The influence of distributor structure on the solids distribution and flow development in circulating fluidized beds.
Circulating fluidized beds (CFB) have been studied intensively during the past two decades in order to continuously improve industrial processes such as Circulating Fluidization Bed Combustion (CFBC) and fluid-catalytic cracking (FCC). Due to its increasing importance, CFB riser reactor has been the subject of many studies (Yerushalmi et al., 1976; Li and Kwauk, 1980; Hartge et al., 1986; Li et al., 1988; Bai et al., 1992; Issangya et al., 1999; Guo and Werther, 2004).
The uniform distribution of solids along the riser is of importance in the successful design of riser reactors. More uniform axial and radial particle flow structure in the riser leads to shorter and more uniform solids and gas residence times, which would potentially lead to a better reactor performance (larger amounts of desired products and/or higher conversion). The flow structures in the entrance regions are critical to the overall particle distribution in the riser. Due to its importance, the distribution of the solids holdup has been the subject of a number of studies as reported in the literature (Li et al., 1988; Bai et al., 1992; Weinstein et al., 1984; Martin et al., 1992; Johnston et al., 1999; Guo and Werther, 2004). However, previous literature concentrated on the knowledge of the detailed solids flow profiles. Very limited studies (Johnston et al., 1999; Guo and Werther, 2004) have been conducted on the influence of distributor structure on the solids distribution. Johnston et al. (1999) studied solids holdup and particle velocity at the entrance region with three types of gas-solid distributor designs, but the study was in a downer, not a riser. Guo and Werther (2004) studied the flow behaviours in a CFB rectangular riser (cross section of 0.3 m x 1 m, 9 m high) with two different distributors. However, solids distribution in a cylindrical riser is quite different from that in a cubical riser. It is therefore necessary to systematically study the influence of distributor structure on the solids distribution in long cylindrical risers.
In order to gain new knowledge about the influence of distributor structure on the solids distribution in long risers, this work investigates the axial and radial solids distribution and flow development by measuring local solids holdups and particle velocities in two long cylindrical risers of different distributors with fluid cracking catalyst particles and over a wide range of operating conditions.
The experiments were conducted in two risers of different distributors with similar diameters (76 and 100 mm inner diameter). The 100 mm diameter riser was 15.1 m long and the 76 mm riser was 10 m long. The solids used in the two columns were the same FCC particles and had a mean particle diameter of 67 [micro]m and a density of 1500 kg/[m.sup.3]. The gas in these experiments was air under near-ambient conditions.
The 100 mm column is shown in Figure 1. Compared with the experimental set-ups used by many other researchers, this riser is much higher and therefore permits gas-solid flow to have a longer distance to develop. The solids in the storage tank are kept around minimum fluidization and a butterfly valve is used to control the flow of the solids into the riser. The solids flow from the storage tank into the riser distributor region where the solids are fluidized by the auxiliary air. The 100 mm solids feed pipe is angled at 40[degrees]C from the horizontal plane and centres at the height of 0.140 m on connecting to the riser. The gas distributor includes a perforated plate with 3% opening of 2.4 mm holes and a series of thirty-seven 10 mm o.d. and 6 mm i.d. tubes uniformly installed on the perforated plate. The 37 tubes give a total opening of 13%, in relation with the total riser cross sectional area. As only minimum gas flow is passed through the bottom perforated plate to maintain just above minimum fluidization, the 3% opening on that plate is not counted. The main riser air enters through a series of tubes extending 0.30 m into the riser bottom to carry the gas-solids suspension up the riser and through a smooth exit into the primary cyclone. The bottom of this cyclone connects to the distributor of the downer. The riser air from the cyclone is sent through a pair of secondary and tertiary cyclones and a bag filter for further cleaning before being exhausted to the atmosphere. From the downer distributor, the gas-solids suspension travels along the downer to the exit where the solids are separated using a fast inertial separator. The gas is then sent through a secondary and a tertiary cyclone and a bag filter for final cleaning. The downer and riser gas flow rates can be set independently from each other. The solids from the fast separator are then returned to the storage tank during normal operation. The diverting valve is used to measure [G.sub.s], when the flow of solids is switched into the measuring tank for a known amount of time. Measuring the height of the accumulated solids allows [G.sub.s] to be determined.
