- Email: [email protected]
density gas–solids circulating fluidized bed
Chemical Engineering Science 108 (2014) 233–243
Contents lists available at ScienceDirect
Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces
Axial and radial development of solids holdup in a high flux/density gas–solids circulating fluidized bed Chengxiu Wang a,b, Jesse Zhu a,n, Shahzad Barghi a, Chunyi Li b a b
Department of Chemical & Biochemical Engineering, University of Western Ontario, London, Ontario, Canada N6A 5B9 State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266555, China
H I G H L I G H T S
G R A P H I C A L
Experiments at high solids flux of 1000 kg/m2 s. Homogenous axial profile of solids holdup with solids holdup up to 0.32. Different radial profiles of solids holdup under high density conditions compared to low solids flux conditions. Better gas–solids contacting under high density conditions in CFBs.
Characteristics of flow structure under extremely high flux/density in a CFB riser.
art ic l e i nf o
a b s t r a c t
Article history: Received 3 October 2013 Received in revised form 25 December 2013 Accepted 27 December 2013 Available online 6 January 2014
Detailed distributions of solids holdup in an extremely high density circulating fluidized bed riser with FCC particles are mapped by an optical fiber probe. The solids circulation rate reaches as high as 1000 kg/ m2 s which has never been achieved before in an academic setting. When solids flux approaches 800 kg/ m2 s, the axial flow structure becomes uniform and the cross-sectional mean solids holdup reaches 0.22 throughout the entire riser; the same reaching 0.32 at 1000 kg/m2 s. Compared to a typical core-annulus structure, the radial distributions of the solids holdup becomes much less uniform with a shrinking core and transits to a monotonic increasing profile towards the wall. Speed of flow development differs at various radial positions with almost instant development in the center even at the highest solids flux of 1000 kg/m2 s and then becoming slower towards the wall. Fluctuations in high density circulating fluidized beds are significantly greater than those in low density ones, leading to more vigorous interactions between gas and solids phases. As a result, better gas–solids contacting and mixing, plus the uniform axial profiles of solids holdup, provide better reactor performance for the high solids flux/ density risers than low flux/density ones. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Circulating fluidized bed Riser Hydrodynamics High flux/density Solids holdup Flow development
A B S T R A C T
1. Introduction n
Corresponding author. Tel.: þ 1 519 661 3807; fax: þ1 519 850 2441. E-mail address: [email protected] (J. Zhu).
0009-2509/$ - see front matter & 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ces.2013.12.042
Circulating fluidized beds (CFBs) have been successfully used in industrial operations such as Fischer–Tropsch synthesis, partial
234
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
oxidation, fluid-catalytic cracking (FCC), and combustion (Reh, 1999; Zhu and Cheng, 2005). While a CFB combustor as a gas– solid reactor operates under low gas velocity and solids flux using larger Group B particles (4 150 mm), the FCC is a gas phase catalytic process which requires significantly higher solids flux with higher gas velocity and employs smaller Group A particles. As one of the most successful and critical processes for conversion of high molecular-weight heavy oil stocks into lighter hydrocarbon products, the FCC process utilizes a riser reactor, where solids flux normally ranges from 400 kg/m2 s to 1200 kg/m2 s and gas velocity from 6 m/s to 28 m/s, increasing with the riser0 s height (Zhu and Bi, 1995). Good knowledge of flow structures in CFB systems is critical for reactor design, modeling and even for the industrial operation. Bi and Zhu (1993) had classified CFBs into high flux and/ or high density (Gs Z200 kg/m2 s, εs Z0.1) circulating fluidized beds (HFCFB/HDCFB) and low density circulating fluidized beds (LDCFB). Despite extensive researches dedicated to gas–solids fluidized bed over the past decades, very limited work has been conducted under solids circulation rates beyond 500 kg/m2 s (Azzi et al., 1991; Martin et al., 1992; Contractor and In Avidan, 1994; and Knowlton, 1995). Recently, studies under high solids flux (Issangya et al., 1999; Grace, 2000; Karri and Knowlton, 1999; Pärssinen and Zhu, 2001a, 2001b; and Yan and Zhu, 2004) had shown that the hydrodynamics are quite
different in comparison with low flux and low density CFB risers operated with Gs o200 kg/m2 s. Issangya (1998) and Issangya et al. (1999) conducted tests in a 6 m high riser under high density conditions (Ug ¼4–8 m/s, Gs ¼200–425 kg/m2 s) and reported that the mean solids holdup was up to 0.1–0.2 without axial variation and the particle downward flow was negligible. Later, Liu et al. (1999) studied gas dispersion in the same system used by Issangya et al. (1999) and found that gas backmixing became lower in the high density riser. More recently, Bi (2004) compared mixing behavior and illustrated that a clear transition of axial mixing appeared from LDCFB to HDCFB operating. In the study by Pärssinen and Zhu, (2001a), (2001b), a high solids flux of 550 kg/m2 s was reached and both axial and radial solids holdup profiles became less uniform under higher solids flux. Issangya et al. (1999, 2000) also reported that radial solids holdup profiles became less uniform at higher Gs (4300 kg/m2 s) with lower solids holdup of less than 0.06 in the center region and 0.4–0.44 near the wall. Understanding the fluid and particle dynamics is evidently of importance to successful modeling of CFB reactors. Flow dynamics also influences pressure drop across the riser, heat transfer (Grace, 1986a, 1986b) as well as erosion rate of surfaces (Zhu et al., 1989). Improved understanding of the flow structures in high flux/ density circulating fluidized bed systems should enable better comprehension of the advantages and limitations of HDCFB
Fig. 1. Schematic diagram of the multifunctional CFB system.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
reactors, in turn leading to more reliable scale-up and more costeffective units (Issangya, 1998). Moreover, it is very helpful to further increase Gs and solids holdup for those industrial processes requiring even higher solids/gas ratios and higher heat exchange. In addition, operating the apparatus with careful design using relatively fine (no-slugging) particles under severe conditions which may induce chocking is possible (Bi et al., 1993; Zhu and Bi, 1995). It had been shown that the dense region at the bottom can extend to the whole riser leading to a high density riser with overall solids holdup of 0.15–0.20 (Contractor and In Avidan, 1994; Issangya et al., 1999). Based on these conditions, it is of interest to study flow structure, operating instability mechanism, particle aggregations and dispersing behaviors under loading conditions with high solids holdup which go beyond those previous researches (Zhu and Bi, 1995). By using an optical fiber probe system, which can simultaneously measure solids holdup and particle velocity, this study is aimed at providing improved experimental investigation in high flux/density CFB riser, including local solids holdup distribution and its evolution with operating conditions. The fluctuations of the local flow structures are also investigated. 2. Experimental details 2.1. CFB system A multifunctional CFB system in this study is shown in Fig. 1. The system includes three circulating fluidized beds, the left hand CFB bed serves as a HDCFB riser (76 mm i.d. and 10 m high). The right hand fluidized beds are two circulating fluidized beds downers (co-current downflow circulating fluidized beds) of different diameters (76 mm i.d. and 5.8 m height and 50 mm and Table 1 Size distribution of the FCC particles. Particle size (mm)
Volume fraction (%)
0–20 20–40 40–60 60–80 80–130 4130
0.61 9.72 26.32 22.80 33.24 7.31
235
4.9 m high, respectively). A downcomer with an inner diameter of 203 mm returns solids during the riser operation. At its bottom there is a solids storage tank with an inner diameter up to 457 mm. The two are used as general solids storage for the entire system. Total solids inventory of FCC particles in the downcomer and storage tank could be up to 450 kg, equivalent to a approximately 6.0 m level of particles which can provide high back pressure in the downcomer and enable high density operating in the CFB riser. In addition, two sets of aeration air are mounted in the recycle loop. One is installed in the inclined solids feeding pipe and the other is fixed at the bottom of the storage tank to maintain the particles at minimum fluidization conditions. The multifunctional circulating fluidized bed can be operated as a CFB riser and downers. For CFB riser operations, particles flowing into the riser bottom are carried upwards by the riser main air. In the outlet of the riser, particles and the riser air are separated by primary, secondary and tertiary cyclones and most of the particles return to the downcomer and further down to the storage tank. Fine particles leaving from the cyclone system are captured in the baghouse and returned periodically to the downcomer. When the system is under downer operating mode, solid particles are first lifted through the riser, separated by the primary cyclone fixed at the top of the downcomer and then fed into the downers. At the top of either downer is a solids distributor where the particles are uniformly distributed along with the downer air to flow down concurrently. At the exit of either downer column, 99% of particles are separated by gravity with fine particles captured by two cyclones installed in series at the top of the exhaust pipeline and the common bagfilter. To eliminate the effects of solids inventory and other influencing parameters on the hydrodynamic characteristics, the whole experimental work in this study was carried out with a constant FCC particle mass of 400 kg stored both in the downcomer and the storage tank for the riser operation and 280 kg stored in the storage tank for the downer mode. The three CFB beds are constructed using aluminum with only small portions made of Plexiglas for visual observation. In order to minimize possible electrostatic charges formed in the columns during the experiments, the whole fluidized bed system is electrically grounded. The gas source used in this study is supplied by a large compressor capable of delivering 283 m3/min at 690 kPa. Equilibrium FCC particles with mean diameter and the particle density of 76 mm and 1780 kg/m3 respectively are used in this study. The particle size distribution measured using BT-9300s laser particle size analyzer is listed in Table 1.
Fig. 2. Schematic diagram of the new optical fiber probe and its working principle.
