Drying 2004 Proceedings of the 14th International Drying Symposium (IDS 2004) So Paulo, Brazil, 22-25 August 2004,

, vol. A, pp. 477-484

DYNAMIC MODEL FOR FREEZE DRYING PROCESS

Edinara Adelaide Boss , Eduardo Coselli Vasco de Toledo and Rubens Maciel Filho 1. State University of Campinas, Chemical Engineering Faculty, Laboratory of Optimization, Design and Advanced Control, P.O. Box 6066, 13083-970 Campinas, SP, Brazil, E-mail: [email protected]
Keywords: Freeze drying, Dynamic Process, Mathematical Model ABSTRACT Freeze drying process involves the following three stages: freezing, primary drying and secondary drying stages. A theoretical model was developed and solved in order to describe quantitatively the dynamic behavior of the primary and secondary drying stages. The solution procedure involves the discretization of the equations by orthogonal collocation in the axial direction and the integration, in relation, to time by differential integrator (LSODE). For the validation procedure, the results of the solution model were compared real values. The model allows seeing design and operating variables more significantly impact the system behavior. Furthermore, the intention was to simulate the process aiming high quality dried product with the minimum time process. The results show that the proposed model presents better predictions than the existing mathematical model, since a more detailed process description is provided. INTRODUCTION Freeze drying process has different applications including pharmaceutical products, foodstuffs (whose organoleptic properties are important and have to be maintained) and other important industrial products (conservation of living microorganism, dehydration or concentration of heat labile). The interesting factors to analyze are the structural rigidity, which facilitates rapid and almost complete rehydration at a later time, little loss of flavor, and minimization of degradative reaction, which normally occur in ordinary drying processes, such as protein denaturation, non-enzymatic browning and enzymatic reactions. The interactions in the variables of such process are complex, and an experimental approach for examining the various operational policies is hard, expensive and time consuming. Alternatively computer simulation may be a convenient and suitable tool to investigate and understand the system behavior. An important point to be considered in such approach is the quality of the model in terms of reliable model predictions.

477

The proposed mathematical model is based on the work of LIAPIS and SADIKOGLU, 1997. The result of the solution model was compared with real values so that the model was validated. The objective was to obtain a mathematical model that could predict the experimental data, so that process simulation is possible to be made. The proposed model takes into account the main process phenomena and allowed to represent well both the industrial data for soluble coffee and pilot plant experimental data for skimmed milk. PROCESS DESCRIPTION Freeze drying is a process by which a solvent is removed from a frozen material or frozen solution by solvent sublimation and by desorption of the sorbed solvent, generally under reduced pressure. This process involves the following three stages: freezing, primary drying and secondary drying stages, as shown in Figure 1:

Freezing
Figure1: Freeze drying stages.

Primary drying

Secondary drying

Freeze dried product

The material to be processed is cooled down to a temperature bellow the solidification. The shape of the pores, the pore size distribution, and the pore connectivity of the porous network of the dried layer formed by the sublimation of the frozen water during the primary drying stage depend on the ice crystals formed during the freezing stage. An excellent option is to obtain large dendritic ice crystals formed and a homogeneous dispersion due to the high mass transfer rate of the water vapor in the dried layer (Liapis et. al, 1996). In such conditions the product could then be dried more quickly. The solvent is removed by sublimation under vacuum and heat is added in the primary drying stage. A significant amount of the sublimation latent heat is also consumed when the water molecules sublimate and enter to the vapor phase. Because of this, the temperature of the frozen product is reduced. It is then necessary to supply heat to the product, which could be provided by conduction, convection and/or radiation. During the secondary drying stage the solvent is removed from the chamber and a small amount of sorbed water can be removed by desorption. The bound water is removed by heating the product under vacuum. PARAMETRIC SENSITIVITY ANALYSIS Due to the relatively large number of variables to be analyzed, it was necessary to submit the process to an initial screening design. The methodology of Plackett-Burman (1946) is an adequate tool for this initial screening, since it makes it possible to determine the influence of various factors with only a small number of trials, instead of using more extensive factorial design, which would furnish more complete information, but would also involve complexity and larger computer effort. The variables analyzed are: model constant dependent upon the structure of the porous medium and relative Knudsen flow permeability (C1, m), constant dependent upon the structure of the porous medium and giving the ratio of bulk diffusivity within the porous medium to the free gas bulk diffusivity (C2, dimensionless), effective dried layer density (Ie, kg/m3), dried layer density (I, kg/m3), frozen layer density (II, kg/m3), thickness of the layer (L, m), plate temperature (Tplate, K) and total pressure (p, N/m2). Each independent variable is tested at two levels, a high (+) and a low (-) level, as shown in Table 1.

