In this study, a pilot-scale countercurrent hot air rotary dryer was used to reduce the moisture content of the sugarcane filter cake. Rotary dryers allow the use of bulk feedstock at high temperatures with short drying times. The response surface methodology was used to optimize and evaluate the effect of hot air temperature and rotational speed on drying efficiency, based on the central composite design. Various temperatures (70, 100 and 130 ℃) and rotation speeds (1, 3 and 5 RPM) were investigated. High drying efficiency was achieved by using high hot air temperatures and low rotational speeds. The error between the model and the experiment was less than 5%. Additionally, seven mathematical drying kinetic models were investigated to describe the drying characteristics of sugarcane filter cake samples. The best-fitted model for the three temperatures considered was the Page model, as assessed by the coefficient of determination (R2), average relative error percentage (P), reduced chi-square (ꭓ2), and root mean square error (RMSE) (R2 = 0.978-1.00). The results demonstrate that the optimization models may be considered for application in an industrial rotary dryer at 70–130 ℃, using the exhausted gas from a power plant stack.
Article
Drying process optimization and kinetic study of sugarcane filter cake using a pilot-scale rotary dryer
https://doi.org/10.21203/rs.3.rs-2606336/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
drying models
filter cake
response surface methodology
rotary dryer
sugarcane
Fossil fuels are currently the primary global energy source; however, the impact of these fuels raises significant environmental problems and resulted in an energy crisis. Therefore, it is necessary to focus on the use of renewable energy to alleviate these issues. Biomass is a primary source of renewable energy, particularly for waste from the sugarcane industry. Thailand currently has 57 sugarcane mills in operation, with an estimated cane output of 74.89 million tons in 2019/2020, which qualifies it as the fourth largest sugarcane producer in the world [1, 2]. Each ton of sugarcane sent to a factory is composed of ~ 104 kg (13%) of sugar, 430 kg (52%) of liquid waste, 231 kg (28%) of bagasse, 26 kg (3%) of molasses and 33 kg (4%) of filter cake [3–5]. Although most by-products, such as bagasse, are used in most sugarcane mills, filter cake is rarely studied and has generated substantial environmental problems [6]. The filter cake is a manufacturing by-product produced from the sugar clarifier process with a moisture content of ~ 70–75%. About 3 million tons of sugarcane filter cake are transported through conveyor belts and dumped into free space, representing a significant problem in waste management.
Filter cake is a black and solid sludge contaminated from sugarcane juice, and fine bagasse with a heating value of ~ 11.71 MJ/kg [5]. It is usually used as a fertilizer source and soil conditioner [7–10], animal feed [11, 12], adsorbent [13], cement [3, 14–16], biogas [17] or returned to cane fields. According to Thailand’s Ministry and Industry [1], there is significant potential in using filter cake as supplemental biomass feedstock in biomass power plants, which usually use bagasse as a sole feedstock. The typical power plants that use bagasse as a sole feedstock can only operate for a maximum of six months to a year. The dried filter cake can be supplied to the operation of a power plant to extend its electricity generation by approximately two to three months, resulting in more economical profitability [5]. The waste heat from the power plant stack, ~ 130 ºC, can be used for the drying process.
Most of the moisture content of the filter cake should be removed before further processing. Tray dryers have been used to predict the drying efficiency of sugarcane bagasse [18–20]. However, rotary dryers are more suitable for continuous processes [21]. The hot air is continuously flowed in the countercurrent direction [22] and a set of lifting flights on the drum's wall lift and enhance heat to the mass [23]. The drying temperature and the rotation speed are the main parameters in the rotary dryer that controls the diffusivity of moisture. In the rotary dryer, high temperatures can be applied for short drying times, allowing for better retention of compounds [24], while the effect of air velocity is less significant [25]. To describe the drying characteristics for different types of biomass, several theoretical and empirical models have been developed and applied [20]. For rotary dryers, several studies have investigated key factors and design parameters, including air velocity, feeding rate, hot air temperature, rotation speed, inclination, pressure, humidity, flow rate, weight, and moisture content (w.b.) for drying various feedstocks, such as olive oil cake [25–27], forestry biomass [23], wood chips [21], vegetable wholesale by-products [28], sweet potato-quinoa-kiwicha flakes [24] and peppermint [29]. Model parameters have to be obtained for settings related to the specific industry. Drying is one of the most important research topics with a lot of publications each year. Although a rotary dryer is a well-known drying process, each unique raw material has to be tested due to differences in the characteristics. From the literature review, there is research on the drying of bagasse [18–20]. However, there is no data on the sugarcane filter cake. RSM is one of the tools for predicting 2–3 factors in a study. In the field of drying, there are several articles that used RSM in different configurations [30–34]. However, there is no drying model for sugarcane filter cake.