236
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
2.2. Measurement of solids holdup Experimental measurements include pressure drop, solids holdup and particle velocity. Twenty pressure taps are mounted along the CFB column and connected with 19 pressure transducers (Omega PX162) to collect the pressure data. Pressure data is mainly used to monitor the operation and to verify the solids holdup measured by an optical fiber probe as discussed below. Local solids holdup and particle velocity can be measured simultaneously using a reflective-type optical fiber probe. Previous researches had shown that this kind of optical fiber probe was effective and accurate for measurement in high velocity fluidized beds and thus had been widely used by many investigators (Johnson et al., 2001; Ellis et al., 2004). The optical fiber probe used in this work is model PV6D, new version developed by the Institute of Processing Engineering, Chinese Academy of Sciences, Beijing, China shown in Fig. 2. The probe with a diameter of 3.8 mm has two subprobes mounted vertically with 1.51 mm effective distance, and the active tip area of each subprobe is 1 1 mm2. Each subprobe consists of many 15 mm quartz fibers for light-emitting and receiving, arranged in alternating arrays. To avoid particles staying in the blind zone, a 0.2 mm thickness glass cover is glued on the probe tip. The underlying theory was elaborated by Liu et al. (2003a), (2003b). Light from the source illuminates a measuring volume of particles through the light-emitting fibers. The received light reflected by the particles is captured by light receiving fibers and processed by a photo-multiplier. The light intensity is then converted into voltage signals and the voltage signals are further amplified and fed into a PC. The voltage signal is then converted into volumetric solids concentration using a calibration equation based on a method developed by Zhang et al. (1998).
From the voltage time series V(t) and the calibration equation, local instantaneous solids holdup, εs(t), can be calculated
εs ðtÞ ¼ f ½V ðtÞ
ð1Þ
where, f is the calibration function. The time–mean solids concentration εs can be given by integrating εs(t) over the time period, T
εs ¼
1 T
Z
T
εs ðtÞdt
0
ð2Þ
The cross-sectionally averaged solids holdup (average solids holdup for short) εs , can be calculated as follow:
εs ¼
Z
1
πR
2
0
R
2π r εs dr ¼
2 2
R
Z
R 0
εs rdr
ð3Þ
In order to map the entire cross-section of the riser, ten axial measuring ports (z ¼0.59, 1.02, 1.94, 2.85, 3.77, 4.78, 5.84, 7.78, 9.61, and 10.09 m above the gas distributor) are installed along the column. At each axial position, measurements are conducted at six radial positions (r/R ¼ 0, 0.316, 0.548, 0.707, 0.837 and 0.950, where r is the distance from the center and R is the riser radius) on each axial level of the CFB riser system. These positions are determined by dividing the column cross-section into five equal areas and determining the mid-point of each of these areas. For the hydrodynamic experiments in the current study, voltage signals from the optical fiber probe are sampled at a frequency of 100 kHz with 1,638,400 data points for each measurement under a wide range of operating conditions so that detailed dynamic nature of the flow structure can be fully collected. To get the valid and repeatable data, all measurements are repeated at least five times.
Fig. 3. Characteristics of flow structures under extremely high flux/density in a CFB riser.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
237
3. Results and discussion 3.1. Achieving high flux/density operating conditions in CFB riser Fig. 3 compares local and cross-sectional mean solids holdup in the CFB riser for solids circulation rates ranging between 100 kg/ m2 s and 1000 kg/m2 s. To the best of our knowledge, such a high solids flux has never been reported in a laboratory scale pilot plant experiments. Firstly, at Gs ¼100 kg/m2 s, the axial profile is approximately exponential in shape with a constant solids holdup (lower than 0.01). By increasing the solids circulation rate (Gs), the axial profile becomes non-uniform with solids concentration (solids holdup) decreasing gradually upwards to the top of the riser. When Gs is extremely high, especially higher than 700 kg/ m2 s, the non-uniform axial flow structure is replaced by a homogenous axial profile with solids holdup higher than 0.2 in the entire riser. Interestingly, the radial distributions of the solids holdup are totally different under different operating conditions. At low solids flux, the radial profile is comparably more uniform and less sensitive to the change of the axial position. Increasing Gs, the areas of the relatively dilute region continue decreasing. The solids holdup in the center of the riser is nearly 0.1 under extremely high Gs. Moving outwards towards the wall, solids holdup increases monotonically. Solids holdup remained greater than 0.4 over a wide cross-sectional area (r/R ¼0.7–1.0, about 60% of the cross-sectional area) even at the top section of the riser. Moreover, flow development is much slower under high flux as the radial profiles continued to change as seen in Fig. 3. The above phenomenon suggests that low solids flux data has very limited usefulness to high solids flux reactor modeling and design, especially for solids fluxes within the industrial range (400–1200 kg/m2 s) in FCC riser. Therefore, there is clearly a need to conduct more fundamental researches to study both axial and radial profiles of solids holdup and flow structures in CFB systems operating at higher flux and/or density. However, achieving high flux/density in a CFB system is extremely difficult in any experimental lab. While a few (only a few) research groups (e.g. University of British Columbian, UBC, Vancouver and University of Western Ontario, UWO, Ontario) have tried to obtain high flux/density operating, the solids fluxes are still far below the practical fluxes in industrial reactor processes. Theoretically, Bi and Zhu (1993) proposed that high densities and high solids fluxes could be accomplished by a combination of high solids inventories, large downcomer-to-riser diameter ratio, a low pressure drop solids feeder, and minimizing pressure drops in solids separation devices and fittings along the CFB loop. Besides, a proper blower and suitable particle size/riser diameter combinations are also of importance. Based on their suggestion, a dual loop CFB was used to reduce total pressure drop of the recycle system at UBC which enabled a high solids flux of abound 400 kg/m2 s to be achieved. A twin-riser CFB with a large downcomer-to-riser ratio was constructed at UWO which achieved high fluxes up to 500 kg/ m2 s. Building further upon the practical experience, we have: (1) Installed a large diameter storage tank at the bottom of the downcomer; (2) replaced the blower with a compressor of high capacity 1000 SCFM at 100 psi; (3) installed an additional air exhaust pipe at the top of the of the downcomer to discharge most of the air flowing upward through the downcomer, minimizing downcomer air flowing into the primary cyclone, so that the pressure drop across the cyclone is significantly reduced. This step increased the available pressure for the riser to achieve higher density; (4) installed two small deflecting plates (see the left insert of Fig. 1) in the solids inlet region, one vertically at the outlet of
Fig. 4. Axial solids holdup distribution for various operating conditions.
the inclined pipe covering 30% of the lower end of the inclined feed pipe joining the riser to prevent the riser air from flowing into the solids feeding pipe which tends to restrict solids downflow. The other deflecting plate was half-way up in the inclined pipe covering 30% of the cross-sectional area of the inclined pipe. It directed particles downwards so as to provide
238
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
a quit “exit route” for the remaining air entering the feed pipe so that the solids movement in the inclined pipe is much faster and steadier. After the modification, the riser can be operated much more steadily at solids circulation rates of up to 1000 kg/m2 s, much
higher than 400 kg/m2 s in the original system reported by Li (2010). This allows us to operate the CFB system under a wide range of operating conditions to obtain a comprehensive map of solids flow in the new solids recycle loop. An example is the data set plotted in Fig. 3 and discussed above.
3.2. Axial profiles of solids holdup
Fig. 5. Comparison of solids holdup profiles under different operating conditions.
Fig. 4 displays the axial distribution of the averaged solids holdup in the CFB riser for superficial gas velocity, Ug, of 5, 7 and 9 m/s and solids flux, Gs, up to 1000 kg/m2 s. As shown in Fig. 4, significantly different axial profiles can be seen under various operating conditions. In general, as solids circulation rate (Gs) increases, the approximately exponential-shaped axial profile is replaced by the relatively non-uniform axial profile with a dense region at the rsier base and a dilute region at the top section and then at even higher solids fluxes, by more uniform axial profiles. In Fig. 4(a) and (b), when Ug is constant the solids holdup increases with increasing Gs, while for a constant Gs, the solids holdup decreases with increasing Ug as plotted in Fig. 4(c). In details, relatively high solids holdups near the bottom of the riser are extensively affected by the particle acceleration and the gas distributor. Above this region, at the lowest Gs of 100 kg/m2 s, the solids holdup decreases exponentially, eventually approaching a constant value up to the riser exit. This exponential shape occurs when particles entering in the riser bottom are immediately
Fig. 6. Radial solids holdup distribution for various operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
entrained without significant. By increasing Gs gradually, a significant dense region at the riser0 s bottom forms, leading to a nonuniform axial flow structure. It can also be seen that the solids holdup along axial elevations is increasing with increasing Gs at a fixed Ug. Meanwhile, the solids holdup is also increasing at each axial level with decreasing Ug when Gs is constant. The dense region occurs between 2 and 4 m heights with the high solids holdup raging from 0.1 to 0.25. Above this region, relatively dilute regions are reached and the average solids holdup becomes independent of the axial variation. It is clear that the 2 melevation lies in the developing flow region with high solids holdups and a fully-developed region with relatively low solids holdup begins at 5 m height. Interestingly, the shape of the axial profile is hardly changed with various operating conditions. It is apparent that the axial profiles move in parallel from low solids holdup towards high solids holdup with increasing Gs and/or decreasing Ug. The solids holdup and the flow development in both regions of the riser do not seem to depend on the height, but are expected to depend on the operating conditions. This may be partly affected by pressure balance in the whole CFB recycling system. If the solids inventory is high, the pressure head in the return system would be sufficiently high. Therefore, it is easy to adjust the pressure in the return system to meet the requirement for pressure balance in the whole loop under various operating conditions. Moreover, this axial profile is different from the heightdependent S-shaped axial profile reported by other authors under
239
high flux and/or high density operating conditions (Issangya et al., 1999; Pärssinen and Zhu, (2001a), (2001b) and Yan and Zhu, 2004). Further increasing Gs, significantly dense solids holdup in the entire height of the riser is achieved ranging from 0.23 to 0.38 when solids circulation rate is extremely high, particularly higher than 700 kg/m2 s. The uniform and dense gas–solids suspension has been achieved along the whole column if ignoring the entrance effect. This homogenous axial structure is similar to other results reported by previous researchers. For instance, Contractor and In Avidan, (1994) found that solids holdup between 0.15 and 0.2 could cover the entire riser for a solids flux up to 685 kg/m2 s and gas velocity up to 5.7 m/s. In contrast to the nonuniformity axial profiles with solids circulation rate lower than 700 kg/m2 s, the dense bed at the base riser of the can persist over the entire column under such high Gs. Axial distribution of solids holdup becomes very uniform. As already noted, being able to achieve this kind of homogenous axial profile depends on being able to provide sufficient pressure head. If the pressure head is high enough, the homogenous axial profile appears to be robust and self-sustaining over a considerable range of operating conditions Information about the effect of Gs on the respective axial profiles in an industrial riser is very useful since the solids flux varies widely from 400 to 1200 kg/m2 s in the industrial reactors. According to the concept of the high density operation as mentioned in the introduction section, it is clear that what have
Fig. 7. Overall view of the solids hold up under different operating conditions.
