The Transition Velocities in a Dual Circulating Fluidized Bed Reactor with
Variation of Temperatures
Myung Won Seo, Jeong Hoi Goo, Sang Done Kim, Jae Goo Lee, Young Tae
Guahk, Nam Sun Rho, Geon Hoe Koo, Do Yeon Lee, Won Chul Cho, Byung
Ho Song
PII:
DOI:
Reference:

S0032-5910(14)00526-9
doi: 10.1016/j.powtec.2014.05.059
PTEC 10314

To appear in:

Powder Technology

Received date:
Revised date:
Accepted date:

30 August 2013
30 April 2014
31 May 2014

Please cite this article as: Myung Won Seo, Jeong Hoi Goo, Sang Done Kim, Jae Goo Lee,
Young Tae Guahk, Nam Sun Rho, Geon Hoe Koo, Do Yeon Lee, Won Chul Cho, Byung
Ho Song, The Transition Velocities in a Dual Circulating Fluidized Bed Reactor with
Variation of Temperatures, Powder Technology (2014), doi: 10.1016/j.powtec.2014.05.059

This is a PDF le of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its nal form. Please note that during the production process
errors may be discovered which could aect the content, and all legal disclaimers that
apply to the journal pertain.

ACCEPTED MANUSCRIPT
The Transition Velocities in a Dual Circulating Fluidized Bed Reactor with

Variation of Temperatures

SC
R

IP

Myung Won Seo1.2, Jeong Hoi Goo3, Sang Done Kim4*, Jae Goo Lee1,2, Young Tae Guahk1, Nam Sun Rho1,
Geon Hoe Koo1, Do Yeon Lee4, Won Chul Cho5, and Byung Ho Song6

MA

NU

Clean Fuel Laboratory, Korea Institute of Energy Research (KIER), 152 Gajeong-ro, Yuseong-gu,
Daejeon, 305-343, Republic of Korea
2
Department of New Energy Technology, UST (University of Science and Technology), 217
Gajeong-ro, Yuseong-gu, Daejeon, 305-333, Republic of Korea
3
Oil & Gas Process Team, SK Engineering and Construction, 100 Euljiro, Jung-gu, Seoul, 100-847,
Republic of Korea
4
Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and
Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 305-701, Republic of Korea
5
Hydrogen Laboratory, Korea Institute of Energy Research (KIER), 152 Gajeong-ro, Yuseong-gu,
Daejeon, 305-343, Republic of Korea
6
Department of Chemical Engineering, Kunsan National University, 1170 Daehangno, Gusan,
Jeonbuk 573-701, Republic of Korea

TE

Tel.: 82-42-350-3913, Fax: 82-42-350-3910, E-mail: kimsd@kaist.ac.kr

CE
P

Abstract

The extra-heavy oil fractions upgrading process, which consisted of a pyrolyzer and a bubbling bed
as a combustor/gasifier, was developed. Prior to operating a dual circulating fluidized bed reactor,

AC

the transition velocities of flow regime were determined at different temperatures using silica sand
particles as the bed material. The transition velocities included minimum fluidization velocity (Umf),
transition velocity to turbulent fluidization (Uc, Uk) and transport velocity (Utr). The minimum
fluidization velocity determined by measuring bed pressured drop with increasing gas velocity was
observed to decrease with increasing temperature. Both the transition velocities from bubbling to
turbulent fluidized bed behavior determined by measuring bed pressure drop fluctuation and
transport velocity from turbulent to fast fluidized bed behavior by emptying time method increased
with increasing temperature. Based on the experimental data, correlations are proposed to predict the
transition velocities at different temperatures. As the actual dual circulating fluidized bed reactor
operation occurs at relatively high temperature, the correlations obtained in this study is directly
applicable to the extra-heavy oil fractions upgrading process. Thus, the transition velocities and flow
regimes in a dual circulating fluidized bed reactor is determined.
Keywords: Extra-heavy oil fractions upgrading process; Dual circulating fluidized bed reactor;
Temperature; Minimum fluidization velocity (Umf); Transport velocity (Utr).
1

1.

Introduction

ACCEPTED MANUSCRIPT

A process for upgrading extra-heavy oil fractions to produce syngas and supply process heat,
which consists of a rapid thermal pyrolyzer (RTP) for extra-heavy oil and a gasifier/combustor for
the RTP, has been developed at the Korea Institute of Energy Research (KIER) [1-2]. The dual

circulating fluidized bed reactor, which consists of a riser and a bubbling fluidized bed, is used in

IP

this extra-heavy oil fractions upgrading process. This concept is widely used in dual fluidized bed
gasifiers (DFBG) [3-4], absorption enhanced reforming (AER), [5] chemical looping combustion

SC
R

(CLC) [6], and cracking gasification process for petroleum residue [7]. The application and recent
studies of dual fluidized bed reactor is well summarized in the latest review [8]. A schematic
diagram of the dual circulating fluidized bed reactor for the process of upgrading extra-heavy oil

NU

fractions is presented in Fig. 1.

