Advanced Powder Technology 19 (2008) 7382

www.brill.nl/apt

Original paper

An Analysis on Modeling of Fluidized Bed Drying of

Granular Material

C. Srinivasakannan a and N. Balasubramanian b

a
Monash University, Sunway Campus, 46150, Petaling Jaya, Selangor, Malaysia
b
AC College of Technology, Anna UniversityChennai, Chennai 600 025, India
Received 1 March 2007; accepted 1 June 2007

Abstract
The drying kinetics corresponding to the falling rate period is commonly modeled using Ficks diffusion
equation in uidized beds. The appropriateness of the use of Ficks diffusion equation to represent the
falling rate drying kinetics is the subject of the present paper. The estimated diffusion coefcient is normally
the single kinetic parameter to assess the kinetics of the drying and is often used to compare the drying
kinetics in different forms of drying. The dependence of the diffusion coefcient on the temperature and the
concentration is well understood from the basic concepts of mass transfer; however, the diffusion coefcient
estimated using Ficks diffusion model is found to vary with additional variables such as the solids holdup.
The dependence of the diffusion coefcient on additional parameters is purely empirical and has not been
justied in the earlier studies. Towards this, experiments are conducted in uidization columns of varying
diameter in order not to vary the quality of uidization while varying the solids holdup on a larger scale.
The estimated diffusion coefcient is found to vary by orders of magnitude with the variation in the column
diameter (solids holdup), necessitating the caution one needs to observe while comparing the kinetics of
uidized beds based on the diffusion coefcient.
Koninklijke Brill NV, Leiden and Society of Powder Technology, Japan, 2008

Keywords
Fluidized bed, diffusion coefcient, drying

Nomenclature

Bim Biot number, KRs /Deff

C moisture content of ragi grain at any time (kg of moisture/kg of dry solid)
Ci initial moisture content of ragi grain (kg of moisture/kg of dry solid)

1To whom correspondence should be addressed. E-mail: chandrasekar.srinivasakannan@eng.monash.

edu.my

Koninklijke Brill NV, Leiden and Society of Powder Technology, Japan, 2008 DOI:10.1163/156855208X291774
74 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382

Ce equilibrium moisture content of ragi grain (kg of moisture/kg of solid)

Dc column diameter (m)
dp particle diameter (m)
Deff effective diffusion coefcient (m2 /s)
D molecular diffusivity of moisture in air (m2 /s)
K mass transfer coefcient across the particle surface (m/s)
r radial coordinate (m)
Re Reynolds number, dp U/
Rs particle radius (m)
RMSE root mean square error
Sh Sherwood number K dp /D
T temperature of heating medium ( C)
t time (s)
U supercial velocity of heating medium (m/s)
Ws solids holdup (kg)
Greek
density of air (kg/m3 )
viscosity of air (kg/ms)

1. Introduction
Fluidized beds are widely used for the drying of granular solids such as grains,
fertilizers, chemicals, pharmaceuticals and minerals, further to specialized applica-
tions such as drying of solutions, pastes and liquids sprayed on to the uidized bed
containing inert materials. Fluidized beds offer advantages such as: (i) the uidity
of the bed facilitating easy handling and transport of solids, (ii) high heat and and
mass transfer rates, (iii) perfect mixing of material in the bed, and (iv) the possibility
of applying other sources of energy such as immersed heating coils, etc.
The mode of operation of a uidized bed could either be batch or continuous.
The batch dryer nds application for small-scale production, while the continuous
systems are used for large-scale production. Prior knowledge on drying kinetics is
essential for the design and sizing of uidized beds. The drying rates in uidized
beds are strongly inuenced by the characteristics of the material and the conditions
of uidization.
 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382 75

