Professional Documents
Culture Documents
A16v29n2 PDF
A16v29n2 PDF
Vol. 29, No. 02, pp. 359 - 369, April - June, 2012
(Submitted: August 22, 2011 ; Revised: September 18, 2011 ; Accepted: December 22, 2011)
Abstract - This study analyzed the influence of the following parameters: overall volumetric heat transfer
coefficient, coefficient of heat loss, drying rate, specific heat of the solid and specific heat of dry air on the
prediction of a model for the fertilizer drying in rotary dryers. The method of parametric sensitivity using an
experimental design was employed in this study. All parameters studied significantly affected the responses of
the drying model. In general, the model showed greater sensitivity to the parameters drying rate and overall
volumetric heat transfer coefficient.
Keywords: Heat and mass transfer; Fertilizer drying; Rotary dryers; Parametric sensitivity.
Feed Shell
Flights
Rotation System
Discharge
Figure 1: The conventional rotary cascade dryer.
Arruda (2008) developed a general model for the granulated single superphosphate (SSPG) obtained
drying process in a conventional rotary dryer that from a rotary dryer in a pilot scale unit.
could be applied to any type of particulate material.
This author found good agreement of the simulated
results from this model with experimental data for MATHEMATICAL MODELLING
fertilizer drying. However, Fernandes et al. (2009)
comment that the deviations of this model Simultaneous Transfer of Mass and Energy in
simulation from their experimental data for an Rotary Dryers
industrial dryer could be explained by the
imprecision of the parameters of the constitutive Modelling of the drying process in moving bed
equations. Actually, the model proposed by Arruda dryers is based on applying the equations of
(2008) is quite general and the parameters of the conservation of mass and energy to the fluid and
constitutive equations were obtained from specific particulate phases in infinitesimal volume elements
studies, such as a thin layer dryer for the drying of the dryer (Barrozo et al., 2001). Figure 2 shows
kinetics equation and a static method for the scheme of the infinitesimal volume element of a
equilibrium moisture content. Thus, this approach conventional rotary dryer operating in countercurrent
had no adjustable parameters, allowing a realistic flow (Arruda, 2008). The modelling approach
analysis of the prediction from the model. considers the fluid dynamic characteristics of the
Therefore, is very important to analyze the dryer, as well as the intrinsic properties of the
sensitivity of the responses of the model to the material, equilibrium moisture and drying kinetics.
variation of the parameters in the main constitutive In the formulation of this model, the following
equations to identify possible weaknesses in the assumptions were made:
model prediction and thus improve the model. The particle axial velocity through the drum is
Figure 2: Schematic of the infinitesimal volume element of the rotary dryer operating at countercurrent flow.
The equation system from the balances of mass with the dry air, which corresponds to only a
and energy between drying gas and particulate fraction of the residence time (Baker, 1983). As
material is presented next, where z is the proposed by Arruda (2008), this fraction refers to
dimensionless length, given by the proportion the effective contact time between the solid and dry
between a given position (x) and the total length of air (ftef) and can be evaluated by the relation
the dryer (L). between the average falling time of the particle and
total time of a cycle. This, in turn, corresponds to
dW R w H * the time spent from material collection by the flight
=- (1) until its return to the particle bed on the bottom of
dz Gf
the drum. This fraction can be evaluated by
Equation (5), where the total number of cycles (NCi)
dM R w H * is given by Equation (6) and the effective contact
=- (2)
dz GS time between the gas and the particles (tef), by
Equation (7). The total solids load in the dryer (H*)
⎡ U va V(Tf -TS )+R w H* (λ+Cp v Tf ) ⎤ is given by Equation (8).
dTf ⎢⎣ +U P πDL(Tf -Tamb ) ⎥⎦
= (3) t q N Ci N Ci t q
dz G f (Cpf +WCp v ) f tef = × = (5)
t Ci N Ci τ
⎡ U va V(Tf -TS )+R w H *CplTS + ⎤
dTS ⎢⎣-R w H[λ+Cp v (Tf -TS )] ⎥⎦
N Ci =
L
=
L
(6)
= (4) l Yqsin(α)
dz GS (CpS +MCpl )
Levels
Parameters
-α -1 0 +1 +α
Rw (min-1) 0.90*Eq.(11) 0.94*Eq.(11) 1.00*Eq.(11) 1.06*Eq.(11) 1.10*Eq.(11)
Uva (kW m-3 oC-1) 0.90*Eq.(12) 0.94*Eq.(12) 1.00*Eq.(12) 1.06*Eq.(12) 1.10*Eq.(12)
Up (kW m-2 oC-1) 0.90*Eq.(13) 0.94*Eq.(13) 1.00*Eq.(13) 1.06*Eq.(13) 1.10*Eq.(13)
Cps (kJ kg-1 °C-1) 0.923 0.968 1.026 1.083 1.128
Cpf (kJ kg-1 °C-1) 0.900 0.944 1.000 1.056 1.100
0.06
a design in which the variance and covariance
0.04
matrices are diagonals, and the parameters are
0.02
uncorrelated.
