Fluidization is the operation by which solid particles are transformed into
fluidlike state through suspension in a gas or liquid. This method of contacting
between solid and fluid has some unusual characteristics, and fluidization engineering
puts them to good use (Kunii and Levenspiel, 1991). The
basic mechanism of a fluidized bed can be seen simply as fluid percolation through
particle interstices via a distributor, in which particles begin to exhibit
fluidlike characteristic upon experiencing sufficient drag force from the fluid.
A large number of industrial processes use the fluidization technique in their
daily operations, namely combustion, gasification of solid fuels, drying of
particles, particle heating, oxidation, metal surface treatments and catalytic
and thermal cracking (Howard, 1989).
Since half a decade, hundreds of fluidized bed concepts and configurations
were patented and came into actual industrial application around the globe.
Most of them are specifically designed to serve a particular application as
mentioned before. Typically, a fluidized bed is cylindrical in shape and operates
using perforated-plate distributor, which holds the particles and distributes
the fluidizing gas into the bed. In these types of beds, the fluidizing gas
velocity entering the bed can be resolved into three components: axial, radial
and tangential components.
The axial and radial components are responsible for the particle movement that
causes bed mixing. The tangential component generates mixing along an annular
path. The fluidizing gas discharged from the conventional distributor possesses
only axial momentum. A deficiency in the radial and tangential momentum leads
to vertical mixing many times faster than the lateral mixing (Geldart,
1986). Thus, it had become a classical problem in a fluidized bed to obtain
a good lateral mixing and to reduce elutriation of particles during operation.
Innovative techniques have been proposed by various researchers to overcome
this issue. One such technique is the swirling fluidization technique or vortexing
bed, or sometimes even designated as toroidal bed as proposed by Shu
et al. (2000) and Wellwood (1997). The swirling
fluidization technique minimizes the axial momentum transferred to the particles
by the fluidizing gas whereas a larger fraction of momentum is now being transferred
radially and tangentially. The particles are now subjected to vigorous mixing
which eventually increases the heat and mass transfer, to and from the bed,
while reducing the elutriation of particles.
Typical methods of achieving swirling fluidization are through the secondary
injection of fluidizing medium into the freeboard tangentially or by using a
distributor which provides inclined injection of the fluidizing medium into
the bottom of the bed. Other known methods are simply by either rotating the
distributor or the bed column at certain angular speed.
The current study focuses on swirling fluidization using an annular blade distributor.
In spite of the potentials for vast industrial applications of the swirling
fluidized bed, there are very few systematic studies of such a bed, though the
principle has already been used in commercial equipment. From the literature
review carried out, there is an apparent lack of reliable experimental data
related to the hydrodynamic regimes and characteristics, particularly those
operating with large particles. Because of the stable mixing regime with the
absence of bubbles and hence, of gas bypassing through slugging, swirl bed has
a bright future in solid-gas processing.
Thus the current research was carried out to investigate the regimes of operation
and hydrodynamic characteristics of such bed with different configurations,
considering two fraction of open area (12.9 and 17.2%), two particles sizes
(5.75 and 9.84 mm), variable bed loadings (0.5 to 2 kg) for increasing superficial
velocities up to 7 m sec-1. Discussion and explanations were also
Many researchers proposed the swirling fluidization technique as a means to
overcome the deficiencies in the conventional beds. Chyang
and Lin (2002) proposed a multi-horizontal-nozzle distributor design to
improve lateral mixing. The distributor offers the fluidizing gas additional
momentum in both radial and tangential directions while minimizing the axial
momentum. They conducted pressure fluctuation and fines elutriation experiments
using glass beads in cold bed before comparing with conventional fluidized bed
(perforated-plate distributor). The authors reported that the multi-horizontal
nozzle distributor produced remarkable improvement in fluidizing quality with
reduced elutriation of fine particles but the presence of dead zone under and
between the nozzles could not be avoided and thus the particles in this region
De Wilde and Broqueville (2008) experimented a new
concept of rotating fluidized. The bed applies the injection of fluidizing gas
tangentially in to the fluidization chamber via multiple entry gas inlet slots
in its cylindrical wall while rotating at a certain speed. The fluidizing gas
is forced to exit from the bed through a centrally positioned chimney. As the
tangential entry gas fluidizes the particles tangentially, a swirling field
is created while the rotation of the bed creates a centrifugal field. The authors
report that sufficiently high solid loading is crucial to obtain a stable and
uniform fluidization. The fluidizing gas flow rate was found to have a minor
effect on fluidization. The performance of the bed is not addressed, but it
is believed to be low. The configuration of the bed is highly complicated, suggesting
a high initial and maintenance cost.