This process consists of two fluidized bed reactors; one is an RTP, and the other one is an RTP

MA

residue gasifier/combustor. Unreacted carbon in the RTP residue from the pyrolyzer will be used as
a feedstock to the gasifier/combustor. The energy generated by gasification/combustion of the RTP
residue in the gasifier/combustor is transferred to the pyrolyzer to supply energy for the endothermic

pyrolysis reaction. The silica sand particles act as the heat carrying bed material that circulates

TE

between the pyrolyzer and RTP residue gasifier/combustor. The pyrolyzer produces synthetic crude
oil (SCO), while the valuable product gas consisting of hydrogen and carbon monoxide can also be

CE
P

obtained. Because the gasifier/combustor requires a longer reaction time, a bubbling fluidized bed
was designed for this process step, while a fast-fluidized bed was designed for the pyrolysis reaction.
To operate a dual circulating fluidized bed reactor with each bed running at a different flow regime,

AC

fundamental data are required, such as the minimum fluidization velocity, transition velocity on
turbulent fluidization, and transport velocity at high temperature.
Although most commercial fluidized bed reactors are operated at high temperature, most studies
have proposed empirical correlations at different operating conditions at room temperature to predict
the fluidization velocity. The measured velocities at higher temperatures have shown a remarkable
discrepancy from the predicted velocities based on correlations reported at room temperature. For
minimum fluidization velocity, Desai et al. [9] indicated that this difference came from the changed
physical properties of gas at the given temperature. Pattipati and Wen [10], Boetterill et al. [11]
measured the minimum fluidization velocity of sand at higher temperature, however correlation was
not proposed. Nakamura et al. [12] and Zheng et al. [13] proposed correlation based on Ergun Eq.
including the high temperature range. Wu and Baeyens [15] reported the minimum fluidization
velocity (Umf) of lime, limestone and sand at high temperature and concluded that the calculated
values from the correlation overestimated Umf at high temperature. Ryu et al. [16] measured the
minimum fluidization velocity of NiO/bentonite particle for the application of CLC. Subramani et al.

ACCEPTED MANUSCRIPT

[17] and Goo et al . [18] neglected the kinetic force term and simplified the correlation between Remf
and Ar in a simple equation. Jiliang et al. [19] investigated effect of particle size distribution on
minimum fluidization velocity. They found that the Umf of particles with wide PSD decreases the
mass fraction of coarse particles due to different inter-particle forces. Previous studies [10-20] on the
minimum fluidization velocity at high temperature are summarized in Table 1.

With increasing gas velocity, bubble size and frequency increase with increasing pressure

IP

fluctuation inside the reactor. The transition velocity to turbulent fluidization (Uc, Uk) is important to
determine onset of the turbulent flow regime. However, to the best of our knowledge, there has been

SC
R

no study on the effect of temperature on the turbulent flow regime.

The transport velocity (Utr) is a transition velocity between the turbulent and the fast fluidization
flow regimes. To operate at a dual circulating fluidized bed reactor having constant solid circulation

NU

rate, the gas velocity of the riser should be greater than Utr [4]. However, most studies on transport
velocity are examined at ambient temperature [21-24]. Only few studies [16,18,25] reported the

MA

transport velocity at high temperature upto 600C. Most correlations are in the form of exponential
function of Ar and Re. Ryu et al. [16] considered the column diameter (Dt) with respect to particle
mean diameter (dp) for small column diameter. Although it is known that the Utr increased with

increasing column diameter, this effect is negligible for the column diameter employed in this study

TE

(0.078m). The particle, experimental conditions and correlations of previous studies [16,18,21-25]
on transport velocity are summarized in Table 2.

CE
P

In the present study, the transition velocities, such as minimum fluidization velocity, transition
velocity to turbulent fluidization, and transport velocity at high temperature, were determined by
pressure drop measurement in a dual circulating fluidized bed. Prior to operating a dual circulating

AC

fluidized bed reactor, the classification of flow regime is crucial for determination of the operating
condition. Based on the experimental data, correlations have been proposed that can predict the
transition velocities at each flow regime for various temperature range.

2.

