Tracking the Flow Dynamics in Circulating Fluidized Bed through High

Subscriber access provided by Umea University Library

Thermodynamics, Transport, and Fluid Mechanics

Tracking the Flow Dynamics in Circulating Fluidized Bed through High-speed Photographing Xiaoyang Wei, and Jesse Zhu Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.9b01751 • Publication Date (Web): 16 Aug 2019 Downloaded from pubs.acs.org on August 20, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Tracking the Flow Dynamics in Circulating Fluidized Bed through High-speed Photographing Xiaoyang Wei, Jesse Zhu* Particle Technology Research Centre, Department of Chemical and Biochemical Engineering, The University of Western Ontario, London, Ontario, Canada N6A 5B9

 Corresponding author. Tel.: +1-519-661-3807; fax: +1-519-850-2241.

E-mail address: [email protected], [email protected] (J. -X. Zhu) ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract In gas-solid circulating fluidized beds (CFBs), particle velocity and solids flux are crucial parameters for the understanding of gas-solids interaction. In the literature, the timeaverage information has been investigated intensely, but the analysis of instantaneous information remains lacking. In this work, flow behavior in a narrow rectangular circulating fluidized bed has been recorded using a high-speed camera. Then, with a verified correlation between grayscale and solids holdup, particle velocity is computed by tracking local solids holdup characteristics using two-dimensional cross-correlation, and solids flux is computed by multiplying particle velocity with particle density and local solids holdup. Time-average information in this work is found to be consistent with previous publications. Instantaneous information of particle velocity and solids flux, including relations with local solids holdup, instantaneous radial distributions, instantaneous solids circulating rates and impacts of macro solids holdup fluctuation, is, for the first time, analyzed systematically and discussed in detail. Keywords: high-speed camera; circulating fluidized bed; cross-correlation; instantaneous solids flux; instantaneous particle velocity;

ACS Paragon Plus Environment

Page 2 of 33

Page 3 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Introduction

The circulating fluidized bed (CFB) has been widely used in both chemical and energy industries, as it offers flexible control of gas/solids flow, high productivity, reduced back mixing, excellent gas-particle contact and so forth.1 In CFBs, particles are entrained upward using high-velocity gas and the flow field corresponds to overall reactor performance, including particle residence time, back mixing, solids flux, gas-solid contact and so forth.2,3 This study focuses on particle velocity and solids flux which are essential parameters for CFB hydrodynamics. Up to now, numerous studies have been conducted to measure the distributions of timeaverage particle velocity and time-average solids flux in CFB risers. The devices employed include laser Doppler velocimetry4-7, optical fiber probe2,8,9 and particle image velocimetry7-11. Many consensuses have already been reached: (1) time-average particle velocity increases from the riser wall to riser center, (2) time-average particle velocity may be negative in the riser wall, (3) time-average particle velocity increases with both the superficial gas velocity and the solids circulating rate, (4) time-average solids flux increases from the riser wall toward riser center, (5) time-average solids flux may be negative in the riser wall. However, the instantaneous flow structure in a CFB riser is very complicated due to the existence of particle aggregations.15,16 Microscopically, particles tend to form clusters due to hydrodynamics. Some clusters, known as crest clusters, are surrounded by coalesced particles and other clusters, known as trough clusters, are immersed in dispersed particles. Macroscopically, crest clusters and the coalesced particles form continuous high solids holdup region, known as the crest phase. Instantaneous particle velocity and solids flux relate directly to gas-particle interaction, but very few studies have been conducted. Instantaneous particle velocity was measured at a specific location in a dilute CFB using laser Doppler velocimetry.6,7,17 Instantaneous cluster velocity was computed with the cross-correlation of cluster signals acquired using an optical fiber probe.18 Instantaneous information of particle velocity and solids flux, including relations with local solids holdup, instantaneous radial distributions, instantaneous solids circulating 1 ACS Paragon Plus Environment


