Effect of mechanical energy on the energy efficiency of spouted beds applied on drying of sorghum [Sorghum bicolor (L) moench]

Accepted Manuscript Title: EFFECT OF MECHANICAL ENERGY ON THE ENERGY EFFICIENCY OF SPOUTED BEDS APPLIED ON DRYING OF SORGHUM [Sorghum bicolor (L) Moench] Authors: R.C. Brito, T.F. P´adua, J.T. Freire, R. B´ettega PII: DOI: Reference:

S0255-2701(16)30697-3 http://dx.doi.org/doi:10.1016/j.cep.2017.03.021 CEP 6957

To appear in:

Chemical Engineering and Processing

Received date: Revised date: Accepted date:

22-12-2016 28-3-2017 29-3-2017

Please cite this article as: R.C.Brito, T.F.P´adua, J.T.Freire, R.B´ettega, EFFECT OF MECHANICAL ENERGY ON THE ENERGY EFFICIENCY OF SPOUTED BEDS APPLIED ON DRYING OF SORGHUM [Sorghum bicolor (L) Moench], Chemical Engineering and Processinghttp://dx.doi.org/10.1016/j.cep.2017.03.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

EFFECT OF MECHANICAL ENERGY ON THE ENERGY EFFICIENCY OF SPOUTED BEDS APPLIED ON DRYING OF SORGHUM [Sorghum bicolor (L) Moench] R. C. BRITO1, T. F. PÁDUA 1, J. T. FREIRE 1 e R. BÉTTEGA1 1

Federal University of São Carlos, Department of Chemical Engineering

Rod. Washington Luiz - Km 235, P.O. Box 676, 13565-905, São Carlos, SP, Brazil. e-mail: [email protected] Graphical Abstract

HIGHLIGHTS    

Analysis of mechanical energy consumption inside a spouted bed dryer; Results for drying of sorghum considering several operational conditions; Statistical analysis of energy consumption and drying efficiency; Importance of accounting mechanical energy for global energy efficiency; ABSTRACT The energy consumption of spouted bed is an important issue that has been studied from various aspects. In order to investigate the energy consumption of the equipment, several experiments of drying of sorghum bicolor [Sorghum bicolor (L) Moench] were conducted and the results were evaluated looking for the energy performance of the spouted bed. A methodology that takes into account the mechanical energy necessary to maintain the spout regime on the global energy efficiency was proposed and the performance was determined based on the drying efficiency, energy efficiency, and specific energy consumption with and without the term corresponding to the

1

mechanical contribution. This methodology permits a more rigorous evaluation of the energy performance of drying equipment. The percentage increase of the specific energy consumption due to inclusion of the mechanical energy was greater at higher loads and at lower temperatures, demonstrating that this term represented a significant portion of the energy consumption of the process.

1.

INTRODUCTION

The search for renewable sources of energy has intensified due to increasing global concerns about the environment. In Brazil, ethanol is almost entirely produced from sugar cane. Nonetheless, sorghum [Sorghum bicolor (L) Moench] has been shown to have considerable potential for the production of ethanol, due to the large amount of sugar stored in its culms. One of its main advantages is the possibility of supplying the ethanol industry during the periods between the sugar cane harvests. For these reasons, studies of the preservation and storage of sorghum have become increasingly important [1,2]. Among the operations involved in preservation and storage, drying is a fundamental step because it reduces the moisture content of the material to suitable levels, slowing its biological activity and avoiding rapid deterioration, enabling prolonged storage without loss of physiological qualities [3]. However, the drying process can account for a significant fraction of the overall energy consumption, depending on the process and the technology used [4,5]. Between the various dryers available, the spouted bed provides high rates of heat and mass transfer, due to the rapid agitation and mixing of the solids, which is characteristic of the fluid dynamics of the system. As a result, the spouted bed is used in a wide range of chemical and physical processes including the drying of pastes and particulate solids [6,7], coating operations[8,9], chemical reactions [10] and pyrolysis [11,12]. The efficient fluid-particle contact provided by the spouted bed ensures high drying rates even at low temperatures, which is advantageous for the processing of thermosensitive materials such as seeds, whose physiological and nutritional qualities are compromised by exposure to high temperatures. In the spouted bed, the cyclic movement of the solids permits shorter periods of direct contact of the material with the drying air[13–15]. Various studies concerning the processing of grains and seeds using spouted beds can be found in the literature. Markowski et al. [16] dried barley in a conical-cylindrical spouted bed, observing the drying kinetics of the grain and estimating the diffusion coefficient as a function of humidity and temperature. Analysis was made of the influence of the conformation assumed by the grain, which had a major influence on determination of the effective diffusivity of the barley. Nascimento et al. [17] evaluated the fluid dynamics of foxtail millet seeds in a conical spouted bed, finding that a greater load led to increases in the values of the characteristic fluid dynamics parameters. Mussi et al. [7] investigated the drying of jambolan residues using a spouted bed, considering the influence of temperature and air velocity on the functional and nutritional properties of the final

2

product. The spouted bed provided satisfactory performance and it was found that despite a decrease in the anthocyanin content, the antioxidant capacity was maintained, irrespective of the values of the variables used in the drying. Chielle et al. [18] used a spouted bed to process papaya seeds, with evaluation of the effects of drying air conditions and time on oil production, concluding that changes in the operating conditions assisted oil extraction without compromising the fatty acids composition. Despite its advantages, the spouted bed has certain limitations that restrict its application on an industrial scale. Among these is the large quantity of energy required to break the packed bed and maintain the characteristic spouting regime. Several studies have addressed the energy consumption of convective dryers (such as the spouted bed) from the thermal perspective. For example, Jittanit et al. [19] compared the energy performance of a fluidized bed with that of a spouted bed for the drying of rice and wheat grains, finding that the spouted bed had a lower specific energy consumption, compared to the fluidized bed, for all the operating conditions and materials used. Jindarat et al. [20] analyzed the specific energy consumption during the drying of coffee beans, using two distinct spouted bed configurations, one conventional and one combined with microwaves, and found that the latter provided lower energy consumption. Golmohammadi et al. [21] dried whole rice grains in a fluidized bed, using an intermittent process in order to (among other objectives) optimize the process in terms of energy usage and identify an optimal drying strategy based on intermission periods and the drying temperature. The use of three intermission periods proved to be a reasonable strategy for the drying of whole rice grains, while the use of intermittence considerably reduced the energy consumption of the process. Nazghelichi et al. [22] analyzed the influence of operating conditions on the energy consumption for drying carrot cubes in a fluidized bed, observing that increases of temperature and the height of the static bed increased energy use, in contrast to the influence of the size of the carrot cubes, with the drying of larger solids resulting in lower energy consumption. In the studies mentioned above, no consideration of the mechanical energy expenditure was included in the energy calculations. However, this term can account for a significant fraction of the total energy used, especially in the case of the spouted bed, where mechanical energy is used to overcome the resistance due to gravity and the downward flow of particles in the annular region, hence maintaining the cyclic movement and a high degree of mixing between the phases. Given the above background, the objective of this work was to analyze the drying of sorghum employing a conical spouted bed, using analysis of the fluid dynamics and kinetics of drying in order to identify the optimal operational conditions, with the main evaluation criterion being the energy consumption of the process. The energy performance of the spouted bed was determined based on the drying efficiency, the energy efficiency, and the specific energy consumption. A comparison was made between the results obtained using the parameters typically employed in spouted bed energy analysis, based only on the thermal energy supplied and consumed in the process, and using the same parameters with addition of a term corresponding to the mechanical energy

