (2005) 2263–2283. [25] S. Barghi, A. Prakash, A. Margaritis, M.A. Bergougnou, Flow regime identification in a slurry bubble column from gas holdup and pressure fluctuations analysis, Can. J. Chem. Eng. 82 (2004) 865–870.