Professional Documents
Culture Documents
Mass Transfer During Forest Biomass Particles Drying in A Fluidised Bed
Mass Transfer During Forest Biomass Particles Drying in A Fluidised Bed
ScienceDirect
Research Paper
article info
The drying of biomass particles is an important industrial process where heat and mass
Article history: transfer simultaneously occur. Following on from our previous work on convective heat
Received 25 September 2019 transfer, here we report experimental results obtained by analysing the mass transfer
Received in revised form between the gas and the surface of the wood particles, during drying in a mechanically
1 August 2020 stirred fluidised bed. It is assumed that, during the constant drying rate period, the
Accepted 3 August 2020 moisture concentration of the gas on the surface of the particle is constant. The results
Published online 27 August 2020 have allowed us to obtain a correlation between the Sherwood number and the Reynolds
number in the range of Rep between 102 and 257; the proposed correlation Shgp ¼ Shgp (Rep)
Keywords: predicts the surface mass transfer coefficient with an average deviation of 30% in relation
Fluidised bed to experimental data.
Mass transfer © 2020 IAgrE. Published by Elsevier Ltd. All rights reserved.
Particles drying
Forestry biomass
Gasesolid fluidised bed
Gasesolid mass transfer
* Corresponding author.
E-mail addresses: rmoreno@uach.cl (R.M. Moreno), greant@eii.uva.es (G. Antolı́n), alejandro.reyes@usach.cl (A.E. Reyes).
https://doi.org/10.1016/j.biosystemseng.2020.08.003
1537-5110/© 2020 IAgrE. Published by Elsevier Ltd. All rights reserved.
164 b i o s y s t e m s e n g i n e e r i n g 1 9 8 ( 2 0 2 0 ) 1 6 3 e1 7 1
where: rg is the density of the air in kg m3; U is the superficial state or transient) and type of pattern of flow (plug flow or
velocity of the air in the bed in m s1; mg the dynamic viscosity back mix flow), as well the method used to calculate differ-
of the air in N s m2; and Dv is the molecular diffusion coef- ences in water vapour concentration in the gas phase adjacent
ficient of water vapour in moist air. to the particles, which appear in the calculation equations.
In the case of a spherical particle located in a gaseous fluid Alvarez and Shene (1996) obtain the effective diffusive co-
moving at an externally imposed velocity U, it has negligible efficients of mass transfer in the interior of the solid using
natural convection effects and the experimental correlation experimental drying curves. It was concluded that the effec-
proposed by Froessling (1938) can be used: tive coefficient is in the range of 2 1011 to 1 109 m2 s1 for
sawdust, soybeans and fish feed.
rg UDp 1=2 mg 1=3 The above studies are focused on the phenomenon of
Shgp ¼ 2:0 þ 0:60 (4)
mg rg Dv moisture flow inside wood particles and wood chips, respec-
for 0.6 < Scg ¼ mg/rg Dv < 2.7 and 2 < Rep ¼ rgUDp/mg < 800. tively, assuming that mass transfer occurs by diffusion and
For fixed beds, Ranz and Marshall (1952) reported an does not cover the analysis of the convective mass transfer
expression for the calculation of the Sherwood number for phenomenon in the interface gasesolid.
large particles (Rep > 80) in liquid and gas systems. When Vasic, Grbavcic, and Radojevic (2014) studied the drying
applied to gaseous systems, the equation can be written as: process through experimental data and proposed calculation
methods for the effective diffusion coefficient.
Shgp ¼ 2:0 þ 1:8Rep1=2 Sc1=3
g (5) The study of mass transport inside solids requires knowl-
edge of the convection mass transfer coefficient kgp. This is
For gaseous system fluidised beds, Kunii and Levenspiel
particularly required when modelling with the Fick equation
(1969) highlighted the complexity of the phenomenon and
and boundary conditions that consider the external resistance
the measurement of the mass transport coefficient due to the
of the gas-solids transfer. This type of models with external
presence of bubbles, whose presence in the bed produces hot
resistance are especially important in processes in which a
air masses that by-pass and do not touch the particles. As in
large part of the drying process takes place under conditions
heat transfer, in the Rep low range, the Sherwood number falls
of constant drying rate, where the kgp convective coefficient
sharply with respect to the 2.0 value as predicted by Eq. (5) and
plays an important role. The results of Turanjanin, Stakie, &
that this corresponds to the limit case of an isolated spherical
Vodnik (1998) and Moreno, Antolin, & Reyes (2016) confirm
particle transferring mass, with the medium in resting con-
that the drying process in wood particles develops mainly
dition (i.e. pure diffusion). For these ranges, the use of the
with constant drying rates.