3

required. The proposed method offers a different way to evaluate the energy performance of the spouted bed. 2.

2.1

MATERIALS AND METHODS

Humidification of the seeds

In order to standardize the initial moisture content, and considering that the sorghum had already been submitted to a drying process prior to purchase and had an average initial moisture content of 12.89±0.02% (wet basis), a prior humidification procedure was necessary. This consisted of placing 200 g of sorghum together with 50 mL of water in polyethylene packages, which were then kept refrigerated at 4 °C for 24 h. The amount of water used was sufficient to achieve a mean humidity of 30% (wet basis), with all the water being absorbed by the grains during the contact period.

2.2

Experimental unit

The fluid dynamics and drying experiments were carried out in the equipment shown in Figure 1. In figure 1, the by-pass system was used to adjust the air flow rate from the 7.5 HP blower. The air flow rate was measured by a Venturi type flow meter. Four resistances of 1000W composed the heater that were controlled by a Flyever FE50SN temperature controller. The drying chamber (spouted bed) consisted of a cylindrical stainless steel column, 60 cm high and 30 cm in diameter, with a 60° conical base of 23 cm in height.

2.3

Experimental procedure

2.3.1

Fluid dynamics characterization

Characterization of the fluid dynamics of the equipment illustrated in Figure 1 was based on the methodology proposed by Mathur and Epstein [24]. Three different sorghum loads (2, 3, and 4 kg) were used, based on the dimensions of the equipment, ensuring that the configuration of the conical spouted bed was maintained. During the procedure, measurements were recorded of the pressure drop and air flow in the equipment, using a Lynx ADS0500 data acquisition system and a program developed in LabVIEW® (National Instruments). Sampling frequencies of 500 Hz were used to obtain the mean pressure drop under each experimental condition, enabling the construction of characteristic fluid dynamics curves and determination of the minimum spout velocity(𝑢𝑚𝑗 ), maximum pressure drop(∆𝑃𝑚 ), and the stable spout pressure drop(∆𝑃𝑚 ). These results were used to define the operational conditions of the drying experiments.

4

Drying process experimental design

2.3.2

The effects of temperature and load on the energy performance of the drying process were investigated using a face-centered central composite design (FCC) with two levels, three replicates at the central point, and four axial points, totaling 11 experiments. The levels of the analyzed variables were established in preliminary tests, using loads analogous to those employed in the fluid dynamics characterization experiments. The coded and real values of the variables used in the design are shown in Table 1. Statistical evaluation of the results obtained was performed using Statistica® v. 7.0 software. The responses (dependent variables used in experimental design) were the drying efficiency and the specific energy consumption, in both cases taking into consideration the mechanical energy required in the process. Table 1: Operating conditions used in the drying process. Experiment

2.3.3

Load (kg)

Temperature (°C)

1

2 (-1)

50 (-1)

2

4 (+1)

50 (-1)

3

3 (0)

50 (-1)

4

2 (-1)

70 (+1)

5

4 (+1)

70 (+1)

6

3 (0)

70 (+1)

7

2 (-1)

60 (0)

8

4 (+1)

60 (0)

9

3 (0)

60 (0)

10

3 (0)

60 (0)

11

3 (0)

60 (0)

Drying process

Prior to each drying process under the different operating conditions, three samples were collected for gravimetric determination of the initial moisture content of the material. The samples were weighed using an analytical balance and then dried in an oven at 105 °C for 24 h, followed by reweighing. After adjustment of the process operating conditions (Table 1) and stabilization of the system, the drying chamber was loaded with the desired amount of sorghum. Samples were removed for determination of moisture content as a function of time, at intervals of 5, 15, and 30 min during an 8 h process. The collections were all made through the top of the equipment, removing relatively small amounts of material so as not to disturb the bed. The moisture contents of samples collected (including the three initial samples) were determined by weight loss after

5

24h in an oven at 105°C. During the drying process, the drying air temperature, the air flow rate, and the pressure drop in the bed were determined using the same measurement system employed in the fluid dynamics characterization. Software developed in LabVIEW was used to monitor the drying process. The kinetics was described, as a function of time, using the dimensionless moisture content (𝑋 ∗ ), calculated using the following equation:

𝑋 ∗ (𝑡) =

𝑋̅𝑡 − 𝑋𝑒𝑞 𝑋 𝑖 − 𝑋𝑒𝑞

(1)

where 𝑋̅𝑡 is the mean moisture at time t, 𝑋 𝑖 is the initial moisture, and 𝑋𝑒𝑞 is the moisture at dynamic equilibrium (all calculated on a dry basis).

2.4

Energy analysis

The energy analysis was performed using equations from the literature [14,20,25]. Energy loss was neglected in energy analysis. Calculation of the energy required to heat the material (𝑄𝑚 ) was based on the product of the wet sorghum load, its specific heat, and the difference between the initial and final temperatures of the material, according to Equation 2:

𝑄𝑚 = 𝑚𝑤𝑠 𝑐𝑝𝑠 (𝑇𝑠,𝑡 − 𝑇𝑠,𝑖 )

(2)

where 𝑚𝑤𝑠 is the mass of wet sorghum, 𝑐𝑝𝑠 is the specific heat of sorghum, 𝑇𝑠,𝑡 and 𝑇𝑠,𝑖 are the mean temperature of sorghum at time t and initial, respectively. The energy required to evaporate the water present in the sample (𝑄𝑤 ) was estimated based on the latent heat of vaporization of water and the mass of water evaporated:

𝑄𝑤 = ∆𝐻𝑣,𝑠 𝑚𝑑𝑠 (𝑋̅𝑡 − 𝑋𝑖 )

(3)