Similar study was done by Wormsbecker et al. (2007).
They investigated the influence of three different types of distributor on a
fluidized bed dryer hydrodynamics. The distributors are perforated-plate, wire
mesh and punched plate. The punched plate consists of hooded openings of 5.75
by 1 mm oriented in a circular pattern which forms rings, separated by 3 mm
from each other. This punched plate distributor was designed to produce a swirling
effect in the bed. Fluidization quality was assessed in terms of drying times
and standard deviation and power spectrum analysis of the pressure fluctuations.
Again, the swirling motion of particles in punched plate distributor is reported
to have superior fluidization performance, but at the expense of high pressure
drop, and thus not suitable for low cost operations.
Lin et al. (1998) reported their findings from
numerical and experimental study on the swirling flow field in a Vortexing Fluidized
Bed Combustor (VFBC). The VFBC was made by a conventional fluidized bed with
four secondary air injection nozzles with tangential entry. The performance
of the flow field was assessed based on the vortex number. The results show
that the vortex number increased with secondary air injection that creates swirling.
However the method is only useful for relatively large beds since the jets from
the nozzles tend to interfere with each other for smaller beds, thus weakening
the swirl effect. The authors also reported that flow turbulence and elutriation
rate could not be established accurately.
Another bed that operates using swirling fluidization technique is the Swirling
Fluidized Bed (SFB), which is the main focus in this study. The bed is annular
type, featuring angular injection of gas and swirling motion of bed material
in a circular path as shown in Fig. 1.
The principle of operation is based on the simple fact that a horizontal component
of gas velocity in the bed creates horizontal motion of the bed particles. A
jet of gas enters the bed at an angle θb to the horizontal.
Due to angular injection, the gas velocity has two components. The vertical
component Uv = U sin θb, causes lifting of the particles.
It is this lifting force that is responsible for fluidization. The horizontal
component Uh = U cos θb, creates a swirling motion
of the particles (Shu et al., 2000; Sreenivasan
and Raghavan, 2002). The bed particles are also likely to undergo a secondary
motion in a toroid-like path and be well mixed in the radial plane.
|| Basic configuration of a swirling fluidized
This variant of fluidized bed provides an efficient means of contacting between
gas and particles. Elutriation of particles which has been a major limiting
factor in the operation of the conventional fluidized bed is reduced significantly,
since the vertical component of velocity is now only a small fraction of the
net gas velocity. The cyclone-like features resulting from the swirling motion
of bed particles also contribute to this low elutriation. Hence it is possible
to fluidize very fine particles and a wide variety of shapes of particles in
this kind of fluidized bed.
Shu et al. (2000) studied a similar bed, termed
as the toroidal bed, which is taken from the overall shape of the bed in the
swirling regime. Relevant hydrodynamic behaviors of the bed are measured with
various inert materials in a pilot scale 400-mm toroidal fluidized bed reactor.
The observed hydrodynamic behavior is found to be essentially predictable at
ambient temperature by conventional hydrodynamic models.
Earlier Wellwood (1997) described the toroidal gassolid
action as horizontal fast fluidization and attempted to predict
the slip velocity in dilute gassolid systems by using an equation analogous
to the ideal gas equation. The study was based on the observed similarities
between microscopic molecules and macroscopic particles. Thus, the author adopted
an assumption that under dilute phase conditions, the analogy is applicable.
He concluded that the experimental results gave general support for the analogy
Sreenivasan and Raghavan (2002) developed an analytical
model on the hydrodynamics of a swirling fluidized bed. The model is put forward
to predict the angular velocity of the swirling bed at given air flow rate and
also the pressure drop of the swirling fluidized bed. In this model, the bed
is treated as a lumped system; the whole bed is a single swirling mass of uniform
angular velocity. The model was developed based on the conservation of angular
momentum principle and the authors validated the model with experimental works.
Recently, Kaewklum and Kuprianov (2010) investigated
the hydrodynamic regimes and characteristics of a conical fluidized bed operating
with annular-blade distributor. Using quartz sand with four different particle
sizes, they compared the pressure drop and minimum fluidization velocity with
four-nozzle tangential entry system which also generates swirling motion in
the bed. From the cold model tests, they concluded that the method of air injection
substantially affected the hydrodynamics and fluidization regimes. However,
the swirling effect can only achieved at higher superficial velocities, particularly
with low bed heights.