The response surface methodology (RSM) is a mathematical and statistical method used for analyzing interactions between multiple factors (independent variables) and responses (dependent variables), for linear and quadratic effects, [24] and for the relationships between the response of interest and several associated variables. It helps to reduce the experimental cost by reducing the number of experimental trials compared to the complete experimental design, especially for full-scale testing [30]. Moreover, it is useful for applications with no reliable mathematical models to establish such a relationship, which can be approximated with a low-grade polynomial model. Many drying kinetic studies have been published. Traditionally, the drying kinetics of food products are described using empirical and semi-empirical models, including the single-parameter Lewis [35], two-parameter Page [36], Henderson–Pabis [37], logarithmic [38], Midilli et al. [39], two-term exponential [40] and Wang–Singh models [37]. Both RSM and kinetic studies are very important for predicting the drying efficiency at a specific time suitable for obtaining data for pilot-scale testing. RSM can be used to predict drying efficiency (DE) as a function of hot air temperature and rotation speed, while the kinetics test can predict moisture ratio (MR) as a function of time.
The present study elucidated the optimum rotation speed and temperature for the DE of the sugarcane filter cake for a pilot-scale rotary dryer. The purpose was to prove the drying process and obtain the drying characteristics of the material for upscaling to industrial scale. The paper focused on the prediction of DE using the RSM technique with two factors, namely temperature and rotation speed, using central composite face-center design (CCF). The experimental independent variables were hot air temperature (70–130 ℃) and rotation speed (1–5 RPM), while the dependent variable was DE. Furthermore, drying kinetic models of the material from the literature were investigated [18–20, 28, 37, 38, 40, 41]. Note that all drying experiments, including RSM and kinetics, were carried out using a pilot-scale rotary dryer with diameter 80 cm and length 6 m, in order to obtain data that approximated the industrial dryer.
The sugarcane filter cake was collected from the juice clarification and vacuum rotary filter process of the sugar industry in Udonthani, Thailand. The mass balance of the sugar manufacturing process is illustrated in Fig. 1. For every ton of sugarcane that enters the process, 4% becomes sugarcane filter cake.
2.1 Pilot-scale rotary dryer
A schematic and photograph of the countercurrent pilot-scale rotary dryer (Ø 0.8 m × length 6.0 m) are depicted in Fig. 2. In short, the dryer consists of three key components. First, it has an inlet, which uses a hopper. The hopper contains the wet material, which is carried into the chamber and the hot air is guided through this hopper. Second, the chamber is driven by a 2.2 kW, 1420 RPM and 220 V induction motor coupled to a gear reduction unit. An inverter is used for rotational speed control. Flights are placed along the chamber. And, thirdly, the outlet area is where the dried product is extracted from the chamber and removed from the process. At one end, the solid material drops into a basket for re-circulating. At the outlet of the chamber, a twelve-coil finned electric heater (15 kW) was installed at the other end of the dryer (opposite side of the inlet). The drying mode was set to be countercurrent hot air flow using an air blower (1.5 kW, 380 V and 2860 RPM) installed after the heater to transfer the heat into the drying chamber. The air velocity was maintained at 2 m/s, controlled by a flow rate control switch. Convective heat was transferred into the inlet of the chamber. Due to the very low air velocity, less turbulence occurred in the experiment. Two type K thermocouples with an accuracy ± 2°C were installed to measure the temperature. The thermocouple sensors were connected with the temperature controller set at the inlet and outlet of the chamber. One thermocouple at the outlet connected to the digital controller was installed to maintain the temperature of the electric hot air coil-finned heater process with within ± 1°C of the set point.
2.2 Experimental procedure
The experiments were carried out with a sugarcane filter cake discharged from the juice clarification and vacuum rotary filter process of the sugar industry in Udonthani, Thailand. The process operated in a semi-continuous mode, in which 30-kg filter cakes were added and re-circulated for 280 mins. The moisture content was monitored every 40 mins. The sugarcane filter cake was not treated with any chemicals or shredded before the experiments. The initial moisture content of the samples was determined by drying duplicate samples with the hot air oven drying method following standard methods AOAC 1990 [43]. Briefly, 2 g of the sample were placed into the crucible and then put into the oven 105°C for 24 hours. The filter cake was dried from 75% to nearly 20% (w.b.) moisture content in each experiment. The filter cake was dried with operating temperatures of 70, 100 and 130°C, a rotation speed of 1 to 5 RPM and a hot air velocity of 2 m/s. Experiments were conducted at each operating temperature after achieving constant temperature. After the dryer reached the target temperature of 70, 100, or 130°C, the sugarcane filter cake was put in the dryer.