240
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
observed in this study in the high flux riser does match the high flux and/or high density operating conditions suggested by Zhu and Bi (1995). In addition, the result is also comparable to the results (0.25–0.30) in C-TFB with a highly uniform and dense suspension flow structure operated at low superficial gas velocity and high solids circulation rate reported by Zhu and Zhu (2008a), (2008b). The uniform high density structure stresses the importance to both uniform gas–solids contacting efficiency and uniform bed-to-wall heat transfer throughout the whole reactor.
3.3. Radial distribution of solids holdup Typical radial profiles of solids holdup in the CFB riser are presented in Fig. 5 under the operating conditions as well as at different axial positions. Under all operating conditions, radial distributions of the solids holdup are nonuniform with dilute and dense regions. The radial profile is relatively flat in the center of the riser and the solids holdup increases towards the wall with the maximum value right at the wall at 0.56. Operating conditions affect dramatically the radial solids holdup distribution. The riser gradually becomes denser from the wall towards the center with increasing solids feeding at a fixed Ug. The solids holdup and its radial distribution in the high density CFB riser are quite different from those in the low density systems. Obviously, the radial profile is a clear “core-annulus” structure when Gs is low. The radial variation can be divided into three parts: a central region up to r/R E0.5–0.6 with a dilute and uniform solids holdup, an intermediate region between r/R E 0.5–0.6 and r/RE 0.8–0.9 where solids holdup appreciably increases, and a wall region when r/ R4 0.9 where the solids holdup is high but not more than 0.35. The results are similar to those reported by other researchers under comparable operating conditions (Liu, 2001; Yan and Zhu, 2004). As Gs increases to higher than 700 kg/m2 s, the dilute region shrinks (r/R ¼0–0.2, less than 20% of the cross-sectional area). After this short region, solids holdup increase gradually towards the wall which can be up to 0.5. The “core-annuals” radial profile is
replaced by the concave parabolic curve under extremely high Gs. More details can be shown in Fig. 6. Regarding the results in columns I, II and V in Fig. 6, it is clear that the superficial gas velocity plays an important role in solids distribution in radial and axial directions. It is apparent that reducing Ug results in an increased solids holdup and thus an increased solids holdup profile. Additionally, superficial gas velocity influences the solids holdup not only in the wall region but also in the center of the riser. For instance, when Ug increases from 5 to 9 m/s, the solids holdup decreases in the center region at different heights and drops more rapidly in the wall region especially at the bottom of the riser. Moreover, the solids holdup distribution is less uniform at lower superficial gas velocity. Through comparison between columns II, III and IV and/or between V, VI and VII, increasing solids circulation rate leads to a higher solids holdup with a less uniform profile. Under high Gs, solids holdup gradient increases and solids distribution becomes less uniform because of the more confined wall effects to the near wall region. Lower Gs leads to a more even radial solids distribution with lower local solids holdup across the column. It is also to see that axial distributions of radial solids holdup vary at different solids fluxes. When Gs is lower than 500 kg/m2 s, the solids holdup in the wall region at the upper section is lower than that in the bottom section. However, the axial variations of radial profiles of solids holdup changed very little under extremely high solids circulating rate of 1000 kg/m2 s.
3.4. Flow development of solids holdup Solids flow can be considered as fully developed if the radial solids distribution remains relatively unchanged with rising axial location (Qi et al., 2009). A detailed review of Figs. 3 and 6 shows that increasing Ug accelerates the solids flow development. On the other hand, when Gs is increased from 300 kg/m2 s to 500 kg/m2 s and then to 700 kg/m2 s at Ug ¼7 m/s, the flow development becomes much slower. A similar trend is also observed when Ug is 9 m/s.
Fig. 8. Solids holdup distribution in different radial regions for various operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
To facilitate the analysis of flow development, three-dimensional profiles of solids holdup are given in Fig. 7. It shows clearly that in the riser center, solids holdup remains nearly invariable in the entire column when Gs o700 kg/m2 s. Similar results are also reported by previous studies under high flux conditions (Issangya et al., 2000; Pärssinen and Zhu, (2001a), (2001b); Yan and Zhu, 2004 and Huang et al., 2007). However, when Gs is up to or higher than 700 kg/m2 s, the flow development becomes extremely slower with a reduced core region of relatively low solids holdups, and the solids holdup is increasing gradually from the core region toward the wall. The radial solids holdup profiles are typically parabolic in shape as aforementioned. In order to further investigate the flow behavior, it is helpful to examine flow development in the different radial locations. To this end, the whole cross-sectional area is arbitrarily divided to three parts: r/R ¼0.0–0.548, 0.548–0.837 and 0.837–1.0. Axial profiles of the solids holdups in these different radial regions are plotted in Fig. 8. As shown in the left of the Fig. 8, in the center where r/R ¼ 0.0– 0.548, the solids holdup is very low with no significant difference in this region throughout the entire riser at various superficial gas velocities. In the intermediate region with r/R ¼0.548–0.837 solids holdup varies near the gas distributor and followed by a flat trend. Near the wall (r/R ¼0.837–1.0), the solids holdup considerably varies with sharply dropping in the entrance and then becomes relatively flat till the top of the column. A similar trend in flow development is also observed when Gs is below 500 kg/m2 s when Ug ¼9 m/s. When Gs is higher than 500 kg/m2 s, especially higher than 700 kg/m2 s, the solids holdup in the center remains at about 0.1 even at the outlet of the column. Solids holdup in the middle region is high up to 0.4 and remains unchanged along the riser up to 4 m-level and then gradually decreases. The solids holdup still remains much higher of up to 0.25 in this intermediate region. While in the wall region, solids suspension density is rather flat with an extremely high value of up to 0.55 along all the axial elevations at the highest solids flux, Gs ¼1000 kg/m2 s. This insensitivity of solids holdup across the whole radial area to the axial variations may suggest that a saturation state is to some extent reached in the extremely high density CFBs. The flow development can also be shown by the radial nonuniformity index (RNI) which was proposed by Zhu and Manyele (2001). Based on their definition, the RNI of solids holdup, RNI(εs), can be defined as RNIðεs Þ ¼
sðεs Þ sð ε s Þ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sðεs Þmax εs ðεs;mf εs Þ
ð4Þ
where s(εs) is the standard deviation of the radial solids holdup, s(εs)max is the maximum possible standard deviation, εs is the average solids holdup, and εs,mf is the highest possible solids holdup. Fig. 9 shows the RNI (εs) along the axial direction in the riser. In general, all RNI (εs) profiles are showing that the value of RNI (εs) is higher at the bottom of the riser followed by a relatively flat curve along the axial direction. Obviously, higher RNI (εs) values at the bottom are due to the development of solids flow. The effect of operating conditions on the RNI (εs) can be inferred from Fig. 9(a). RNI (εs) decreases with increasing superficial gas velocity, showing that the radial profiles of solids holdup become more uniform. High gas velocity can enhance the gas– particle interactions. Solids flow development would be accelerated by high drag force impacted by high speed gas so that the solids flow structure becomes more uniform. At a given Ug, high Gs results in higher RNI (εs) because of the increased solids holdup which may induce the formation of the particle aggregation and thus increase the radial nonuniformity. Moreover, Fig. 9(b) shows a
241
Fig. 9. RNI of solids holdup the riser under different operating conditions (a) RNI (εs) and (b) relationship between RNI and solids holdup.
relationship between the RNI (εs) and average solids holdup. It shows that the RNI (εs) values increase with the average solids holdups the radial profiles of solids holdup become steepened with increasing average solids holdup (Zhu and Manyele, 2001). It is obvious that stable RNI (εs) is achieved at higher levels along the riser with increasing of Gs or decreasing of Ug. Take Ug ¼9 m/s as an example, an increase in solids circulation rate from 100 to 600 kg/ m2 s, extends the distance from 1 m to about 6 m. Further increasing Gs, the distance for the stabilization of RNI (εs) appears to extend beyond the total length of the riser. The phenomenon again verifies the acceleration process to achieve complete flow development in the CFBs. 3.5. Flow fluctuations in the high flux/density riser From the foregoing analysis, it is obvious that the local solids holdups vary in the radial direction. In order to gain a better understanding of the local solids flow, it is helpful to examine the flow fluctuations under different operating conditions, especially in high density (HDCFB) and low density circulating fluidized beds (LDCFB). Fluctuations can be quantified by the standard deviation and the intermittency indices (γ) of solids holdups. Definition of γ in details can be found in the previous paper (Brereton and Grace, 1993). Besides, the intermittency index can also be used to
242
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
describe the phenomenon of segregation between gas and solids. The value of intermittency index is equal to one if the gas–solids suspension is considered as an “ideal cluster flow”. On the other hand, the value of intermittency index is zero if there is a “perfect core-annulus flow”. Fig. 10 compares the radial profiles of local solids holdup, standard deviation and the corresponding intermittency index at two axial heights. Generally, the local solids holdup is lowest and has a relatively flat radial distribution under the low flux of 100 kg/m2 s. With increasing solids flux or decreasing gas velocity, the solids holdup increases, and the radial distribution becomes steeper. The standard deviation and γ increase with increasing Gs or decreasing Ug. This suggests increased flow fluctuation with the increase of Gs or decrease of Ug. The fluctuation of low flux solids holdups increases outward, reaching maxima at the outer wall. For high flux conditions, the fluctuation increases gradually to the highest value at r/R ¼0.6–0.7 and thereafter decreases towards the wall. The greatest fluctuation always occurs in the middle region of the column under high flux operating conditions which are also reported under high density operating conditions in many papers (Issangya et al., 2000; Zhu and Zhu, 2008a, 2008b, Qi et al., 2009). This phenomenon has been considered as one of the most important distinguishing features in HDCFBs. When Gs is lower than 500 kg/m2 s, the
standard deviation and the corresponding intermittency index remains a very low level around 0.02–0.05 up to r/R E 0.4–0.6. Considering its solids holdup profile, this indicates that a wide dilute region still occupies the center surrounded by a dense annulus region at the wall. Lower solids holdups in the center of the column and higher solids concentrations near the wall make fewer fluctuations. It is found that the standard deviations and intermittency indices under extremely high solids flux of 1000 kg/ m2 s are much higher than those under other low high solids flux conditions at almost all radial positions. This highest magnitude of fluctuations may due to the relatively higher solids holdup suggesting that the gas–solids and inter-particle interaction in this particularly high flux CFB are more vigorous than those in relatively low solids flux CFBs operations. The dramatic fluctuations probably induce better gas–solids contacting and mixing, improving consequently the reactor performance. Additionally, for each solids circulation rate, there is a clear trend for the development of standard deviation and γ with axial variations. Clearly, both parameters tend to decrease with the increasing axial level. This illustrates that local solids flow is liable to change with increasing height from one where cluster-flow structures are more common to one where the core-annulus structure becomes predominant. The same results are also presented by Brereton and Grace (1993).
