Skip to Content
ProcessesProcesses
  • Article
  • Open Access

10 May 2026

17 Pages

Sorghum and Wheat-Based Extruded Aquatic Feed—Impact of Drying Parameters on Pellet Quality and Energy Efficiency

,
and
Department of Grain and Food Science, Kansas State University, Manhattan, KS 66506, USA
*
Author to whom correspondence should be addressed.

Abstract

Energy consumption and different methods for determining energy efficiency were evaluated for drying of extruded rainbow trout feed pellets using a pilot-scale, heated air, integrated conveyor dryer and cooler. Impact of drying parameters on product quality, especially final moisture and pellet durability index (PDI), was also studied. From an initial moisture of 21.2 to 22.1% wet basis (wb), the drying–cooling process reduced the pellet moisture to 3.5 to 5.0% wb. Dryer throughput (82–121 kg/h) did not have statistically significant impact on final moisture (p = 0.0965) although the highest throughput corresponded to highest moisture; but increase in drying temperature from 93 to 115 °C led to a significant decrease in final moisture (p = 0.0285). Increase in dryer throughput led to a significant increase in PDI from 82.8 to 88.0% (p = 0.0003), while increase in drying temperature resulted in a slight decrease in PDI from 84.3 to 83.6%, although not statistically significant (p = 0.0811). Sorghum-based aquatic feed had a slightly lower PDI than wheat-based feed (82.8 versus 83.7%, respectively), but the difference was not statistically significant (p = 0.3009). Differences in pellet durability were attributed primarily to structural weakness induced by product shrinkage during drying, which in turn was impacted by drying rates. Specific energy consumption (SEC) during drying decreased from 136.3 to 101.1 MJ/kg-water with increase in throughput and increased from 122.5 to 150.1 MJ/kg-water with increase in drying temperature. An inverse trend was observed for various measures of dryer energy efficiency, with increase in efficiency for higher throughput and decrease for higher temperature. Sorghum-based aquatic feed had a higher drying SEC as compared to wheat-based feed and a lower energy efficiency. Overall, the results highlighted trade-offs between throughput, drying efficiency and pellet quality during drying of aquatic feeds.