Experimental

Experiments were carried out in a dual circulating fluidized bed reactor as shown in Fig. 2. It is
composed of a fast fluidized bed riser (0.078 m I.D. 8.5 m high) and a bubbling fluidized bed
gasifier/combustor (0.2 m I.D. 2.1 m high). To control the solid circulation rate and prevent the gas
bypassing between riser and bubbling fluidized bed, upper and lower loop-seals (0.078 m I.D. 0.4
m high) were placed. Four different sizes (135, 210, 270, and 385 m) of silica sand particle with a
density of 2500 kg/m3 and sphericity (s) of 0.86 were used as the bed material. This material is
classified as Geldart group B, which is used in extra-heavy oil fractions upgrading process. The total
solid inventory inside the reactor was maintained as 40 kg. The fluidizing air was supplied from the

ACCEPTED MANUSCRIPT

compressor, and then passed through air filter and controlled by a regulator. The entrained particle
from the riser were separated in a cyclone and transferred to the upper loop-seal. The particle move
to the bubbling fluidized bed and then circulated back to the riser though a lower loop-seal. The
aerated air in the lower loop-seal controls the solid circulation rate inside the reactor.
Experimental variables and operational range are presented in Table 3. Online pressure transducers

(DPLH series, Sensys, Korea) were connected to pressure taps at different heights of the riser and

IP

bubbling fluidized bed to measure absolute and differential pressures inside the reactor. The
minimum fluidization velocities (Umf) and the transition velocity to turbulent fluidization (Uc, Uk) of

SC
R

different particles were measured at different temperatures (25 800C) in the bubbling fluidized
bed. The minimum fluidization velocity is determined by measuring bed pressured drop, whereas the
transition velocities to turbulent fluidized bed behavior is determined by measuring bed pressure

NU

drop fluctuation with increasing gas velocity. The bed pressure drop at each gas velocity is
monitored around 3 min and stored in a data acquisition system (midi Logger GL820, Graphtec,

MA

Japan). All experimental runs were conducted more than three times and averaged to compensate
experimental reproducibility.

The transport velocity was determined by the emptying time method. The procedure to determine

Utr is given below. First, the particles were circulated until reactor reached steady state at a given

TE

riser velocity and temperature. When it reached a steady-state, the bed height in the bubbling
fluidized bed became constant. When the pressure drop in the riser became 3 kPa (approximately 0.6

CE
P

m height), air being fed into the lower loop-seal was stopped, and the time required entrain all of the
particles from the riser was measured.
As all experiments were carried out in a high temperature range, the air flow rate at each

AC

temperature was calibrated. The properties of the fluidizing gas, such as density and viscosity, are
calculated from the following relations [26].

g 1.2

293
T

T 1.504
g 1.46 10
T 120
6

3.
3.1.

(1)
(2)

Results and Discussion

Minimum fluidization velocity (Umf)

The minimum fluidization velocities (Umf) from the pressure drop in the bed of sand particle (dp =
270 m) at different temperatures are shown in Fig. 3. As the gas velocity increased from the fixed
bed, the pressure drop increased and later became constant at a certain gas velocity [9-20]. The

ACCEPTED MANUSCRIPT

minimum fluidization velocity was determined from the gas velocity by extrapolating the pressure
drop across the fixed bed and that of maximum theoretical pressure drop.
The Umf of silica sand particles (dp = 135, 210, 270, and 385 m) at various temperatures are shown
in Fig. 4. As seen, Umf increase with increasing particle size and decreasing temperature. For the
largest particle (dp = 385 m), the decrease of Umf is larger than that of the other particle sizes. In the

high temperature range, it is confirmed that the change of Umf with increasing temperature is

IP

marginal and becomes constant at high temperatures.

mf 1

SC
R

The bed porosity at the minimum fluidization velocity (mf) is determined by the following Eq.:

Pb
L( s g )g

(3)

As is shown in Fig. 5, smaller particles have higher porosity than large particles. The bed porosity

NU

increases with increasing temperature. It can be noticed that the maximum pressure drop tends to
decrease with increasing temperature, as seen in Fig. 3. Therefore, the bed porosity increases with

MA

increasing temperature since the bed pressure drop (Pb) decreased with increasing temperature.
Subramani et al. [17] also observed a similar trend with ilmenite and sand particles. It can be
concluded that porosity increases with increasing temperature but decreases with particle size.