Fig. 10. Radial profiles of solids holdup, standard deviation and intermittency indices along the riser under different operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
4. Conclusions The current work presented a comprehensive insight into the solids holdup and flow development phenomena in a high flux/ density (up to 1000 kg/m2 s) riser, using an optical fiber. Solids suspension having a solids holdup of up to 0.2–0.3 can be maintained throughout the entire high flux/density riser. A homogenous axial flow structure is observed at Gs ¼1000 kg/m2 s. Radial distributions of the solids holdup are nonuniform with a dilute region and a dense region. For Gs 4700 kg/m2 s, the dilute core region shrinks to about r/R ¼0–0.2, less than 20% of the crosssectional area. Solids holdups thereafter increase monotonically towards the wall reaching up to 0.56. The speed of flow development varies in different crosssectional areas. Flow development is almost instant at the center through the entire riser even at the high density conditions, while it becomes much slower in the wall region. Increasing solids flux prolongs the solids flow development. Better gas–solids contacting and mixing indicated by standard deviation and intermittency index of the solids holdup over the entire cross-sectional area under extremely high solids flux can greatly lead to vigorous interactions between gas and solids phases, improving the reactor performance. Notation f Gs Le r/R RNI(εs) t T Ug V(t) z
Function between the voltage and solids holdup (dimensionless) solids circulation rate (kg/m2 s) effective distance between the two subprobes (m) normalized radial position radial nonuniformity index of solids holdup time period of particles through the two subprobes (s) time period of testing (s) superficial gas velocity (m/s) voltage time series (V) distance above the gas distributor (m)
Greek letters
εs εs,mf εs(t) εs τ s(εs) s(εs)max
solids holdup (dimensionless) the highest solids holdup (dimensionless) instantaneous solids holdup (dimensionless) average solids holdup in the entire column (dimensionless) lag time (s) standard deviation of the radial solids holdup (dimensionless) maximum possible standard deviation (dimensionless)
Subscripts 1, 2 g p s
subprobe 1 and 2 of optical fiber probe gas particle solids
References Azzi, M., Turlier, P., Betrand, J.R., 1991. Mapping solids concentration in a circulating fluidized bed using gammametry. Powder Technol. 67, 249–258.
243
Bi, H.T., Zhu, J.X., 1993. Static instability analysis of circulating fluidized beds and concept of high-density risers. AIChE J. 39, 1272–1280. Bi, H.T., 2004. Gas and solid mixing in high-density CFB risers. Int. J. Chem. React. Eng. 2, A12. Bi, H.T., Grace, J.R., Zhu, J.X., 1993. Types of choking in vertical pneumatic systems. Int. J. Multiph. Flow 19, 1077–1092. Brereton, C.M., Grace, J.R., 1993. Microstructural aspects of the behavior of circulating fluidized beds. Chem. Eng. Sci. 48 (14), 2565–2572. Contractor, R.M., In Avidan, A.A., 1994. Circulating Fluidized Bed Technology IV. American Institute of Chemical EngineersNew York, pp. 387–391. Ellis, N., Bi, H.T., Lim, C.J., Grace, J.R., 2004. Influence of probe scale and analysis method on measured hydrodynamic properties of gas-fluidized beds. Chem. Eng. Sci. 59 (8-9), 1841–1851. Grace, J.R., 1986a. Contacting modes and behavior classification of gas-solids and other two-phase suspensions. Can. J. Chem. Eng. 64, 353–363. Grace, J.R., 1986b. Heat transfer in circulating fluidized beds. In: Basu, P. (Ed.), Circulating Fluidized Bed TechnologyPergamon Press, Toronto, pp. 63–81. Grace, J.R., 2000. Reflections on turbulent fluidization and dense suspension upflow. Powder Technol. 113, 242–248. Huang, W.X., Yan, A.J., Zhu, J.X., 2007. Hydrodynamics and flow development in a 15.1 m circulating fluidized bed riser. Chem. Eng. Technol. 30 (4), 460–466. Issangya, A.S., 1998. Flow Dynamics in High Density Circulating Fluidized Beds (Ph. D. diss.). Univ. of British Columbia, Vancouver, BC. Issangya, A.S., Bai, D.R., Bi, H.T., Lim, K.S., Zhu, J., Grace, J.R., 1999. Suspension densities in a high-density circulating fluidized bed riser. Chem. Eng. Sci. 54, 5451–5460. Issangya, A.S., Grace, J.R., Bai, D.R., Zhu, J.X., 2000. Further measurements of flow dynamics in a high-density circulating fluidized bed riser. Powder Technol. 111, 104–113. Johnsson, H., Johnsson, F., 2001. Measurements of local solids volume-fraction in fluidized bed boilers. Powder Technol 115, 13–26. Karri, S.B.R., Knowlton, T.M., 1999. A comparison of annulus solids flow direction and radial solids mass flux profiles at low and high mass fluxes in a riser. In: Werther, J. (Ed.), Circulating Fluidized Bed Technology VI. Frankfurt, Dechema, pp. 71–76. Knowlton, T., 1995. Interaction of pressure and diameter on CFB pressure drop and holdup. Paper for Workshop, Modeling and Control of Fluidized Bed Systems. Hamburg, May 22–23. Li, D.B., 2010. Investigation of circulating Fluidized Bed Riser and Downer Reactor Performance for Catalytic Ozone Decomposition (Ph.D. diss.), University of Western Ontario, London, ON. Liu, J.Z., Grace, J.R., Bi, H.T., Morikawa, H., Zhu, J.X., 1999. Gas dispersion in fast fluidization and dense suspension upflow. Chem. Eng. Sci. 54 (22), 5441–5449. Liu, J.Z., 2001. Particle and Gas Dynamics of High Density Circulating Fluidized Beds (Ph.D. diss.). Univ. of British Columbia, Vancouver, BC. Liu, J.Z., Grace, J.R., Bi, H.T., 2003a. Novel multifunctional optical-fiber probe: development and validation. AIChE J. 49 (6), 1405–1420. Liu, J.Z., Grace, J.G., Bi, H.T., 2003b. Novel multifunctional optical-fiber probe: highdensity CFB measurements. AIChE J. 49 (6), 1421–1432. Martin, M.P., Turlier, P., Bernard, J.R., 1992. Gas and solid behavior in cracking circulating fluidized beds. Powder Technol. 70, 249–258. Pärssinen, J.H., Zhu, J.X., 2001a. Axial and radial solids distribution in a long and high-flux CFB riser. AIChE J. 47, 2197–2205. Pärssinen, J.H., Zhu, J.X., 2001b. Particle velocity and flow development in a long and high-flux circulating fluidized bed riser. Chem. Eng. Sci. 56, 5295–5303. Qi, X.B., Zhu, H.Y., Zhu, J., 2009. Demarcation of a new circulating turbulent fluidization regime. AIChE J. 55 (3), 594–611. Reh, L., 1999. Challenges of circulating fluid-bed reactors in energy and raw materials industries. Chem. Eng. Sci. 54 (22), 5359–5368. Yan, A.J., Zhu, J.X., 2004. Scale-up effect of riser reactors (1): axial and radial solids concentration distribution and flow development. Ind. Eng. Chem. Res. 43, 5810–5819. Zhang, H., Johnston, P.M., Zhu, J.X., de Lasa, H.I., Bergougnou, M.A., 1998. A novel calibration procedure for a fiber optic solids concentration probe. Powder Technol. 100, 260–272. Zhu, H.Y., Zhu, J.X., 2008a. Gas-solids flow structures in a novel circulatingturbulent fluidized bed. AIChE J. 54 (5), 1212–1223. Zhu, H.Y., Zhu, J.X., 2008b. Comparative study of flow structures in a circulatingturbulent fluidized bed. Chem. Eng. Sci. 63, 2920–2927. Zhu, J.X., Grace, J.R., Lim, C.J., 1989. Erosion-causing particle impacts on tubes in fluidized beds. In: Grace, V.I., Schemilt, J.R., Bergougnou, L.W., M.A. (Eds.), FluidizationEngineering Foundation, New York, pp. 613–620. Zhu, J.X., Bi, H.T., 1995. Distinctions between low density and high density circulating fluidized beds. Can. J. Chem. Eng. 73, 644–649. Zhu, J.X., Manyele, S.V., 2001. Radial nonuniformity index (RNI) in fluidized beds and other multiphase flow systems. Can. J. Chem. Eng. 79, 202–211. Zhu, J.X., Cheng, Y., 2005. Fluidized-bed reactors and applicationsIn: Clayton, Crowe (Ed.), Multiphase Flow Handbook. CRC Press, New York5.55–5.93 (Chapter 5.3).