1. Introduction

As the world-wide demand for high-quality protein rises, fish consumption has increased by 3% annually from 1961 to 2019, which is nearly double the population growth rate [1]. Aquatic species are now the source of approximately 15% of the animal protein globally, and more than 3 billion people get at least 20% of their daily protein from fish [2]. Aquaculture supplies 130.9 million tons out of a total of 185.4 million tons production of aquatic animals in the world, with the share of the former continuing to grow each year due to ocean stock depletion and stagnation of capture fisheries.
Aquatic feed constitutes a large percentage (50–70%) of the total cost of production in aquaculture [3]. Approximately 70% of aquaculture farms use commercially produced feeds, and the aquafeed industry with an annual growth rate of 2.7% is outperforming most other animal feed sectors [4]. Extrusion has become the preferred processing method for producing pelletized aquatic feed that is highly durable, water stable and tailored to varying requirements such as floating versus sinking based on target species. The extrusion process blends raw materials with steam and water, applies high temperature, pressure and shear, and gelatinizes starches and denatures proteins, resulting in a cooked, plasticized dough that is forced out through die holes to get shaped into pellets with varying degrees of expansion and porosity depending on the energy input [5,6].
Depending on the process requirements, the in-barrel product moisture during extrusion of aquatic feed can range up to 40% wet basis [7,8]. Vapor flash-off induced by high temperatures reduces the product moisture on discharge from the extruder die, and a further reduction is observed due to contact with room temperature air as the aquatic feed pellets are taken away either pneumatically or via conveyor belt [9]. This initial moisture removal is inadequate, and further drying is essential for bringing the final pellet moisture down to 8–12% in order to inhibit microbial growth and increase resistance to mechanical damage [10]. Thus, the drying step ensures adequate shelf-life and prevents deterioration of the aquatic feed during transport, handling and storage. Removal of water also creates physical space in the pores and a non-polar pellet matrix, which facilitates oil absorption during the subsequent coating step [11].
Extensive research has been conducted on drying of various foods and its relationship to product quality [12,13,14]. The polymer science phenomenon of glass transition is often used to explain certain aspects of food quality [15,16]. Application of this concept to drying, a process involving transition of product from a malleable “rubbery” state to a rigid “glassy” state, has been helpful in understanding resultant quality of food products and in a more recent study on aquatic feed as well [9,12]. The specific conditions under which glass transition occurs in the dryer, including the temperature, humidity and velocity of the drying air, dictate internal stresses, final microstructure, surface characteristics, mechanical strength and hydrodynamic properties of aquatic feed pellets. For example, a poorly controlled drying process can introduce micro-fractures and create a brittle structure, resulting in a final product of inferior quality. Other studies have also focused on the impact of drying on extruded aquatic feed quality. Besides the durability of pellets, drying temperature and air velocity have been shown to strongly influence final moisture, porosity, bulk density, pellet buoyancy and water stability [10,11,17,18,19,20,21]. Factors such as product size, bed depth, humidity and air flow distribution are also important considerations during drying of aquatic feed [17,20,22,23].
It is obvious from the above discussion that the impact of drying on technical quality of food and feed products is well researched. On the other hand, energy consumption and efficiency of the drying process for agricultural products in general and extruded aquatic feed in particular is relatively less understood. Quantification of energy input during extrusion is a standard practice, especially mechanical energy and more recently the harder to measure thermal energy as well pioneered by research at Kansas State University [8,24,25,26,27]. Similar focus on energy metrics during post-extrusion drying and also benchmarking and improvement of dryer efficiency is lacking but is sorely needed as moisture removal is one of the most energy-intensive operations in food and feed processing [28].
Energy consumed during drying of agricultural commodities in various studies using different types of dryers was reviewed [29]. A convective hot air parallel air flow dryer for cherry fruits was one of the systems described. Energy delivered by the electric heaters was directly related to air velocity but decreased with an increase in drying temperature. Similar results with respect to air temperature and velocity were observed for drying of chamomile flowers in a convective hot air dryer [30]. Energy required for drying of celery leaves in a ventilated oven was also found to decrease with air temperature [31]. It should be noted that the dryers used in these studies were batch-type. In another research, energy consumption in a continuous dryer for hops used in brewing was calculated based on heating fuel used by the boiler [32]. The dryer in this case consisted of three conveyor-belts. Energy consumption data by itself, however, does not provide an idea of dryer energy efficiency for facilitation of improvement in performance. Fewer studies have focused on the latter aspect. Although there is no standard method for characterizing dryer efficiency, it is broadly defined as the ratio between the ideal heat load or the theoretical minimum energy required to dry products and actual energy consumption [9,29]. There can be different ways for calculating either parameter in this ratio, leading to varying estimates of efficiency. For example, the heat load can be based on latent heat of vaporization alone or incorporate the sensible heating of air as well, and the actual energy consumption can just be based on energy used for heating or also include energy consumption for mechanical parts such as blowers [30,31].
Literature search specifically for dryer energy efficiency for extruded aquatic feed yielded a study using batch, fluidized bed dryer [20]. However, the method used for measuring efficiency was not elaborated clearly. The current research aims to fill this gap by exploring energy consumption and also different methods for determining energy efficiency during drying of extruded rainbow trout feed pellets. As discussed above, the drying process can also significantly impact aquatic feed quality. This work goes a step further than most dryer studies on a range of agricultural products by examining the trade-off or balance between product quality and energy efficiency, which can lead to improvement of dryer design and optimization of the drying process. This approach was aimed at the implementation of energy-saving strategies while mitigating the risk of compromising critical quality parameters such as mechanical durability of the pellets, which is a daunting engineering problem for the aquafeed industry [9]. Based on results from the literature summarized before, it was hypothesized that an increase in dryer air temperature would improve dryer efficiency and pellet durability, while an increase in product bed depth in the dryer would have a negative impact on both aspects due to higher load and poor air flow. It was also hypothesized that composition of the aquatic feed, specifically the primary component or grain used for carbohydrate source, would have an impact on drying and pellet quality due to differences in physicochemical properties and water binding [8,33]. A pilot-scale, heated air, continuous, two-belt conveyor dryer and single-belt cooling system was used for drying of aquatic feed produced using a single-screw extruder. Wheat and grain sorghum were the two carbohydrate sources that were studied as part of the feed formulation.