rates and impacts of macro solids holdup fluctuation, has not been discussed yet in the literature. As a result, more studies on the instantaneous particle velocity and solids flux are needed for a thorough understanding of the CFB riser. In this work, gas-solids suspension across a CFB riser was well recorded with a resolution of 190 μm using a high-speed camera. Particle velocity is computed based on two-dimensional cross-correlation of image blocks. With a verified correlation between grayscale and solids holdup19,20, solids flux is further calculated by multiplying block velocity with solids holdup and particle density. Then, time-average data was first computed and the trend, as well as magnitude, is consistent with previous studies. After that, instantaneous flow characteristics are analyzed systematically and discussed in detail.

2

Experimental 2.1 Circulating fluidized bed

The circulating fluidized bed employed in this work has a rectangular riser (19 mm thick, 114 mm wide and 7.6m high), two cyclones and a bag filter in series, a downcomer and an inclined pipe connecting the riser and downcomer (see Figure 1). In industry, risers usually serve as the main place for chemical reactions and downcomer are the auxiliary device for solids circulation. The top section of downcomer is cylindrical and can be used to measure solids circulation rate by adjusting flapper valves. The bottom section of downcomer is used to store particles. During operation, gas is introduced into the riser through a distributor and particles enter the riser through the inclined pipe. The gas-solid mixture flows concurrently upward in the CFB riser and gets separated in cyclones followed by a bag filter with gas released to atmosphere and particles eventually returned to the bottom section of the downcomer. Then, particles will enter the riser for a new cycle. Particles circulating in this CFB were FCC with 67μm in diameter and 1877kg/m3 in density. Superficial gas velocity, ranging from 3.0 to 9.0 m/s, is controlled using a rotameter, while solids circulation rate, ranging from 50 to 150 kg/m2s is regulated using a valve installed in the middle of the inclined pipe. The experiment measurement of solids circulation rate is achieved with a divider installed in the cylindrical section of the 2 ACS Paragon Plus Environment

Page 4 of 33

Page 5 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


downcomer(see Figure 1). During the measurement, the top valve is inclined 45° to the upper-left and the bottom valve stays horizontal. In this case, particles from the riser go to one side of the divider. By measuring how much particles accumulate in a given period, the solids circulation rate can be calculated with the particle bulk density. Details of this circulating fluidized bed are elaborated in previous papers.19, 20

2.2 Visualization system Vertically, there are two regions along a CFB riser, known as the acceleration region and the fully developed region. In the acceleration region, particles get accelerated by the flowing gas and the mean solids holdup decreases along the riser.21 In the fully developed region, particles have been fully accelerated and the mean solids holdup remains constant along the riser.21 As the flow in the acceleration region is apparatus sensitive, the visualization system is placed in the fully developed region of the riser in this work (H=5.33m). In a previous study, the axial distribution of this CFB system has been analyzed using an optical fiber probe and H=5.33m locates in the fully developed region.21 A lamp, installed with a 500W quartz halogen bulb, is placed facing the CFB riser and provides uniform and consistent illumination with the help of a diffusion panel (see Figure 2). A high-speed camera (MotionScopeM2 from Redlake) is placed on the other side and covered in a blank box to avoid light pollution from other sources. If particles appear in the riser, the light would be blocked, resulting in low grayscales. A calibration between the grayscale and solids holdup (see Equation 1) had been developed and validated with an optical fiber probe.19,20 Each pixel denotes a 190μm × 190 μm area physically and its solids holdup can be calculated based on the verified correlation. For the imaging method, there is always a compromise between resolution and the measurement scope. Aiming to capture the instantaneous flow dynamics across the CFB riser, the particle velocity in this work is acquired by tracing the motion of a small region not by tracing one particle. The camera, placing at 5.33m of the CFB riser, remains stationary during recording. The frame rate is set to be 2000 fps, indicating that time interval for adjacent images is 1/2000 s. The shift in the image sequence reflects the 3 ACS Paragon Plus Environment