Using the same bed with annular-blade distributor, (Kaewklum
et al., 2009) again reported hydrodynamics, now together with combustion
and emission characteristics of rice husks. Four regimes of operation were obtained,
favoring the fully swirling regime where the combustion temperature profiles
were rather uniform, suggesting intense heat and mass transfer in the radial
direction. The authors concluded that the swirling fluidization in the conical
bed yielded very high combustion efficiencies coupled with reduced emissions.
Though comprising such merits, the SFB also comes with several drawbacks. Towards
attending to these deficiencies through the proposal of a novel distributor,
it is imperative to first understand fully the existing bed characteristics
and other bed configurations which have not been addressed in the literature.
MATERIALS AND METHODS
A well-planned methodology is important to meet the desired project objectives.
Here, details of the experimental apparatus including distributor design, blower
selection, flow measurement and experimental procedure for pressure drop measurement
Annular distributor: The distributor assembly consists of lower and
upper flanges together with inner plates respectively, holding the blades firmly
to form the annular distributor, somewhat similar to that used by Shu
et al. (2000), Sreenivasan and Raghavan (2002),
Kaewklum and Kuprianov (2010) and Kaewklum
et al. (2009). The centre body is important to avoid the possible
creation of dead zone at the centre of the bed during operation
with bed materials. Air enters the plenum chamber via tangential entry and expands
before entering the annular blade distributor.
Particle and blades description: Particles used in this experiment are
large spherical PVC beads, which fall in Geldart type D particles as proposed
by Geldart (1986).
|| Blades used to form annular
|| Blade angle and blade overlapping angle
|| Particle density and size
Two different sizes of particles are used, with their respective density and
diameters are shown in Table 1.
Similarly, two versions of blades with different geometry are used in the current
work as shown in Fig. 2. The blades resemble the shape of
truncated sectors of a circle, having angles of 15° and 18°, respectively.
The overlapping length between two successive blades helps to direct the air
at the designed angle. Larger overlapping lengths of blades may result in higher
distributor pressure drop. This is because larger overlapping lengths constrain
the flow of air much longer between the blade s before it enters the bed. Since
the blades resemble the shape of sectors as shown above, the over-lapping length
is more appropriate to be termed as blade overlapping angle as shown in Fig.
3. Larger angular value naturally represents larger over-lapping area of
blade. Thus, the 15° and 18° blades have the overlapping angles of 9°
and 12°, respectively.
RESULTS AND DISCUSSION
The findings from hydrodynamic characteristics for various bed configurations
are presented and the regimes of operation observed are also discussed.
||Variation of bed loading (0.5 to 2 kg)
||Variation of particle sizes (9.84 and 5.75 mm)
Variation of FOA (12.9 and 17.2%)
||Variation distributor of blade overlapping angle (9° and
Findings from this study were also compared with those of a conventional fluidized
bed with perforated plates which has the same fraction of open area and bed
Regimes of operation: Typical regimes of operation in a conventional
fluidized bed include packed bed, minimum fluidization, bubbling, slugging and
finally elutriation. While operating a SFB, one can distinguish different regimes
of operation as shown in Fig. 4. Though the packed regime
(Regime I) still exists, progressive increase of fluidizing gas flow rate upon
minimum fluidization led to a condition suitably designated as the minimum swirling
condition (Regime II) where the bed almost swirls. Few particles even started
swirling gently at this point. Further increase in fluidizing gas flow rate
results in the desired swirling motion of the bed (Regime III). At this condition,
the bed is subjected to both fluidization and swirling where vigorous mixing
occurs and interaction between gas and particles are intense. This regime was
the largest regime where the particles tend to swirl faster with the increase
of fluidizing gas (thus increasing pressure drop) until finally reaching elutriation
(Regime IV) for shallow beds (bed weight less than 1000 g).
For deeper beds, 1500 g bed loading for instance, a two-layer bed is observed
as reported by Sreenivasan and Raghavan, (2002), Chyang
and Lin, (2002), Kaewklum and Kuprianov, 2010).
In a two-layer bed which occurs at a static bed height greater than 45 mm, a
thin, continuously swirling bottom layer and a vigorously bubbling top layer
are visible upon minimum swirling velocity. This is because the horizontal component
of the velocity is attenuated and finally vanishes at the interface between
the two layers as a result of continuous momentum transfer inside the bed. This
regime is shown in Fig. 5.
|| Regimes of operation in the SFB for shallow bed (1000 g bed
|| Regimes of operation in the SFB for deep bed (1500 gram bed
Effect of various bed configurations: Various configurations of the
swirling fluidized bed as outlined in the previous section are studied through
batch experiments and presented in the following section. Apart from pressure
drop and minimum fluidization velocity, quality of fluidization is determined
qualitatively through observation.