Most of the correlations to predict Umf are based on the Ergun Eq., as

TE

1.75C1 Re 2mf 150C 2 Re mf Ar ,

C1

mf3 s

1 mf

CE
P

where

C2

mf3 s2

(4)

(5)

where Ar is the Archimedes number, and Remf is the Reynolds number at the minimum fluidization

AC

condition. C1 and C2 are constant determined by voidage and shape factors. These value are
approximated as 14 (C1) and 11 (C2 ) by Wen and Yu [27] based on the empirical data obtained with
different properties of gas and solid particles. However, at lower Reynolds numbers (Remf < 5), the
first term of the left side (kinetic force) of Eq. (4) can be neglected [8] so that Eq. (4) can be
rewritten as [17-18]:

150C2 Re mf Ar

(6)

Therefore, the minimum fluidization velocity can be represented as

U mf

d p2 (s g )g
150C2

(7)

In Eq. (7), g is notably small compared with s; therefore, the minimum fluidization velocity of
particles at different temperature is a function of the gas viscosity. Therefore, the minimum
fluidization velocity decreases with increasing temperature due to increasing gas viscosity.

ACCEPTED MANUSCRIPT

Based on the experimental data of this and previous studies [10-11,16], the relationship between
Remf and Ar is plotted in Fig. 6. The C2 value in Eq. (6) was determined to be 11.58, which
corresponds well to the approximation of Wen and Yu [27] as

Remf

Ar
1737

(8)

d p2 ( s g )g

IP

U mf

and the resulting correlation can be written as

1737

(9)

SC
R

in the range of 0.01 < Remf < 5.2 and 10 < Ar < 9287 with a correlation coefficient of 0.996.

NU

3.2. Transition velocity to turbulent fluidization (Uc, Uk)

With increasing gas velocity in the bubbling fluidized bed, the pressure drop fluctuation increased
as bubble size and frequency increased [28]. The fluctuation reaches its maximum value at a gas

MA

velocity of Uc and later tends to decrease with a further increase in gas velocity. This decrease is due
to the increase of bed uniformity with the formation of small bubbles, which is due to the collapse of
large bubbles and slugs. The gas velocity subsequently reaches Uk, which exhibits stabilized pressure

fluctuation with further increasing gas velocity. However, the existence of Uk is controversial

TE

because it depends on the physical properties of both particle and column diameters. The transition
velocity into the turbulent fluidization is therefore defined as either Uc or Uk.

CE
P

The calculation of Uc for silica sand particles (dp = 270 m) at different temperature in the riser is
presented in Fig. 7. As seen, the pressure fluctuation reaches a maximum and later decreases with
increasing gas velocity. The gas velocity at the maximum pressure fluctuation can be defined as Uc.

AC

However, the velocity with a constant pressure fluctuation (Uk) is not observed in this study. This
finding is similar to results observed with coal (dp = 730 m, 1030 m) [29] and Glass beads (dp =
362 m) [30] and is different from Yerusalmi and Cankurts results [31-32], which indicated the
distinct existence of Uk.
Fig. 8 shows Uc for silica sand particles (dp = 270 m) from the experimental data, as well as
correlations from previous studies [30, 33-34] with a variation of temperature. As seen, Uc increases
linearly with increasing temperature. The value of Uc is 1.0 m/s at room temperature and later
increases from 1.1 to 2.0 m/s with increasing temperature from 200C to 800C. This result is
opposite to the minimum fluidization velocity, which decreases with increasing temperature. The
increasing trend of Uc with the variation of temperature can be explained by the change of the
particle entrainment rate, which is related to the transport velocity. It has been shown that the Uc is a
function of the physical properties of the particles and the column diameter; however, the effect of
temperature on Uc is rarely reported. In Fig. 8, the experimental data are compared with Uc that has
been calculated from the correlations [30, 33-34] at room temperature. As shown in Fig. 8, the

ACCEPTED MANUSCRIPT

calculated value from Cai et al.s correlation [34] is in good agreement with the experimental data.
The correlation of Bi and Grace [33] overestimated Uc value. Comprising Uc values obtained in the
present and previous studies [30, 33-34], the correlations have been correlated, as presented in Fig. 9,
as

Rec 0. 973 Ar 0.36

(10)

This Eq. covers the range of variables 1.1 < Rec < 32 and 0.55 < Ar < 8613 with a correlation

IP

coefficient of 0.96.

SC
R

Fig. 10 shows the bed voidage as a function of gas velocity at room temperature and at elevated
temperature (200C and 600C). With increasing the gas velocity beyond the minimum fluidization
velocity, the bed voidage remains constant and later suddenly rises at a given gas velocity (Uc). This

NU

increase is due to the increase of turbulence in the bed, resulting in bubble collapse and particle
entrainment. As seen, Uc can be measured by bed voidage with increasing gas velocity. The
increasing trend of Uc with increasing temperature is also observed. This value is equivalent to the

MA

Uc value determined by pressure fluctuation, as shown in Fig. 7.