6

where 𝑚𝑑𝑠 is the mass of dry sorghum and ∆𝐻𝑣,𝑠 is the latent heat of vaporization of sorghum. The specific heat of the sorghum was calculated as a function of the change in the moisture content, using the following expression [26]:

̅𝑡 𝑐𝑝𝑠 = 1.390 + 0.0322𝑀

(4)

̅𝑡 is the mean moisture at time t in wet basis. where 𝑀 Determination of the latent heat of vaporization required taking account of the fact that hygroscopic materials such as sorghum require a greater amount of energy to vaporize the moisture present, compared to the amount of energy required for the vaporization of free water, because some of the moisture is attached to the structure of the material. Hence, the following equation was used, which is specific for estimation of the latent heat of vaporization of sorghum and is a function of the moisture content (dry basis, decimal) and the grain temperature (°C) [25,26]:

𝛥𝐻𝑣,𝑤 = (2502.2 − 2.39 𝑇𝑠,𝑡 )[1 + 1.006 exp(−19.650𝑋̅𝑡 )]

(5)

The thermal energy supplied to the system (𝑄) was considered function of the difference between the temperature of the air at the entrance of the equipment and the ambient temperature, as follows:

𝑄 = 𝑚̇𝑐𝑝 (𝑇𝑑 − 𝑇𝑎 )

(6)

where 𝑚̇ is the mass flow rate of air, 𝑐𝑝 is the specific heat of air, 𝑇𝑑 and 𝑇𝑎 are the temperature of drying air and ambient, respectively. The spouted bed requires a large amount of mechanical energy to maintain the characteristic regime, both to overcome the parallel resistances caused by the upward flow of air in the jet region and the downward flow of the solids in the annular region, as well as to disrupt the packed bed. Hence, a term was added to represent this portion of energy in the calculations, coupling the pressure drop and the kinetic energy involved in the process, and ignoring the potential energy. A mechanical energy balance was applied, with the drying chamber as the control volume, resulting in the following expression:

7

𝑊𝑚 = 𝑚̇ (

∆𝑣 2 ∆𝑃 + ) 2 𝜌

(7)

where 𝑊𝑚 represents the mechanical energy required to maintain the characteristic flow regime, ∆𝑣 is the variation of air velocity in the inlet and outlet position, ∆𝑃 is the pressure drop and 𝜌 is the specific mass of air. The parameters used to analyze the energy performance of the spouted bed were the energy efficiency (𝐸𝐸), the drying efficiency (𝐷𝐸), and the specific energy consumption (𝑆𝐸𝐶). The energy efficiency was calculated using Equations 8 and 9, considering the ratio between the energy required to evaporate the water and the total energy supplied to the system. Two different forms of the total energy supplied were considered, the thermal energy alone (Equation 8), and the combined thermal and mechanical energy (Equation 9).

𝑄𝑤 𝑄. 𝑡

𝐸𝐸 =

𝐸𝐸 ∗ =

𝑄𝑤 (𝑄 + 𝑊𝑚 )𝑡

(8)

(9)

In addition to the energy required to evaporate the water, the drying efficiency also includes (in the term referring to the consumption) the energy required to heat the material, as estimated using Equations 10 and 11. In the case of the total energy, a distinction analogous to that used for the energy efficiency was made.

𝐷𝐸 =

𝑄𝑚 + 𝑄𝑤 (𝑄)𝑡

𝐷𝐸 ∗ =

𝑄𝑚 + 𝑄𝑤 (𝑄 + 𝑊𝑚 )𝑡

(10)

(11)

8

Finally, Equations 12 and 13 were used to obtain the specific energy consumption. As in the calculations of energy and drying efficiency, a parameter was used that considered only the thermal energy, together with another parameter that represented the sum of the thermal and mechanical energies.

𝑆𝐸𝐶 =

𝑆𝐸𝐶 ∗ =

(𝑄)𝑡 𝑚𝑑𝑠 (𝑋̅𝑡 − 𝑋𝑖 )

(𝑄 + 𝑊𝑚 )𝑡 𝑚𝑑𝑠 (𝑋̅𝑡 − 𝑋𝑖 )

3.

RESULTS AND DISCUSSION

3.1

Fluid dynamics

(12)

(13)

The results of characterization of the fluid dynamics of the process are illustrated in Figure 2. The tests were performed in triplicate using sorghum loads of 2, 3, and 4 kg.

9

600

600

550 500 450

550

(a)

4 kg (experiment 1) 4 kg (experiment 2) 4 kg (experiment 3)

500 450 400

Pressure drop (Pa)

400

Pressure drop (Pa)

2 kg (Forwards) 3 kg (Forwards) 4 kg (Forwards) 2 kg (Backwards) 3 kg (Backwards) 4 kg (Backwards)

350 300 250 200

350 300 250 200

150

150

100

100

50

50

0 0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

(b)

0 0.000

0.005

0.010

Flow rate (m³/s)

0.015

0.020

0.025

0.030

0.035

Flow rate (m³/s)

Figure 2: (a) Pressure drop in the spouted bed as a function of the air flow for the 4 kg loads in experiments 1, 2 and 3; (b) Pressure drop in the spouted bed as a function of the air flow for the 2, 3 and 4 kg loads.

It can be seen from Figure 2(a) that the curves showed the qualitative behavior typical of a spouted bed, as described by Mathur and Epstein [24]. In terms of the influence of the load on the fluid dynamics, shown in Figure 2(b), there were increases in the maximum pressure drop and the minimum spout flow as the sorghum loading was increased, as expected. This effect was due to the greater flow resistance caused by the increased load, resulting in a greater pressure drop and a higher minimum spout velocity. Mathur and Epstein [24] reported that the pressure drop in the bed is directly proportional to the height of the static bed divided by its cross-sectional area. The quantitative influence of the load can be seen more clearly in Table 2, which shows the values of the fluid dynamics parameters estimated from the characteristic curves.

Table 2: Data of the parameters obtained from the fluid dynamics tests. Load (kg)

ΔPm (Pa)

umj (m.s-1)

ΔPs (Pa)

2

358.36 ± 21.95

21.66 ± 0.06

108.17 ± 19.86

10

3

443.12 ± 46.23

25.44 ± 0.34

156.64 ± 7.12

4

510.30 ± 9.13

28.55 ± 0.55

171.71 ± 11.76

Similar results concerning the influence of the bed height were obtained by Nascimento et al. [17] and Xavier et al. [12], who analyzed the fluid dynamics of foxtail millet and macadamia seeds, respectively, in conical spouted beds. Another feature that could be observed from the fluid dynamics curves was that after a certain flow rate in the region corresponding to a decrease in pressure of the stable spout, where an increase of the flow should not, in principle, significantly influence the pressure drop, the curves presented slight decreasing pressure drop trends. This could be explained by the particular characteristics of the conical configuration, whereby the airflow at the bed inlet can be increased substantially over a wide operational range without losing the characteristic cyclic movement of the solids. As a result, the process shifts from a stable spouted state to a transition state, where the pressure drop begins to decrease until reaching a state known as a jet spouted bed, where once again the flow increase does not have any significant effect on the pressure drop. The jet spouted bed is characterized by a high flow rate and expansion of the annular region, resulting in increased porosity [28–31].