Contents lists available at ScienceDirect
Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces
Axial and radial development of solids holdup in a high flux/density gas–solids circulating fluidized bed Chengxiu Wang a,b, Jesse Zhu a,n, Shahzad Barghi a, Chunyi Li b a b
Department of Chemical & Biochemical Engineering, University of Western Ontario, London, Ontario, Canada N6A 5B9 State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266555, China
H I G H L I G H T S
G R A P H I C A L
Experiments at high solids flux of 1000 kg/m2 s. Homogenous axial profile of solids holdup with solids holdup up to 0.32. Different radial profiles of solids holdup under high density conditions compared to low solids flux conditions. Better gas–solids contacting under high density conditions in CFBs.
Characteristics of flow structure under extremely high flux/density in a CFB riser.
art ic l e i nf o
a b s t r a c t
Article history: Received 3 October 2013 Received in revised form 25 December 2013 Accepted 27 December 2013 Available online 6 January 2014
Detailed distributions of solids holdup in an extremely high density circulating fluidized bed riser with FCC particles are mapped by an optical fiber probe. The solids circulation rate reaches as high as 1000 kg/ m2 s which has never been achieved before in an academic setting. When solids flux approaches 800 kg/ m2 s, the axial flow structure becomes uniform and the cross-sectional mean solids holdup reaches 0.22 throughout the entire riser; the same reaching 0.32 at 1000 kg/m2 s. Compared to a typical core-annulus structure, the radial distributions of the solids holdup becomes much less uniform with a shrinking core and transits to a monotonic increasing profile towards the wall. Speed of flow development differs at various radial positions with almost instant development in the center even at the highest solids flux of 1000 kg/m2 s and then becoming slower towards the wall. Fluctuations in high density circulating fluidized beds are significantly greater than those in low density ones, leading to more vigorous interactions between gas and solids phases. As a result, better gas–solids contacting and mixing, plus the uniform axial profiles of solids holdup, provide better reactor performance for the high solids flux/ density risers than low flux/density ones. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Circulating fluidized bed Riser Hydrodynamics High flux/density Solids holdup Flow development
A B S T R A C T
1. Introduction n
Corresponding author. Tel.: þ 1 519 661 3807; fax: þ1 519 850 2441. E-mail address: [email protected] (J. Zhu).
0009-2509/$ - see front matter & 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ces.2013.12.042
Circulating fluidized beds (CFBs) have been successfully used in industrial operations such as Fischer–Tropsch synthesis, partial
234
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
oxidation, fluid-catalytic cracking (FCC), and combustion (Reh, 1999; Zhu and Cheng, 2005). While a CFB combustor as a gas– solid reactor operates under low gas velocity and solids flux using larger Group B particles (4 150 mm), the FCC is a gas phase catalytic process which requires significantly higher solids flux with higher gas velocity and employs smaller Group A particles. As one of the most successful and critical processes for conversion of high molecular-weight heavy oil stocks into lighter hydrocarbon products, the FCC process utilizes a riser reactor, where solids flux normally ranges from 400 kg/m2 s to 1200 kg/m2 s and gas velocity from 6 m/s to 28 m/s, increasing with the riser0 s height (Zhu and Bi, 1995). Good knowledge of flow structures in CFB systems is critical for reactor design, modeling and even for the industrial operation. Bi and Zhu (1993) had classified CFBs into high flux and/ or high density (Gs Z200 kg/m2 s, εs Z0.1) circulating fluidized beds (HFCFB/HDCFB) and low density circulating fluidized beds (LDCFB). Despite extensive researches dedicated to gas–solids fluidized bed over the past decades, very limited work has been conducted under solids circulation rates beyond 500 kg/m2 s (Azzi et al., 1991; Martin et al., 1992; Contractor and In Avidan, 1994; and Knowlton, 1995). Recently, studies under high solids flux (Issangya et al., 1999; Grace, 2000; Karri and Knowlton, 1999; Pärssinen and Zhu, 2001a, 2001b; and Yan and Zhu, 2004) had shown that the hydrodynamics are quite
different in comparison with low flux and low density CFB risers operated with Gs o200 kg/m2 s. Issangya (1998) and Issangya et al. (1999) conducted tests in a 6 m high riser under high density conditions (Ug ¼4–8 m/s, Gs ¼200–425 kg/m2 s) and reported that the mean solids holdup was up to 0.1–0.2 without axial variation and the particle downward flow was negligible. Later, Liu et al. (1999) studied gas dispersion in the same system used by Issangya et al. (1999) and found that gas backmixing became lower in the high density riser. More recently, Bi (2004) compared mixing behavior and illustrated that a clear transition of axial mixing appeared from LDCFB to HDCFB operating. In the study by Pärssinen and Zhu, (2001a), (2001b), a high solids flux of 550 kg/m2 s was reached and both axial and radial solids holdup profiles became less uniform under higher solids flux. Issangya et al. (1999, 2000) also reported that radial solids holdup profiles became less uniform at higher Gs (4300 kg/m2 s) with lower solids holdup of less than 0.06 in the center region and 0.4–0.44 near the wall. Understanding the fluid and particle dynamics is evidently of importance to successful modeling of CFB reactors. Flow dynamics also influences pressure drop across the riser, heat transfer (Grace, 1986a, 1986b) as well as erosion rate of surfaces (Zhu et al., 1989). Improved understanding of the flow structures in high flux/ density circulating fluidized bed systems should enable better comprehension of the advantages and limitations of HDCFB
Fig. 1. Schematic diagram of the multifunctional CFB system.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
reactors, in turn leading to more reliable scale-up and more costeffective units (Issangya, 1998). Moreover, it is very helpful to further increase Gs and solids holdup for those industrial processes requiring even higher solids/gas ratios and higher heat exchange. In addition, operating the apparatus with careful design using relatively fine (no-slugging) particles under severe conditions which may induce chocking is possible (Bi et al., 1993; Zhu and Bi, 1995). It had been shown that the dense region at the bottom can extend to the whole riser leading to a high density riser with overall solids holdup of 0.15–0.20 (Contractor and In Avidan, 1994; Issangya et al., 1999). Based on these conditions, it is of interest to study flow structure, operating instability mechanism, particle aggregations and dispersing behaviors under loading conditions with high solids holdup which go beyond those previous researches (Zhu and Bi, 1995). By using an optical fiber probe system, which can simultaneously measure solids holdup and particle velocity, this study is aimed at providing improved experimental investigation in high flux/density CFB riser, including local solids holdup distribution and its evolution with operating conditions. The fluctuations of the local flow structures are also investigated. 2. Experimental details 2.1. CFB system A multifunctional CFB system in this study is shown in Fig. 1. The system includes three circulating fluidized beds, the left hand CFB bed serves as a HDCFB riser (76 mm i.d. and 10 m high). The right hand fluidized beds are two circulating fluidized beds downers (co-current downflow circulating fluidized beds) of different diameters (76 mm i.d. and 5.8 m height and 50 mm and Table 1 Size distribution of the FCC particles. Particle size (mm)
Volume fraction (%)
0–20 20–40 40–60 60–80 80–130 4130
0.61 9.72 26.32 22.80 33.24 7.31
235
4.9 m high, respectively). A downcomer with an inner diameter of 203 mm returns solids during the riser operation. At its bottom there is a solids storage tank with an inner diameter up to 457 mm. The two are used as general solids storage for the entire system. Total solids inventory of FCC particles in the downcomer and storage tank could be up to 450 kg, equivalent to a approximately 6.0 m level of particles which can provide high back pressure in the downcomer and enable high density operating in the CFB riser. In addition, two sets of aeration air are mounted in the recycle loop. One is installed in the inclined solids feeding pipe and the other is fixed at the bottom of the storage tank to maintain the particles at minimum fluidization conditions. The multifunctional circulating fluidized bed can be operated as a CFB riser and downers. For CFB riser operations, particles flowing into the riser bottom are carried upwards by the riser main air. In the outlet of the riser, particles and the riser air are separated by primary, secondary and tertiary cyclones and most of the particles return to the downcomer and further down to the storage tank. Fine particles leaving from the cyclone system are captured in the baghouse and returned periodically to the downcomer. When the system is under downer operating mode, solid particles are first lifted through the riser, separated by the primary cyclone fixed at the top of the downcomer and then fed into the downers. At the top of either downer is a solids distributor where the particles are uniformly distributed along with the downer air to flow down concurrently. At the exit of either downer column, 99% of particles are separated by gravity with fine particles captured by two cyclones installed in series at the top of the exhaust pipeline and the common bagfilter. To eliminate the effects of solids inventory and other influencing parameters on the hydrodynamic characteristics, the whole experimental work in this study was carried out with a constant FCC particle mass of 400 kg stored both in the downcomer and the storage tank for the riser operation and 280 kg stored in the storage tank for the downer mode. The three CFB beds are constructed using aluminum with only small portions made of Plexiglas for visual observation. In order to minimize possible electrostatic charges formed in the columns during the experiments, the whole fluidized bed system is electrically grounded. The gas source used in this study is supplied by a large compressor capable of delivering 283 m3/min at 690 kPa. Equilibrium FCC particles with mean diameter and the particle density of 76 mm and 1780 kg/m3 respectively are used in this study. The particle size distribution measured using BT-9300s laser particle size analyzer is listed in Table 1.
Fig. 2. Schematic diagram of the new optical fiber probe and its working principle.