Page 6 of 33

motion of particles in the CFB riser. In other words, particle velocity can be determined according to the flow shift between images and solids flux can be computed by considering the solids holdup. Details of this visualization system were elaborated in previous papers. 19, 20 𝐆 = 𝟐𝟖.𝟎 + 𝟐𝟐𝟖.𝟑𝐞

( ―𝟏𝟗.𝟔𝟐𝛆𝐬)

Equation 1

G is grayscale; εs is solids holdup;

2.3 Data processing The acquired image sequence was read into MATLAB and a 2-D average kernel (3 pixels × 3 pixels) is first applied to remove noise. Then, an image block, 11.4mm in height and 5.7mm in width, is picked up from an image and its information, including pixel-wise grayscales and the coordinate (x1, y1), is stored. After that, the picked image block is used as a reference to search its location in the following images. The searching scope is around the original coordinate (x1, y1), but in a large region (25 pixels more to the left and right, 50 pixels more upward, 40 pixels more downward) (see Figure 3). The maximum velocity, which can be determined with this searching scope, is 4.75 m/s horizontally, 19 m/s upward and 3.8 m/s downward. The similarity of image block and target-area is evaluated using two-dimensional cross-correlation.22 The correlation coefficient ranges from -1 to +1. The closer the correlation coefficient to +1, the more similarity the two signals share. Two signals are believed to have a good similarity if

their correlation coefficient is above 0.6.2 The probability distribution of correlation coefficients is shown in Figure 4. The distribution has an exponential shape and the majority is above 0.8, implying that the searching algorithm based on two-dimensional cross-correlation is reliable and accurate. Low correlation coefficients still occur in low probabilities. That may owe to the grayscale change caused by disintegration or collisions of particle aggregations.15 If only one most similar target-area is located in the searching scope and its correlation coefficient is higher than 0.6, the searching is effective and the coordinate (x2, y2) of target-area is stored. Particle velocity can be calculated based on 4 ACS Paragon Plus Environment

Page 7 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the pixel shift using Equation 2 and 3. Grayscales in that image block are converted to solids holdup based on a verified calibration between the grayscale and solids holdup pixel-wisely.16.17 Then, the solids flux is calculated by multiplying vertical velocity with block-average solids holdup and particle density (see Equation 4). To have a better velocity resolution, the target area is searched in the image after the adjacent image in this work. There are two direction components for both particle velocity and solids flux. In this work, it has been found that the horizontal motion is neglectable and the vertical motion is dominant, which is consistent with previous publication.17 In the following analysis, velocity and solids flux denote to their components in the vertical direction. An example of the probability density distribution has been plotted for particle velocity and solids flux, respectively (see Figure 5). The data was acquired at y/Y=0, Ug = 5 m/s, Gs = 150kg m2s. Both instantaneous particle velocity and instantaneous solids flux are not normally distributed and have wide distributions, which may relate with the appearance of particle aggregations and the macroscopic fluctuations. In the data processing, we have computed both the “Mean” average and “Median” average. They have the same lateral distributions, indicating that the choice of Mean and Median does not influence the discussion in this manuscript. However, “Median” average could eliminate the influences of extreme values and better represents the overall information. In the following discussion, the “Median” average is taken as the time-average information for each measurement location. 𝑽𝒑𝒗 = 𝑽𝒑𝒉 =

𝐲𝟏 ― 𝐲𝟐 ∆𝐭

𝐱𝟏 ― 𝐱𝟐 ∆𝐭

d

Equation 2

d

Equation 3

𝐅𝐬 = 𝐕𝐩𝐯 × 𝛆𝐬 × 𝛒𝐩 (x1, y1) is the coordinate of image block; (x2, y2) is the coordinate of the target area; d is the actual size of one pixel, [μm]; 5 ACS Paragon Plus Environment