3.3. Transport velocity (Utr)

TE

By increasing the gas velocity beyond the turbulent flow regime, the fast fluidization regime is
reached. The transport velocity (Utr) is defined at the velocity transporting all particles that were

CE
P

injected into the bed. The experimental methods for measuring transport velocity can be classified
into two ways, either using the solid concentration or the particle entrainment velocity [25]. With the
solid concentration, Yerushalmi and Cankurt [32] observed a pressure drop inside the bed with a
variation of solid circulation rate (Gs) and gas velocity (U) known as P/L Gs U diagram. The

AC

flooding point method [32] measured the suspension density with a solid circulation rate and gas
velocity. Using the particle entrainment velocity, Schnitzlein and Weinstein [35] extrapolated Utr
linearly from the maximum solid circulation rate at each gas velocity. The most common method to
determine Utr is the emptying time method, which is related to the kink point between the lines
from emptying times for all the particles transported at lower gas velocities and higher gas velocities
[18, 24-25]. The transport velocity (Utr) determination with different methods, such as the emptying
time method, the P/L Gs U diagram, and the flooding point method for silica sand particles (dp
= 270 m) at room temperature, is shown in Fig. 11. The obtained Utr values are 2.55, 2.5 2.7, 2.6
m/s for each methods, respectively. As these values are similar, the following Utr are determined
from the emptying time methods as the temperature varies.
The Utr values at different particle sizes and temperatures are shown in Fig. 12. As seen, Utr also
increases with increasing particle size. With increasing particle size, the Utr increases with
decreasing particle entrainment velocity. As seen, the rate of increase in Utr is remarkable for the

ACCEPTED MANUSCRIPT

largest particle (dp = 385 m), which would be due to the large decrease of minimum fluidization
velocity of the largest particle with increasing temperature.
Bae et al. [25] reported that higher gas velocity is needed for transition into fast fluidization
because the terminal gas velocity (Ut) increases and particle entrainment velocity decreases with
increasing temperature. A fluidized solid particle in fluid is influenced by the drag force, buoyancy
4 gd p ( p g )
3 g C D

]1/ 2

IP

Ut [

force, and gravity; therefore, the Ut of a particle is determined from the following Eq. (11).
(11)

SC
R

where CD is the drag coefficient to be determined experimentally and can be expressed according to
the flow region.

24
Re p

for Rep < 5.76 Stokes' region

CD

10
Re1p/ 2

for 5.76 < Rep < 541 Intermediate region

MA

NU

CD

C D 0.43

for 541 < Rep < 200,000 Newton's region

(12)
(13)
(14)

From the above Eqs, it can be noted that the drag coefficient is a function of Rep except Newtons

TE

region. The tendency of the terminal velocity to change can be aggregated to changes in CDg and
can be represented at each flow region according to Rep as

CD g

CD g g

(15)

for 5.76 < Rep < 541 Intermediate region

(16)

for 541 < Rep < 200,000 Newton's region

(17)

AC

C D g ( g ) 0.5

CE
P

for Rep < 5.76 Stokes' region

It is known that the density of air decreases but the viscosity of air increases with increasing gas
temperature. At a high gas velocity range (Newtons region), CDg is proportional to the density of
gas as shown in Eq. (14), which decreases with increasing temperature. Therefore, the particle
terminal velocity in Eq. (11) increases and particle entrainment velocity decreases, resulting in
increasing Utr at higher temperature.
Based on Utr values obtained in this and previous studies [16, 25] of different particle size and
temperatures have been correlated as shown in Fig. 13 as

Retr 0. 611Ar 0.586

(18)

This Eq. covers the variable range 2 < Retr < 79 and 10 < Ar < 4692 with a correlation coefficient
of 0.985. Compared with other correlations as shown in Table 2, the coefficient (0.611) is quite
small. The proposed correlation is applicable for ordinary fluidized bed reactors including dual
circulating fluidized bed reactor.

4.

ACCEPTED MANUSCRIPT

Conclusions

The fundamental data of the transition velocities, such as minimum fluidization velocity, turbulent
fluidization velocity and transport velocity in the dual circulating fluidized bed reactor, were
determined for different particle sizes as a function of temperature. With increasing temperature, the

minimum fluidization velocity determined by measuring bed pressured drop with increasing gas

IP

velocity decreases due to the increase of gas viscosity. Conversely, the transition velocities from
bubbling behavior to turbulent fluidized bed behavior by measuring bed pressure drop fluctuation,

SC
R

and from turbulent to fast fluidized bed conditions from emptying time method, increases with
increasing temperature. In this work, a correlation in terms of Reynolds and Archimedes numbers at
wide temperature range of room temperature to 800oC is proposed. The correlation obtained in this

NU

study can be used to determine the operating condition in a dual circulating fluidized bed reactor.