236
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
2.2. Measurement of solids holdup Experimental measurements include pressure drop, solids holdup and particle velocity. Twenty pressure taps are mounted along the CFB column and connected with 19 pressure transducers (Omega PX162) to collect the pressure data. Pressure data is mainly used to monitor the operation and to verify the solids holdup measured by an optical fiber probe as discussed below. Local solids holdup and particle velocity can be measured simultaneously using a reflective-type optical fiber probe. Previous researches had shown that this kind of optical fiber probe was effective and accurate for measurement in high velocity fluidized beds and thus had been widely used by many investigators (Johnson et al., 2001; Ellis et al., 2004). The optical fiber probe used in this work is model PV6D, new version developed by the Institute of Processing Engineering, Chinese Academy of Sciences, Beijing, China shown in Fig. 2. The probe with a diameter of 3.8 mm has two subprobes mounted vertically with 1.51 mm effective distance, and the active tip area of each subprobe is 1 1 mm2. Each subprobe consists of many 15 mm quartz fibers for light-emitting and receiving, arranged in alternating arrays. To avoid particles staying in the blind zone, a 0.2 mm thickness glass cover is glued on the probe tip. The underlying theory was elaborated by Liu et al. (2003a), (2003b). Light from the source illuminates a measuring volume of particles through the light-emitting fibers. The received light reflected by the particles is captured by light receiving fibers and processed by a photo-multiplier. The light intensity is then converted into voltage signals and the voltage signals are further amplified and fed into a PC. The voltage signal is then converted into volumetric solids concentration using a calibration equation based on a method developed by Zhang et al. (1998).
From the voltage time series V(t) and the calibration equation, local instantaneous solids holdup, εs(t), can be calculated
εs ðtÞ ¼ f ½V ðtÞ
ð1Þ
where, f is the calibration function. The time–mean solids concentration εs can be given by integrating εs(t) over the time period, T
εs ¼
1 T
Z
T
εs ðtÞdt
0
ð2Þ
The cross-sectionally averaged solids holdup (average solids holdup for short) εs , can be calculated as follow:
εs ¼
Z
1
πR
2
0
R
2π r εs dr ¼
2 2
R
Z
R 0
εs rdr
ð3Þ
In order to map the entire cross-section of the riser, ten axial measuring ports (z ¼0.59, 1.02, 1.94, 2.85, 3.77, 4.78, 5.84, 7.78, 9.61, and 10.09 m above the gas distributor) are installed along the column. At each axial position, measurements are conducted at six radial positions (r/R ¼ 0, 0.316, 0.548, 0.707, 0.837 and 0.950, where r is the distance from the center and R is the riser radius) on each axial level of the CFB riser system. These positions are determined by dividing the column cross-section into five equal areas and determining the mid-point of each of these areas. For the hydrodynamic experiments in the current study, voltage signals from the optical fiber probe are sampled at a frequency of 100 kHz with 1,638,400 data points for each measurement under a wide range of operating conditions so that detailed dynamic nature of the flow structure can be fully collected. To get the valid and repeatable data, all measurements are repeated at least five times.
Fig. 3. Characteristics of flow structures under extremely high flux/density in a CFB riser.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
237
3. Results and discussion 3.1. Achieving high flux/density operating conditions in CFB riser Fig. 3 compares local and cross-sectional mean solids holdup in the CFB riser for solids circulation rates ranging between 100 kg/ m2 s and 1000 kg/m2 s. To the best of our knowledge, such a high solids flux has never been reported in a laboratory scale pilot plant experiments. Firstly, at Gs ¼100 kg/m2 s, the axial profile is approximately exponential in shape with a constant solids holdup (lower than 0.01). By increasing the solids circulation rate (Gs), the axial profile becomes non-uniform with solids concentration (solids holdup) decreasing gradually upwards to the top of the riser. When Gs is extremely high, especially higher than 700 kg/ m2 s, the non-uniform axial flow structure is replaced by a homogenous axial profile with solids holdup higher than 0.2 in the entire riser. Interestingly, the radial distributions of the solids holdup are totally different under different operating conditions. At low solids flux, the radial profile is comparably more uniform and less sensitive to the change of the axial position. Increasing Gs, the areas of the relatively dilute region continue decreasing. The solids holdup in the center of the riser is nearly 0.1 under extremely high Gs. Moving outwards towards the wall, solids holdup increases monotonically. Solids holdup remained greater than 0.4 over a wide cross-sectional area (r/R ¼0.7–1.0, about 60% of the cross-sectional area) even at the top section of the riser. Moreover, flow development is much slower under high flux as the radial profiles continued to change as seen in Fig. 3. The above phenomenon suggests that low solids flux data has very limited usefulness to high solids flux reactor modeling and design, especially for solids fluxes within the industrial range (400–1200 kg/m2 s) in FCC riser. Therefore, there is clearly a need to conduct more fundamental researches to study both axial and radial profiles of solids holdup and flow structures in CFB systems operating at higher flux and/or density. However, achieving high flux/density in a CFB system is extremely difficult in any experimental lab. While a few (only a few) research groups (e.g. University of British Columbian, UBC, Vancouver and University of Western Ontario, UWO, Ontario) have tried to obtain high flux/density operating, the solids fluxes are still far below the practical fluxes in industrial reactor processes. Theoretically, Bi and Zhu (1993) proposed that high densities and high solids fluxes could be accomplished by a combination of high solids inventories, large downcomer-to-riser diameter ratio, a low pressure drop solids feeder, and minimizing pressure drops in solids separation devices and fittings along the CFB loop. Besides, a proper blower and suitable particle size/riser diameter combinations are also of importance. Based on their suggestion, a dual loop CFB was used to reduce total pressure drop of the recycle system at UBC which enabled a high solids flux of abound 400 kg/m2 s to be achieved. A twin-riser CFB with a large downcomer-to-riser ratio was constructed at UWO which achieved high fluxes up to 500 kg/ m2 s. Building further upon the practical experience, we have: (1) Installed a large diameter storage tank at the bottom of the downcomer; (2) replaced the blower with a compressor of high capacity 1000 SCFM at 100 psi; (3) installed an additional air exhaust pipe at the top of the of the downcomer to discharge most of the air flowing upward through the downcomer, minimizing downcomer air flowing into the primary cyclone, so that the pressure drop across the cyclone is significantly reduced. This step increased the available pressure for the riser to achieve higher density; (4) installed two small deflecting plates (see the left insert of Fig. 1) in the solids inlet region, one vertically at the outlet of
Fig. 4. Axial solids holdup distribution for various operating conditions.
the inclined pipe covering 30% of the lower end of the inclined feed pipe joining the riser to prevent the riser air from flowing into the solids feeding pipe which tends to restrict solids downflow. The other deflecting plate was half-way up in the inclined pipe covering 30% of the cross-sectional area of the inclined pipe. It directed particles downwards so as to provide
238
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
a quit “exit route” for the remaining air entering the feed pipe so that the solids movement in the inclined pipe is much faster and steadier. After the modification, the riser can be operated much more steadily at solids circulation rates of up to 1000 kg/m2 s, much
higher than 400 kg/m2 s in the original system reported by Li (2010). This allows us to operate the CFB system under a wide range of operating conditions to obtain a comprehensive map of solids flow in the new solids recycle loop. An example is the data set plotted in Fig. 3 and discussed above.
3.2. Axial profiles of solids holdup
Fig. 5. Comparison of solids holdup profiles under different operating conditions.
Fig. 4 displays the axial distribution of the averaged solids holdup in the CFB riser for superficial gas velocity, Ug, of 5, 7 and 9 m/s and solids flux, Gs, up to 1000 kg/m2 s. As shown in Fig. 4, significantly different axial profiles can be seen under various operating conditions. In general, as solids circulation rate (Gs) increases, the approximately exponential-shaped axial profile is replaced by the relatively non-uniform axial profile with a dense region at the rsier base and a dilute region at the top section and then at even higher solids fluxes, by more uniform axial profiles. In Fig. 4(a) and (b), when Ug is constant the solids holdup increases with increasing Gs, while for a constant Gs, the solids holdup decreases with increasing Ug as plotted in Fig. 4(c). In details, relatively high solids holdups near the bottom of the riser are extensively affected by the particle acceleration and the gas distributor. Above this region, at the lowest Gs of 100 kg/m2 s, the solids holdup decreases exponentially, eventually approaching a constant value up to the riser exit. This exponential shape occurs when particles entering in the riser bottom are immediately
Fig. 6. Radial solids holdup distribution for various operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
entrained without significant. By increasing Gs gradually, a significant dense region at the riser0 s bottom forms, leading to a nonuniform axial flow structure. It can also be seen that the solids holdup along axial elevations is increasing with increasing Gs at a fixed Ug. Meanwhile, the solids holdup is also increasing at each axial level with decreasing Ug when Gs is constant. The dense region occurs between 2 and 4 m heights with the high solids holdup raging from 0.1 to 0.25. Above this region, relatively dilute regions are reached and the average solids holdup becomes independent of the axial variation. It is clear that the 2 melevation lies in the developing flow region with high solids holdups and a fully-developed region with relatively low solids holdup begins at 5 m height. Interestingly, the shape of the axial profile is hardly changed with various operating conditions. It is apparent that the axial profiles move in parallel from low solids holdup towards high solids holdup with increasing Gs and/or decreasing Ug. The solids holdup and the flow development in both regions of the riser do not seem to depend on the height, but are expected to depend on the operating conditions. This may be partly affected by pressure balance in the whole CFB recycling system. If the solids inventory is high, the pressure head in the return system would be sufficiently high. Therefore, it is easy to adjust the pressure in the return system to meet the requirement for pressure balance in the whole loop under various operating conditions. Moreover, this axial profile is different from the heightdependent S-shaped axial profile reported by other authors under
239
high flux and/or high density operating conditions (Issangya et al., 1999; Pärssinen and Zhu, (2001a), (2001b) and Yan and Zhu, 2004). Further increasing Gs, significantly dense solids holdup in the entire height of the riser is achieved ranging from 0.23 to 0.38 when solids circulation rate is extremely high, particularly higher than 700 kg/m2 s. The uniform and dense gas–solids suspension has been achieved along the whole column if ignoring the entrance effect. This homogenous axial structure is similar to other results reported by previous researchers. For instance, Contractor and In Avidan, (1994) found that solids holdup between 0.15 and 0.2 could cover the entire riser for a solids flux up to 685 kg/m2 s and gas velocity up to 5.7 m/s. In contrast to the nonuniformity axial profiles with solids circulation rate lower than 700 kg/m2 s, the dense bed at the base riser of the can persist over the entire column under such high Gs. Axial distribution of solids holdup becomes very uniform. As already noted, being able to achieve this kind of homogenous axial profile depends on being able to provide sufficient pressure head. If the pressure head is high enough, the homogenous axial profile appears to be robust and self-sustaining over a considerable range of operating conditions Information about the effect of Gs on the respective axial profiles in an industrial riser is very useful since the solids flux varies widely from 400 to 1200 kg/m2 s in the industrial reactors. According to the concept of the high density operation as mentioned in the introduction section, it is clear that what have
Fig. 7. Overall view of the solids hold up under different operating conditions.
