Professional Documents
Culture Documents
00
Printed in Great Britain. Pergamon Journals Ltd.
and
L. T. FAN
Department of Chemical Engineering, Kansas State University, Manhattan, KS 66506, U.S.A.
Abstract-A fairly rigorous mechanistic model of a continuous fluidized-bed dryer has been developed. It
depicts the dynamic interactions between gaseous and solid phases in detail. The performance of the dryer has
been simulated numerically based on the model. The effects of the operating parameters on the performance
characteristics of the dryer have been investigated. These parameters include the superficial gas velocity, the
inlet temperature of the drying gas, the mean residence time of solids and the dryer-wall temperature. The
results of simulation based on the present model are compared with those based on an existing model. This
comparison shows that the former is a substantial improvement over the latter.
INTROD%JCTION moisture content of the outlet gas and also the average
The fluid&d-bed dryer possesses many significant moisture content and temperature-of the solids at the
features over the conventional packed-bed or moving- exit. It will be amply demonstrated that the proposed
bed dryer [see, for example, Vlnecbk et al. (1966), model represents a significant improvement over an
Nonhebei and Moss (197 l)]. These include the follow- existing mechanistic model for the continuous
ing: (i) drying gas is locally mixed intensively during its fluidized-bed dryer proposed by Palancz (1983).
passage through the bed; consequently the rate of mass
and heat transfer between the gas and solids are high, MATHEMATICAL MODELLING
(ii) the extremely rapid heat transfer enables a rela- A schematic diagram of the model is shown in Fig. 1.
tively high inlet gas temperature to be used; (iii) the The present model is based on the two-phase theory of
time of drying is relatively short. fluidization (see, for example, Davidson and Harrison,
Because of its numerous advantages, fluidized-bed 1963). The underlying assumptions of this theory are
drying has been increasingly applied in diverse indus-
tries in either the batch or continuous mode (VaniZccCk
et al., 1966, Viswanathan et al., 1982). In fact, several OUTLET GAS
INLET GAS
2419
2420 F. S. LAI et al.
that the bed is divided into two phases, a bubble phase 1. The bed fraction of the bubble phase, 6,:
and an emulsion phase (which remains in minimum
fluidization conditions), and that the excess flow of the a,= 1-9 (4)
fluidizing fluid above minimum fluidization conditions r
passes through the bed as bubbles. The fluid in the where H,/H,, is given by (Babu et al., 1978)
bubble and emulsion phases and the solid particles are
H 14.311 (U, - Umf)o~‘3sd~~oo6ppo.‘36
considered to be continua. Additional simplifying as- r= 1+
sumptions imposed in deriving the present model are H mf ~0.937
mf Pg
0.126
as follows: (5)
Tp = iz ;exp(--t) T,dt,.
ip =~~(~~-Trer)+x~cw(*~-*r~~) (49)
(41)
i,, = YO + c, CT, - *,,) . (50)
c
We define the average temperature of gas bubbles, Tb,
The energy balance around the stagnant film sur-
as
rounding the particle yields (see Fig. 3)
o= 9 f
{c*co- T,‘) + x0 cc,,(j-0 - T,,)
+ YOI-cgCTe - Tre,-)
-x,~c,“(T,-~T,,,)+~ol}+6,(H,),(~b--T,)
+(1--6,)(~--s,,)~(~~-Xx,)[C,(~p-TT,,)+g,,
P
STAGNANT FILM
EMULSION
+awhw(*w-_T,)-_(l --6b)(i -C,,,,)$$,(q-q)
P
-P,Sb(K,)b(x,--b)[c,(r,-*,f)+yo] (9
or
~(C.(r,-r,)+(x,-x,)yo+c~~(r,--T,,)x,
r
Fig. 3. Energy balance around the stagnant film surrounding
-(*o - ~ef)Xol} = &,(&),(T-- T,)+ (I -6,) a solid particle.
Continuous fluidized-bed dryer 2423
(62)
and
1
(53) T,",= (U,,Ccs T, + x, (c, T, + ro)l
Uo(c, +%"I=,,)
or
- ~OX,“,YOI- (63)
NUMERICALSIMULATION
The solution of the model equations is obtained
through a two-dimensional trial-and-errorprocedure.