Equation 4


∆t is the time interval, [s]; 𝑉𝑝𝑣 is the particle velocity in vertical direction, [m/s]; 𝑉𝑝ℎ is the particle velocity in the horizontal direction, vertical velocity, [m/s]; Fs is the solids flux, [kg/m²s]; ε𝑠 is solids holdup of image block; 𝜌𝑝 is particle density, [kg/m³];

3

Results and discussion 3.1 Time-average particle velocity

In the fully developed region, time-average particle velocity was plotted versus lateral positions (see Figure 6). Particle velocity increases monotonically from the wall region to the center. In the center region, particle velocity increases with superficial gas velocity and time-average particle velocity is higher than superficial gas velocity. In the wall region, particle flows downward at low gas velocity (Ug = 3 m s, Ug = 5 m s ) and upward at high gas velocities (Ug = 9 m s ). Downflowing velocity is found to be less than 2 m/s in magnitude and thickness of down-flowing annulus shrinks with increasing gas velocity. The overall lateral distribution is not symmetric, and this may owe to that particles enter the riser from one side. These characteristics are all consistent with previous studies, indicating that the analysis of the flow field is reliable in this work. 2, 2325

The apparent slip velocity is the difference between gas velocity and average particle velocity and computed using Equation 5 in this work. The apparent slip velocity was plotted versus superficial gas velocity in Figure 7. The apparent slip velocity increases with superficial gas velocity, which agrees well with previous papers. 3, 26 The reason, why apparent slip velocity increases with superficial gas velocity, may be explained with that the gas flow rate penetrating through cluster decreases with increasing superficial gas 6 ACS Paragon Plus Environment

Page 8 of 33

Page 9 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


velocity.27 However, this trend may also relate with the lateral distribution of solids holdup and further investigation is needed for a comprehensive understanding. 𝑼𝒈

𝑽𝒔𝒍𝒊𝒑 = 𝟏 ― 𝜺 ― 𝑽𝒔 𝒔

Equation 5

𝑉𝑠𝑙𝑖𝑝 is the apparent slip velocity, [m/s]; 𝑈𝑔 is the superficial gas velocity, [m/s]; 𝑉𝑠 is the lateral average velocity, [m/s]; 𝜀𝑠 is the spatial average solids holdup;

3.2 Time-average solids flux In the fully developed region, time-average solids flux was plotted versus lateral positions (see Figure 8). Time-average solids flux increases from the riser wall to the center monotonically. In the riser wall, solids flux is negative at low gas velocity and positive at high gas velocity. These are consistent with previous research.2 At Ug = 5 m s, solids flux in the center at Gs = 100kg m2s is lower than that at Gs = 150kg m2s. At Gs = 150kg m2s, solids flux at Ug = 5 m s is higher in the center and lower in the wall region comparing with that at Ug = 3 m s.

3.3 Instantaneous information Data in different operating conditions ( Ug = 3 m s, 5 m s, 7 m s,9 m s; Gs = 50kg m2s , 100kg m2s,150kg m2s) has been analyzed to draw the conclusions on the instantaneous information and the figures presented in this section are examples for demonstration. Relationship with local solids holdup Instantaneous particle velocity (r/R=0, Ug = 5m s, Gs = 150kg m2s) was plot versus its corresponding block-average solids holdup (see Figure 9). The particle velocity concentrates around 5 m/s and has a wide velocity distribution from 2.5 m/s to 6.5 m/s. 7 ACS Paragon Plus Environment