478

Table 1: Level of variables for skimmed milk process.

II 773 1290

L 0.015 0.025

P 3.8 6.34

Tplate 234.9 390.6

Bearing this in mind, the results of Figure 2 depict that the variables with larger effects are C2, Ie and L. The variable P was also included in the factorial since it is an important operational variable of the process. Once the relevant variables were selected by screening, simulations have been planned to obtain a quadratic model, consisting of 24 trials plus a star configuration with a central point. Table 2 presents the range of the variables for the complete factorial design.
2 1
-2 PERCENTUAL EFFECT (%) -4 -6 -8 -10 -12

0 PERCENTUAL EFFECT (%) -1 -2 -3 -4 -5 -6 -7 -8 -9

C1 C2 ROIE ROI ROII L P TPLACA

C2 ROIE L P ROIEL
2 TIME (H) 4

-10

Figure 2: Percentual effect for amount of removed water. C1=C1.C2=C2. ROIE=Ie. ROI=I. ROII=II. TPLACA= Tplate.

Figure 3: Percentual effect for amount of removed water. C2=C2. ROIE=Ie.

Table 2: Range of variability for complete factorial design.

C2 L Ie 0.76 161 0.015 1.27 269 0.025 1.01 215 0.02 0.5 107.5 0.01 1.52 322.5 0.03 a 2 = (24)1/4 (star point): where 4 is the number of variables.

P 3.8 6.34 5.07 2.535 7.605

Figure 3 shows the effects of the variables from the complete factorial design. As can be seen the most L gives relevant effect. When the value of L was increased from the negative disturbance to the positive disturbance, the amount of removed water decreases. The same occurs with Ie and the opposite is observed when the value of Ie and L were increased from the negative disturbance to the positive disturbance at the same time.

479

Throughout the use of this analysis and from extensive of simulation was possible to obtain a term that was included on the model. The explanation about this is presented as follow. MATHEMATICAL MODEL A mathematical model based on fundamental mass and energy balance equations have been proposed, based on an original model (Liapis and Sadikoglu, 1997), and used to know the amount of removed and residual water. The assumptions are those considered by Millman et. al. (1985) and Lichtfield and Liapis (1979): - One-dimensional heat and mass transfer; - Sublimation occurs at an interface parallel to surface; - The solid layer is semi-infinitive; - A binary mixture of water vapor and inert gas flows through the dried layer; - At the interface, the concentration of water vapor is in equilibrium with the ice; - In the porous region, the solid matrix and the enclosed gas are in thermal equilibrium.; The frozen region is considered to be homogeneous, with uniform thermal conductivity, density, and specific heat, and to contain a significant proportion of dissolved gases. The energy balance of the developed dynamic model is shown in equations (1) and (2). The mass balance equations are described in equations (3) and (4).
Cp g TI H v I Csw (N t TI ) C1.P = . t x Ie .CpIe t Ie .CpIe 9.5(C2 . Ie .L )

(1)

This equation is valid for the primary a secondary dried periods, in accord to the application intervals and data to be used, which depend on the dried period that they refer to. For the first period, the interval is 0 x X and for the secondary period is 0 x L. In such equations x is the spatial coordinator, which is the thickness of the product on the tray, L the thickness of the product on the tray and X the position of frozen interface.

1 pw R t TI

TII 2TII = II , X x L t x 2 C sw 1 N w ,0xX = I M w x M w t

(2)

(3) (4)

1 p in 1 N in , 0 x X = R t TI M in x The term

Csw accounts for changes in the concentration of sorbed or bounded water with t time. TI is the dried layer temperature (K), Hv is the enthalpy of vaporization of sorbed water (J/kg), I is the dried layer density (kg/m3), Ie is the effective dried layer density (kg/m3), CpIe is the effective dried layer heat capacity (J/(kg K)), Cpg is the gas heat capacity (J/(kg K)), Nt is the total mass transfer flux (kg/(m2 s)), TII is the frozen layer temperature (K), II is the frozen layer thermal diffusivity (m2/s), R is the Universal gas constant, pw is the partial pressure of the water pressure (N/m2) pin is the partial pressure of the inert gas (N/m2), MW is the molecular weight of the water vapor, Min is the molecular weight of the inert gas, NW is the water vapor mass transfer flux (kg/(m2 s)), Nin is the inert gas mass transfer flux (kg/(m2 s)) and is the voidage fraction.