For simplification, first we seek to reduce the integro-
differential equations to a set of first-order differential
y0 in eq. (54) is to be evaluated at T,,. It can be related equations. This is achieved by introducing three new
to the heat of vaporization at any arbitrary tempera- intermediate variables.
ture, T, as follows:
Or
(SW
with the boundary condition
dr,” TP (65b)
dt, = TS-exp
Tp = Tpo at t,=O (58b) with the boundary condition
and y. to be evaluated at T = 0°C. The average Tp” =0 at t, = 0; (65~)
temperature of particles, T, can be evaluated from
and
Tb dr (66a)
When t. exceeds a certain value, e.g. tf , xz and T in eqs X:lr,_r,Ur T;lcI=fz and Tb*lrcHr
(64a) and (65a), respectively, remain constant; &en, we through eqs (64b), (65b) and (66b) with the cor-
have responding boundary conditions by using the
Runge-Kutta method.
%;=ilmx;exp(-t)dc Calculate $, L$ and Tb using eqs (70), (71) and
(5)
(69), respectively.
Evaluate x, and T, from eqs (17) and (46),
=i l’x;exp( -~)d~~+~~“x;e~~(-5>d~~ (@
respectively.
(7) Compare x, and T, calculated in step (6) with the
= Xp* I,.=t; + xp*It,= tz exp . (70) initially guessed values XL and TL. If they are not
identical, determine a new pair of initial values of
xe and T, and repeat steps (1 j(7).
Similarly,
(8) Stop when XL, TL and xc, T, are identical.
^.
The stopping criteria used in the present study are
TP = T~(r,=r,O+TpIt,=r~exp -5 . (71)
( s> 1x:-x,/ < 10-4 and IT:-Tel < lo-‘.
Thus, the solution of the governing equations, eqs (3),
For illustration, the following data are considered
(17), (33a), (46), (27a) and (58a), can be obtained by (Palancz
1983): UO= lms-I, To = 25O”C, x0
solving only a set of first-order differential equations
= 0.015,’ ps = 1 kgmP3, ps = 2500 kgm-j, pw
along with several algebraic equations. The calculation
= 1000kgm-3, p(B= 2 x 10e5 kgm-‘s-‘, k, = 2.93
procedure is described below.
x 10-z Jm-‘s-‘“C-‘, cs= 1.06kJkg-‘“C-I, cP
(1) Input data. = 1.26kJkg-‘“C-‘, c,, = 1.93kJkg-‘“C-l, cw
(2) Assume the initial values XL for xc and TL for T,. = 4.19 kJkg- 1 “C- I, y,, = 2.5 x lo3 kJkg- l, xPO
Fig. 4. Effect of the inlet-gas temperature. T,,, T 105”C, Tpo = 2O”C, U. = 1 m/s, xpo = 0.35, x,, = 0.015,
t, = 300s.
Continuous fluid&d-bed dryer 2425
=0.35, <=3OOs, H,=0.5m, D,=O.l5m, d,=2 curve is a linearly declining section. The remaining
x lo-‘m, Ds = 2 x lo-’ m2s-i, Tpo = 2O”C, T, portion of each of the two curves represents the falling-
=105”C,x,=0.2,n=3,K=lx10-2. rate drying period in which the temperature and
moisture content of the particle approach gradually
RESULTS AND DISCUSSION their respective equilibrium values.
The average moisture content and temperature of
particles at the exit are related to the inlet-gas Eflectsof the operating parameters
temperature in Fig. 4. Figures 5-7 show the effects of The performance characteristics of the dryer under
various operating parameters on variations of the various To are revealed in Fig. 4. The higher the
temperature and moisture content of a single particle temperature of the inlet gas, the higher the temperature
as functions of time. In Fig. 8, the temperature and of the gas in the bubble and emulsion phases, thus
moisture content of a particle based on Palann’s enhancing the rate of evaporation. This, in turn, results
model are compared with those based on the present in an increase in the average temperature and a
model. The three stages of drying can be clearly decrease in the moisture content of particles at the exit.
identified in the x,(f,) and T,(t,) curves in Figs 5-8. Note that the Tr curve in Fig. 4 with To less than a
The rather short initial stage of the T,(Q) curves, each certain value (250°C in this example) has a relatively
with a steep positive slope, involves the preheating of a small gradient with respect to To_ This implies that the
particle, resulting in a sharp rise in its temperature dryer is not highly sensitive to the change in To. TO
from the inlet value. The subsequent horizontal section prevent burning or cracking of particles, the drying
represents the constant-rate drying period with the operation needs to be conducted within this range,
temperature of the particle equal to the wet-bulb where moderate fluctuations in To will not cause
temperature. The corresponding portion of the xp (t,) overdrying.