2. Materials and Methods

2.1. Aquatic Feed Formulations

Two aquatic feed diets were formulated to assess the impact of different carbohydrate sources on post-extrusion drying, besides dryer process parameters. The diets were designed to be nutritionally complete for feeding rainbow trout, with approximately 45% crude protein and 18% crude lipids, and containing either wheat or grain sorghum as the carbohydrate source at 25.9% inclusion level. The remaining portion of the diets was exactly identical and consisted of poultry by-product meal and Menhaden fishmeal (15.6% each), soybean meal (13.8%), corn protein concentrate, blood meal and feather meal (5.7% each), and minor ingredients (11.9%) that were primarily sources for micronutrient such as vitamins, minerals and amino acids. Further information, which is not directly relevant to this study, including detailed formulation, ingredient sources and methods for proximate composition analyses of diets were described previously [33].

2.2. Experimental Design

The study involved three experiments focusing on impact of dryer throughput and consequently product bed depth (experiment 1), drying temperature (experiment 2) and different carbohydrate or grain source (experiment 3) on post-extrusion drying performance, energy usage, efficiency and final aquatic feed pellet quality. Experiment 1 consisted of three treatments corresponding to different levels of dryer throughput. The dryer air temperature or drying temperature was kept constant at 104.4 °C (or 220 °F) and extruded sorghum-based aquatic feed was the product. The throughput was controlled via speed of the feeder screw, set at 10, 12 and 14 rpm, which delivered raw diets to the extrusion system. This corresponded to feed rates of 82, 108 and 121 kg/h, respectively. For obtaining the feed rates, the volumetric feed delivery system was calibrated for mass throughput at different screw speeds. Measurements were made in duplicate and averaged over both high and low fill levels of the live-bottom feed hopper. Henceforth, these raw material feed rates to the extruder will be used for the dryer throughput, although the latter would actually be higher due to moisture gain during extrusion. In experiment 2, three treatments corresponding to different levels of drying temperature, viz., 93.3°, 104.4° and 115.5 °C (200, 220 and 240 °F, respectively) were studied with wheat-based extruded aquatic feed as the product. The dryer throughout was kept constant at 79 kg/h for these treatments, which corresponded to extrusion feeder screw speed of 10 rpm. Experiment 3 consisted of just two treatments and involved comparison of drying performance and related metrics for sorghum- versus wheat-based extruded aquatic feed at constant throughput (79–82 kg/h) and dryer temperature (104.4 °C).

2.3. Aquatic Feed Extrusion

The focus of this study was post-extrusion drying. Therefore, only a brief description of aquatic feed production steps prior to drying is provided here including extrusion processing. Detailed information on methodology and equipment can be obtained elsewhere [33]. Wheat and sorghum grain were coarse ground using a pilot-scale hammer mill (Model D Comminuter, Fitzpatrick, Westwood, MA, USA), followed by mixing of all ingredients using a double-ribbon batch blender (Model 207SM20, Wenger Manufacturing, Sabetha, KS, USA) and post-mix grinding on the hammer mill using a 1.02 mm screen. Formulations were processed using a 37.3 kW (50 hp) pilot-scale single-screw extruder (Model X-20, Wenger Manufacturing, Sabetha, KS, USA) with a screw diameter of 82.6 mm (3.25 in) and L/D ratio of 10:1. The raw material mash was conditioned to 95–99 °C in a differential diameter–differential speed pre-conditioner (Model 2 DDC, Wenger Manufacturing) with water and steam injection at 5–7 kg/h and 14 kg/h, respectively. Water was injected into the extruder barrel at rate of 4–10 kg/h. The water injection rates into the system were adjusted according to the feed rate for achieving a final in-barrel moisture of 28–29% wb. A die with five 2.50 mm circular cross-section openings was used for shaping the product at the extruder discharge, followed by cutting with a rotary knife system. A negative pressure pneumatic system was used to convey aquatic feed pellets from the extruder discharge to a pilot-scale continuous drying and cooling system, which is described in detail in Section 2.4.