MA

Acknowledgment

This work was conducted under the framework of Research and Development Program of the

Korea Institute of Energy Research (KIER) (B4-2433-02, B4-2434-05). Additionally, this work was

TE

supported by the project Development of the design technology of a Korean 300 MW class IGCC
demonstration plant, funded by the Korea Institute of Energy Technology Evaluation and Planning

CE
P

(KETEP), affiliated with the Ministry of Knowledge Economy of the Korean Government

AC

(No.2011951010001B).

References

ACCEPTED MANUSCRIPT

[1] M.W. Seo, Y.T. Guahk, H.W. Ra, S.J. Yoon, N.S. Rho, J.G. Lee, S.D. Kim, Hydrodynamic
characteristics of 1 BPD extra-heavy oil RTP process. Proc. KIChE Autumn Meeting, Busan, 2012.

[2] M.W. Seo, Y.T. Guahk, N.S. Rho, S.J. Yoon, H.W. Ra, G.H. Koo, Y.K. Kim, J.H. Kim, J.G.
Lee, S.D. Kim, Gasification characteristics of Rapid Thermal Pyrolyzer residue in a fluidized bed
reactor. Energy Fuels (2014) (DOI: 10.1021/ef500258v)

IP

[3] C. Pfeifer, R. Rauch, H. Hofbauer, In-Bed Catalytic tar reduction in a dual fluidized bed biomass
steam gasifier, Ind. Eng. Chem. Res 43 (2004) 1634-1640.

SC
R

[4] M.W. Seo, T.C.B. Nguyen, Y.I. Lim, S.D. Kim, S.W. Park, B.H. Song, Y.J. Kim, Solid
circulation and pressure drop characteristics of a dual circulating fluidized bed reactor: experiments
and CFD simulation, Chem. Eng. J 168 (2011) 803-811.

NU

[5] S. Koppatz, C. Pfeifer, R. Rauch, H. Hofbauer, T. Marquard-Moellenstedt, M. Specht, H2 rich

product gas by steam gasification of biomass with in situ CO2 absorption in a dual fluidized bed
system of 8 MW fuel input, Fuel Proc. Tech 90 (2009) 914-921.

MA

[6] T. Prll, P. Kolbitsch, J. Bolhr-Nordenkampf, H. Hofbauer, A novel dual circulating fluidized

bed system for chemical looping processes, AlChE J 55 (2009) 3255-3266.

[7] Y. Zhang, D. Yu, W. Li, S. Gao, G. Xu, H. Zhou, J. Chen, Fundamental study of cracking
gasification process for comprehensive utilization of vacuum residue, Appl. Energy 112 (2013)
1318-1325.

TE

[8] J. Zhang, R. Xu, G. Zhang, C. Yao, Y. Zhang, Y. Wang, G. Xu, Recent studies on chemical
engineering fundamentals for fuel pyrolysis and gasification in dual fluidized bed, Ind. Eng. Chem.
Res 52 (2013) 6283-6302.

CE
P

[9] A. Desai, H. Kikukawa, A.H. Pulsifer, The effect of temperature upon minimum fluidization
velocity, Powder Technol 16 (1977) 143-144.

AC

[10] R.R. Pattipati, C.Y. Wen, Minimum fluidization velocity at high temperatures, Ind. Eng. Chem.
Process Des. Dev 20 (1981) 705-708.
[11] J.S.M. Botterill, Y. Teoman, K.R. Yuregir, The effect of operating temperature on the velocity
of minimum fluidization, bed voidage and general behavior, Powder Technol 31 (1982) 101-110.
[12] M. Nakamura, Y. Hamada, S. Toyama, An experimental investigation of minimum fluidization
velocity at elevated temperatures and pressures, Can. J. Chem. Eng 63 (1985) 8-13.
[13] Z.X. Zheng, R. Yamazaki, G. Jimbo, Minimum fluidizing velocity of large particles at elevated
temperatures, Kagaku Kogaku Rombun 11 (1985) 115.
[14] A. Mathur, S.C. Saxena, Z.F. Zhang, Hydrodynamic characteristics of gas-fluidized beds over a
broad temperature range, Powder Technol 47 (1986) 247-256.
[15] S.Y. Wu, J. Baeyens, Effect of operating temperature on minimum fluidization velocity.
Powder Technol 67 (1991) 217-220.
[16] H.J. Ryu, N.Y. Lim, D.H. Bae, G.T. Jin, Minimum fluidization velocity and transition velocity
to fast fluidization of oxygen carrier particle for chemical-looping combustor, HWAHAK
KONGHAK. 41 (2003) 624-631.
[17] H. Subramani, M.B.M. Balaiyya, L.R. Miranda, Minimum fluidization velocity at elevated
temperatures for Geldart's group-B powders, Exp. Therm. Fluid. Sci. 32 (2007) 166-173.