3.2 Drying kinetics

Figure 3 shows the curves for the decrease of moisture content (dimensionless) as a function of time, under the operating conditions employed.

11

2 kg 3 kg 4 kg

1.0

0.8

(b) dimensionless moisture (-)

dimensionless moisture (-)

(a)

0.6

0.4

0.2

50°C 60°C 70°C

1.0

0.8

0.6

0.4

0.2

0.0

0.0 0

100

200

300

400

500

0

100

time (min)

200

300

400

500

time (min)

Figure 3: (a) Dimensionless moisture as a function of the drying time for temperatures of 70°C, the load being varied in 2, 3 e 4 kg; (b) Dimensionless moisture as a function of the drying time for loads of 4 kg, the temperature being varied in 50, 60 and 70°C.

It can be seen from Figure 3 that for all the operational conditions evaluated, there were marked decreases in moisture content during the initial phases of drying. This was indicative of the removal of free water under a high mass transport gradient. The curves generally showed behavior typical of the predominance of internal mass transfer mechanisms. Such behavior is characterized by an exponential decrease of moisture as a function of time, suggesting the existence of a period of decreasing mass transfer rate. The highly heterogeneous nature of the porous medium, which provided obstacles to the flow of moisture, together with the fact that the internal moisture was likely to be strongly associated with the structure of the material, resulted in a continuous and marked decrease in the drying rate. The influence of temperature is shown in Figure 3(b), from which it can be seen that a higher temperature resulted in a shorter drying time in order to reach the same moisture content, due to the greater amount of thermal energy supplied to the system. The influence of the load is shown in Figure 3(a), where the curves are almost superimposed throughout the drying process, up to 480 min, and tend towards a dynamic equilibrium. An increase in the load acted to

12

reduce the circulation rate of the solids, potentially hindering the removal of moisture and slowing the drying rate. However, two factors acted to minimize this effect, providing a possible explanation for the observed results. One was the predominance of internal mass transfer mechanisms. Since the process was controlled by the internal resistance, the characteristics of the medium exerted little influence on the drying rate, which was mainly determined by the characteristics of the porous structure of the material. The other factor was the high degree of agitation of the particles in the spouted bed, which promoted high rates of heat and mass transfer throughout the active volume, hence maintaining the temperature relatively uniform during the drying process, minimizing the influence of the load.

3.3 Energy and drying efficiency

The drying and energy efficiency results for experiments 4, 5, and 6 are presented in Figure 4, which show the variation of the parameters over time during the processes.

1.0

1.0

(a)

Drying efficiency

0.8

0.6

0.4

0.2

4 kg 70 °C (EE*) 3 kg 70 °C (EE*) 2 kg 70 °C (EE*) 4 kg 70 °C (EE) 3 kg 70 °C (EE) 2 kg 70 °C (EE)

(b) 0.8

Energy efficiency

4 kg 70 °C (DE*) 3 kg 70 °C (DE*) 2 kg 70 °C (DE*) 4 kg 70 °C (DE) 3 kg 70 °C (DE) 2 kg 70 °C (DE)

0.6

0.4

0.2

0.0

0.0 0

100

200

300

400

500

0

100

time (min)

200

300

400

500

time (min)

Figure 4: (a) Drying efficiency as a function of time and (b) Energy efficiency as a function of time for temperatures of 70°C and loads of 2, 3 e 4 kg.

13

The behavior of the curves (which were similar for all the other operating conditions used) reflected an exponential decrease in the drying efficiency over time. The drying and energy efficiencies decreased markedly from the start of the process up to approximately 100 min, after which the decreases became less pronounced. At the start of the process, the high moisture content meant that the energy supplied to the system was essentially used to evaporate the water present in the material that was not strongly bound to the structure of the solid. As the moisture content decreased, its removal became increasingly difficult because the water was intimately associated with the porous structure of the material. Hence, some of the energy supplied was used to heat the solid and physically dissociate the water, resulting in lower energy efficiency. Due to these factors, high drying and energy efficiencies were observed in the initial stages of the process. The efficiency was therefore mainly influenced by the moisture content of the material and was proportional to the drying rate of the process. The explanation for the behavior observed in Figure 4 is supported by the findings of previous studies that have analyzed the energy used during the drying process [22,32,33]. In the drying of carrots using a fluidized bed, Nazghelichi et al. [22] observed that the amount of energy used was greater at the start of the process and decreased continuously with the loss of moisture. Hence, assuming that such behavior applied in the present case, the findings of Nazghelichi et al. [22] were in agreement with those obtained in this work. During the drying of paper, Kudra [4] observed that the drying and energy efficiencies showed a period of stabilization around the maximum values, subsequently decreasing rapidly. Curves plotted of the decrease in moisture as a function of time revealed that the stabilization period corresponded exactly to the period during which the process was controlled by the external mass transfer resistance and the drying rate was constant. The decrease started at the point where the internal mass transfer resistance began to limit the process, causing the drying rate to decrease. In the present work, the period of stability reported by Kudra [4] for drying and energy efficiency was not observed. This was in agreement with the fact that the kinetic curves for the drying process did not show any evidence of a linear decrease in moisture over time, which would be a typical characteristic of a period with a constant drying rate. It can be seen from Figure 4 that as the load was increased, the drying and energy efficiencies also increased, for all the operational conditions analyzed. Hence, the processes performed with 4 kg loads presented the highest drying and energy efficiencies. In explaining the influence of the load, it should be noted that the kinetic curves of the drying process indicated a non-significant influence of the load. Therefore, it appears that as the load was increased, the removal of a greater amount of water resulted in better energy utilization, hence increasing efficiency. When the term corresponding to the mechanical energy was included in the calculations of drying and energy efficiency, lower efficiencies were obtained, compared to use of the standard parameters. When the

14

mechanical energy supplied to the system to maintain the spouted regime was considered in the energy calculation, there was an increase of the term corresponding to the energy supply. However, the consumption term remained unaltered, since this term was based only on the energy directly involved in heating the material and evaporating the water (the thermal energy), while the mechanical energy was used to overcome the resistances due to gravity, friction, and the descending flow of particles, hence maintaining the movement of the material in the equipment. Based on the characteristics of the spouted bed, where the mechanical energy is predominant, energy analysis without consideration of this type of energy would lead to overestimation of the drying efficiency. Evaluation of the influence of temperature on the energy and drying efficiencies was performed using a load of 4 kg submitted to drying with air at 50, 60, and 70 °C. The results are shown in Figure 5.