Drying of solids is generally understood to follow two distinct drying zones

known as the constant rate period and falling rate period demarcated with critical
moisture content. The critical moisture content is reported to vary with operating
parameters and with the type of drying equipment. The constant rate period is un-
derstood to have a maximum drying rate, which remains constant until the critical
moisture content with the resistance for moisture transfer in the gas phase. The
rate of diffusion of moisture to the surface of solids becomes the limiting factor
for moisture transfer as far as the falling rate period is concerned. The extent of
drying zones is decided based on the type of material, with materials like sand, ion
exchange resin, glass beads, etc., reported to have a larger duration of constant rate
period and short linear falling rate period, while brous grains such as mustard,
pepper, ragi, poppy seeds, etc., are reported to have a very short duration constant
rate period and a longer curvilinear falling rate period [1].
The complex hydrodynamics and process calculations are material and dryer spe-
cic, and hence numerous mathematical models have been developed to estimate
the drying kinetics. These range from analytical models solved with a variety of
simplied assumptions to purely empirical models, often built by regression of ex-
perimental data.
The drying rate in the constant rate period in uidized bed drying is modeled
using (i) a simple empirical correlation relating drying rate to the inuencing pa-
rameters or utilizing the heat/mass transfer coefcient between solids and gas in
uidized bed [26] or (ii) either two-phase or three-phase bubbling bed models,
which assume the bed to be made of a bubble phase, emulsion phase or a dense
phase with the exchange of mass and energy between these phases [713].
Similarly, drying kinetics in the falling rate period is modeled using simple
models considering the rate of drying to linearly reduce from the critical moisture
content to the equilibrium moisture content [14], which form characteristics of non-
porous material as well as material with very high porosity. Materials such as food
grains that offer high resistance for moisture diffusion exhibit a curvy falling rate
period that could be modeled using simple exponential time decay models (Newton
model, Page model, Henderson and Pabis model, two-term exponential model and
approximate diffusion model) which are purely empirical in nature [1517]. The
model that best ts the experimental data is chosen as the appropriate model. On
the other hand, complex models that serve the purpose of improving the fundamen-
tal understanding of drying kinetics are widely based on Ficks diffusion equation
[1823]. The drying rate is estimated by solving the partial differential equation
either using analytical solutions or using numerical techniques depending on the
boundary conditions selected to represent the experimental conditions. The evalu-
ated diffusion coefcients could well serve as a single kinetic parameter indicating
the rate of drying, which could well be extrapolated beyond the experimental con-
ditions within acceptable errors.
The use of Ficks diffusion equation to represent the drying kinetics in a uidized
bed assumes all the solids in the uidized bed to be exposed to uniform conditions.
76 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382

The moisture in the solids is assumed to be uniformly distributed initially and the
model predicts the varying moisture concentration prole in the solids with respect
to time. The use of Ficks diffusion equation is also used as an appropriate mod-
eling tool in representing the drying kinetics in other kinds of dryers such as tray
dryers, wherein the shape of the solid slab is taken into consideration. Irrespective
of the kind of dryer, the diffusion coefcient estimated serves as an indicator for
the estimation of the drying kinetics and often comparison of the performances of
different dryers is based on the estimated diffusion coefcient.
It is the basic objective of this paper to test the reliability of the estimated kinetic
parameter (diffusion coefcient) to serve the good purpose of predicting the drying
kinetics of granular material in particular to uidized beds. An inconsistency in the
estimated diffusion coefcient from the experimental data is suspected, since the
model only considers single particles in the bed, ignoring in the quantum of ma-
terial in the bed. The dependence of the diffusion coefcient on the temperature
and the concentration is well documented and understood from the basic concepts
of mass transfer. However, the reported dependency on solids holdup could not be
reasoned with any justication. This is basically due to the inability of the model
to consider the geometry of the bed, resulting in the estimated diffusion coefcient
showing a dependency on solids holdup. Towards this, experiments are conducted
in columns of 0.148 and 0.245 m diameter, with the food grain ragi (Eleusine coro-
cana), a granular material, by varying the solid holdup in a broad range from 0.3
to 2.6 kg. The variation in the column diameter is to ensure that the increase in the
solids holdup does not affect the quality of uidization.

2. Experimental
Drying experiments were conducted using uidized columns of 0.148 and 0.245 m
internal diameter with a height of 1.2 m. The schematic diagram of the experimental
setup is shown in Fig. 1. The experimental setup consists of an air supply line from
an air bower, control valve (1), orice plate with monometer (2), electric heater
with temperature control (3), thermocouple (4), plenum chamber (5), perforated
air distributor plate (6), uidization column (7) and cyclone separator (8). The gas
distributor was 2 mm thick with 2-mm perforations having 13% free area. A ne
wire mesh was spot welded over the distributor plate to arrest the ow of solids
from the uidized bed to the air chamber. Air from the blower was heated and
fed to the uidization column through the air chamber (plenum). The electrical
heater consisted of multiple heating elements each of 2 kW rating. A temperature
controller, provided to the air chamber, facilitated control the of the air temperature
within 3 C of the set temperature for the entire operating range of 30110 C.
Airow was measured using a calibrated orice meter.
Table 1 shows the physical characters of ragi as well as the experimental condi-
tions covered in the present study. A good uidization behavior in terms of perfect
mixing of the bed material was observed visibly. This was substantiated with low
 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382 77

Figure 1. Schematic diagram of the experimental setup: 1, control valve for air from blower; 2, ori-
ce plate with monometer; 3, electric heater with temperature control; 4, thermocouple; 5, plenum
chamber; 6, perforated air distributor plate; 7, uidization column; 8, cyclone separator.