10

ACCEPTED MANUSCRIPT

[18] J.H. Goo, M.W. Seo, S.D. Kim, B.H. Song, Effect of temperature and particle size on minimum
fluidization and transport velocities in a dual fluidized bed, in: Proceedings of the 20th International
Conference, Xian, 2009, 305-309.
[19] M. Jiliang, C. Xiaoping, L. Daoyin, Minimum fluidization velocity of particles with wide size
distribution at high temperatures, Powder Technol 235 (2013) 271-278.

[20] Z.D. Chen, X.P. Chen, Y. Wu, Study on minimum fluidization velocity at elevated temperature,
Proceedings of the CSEE 30 (2010) 21-25.

IP

[21] J.F. Perales, T. Coll, M.F. Llop, L. Puigjaner, J. Arnaldos, J. Cassal, Circulating Fluidized Bed
Technology III, Pergamon Press, New York, 1990,73-78.

SC
R

[22] G.S. Lee, S.D. Kim, Bed expansion characteristics and transition velocity in turbulent fluidized
beds, Powder Technol 62 (1990) 207-215.

NU

[23] M. Tsukada, D. Nakanishi, M. Horio, The effect of pressure on the phase transition from
bubbling to turbulent fluidization, Int. J. Multiphase Flow 19 (1993) 27-34.
[24] J. Adnez, L.F. de Diego, P. Gayan, Transport velocities of coal and sand particles, Powder
Technol 77 (1993) 61-68.

MA

[25] D.H. Bae, H.J. Ryu, D.W. Shun, G.T. Jin, D.K. Lee, Effect of temperature on transition velocity
from turbulent fluidization to fast fluidization in a gas fluidized bed, HWAHAK KONGHAK 39
(2001) 456-464.

TE

[26] K. Svoboda, M. Hartman, Influence of temperature on incipient fluidization of limestone, lime,

coal ash, and corundum, Ind. Eng. Chem. Process Des. Dev 20 (1981) 319-326.
[27] C.Y. Wen, Y.H. Yu, A generalized method for predicting the minimum fluidization velocity,
Chem. Eng. Prog. Symp. Ser 62 (1966) 100-111.

CE
P

[28] D. Kunii, O. Levenspiel, Fluidization Engineering, 2nd Ed. Butterworth-Heinmann, MA,

U.S.A. (1991).

AC

[29] G.Y. Han, G.S. Lee, S.D. Kim, Hydrodynamic characteristics of a circulating fluidized bed,
Korean J. Chem. Eng 2 (1985) 141-147.
[30] G.S. Lee, S.D. Kim, Pressure fluctuations in turbulent fluidized beds. J. Chem. Eng. Jpn 21
(1988) 515-521.
[31] J. Yerushalmi, N.T. Cankurt, High-velocity fluid beds, Chem. Tech. 8 (1978) 564-572.
[32] J. Yerushalmi,N.T. Cankurt, Further studies of the regimes of fluidization, Powder. Technol 24
(1979) 187-205.
[33] H.T. Bi, J.R. Grace, Flow regime diagrams for gas-solid fluidization and upward transport, Int.
J. Multiphase Flow 21 (1995) 1229-1236.
[34] P. Cai, S.P. Chen, Y. Jin, Z.Q. Yu, Z.W. Wang, Effect of operating temperature and pressure on
the transition from bubbling to turbulent fluidization, AIChE Symposium series 85 (1989) 37-43.
[35] M.G. Schnitzlein, H. Weistein, Flow characterization in high-velocity fluidized beds using
fluctuations, Chem. Eng. Sci. 43 (1988) 2605-2614.

11

Figure Captions

ACCEPTED MANUSCRIPT

Figure 1 Schematic diagram of the extra-heavy oil fractions upgrading process.

Figure 2 Schematic diagram of the dual circulating fluidized bed reactor: 1. Riser, 2. Cyclone, 3.
Downcomer, 4. Upper loop-seal, 5. Bubbling fluidized bed, 6. Lower loop-seal, 7. Electric heater.
Figure 3 Determination of minimum fluidization velocity at various temperatures (dp= 270 m).

Figure 4 Effect of temperature on minimum fluidization velocity at various particle sizes.

IP

Figure 5 Effect of temperatures on bed voidage at minimum fluidization velocity with various

SC
R

particle sizes.

Figure 6 The minimum fluidization velocity correlation between Remf and Ar compared with other
studies (: Pattipati and Wen [10], : Botterill et al. [11], : Ryu et al. [16], : present study).
Figure 7 Determination of Uc at various temperatures (dp= 270 m).