Effect of variable of bed loading: Bed loadings were increased from
500 to 2000 g in steps of 500 g to investigate the effect of variable bed loading,
which also corresponds to its respective bed height. Figure 6
shows ΔPb against Vs with cone as centre body. As
mentioned earlier, ΔPb increased with the increase of Vs
upon minimum fluidization. This distinct feature differentiates the SFB from
conventional beds. The reason for this feature is the increase of centrifugal
bed weight, which results in higher wall friction as proposed by Sreenivasan
and Raghavan (2002), apart from increasing friction between particles.
||Bed pressure drop against superficial velocity for variable
|| Bed pressure drop against superficial velocity for different
For deep beds, i.e., higher bed loadings, a two layer bed appeared, a swirling
bottom layer and bubbling top layer as discussed earlier. Higher bed loading
naturally impose higher pressure drops. With cylinder as centre body, slightly
higher pressure drops were obtained for all bed loadings.
Effect of particle size: Batch experiments were conducted with two
different particles, 5.75 and 9.84 mm for two different bed weights as in Fig.
It can be seen that in the packed region, larger particles have lower pressure
drop for both bed weights. This is due to the fact that smaller particles actually
have a larger surface area. Larger particles, on the other hand, are capable
of withstanding higher superficial velocity and hence longer swirling. Similar
trends are found for 1500 g bed weight for both particle sizes.
|| Bed pressure drop against superficial velocity for different
blade overlapping angle
|| Bed pressure drop against superficial velocity for different
number of blades
Effect of distributor blade overlapping angle and number of blades:
To bring out the effect of blade overlapping angle, two sets of distributor
blades, having 9° and 12° overlapping angles were investigated. Though
higher overlapping angle was expected to impose higher pressure drop since air
is forced to flow through longer blade opening, thus higher resistance, the
findings yield that this is not true to all bed configuration. Figure
8 show that higher overlapping angle results higher pressure drop only for
5.75 mm particles, while experiments with 9.84 mm particles suggest otherwise.
Less pressure drop consumed with larger overlapping length of blade, which is
actually good in terms of energy saving during operation.
||Bed pressure drop against superficial velocity with cone as
In Fig. 9, the number of blades is varied to observe the
effect fraction of open area (FOA). The FOA for 60 blades is calculated to be
12.89% from total distributor area, while for 30 blades, the FOA is 17.2%.
By using smaller number of blades, which is 30 blades in our case, the distributor
has larger FOA compared to that of 60 blades. Thus, larger momentum is now transferred
to the bed, resulting in higher pressure drop values in bed as depicted in the
Fig. 9 above. Operating with smaller particles also exhibits
similar feature where the bed pressure drop is higher for beds with smaller
number of blades (higher FOA).
Comparison between conventional fluidized bed and SFB: Bed pressure
drop against superficial velocity for both conventional fluidized bed and SFB
is also made, having the same fraction of open area, as in Fig.
Figure 10 clearly shows that SFB requires only half of the
potential energy for fluidization compared to the conventional bed. This is
true for all SFB configurations, including both small and large particle size.
Apart from that, the minimum fluidization velocities are also lower for SFB.
Therefore, it leads to a conclusion whereby SFB is superior compared to the
conventional fluidized bed. Therefore, we can expect better output when dealing
with actual processes involving solid-fluid contact when operating with SFB.
However, more experiments are needed to support this finding.
In conclusion, the SFB has been investigated through batch operations. The
findings indicate that the sequence of flow regimes in swirling fluidized bed
are packed bed, minimum fluidization, swirling regime, two-layer regime and
finally elutriation or transport regime. Deep beds are prone to form partially
fluidized regime and two-layer beds. Various configurations were also investigated
and the study concludes as below:
||The hydrodynamics of swirling fluidized bed are different
from other conventional fluidized bed, in which the pressure drop increases
with the mass flow rate of fluidizing gas
||Larger particles have lower pressure drop and capable of withstanding
higher superficial velocity and hence, larger swirling regime
||Larger overlapping angle imposes additional pressure drop,
particularly at the distributor since the air is now forced to flow through
higher resistance. But, larger overlapping angle also delays the presence
of two-layer as well as reducing the elutriation by expanding the swirling
||Particle size, bed weight and the number of blades (FOA) are
the most important variables that have more influence on the bed behavior.
The blade geometry has relatively smaller effect on the bed behavior.
The authors would like to express sincere gratitude to Universiti Teknologi
Petronas (UTP), Universiti Tun Hussein Onn Malaysia (UTHM) and the Ministry
of Higher Education (MOHE) for the opportunity and sponsorship to carry out
the entire research.