240
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
observed in this study in the high flux riser does match the high flux and/or high density operating conditions suggested by Zhu and Bi (1995). In addition, the result is also comparable to the results (0.25–0.30) in C-TFB with a highly uniform and dense suspension flow structure operated at low superficial gas velocity and high solids circulation rate reported by Zhu and Zhu (2008a), (2008b). The uniform high density structure stresses the importance to both uniform gas–solids contacting efficiency and uniform bed-to-wall heat transfer throughout the whole reactor.
3.3. Radial distribution of solids holdup Typical radial profiles of solids holdup in the CFB riser are presented in Fig. 5 under the operating conditions as well as at different axial positions. Under all operating conditions, radial distributions of the solids holdup are nonuniform with dilute and dense regions. The radial profile is relatively flat in the center of the riser and the solids holdup increases towards the wall with the maximum value right at the wall at 0.56. Operating conditions affect dramatically the radial solids holdup distribution. The riser gradually becomes denser from the wall towards the center with increasing solids feeding at a fixed Ug. The solids holdup and its radial distribution in the high density CFB riser are quite different from those in the low density systems. Obviously, the radial profile is a clear “core-annulus” structure when Gs is low. The radial variation can be divided into three parts: a central region up to r/R E0.5–0.6 with a dilute and uniform solids holdup, an intermediate region between r/R E 0.5–0.6 and r/RE 0.8–0.9 where solids holdup appreciably increases, and a wall region when r/ R4 0.9 where the solids holdup is high but not more than 0.35. The results are similar to those reported by other researchers under comparable operating conditions (Liu, 2001; Yan and Zhu, 2004). As Gs increases to higher than 700 kg/m2 s, the dilute region shrinks (r/R ¼0–0.2, less than 20% of the cross-sectional area). After this short region, solids holdup increase gradually towards the wall which can be up to 0.5. The “core-annuals” radial profile is
replaced by the concave parabolic curve under extremely high Gs. More details can be shown in Fig. 6. Regarding the results in columns I, II and V in Fig. 6, it is clear that the superficial gas velocity plays an important role in solids distribution in radial and axial directions. It is apparent that reducing Ug results in an increased solids holdup and thus an increased solids holdup profile. Additionally, superficial gas velocity influences the solids holdup not only in the wall region but also in the center of the riser. For instance, when Ug increases from 5 to 9 m/s, the solids holdup decreases in the center region at different heights and drops more rapidly in the wall region especially at the bottom of the riser. Moreover, the solids holdup distribution is less uniform at lower superficial gas velocity. Through comparison between columns II, III and IV and/or between V, VI and VII, increasing solids circulation rate leads to a higher solids holdup with a less uniform profile. Under high Gs, solids holdup gradient increases and solids distribution becomes less uniform because of the more confined wall effects to the near wall region. Lower Gs leads to a more even radial solids distribution with lower local solids holdup across the column. It is also to see that axial distributions of radial solids holdup vary at different solids fluxes. When Gs is lower than 500 kg/m2 s, the solids holdup in the wall region at the upper section is lower than that in the bottom section. However, the axial variations of radial profiles of solids holdup changed very little under extremely high solids circulating rate of 1000 kg/m2 s.
3.4. Flow development of solids holdup Solids flow can be considered as fully developed if the radial solids distribution remains relatively unchanged with rising axial location (Qi et al., 2009). A detailed review of Figs. 3 and 6 shows that increasing Ug accelerates the solids flow development. On the other hand, when Gs is increased from 300 kg/m2 s to 500 kg/m2 s and then to 700 kg/m2 s at Ug ¼7 m/s, the flow development becomes much slower. A similar trend is also observed when Ug is 9 m/s.
Fig. 8. Solids holdup distribution in different radial regions for various operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
To facilitate the analysis of flow development, three-dimensional profiles of solids holdup are given in Fig. 7. It shows clearly that in the riser center, solids holdup remains nearly invariable in the entire column when Gs o700 kg/m2 s. Similar results are also reported by previous studies under high flux conditions (Issangya et al., 2000; Pärssinen and Zhu, (2001a), (2001b); Yan and Zhu, 2004 and Huang et al., 2007). However, when Gs is up to or higher than 700 kg/m2 s, the flow development becomes extremely slower with a reduced core region of relatively low solids holdups, and the solids holdup is increasing gradually from the core region toward the wall. The radial solids holdup profiles are typically parabolic in shape as aforementioned. In order to further investigate the flow behavior, it is helpful to examine flow development in the different radial locations. To this end, the whole cross-sectional area is arbitrarily divided to three parts: r/R ¼0.0–0.548, 0.548–0.837 and 0.837–1.0. Axial profiles of the solids holdups in these different radial regions are plotted in Fig. 8. As shown in the left of the Fig. 8, in the center where r/R ¼ 0.0– 0.548, the solids holdup is very low with no significant difference in this region throughout the entire riser at various superficial gas velocities. In the intermediate region with r/R ¼0.548–0.837 solids holdup varies near the gas distributor and followed by a flat trend. Near the wall (r/R ¼0.837–1.0), the solids holdup considerably varies with sharply dropping in the entrance and then becomes relatively flat till the top of the column. A similar trend in flow development is also observed when Gs is below 500 kg/m2 s when Ug ¼9 m/s. When Gs is higher than 500 kg/m2 s, especially higher than 700 kg/m2 s, the solids holdup in the center remains at about 0.1 even at the outlet of the column. Solids holdup in the middle region is high up to 0.4 and remains unchanged along the riser up to 4 m-level and then gradually decreases. The solids holdup still remains much higher of up to 0.25 in this intermediate region. While in the wall region, solids suspension density is rather flat with an extremely high value of up to 0.55 along all the axial elevations at the highest solids flux, Gs ¼1000 kg/m2 s. This insensitivity of solids holdup across the whole radial area to the axial variations may suggest that a saturation state is to some extent reached in the extremely high density CFBs. The flow development can also be shown by the radial nonuniformity index (RNI) which was proposed by Zhu and Manyele (2001). Based on their definition, the RNI of solids holdup, RNI(εs), can be defined as RNIðεs Þ ¼
sðεs Þ sð ε s Þ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sðεs Þmax εs ðεs;mf εs Þ
ð4Þ
where s(εs) is the standard deviation of the radial solids holdup, s(εs)max is the maximum possible standard deviation, εs is the average solids holdup, and εs,mf is the highest possible solids holdup. Fig. 9 shows the RNI (εs) along the axial direction in the riser. In general, all RNI (εs) profiles are showing that the value of RNI (εs) is higher at the bottom of the riser followed by a relatively flat curve along the axial direction. Obviously, higher RNI (εs) values at the bottom are due to the development of solids flow. The effect of operating conditions on the RNI (εs) can be inferred from Fig. 9(a). RNI (εs) decreases with increasing superficial gas velocity, showing that the radial profiles of solids holdup become more uniform. High gas velocity can enhance the gas– particle interactions. Solids flow development would be accelerated by high drag force impacted by high speed gas so that the solids flow structure becomes more uniform. At a given Ug, high Gs results in higher RNI (εs) because of the increased solids holdup which may induce the formation of the particle aggregation and thus increase the radial nonuniformity. Moreover, Fig. 9(b) shows a
241
Fig. 9. RNI of solids holdup the riser under different operating conditions (a) RNI (εs) and (b) relationship between RNI and solids holdup.
relationship between the RNI (εs) and average solids holdup. It shows that the RNI (εs) values increase with the average solids holdups the radial profiles of solids holdup become steepened with increasing average solids holdup (Zhu and Manyele, 2001). It is obvious that stable RNI (εs) is achieved at higher levels along the riser with increasing of Gs or decreasing of Ug. Take Ug ¼9 m/s as an example, an increase in solids circulation rate from 100 to 600 kg/ m2 s, extends the distance from 1 m to about 6 m. Further increasing Gs, the distance for the stabilization of RNI (εs) appears to extend beyond the total length of the riser. The phenomenon again verifies the acceleration process to achieve complete flow development in the CFBs. 3.5. Flow fluctuations in the high flux/density riser From the foregoing analysis, it is obvious that the local solids holdups vary in the radial direction. In order to gain a better understanding of the local solids flow, it is helpful to examine the flow fluctuations under different operating conditions, especially in high density (HDCFB) and low density circulating fluidized beds (LDCFB). Fluctuations can be quantified by the standard deviation and the intermittency indices (γ) of solids holdups. Definition of γ in details can be found in the previous paper (Brereton and Grace, 1993). Besides, the intermittency index can also be used to
242
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
describe the phenomenon of segregation between gas and solids. The value of intermittency index is equal to one if the gas–solids suspension is considered as an “ideal cluster flow”. On the other hand, the value of intermittency index is zero if there is a “perfect core-annulus flow”. Fig. 10 compares the radial profiles of local solids holdup, standard deviation and the corresponding intermittency index at two axial heights. Generally, the local solids holdup is lowest and has a relatively flat radial distribution under the low flux of 100 kg/m2 s. With increasing solids flux or decreasing gas velocity, the solids holdup increases, and the radial distribution becomes steeper. The standard deviation and γ increase with increasing Gs or decreasing Ug. This suggests increased flow fluctuation with the increase of Gs or decrease of Ug. The fluctuation of low flux solids holdups increases outward, reaching maxima at the outer wall. For high flux conditions, the fluctuation increases gradually to the highest value at r/R ¼0.6–0.7 and thereafter decreases towards the wall. The greatest fluctuation always occurs in the middle region of the column under high flux operating conditions which are also reported under high density operating conditions in many papers (Issangya et al., 2000; Zhu and Zhu, 2008a, 2008b, Qi et al., 2009). This phenomenon has been considered as one of the most important distinguishing features in HDCFBs. When Gs is lower than 500 kg/m2 s, the
standard deviation and the corresponding intermittency index remains a very low level around 0.02–0.05 up to r/R E 0.4–0.6. Considering its solids holdup profile, this indicates that a wide dilute region still occupies the center surrounded by a dense annulus region at the wall. Lower solids holdups in the center of the column and higher solids concentrations near the wall make fewer fluctuations. It is found that the standard deviations and intermittency indices under extremely high solids flux of 1000 kg/ m2 s are much higher than those under other low high solids flux conditions at almost all radial positions. This highest magnitude of fluctuations may due to the relatively higher solids holdup suggesting that the gas–solids and inter-particle interaction in this particularly high flux CFB are more vigorous than those in relatively low solids flux CFBs operations. The dramatic fluctuations probably induce better gas–solids contacting and mixing, improving consequently the reactor performance. Additionally, for each solids circulation rate, there is a clear trend for the development of standard deviation and γ with axial variations. Clearly, both parameters tend to decrease with the increasing axial level. This illustrates that local solids flow is liable to change with increasing height from one where cluster-flow structures are more common to one where the core-annulus structure becomes predominant. The same results are also presented by Brereton and Grace (1993).