Table 1.
Characteristics of the material and the details of the experimental conditions

Name of material ragi (E. corocana)

Shape of material spherical
Size, dp 103 (m) 1.48
Particle density (kg/m3 ) 1200
Diameter of uidized bed (m) 0.148, 0.245
Minimum uidization velocity, Umf (m/s) 0.47
Terminal velocity, Ut (m/s) 6.9
Temperature of uidizing air ( C) 80
Fluidizing air velocity (m/s) 1.2
Solid holdup (kg) 0.302.6

uctuation in the bed pressure drop, which is an indication for smooth uidization
without the formation slugs. The minimum uidization velocity was not found to
vary with the temperature within the range of temperatures covered in the present
study.
Air at the desired temperature and ow rate was allowed to ow through the
uidization column. A known quantity of ragi with a known initial moisture con-
tent was introduced into the column after ensuring the steady temperature and the
air ow rate. As uidization continued, ragi samples of approximately 1 g were
scooped out of the bed each time, at regular intervals of time, for moisture con-
78 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382

tent estimation. The samples were weighed and kept in an air oven at 105 C until
no further weight change of the samples was observed. The ratio of difference in
weight of the sample to the nal weight of the sample after drying at 105 C is the
moisture content of the sample (dry basis). The experimental data was checked for
reproducibility and was found to deviate within 2%. The equilibrium moisture
content was estimated by keeping the samples in an air oven at the desired tem-
peratures until no further weight change. The equilibrium moisture content is the
ratio of the difference in weight of samples kept at the desired temperature until no
further weight change to the bone dry weight of the samples, to the bone dry weight
of the samples (dry basis).

3. Results and Discussion

The acceptability of the effective diffusivity coefcient estimated using Ficks dif-
fusion equation to serve as a single parameter to assess the drying kinetics in a
uidized bed is the subject of analysis in the present paper. The analysis is per-
formed based on the experimental data generated in the present study.
Ficks diffusion equation assumes that the moisture diffuses from inside the par-
ticle to the surface of the particle and evaporates at the surface, and that all the
particles are uniform in size and spherical in shape. The uidized beds are perfectly
mixed beds and the solids at any point in the beds are exposed to same drying condi-
tions. The general form of Ficks diffusion equation to estimate the drying kinetics
for spherical particles is as given as:
 2 
C C 2 C
= Deff + . (1)
t r 2 r r
Different boundary conditions are utilized, which are varied and are selected to
reect the experimental conditions. One of the common boundary conditions, which
also accounts for the external mass transfer resistance, is:
at t = 0; 0 < r < Rs ; C = Ci ,
at t > 0; r = 0; C/r = 0,
at t > 0; r = Rs ; Deff (C/r) = K(Csj Cbe ),
where Csj is the moisture concentration just within the sphere and Cbe is the con-
centration required to maintain equilibrium with the surrounding atmosphere. The
analytical solution for equation (1) for the above boundary conditions was provided
by Crank [24]:

C Ce  6Bi2m exp(n2 Deff t/Rs2 )
= , (2)
Ci Ce n2 (n2 + Bim (Bim 1))
n=1
where n are the roots of the equation:
n cot n + Bim 1 = 0. (3)
 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382 79

The mass Biot number (Bim ) is dened as KRs /Deff and the mass transfer coef-
cient (K) can be calculated based on correlations reported in literature [2527].
A compilation of a number of mass transfer coefcients correlations was due to
Kunii and Levenspiel [28]. However, the equation due to Richardson and Szekely
[29] is utilized in the present study as the mass transfer coefcient was evaluated
for shallow beds:

Kdp 1.16 for 0.1 < Re < 15,
Sh = = 0.374Re0.5 (4)
D 2.01Re for 15 < Re < 250.
The Sherwood number is the ratio of external mass transfer resistance to the mole-
cular diffusivity, while the Biot number is the ratio of external mass transfer resis-
tance to the overall mass transfer resistance. The effective diffusivity is estimated
by minimizing the error between the experimental data and the model prediction.
It is not the objective of the present paper to estimate the effect of various
parameters on the kinetics of drying. The experiments are designed to capture
the variation in diffusion coefcient with the solids holdup by varying the size of the
uidization column to have identical conditions of uidization. Figure 2 shows the
plot of relative moisture content with respect to time for solids holdup from 0.30
to 2.6 kg. A good uidization condition was maintained by having a column di-
ameter of 0.148 m for solids holdup until 0.6 kg and a 0.245-m column for solids
holdup above 1.3 kg. The moisture content is found to decrease with an increase in
the time of drying irrespective of the solids holdup until attaining the equilibrium