However, particle velocity has a holistic decrease at high solids holdup, indicating that particle aggregations move more slowly than dispersed particles. But, particle aggregations differ not only in solids holdup, but also in size and shape, resulting in the data in figure 9 is scattered. In the previous study, it has also been found that there is no clear relationship between cluster velocity and cluster size.18 To investigate the relationship between aggregation properties and velocity, it is suggested to take shape, size and solids holdup into consideration. Instantaneous solids flux (r/R=0, Ug = 5m s, Gs = 150kg m2s) was also plotted versus its corresponding solids holdup (see Figure 10). Solids flux is calculated by multiplying solids holdup with particle velocity and particle density. This explains why solids flux increases with solids holdup. Their relationship is close to linear for low solids holdup and tends to be chaotic for high solids holdup, indicating that gas-particle interaction has an abrupt change due to the increase of solids holdup. High solids holdup may correspond to aggregations and the velocity varies with their solids holdup, size and shape, resulting in the chaotic relationship for high solids holdup. Instantaneous lateral distribution Instantaneous solids holdup distributions across the riser have already been discussed in the previous study.15 Here, lateral distributions of instantaneous flow dynamics are discussed both for different moments and heights. Instantaneous particle velocity was computed and plotted versus lateral positions (see Figure 11). For 11a, image blocks were sampled laterally at adjacent heights in the same image (Ug = 5m s, Gs = 150kg m2s). Their instantaneous particle velocity shares a similar pattern but differs slightly in magnitude, indicating that collision happens in the riser. Generally, instantaneous particle velocity increases from the wall region to the center region. This increase is not monotonical and drops may occur, which may owe to the existence of particle aggregations. For 11b, instantaneous particle velocity was plotted versus lateral positions with a time gap of 1 second for the same height (Ug = 5m s, Gs = 100kg m2s). The instantaneous distributions share similar characteristics with the distributions in 11a. At

8 ACS Paragon Plus Environment

Page 10 of 33

Page 11 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


different moments, instantaneous lateral distributions of particle velocity change significantly in perspectives of both pattern and magnitude. Lateral distributions of instantaneous solids flux were shown in Figure 12. Solids flux increases from the riser wall to riser center and this increase is not monotonical due to the appearance of severe drops. The lateral distributions of instantaneous solids flux are more chaotic than that of instantaneous particle velocity, owing to that solids flux includes the fluctuations of both local solids holdup and particle velocity. At different moments, the instantaneous lateral distribution of solids flux also differs in both pattern and magnitude, especially in the riser center. Instantaneous solids circulating rate was further computed by taking the average of solids flux across the riser. Many operating conditions have been analyzed and the example data at Ug = 5m s, Gs = 100kg m2s is shown in Figure 13. From the image tracking, the average solids circulation rate over is 91 kg m2s, which is close to the experimental measurement (100 kg m2s). It has been found that instantaneous solids circulating rate keeps fluctuating significantly around experimental solids circulation rate. These may owe to (1) the pulsed particle feed from inclined solids return pipe, (2) the dense suspension scattered from the wall region to the center, (3) the appearance of crest phase in macroscope.15 Even for adjacent images, there are still variations of solids circulating rate, which may owe to the existence of particle clusters. The differential pressure along risers has been found to fluctuate in low frequency.28 The significant fluctuation of instantaneous solids circulating rate may be a main cause of the pressure fluctuation.

3.4 Impact of macro fluctuation Based on the previous analysis, solids holdup across the riser fluctuates significantly.15, 2833 To

investigate the impacts of solids holdup fluctuation, top 20% dense images in the

fully developed region (H=5.33m) are considered as dense moments and top 20% dilute images in the fully developed region (H=5.33m) are considered as dilute moments. The lateral distribution of time-average solids holdup for dense moments and dilute moments are shown in Figure 14. Solids holdup in dense moments is significantly higher than that in dilute moments across the riser, indicating this classification is reasonable. 9 ACS Paragon Plus Environment