2.4. Extrudate Drying and Cooling

A sloped oscillating spout spreader at the end of the pneumatic conveying system was used to discharge the wet aquatic feed pellets from the extruder into a two-pass gas-fired continuous dryer with a single drying chamber attached to a single-pass cooler (Series 4800, Wenger Manufacturing). Product was conveyed in the dryer and cooler on stainless-steel woven screen belt conveyors. The dryer air circulation was designed to follow a dual-plenum, central air flow pattern. This was achieved via two air flow plenums on either side of the drying chamber to allow circulation between the burner/ heater, dryer air intake and exhaust, and to and from the drying chamber. The dry, heated air entered the drying chamber at one end and simultaneously passed through the top drying belt (first product pass) from below and the bottom drying belt (second product pass) from above. Spent air emerged from the top belt from above the product bed and the bottom belt from below the product bed and exited the other end of the drying chamber close to the dryer exhaust. The dryer and cooler residence time, controlled via the speeds of the conveyor belts, were kept constant for all experiments and so were other drying parameters such as air velocity. The residence time for the first and second product passes through the dryer was set at 8 and 9.5 min, respectively. Cooling was done with room temperature air with a residence time of 4.5 min.
Ambient air and dryer exhaust air temperature and humidity data were collected using relative humidity and temperature probes (Model THWD-5, Amprobe, Everett, WA, USA). The probe for ambient air was placed at the dryer intake below the feed spreader, while the exhaust air probe was inserted through a hole in the outlet duct. The probe maximum temperature rating was higher than the exhaust air temperature, therefore an assumption was made for the latter. Volumetric flow rates of the cooler inlet air and dryer exhaust air (in cubic feet per minute or ft3/min) were based on average data for a typical drying experiment for extruded aquatic feed and collected at the same locations as the air temperature and humidity data described above using an anemometer (Model K1-10001, Koselig LLC, Providence, RI, USA). Natural gas usage (in cubic feet or ft3) was determined by visual observation of the gas meter attached to the dryer supply. All readings were taken at the beginning and end of each treatment, but in some cases were also recorded continuously using a video camera and subsequently transcribed and averaged. Product was collected off the extruder, off the spreader and off the integrated dryer–cooler system for moisture content analysis.

2.5. Moisture Content

Moisture content was measured using the AACC 44-19.01 method [34]. Approximately 2 g of samples were dried at 135 °C for 2 h and the difference in masses before and after drying was used to calculate the moisture content on wet basis (wb) or as is basis. Measurements were done in triplicate.

2.6. Drying Performance Metrics

2.6.1. Mass Flow Rates

Flow rate or throughput of raw material into the extrusion system was known from the feeder screw calibration. The flow rates of the raw material and product at various other points of the aquatic feed processing system were however unknown and were calculated from the measured moisture content using principles of mass balance as described previously [27]. The following equation for mass balance of solids was employed for this purpose, assuming negligible material losses.
  m i + 1 k g h = m i · 1 − X w i 1 − X w i + 1
where mi is the product mass flow rate (kg/h) at point i of the system and is known; mi+1 is the unknown product mass flow rate (kg/h) at the next point of interest; Xwi and Xwi+1 are moisture contents of the product stream (wet basis fraction) at points i and i + 1, respectively; and location i = 1 represents the feeder screw discharge, i = 2 the extruder discharge, i = 3 the spreader discharge and i = 4 final product discharge from the dryer/cooler system.
The difference between mass flow rates at two consecutive points gives the moisture removal rate from the portion of the system in between, after taking into account any mass flow streams into the system such as water and steam injection. Air flow into and out of the dryer/cooler need not be accounted as they are assumed to be the same and balance or cancel each other out.

2.6.2. Specific Energy Consumption

For quantifying dryer performance, the specific energy consumption (SEC) is a commonly used parameter [29,30,31]. SEC was calculated using the following equation.
S E C M J k g − w a t e r = Q g 1000 · m r
where mr is the rate of moisture removal from the dryer–cooler system (kg/h) calculated from m4-m3 using Equation (1); and Qg is the rate of natural gas energy consumed by the dryer burner (kJ/h). Qg was calculated as shown below, assuming 100% conversion, so it represents a theoretical value.
Q g k J h = V g · H v
where Vg is the volumetric consumption rate of natural gas from the gas meter (ft3/h) and the Hv is the standard volumetric heat value of natural gas (1055 kJ/ft3). A related quantity, which is the gas energy consumption normalized to the throughput or simply the normalized gas energy consumption (Qg,norm) was also calculated as follows.
Q g , n o r m k J k g − r a w = Q g m 1
where m1 is the material flow rate (kJ/kg-raw) at the feeder screw discharge commonly referred to in this study as the throughput.