0.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Experiment
RESULTS AND DISCUSSION Figure 3: Experimental and simulated results for the
solids moisture content at the exit of the conventional
Figure 3 show the results of the comparisons dryer for each experiment (see conditions in Table 3).
between experimental data (Arruda 2008) for the
solids moisture content, obtained at the end of the The resolution of the differential equation system
dryer (z=1), and those computed by the model with (Equations (1) to (4)) was obtained by the technique
the parameters in the central levels of Table 2. The of normal collocation with 10 placement points for
experimental data were obtained by Arruda (2008) the 4th order polynomial approximation, using the
with a conventional rotary dryer equipped with sub-routine “bvp4c” of MATLAB® software. The
three-segmented angular flights. The flights had the relative tolerance used was 10-6. The simulated
dimensions: first segment = 0.02 m, second and third results (values of response variables at the end of the
segment = 0.007 m; the length of the flights was 1.5 m, countercurrent rotary dryer, i.e., for the solid at z = 1)
with an angle of 135o between segments. The were analyzed statistically using STATISTICA®
operating conditions of the experiments performed software, with each of the responses subjected to an
by Arruda (2008) are presented in Table 3. analysis of variance. The parameters whose level of
A good agreement between the simulated values significance exceeded 5% were neglected (in a
of solids moisture content and those obtained hypothesis test using Student’s distribution) and the
experimentally in the rotary cascading dryer can be respective parameters were considered irrelevant.
observed in the results presented in Figure 3. The Equations (19) and (20) represent the matrix forms
average deviation of the simulation results in relation of the equations fitted for the solid moisture content
to the experimental data was 7.7 %. These deviations (M) and solid temperature (Ts) in condition 2,
Brazilian Journal of Chemical Engineering Vol. 29, No. 02, pp. 359 - 369, April - June, 2012
364 M. G. Silva, T. S. Lira, E. B. Arruda, V. V. Murata and M. A. S. Barrozo
respectively, as functions of the dimensionless are the most important parameters to be considered
parameters (overall volumetric heat transfer in the model. Moreover, all coefficients were
coefficient: x1; coefficient of heat loss: x2; drying retained in Equation (19), because all parameters
rate: x3; specific heat of solid: x4; and specific heat were significant for the solid moisture content, i.e.,
of dry air: x5). In the equations, the interactions and the levels of significance did not exceed 5% as
quadratic effects (r = 0.999) are present, beyond the shown in Table 4. In Equation (20), the coefficients
main effects. It was observed that the highest values of the quadratic term (x22) and interaction terms
of the coefficients of the main effects in Equations (x2x4 and x4x5) were equal to zero because these
(19) and (20) were for the parameters drying rate coefficients were irrelevant for the solid
(x3) and overall volumetric heat transfer coefficient temperature, i.e., the levels of significance
(x1). These results demonstrate that both Uva and Rw exceeded 5% (see Table 5).
where:
⎡ x1 ⎤ ⎡-0.000200 ⎤
⎢x ⎥ ⎢ +0.000109 ⎥
⎢ 2⎥ ⎢ ⎥
x = ⎢ x 3 ⎥ ; b = ⎢-0.000644 ⎥ and
⎢x ⎥ ⎢ +0.000100 ⎥
⎢ 4⎥ ⎢ ⎥
⎢⎣ x 5 ⎥⎦ ⎢⎣-0.000191 ⎥⎦
where:
⎡ x1 ⎤ ⎡ +1.16796 ⎤
⎢x ⎥ ⎢-0.11462 ⎥
⎢ 2⎥ ⎢ ⎥
x = ⎢ x 3 ⎥ ; b = ⎢-0.69211 ⎥ and
⎢x ⎥ ⎢-0.36060 ⎥
⎢ 4⎥ ⎢ ⎥
⎢⎣ x 5 ⎥⎦ ⎢⎣ +0.20112 ⎥⎦
Table 4: Regression coefficients and statistical 4 and 5, respectively). Even though most of the
tests for the solid moisture content (M). quadratic and interaction coefficients are significant,
the response surfaces look rather flat, because the
Factor
Regression
Std. Err. t(25) p
quadratic and interaction coefficients are orders of
Coeff. magnitude smaller than the linear coefficients.