80 0.4
70
60 0.3
50
40 a2
1.0
i -\, ---_-_--__---.
\
30 - \
1.2
\ ---____--- ---
20 0.1
Parameter: U, [m/we]
IO
0 0.0
0 100 200 300 400
time, ts [sec.]
Fig. 5. Effect of the superficial gas velocity. To = 25O”C, TM = 2O”C, T, = 5O”C, x,,,, = 0.35, x0 = 0.015,
t, = 300s.
2426 F. S. LAI et al.
The influence of the superficial gas velocity on the The effect of the dryer-wall temperature on the
performance characteristics of the dryer can be dis- variations of the moisture content and temperature of
cerned in Fig. 5. When U,, increases, the average an individual particle as functions of time can be
temperature of particles at the exit increases ap- observed in Fig. 6. Naturally, a rise of wall temperature
preciably while their average moisture content reduces increases the rate of heat transfer to the emulsion gas.
sharply. This can be attributed to the intensified mass This leads to an increase in temperature of the
and heat transfer among bubbles, emulsion gas and emulsion gas, thereby enlarging the driving force for
solids. Figure 5 shows that the gradients of the T,(t,) evaporation of moisture from the particle.
and xp(t,) curves are substantially increased in the Consequently, the average temperature of particles at
constant-rate drying period. It is worth noting that the exit increases while their average moisture content
these gradients are not affected as significantly by the decreases.
change in UO in the falling-rate period as they are in the The effect of the mean residence time of particles on
constant-rate drying period. This phenomenon sug- the dryer performance is illustrated in Fig. 7. With the
gests that the fluidized-bed dryer is effective in enhanc- bed height fixed, the smaller the mean residence time,
ing the drying rate mainly in the constant-rate drying the larger the feed flow rate of solids and the shorter
period. The relationship between the superficial gas the contact time between the particles and drying gas.
velocity and the length of the constant-rate drying This results in a relatively low average temperature and
period can be roughly approximated by the expression a high moisture content for the particles at the exit.
rs = 4 .8 x lo4 e--6.*“0
Comparison with an existing mechanistic model
which should be of practical use in the design of the A mechanistic model proposed by Palancz (1983)
fluid&d-bed dryer. gives a comprehensive description of the heat and mass
9c
8C
7c
60
50
40
-1 -_------ 105 ___-
30
Curve
-
I Vorioble
iT, (t-1 1
20
IO
50 100 150
time, ts bl
Fig. 6. Effect of’the dryer-wall temperature. To = 25O”C, TM = 2O”C, U0 = 1 m/s, xpo = 0.35, x0 = 0.015,
r, = 300 s.
Continuous fluid&d-bed dryer 2427
80
0.3
60
50
40
30
20
-, , , ,Eij
IO
0 50 100 150 zoo0
time, 1, [WC]
Fig. 7. Effect of the mean residence time of particles. T, = 25o”C, T, = lOS’C, TFo = 2O”C, U0 = 1 m/s,
xpo = 0.35, xg = 0.015.
transfer among gaseous and solid phases in a continu- drying gas from solids, an appreciable amount of
ous fluid&d-bed dryer. It is free of the assumptions moisture migrates from the emulsion gas to the
that the drying gas is homogeneous and that exit bubbles; it is not plausible that its accompanying
streams are in equilibrium. Palancz’s model appears to thermal energy can be neglected. A consequence of this
be the only existing model comparable to the present assumption is that in Palancz’s model, the energy
one. In fact, the present model is an exhaustive conservation equation for the bubble phase, which
amendment and a substantial extension of Palancz’s corresponds to eq. (33a), is linear and only contains the
model. The major differences between the two models first term on the right-hand side of the equation.
are as follows: Subsequently, in his energy conservation equation for
1. To simplify the governing equations of his model the emulsion gas, the term designating the energy
and to facilitate its solution, Palancz has imposed an transfer accompanied by the evaporation of moisture
assumption that the specific heat of the drying gas contains only T, instead of Tp - T,. This means that the
remains constant throughout the entire drying process. energy conservation equation depends on the choice of
In other words, reference temperature, which is impossible.