[FIGURE 1 OMITTED]
The 76 mm riser is paired with a 203 mm diameter riser of the same length (10 m) in a twin riser apparatus, as shown in Figure 2. The solids in the storage tank are kept around minimum fluidization and flow into the risers through the inclined pipe. The flow of solids is controlled using two separate butterfly valves for the two risers. The configurations of these twin risers are exactly the same except for the diameter. Only one riser is operated at a time. Each riser distributor was a perforated plate. The distributor has an opening area of 13% with 127, 2.4 mm i.d. holes. The 76 mm solids feed pipe is also angled at 40[degrees]C from the horizontal plane and centres at the height of 0.140 m on connecting to the riser. The gas-solids mixture travels up the riser and is separated using a series of three cyclones. The solids are returned to the storage tank and the gas is further cleaned using a bag filter. At the top of the storage tank, a section of the column with two half valves at each end is used to measure [G.sub.s]. This measuring section is a cylinder with a partition in the middle and the two half valves installed at the top and bottom of this section. When [G.sub.s] is being measured, the upper half valve diverts the solids flow down half of the cylinder, the bottom of which is sealed by the lower valve. After given time, the upper half is flipped to the other open side to allow the solids to flow through and also allow the solids collected on the measuring side to settle. The height of the accumulated solids is measured, which allows [G.sub.s] to be determined with the known time. Please note that the solids inlet arrangements for the two risers are similar: both solids feed tubes are angled at 40[degrees]C from the horizontal plane and centres at the height of 0.140 m on connecting to the riser. This shall eliminate any effect due to the positioning of the solids inlet tubes on the results from this study.
Solids concentration measurements were conducted with a reflective-type fibre optic concentration-probe. The active area in the probe tip was approximately 2 mm x 2 mm, consisting of approximately 8000 emitting and receiving quartz fibres, each having a diameter of 15 [micro]m. More details of this probe including its calibration procedure can be found from Zhang et al. (1998). Measurements were taken at 11 radial positions (r/R = 0.00, 0.16, 0.38, 0.50, 0.59, 0.67, 0.74, 0.81, 0.87, 0.92, and 0.98) and at eight levels for the two risers as listed in Table 1. A five-fibre optic velocity probe was used to measure the particle velocities. The five-fibre optic probe consists of two light emitting fibres (B and D) and three light detecting fibres (A, C, and E) arranged precisely in the same line. A particle flowing by the centre point between any two neighbouring fibres will produce a reflective signal to a detection fibre. By counting the time difference between the two signals from A-B and B-C (or C-D and D-E), the velocity of a particle passing along the array of the five fibres can be determined. Details of the five-fibre optic probe have been presented previously (Zhu et al., 2001). The particle velocity was measured at the same locations as those for the solids concentration measurements.
[FIGURE 2 OMITTED]
The superficial gas velocity, [U.sub.g], ranges from 3.5 to 8 m/s and solids circulation rate, [G.sub.s], ranges from 50 to 200 kg/([m.sup.2] s). All experiments were conducted under ambient temperature and pressure.
RESULTS AND DISCUSSION
Axial Distributions of Solids Holdup
Figure 3 shows axial profiles of cross-sectional average solids holdup in both the 15.1 m high, 100 mm-diameter riser with the multi-tube distributor (for simplicity, referred as Riser-Tube and Distributor-Tube in this paper) and the 10 m high, 76 mm-diameter riser with the perforated multi-orifice distributor (for simplicity, referred as Riser-Orifice and Distributor-Orifice in this paper). In both risers, the solids circulation rate ranges from 50 to 200 kg/([m.sup.2] s) and superficial gas velocity ranges from 3.5 to 8.0 m/s. Clearly seen from Figure 3, the axial distribution of the Riser-Tube is much more uniform than that of the Riser-Orifice, especially at the entrance region. Furthermore, the fully developed region is reached early with the multi-tube distributor. Therefore, the structure of the riser distributor has a very important influence on the axial distribution of solids holdup. At the meantime, it is also obviously shown in Figure 3 that increasing solids fluxes and/or decreasing the superficial gas velocity causes an increase in cross-sectional average solids holdup, as expected from previous reported data (Li and Kwauk, 1980; Bai et al., 1992; Martin et al., 1992; Karri and Knowlton, 1998, 1999).