Figure 2. Variation in relative moisture content with drying time (T , 80 C; Ci , 0.210 kg/kg;
U , 1.2 m/s; Ce , 0.030).
80 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382

moisture content. The rate of drying is found to decrease with the increase in the
solids holdup.
The temperature of the bed at any given time during the duration of drying is
found to be higher at a lower solids holdup as compared to drying with higher solids
holdup. This possibly explains the reason for the higher drying rate with lower
solids holdup. As the amount of solids is lower, the amount of moisture that diffuses
from the solids to the gas phase is lower, resulting in a higher bed temperature.
The higher bed temperature indirectly increases the rate of diffusion of moisture,
thereby increasing the drying rate. Although the bed temperature is bound to vary
from near the wet bulb temperature to the inlet temperature of the drying medium
with corresponding variation in the diffusion of moisture from the particles, the
model predicts only a single average constant diffusivity for the entire period of
drying.
Figure 2 shows a signicant variation in the drying kinetics with the increase in
solids holdup. The effective diffusivity coefcient for each of the drying curves is
estimated by minimizing the error between the model prediction and the experimen-
tal data. The estimated diffusivity coefcients are listed in Table 2. It can be seen
(Table 2) that the effective diffusivity coefcient increases with the reduction in
solids holdup. The variation of diffusivity coefcient with solids holdup is found to
vary by orders of magnitude. This large variation in effective diffusivity coefcient
cautions against the use of the effective diffusivity coefcient as a single parameter
to compare the magnitudes of drying and its use in estimation of drying kinetics.
Further, the empirical dependence of the model with additional parameters denies
extrapolation beyond the experimental range, very much limiting its application.
The application of Ficks diffusion equation for cases where the entire bed of
material is considered as single particles with known dimensions does not show any
additional dependence of the diffusivity coefcient. This nds wider application in
xed bed or static dryers/adsorbers. The application of Ficks diffusion equation to
uidized beds does not consider the whole bed, but rather considers the individual
particles that make the bed of material, with the assumption that all the materials
are exposed to uniform conditions. The variation of the number of particles with the
increase in the solids holdup could not be accounted for by the model, which brings
in the additional dependence of the model parameters with the solids holdup.

Table 2.
Evaluated effective diffusivity coefcient with the variation in the solids holdup

Dc (m) Ws (kg) Deff 1011 (m2 /s) RMSE 102

0.148 0.3 8.2 3.3

0.148 0.6 6.1 4.6
0.245 1.3 3.2 4.2
0.245 2.6 2.4 5.5
 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382 81

4. Conclusions
The use of Ficks diffusion equation to model the drying kinetics of the granular
material in a uidized bed shows a larger dependence of the evaluated effective
diffusivity coefcient on solids holdup. The dependence of the effective diffusivity
coefcient on solids holdup is due to the assumption of single particles in the bed,
rather than representing the whole geometry of the bed. The variation of solids
holdup causes the bed temperature to vary due to the variation in the number of
particles in the bed. Since the model does not have the ability to capture the variation
in the bed temperature due to the number of particles, an empirical dependence of
the effective diffusivity with the solids holdup results.
With the fallacies in the use of the basic Ficks diffusion equation in modeling
the drying kinetics of the uidized bed well known, the use of a simple empirical
approach makes good sense, as the complex models require numerical computation.
Further improvement of the basic models in terms of ability to account for the
variation in bed temperature with time is necessary before the models could be
utilized with a certain degree of condence.