1.0

4 kg 50 °C (DE*) 4 kg 60 °C (DE*) 4 kg 70 °C (DE*) 4 kg 50 °C (EE*) 4 kg 60 °C (EE*) 4 kg 70 °C (EE*)

Drying and energy efficiency

0.8

0.6

0.4

0.2

0.0 0

100

200

300

400

500

time (min)

Figure 5: Drying and energy efficiency as a function of time for loads of 4 kg and temperatures of 50, 60 e 70°C.

The highest drying and energy efficiencies were achieved for the process conducted at 50 °C. No significant differences were observed between the efficiencies obtained at 60 and 70 °C.

15

These observations can be explained by the fact that with an increase in the drying air temperature, there was also an increase in the amount of thermal energy supplied to the system. Another effect was that at a higher temperature, there was a greater temperature gradient between the air and the wall of the equipment, resulting in increased heat loss. As a result, the heat lost from the system to the environment through the bed wall was greater, compared to use of air at a lower temperature, resulting in reduced efficiency of energy utilization in the drying process. In addition, increasing the temperature of the drying air and, consequently, the amount of thermal energy supplied, resulted in poorer utilization of this energy in the process, because a larger fraction of this energy was lost together with the heated air that exited the top of the equipment. This is a characteristic of the spouted bed, since most of the heated gaseous phase flows through the spout region, where the heat transfer coefficient was higher, compared to the annular region. As a result, most of the heat transfer occurs in the spout region, with the annular region representing a reservoir of the heat exchanged between the drying air and the solids. However, the contact time of the material with the drying air is shorter in the spout region, compared to the annular one, so that a smaller amount of energy is actually employed in the heat transfer, while the other part of this energy is lost from the top of the equipment [31]. It can therefore be seen that despite the higher drying rates provided by the air at higher temperature, which reduced the processing time, use of a higher temperature also increased the thermal losses of the process. As a result, a smaller portion of the total energy supplied to the system was effectively employed in the drying process. The values obtained for the energy efficiency were lower, compared to the drying efficiency, for the 4 kg load at all three temperatures analyzed. This was because in the consumption term, the energy efficiency only considered the energy effectively used to evaporate water, ignoring the portion of energy used to heat the material. Given the characteristics of sorghum, which is a biological material with an irregular porous structure, as well as the results obtained, which confirmed the predominance of internal mass transfer mechanisms, the energy efficiency was not the most suitable parameter for energy analysis of the process. This was because in processes governed by the internal resistance to mass transfer, where the drying rate is low, a significant fraction of the energy supplied to the system is used to heat the material. Therefore, failure to consider this energy consumption leads to a lower apparent efficiency, because not all the energy actually consumed in the process is considered. In a real drying operation, the process needs to be continued until the product specification is reached, implying different total drying times for different drying air temperatures. Therefore, an analysis of the drying efficiency was performed for the operational conditions established, considering the time required for the moisture content to reach 12-13% (wet basis), which is generally considered suitable for storage of the material. The aim was to obtain a drying efficiency for the process as a whole, rather than a comparison of the values obtained under the different conditions after a fixed period. The results are shown in Table 3.

16

Table 3: Drying efficiency data obtained for a moisture range of 12-13%. DE* (%) Load (kg)

DE (%)

DE* (%)

70°C

DE (%)

DE* (%)

60°C

DE (%) 50°C

2

27.50

28.87

24.20

25.60

25.96

27.75

3

31.08

33.07

32.59

35.21

26.46

28.78

4

36.05

38.79

30.67

33.49

29.91

33.29

Different from the results for the drying efficiencies achieved considering the same time interval, the highest drying efficiency for a 4 kg load was obtained at 70 °C, due to the shorter time required for the moisture content to reach 12-13% (wet basis).

3.4 Specific energy consumption

The results obtained for the specific energy consumption are shown in Table 4 and Figure 6, for the operating conditions listed in Table 1. In this case, the time taken for the moisture content to reach 12-13% (wet basis) was also adopted in the specific energy consumption calculations.

Table 4: Specific energy consumption obtained for the established operating conditions. SEC*

SEC

SEC*

SEC

SEC*

SEC

(kJ/kg)

(kJ/kg)

(kJ/kg)

(kJ/kg)

(kJ/kg)

(kJ/kg)

Load (kg)

70°C

60°C

50°C

2

9892.64

9422.52

10999.03

10400.98

14514.01

13580.07

3

7618.23

7159.16

8235.56

7622.94

14019.20

12888.22

4

7357.59

6837.19

7902.93

7238.69

11019.87

9900.65

17

20000

SEC* (70°C) SEC (70°C) SEC* (60°C) SEC (60°C) SEC* (50°C) SEC (50°C)

18000 16000

SEC (kJ/kg)

14000 12000 10000 8000 6000 4000 2000 0 2

3

4

load (kg)

Figure 6: Variation of the Specific Energy Consumption (SEC e SEC*) for the different operating conditions employed.

For all the temperatures evaluated, the lowest specific energy consumption values were obtained for the processes with loads of 4 kg, for both the standard parameter and the proposed parameter with inclusion of the mechanical energy. This indicated that in addition to greater productivity in use of the equipment, a higher load in the dryer resulted in better use of energy in the drying process. The lower energy consumption values obtained with the higher loads could be explained from the drying kinetics curves, which indicated a negligible influence of the load, with close similarity of the curves and the drying times required for attainment of the same moisture content using different loads. The finding that a higher load led to lower specific energy consumption was also reported by Jindarat et al. [20] and Wachiraphansakul and Devahastin [32], who attributed the reduction to faster evaporation of water. When the mechanical energy was included in the consumption term, higher energy consumption was obtained, under all the operational conditions tested. These increases were estimated by calculation of the relative increases under the conditions employed, according to Equation 14. The values obtained are provided in Table 5.

18

𝑅𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑖𝑛𝑐𝑟𝑒𝑎𝑠𝑒 (%) =

𝑆𝐸𝐶 ∗ − 𝑆𝐸𝐶 . 100 𝑆𝐸𝐶

(14)

Table 5: Relative increase data (%) for the specific energy consumption.