480

In the proposed model, energy balance equation differs from the original model due to the introduction of a new term and neglecting of the quadratic term. The quadratic term takes account two phenomena: diffusivity and convectivity. The convectivity is the predominant phenomena.
The initial and boundary conditions for equations 1 to 4 for the primary stage are:

TI = TII = Tx = T0 , t = 0, 0 x L 0 0 pw = pw , p in = p in , t = 0, 0 x X
0 C sw = C sw , t = 0, 0 x L T q I = k Ie I x =0 , x = 0, t > 0 x

(5) (6) (7) (8)

and,

4 q I = .F TUP (TI (t ,0 )) , t > 0 4

(9) (10)

for the radiation heat transfer to the upper dried surface, T T k II II x = X k Ie I x = X + V ( II cp II TII x = X I cp I TI x = X ) + N t cp g T = H s N t , x = X, 0 < x x t t X =L TI = Tx = TII , x = X, t>0
TII x = L , x = L, t > 0 x 4 4 q II = .F TLP (TII (t ,0 )) , t > 0 q II = k II

(11) (12) (13)

q II = k f TLP (TII

) (t , L )) ) , t > 0

(14)

where is the Stefan-Boltzmann constant (kW/(m2 K4)), TUP is the upper plate temperature (K) and TLP is the lower plate temperature (K), kf film thermal conductivity.
The pressure can be calculated as follow:

p w = p w0 , pin = pin 0 = P0 Pw0 , x = 0, t>0

(15)

pw = f (Tx) , x=X, 0 < t t X = L pin x = X = 0 , x = X, 0 < t t X = L x

(16) (17)

V =

N w x= X dX = dt II I

(18)

The initial condition for the equation (18): X=0, t=0

(19)

The initial and boundary condition for the equation (1) for the secondary stage may be written as: (20) TI = (x ) , t = tX=L , 0 x L T (21) q I = k Ie I x =0 , x = 0, t > t X=L x (22) TI q II = k Ie x = L , x = L, t > t X=L x

481

and for the radiation heat transfer: 4 4 q I = .F TUP (TI (t ,0)) , t > t X=L

(23) (24) (25) (26) (27) (28) (29) of the of the

The initial and boundary conditions for equations (1) for the secondary stage are: pw = (x), t = t X=L, 0 x L pin = (x), t = t X=L , 0 x L Csw = (x), t = t X=L , 0 x L p w = p w0 , pin = pin 0 = P0 Pw0 , x = 0, t = t X=L p w x = L = 0 , x=L, t < t X=L x p in x = L = 0 , x=L, t > t X=L x The functions (x), (x), (x) and (x) provide the profiles of TI, pw, pin e Csw at the end primary drying stage or at the beginning of the secondary drying stage obtained by the solution model equation for the primary drying stage. RESULTS AND DISCUSSION

The predictions from the original model (Liapis and Sadikoglu, 1997), from the proposed mathematical model and the experimental (pilot plant of skimmed milk) data are shown in Figures 4 and 5. It is observed that the proposed mathematical model presents better predictions compared to the original model in relation to the real data. Figures 6 and 7 show the amount of removed water and the residual water compared to the data from soluble coffee a continuous operated industrial plant. The values of the parameters required to the mathematical model were from the real plant but cannot be disclosed. As can be seen the proposed model presented an excellent performance allowing representing the real data very closely. A detailed deterministic model based on mass and energy balances was proposed. Taking into account a mathematical model from literature (Liapis and Sdikoglu,1997) new terms were adopted based on extensive analysis of the process.
0,15
0,6 AMOUNT OF REMOVED WATER (KG) CURVE1 REALDATA CURVE2

0,5

AMOUNT OF RESIDUAL WATER (KG)

CURVE1 REALDATA CURVE2 0,10

0,4

0,3

0,2

0,05

0,1

0,0 2 4 6 8 TIME (H) 10 12

14

16

18 TIME (H)

20

22

Figure 4:Amount of removed water versus the time during the primary period of freeze drying of the skimmed milk. CURVE1: original model. REALDATA: real data. CURVE2: proposed model.