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Radial Distributions of Solids Holdup
Figure 4 shows the detailed radial solids distributions at the first two levels above the distributor for both risers (Riser-Tube, with the multi-tube distributor, z = 0.95 m and z = 2.59 m; Riser-Orifice, with the multi-orifice distributor, z = 1.53 m and z = 2.73 m) under several different operating conditions. Seen from Figure 4, under the same conditions, the radial profiles of the Riser-Tube, is much more uniform than those of the Riser-Orifice in the distribution region. Comparing Figure 4a and b, the solids holdups in the wall region of the Riser-Orifice are much higher than those in the wall region of the Riser-Tube. In Figure 4c and d, the solids holdups in the wall region of the Riser-Orifice are also higher than those in the wall region of the Riser-Tube, although the difference is not as obvious as between Figure 4a and b. This is due to the distributor structure. The effect of the multi-tube distributor on the solids distribution of the riser is much better than that of the multi-orifice distributor.
Figure 5 compares radial solids distributions in the middle sections of the two risers (Riser-Tube: z = 4.51 m and z = 6.34 m; Riser-Orifice: z = 3.96 m and z = 6.34 m) under several different operating conditions. Seen from Figure 5, under the same conditions, the radial profiles of the Riser-Tube are very close to those of the Riser-Orifice. Comparing Figure 5a and b, the solids holdups in the wall region of the Riser-Orifice are just a slightly higher than those at the wall region of the Riser-Tube. As observed in Figure 5c and d, the solids holdups in the wall region of the Riser-Orifice are also close to those in the wall region of the Riser-Tube, with the difference being much less than that between Figure 4a and b. When carefully examining Figure 5a and c, the radial profiles of solids holdup for the Riser-Tube at z = 4.51 m and z = 6.34 m are almost the same, which is different from those at lower levels. This means that the radial profiles of solids holdup for the Riser-Tube at z = 4.51 m and z = 6.34 m are in the fully developed region. On the other hand, some of the radial profiles of solids holdup for the Riser-Orifice at z = 3.96 m and z = 6.34 m are still different, which indicates that the fully developed region is not reached for some operating conditions. This clearly shows that the distributor structure has great influence on the solids distribution of the riser and the flow development. The tubes in the multi-tube distributor led to high air velocity at the tube exit and also a well distributed gas flow, which leads to a better and more even particle distribution. Therefore, the particles reached the fully developed region at a lower level with the multi-tube distributor. That is, the flow development is much faster for riser distributor designed with a series of tubes.
To further analyze the solids flow development, the axial profiles of solids holdups in the three radial regions, r/R = 0.0-0.632, 0.632-0.894 and 0.894-1.0, for both risers, are plotted in Figure 6. The figure reveals the difference in the flow development in the three radial regions. In the central region (r/R = 0.0-0.632, 40% of the cross-sectional area), the solids holdup is very low and nearly constant all the way from the riser bottom to the top for both risers. There is no significant difference for the two risers in this region. In the middle region (r/R = 0.632-0.894, 40% of the cross-sectional area), the profile is still flat, except for the situation with [G.sub.s] = 100 kg/[m.sup.2] s and [U.sub.g] = 5.5 m/s for the Riser-Orifice, for which the solids holdup profile is not fiat at entrance region until approximately 4 m and then becomes flat towards the riser top. Most significant variations of the solids holdup happen in the wall region (r/R = 0.894-1.0, 20% of the cross-sectional area), where the solids concentration of the Riser-Orifice is rather high at the bottom and drops sharply with increasing height till approximately 4 m, then becomes flat towards the top. However, the profiles of solids holdup are rather flat for the Riser-Tube. That is, the flow development of the riser with the multi-tube distributor is faster than with the multi-orifice distributor. Figure G also shows that increasing [G.sub.s] significantly slows down the flow development process while increasing [U.sub.g] accelerates it. The flow develops first in the riser centre region, and then gradually and progressively closer to the wall as the solids pass through the riser.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Radial Distributions of Particle Velocity
Figure 7 compares the radial profiles of particle velocity for these two different distributors under five different conditions in the lower section of the risers (Riser-Tube: z = 0.95 m; Riser-Orifice: z = 1.53 m). This plot shows that the radial profiles of particle velocity for the Riser-Tube are more uniform than those for the Riser-Orifice. Meanwhile, with the increase of the solids circulation rate, the difference in radial profiles of particle velocity between the risers with different distributors is increased. Under the same solids circulation rate, the difference of radial profiles of particle velocity between the two risers with different distributors is almost the same for each of the three tested superficial gas velocities. In the multi-tube distributor, the tubes evenly distribute the gas, which offers the same energy to the particles at the same level. However, there are no such tubes in the multi-orifice distributor. As a result, the radial profiles of particle velocity for the Riser-Tube are more uniform than those for the Riser-Orifice.