2.3 Selection of optimum hot air temperature and rotation speed
The RSM was used to optimize the sugarcane filter cake drying conditions. The input factors of the experiments were conducted using the hot air temperature and rotation speed as independent variables. The RSM is a mathematical and statistical procedure that represents the participation of variables and affects the independent value by choosing the least experiments that can show the best connection results. The second polynomial order of the RSM in this study can be expressed as [44]:
$$\text{Y = }{\text{β}}_{\text{0}}\text{+ }\sum _{\text{i=1}}^{\text{k}}{\text{β}}_{\text{i}}{\text{x}}_{\text{i}}\text{+ }\sum _{\text{i=1}}^{\text{k}}{\text{β}}_{\text{ii}}{\text{x}}_{\text{i}}^{\text{2}}\text{+ }\sum _{\text{i}}\sum _{\text{j}}{\text{β}}_{\text{ij}}{\text{x}}_{\text{i}}{\text{x}}_{\text{j}}\text{ }\text{ }\text{ }\text{ }\text{ (1)}$$
where Y is the dependent variable (response), xi is the independent variable (factor) of hot air temperature and xj is the independent variable (factor) of rotation speed. The regression coefficients are \({\text{β}}_{\text{0}}\) for the intercept and \({\text{β}}_{\text{i}}\), \({\text{β}}_{\text{ii}}\) and \({\text{β}}_{\text{ij}}\) for the linear, quadratic and interaction terms, respectively [45].
A quadratic CCD is represented by a standard 2k factorial. In the present study, the experiment was generated using a CCF (Central Composite Face-centered) with two numeric factors on three levels (-1, 0 and 1) by five center points, four axial points and the value of α = 1 to analyze the independent factors of hot air temperature (70–130 ℃) with rotation speed (1–5 RPM) on the drying efficiency. The experimental drying efficiency (DE(%)) was determined by Eq. (2) [31]:
where DE(%) is the drying efficiency and M0 and Me are the moisture content of the samples before and after drying (280 min), respectively. A set of 13 experimental treatments, including five replicates at the center point was investigated. The RSM software Minitab®18 (Faculty of Engineering, Khon Kaen University) was utilized. The results were tested by the analysis of variance (ANOVA) and the process parameters were thus at the significant level of P-value ≤ 0.05. Table 1 shows the ranges and levels of a variable in operation. The two independent variables were hot air temperature and rotation speed. The response variable, DE(%), was evaluated by CCF.
|
Variable |
Level |
|||
|---|---|---|---|---|
|
-1 |
0 |
1 |
||
|
Hot air temperature (℃) |
70 |
100 |
130 |
|
|
Rotation speed (RPM) |
1 |
3 |
5 |
2.4 Mathematical kinetic modelling
To assess the characteristics of sugarcane filter cake in the drying kinetic modelling, seven drying models were used with the experimental data obtained from the pilot-scale setup. The moisture ratio (MR) represents the amount of moisture content remaining to analyze the experimental drying during the sugarcane filter cake experiment following:
$$\text{MR = }\frac{{\text{M}}_{\text{t}}\text{-}{\text{M}}_{\text{e}}}{{\text{M}}_{\text{o}}\text{-}{\text{M}}_{\text{e}}}\text{ (3)}$$
where M0 is the initial moisture content, Mt is the moisture content at drying time t (min) and Me is the equilibrium moisture content at each temperature. Since Me is relatively small compared with M0 [41], Eq. (3) can be simplified to:
$$\text{MR = }\frac{{\text{M}}_{\text{t}}}{{\text{M}}_{\text{o}}}\text{ }\text{ }\text{ }\text{ (4)}$$
3.1 Optimum temperature and rotation speed
3.1.1 The experiment design and nonlinear regression model