Load (kg)

RELATIVE INCREASE IN SPECIFIC ENERGY CONSUMPTION (%) 70°C

60°C

50°C

2

4.99

5.75

6.88

3

6.41

8.04

8.77

4

7.61

9.18

11.30

For all three temperatures used, the relative increase was higher in the presence of a greater load. This can be explained assuming that a greater load increased the resistance to the flow and consequently increased the pressure drop. Hence, additional mechanical energy was required to maintain the spout regime, which increased the contribution of the mechanical energy term in the energy consumption calculation. Despite the clear predominance of thermal energy, the significance of mechanical energy increased with the mass of solids processed, highlighting the importance of including this term in energy calculations, especially at larger scales. On the other hand, this percentage decreased when a higher temperature was used (at the same load). Since the temperature did not have any significant influence on the fluid dynamics, its variation did not significantly affect the mechanical energy required for the process. However, a higher temperature meant that more thermal energy was supplied to the system, leading to an even greater significance of the thermal portion, relative to the mechanical energy, resulting in a reduction in the percentage contribution of mechanical energy. The lowest energy consumptions (using both the standard and proposed parameters) were obtained for the processes conducted at 70 °C. A trend could be observed whereby a higher temperature led to a decrease in the specific energy consumption. This could be attributed to the shorter drying time due to the faster rate of evaporation caused by the greater amount of thermal energy supplied to the system. The effects on specific energy consumption were more sensitive at temperatures of 50 and 60 oC, indicative of a tendency towards saturation with increasing temperature.

19

3.5 Statistical analysis of the experimental data

Statistical analysis of the experimental data was used to evaluate the effects of the operating conditions (temperature and load) on DE* and SEC*. The Pareto diagrams obtained for DE* and SEC* are shown in Figures 7 and 8, respectively, using a significance level of α = 0.10 due to the high variability of the drying process. In addition to the influence of the linear terms (L), the FCC applied enabled evaluation of the effects of the quadratic (Q) and interaction terms.

(2)Load(L)

(1)Temperature(L)

Load(Q)

Temperature X Load

Temperature(Q)

p=,1 Standardized Effect Estimate (Absolute Value)

Figure 8: Pareto diagram for specific energy consumption with a significance level of α=0.10. In the case of the drying efficiency, the linear terms for temperature and load were significant, with a greater influence of the load on the response. No significant influences on the drying efficiency were observed for the

20

temperature and load quadratic terms, or for the term representing the interaction between the independent variables. A model (Equation 15) was therefore obtained from the coded variables, correlating the drying efficiency with the significant terms (the linear terms for temperature and load) identified from the Pareto diagram.

𝐷𝐸 ∗ = 30.50 + 2.22𝑍 + 3.16𝑌

(15)

where 𝑍 and 𝑌 represents the coded temperature and coded load, respectively. The Pareto diagram for the specific energy consumption revealed a strong influence of the linear temperature term, relative to the other terms (which were also statistically significant, but to lesser extents). Considering all the statistically significant terms observed in the Pareto diagram, a model was constructed using the coded variables (Equation 16), relating the specific energy consumption to the temperature and the load.

𝑆𝐸𝐶 ∗ = 8621.37 − 2447.44Z + 1816.43𝑍 2 − 1520.88𝑌 + 448.69𝑌 2 + 239.77 𝑍 𝑌

(16)

Statistical evaluation of the models obtained for DE* and SEC* (Equations 15 and 16) was performed using analysis of variance (ANOVA), as shown in Tables 6 and 7.

Table 6: ANOVA for drying efficiency (DE*)

Source of variation

Sum of squares

Regression

85.19

Residual

Degrees of

Mean square

Fcalc

p-value

2

42.60

11.51

0.0044

29.61

8

3.70

Lack of fit

27.01

6

4.50

3.46

0.2412

Pure error

2.60

2

1.30

Total

114.80

10

freedom

R²=74.20%; Ftab(regression/residual) =3.11; Ftab(lack of fit/pure error)=9.33.

21

Table 7: ANOVA for specific energy consumption (SEC*)

Source of variation

Sum of squares

Regression

58916609

Residual

Degrees of

Mean square

Fcalc

p-value

5

11783322

22.16

0.0019

2659058

5

531811.5

Lack of fit

2622410

3

874137

47.70

0.0206

Pure error

36648

2

18324

Total

63440063

10

freedom

R²=95.81%; Ftab(regression/residual) =3.45; Ftab(lack of fit/pure error)=9.16.

The ANOVA results for the drying efficiency showed a coefficient of determination (R²) that indicated a good fit of the statistical model to the experimental data, considering that the drying process involved several uncontrolled external variables, such as the psychrometric state of the inlet air, which contributed to greater experimental variability. According to the literature [24,33] the calculated value of F for the regression should be about 3 times higher than the tabulated value, for the model to be considered statistically significant. This requirement was fulfilled, since Fcal(regression/residual) = 3.70Ftab(regression/residual). In addition, the model did not show any lack of fit, given that Fcalc(lack of fit/pure error) < Ftab(lack of fit/pure error). In the ANOVA corresponding to the specific energy consumption, both the R² value and the calculated F value for the regression satisfied the criteria, indicating that the model provided a satisfactory fit to the experimental data. However, analysis of F calculated for the lack of fit revealed that Fcalc(lack of fit/pure error) > Ftab(lack of fit/pure error),

showing that the model presented a lack of fit. However, this could also result from a low lack of fit and

a tendency towards zero of the pure error. Figure 9 shows the distribution of residuals for SEC*, from which it can be seen that despite the ANOVA results, the model was able to adequately represent the data, since the residuals were randomly distributed.

22

1000 800 600

Raw Residuals

400 200 0 -200 -400 -600 -800 -1000 6000

7000

8000

9000

10000

11000

12000

13000

14000

15000

16000

Observed Values

Figure 9: Distribution of residuals for SEC*.

The operating conditions that maximized efficiency and minimized energy consumption were evaluated by the construction of response surfaces based on the variables and levels analyzed, shown in Figures 10 and 11.

23

Figure 11: Response surface for specific energy consumption.

The results illustrated in the response surfaces were in agreement with the findings of the previous experiments, described above. In the case of the drying efficiency, the highest values corresponded to the higher loads and temperatures, with the lowest efficiency obtained at the extreme point corresponding to the lowest load and temperature, revealing the existence of a linear relationship. In the case of the specific energy consumption, the value decreased as the temperature and the load increased, with a tendency towards stabilization at temperatures between 63 and 69 °C and at loads between 3.8 and 4.2 kg.

4.