2.6.3. Energy Efficiency

Energy efficiency (Eenergy) was defined as the fraction of the theoretical natural gas energy expended that was consumed for useful purposes, including heating up the product, dryer and cooler air, and for evaporation of water. Eenergy was calculated using Equation (5).
E e n e r g y % = ( Q 4 − Q 3 ) + ( Q d a , e x − Q d a , i n ) + ( Q c a , e x − Q c a , i n ) Q g · 100
where Q3 = rate of energy (or heat) associated with the product entering the dryer–cooler (kJ/h), Q4 = rate of energy associated with the product exiting the dryer–cooler (kJ/h), Qda,ex = rate of energy associated with dryer exhaust air (kJ/h), Qda,in = rate of energy associated with dryer inlet air (kJ/h), Qca,ex = rate of energy associated with cooler exhaust air (kJ/h) and Qca,in = rate of energy associated with cooler inlet air (kJ/h). Calculations of these energy flows are described below.
The energy flows associated with the product (Q3 and Q4) were calculated using the following equation.
Q i k J h = m i · C p i · T i
where i = 3 and 4 for product entering and exiting the dryer–cooler, respectively; T is the product temperature (40 °C and 26.7 °C, for product inlet and exit, respectively) based on average data for a typical drying experiment for extruded aquatic feed; and Cp is the specific heat of the product, which was calculated as follows.
C p i k J k g ℃ = X w i · C p w + ( 1 − X w i ) · C p s
where Cpw is the specific heat of water (4.18 kJ/kg°C) and Cps is the average specific heat of all non-water components (1.5 kJ/kg°C).
The energy flows associated with all air streams were calculated as follows from enthalpies obtained using an online psychometrics calculator [35].
Q a k J h = m a · h a
where ha is the enthalpy of air stream in kJ/kg obtained from the psychometric calculator using the measured air temperature and humidity; and ma is the mass flow rate of air stream (kg/h) obtained from Equation (9).
m a k g h = ρ a · V a · 0.0283
where ρa is air density obtained from the online psychometric calculator; and Va is the volumetric flow rate of air (ft3/h), the constant 0.0283 (m3/ft3) is a conversion factor.

2.6.4. Evaporation Efficiency

Dryer efficiency based on theoretical energy required to evaporate water calculated using its latent heat of evaporation and expressed as a fraction of the natural gas energy expended is defined as the evaporation efficiency (Eevap). It was calculated from Equation (10) as follows [30,31].
E e v a p % = m r · λ w Q g · 100
where λw is the latent heat of evaporation for water. For the temperature range of 93–115 °C in this study the standard value of 2260 kJ/kg-water for free water at 100 °C was used for simplicity. This was an approximation because free or unbound water is only a fraction of the water in the product and the latent heat of bound water is much higher and evaporation occurs even before the product temperature reaches close to the drying air temperature and continues in the cooler.