correlations derived by Richardson and Szekely (1961) are
proposed:
2.2. Drying of forest biomass
Shgp ¼ 0:374Re1:18
p para 0:1 < Rep < 15 (6)
One of the few reported studies on sawdust drying in units of
Shgp ¼ 2:01Re0:5
p para 15 < Rep < 250 (7) particle beds was carried out in the USA by Malte, Dorri,
Emery, Cox, and Robertus (1983). In this research, particles
For gaseous fluidised bed systems, there is some research larger than those considered in this study were studied,
on the coefficients of mass transfer, such as the work of reaching particles of up to 10 mm and operating temperatures
Ciesielczyk (1996) who determined mass transfer coefficients between 350 and 550 K. Turanjanin, Stakie, & Vodnik (1998)
in the constant drying rate period of the drying process. experimentally obtained drying curves in fluidised bed
In a subsequent work investigating the drying of particles under different operating conditions and with different spe-
in a vacuum fluidised bed, Kozanoglu, Vilchez, Casal, and cies found in Yugoslavia. The initial moisture content was
Arnaldos (2001) report a correlation of the form: 1 kg kg1 d.b. and concluded that the kinetics of sawdust
drying in fluidised bed units is strongly influenced by the
Shgp ¼ 5882Re3:07
p Knp
2:94
(8)
operating temperature, and that there are no significant dif-
where Knp corresponds to the Knudsen number defined as: ferences when experimenting with different materials.
Zabaniotou (2000) studied the drying of forest biomass in
l
Knp ¼ (9) Greece, in a rotary dryer, using a combination of experimental
Dp
results with a mathematical model composed of three stages:
where: l is the mean free path of the molecules in m. a preheating stage, a drying stage at a constant drying rate and
The dependence of the mass transfer on the Knudsen another stage with a rate of decreasing drying.
number replacing the normal dependence with the Schmidt Pang (2000) simulated the drying process of wood fibres
number can be justified in this case by the low pressures to found in New Zealand, to obtain fibres with a moist content
which the product is subjected during the drying, since, as it percentage close to 9e11% that are used in the manufacture of
decreases the pressure, the free path of the molecules is also MDF boards (medium density fibreboard). It was shown that
modified. the drying can be divided into three stages: for moist content
As in heat transfer, in the phenomenon of mass transport, above 50% the liquid moisture evaporates on the surface of the
there are large discrepancies between the correlations, due to fibre and the vapour then diffuses into the air stream. For
variations in the properties of the fluidised solids, to as- moist contents between 20 and 50% moisture evaporates in-
sumptions regarding the type of fluidisation regime (steady side the material, decreasing the rate of drying. Finally, if the
166 b i o s y s t e m s e n g i n e e r i n g 1 9 8 ( 2 0 2 0 ) 1 6 3 e1 7 1
In mass transfer, mean values of the mass transfer co- cv;s cv;0
Dcml ¼ (16)
efficients for the whole bed can be calculated using results cv;s cv;i
ln cv;s cv;o
obtained experimentally. It is assumed here that, in analogy
with heat transfer, the moisture concentration in the moist air Since the equation applies to the entire bed, then Ap ¼ SpVp
(cv,∞) shown in Eq. (1), at a given bed height, is uniform across and
the cross section of the bed and that the mass transfer is
completed in a very small distance on the distributor. mv
kgp ¼ (17)
To determine the values of the convective mass transfer Sp Vp Dcml
coefficient in the biomass-air system, for the period of constant
In terms of the drying rate (-dw/dt) Eq. (17) can be written
drying rate, a mass balance can be made. Thus, for a bed element
as:
of thickness dl, the differential flow of moisture from the surface
of the particles to the moist air, according to Eq (1) is:
rp;0 dw
dt
dmv ¼ kgp ðcv;s cv ÞSp Adl (10) kgp ¼ (18)
Sp Dcml
where: dmv is the differential mass flow of water vapour For the experimental analysis of the mass transfer coeffi-
transferred between the solid and the gas in a bed differential cient over the period of the constant drying rate, it was
element of the bed dl in kg s1; cv is the mass concentration of assumed that the concentration of water on the surface of the
moisture in the moist air in kg m3; Sp is the particle surface particles corresponds to that of the saturated air that is at the
per unit bed volume in m2 m3; A is the cross section of the wet bulb temperature of the moist air (Ciesielczyk, 1996), since
empty bed, in m2; and dl is a bed differential height in m. in this period the product shows a wet surface.