Mean/Interc. 0.123719 0.000001 244684.0 <10-6
x1 -0.000200 <10-6 -1075.0 <10-6
x12 0.000007 <10-6 26.5 <10-6
x2 0.000109 <10-6 585.5 <10-6
x22 -0.000001 <10-6 -2.2 0.035841
x3 -0.000644 <10-6 -3471.1 <10-6
x32 0.000007 <10-6 25.8 <10-6
x4 0.000100 <10-6 537.4 <10-6
x42 -0.000002 <10-6 -6.0 0.000003
x5 -0.000191 <10-6 -1027.0 <10-6
x52 0.000009 <10-6 35.3 <10-6
x1x2 0.000002 <10-6 9.1 <10-6
x1x3 -0.000012 <10-6 -59.7 <10-6
x1x4 0.000001 <10-6 6.5 0.000001
x1x5 -0.000005 <10-6 -25.9 <10-6
x2x3 0.000005 <10-6 23.5 <10-6
x2x4 -0.000001 <10-6 -4.2 0.000280
x2x5 -0.000004 <10-6 -20.5 <10-6 Figure 4: Response surface of solid moisture content
x3x4 0.000003 <10-6 15.0 <10-6 (M) as a function of the drying rate (Rw or x3) and the
x3x5 -0.000011 <10-6 -51.8 <10-6 overall volumetric heat transfer coefficient (Uva or x1)
x4x5 0.000001 <10-6 5.8 0.000005
for condition 2.
Table 5: Regression coefficients and statistical
tests for the solid temperature (Ts).
Regression
Factor Std. Err. t(25) p
Coeff.
Mean/Interc. 47.71602 0.000751 63499.97 <10-6
x1 1.16796 0.000276 4233.50 <10-6
x12 -0.02749 0.000380 -72.41 <10-6
x2 -0.11462 0.000276 -415.45 <10-6
x22 0.00057 0.000380 1.49 0.147741
x3 -0.69211 0.000276 -2508.67 <10-6
x32 0.00611 0.000380 16.10 <10-6
x4 -0.36060 0.000276 -1307.07 <10-6
x42 0.00360 0.000380 9.48 <10-6
x5 0.20112 0.000276 729.01 <10-6
x52 -0.00965 0.000380 -25.41 <10-6 Figure 5: Response surface of solid temperature (Ts)
x1x2 -0.00466 0.000302 -15.42 <10-6 as a function of the drying rate (Rw or x3) and the
x1x3 -0.00126 0.000302 -4.18 0.000315 overall volumetric heat transfer coefficient (Uva or x1) for
x1x4 -0.00956 0.000302 -31.66 <10-6
condition 2.
x1x5 0.01025 0.000302 33.95 <10-6
x2x3 0.00467 0.000302 15.45 <10-6
x2x4 0.00048 0.000302 1.59 0.123429 As expected, the lowest values of solid moisture
x2x5 0.00455 0.000302 15.05 <10-6 content (M) were achieved for the highest values of
x3x4 0.01865 0.000302 61.73 <10-6 drying rate (Rw or x3) and overall volumetric heat
x3x5 -0.00585 0.000302 -19.36 <10-6 transfer coefficient (Uva or x1). Lower values of solid
x4x5 -0.00050 0.000302 -1.65 0.111414 temperature (Ts) were achieved for higher values of
drying rate and lower values of overall volumetric
Response surfaces were generated from heat transfer coefficient. However, from these results
Equations (19) and (20) for the solid moisture it is possible to quantify the effects of these
content (M) and the solid temperature (Ts) as important drying parameters and their respective
functions of the most important parameters (Figures sensitivities on the drying responses.
Brazilian Journal of Chemical Engineering Vol. 29, No. 02, pp. 359 - 369, April - June, 2012
366 M. G. Silva, T. S. Lira, E. B. Arruda, V. V. Murata and M. A. S. Barrozo
Additional simulations with perturbations of ± 20% (M) profiles as a function of the dimensionless
in the values of the parameters (Uva, Up, Rw, Cps, Cpf) length of the dryer (z) for several values of the
were performed. The impact (in percentage) of the overall volumetric heat transfer coefficient (Uva) and
perturbations (± 10% and ± 20%) in solid temperature drying rate (Rw). Figures 5 and 7 show that the
(Ts) at the end of the dryer are shown in Table 6 for the drying rate (Rw) has a much greater influence on
three operating conditions presented in Table 1. In solid moisture content (M) than the overall
general, it is observed that the variations of solid volumetric heat transfer coefficient (Uva). Figures 8 e 9
temperature (Ts) were up to -10.5%, i. e., values show the solid temperature (Ts) profiles as a function
greater than the experimental uncertainty. Again, the of the dimensionless length of the dryer (z) for
parameters that affected the solid temperature (Ts) the several values of the overall volumetric heat transfer
most were the drying rate (Rw) and the overall coefficient (Uva) and drying rate (Rw). It was
volumetric heat transfer coefficient (Uva). observed again the great influence of both Rw and
Figures 6 and 7 show the solid moisture content Uva on the solid temperature (Ts).