2.6.5. Zero Heat Loss Efficiency

A new method was proposed in this study for calculating drying efficiency based on a maximum theoretical throughput in a scenario where there are no heat losses. The various energy flow terms in Equation (5) can rearranged to obtain an energy balance equation for the dryer–cooler after introducing an additional term Qloss to represent total losses as shown below.
Q 3 + Q d a , i n + Q c a , i n + Q g = Q 4 + Q d a , e x + Q c a , e x + Q l o s s
Qloss (kJ/h) represents the rate of ‘non-useful’ energy consumed, including heat absorbed by the dryer frame, heat loss from dryer surfaces to ambient, air and product losses, and also the conversion efficiency of gas into thermal energy. Although not possible in practice, maximum efficiency in the drying process can be obtained if the Qloss is minimized or reduced to zero. This ideal scenario would allow a much higher theoretically maximum throughput (m1′max) and correspondingly maximum water removal rate (mr′max). These were calculated using the following procedure while keeping unchanged the incoming and discharge product moisture contents into the dryer–cooler (Xw3 and Xw4, respectively), corresponding product temperatures (T3 and T4, respectively), gas energy consumption (Qg), air inlet and exhaust temperatures for dryer (Tda,in and Tda,ex, respectively) and cooler (Tca,in and Tca,ex, respectively). Some of these assumptions are of course very simplistic and would not reflect real conditions. Rearranging the terms in Equation (11) under this scenario and putting Qloss = 0, the following is obtained.
( Q 3 ′ − Q 4 ′ ) + Q g = ( Q d a , e x ′ − Q d a , i n ′ ) + ( Q c a , e x ′ − Q c a , i n ′ )
After elaborating each term in the above equation, the dryer energy balance equation assuming zero losses is as follows.
m 3 ′ C p 3 T 3 − m 4 ′ C p 4 T 4 + Q g = ( m h ) d a , e x ′ − ( m h ) d a , i n ′ + ( m h ) c a , e x ′ − ( m h ) c a , i n ′
The above equation was solved iteratively, by incrementally increasing m1′, calculating the corresponding m3′ and m4′ using mass balance, and obtaining the new moisture removal rate mr′ from the difference between m3′ and m4′. This was then used to calculate the new absolute humidity of the cooler and dryer exhaust air flows, and from there the corresponding air enthalpies (ha′) were obtained using the online psychometric calculator mentioned above. The iterative process was stopped when both sides of Equation (13) were equal, and the maximum theoretical throughput (m1′max) thus obtained and the actual throughput (m1) were used to calculate zero-loss efficiency (Ezero-loss) using Equation (14).
E z e r o − l o s s % = m 1 m 1 m a x ′ · 100

2.7. Product Quality Measurements

Extruded product bulk density (g/L) was measured by recording the mass of pellets filling a 1 L cup to the brim. Duplicate measurements were made for both wet product directly off the extruder discharge and dry product after the dryer–cooler. Pellet durability index (PDI) is the standard for measuring the number of intact pellets versus broken pellets, dust and fines that are generated through repeated handling. PDI was tested using a Holmen NPH100 pneumatic tester (Tekpro, Norfolk, VA, USA) as per the method described previously [36]. Pellets were first sieved using a U.S. No. 10 screen to remove broken pellets and fines. One hundred grams of sieved pellets were tested. Samples were placed in the chamber and then air-agitated for 60 s using 70 mbar of pressure. Upon test completion, the fines were automatically sieved from the tested sample. PDI was defined as the percentage of pellets remaining after the conclusion of the test. Three replicate tests were conducted for each treatment.

2.8. Statistical Analysis

Statistical analyses were performed using SAS version 9.4 software (SAS Institute, Inc., Cary, NC, USA). One-way ANOVA (p < 0.05) using the general linear model was employed to determine statistical difference in treatment. Tukey’s post hoc test (p < 0.05) was performed for pairwise comparison between means after a significant ANOVA result.