Figure 8 compares the radial profiles of particle velocity for these two risers under five different conditions at 2.5-3.0 m above the distributor (Riser-Tube: z = 2.59 m; Riser-Orifice: z = 2.73 m). In Figure 8, the radial profiles of particle velocity for the Riser-Tube are very close to those for the Riser-Orifice. This indicates that at this level, the influence of riser distributor on the solids distribution is much less than at the lower level. This further indicates that the difference in the results of Figure 7 is indeed from the influence of riser distributor. In other words, the distributor design influences only the lower section of the riser. Nonetheless, the distributor design does have a strong effect to the lower portion of the riser and is therefore an important factor for the design of circulating fluidized bed reactors.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
Figure 9 plots the local particle velocities in three radial regions on eight axial elevations for both risers. In this figure, the difference in the flow development in the three radial regions is clearly revealed. In the riser centre (r/R = 0.0-0.632, 40% of the cross-sectional area), the particles already gained a fairly high velocity before the height of 1.5 m, whereas a maximum velocity is finally reached at the height of 3-4 m. There is no significant difference for the two risers in this region. In the middle region (r/R = 0.632-0.894, 40% of the cross-sectional area), the particle velocity is increasing throughout the Riser-Orifice; however, the particle velocity of the Riser-Tube becomes constant at around 4 m in this middle region. That is, the flow development is faster in the riser with the multi-tube distributor. In the wall region (r/R = 0.894-1.0, 20% of the cross-sectional area), the particle velocity remains low up to about 4-6 m and then increases slowly towards the top of the Riser-Orifice; the particle velocity of the Riser-Tube just has a rather small change. The flow development in this region is significantly slower than the two other regions. Moreover, it is also clearly shown that the flow development of the Riser-Orifice is slow than that of the Riser-Tube in this region. Figure 9 clearly shows that particle acceleration (therefore flow development) first starts from the centre and then extends to the wall. This dictates the development of the radial profiles of particle velocity. The flow development is faster for the riser with the multi-tube distributor than with the multi-orifice distributor.
[FIGURE 10 OMITTED]
Figure 10 shows a schematic comparison of solids holdup and particle velocity distribution at the bottom of the two risers. At the bottom of the Riser-Orifice in Figure 10a, passing through the orifices, gas is roughly evenly distributed into the riser. Due to the wall effect, gas meets greater resistance in the wall region than at the centre region. More gas tends to go through the centre region of the riser and there is a sharp difference of gas velocities across the cross-section. As a result, the particles are accelerated to different degrees across the cross-section. That is, particle distribution in the Riser-Orifice is not even over the cross-section. More particles are accumulated at the wall region than at the centre region. The radial distributions of both solids holdup and particle velocity are not flat for the Riser-Orifice as shown in Figure 10a. This is a typical core-annulus structure (Bai et al., 1995).