The relationship between the two independent variables (hot air temperature and rotation speed) and the dependent variable (DE (%)) was studied. The results shown in Table 2 for the CCF experimental matrix represent the experimental runs arranged in randomize order according to the RSM model design [43]. From Table 3, the predicted values indicate that the linear, quadratic and interaction terms, compared with experimental values, played a significant (P-value < 0.05) role in the ANOVA and the lack of fit testing was insignificant. The quadratic term for the rotation speed was more significant than the hot air temperature with an F-value of 23.76 compared to 0.12 for the temperature. It was also found that the rotation speed had an effect on the DE% greater than the hot air temperature due to the high F-value of 44.5. In contrast, the quadratic term of the hot air temperature had an insignificant effect on the DE% with a P-value of 0.735 compared to the F-value of 0.05 [46]. The fitted regression model was explained in terms of the coefficient of determination (R2, 0.94), adjusted R2 (0.90) and predicted R2 (0.65). The correlation of predicted against actual results, and the normal plot residuals of the drying efficiency response are shown in Fig. 3. The developed models are fitted with the data, as shown from the normal probability plotline that closes the standardized residuals. The regression model to the response in terms of uncoded units determined by the software is given by:
DE (%) = -62.2 + 1.279 hot air temperature (°C) + 34.57 rotation speed (RPM) – 0.00162 hot air temperature (°C) × hot air temperature (°C) – 5.06 rotation speed (RPM) × rotation speed (RPM) – 0.1362 hot air temperature (°C) × rotation speed (RPM) (5)
| Run | Hot air temperature (℃) | Rotation speed (RPM) | Drying efficiency (%) experiment |
|---|---|---|---|
| 1 | 70.0 | 1.0 | 39.68 |
| 6 | 130.0 | 3.0 | 91.28 |
| 2 | 130.0 | 1.0 | 82.49 |
| 7 | 100.0 | 1.0 | 71.03 |
| 10 | 100.0 | 3.0 | 55.12 |
| 3 | 70.0 | 5.0 | 20.84 |
| 13 | 100.0 | 3.0 | 68.55 |
| 11 | 100.0 | 3.0 | 67.81 |
| 12 | 100.0 | 3.0 | 67.63 |
| 4 | 130.0 | 5.0 | 30.96 |
| 8 | 100.0 | 5.0 | 28.73 |
| 9 | 100.0 | 3.0 | 68.42 |
| 5 | 70.0 | 3.0 | 46.01 |
3.1.2 Drying efficiency optimization
To investigate the relationship of the hot air temperature and rotation speed to the DE(%), contour plots are given in Fig. 4. These plots help to analyze the interactions between parameters (hot air temperature and rotation speed) and identify appropriate conditions to maximize the DE(%).
| Source | DF | F-value | P-value | Significant (< 0.05) |
|---|---|---|---|---|
| Model | 5 | 22.68 | < 0.001 | Significant |
| Linear | 2 | 39.16 | < 0.001 | Significant |
| Hot air temperature (℃) | 1 | 33.81 | < 0.001 | Significant |
| Rotation speed (RPM) | 1 | 44.50 | < 0.001 | Significant |
| Quadratic | 2 | 14.74 | 0.003 | Significant |
| Hot air temperature (℃) × hot air temperature (℃) | 1 | 0.12 | 0.735 | Insignificant |
| RPM × RPM | 1 | 23.76 | 0.002 | Significant |
| Two-way interaction | 1 | 5.62 | 0.050 | Significant |
| Hot air temperature (℃) × RPM | 1 | 5.62 | 0.050 | Significant |
| Error | 7 | |||
| Lack-of-fit | 3 | 1.94 | 0.265 | Insignificant |
| Pure error | 4 | |||
| Total | 12 |
| Experiment | Hot air temperature (℃) | RPM | DE(%) | |
|---|---|---|---|---|
| 1 | 80 | 1 | Actual Predicted %Error | 49.76 48.37 2.88 |
| 2 | 90 | 1 | Actual Predicted %Error | 59.85 57.04 4.93 |
| 3 | 115 | 1 | Actual Predicted %Error | 76.21 77.31 1.42 |
The lowest and highest DE(%) values are denoted by blue and red in Fig. 4, respectively. The interaction between the hot air temperature and rotation speed parameters is very significant for the DE(%). The predicted operation at low hot air temperatures with high rotation speeds results in low DE(%), whereas high hot air temperatures combined with low rotation speeds result in high DE(%) [47]. Increasing the rotation speed leads to a decrease in the length of the drying rate zone due to a decrease in residence time. Therefore, reducing the moisture content of the sugarcane filter cake [23] affects the DE(%) due to the longer drying time. Confirmation experiments were performed to verify the model fitted using test conditions with three experimental ranges. The exponential condition, actual values, predicted values and error percentage are summarized in Table 4. The experimental actual and predicted values were found with tolerances that can be accepted for different errors below the 5% level [30, 47, 48].