The water vapour flow can be related to the moist content The calculation of the diffusivity of the vapour in the gas,
of the moist air through: Dv, was made using the equation of Slattery and Bird (1958)
which is valid for mixtures of gases A and B at low pressures:
dmv ¼ mda dwa (11) b
pDv T
where:
1
is the mass flow of dry air in kg s- and dwa is the
mda 1=2 ¼ a pffiffiffiffiffiffiffiffiffiffiffiffiffiffi (19)
1=3 TcA TcB
differential variation of the humidity ratio of the moist air in a pcA pcB ðTcA TcB Þ5=12 1
PMA
þ PM1 B
bed differential element dl in kg kg1.
where: p is the gas pressure in Pa; pcA is the critical pressure of
Then:
the component A in atmospheres; pcB is the critical pressure of
mda dwa ¼ kgp ðcv;s cv ÞSp Adl (12) the component B in atmospheres; TcA is the critical tempera-
ture of the component A in K; TcB is the critical temperature of
If incompressible flow is considered, then: the component B in K; PMA is the molecular weight of the
b i o s y s t e m s e n g i n e e r i n g 1 9 8 ( 2 0 2 0 ) 1 6 3 e1 7 1 167
component A in kg kmol1; PMB is the molecular weight of the mechanism was coupled to a variable speed motor allowing
component B in kg kmol1; T is the temperature of the gas in K; the rotational speed of the agitator inside the bed to be varied.
and the constants have the following values: The mechanism had a 15 mm height (La) to avoid entrapment
a ¼ 3.64 104 y b ¼ 2.334. of particles in the area close to the distributor (Moreno et al.,
For the calculation of the mass transfer coefficient, using 2016).
an analysis of the aerodynamics of the biomass particles in Temperature sensors were installed and connected to a
contact with the air from Moreno, Antolı́n, and Reyes (2009) data acquisition system, consisting of a Digi-Sense Scanning
the particle surface area Sp was determined: Thermometer of 12 channels (ColeeParmer Instrument
Company, Model 92000-05, Barrington, IL USA), resolution
Sp ¼ 6737 d0:1237
p ; R2 ¼ 0:960; w ¼ 0:15kg:kg1 ; 0:89mm dp 0.1 K and ±0.5 K accuracy, with a RS-232 output to a PC, to
3:56mm collect and to analyse the information in Windows using a
(20) ScanLink 2.0 software. Data was collected every 4 s. The dryer
temperature control system consists of a PT100 temperature
and
sensor placed in the pre-chamber, under the distributor and
Sp ¼ 3778 dp0:2432 ; R2 ¼ 0:972; w ¼ 2:0 kg:kg1 ; 1:44 mm dp connected to a Fuzzy Logic micro-processor controller.
Experiments were carried out using a bed of Pinus Radiata
3:56mm
biomass particles with the inlet temperature to the dryer and
(21)
the gas outlet were recorded during drying. Because it was
necessary to place thermocouples at the inlet and outlet of the
3.2. Experimental work
bed, the gas sensor on the exit of the bed or freeboard (Fig. 2),
was protected with a filter during each drying test to prevent
The experiments were carried out in a fluid bed particles
particles influencing the outlet gas temperature measure-
dryer, shown in Fig. 1. The drying chamber had a circular
ment, as shown in Fig. 3.
section diameter of 0.3 m.
Moreover, to avoid errors caused by the heat transferred
The chamber allowed, for a bed height of 0.17 m and a
from the air to the distributor, the gas temperature was
moisture content of 0.12 kg kg1 d.b., to be loaded with
measured at the inlet of the bed, i.e., at a point just above the
approximately 2 kg of particulate biomass.
air distributor. The temperature of gas flowing through the
The chamber had an internal agitator built of steel with a
distributor had a rapid temperature drop when it came into
vertical axis and two mixers with two blades each. The
contact with the particles and reached equilibrium at a height
Fig. 1 e Schematic diagram of experimental equipment: (1) air blower; (2) motor; (3) frequency converter; (4) air heater; (5)
drying chamber; (6) air distributor; (7) agitator; (8) cyclone; (9) water manometer; (10) Pitot tube; (11) Bourdon manometer; (12)
Thermometer; (13) room thermocouple probe; (14) PT100 probe; (15) drying and bed temperature probes; (16) scanning
thermometer; (17) temperature/relative humidity probe; (18) humidity logger; (19) RS-232 cable; (20) computer; (21) fuzzy-
logic micro-processor, and (22) electric power.
168 b i o s y s t e m s e n g i n e e r i n g 1 9 8 ( 2 0 2 0 ) 1 6 3 e1 7 1
Fig. 4 e Biomass drying curve in an agitated fluidised bed; dp ¼ 2.18 mm; N ¼ 2 s¡1; Umf ¼ 0.82 m s¡1; M0 ¼ 2.0 kg;
wi ¼ 1.5 kg kg¡1; T ¼ 423 K.