CONCLUSIONS

24

For all the operational conditions evaluated, there were substantial decreases in moisture content during the initial phase of the process. The temperature exerted a significant influence on the process. In contrast, there was no pronounced influence of solids loading on the drying kinetics. The energy and drying efficiencies showed exponential decreases over time, with similar qualitative behavior under all the operational conditions employed. The energy and drying efficiencies increased in line with the solids loading, in both cases with the highest values observed for the 4 kg load. In evaluation of the influence of temperature, considering the efficiencies based on the time required to reach a moisture content of 12-13% (wet basis) at the different temperatures used, higher efficiencies were obtained for the process performed at 70 °C, due to the shorter overall process time. The lowest specific energy consumptions was obtained in the processes with loads of 4 kg. A higher temperature resulted in lower specific energy consumption, with the lowest for the process performed using a 4 kg load and temperature of 70 ºC. The best operational conditions happened in higher loads and temperatures and the specific energy consumption revealed a tendency towards stabilization. Inclusion of the term corresponding to mechanical energy in determination of the drying and energy efficiencies resulted in lower efficiencies in comparison with the standard efficiency. The inclusion of the mechanical energy term in the calculation resulted in higher values, under all the conditions tested. The percentage increase of the specific energy consumption due to inclusion of the mechanical indicates that this term represented a significant portion of the energy consumption of the process.

NOMENCLATURE 𝑐𝑝

specific heat of solids

[kJ kg-1 K-1]

𝑐𝑝𝑠

specific heat of air

[kJ kg-1 K-1]

𝐷𝐸

drying efficiency

[-]

𝐸𝐸

energy efficiency

[-]

∆𝐻𝑣,𝑠

latent heat of vaporization of sorghum

[kJ kg-1]

25

̅ 𝑀

mean moisture of solids in wet basis

[kg kg-1]

𝑚̇

mass flow rate of air

[kg s-1]

𝑚𝑑𝑠

mass of dry solids

[kg]

𝑚𝑤𝑠

mass of wet solids

[kg]

∆𝑃

pressure drop

[Pa]

∆𝑃𝑚

maximum pressure drop

[Pa]

∆𝑃𝑠

stable spout pressure drop

[Pa]

𝑄

thermal energy supplied

[kJ s-1]

𝑄𝑚

energy required to heat the material

[kJ]

𝑄𝑤

energy required to evaporate the water

[kJ]

𝑆𝐸𝐶

specific energy consumption

[kJ kg-1]

𝑡

time

[s]

𝑇

temperature

[K]

𝑇𝑎

ambient temperature

[K]

𝑇𝑑

drying air temperature

[K]

𝑢𝑚𝑗

minimum spout velocity

[m s-1]

∆𝑣

variation of air velocity

[m s-1]

𝑊𝑚

mechanical energy

[kJ s-1]

𝑋∗

dimensionless moisture

[-]

𝑋̅

mean moisture in dry basis

[kg kg-1]

𝑋𝑖

initial moisture in dry basis

[kg kg-1]

𝑋𝑒𝑞

moisture at dynamic equilibrium in dry basis

[kg kg-1]

𝑌

coded load

[-]

𝑍

coded temperature

[-]

Subscripts 𝑖

initial

𝑠

solid

𝑡

time

Greek symbols 𝜌

specific mass of air

[kg m-3]

ACKNOWLEDGEMENT

26

Authors appreciate the financial support from CNPQ (454475/2014-4).

REFERENCES

[1]

I.A. Olukoya, D. Bellmer, J.R. Whiteley, C.P. Aichele, Evaluation of the environmental impacts of ethanol

production

from

sweet

sorghum,

Energy

Sustain.

Dev.

24

(2015)

1–8.

doi:10.1016/j.esd.2014.10.004. [2]

M. Wang, Y. Chen, X. Xia, J. Li, J. Liu, Energy efficiency and environmental performance of bioethanol production from sweet sorghum stem based on life cycle analysis, Bioresour. Technol. 163 (2014) 74–81. doi:10.1016/j.biortech.2014.04.014.

[3]

M.A.S.A.S. Barrozo, A. Mujumdar, J.T.T. Freire, Air-Drying of Seeds: A Review, Dry. Technol. 32 (2014) 1127–1141. doi:10.1080/07373937.2014.915220.

[4]

T. Kudra, Energy aspects in drying, Dry. Technol. 22 (2004) 917–932. doi:10.1081/DRT120038572.

[5]

T. Kudra, Energy Performance of Convective Dryers, Dry. Technol. 30 (2012) 1190–1198. doi:10.1080/07373937.2012.690803.

[6]

A.D.A. Araújo, R.M.D. Coelho, C.P.M.L.C.P.M.L. Fontes, A.R.A. Silva, J.M.C.J.M.C. Da Costa, S. Rodrigues, A.D.A. Ara??jo, R.M.D. Coelho, C.P.M.L.C.P.M.L. Fontes, A.R.A. Silva, J.M.C.J.M.C. Da Costa, S. Rodrigues, Production and spouted bed drying of acerola juice containing oligosaccharides, Food Bioprod. Process. 94 (2015) 565–571. doi:10.1016/j.fbp.2014.08.005.

[7]

L.P. Mussi, A.O. Guimarães, K.S. Ferreira, N.R. Pereira, Spouted bed drying of jambol??o (Syzygium cumini) residue: Drying kinetics and effect on the antioxidant activity, anthocyanins and nutrients contents, LWT - Food Sci. Technol. 61 (2015) 80–88. doi:10.1016/j.lwt.2014.11.040.

[8]

N. Ali, T. Al-Juwaya, M. Al-Dahhan, An advanced evaluation of spouted beds scale-up for coating TRISO nuclear fuel particles using radioactive particle tracking (RPT), Exp. Therm. Fluid Sci. (2016). doi:10.1016/j.expthermflusci.2016.08.002.

[9]

M.G.. Vieira, S.C.. Rocha, Influence of the liquid saturation degree on the fluid dynamics of a spouted-bed

coater,

Chem.

Eng.

Process.

Process

Intensif.

43

(2004)

1275–1280.

doi:10.1016/j.cep.2003.12.004. [10]

D. Liu, E.P.L. Roberts, A.D. Martin, S.M. Holmes, N.W. Brown, A.K. Campen, N. de las Heras,

27

Electrochemical regeneration of a graphite adsorbent loaded with Acid Violet 17 in a spouted bed reactor, Chem. Eng. J. 304 (2016) 1–9. doi:10.1016/j.cej.2016.06.070. [11]

J. Alvarez, M. Amutio, G. Lopez, J. Bilbao, M. Olazar, Fast co-pyrolysis of sewage sludge and lignocellulosic biomass in a conical spouted bed reactor, Fuel. 159 (2015) 810–818. doi:10.1016/j.fuel.2015.07.039.