3.2 Evaluation of drying kinetic models
Table 5 shows the installation of the seven thin-layer drying models, in which the moisture content data are calculated in the moisture ratio format. According to Table 5, the constants k, a, b and n are the adjustable parameters of the model. The Levenberg–Marquardt method in Origin® 8.0 is used for evaluating the variables in the model equations. The coefficient of determination (R2), the average relative error percentage (P), the reduced chi-square (ꭓ2) and the root mean square error (RMSE) are shown in Eqs. (6)–(9) [20, 41]:
| Model | Representation model | References |
|---|---|---|
| Newton | MR = exp(-kt) | [[20, 35, 36]] |
| Page | MR = exp(-ktn) | [[29, 36]] |
| Henderson–Pabis | MR = a⸱exp(-ktn) | [[37, 49]] |
| Logarithmic | MR = a⸱exp(-kt) + b | [[38]] |
| Midilli–Kucuk | MR = a⸱exp(-ktn) + bt | [[39]] |
| Two-term exponentials | MR = a⸱exp(-kt) + (1-a)⸱exp(-kat) | [[40]] |
| Wang–Singh | MR = 1 + at + bt2 | [[37]] |
$${\text{R}}^{\text{2}} \text{= 1-}\frac{{\sum _{\text{i=1}}^{\text{N}}\text{(}{\text{MR}}_{\text{pre,i}}\text{-}{\text{MR}}_{\text{exp,i}}\text{)}}^{\text{2}}}{{\sum _{\text{i=1}}^{\text{N}}\text{(}{\stackrel{\text{-}}{\text{MR}}}_{\text{pre}}\text{-}{\text{MR}}_{\text{exp,i}}\text{)}}^{\text{2}}}$$
6
$$\text{P}\text{ = }\frac{\text{100}}{\text{N}}\sum _{\text{i=1}}^{\text{N}}\frac{\left|{\text{MR}}_{\text{exp,i}}\text{-}{\text{MR}}_{\text{pre,i}}\right|}{{\text{MR}}_{\text{exp,i}}}$$
7
$${\text{ꭓ}}^{\text{2 }}\text{= }\frac{\sum _{\text{i=1}}^{\text{N}}\text{(}{{\text{MR}}_{\text{exp,i}}\text{-}{\text{MR}}_{\text{pre,i}}\text{)}}^{\text{2}}}{\text{N-Z}}$$
8
$$\text{RMSE =}{\text{ }\left[\frac{\text{1}}{\text{N}}{\sum _{\text{i=1}}^{\text{N}}\text{(}{\text{MR}}_{\text{pre,i}}\text{-}{\text{MR}}_{\text{exp,i}}\text{)}}^{\text{2}}\right]}^{\text{1}\text{/}\text{2}}$$
9
where MRexp,I is for the experimental moisture ratio data, MRpre,I is for the predicted moisture ratio data, the number of observations is N and the number of drying constants is z [50].
Table 6. Statistical computing results.
| Temp. | Model | Equation | R2 | P | ꭓ2 | RMSE | |
| 70 ºC | Newton | MR = exp(-0.0019⸱t) | 1.0000 | 3.0918 | < 0.0001 | 0.0045 | |
| Page | MR = exp(-0.0033⸱t0.8970) | 1.0000 | 2.7828 | < 0.0001 | 0.0031 | ||
| Henderson–Pabis | MR = 1.1889⸱exp(-0.033⸱t0.535) | 0.9988 | 4.5175 | 0.0045 | 0.0688 | ||
| Logarithmic | MR = 0.5949⸱exp(-0.0041⸱t)-0.4144 | 1.0000 | 2.5153 | < 0.0001 | < 0.0001 | ||
| Midilli–Kucuk | MR = 1.0095⸱exp(-0.0029⸱t) + 0.0006⸱t | 1.0000 | 2.5203 | < 0.0001 | < 0.0001 | ||
| Two-term exponentials | MR = 0.188975⸱exp(-0.0067⸱t)+(1-0.1889) ⸱exp(-0.0067⸱0.1889⸱t) | 1.0000 | 2.6107 | < 0.0001 | 0.0034 | ||
| Wang–Singh | MR = 1-0.0022⸱t + 2.74⸱10− 6⸱t2 | 1.0000 | 2.4543 | < 0.0001 | 0.0042 | ||
| 100 ºC | Newton | MR = exp(-0.0033⸱t) | 0.9724 | 16.5929 | 0.1057 | 0.3251 | |
| Page | MR = exp(-0.0003⸱t1.4578) | 0.9776 | 16.0276 | 0.0853 | 0.2920 | ||
| Henderson–Pabis | MR = 0.9799⸱exp(-0.000182⸱t1.5440) | 0.9763 | 16.3200 | 0.0906 | 0.3011 | ||
| Logarithmic | MR = 99.8340⸱exp(-0.000025⸱t)-98.8124 | 0.9772 | 16.4394 | 0.0870 | 0.2949 | ||
| Midilli–Kucuk | MR = 1.0213⸱exp(-1.19⸱10− 9⸱t)-0.0025⸱t | 0.9772 | 16.4331 | 0.0870 | 0.2950 | ||
| Two-term exponentials | MR = 1.9322⸱exp(-0.058⸱t)+(1-1.9322) ⸱exp(-0.058⸱1.9322⸱t) | 0.9774 | 15.9603 | 0.0861 | 0.2934 | ||
| Wang–Singh | MR = 1-0.0021⸱t-1.4⸱10− 6⸱t2 | 0.9765 | 16.0415 | 0.0895 | 0.2992 | ||
| 130 ºC | Newton | MR = exp(-0.0033⸱t) | 0.9717 | 16.8153 | 0.1086 | 0.3295 | |
| Page | MR = exp(-7.27⸱10− 6⸱t2.162) | 0.9817 | 18.7839 | 0.0696 | 0.2639 | ||
| Henderson–Pabis | MR = 0.9962⸱exp(-6.41⸱10− 6⸱t2.184) | 0.9813 | 18.7607 | 0.0709 | 0.2663 | ||