Fig. 5 e Biomass drying curve in an agitated fluidised bed; dp ¼ 1.85 mm; N ¼ 2 s¡1; Umf ¼ 0.81 m s¡1M0 ¼ 2.0 kg;
wi ¼ 1.5 kg kg¡1; T ¼ 423 K.
of a complete mixing flow pattern, which was foreseeable as Moreno, R. (2005). Secado de partı́culas de Biomasa Forestal en un
the drying chamber is a fluidised bed. According to experi- lecho fluidizado. Doctoral dissertation thesis. Valladolid,
Espan ~ a: Universidad de Valladolid.
mental measurements carried out on the distributor plate, the
Moreno, R., Antolı́n, G., & Reyes, A. (2007). Thermal behaviour of
entrance zone extended approximately 20 mm above the
forest biomass drying in a mechanically agitated fluidized bed.
distributor, confirming the results of other authors. Latin American Applied Research, 37, 105e113.
Moreno, R., Antolı́n, G., & Reyes, A. (2009). Aerodynamics of
a fluidized bed of forestry biomass particles with
Declaration of competing interest mechanical agitation. Latin American Applied Research, 39,
11e18.
The authors declare that they have no known competing Moreno, R., Antolı́n, G., & Reyes, A. (2016). Heat transfer during
forest biomass particles in an agitated fluidised bed.
financial interests or personal relationships that could have
Biosystems Engineering, 151, 65e71.
appeared to influence the work reported in this paper. Pang, S. (2000). Mathematical modelling of MDF fiber
drying: Drying optimization. Drying Technology, 18(7),
1433e1448.
Ranz, W. E., & Marshall, W. R., Jr. (1952). Evaporation from drops.
Acknowledgments Part I. Chemical Engineering Progress, 48(3), 141e146.
Reina, J., Velo, E., & Pigjaner, L. (2001). Fluidization of waste-wood
The authors acknowledge the financial support of Direccio n particles with mechanical agitation of the bed. Industrial &
n y Desarrollo of the Universidad Austral de
de Investigacio Engineering Chemistry Research, 40(1), 393e397.
Chile and to Universidad de Valladolid de Espan~ a. Renstro€ m, R., & Berghel, J. (2002). Drying of sawdust in an
atmospheric pressure spouted bed steam dryer. Drying
Technology, 20(2), 449e464.
references Richardson, J. F., & Szekely, J. (1961). Mass transfer in a fluidized
bed. Transactions of the Institution of Chemical Engineers, 39,
212e222.
Slattery, J. C., & Bird, R. B. (1958). Calculation of the diffusion
Adams, R. L. (1984). Heat transfer in large particle bubbling coefficient of dilute gases and of the self-diffusion coefficient
fluidized beds. Journal of Heat Transfer, 106, 85e90. of dense gases. AIChE Journal, 4(2), 137e142.
Alvarez, P., & Shene, C. (1996). Experimental study of the heat and Temple, S. J., & Van Boxtel, A. J. B. (1999). Modelling of fluidized-
mass transfer during drying in a fluidized bed dryer. Drying bed drying of black tea. Journal of Agricultural Engineering
Technology, 14(3&4), 79e86. Research, 74, 203e212.
Banerjee, S. (2001). Mechanisms of terpene release during Turanjanin, V., Stakie , M., & Vodnik, J. (1998). Experimental
sawdust and flake drying. Holzforschung, 55(4), 413e416. investigation of sawdust drying process in a fluidized bed. In
Ciesielczyk, W. (1996). Analogy of heat and mass transfer during Proceedings of the 11th International Drying Symposium, B (pp.
constant rate period in fluidized bed drying. Drying Technology, 1645e1651).
14(2), 217e230. Vanecek, V., Markvart, M., & Drbohlav, R. (1966). Fluidized bed
Froessling, N. (1938). Ueber die verdunstung fallender tropfen. drying. New York: Leonard Hill.
Gerlands Beitrage Geophys, 52, 170e216. Vasic, M., Grbavcic, Z., & Radojevic, Z. (2014). Chemical Engineering
Kozanoglu, B. U., Vilchez, J. A., Casal, J., & Arnaldos, J. (2001). Mass and Processing, 76, 33e44.
transfer coefficient in vacuum fluidized bed drying. Chemical Watano, S., Yeh, N., & Miyanami, K. (1999). Heat transfer and the
Engineering Science, 56, 3899e3901. mechanism of drying in agitation fluidized bed. Chemical &
Kunii, D., & Levenspiel, O. (1969). Fluidization engineering. New Pharmaceutical Bulletin, 47(6), 843e846.
York: John Wiley & Sons, Inc. Zabaniotou, A. A. (2000). Simulation of forestry biomass
Malte, P. C., Dorri, B., Emery, A. F., Cox, R. B., & Robertus, R. J. drying in a rotary dryer. Drying Technology, 18(7),
(1983). The drying of small wood particles. Drying Technology, 1415e1431.
1(1), 83e115.