At the bottom of the Riser-Tube in Figure 10b, passing mostly through a series of tubes, gas is evenly distributed into the riser. At the tip of each tube, the same high gas velocity is reached so that a fairly uniform gas distribution is maintained at this level. Compared with Figure 10a, the gas velocity at the wall region of the Riser-Tube is higher than that of the Riser-Orifice because the gas is distributed more evenly to the wall region through high-velocity jets from the tubes. An increase in gas velocity will lead to an increase in the turbulence of solids movement and a decrease in the tendency of cluster formation. From the viewpoint of the boundary layer theory, the thickness of boundary layer will be reduced with increasing degree of turbulence of the gas-solids flow. As a result, the core region in the Riser-Tube becomes greater than in the Riser-Orifice. Particle distribution in the Riser-Tube is much more even than that in the Riser-Orifice. That is, the radial distributions of both solids holdup and particle velocity are much flatter for the Riser-Tube than those for the Riser-Orifice, as observed in Figures 4-9.
[FIGURE 11 OMITTED]
It is worth mentioning that the above results are from two risers of slightly different sizes, so that one may question whether there is any effect of scale. This research group has conducted a series of scale-up studies using the same 76 mm and another 203 mm riser in the same laboratory (Yan and Zhu, 2004; Yan et al., 2005). Our earlier results (Yan and Zhu, 2004; Yan et al., 2005), as plotted in Figure 11, have shown that the larger diameter riser (203 mm) lead to significantly higher solids holdups in all regions of the riser. As our current results with the Riser-Tube has a larger diameter but a lower solids concentration, in reserve direction of the scale-up effect we found earlier, the differences presented in this paper are truly due to the effects of the different distributor design.
The distributor structure has an important effect on the solids distribution and flow development in the circulating fluidized beds, especially for the riser bottom section. The axial distribution of solids holdups is much more uniform for the riser with a multi-tube distributor than with a multi-orifice distributor. Meanwhile, the radial profiles of solids holdups for the riser with the multi-orifice distributor are sharper than those with the multi-tube distributor, especially at the riser bottom. The radial profiles of particle velocities for the riser with the multi-tube distributor are also more uniform than those for riser with the multi-orifice distributor. In the riser with the multi-tube distributor, both gas and particles are distributed more uniformly across the section, so that the solids acceleration and flow development is much more uniform and faster. The distributor design is an important factor for the design of circulating fluidized bed reactor.
NOMENCLATURE D riser diameter (m) [G.sub.s] overall solids circulating rate (kg/[m.sup.2]s) r/R reduced radial position [U.sub.g] superficial gas velocity (m/s) z distance from the riser distributor (m) [[epsilon].sub.s] apparent solids holdup [[bar.epsilon].sub.s] average solids holdup over the riser
The authors would like to acknowledge the National Science and Engineering Research Council of Canada for financial support.
Manuscript received March 27, 2008; revised manuscript received July 13, 2008; accepted for publication August 27, 2008.
Bai, D.-R., Y. Jin, Z.-Q. Yu and J.-X. Zhu, "The Axial Distribution of the Cross-Sectionally Averaged Voidage in Fast Fluidized Beds," Powder Technol. 71, 51 (1992).
Bai, D., E. Shibuya, Y. Masuda, K. Nishio, N. Nakagawa and K. Kato, "Distinction between Upward and Downward Flows in Circulating Fluidized Beds," Powder Technol. 84(1), 75 (1995).
Guo, Q. and J. Werther, "Flow Behaviours in a Circulating Fluidized Bed with Various Bubble Cap Distributors," Ind. Eng. Chem. Res. 43, 1756 (2004).
Hartge, E.-U., Y. Li and J. Werther, "Flow Structures in Fast Fluidized Beds," in K. Ostergaard and A. Sorensen, Eds., "Fluidization V," Engineering Foundation, New York (1986), p. 345.
Issangya, A., D. Bai, H. T. Bi, K. S. Lim, J. R. Grace and J. Zhu, "Suspension Densities in a High-Density Circulating Fluidized Bed Riser," Chem. Eng. Sci. 54, 5451 (1999).
Johnston, P. M., J. Zhu, H. I. de Lasa and H. Zhang, "Effect of Distributor Designs on the Flow Development in Downer Reactor," AIChE J. 45, 1587 (1999).