Figure 5: Amount of residual data versus the time during the apprenticeship of primary period of freeze drying of the skimmed milk. CURVE1: original model. REALDATA: real data. CURVE2: proposed model.

482

0,6 AMOUNT OF REMOVED WATER (KG)

0,5

0,4

0,3

0,2

0,1 1 2 3 4 5 TIME (H) 6

REALDATA MODEL
7 8 9

0,090 0,085 0,080 0,075 0,070 0,065 0,060 0,055 0,050 0,045 0,040 0,035 0,030 0,025 0,020 0,015 0,010 0,005 0,000 -0,005 8 10 12 14 TIME (H)

AMOUNT OF RESIDUAL WATER (KG)

REALDATA MODEL

16

18

20

Figure 6: Amount of removed water versus time Figure 7: Amount of residual water versus time during the primary drying stage of freeze drying of during the secondary drying stage of freeze drying soluble coffee. of soluble coffee. CONCLUSION The results show that the proposed mathematical model represents well the process. The methodology of Plackett-Burman was shown to be very useful for the determination of significant variables for further process optimization. In such way, it is possible to consider a large number of variables and to avoid the loss of information, which might be essential for the model development as well as in the optimization of the process.The use of this technique reduced the number of variables to be analyzed, for the considered case study from eight to four. ACKNOWLEDGEMENT The authors acknowledge the financial support from FAPESP. Process: 99/11333-1. NOTATION C1 C2 constant dependent only upon structure of porous medium and giving relative Knudsen flow permeability constant dependent only upon structure of porous medium and giving the ratio of bulk diffusivity within the porous medium to the free gas bulk diffusivity, dimensionless heat capacity weight fraction of bound water in dried layer thermal conductivity thickness molecular weight mass flux partial pressure total pressure heat transfer rate time temperature interface velocity. position of frozen interface m Dimensionless

Cp
C sw

k L M N p P q t T V X

J/kg K kg of water/kg of solid kW/(m K) M kg/kmol kg/(m2 s) N/m2 N/m2 KJ/s S K M/s M

483

x
Hs Hv Greek Symbols

position thermal diffusivity pw at the end of primary drying period. voidage fraction. enthalpy of sublimation of frozen water enthalpy of vaporization of sorbed water TI at the end of primary drying period. Csw at the end of primary drying period. pin at the end of primary drying period. density Stefan-Boltzmann constant Effective value. Gas. Inert. Lower plate. total Water vapor. Interfacial value. Upper plate. Surface value. Dried region. Frozen region.
Superscripts 0

M M2/s N/m2 Dimensionless kJ/kg kJ/kg K kg of water/kg of solid N/m2 kg/m3 kW/(m2 K4) Initial value at time zero.

Subscripts e g in LP t w X UP 0 I II

REFERENCES LIAPIS, A. I. and SADIKOGLU, H. (1997) Mathematical Modeling of the Primary and Secondary Drying Stages of Bulk Solution Freeze-Drying in Trays: Parameter Estimation and Model Discrimination by Comparison of the Theoretical Results with Experimental Data. Drying Technology. Missouri, v. 15, n 3-4, p. 791-810. LIAPIS, A. I., BRUTTINI, R., PIKAL, M.J. (1996) Research and Development Needs and Opportunities in Freeze Drying. Drying Technology, v. 14, n. 6, p. 1265-1300. LICHTFIELD, R. J., LIAPIS, A I. (1979)An absorption-sublimation model for a freeze dryer. Chemical Engineering Science. Great Britain, v. 34, n. 9, p. 1085- 1090. MILLMAN, M. J., LIAPIS, A. I. and MARCHELLO, J. M. (1985) An Analysis of the Lyophilization Process Using a Sorption-Sublimation Model and Various Operational Policies. AIChE Journal. Missouri, v.31, n.10, p. 1594-1604. PLACKETT, R. L, BURMAN, J. P. The Design of Optimum Multifactorial Experiments. Biometrika. 1946, v.33, p.305-325.

484