| Logarithmic | MR = 380.0125⸱exp(-7.72⸱10− 6⸱t)-378.922 | 0.9812 | 19.4917 | 0.0716 | 0.2675 | ||
| Midilli–Kucuk | MR = 1.0902⸱exp(-8.64⸱10− 10⸱t)-0.0029⸱t | 0.9812 | 19.4931 | 0.0716 | 0.2675 | ||
| Two-term exponentials | MR = 2.2141⸱exp(-0.007⸱t)+(1-2.2141) ⸱exp(-0.007⸱2.2141⸱t) | 0.9829 | 19.4512 | 0.0651 | 0.2550 | ||
| Wang–Singh | MR = 1-0.0009⸱t-7.1⸱10− 6⸱t2 | 0.9798 | 17.8002 | 0.0768 | 0.2771 | ||
Table 6 shows the fitting model computation, with the best-fitted models based on the highest R2 values and the lowest P, ꭓ2 and RMSE values. The best models selected on R2 values were more than 0.97 for all models. The Page model has high R2 values and low P, ꭓ2, and RMSE values at all temperatures. Generally, the R2, P, ꭓ2 and RMSE values were found in the range of 0.9776–1, 2.7828–18.7839, < 0.0001–0.0853 and 0.0031–0.2920, respectively. The experimental kinetic data and the fitted models are shown in Figs. 5 and 6, respectively. The statistical results were calculated by the experiment and the predicted MR with constant values was plotted. The MR decreased exponentially with an elapsed drying duration and decreased faster at a higher temperature in all cases, as expected. Figure 7 confirms that this model is a good fit between the experimental and predicted MR of sugarcane filter cake characteristic as R2 = 0.9696 and slope = 0.9627 [41].
A rotary dryer is shown to be an effective method for drying sugarcane filter cake. The hot air temperature and rotation speed were significantly effective parameters on the drying efficiency, as shown by the contour plot. Increased temperature and reduced rotation speed resulted in a high percentage of drying efficiency. From the seven mathematical drying kinetic models, the Page kinetic model was suitable for predicting the drying characteristics at a drying temperature of 70–130 ºC. Sugarcane filter cake can be used as a boiler fuel in sugar factories using the optimization tool and kinetic data. For the boiler's heat to be used as pre-heated, the operating cost can be further reduced.
Acknowledgments
This research project was financially supported by the Center for Alternative Energy Research and Development, Khon Kaen University, Thailand.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
- The Cane and Sugar Board, Ministry of Industry of Thailand, 2020.
- W. Pornprakun, S. Sungnul, C. Kiataramkul, E.J. Moore, Determining optimal policies for sugarcane harvesting in Thailand using bi-objective and quasi-Newton optimization methods, Adv. Differ. Equ., 257 (2019).
- N. Makul, G. Sua-iam, Characteristics and utilization of sugarcane filter cake waste in the production of lightweight foamed concrete, J. Clean. Prod., 126 (2016) 118-133.
- P.A. Ochoa George, J.J.C. Eras, A.S. Gutierrez, L. Hens, C. Vandecasteele, Residue from Sugarcane Juice Filtration (Filter Cake): Energy Use at the Sugar Factory, Waste Biomass Valorization, 1 (2010) 407-413.
- K. Pajampa, T. Wongwuttanasatian, Biomass pellets produced from filler cake as waste to energy in sugar industry, IOP Conf. Ser. Earth Environ. Sci., 257 (2019).
- A.V. Delgado, C.d.A. Casanova, Sugar processing and by-products of the sugar industry, Agric. Serv. Bull. (F. A. O.), 144 (2001).
- E.M. Ossom, R.L. Rhykerd, Response of Ipomoea batatas (L.) Lam. to soil fertilization with filter cake, Trans. Ill. State Acad. Sci., 100 (2007) 197-208.
- S.R. Utami, S. Kurniawan, B. Situmorang, N.D. Rositasari, Increasing P-Availability and P-Uptake Using Sugarcane Filter Cake and Rice Husk Ash to Improve Chinesse Cabbage (Brassica Sp) Growth in Andisol, East Java, J. Agric. Sci., 4 (2012).