Fig. 10. Radial profiles of solids holdup, standard deviation and intermittency indices along the riser under different operating conditions.
C. Wang et al. / Chemical Engineering Science 108 (2014) 233–243
4. Conclusions The current work presented a comprehensive insight into the solids holdup and flow development phenomena in a high flux/ density (up to 1000 kg/m2 s) riser, using an optical fiber. Solids suspension having a solids holdup of up to 0.2–0.3 can be maintained throughout the entire high flux/density riser. A homogenous axial flow structure is observed at Gs ¼1000 kg/m2 s. Radial distributions of the solids holdup are nonuniform with a dilute region and a dense region. For Gs 4700 kg/m2 s, the dilute core region shrinks to about r/R ¼0–0.2, less than 20% of the crosssectional area. Solids holdups thereafter increase monotonically towards the wall reaching up to 0.56. The speed of flow development varies in different crosssectional areas. Flow development is almost instant at the center through the entire riser even at the high density conditions, while it becomes much slower in the wall region. Increasing solids flux prolongs the solids flow development. Better gas–solids contacting and mixing indicated by standard deviation and intermittency index of the solids holdup over the entire cross-sectional area under extremely high solids flux can greatly lead to vigorous interactions between gas and solids phases, improving the reactor performance. Notation f Gs Le r/R RNI(εs) t T Ug V(t) z
Function between the voltage and solids holdup (dimensionless) solids circulation rate (kg/m2 s) effective distance between the two subprobes (m) normalized radial position radial nonuniformity index of solids holdup time period of particles through the two subprobes (s) time period of testing (s) superficial gas velocity (m/s) voltage time series (V) distance above the gas distributor (m)
Greek letters
εs εs,mf εs(t) εs τ s(εs) s(εs)max
solids holdup (dimensionless) the highest solids holdup (dimensionless) instantaneous solids holdup (dimensionless) average solids holdup in the entire column (dimensionless) lag time (s) standard deviation of the radial solids holdup (dimensionless) maximum possible standard deviation (dimensionless)
Subscripts 1, 2 g p s
subprobe 1 and 2 of optical fiber probe gas particle solids
References Azzi, M., Turlier, P., Betrand, J.R., 1991. Mapping solids concentration in a circulating fluidized bed using gammametry. Powder Technol. 67, 249–258.
243
Bi, H.T., Zhu, J.X., 1993. Static instability analysis of circulating fluidized beds and concept of high-density risers. AIChE J. 39, 1272–1280. Bi, H.T., 2004. Gas and solid mixing in high-density CFB risers. Int. J. Chem. React. Eng. 2, A12. Bi, H.T., Grace, J.R., Zhu, J.X., 1993. Types of choking in vertical pneumatic systems. Int. J. Multiph. Flow 19, 1077–1092. Brereton, C.M., Grace, J.R., 1993. Microstructural aspects of the behavior of circulating fluidized beds. Chem. Eng. Sci. 48 (14), 2565–2572. Contractor, R.M., In Avidan, A.A., 1994. Circulating Fluidized Bed Technology IV. American Institute of Chemical EngineersNew York, pp. 387–391. Ellis, N., Bi, H.T., Lim, C.J., Grace, J.R., 2004. Influence of probe scale and analysis method on measured hydrodynamic properties of gas-fluidized beds. Chem. Eng. Sci. 59 (8-9), 1841–1851. Grace, J.R., 1986a. Contacting modes and behavior classification of gas-solids and other two-phase suspensions. Can. J. Chem. Eng. 64, 353–363. Grace, J.R., 1986b. Heat transfer in circulating fluidized beds. In: Basu, P. (Ed.), Circulating Fluidized Bed TechnologyPergamon Press, Toronto, pp. 63–81. Grace, J.R., 2000. Reflections on turbulent fluidization and dense suspension upflow. Powder Technol. 113, 242–248. Huang, W.X., Yan, A.J., Zhu, J.X., 2007. Hydrodynamics and flow development in a 15.1 m circulating fluidized bed riser. Chem. Eng. Technol. 30 (4), 460–466. Issangya, A.S., 1998. Flow Dynamics in High Density Circulating Fluidized Beds (Ph. D. diss.). Univ. of British Columbia, Vancouver, BC. Issangya, A.S., Bai, D.R., Bi, H.T., Lim, K.S., Zhu, J., Grace, J.R., 1999. Suspension densities in a high-density circulating fluidized bed riser. Chem. Eng. Sci. 54, 5451–5460. Issangya, A.S., Grace, J.R., Bai, D.R., Zhu, J.X., 2000. Further measurements of flow dynamics in a high-density circulating fluidized bed riser. Powder Technol. 111, 104–113. Johnsson, H., Johnsson, F., 2001. Measurements of local solids volume-fraction in fluidized bed boilers. Powder Technol 115, 13–26. Karri, S.B.R., Knowlton, T.M., 1999. A comparison of annulus solids flow direction and radial solids mass flux profiles at low and high mass fluxes in a riser. In: Werther, J. (Ed.), Circulating Fluidized Bed Technology VI. Frankfurt, Dechema, pp. 71–76. Knowlton, T., 1995. Interaction of pressure and diameter on CFB pressure drop and holdup. Paper for Workshop, Modeling and Control of Fluidized Bed Systems. Hamburg, May 22–23. Li, D.B., 2010. Investigation of circulating Fluidized Bed Riser and Downer Reactor Performance for Catalytic Ozone Decomposition (Ph.D. diss.), University of Western Ontario, London, ON. Liu, J.Z., Grace, J.R., Bi, H.T., Morikawa, H., Zhu, J.X., 1999. Gas dispersion in fast fluidization and dense suspension upflow. Chem. Eng. Sci. 54 (22), 5441–5449. Liu, J.Z., 2001. Particle and Gas Dynamics of High Density Circulating Fluidized Beds (Ph.D. diss.). Univ. of British Columbia, Vancouver, BC. Liu, J.Z., Grace, J.R., Bi, H.T., 2003a. Novel multifunctional optical-fiber probe: development and validation. AIChE J. 49 (6), 1405–1420. Liu, J.Z., Grace, J.G., Bi, H.T., 2003b. Novel multifunctional optical-fiber probe: highdensity CFB measurements. AIChE J. 49 (6), 1421–1432. Martin, M.P., Turlier, P., Bernard, J.R., 1992. Gas and solid behavior in cracking circulating fluidized beds. Powder Technol. 70, 249–258. Pärssinen, J.H., Zhu, J.X., 2001a. Axial and radial solids distribution in a long and high-flux CFB riser. AIChE J. 47, 2197–2205. Pärssinen, J.H., Zhu, J.X., 2001b. Particle velocity and flow development in a long and high-flux circulating fluidized bed riser. Chem. Eng. Sci. 56, 5295–5303. Qi, X.B., Zhu, H.Y., Zhu, J., 2009. Demarcation of a new circulating turbulent fluidization regime. AIChE J. 55 (3), 594–611. Reh, L., 1999. Challenges of circulating fluid-bed reactors in energy and raw materials industries. Chem. Eng. Sci. 54 (22), 5359–5368. Yan, A.J., Zhu, J.X., 2004. Scale-up effect of riser reactors (1): axial and radial solids concentration distribution and flow development. Ind. Eng. Chem. Res. 43, 5810–5819. Zhang, H., Johnston, P.M., Zhu, J.X., de Lasa, H.I., Bergougnou, M.A., 1998. A novel calibration procedure for a fiber optic solids concentration probe. Powder Technol. 100, 260–272. Zhu, H.Y., Zhu, J.X., 2008a. Gas-solids flow structures in a novel circulatingturbulent fluidized bed. AIChE J. 54 (5), 1212–1223. Zhu, H.Y., Zhu, J.X., 2008b. Comparative study of flow structures in a circulatingturbulent fluidized bed. Chem. Eng. Sci. 63, 2920–2927. Zhu, J.X., Grace, J.R., Lim, C.J., 1989. Erosion-causing particle impacts on tubes in fluidized beds. In: Grace, V.I., Schemilt, J.R., Bergougnou, L.W., M.A. (Eds.), FluidizationEngineering Foundation, New York, pp. 613–620. Zhu, J.X., Bi, H.T., 1995. Distinctions between low density and high density circulating fluidized beds. Can. J. Chem. Eng. 73, 644–649. Zhu, J.X., Manyele, S.V., 2001. Radial nonuniformity index (RNI) in fluidized beds and other multiphase flow systems. Can. J. Chem. Eng. 79, 202–211. Zhu, J.X., Cheng, Y., 2005. Fluidized-bed reactors and applicationsIn: Clayton, Crowe (Ed.), Multiphase Flow Handbook. CRC Press, New York5.55–5.93 (Chapter 5.3).