Lateral distributions of time-average particle velocity have been computed for dense moments and dilute moments, respectively (see Figure 15). At Gs = 100kg m2s, particle velocity in dense moments is lower in the center region and higher in the wall region, compared with the that for dilute moments. At Gs = 150kg m2s, particle velocity in dilute and dense moments becomes similar. Lateral distributions of time-average solids flux have also been computed in the fully developed region (H=5.33m) for dense moments and dilute moments (see Figure 16). In the center, the solids flux in dense moments is much higher than that in dilute moments. In the wall region, whether the solids flux in dense moments is higher is uncertain. The particle velocity in dense moments is either low or similar than that for dilute moments (see Figure 15), but dense moments exhibit much higher solids holdup. Solids flux is the particle weight passed through unit cross-section area per second and is the product of solids holdup, particle velocity and particle density. As a result, instantaneous solids flux in dense moments is significantly higher than that in dilute moments. In the previous study, it has been found dense moments and dilute moments appear alternatingly in the CFB riser, which explains why the instantaneous solids circulation rate keeps fluctuating significantly.15

4

Conclusions

Gas-solid flow in the CFB has been well recorded using the high-speed camera. This work provides a reliable way to detect flow dynamics across the riser using twodimensional cross-correlation. Particle velocity was computed by tracking image blocks in the image sequence. With a verified correlation between solids holdup and grayscale, solids flux was calculated by multiplying particle velocity with solids holdup and particle density. First, time-average information is found to be consistent with the literature, indicating the tracking of flow dynamics in this work are reliable. Then, instantaneous flow information has been, for the first time, studied systematically. 1. In general, instantaneous particle velocity and solids flux increases from the riser wall to riser center, but not monotonically due to the existence of particle aggregations. 2. Across the CFB riser, the instantaneous lateral distribution of particle velocity and solids flux varies significantly both in pattern and magnitude. 10 ACS Paragon Plus Environment

Page 12 of 33

Page 13 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3. Instantaneous particle velocity decreases with local solids holdup holistically and instantaneous solids flux increases with local solids holdup. 4. Due to the alternative appearance of dilute moments and dense moments, instantaneous solids circulation rate fluctuates significantly. 5. At Gs = 100kg m2s, particle velocity in dense moments is lower in the center region and higher in the wall region, compared with the that for dilute moments. At Gs = 150kg m2s, particle velocity in dilute and dense moments becomes similar. 6. Solids flux in dense moments is much higher than that in dilute moments in the center region. In the wall region, whether the solids flux in dense moments is higher is uncertain.

5

Acknowledgment

The authors are grateful to the financial support from Natural Science and Engineering Research Council of Canada, and the sponsorship of the China Scholarship Council.



6

References

(1) Zhu, J.; Cheng, Y. Applications of Fluidized Bed Reactors. In Multiphase Flow Handbook; Michaelides, E., Crowe, C. T., Schwarzkopf, J. D., Eds.; CRC Press, 2016; pp.1029. (2) Wang, C.; Zhu, J.; Li, C.; Barghi, S. Detailed measurements of particle velocity and solids flux in a high density circulating fluidized bed riser. Chem. Eng. Sci. 2014, 114, 9. (3) Wang, C.; Li, C.; Zhu, J.; Barghi, S. A comparison of flow development in high density gas ‐ solids circulating fluidized bed downer and riser reactors. AlChE J. 2015, 61, 1172. (4) Yang, Y. L.; Jin, Y.; Yu, Z. Q.; Wang, Z. W. Investigation on slip velocity distributions in the riser of dilute circulating fluidized bed. Powder Technol. 1992, 73, 67. (5) Wei, F.; Lin, H.; Cheng, Y.; Wang, Z.; Jin, Y. Profiles of particle velocity and solids fraction in a high-density riser. Powder Technol. 1998, 100, 183. (6) Breault, R.W.; Yue, P.; Ludlow, J.C. Cluster particle number and granular temperature for cork particles at the wall in the riser of a CFB. Powder Technol. 2005, 149, 68. (7) Breault, R. W.; Guenther, C.; Shadle, L.J. Velocity fluctuation interpretation in the near wall region of a dense riser. Powder Technol. 2008, 182, 137. (8) Breault R.W.; Casleton, E. M.; and Guenther, C. P. Chaotic and Statistical tests on Fiber optic dynamic data taken from the riser section of a circulating fluidized bed. Powder Technol. 2012, 220, 151. (9) Pärssinen, J. H.; Zhu, J. X. Particle velocity and flow development in a long and highflux circulating fluidized bed riser. Chem. Eng. Sci. 2001, 56, 5295. (10) Matsuda, S.; Hatano, H.; Takeuchi, H.; Pyatenko, A. T.; Tsuchiya, K. Motion of individual solid particles in a circulating fluidized bed riser. In Circulating Fluidized Bed Technology V; Kwauk, M., and Li, J.H., Eds.; Science Press: Beijing, 1996; pp. 176. (11) Shi, H.; Wang, Q.; Xu, L.; Luo, Z.; Cen, K. Visualization of clusters in a circulating fluidized bed by means of particle-imaging velocimetry (PIV) technique. In Proceedings of the 9th International Conference on Circulating Fluidized Beds;Hamburg, Germany, 2008; pp. 1013. 12 ACS Paragon Plus Environment