- S. Meunchang, S. Panichsakpatana, R.W. Weaver, Co-composting of filter cake and bagasse; by-products from a sugar mill, Bioresour. Technol., 96 (2005) 437-442.
- R. Prado, G. Caione, C. Campos, Filter Cake and Venessa as Fertilizers Contributing to Conservation Agriculture, Appl. Environ. Soil Sci., 2013 (2013) 8.
- A. Jala, S. Chakkrawan, Bedding material mixed with different ratio of filter cake from sugar cane which effected on growth of 2 types of earthworm, Tjst., 5 (2016) 43-55.
- I.M. Lima, M. Wright, Microbial stability of worm castings and sugarcane filter mud compost blended with biochar, Cogent food agric., 4 (2018).
- W.T. Tsai, H.P. Chen, C.W. Lai, K.J. Hsien, M.S. Lee, J.M. Yang, Preparation of adsorbents from sugarcane manufacturing by-product filter-mud by thermal activation, J Anal Appl Pyrolysis, 70 (2003) 399-411.
- H. Li, W. Xu, X. Yang, J. Wu, Preparation of Portland cement with sugar filter mud as lime-based raw material, J. Clean. Prod., 66 (2014) 107-112.
- H. Li, J. Xu, J. Wu, W. Xu, Y. Xu, Influence of sugar filter mud on formation of portland cement clinker, J. Wuhan Univ. Technol. Mater. Sci. Ed., 28 (2013) 746-750.
- G. Sua-Iam, N. Makul, Effect of incinerated sugarcane filter cake on the properties of self-compacting concrete, Constr Build Mater., 130 (2017) 32-40.
- L. Janke, A. Leite, M. Nikolausz, T. Schmidt, J. Liebetrau, M. Nelles, W. Stinner, Biogas Production from Sugarcane Waste: Assessment on Kinetic Challenges for Process Designing, Int J Mol Sci, 16 (2015) 20685-20703.
- J.M.G. Costa, J.L.G. Corrêa, B.E. Fonseca, F.M. Borém, S.V. Borges, Drying and Isotherms of Sugar Cane Bagasse, Eng. Agric., 23 (2015) 128-142.
- M.A. Mazutti, G. Zabot, G. Boni, A. Skovronski, D. de Oliveira, M. Di Luccio, J.V. Oliveira, M.I. Rodrigues, H. Treichel, F. Maugeri, Mathematical modeling of thin-layer drying of fermented and non-fermented sugarcane bagasse, Biomass Bioenergy, 34 (2010) 780-786.
- F.B. Scheufele, C. Ribeiro, A.N. Módenes, F.R. Espinoza-Quiñones, R. Bergamasco, N.C. Pereira, Assessment of drying temperature of sugarcane bagasse on sorption of reactive blue 5G dye, Fibers Polym., 16 (2015) 1646-1656.
- J. Zabski, P. Lampart, S. Gumkowski, Biomass drying: Experimental and numerical investigations, Arch. Thermodyn., 39 (2018) 39-73.
- M.G. Ortega, F. Castaño, M. Vargas, F.R. Rubio, Multivariable robust control of a rotary dryer: Analysis and design, Control Eng. Pract., 15 (2007) 487-500.
- A. Zabaniotou, Simulation of forestry biomass drying in a rotary dryer, Dry. Technol., 18 (2000) 1415-1431.
- V. Vásquez Villalobos, O. Hernández-Bracamonte, J. Rojas-Naccha, V. Ninaquispe-Zare, C. Rojas, J. Vásquez-Angulo, Moisture prediction of sweet potato-quinoa-kiwicha flakes dried by rotary drum dryer using artificial intelligence, Sci. Agropecu., 9 (2018) 83-91.
- M.K. Krokida, Z.B. Maroulis, C. Kremalis, Process Design of Rotary Dryers for Olive Cake, Dry. Technol., 20 (2006) 771-788.
- A.G. Chilka, V.V. Ranade, CFD modelling of almond drying in a tray dryer, Can. J. Chem. Eng., 97 (2019) 560-572.
- F.J. Gómez-de la Cruz, P.J. Casanova-Peláez, J.M. Palomar-Carnicero, F. Cruz-Peragón, Modeling of olive-oil mill waste rotary dryers: Green energy recovery systems, Appl. Therm. Eng., 80 (2015) 362-373.
- A. Iguaz, A. Esnoz, G. Martı́nez, A. López-Gómez, P. Vírseda, Mathematical modeling and simulation for the drying process of vegetable wholesale by-products in a rotary dryer, J. Food Eng., 59 (2003) 151-160.