References
1. C. Strumillo and T. Kudra, Drying: Principles, Applications and Design. Gordan & Breach, New
York (1986).
2. N. I. Syromyatnikov, P. K. Vasanova and D. N. Shimanskii, Heat and Mass Transfer in Fluidised
Bed. Khimiya, Moscow (1967).
3. D. Kunni and O. Levenspiel, Fluidization Engineering. Wiley, New York (1969).
4. N. Anantharaman and S. H. Ibrahim, Fluidised bed drying of pulses and cereals, in: Recent Ad-
vances in Particulate Science and Technology, S. H. Ibrahim (Ed.). Indian Institute of Technology,
Madras (1982).
5. A. N. Chandran, S. Subbarao and Y. B. G. Varma, Fluidised bed drying of solids, AIChE J. 32,
2938 (1990).
6. W. Ciesielczyk and J. Iwanowski, Analysis of udized bed drying kinetics on the basis of inter-
phase mass transfer coefcient, Drying Technol. 24, 11531157 (2006).
7. J. H. B. J. Hoebink and K. Rietema, Drying of granular solids in a uidised bed I, Chem. Eng. Sci.
35, 2135 (1980).
8. J. H. B. J. Hoebink and K. Rietema, Drying of granular solids in a uidised bed II, Chem. Eng.
Sci. 35, 2257 (1980).
9. J. B. Alebregtse, Fluidised bed drying: A mathematical model for hydrodynamics and mass trans-
fer, in: Heat and Mass Transfer in Fixed and Fluidised Bed (W. P. M. van Swaajj and N. H. Afgan,
Eds), pp. 511521. Hemisphere, New York (1986).
10. B. A. Palancz, Mathematical model for continuous uidised bed drying, Chem. Eng. Sci. 38, 1045
(1983).
11. M. Parti, Evaluation of selected mathematical model for grain drying, in: Drying 91 (A. S. Mu-
jumdar and I. Filkova, Eds), pp. 369378. Elsevier, Amsterdam (1991).
12. V. V. Kafarov and I. N. Dorokhov, Modeling and optimization of drying process, Int. Chem. Eng.
32, 475 (1992).
82 C. Srinivasakannan, N. Balasubramanian / Advanced Powder Technology 19 (2008) 7382

13. A. Topuz, M. Gur and M. Z. Gul, An experimental and numerical study of uidized bed drying of
hazelnut, Appl. Thermal Eng. 24, 15341547 (2004).
14. R. B. Keey, Drying of Loose and Particulate Materials. Hemisphere, New York (1991).
15. C. Srinivasakannan and N. Balasubramanian, An experimental and modeling investigation on the
drying of ragi (Eleusine carocane) in uidized bed, Drying Technol. 24, 16831689 (2006).
16. W. Senadeera, B. R. Bhandari, G. Young and B. Wijesinghe, Inuence of shapes of selected veg-
etable material on drying kinetics during uidized bed drying, J. Food Eng. 58, 277283 (2003).
17. M. Nitz and O. P. Taranto, Drying of beans in a pulsed uidized bed dryer: drying kinetics, uid
dynamic study and comparison with conventional uidization, J. Food Eng. 80, 249256 (2007).
18. K. M. Kundu, R. Das, A. B. Datta and P. K. Chatterjee, On analysis of drying process, Drying
Technol. 23, 10931105 (2005).
19. G. Ukan and S. Ulku, Drying of corn grains in batch uidized bed, in: Drying of Solids (A. S.
Mujumdar, Ed.), pp. 9196. Wiley, New Delhi (1986).
20. H. Feng, Analysis of microwave assisted uidized bed drying of particulate product with simpli-
ed heat and mass transfer model, Int. Commun. Heat Mass Transfer 29, 10211028 (2002).
21. S. Syahrul, F. Hamdullahpur and I. Dicer, Energy analysis of uidized bed drying of moist parti-
cles, Exergy 2, 8798 (2002).
22. H. G. Wang, T. Dyakowski, P. Senior, R. S. Raghavan and W. Q. Yang, Modelling of batch u-
idized bed drying of pharmaceutical granules, Chem. Eng. Sci. 62, 15241535 (2007).
23. T. Kudras and G. I. Efremov, A quasi-stationary approach to drying kinetics in uidized particulate
materials, Drying Technol. 21, 10771090 (2003).
24. J. Crank, The Mathematics of Diffusion. Oxford University Press, Oxford (1975).
25. K. N. Kettenring, E. L. Mandereld and J. M. Smith, Heat and mass transfer in uidized systems,
Chem. Eng. Prog. 46, 139 (1950).
26. E. V. Schlunder, M. Martin and P. Gummel, Drying Fundamentals and Technology. McGill Uni-
versity, Montreal (1977).
27. W. Resnic and R. R. White, Mass transfer in systems of gas solid uidized solids, Chem. Eng.
Prog. 45, 377 (1949).
28. D. Kunii and O. Levenspiel, Fluidization Engineering. Wiley, New York (1969).
29. J. F. Richardson and J. Szekely, Mass transfer in a uidized bed, Trans. Inst. Chem. Eng. 39,
212217 (1961).