NU

Figure 8 Effect of temperature on Uc and comparison with other correlations (: Bi and Grace [33],
: Lee and Kim [30], ---: Cai et al. [34], : present study, dp= 270 m).

MA

Figure 9 Correlation of Uc in terms of Rec and Ar compared with previous studies ( : Bi and Grace
[33], : Lee and Kim [30], : Cai et al. [34], : present study, dp= 270 m).
temperature (200oC and 600C)

Figure 10 The bed voidage with a variation of gas velocity at room temperature and elevated

TE

Figure 11 Determination of Utr with different methods; (a) emptying time method, (b) P/L-Gs-U
diagram method, (c) flooding point

CE
P

Figure 12 Effect of temperature on Utr at various particle sizes

Figure 13 Correlation of Utr in terms of Ar and Retr [ : Sand (dp= 256 m) from Bae et al. [25], :

AC

NiO/bentonite (dp= 181 m) from Ryu et al. [16], : present study).]

12

CE
P
AC

Fig. 1

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

13

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

14

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

15

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

16

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

17

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

18

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

19

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

20

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

21

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

22

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

23

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

24

AC

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

25

ACCEPTED MANUSCRIPT
Table 1. Summary of minimum fluidization velocity (Umf) at high temperature
Materials

dp (m)

T (oC)

Correlation

Pattipati and Wen [10]

Sand

240

20-750

Botterill et al. [11]

Sand

380, 460

20-815

Nakamura et al. [12]

Glass beads

200-4000

7-527

Re mf (33.9532 0.0465Ar)0.5 33.953

IP
SC
R

Glass beads

Zheng et al. [13]

20-700

Sand

559-3788

Wu and Bayens [15]

Lime,
Limestone,
Sand

134-939

25-1000

MA

Mathur et al. [14]

20-400

Re mf 7.33 105 10(8.24 log10 Ar 8.81)

1/ 2

U mf

2.997 10 3 d 1p.636 ( p g )1.128 g

181

25-1000

Subramani et al. [17]

Ilmenite, Sand,
Limestone,
Quartz
Magnetite

128-200

25-700

Re mf

Ar
1502

Goo et al. [18]

Sand

210-380

25-800

Re mf

Ar
1667

Quartz sand,
Bottom ash
particles

500-1290

30-600

Sand

134-939

CE
P

AC

Jiliang et al. [19]

Chen et al. [20]

TE

NiO/bentonite

Re mf (18.752 0.00315Ar)0.5 18. 75

NU

Sand

Ryu et al. [16]

Authors

20-400

g0.0247 0.446

i 1

U mf 0. 28 xi di0.599(

)0.533 /0.066

Re mf 0. 01036Ar 0.7107

26

ACCEPTED MANUSCRIPT
Table 2. Summary of studies on transport velocity (Utr)
Materials

dp (m)

T (oC)

Correlation

Perales et al. [21]

FCC, sand

80
120-1200

A.C

Re tr 1.41 Ar 0.483

Lee and Kim [22]

Cement, raw meal

23.6

A.C.

Re tr 2.916 Ar 0.354

Tsukada et al. [23]

FCC

46

A.C.

Re tr 1.806 Ar 0.458

Adnez et al. [24]

Sand, coal

170-894
316-894

A.C

Bae et al. [25]

Sand

256

Ryu et al. [16]

NiO/bentonite

181

Goo et al. [18]

Sand

210-380

IP

SC
R
25-600
25-600
25-600

Retr 2.078 Ar 0.463

Re tr 0.0428 Ar 0.5866 (

Dt

dp

) 0.5208

Retr 2. 001 Ar 0.405

AC

CE
P

TE

MA

NU

Authors

27

ACCEPTED MANUSCRIPT
Table 3. Experimental variables and operational range
Operational range

Mean particle size (dp, m)

135, 210, 270, 385

Temperature (oC)

25 - 800

Gas velocity to the riser (Ug,r m/s)

1.7 5.6

Experimental variables

IP

0 0.075

AC

CE
P

TE

MA

NU

SC
R

Gas velocity to the bubbling fluidized bed (Ug,b, m/s)

28

AC

Graphical abstract

CE
P

TE

MA

NU

SC
R

IP

ACCEPTED MANUSCRIPT

29

ACCEPTED MANUSCRIPT
Highlights
The flow regime velocities were determined in a dual circulating fluidized bed.
The minimum fluidization velocity decreases with increasing temperature.
The transition and transport velocity increase with increasing temperature.

AC

CE
P

TE

MA

NU

SC
R

IP

We propose correlations for the flow regime velocities at different temperatures.

30