cg = X,C, = constant 2. To evaluate the equilibrium moisture content of
and the drying gas on the surface of a particle, Palancz’s
cg = xbc,, = constant. model resorts to the approximate expression
and CONCLUSION
SO 0.4
70
60
I Curve I Variable
0 50 100 150
time, ts [set]
Fig. 8. Comparison of the present model with Palann’s model. To = 25O”C, T, = 5O”C, T* = 2O”C,
U0 = 1 m/s, xpo = 0.35, x0 = 0.015, t, = 300s.
Continuous fluid&d-bed dryer 2429
Palancz, which assumes a constant specific heat for the bubble and cloud-wake regions based on the
drying gas. The results of simulation have proved that volume of bubbles, s-l
this assumption leads to an overestimation of the (Kbc)b coefficient of gas interchange between the
temperature and moisture content of particles. bubble and emulsion phases based on the
It is unlikely that the moisture content of the drying volume of bubbles, s- 1
gas on the surface of a particle can undergo a &e)b coefficient of gas interchange between the
discontinuity as suggested in Palancz’s model. The cloud-wake region and the emulsion phase
present model does not contain such a discontinuity, based on the volume of bubbles, s-l
and thus should be more rational in expressing heat k, thermal conductivity of the drying gas,
and mass-transfer relationships between the drying gas
Jm--'"C--'
on the surface of a particle (dry basis), fluidization velocity and fluid&d bed expansion ratio.
dimensionless A.I.Ch.E. Symp. Ser. 74. 176-186.
moisture content of inlet particles (dry basis), Broadhurst. T. E and Becker, H. A., 1975. Onset of’ Auidiz-
%a ation and slugging in beds of uniform particles. A.1.Ch.E. J.
dimensionless 21, 238-247.
5 critical moisture content of a particle (dry Davison, J. F. and Harrison, D., 1963, Fluidized Particles,
basis), dimensionless Chap. 1, pp. 1%20. Cambridge University Press,
z elevation, m Cambridge.
Kato, K., Omura, S., Taneda, D., Onozania, I. and Iijima, A.,
198 1, Drying characteristics in a packed fluidized bed dryer.
Greek letters J. them. Engng Jap. 14, 365371.
YO heat of vaporization, kJ kg-’ Kunii, D. and Levenspiel, O., 1969, Fluidization Engineering,
fraction of the fluidized bed consisting of Chap. 7, Wiley, New York.
6,
bubbles, dimensionless Li, C. H. and Finlayson, B. A., 1977, Heat transfer in
packed beds-a reevaluation. Chem. Engng Sci. 38,
% void fraction in the emulsion phase, 147-l 53.
dimensionless Nonhebel, G. and Moss, A. A. H., 1971, Drying of Solids in the
Emf void fraction at minimum fluidizing con- Chemical Industry, Chap. 11. Butterworth, London.
ditions, dimensionless Palancz, B., 1983, A mathematical model for continuous
fluidized bed drying. Chem. Engng Sci. 38, 1045-1059.
PI5 viscosity of gas, kgm- ’ s- r Palancz, B. and Parti, M., 1973, Examination of the heat-and-
&3 density of gas, kg m- 3 moisture-content variations in granular bed types. Acta
PS density of dry solids, kg m- 3 Tech. Acad. Sci. Hung. 74, 441461.
density of water, kg m- 3 Van&c&k,V., Markvart, M. and Drbohiav, R., 1966, Fluidized
PW
Bed Drying (Translated by Landau, J.). Leonard Hill,
PWS density of wet solids, kg m- 3
London.
CJ evaporation coefficient, kg m- 2 s- l Viswanathan, K., Subba Rao, D. and Raychaudhury, B. C.,
4, sphericity of a particle, dimensionless 1982, Coherent representation of the drying of gas and
solids in fluidized beds. Ind. them. Engq- 14, 12-23.
Wen, C. Y. and Yu, Y. H., 1966, A generalized method for
REFERENCES
predicting the minimum guidrzation velocity. A.1.Ch.E. J.
Babu, S. P., Shah, B. and TalwaJkar, A., 1978, Fluidization 12, 610-612.
correlation for coal gasification materials-minimum