Page 14 of 33

Page 15 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(12) Gopalan, B.; Shaffer, F. A new method for decomposition of high speed particle image velocimetry data. Powder Technol. 2012, 220, 164. (13) Shaffer, F.; Gopalan, B.; Breault, R. W.; Cocco, R.; Karri, S. R.; Hays, R.; Knowlton, T. High speed imaging of particle flow fields in CFB risers. Powder Technol. 2013, 242, 86. (14) McMillan, J.; Shaffer, F.; Gopalan, B.; Chew, J. W.; Hrenya, C.; Hays, R.; Cocco, R. Particle cluster dynamics during fluidization. Chem. Eng. Sci. 2013, 100, 39. (15) Wei, X.; Zhu, J. Capturing the instantaneous flow structure in gas-solid circulating fluidized bed using high-speed imaging and fiber optic sensing. Chem. Eng. Sci. 2019, 207, 713. (16) Zou, B.; Li, H.; Xia, Y.; Ma, X. Cluster structure in a circulating fluidized bed. Powder Technol. 1994, 78, 173. (17) Van den Moortel, T.; Azario, E.; Santini, R.; Tadrist, L. Experimental analysis of the gas–particle flow in a circulating fluidized bed using a phase Doppler particle analyzer. Chem. Eng. Sci. 1998, 53, 1883. (18) Xu, J.; Zhu, J. A new method for the determination of cluster velocity and size in a circulating fluidized bed. Ind. Eng. Chem. Res. 2011, 51, 2143. (19) Yang, J.; Zhu, J. A novel method based on image processing to visualize clusters in a rectangular circulating fluidized bed riser. Powder Technol. 2014, 254, 407. (20) Yang, J.; Zhu, J. An alternative method for mapping solids holdup in a narrow rectangular CFB riser through image calibration. Can. J. Chem. Eng. 2014, 92, 2202. (21) Xu, J.; Zhu, J. X. Experimental study on solids concentration distribution in a twodimensional circulating fluidized bed. Chem. Eng. Sci. 2010, 65, 5447. (22) Shapiro, L.; Haralick, R. Computer and robot vision ； Reading, MA: AddisonWesley, 1992. (23) Griffith, A. E.; Louge, M. Y. The scaling of cluster velocity at the wall of circulating fluidized bed risers. Chem. Eng. Sci. 1998, 53, 2475. (24) Noymer, P. D.; Glicksman, L. R. Descent velocities of particle clusters at the wall of a circulating fluidized bed. Chem. Eng. Sci. 2000, 55, 5283.