Karri, S. B. R. and T. M. Knowlton, "Flow Direction and Size Segregation of Annulus Solids in a Riser," in L.-S. Fan and T. M. Knowlton, Eds., "Fluidization IX," Engineering Foundation, New York (1998), p. 189.
Karri, S. B. R. and T. M. Knowlton, "Comparison of Annulus Solids Flow Direction and Radial Solids Mass Flux Profiles at Low and High Mass Fluxes in a Riser," in J. Werther, Ed., "Circulating Fluidized Bed Technology VI," Frankfurt, Dechema (1999), p. 71.
Li, Y. and M. Kwauk, "The Dynamics of Fast Fluidization," in J. R. Grace and J. M. Matsen, Eds., "Fluidization," Plenum Press, New York (1980), p. 537.
Li, J., Y. Tung and M. Kwauk, "Axial Voidage Profile of Fluidized Beds in Different Operating Regions," in P. Basu and J. F. Large, Eds., "Circulating Fluidized Bed Technology II," Pergamon Press, Toronto (1988), p. 193.
Martin, M. P., P. Turlier and J. R. Bernard, "Gas and Solids Behaviour in Cracking Circulating Fluidized Beds," Powder Technol 70, 249 (1992).
Weinstein, H., R. A. Graff, M. Meller and M. J. Shao, "The Influence of the Imposed Pressure Drop across a Fast Fluidized Bed," in D. Kunii and R. Toei, Eds., "Fluidization VI," Engineering Foundation, New York (1984), p. 299.
Yan, A.-J. and J.-X. Zhu, "Scale-Up Effect of Riser Reactors (1)--Axial and Radial Solids Concentration Distribution and Flow Development," Ind. Eng. Chem. Res. 43(19), 5810-5819 (2004).
Yan, A.-J., J. S. Ball and J.-X. Zhu, "Scale-up Effect of Riser Reactors (3)--Axial and Radial Solids Flux Distribution and Flow Development," Chem. Eng. J. 109(1-3), 97-106 (2005).
Yerushalmi, J., D. H. Turner and A. M. Squires, "The Fast Fluidized Bed," I&EC, Proc. Des. Dev. 15, 47 (1976).
Zhang, H., P. M. Johnston, J.-X. Zhu, H. I. de Lasa and M. A. Bergougnou, "A Novel Calibration Procedure for a Fibre Optic Concentration Probe," Powder Technol. 100, 260 (1998).
Zhu, J.-X., G.-Z. Li, S.-Z. Qin, F.-Y. Li, H. Zhang and Y.-L. Yang, "Direct Measurements of Particle Velocities in Gas-Solids Suspension Flow Using a Novel 5-Fiber Optical Probe," Powder Technol. 115(2), 184 (2001).
Aijie Yan, (1) Weixing Huang (1,2) and Jesse (Jingxu) Zhu (1) *
(1.) Department of Chemical and Biochemical Engineering, University of Western Ontario, London, Ontario, Canada N6A 5B9
(2.) School of Chemical Engineering, Sichuan University, Chengdu, Sichuan 610065, China
* Author to whom correspondence may be addressed.
E-mail address: firstname.lastname@example.org
Table 1. Locations of solids concentration measurement along the two risers 15.1 m riser 10 m riser Position Height, z, Position Height, z, axial distance from axial distance from the distributor (m) the distributor (m) 1 0.95 1 1.53 2 2.59 2 2.73 3 4.51 3 3.96 4 6.34 4 5.13 5 8.16 5 5.9 6 10.00 6 6.34 7 12.28 7 8.74 8 14.08 8 9.42
|Printer friendly Cite/link Email Feedback|
|Author:||Yan, Aijie; Huang, Weixing; Zhu, Jesse "Jingxu"|
|Publication:||Canadian Journal of Chemical Engineering|
|Date:||Dec 1, 2008|
|Previous Article:||Steam fingering at the edge of a steam chamber in a heavy oil reservoir.|
|Next Article:||Experiences in applying data-driven modelling technology to steelmaking processes.|