[12]

T.P. Xavier, B.P. Libardi, T.S. Lira, M.A.S. Barrozo, Fluid dynamic analysis for pyrolysis of macadamia shell in a conical spouted bed, Powder Technol. 299 (2016) 210–216. doi:10.1016/j.powtec.2016.05.034.

[13]

S. Sahin, G. Sumnu, F. Tunaboyu, Usage of solar-assisted spouted bed drier in drying of pea, Food Bioprod. Process. 91 (2013) 271–278. doi:10.1016/j.fbp.2012.11.006.

[14]

M.L. Passos, A.S. Mujumdar, G. Vijaya, V.G.S. Raghavan, Spouted and Spout-Fluidized Beds for Gram Drying, 1989. doi:10.1080/07373938908916621.

[15]

W.S.

Peterson,

Spouted

bed

drier,

Can.

J.

Chem.

Eng.

40

(1962)

226–230.

doi:10.1002/cjce.5450400512. [16]

M. Markowski, I. Białobrzewski, A. Modrzewska, Kinetics of spouted-bed drying of barley: Diffusivities

for

sphere

and

ellipsoid,

J.

Food

Eng.

96

(2010)

380–387.

doi:10.1016/j.jfoodeng.2009.08.011. [17]

L.D. do Nascimento, L.G. Corumbá, S.C. dos S. Rocha, O.P. Taranto, C.M.L. Costa, L.J.G. de Faria, Fluid-dynamics evaluation in a conical spouted bed and characterization of foxtail millet seeds, Particuology. 23 (2015) 1–7. doi:10.1016/j.partic.2015.01.009.

[18]

D.P. Chielle, D.A. Bertuol, L. Meili, E.H. Tanabe, G.L. Dotto, Spouted bed drying of papaya seeds for oil production, LWT - Food Sci. Technol. 65 (2016) 852–860. doi:10.1016/j.lwt.2015.09.022.

[19]

W. Jittanit, G. Srzednicki, R.H. Driscoll, Comparison Between Fluidized Bed and Spouted Bed Drying for Seeds, Dry. Technol. 31 (2013) 52–56. doi:10.1080/07373937.2012.714827.

[20]

W. Jindarat, S. Sungsoontorn, P. Rattanadecho, Analysis of Energy Consumption in a Combined Microwave–hot Air Spouted Bed Drying of Biomaterial: Coffee Beans, Exp. Heat Transf. 28 (2014) 107–124. doi:10.1080/08916152.2013.821544.

[21]

M. Golmohammadi, M. Assar, M. Rajabi-Hamaneh, S.J. Hashemi, Energy efficiency investigation of intermittent paddy rice dryer: Modeling and experimental study, Food Bioprod. Process. 94 (2015) 275–283. doi:10.1016/j.fbp.2014.03.004.

28

[22]

T. Nazghelichi, M.H. Kianmehr, M. Aghbashlo, Thermodynamic analysis of fluidized bed drying of carrot cubes, Energy. 35 (2010) 4679–4684. doi:10.1016/j.energy.2010.09.036.

[23]

J.T. Perazzini, M.T.B, Freire, F.B., Freire, Influence of bed geometry on the drying of skimmed milk in a Spouted Bed, (2015) 447–460.

[24]

K.B. Mathur, N. Epstein, Spouted Beds, Academic Press, New York, 1974.

[25]

M.G. a. Vieira, L. Estrella, S.C.S. Rocha, Energy Efficiency and Drying Kinetics of Recycled Paper Pulp, Dry. Technol. 25 (2007) 1639–1648. doi:10.1080/07373930701590806.

[26]

D.B. Brooker, F.W. Bakker-Arkema, C.W. Hall, Drying and Storage of Grains and Oilseeds, Van Nostrand Reinhold, New York, 1992.

[27]

S. Cenkowski, D.S. Jayas, D. Hao, Latent heat of vaporization for selected foods and crops, Can. Agric. Eng. 34 (1992).

[28]

M. Olazar, G. Lopez, H. Altzibar, J. Bilbao, Modelling batch drying of sand in a draft-tube conical spouted bed, Chem. Eng. Res. Des. 89 (2011) 2054–2062. doi:10.1016/j.cherd.2011.01.012.

[29]

S. Devahastin, R. Tapaneyasin, A. Tansakul, Hydrodynamic behavior of a jet spouted bed of shrimp, J. Food Eng. 74 (2006) 345–351. doi:10.1016/j.jfoodeng.2005.03.012.

[30]

A.. Markowski, W. Kaminski, Hydrodynamic characteristics of jet-spouted beds, Can. J. Chem. Eng. 61 (1983) 377–381. doi:10.1002/cjce.5450610318.

[31]

M.J.S. José, S. Alvarez, A.O. De Salazar, M. Olazar, J. Bilbao, Influence of the particle diameter and density in the gas velocity in jet spouted beds, Chem. Eng. Process. Process Intensif. 44 (2005) 153–157. doi:10.1016/j.cep.2004.03.011.

[32]

E.K. Akpinar, Energy and exergy analyses of drying of red pepper slices in a convective type dryer, Int.

Commun.

Heat

Mass

Transf.

31

(2004)

1165–1176.

doi:10.1016/j.icheatmasstransfer.2004.08.014. [33]

O. Corzo, N. Bracho, A. Vásquez, A. Pereira, Energy and exergy analyses of thin layer drying of coroba slices, J. Food Eng. 86 (2008) 151–161. doi:10.1016/j.jfoodeng.2007.05.008.

[34]

H.A. Becker, H.R. Sallans, Drying wheat in a spouted bed, Chem. Eng. Sci. 18 (1961) 97–112.

[35]

S. Wachiraphansakul, S. Devahastin, Drying kinetics and quality of okara dried in a jet spouted bed of sorbent particles, LWT - Food Sci. Technol. 40 (2007) 207–219. doi:10.1016/j.lwt.2005.11.010.

[36]

G.E.P. Box, J.S. Hunter, W.G. Hunter, Statistics for Experimenters: Design, Innovation and

29

Discovery, 2 Ed., John Wiley & Sons, Hoboken, New Jersey, 2005.

Figure 1: (1) blower; (2) by-pass system; (3) Venturi type flow meter; (4) heater; (5) temperature controller; (6) Drying chamber; (7) data acquisition system (adapted of PERAZZINI, et al. [23]).

Figure 7: Pareto diagram for drying efficiency with a significance level of α=0,10.

30

(1)Temperature(L)

(2)Load(L)

Temperature(Q)

Load(Q)

Temperature X Load

p=,1 Standardized Effect Estimate (Absolute Value)

Figure 10: Response surface for drying efficiency

31

32