- S. Tarhan, İ. Telci, M.T. Tuncay, H. Polatci, Product quality and energy consumption when drying peppermint by rotary drum dryer, Ind Crops Prod, 32 (2010) 420-427.
- A.K. Khare, A.K. Biswas, S. Balasubramanium, M.K. Chatli, J. Sahoo, Optimization of meat level and processing conditions for development of chicken meat noodles using response surface methodology, J Food Sci Technol, 52 (2015) 3719-3729.
- R. Anyanwu, C. Rodriguez, A. Durrant, A.G. Olabi, Optimisation of Tray Drier Microalgae Dewatering Techniques Using Response Surface Methodology, Energies, 11 (2018) 2327.
- H. Majdi, J. Esfahani, M. mohebbi, Optimization of convective drying by response surface methodology, Comput Electron Agric, 156 (2019) 574-584.
- C.E. Onu, P. K. Igbokwe, J. T. Nwabanne, C. O. Nwajinka, P. E. Ohale, Evaluation of optimization techniques in predicting optimum moisture content reduction in drying potato slices, Artif Intell, 4 (2020) 39-47.
- P.I. Moloto, M. Mosala, A.O. Omolola, A.I.O. Jideani, S.M. Laurie, Optimization of hot-air drying conditions on functional properties of flour from dried South African sweet potato cultivars (Impilo and Bophelo) using the response surface methodology, J Food Sci Technol, 41 (2021) 39-46.
- W.K. Lewis, The Rate of Drying of Solid Materials, J Ind Eng Chem, 13 (1921) 427-432.
- D.M. Bruce, Exposed-layer barley drying: Three models fitted to new data up to 150°C, J. Agric. Eng. Res., 32 (1985) 337-348.
- Z. Wang, J. Sun, X. Liao, F. Chen, G. Zhao, J. Wu, X. Hu, Mathematical modeling on hot air drying of thin layer apple pomace, Food Res. Int., 40 (2007) 39-46.
- O. Yaldiz, C. Ertekin, H.I. Uzun, Mathematical modeling of thin layer solar drying of Sultana grapes, Energy, 26 (2001) 457-465.
- A. Midilli, H. Kucuk, Z. Yapar, A NEW MODEL FOR SINGLE-LAYER DRYING, Dry. Technol., 20 (2002) 1503-1513.
- E.K. Akpinar, Y. Bicer, C. Yildiz, Thin layer drying of red pepper, J. Food Eng., 59 (2003) 99-104.
- S. Kayran, İ. Doymaz, Infrared Drying and Effective Moisture Diffusivity of Apricot Halves: Influence of Pretreatment and Infrared Power, J. Food Process. Preserv., 41 (2017).
- O. Almazan, Gonzalez, L., Galvez, L., The sugar cane, its byproducts and co-products, Proceedings of the Third Annual Meeting of Agricultural Scientists, Réduit, Mauritius, 1998.
- AOAC, Official methods of analysis of the Association of Official Analytical Chemists, The Association, Arlington, VA, 1990.
- D.C. Montgomery, Design and analysis of experiments, Ninth edition. John Wiley & Sons, New York, USA.2017.
- E. Lazzari, A.d.S. Polidoro, B. Onorevoli, T. Schena, A.N. Silva, E. Scapin, R.A. Jacques, E.B. Caramão, Production of rice husk bio-oil and comprehensive characterization (qualitative and quantitative) by HPLC/PDA and GC × GC/qMS, Renew. Energy, 135 (2019) 554-565.
- S.H. Shah, Z. Abdul Haiyee, K. Ismail, N. Hashim, W.I. Wan Ismail, Optimization of curcuma longa L. Rhizome supercritical carbon dioxide extraction (SC-CO2) by response surface methodology (RSM), J. Teknol., 78 (2016) 87-92.
- N. Kumar, S.K. Singh, G. Gautam, A.K. Padap, A. Mohan, S. Mohan, Synthesis and statistical modelling of dry sliding wear of Al 8011/6 vol.% AlB2in situcomposite, Mater. Res. Express, 4 (2017).
- M.P. Gundupalli, D. Bhattacharyya, Ethanol Production from Acid Pretreated Food Waste Hydrolysate Using Saccharomyces cerevisiae 74D694 and Optimizing the Process Using Response Surface Methodology, Waste Biomass Valorization, 10 (2017) 701-708.
- D.S. Boyce, Grain moisture and temperature changes with position and time during through drying, J. Agric. Eng. Res., 10 (1965) 333-341.
- B. Alaei, R. Amiri Chayjan, Drying Characteristics of Pomegranate Arils Under Near Infrared-Vacuum Conditions, J. Food Process. Preserv., 39 (2015) 469-479.
No competing interests reported.