Figure 6: Solid moisture content (M) profiles for Figure 7: Solid moisture content (M) profiles for
several values of overall volumetric heat transfer several values of drying rate (Rw) in condition 3.
coefficient (Uva) in condition 3.
Figure 8: Solid temperature (Ts) profiles for Figure 9: Solid temperature (Ts) profiles for
several values of overall volumetric heat transfer several values of drying rate (Rw) in condition 1.
coefficient (Uva) in condition 3.
The results of this study showed that the high of heat loss (Up), drying rate (Rw), specific heat of
sensitivity of the drying rate indicates a need to be the solid (Cps) and specific heat of dry air (Cpf) on
careful in the choice of the equation to represent the the prediction of the model proposed by Arruda
drying kinetics and in the reliability of its (2008) for fertilizer drying in rotary dryers. All
parameters. Likewise, the choice and accuracy of the variables studied significantly affected the model,
parameters of the equation for the overall heat although some of the parameters are very small (e.g.,
transfer coefficient are of fundamental importance. second order parameters) compared with the linear
effect of each variable.
The statistical results indicated that the drying
CONCLUSIONS rate (Rw) and the overall volumetric heat transfer
coefficient (Uva) are the parameters that most
The method of parametric sensitivity using strongly influence the solid moisture content (M) and
experimental design was employed to analyze the solid temperature (Ts). The sensitivities of these
influence of the following parameters: overall parameters were quantified. Thus, the choices of the
volumetric heat transfer coefficient (Uva), coefficient equation to represent the drying kinetics and the
Brazilian Journal of Chemical Engineering Vol. 29, No. 02, pp. 359 - 369, April - June, 2012
368 M. G. Silva, T. S. Lira, E. B. Arruda, V. V. Murata and M. A. S. Barrozo
Iguaz, A., Esnoz, A., Martinez, G., López, A., Renaud, M., Thibault, J., Trusiak, A., Solids
Vírseda, P., Mathematical modeling and transportation model of an industrial rotary dryer.
simulation for the drying process of vegetable Dry. Technol., 18, 843-865 (2000).
wholesale by-products in a rotary dryer. J. Food Ribeiro, J. A., Oliveira, D. T., Passos, M. L. A.,
Eng., 59, 151-160 (2003). Barrozo, M. A. S., The use of nonlinearity
Kemp, I. C., Comparison of Particles Motion measures to discriminate the equilibrium moisture
Correlations for Cascading Rotary Dryers. Proc. equations for Bixa orellana seeds. J. Food Eng.,
14th Int. Dry. Symp. (IDS), São Paulo, Brazil, B, 66, (1), 63-68 (2005).
790-797 (2004). Shahhosseini, S., Cameron, I. T., Wang, F. Y., A
Kemp, I. C., Oakley, D. E., Simulation and scale-up simple dynamic model for solid transport in
of pneumatic conveying and cascading rotary rotary dryers. Dry. Technol., 18, 867-886 (2000).
dryers. Dry. Technol., 15, 1699-1710 (1997). Sheehan, M. E., Britton, P. F., Schneider, P. A., A
Krokida, M. K., Maroulis, Z. B., Kremalis, C., model for solids transport in flighted rotary dryers
Process design of rotary dryers for olive cake. based on physical considerations. Chem. Eng.
Dry. Technol., 20, 771-788 (2002). Sci., 60, 4171-4182 (2005).
Lisboa, M. H., Vitorino, D. S., Delaiba, W. B., Sherrit, R. G., Caple, R., Behie, L. A. and Mehrotra, A.
Finzer, J. R. D., Barrozo, M. A. S., A study of K., Movement of solids through flighted rotating
particle motion in rotary dryer. Braz. J. Chem. drums. J. Chem. Eng., 71, 337-346 (1993).
Eng., 24 (3), 265-374 (2007). Song, Y., Thibault, J., Kudra, T., Dynamic
Lobato, F. S., Steffen Jr., V., Arruda, E. B., Barrozo, characteristics of solids transportation in rotary
M. A. S., Estimation of drying parameters in dryers. Dry. Technol., 21, 775-773 (2003).
rotary dryers using differential evolution. J. Phys. Xu, Q., Pang, S., Mathematical modeling of rotary
Conf. Ser., 135, 1-8 (2008). drying of woody biomass. Dry. Technol., 26,
Page, G. E., Factors influencing the maximum 1344-1350 (2008).
rates of air drying shelled corn in thin-layer. Zabaniotou, A. A., Simulation of forestry biomass
M.Sc. Thesis, Indiana-USA, Purdue University drying in a rotary dryer. Dry. Technol., 18, 1415-
(1949). 1431 (2000).
Brazilian Journal of Chemical Engineering Vol. 29, No. 02, pp. 359 - 369, April - June, 2012