(25) Harris, A. T.; Davidson, J. F.; Thorpe, R. B. The prediction of particle cluster properties in the near wall region of a vertical riser (200157). Powder Technol. 2002, 127, 128. (26) Islam, M. A.; Krol, S.; de Lasa, H. I. Slip velocity in downer reactors: drag coefficient and the influence of operational variables. Ind. Eng. Chem. Res. 2009, 49, 6735 (27) Yunhau, Z.; Huilin, L.; Yurong, H.; Ding, J.; Lijie, Y. Numerical prediction of combustion of carbon particle clusters in a circulating fluidized bed riser. Chem. Eng. J. 2006, 118, 1. (28) Anantharaman, A.; Karri, S. R.; Findlay, J. G.; Hrenya, C. M.; Cocco, R. A.; Chew, J. W. Interpreting Differential Pressure Signals for Particle Properties and Operating Conditions in a Pilot-Scale Circulating Fluidized Bed Riser. Ind. Eng. Chem. Res. 2016, 55, 8659. (29) Brown, R. C.; Brue, E. Resolving dynamical features of fluidized beds from pressure fluctuations. Powder Technol. 2001, 119, 68. (30) Yang, T. Y.; Leu, L. P. Multiresolution analysis on identification and dynamics of clusters in a circulating fluidized bed. AlChE J. 2009, 55, 612. (31) Chew, J. W.; Hays, R.; Findlay, J. G.; Knowlton, T. M.; Karri, S. R.; Cocco, R. A.; Hrenya, C. M. Cluster characteristics of Geldart Group B particles in a pilot-scale CFB riser. I. Monodisperse systems. Chem. Eng. Sci. 2012, 68, 72. (32) Chew, J. W.; Hays, R.; Findlay, J. G.; Knowlton, T. M.; Karri, S. R.; Cocco, R. A.; Hrenya, C. M. Cluster characteristics of Geldart group B particles in a pilot-scale CFB riser. II. Polydisperse systems. Chem. Eng. Sci. 2012, 68, 82. (33) Chew, J. W.; Parker, D. M.; Cocco, R. A.; Hrenya, C. M. Cluster characteristics of continuous size distributions and binary mixtures of Group B particles in dilute riser flow. Chem. Eng. J. 2011, 178, 348.


Page 16 of 33

Page 17 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 Schematic diagram of the circulating fluidized bed



Figure 2 Schematic diagram of visualization system


Page 18 of 33

Page 19 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3 Demonstration of image tracking



Figure 4 Probability density distribution of cross-correlation coefficient


Page 20 of 33

Page 21 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

(b) Figure 5 Probability density distribution of particle velocity (a) and solids flux (b)



Figure 6 Lateral distributions of vertical velocity (H=5.33m)


Page 22 of 33

Page 23 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7 Slip velocity versus superficial gas velocity



Figure 8 Lateral distributions of solids flux (H=5.33m)


Page 24 of 33

Page 25 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9 Instantaneous particle velocity versus solids holdup



Figure 10 Instantaneous solids flux versus solids holdup


Page 26 of 33

Page 27 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

(b) Figure 11 Lateral distributions of instantaneous particle velocity (a: adjacent heights, b: time gap of 1 second)



900 800

Solids circulation rate, kg/m²s

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 33

t=0.25s

700 600

t=0.30s

500

t=0.40s

400 300 200 100 0 -100 -200 -300 -400 -1

-0.75

-0.5

-0.25

0

0.25

0.5

0.75

Radial position, y/Y

Figure 12 Lateral distributions of instantaneous solids flux


1

Page 29 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 13 Fluctuation of solids circulation rate



Figure 14 Lateral distributions of solids holdup for dense moments and dilute moments


Page 30 of 33

Page 31 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 15 Lateral distributions of particle velocity for dense moments and dilute moments



Figure 16 Lateral distributions of solids flux for dense moments and dilute moments


Page 32 of 33

Page 33 of 33 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For Table of Contents Only The gas-solid flow was tracked in the image sequence using two-dimensional crosscorrelation.


Tracking the Flow Dynamics in Circulating Fluidized Bed through High

Recommend Documents