3. Results and Discussion

3.1. Drying Process and Changes in Pellet Moisture

Aquatic feed pellet moisture content at various stages of drying is shown in Table 1. At the extruder discharge (off-extruder), moisture content ranged from 21.7 to 23.9% wet basis and was lowered slightly to 21.2 to 22.1% wb (off-spreader) in the pneumatic system used for conveying the pellets to the dryer. This reduction was due to surface drying of hot product (~120–135 °C) from the extruder discharge as it came into contact with the air flow at room temperature in the pneumatic system. The product in turn was expected to cool down in this conveying process before entering the dryer.
Table 1. Moisture content (MC; % wet basis) of aquatic feed at various stages of drying as affected by product throughput, dryer air temperature and grain used in formulation.
Subsequent operations involved two-pass hot air drying with a total residence time of 17.7 min followed by single-pass cooling using room temperature air with residence time of 4.5 min, as described earlier. The dryer and cooler inlet air was at room temperature (~27–29 °C) and relative humidity (RH) of 66–83%, with flow rate of the two being 1500 cu-ft/h and 2500 cu-ft/h, respectively, based on average measurements for a typical operation. The dryer exhaust air RH increased slightly from 4.2 to 5.3% as the product throughput increased due to greater amount of moisture removed, as discussed later. The RH decreased from 5.7 to 3.2% as the drying temperature increased from 93 to 115 °C. This is because of higher moisture holding capacity of the air with increase in temperature, as per basic principles of psychometrics. The product temperature measured at the end of the second dryer pass, before passing onto the cooler, ranged from 81.9 to 86.2 °C for the experiment with varying throughput with no particular trend observed. On the other hand, post-dryer product temperature increased from 71.5 to 88.5 °C as the dryer air temperature increased from 93 to 115 °C, which was expected. The post-cooling product temperature was close to room temperature. The cooler exhaust air temperature was unchanged for most treatments at 45 °C but was slightly higher (48 °C) for the treatment with highest drying temperature. The corresponding RH ranged from 25.9 to 32.5%, with the lowest observed for the highest drying temperature of 115 °C and highest for the largest throughput of 121 kg/h. These trends were also consistent with psychometrics, with the drying temperature impacting product temperature in the cooler and consequently the air temperature and the amount of water removed by the air being directly related to the throughput. The data clearly indicated some degree of moisture removal during the cooling process as heat was transferred from the product to the air.
The drying and cooling processes were for the most part treated as one composite unit operation in this study. Henceforth all references to dryer or drying, unless otherwise mentioned, imply the composite drying–cooling operation as both sub-processes contributed to reduction in pellet moisture although the contribution of cooling was likely to be substantially less. After pneumatic conveying, the extruded pellets were delivered to the dryer via the spreader. The final product exiting the dryer–cooler had moisture in the range of 3.5 to 5.0% wb (Table 1), which represented adequate drying for a shelf-stable product. The final moisture was not significantly affected by the throughput and was in a relatively narrow range of 3.9–5.0% wb. However, it appeared that the lower bed depth on the dryer belts at throughputs of 82 and 108 kg/h exposed the product to more air and heat, resulting in a relatively lower moisture content as compared to the higher throughput of 121 kg/h. Similar results were reported previously during drying of extruded aquatic feed (initial moisture 20.8%) designed for catfish using an experimental continuous belt dryer that mimicked commercial drying operations [22]. Post-drying moisture increased from 5.6% to 10.2%, as the product bed thickness on the dryer belts was raised progressively from 20 mm to 60 mm. The slowing of the drying process with increased bed thickness was attributed to increase in air flow resistance due to decreased void space, which in turn led to lowering air flow velocity as also confirmed by computations fluid dynamics based simulations. It should be noted that the differences in final product moisture were more pronounced than in the current study, with the smaller pellet size (~1.2 mm diameter) of the catfish feed as one of the contributing factors due to greater packing density.
Drying temperature had a more consistent and also statistically significant (p = 0.0285) impact on the final product moisture, which decreased from 5.0% at 93 °C to 3.5% at 115 °C. This was expected due to the additional heat input as drying temperature increased and has also been reported previously for drying of extruded aquatic feed using a batch fluidized bed dryer [20] and a continuous belt dryer [19]. The type of grain (sorghum versus wheat) used in the aquatic feed diets did not have a significant impact on the off-cooler or final product moisture, which were very similar (4.1 versus 3.9% wb, respectively). Further analysis of drying performance, including product quality, energy consumption and efficiency, revealed some differences between the grains used. This is discussed in more detail in Section 3.2, Section 3.3 and Section 3.4.

3.2. Impact on Pellet Quality

Aquatic feed bulk density data are shown in Table 2. Off-extruder or OE bulk density ranged from 400 to 453 g/L, while off-dryer or OD bulk density ranged from 432 to 462 g/L. Typically, during production of aquatic feed via extrusion the bulk density is inversely related to throughput. This relationship is observed when all other processing factors are constant and is primarily because of higher pressure generated at the die with increase in throughput, which in turn leads to greater transformation of starch in the product matrix and higher expansion. In this study, however, the preconditioning intensity or the ratio of steam injection to throughput was not kept constant, which contributed to lack of a significant impact of throughput on off-extruder bulk density. This was actually desirable as varying bulk density and porosity of the product can impact drying characteristics and efficiency and thus obfuscate the effect of product bed depth on dryer belts that was the primary intent of having throughput as an independent variable. On the other hand, off-dryer bulk density was significantly impacted by throughput (p = 0.0044). For the least bed depth, corresponding to the lowest throughput of 82 kg/h, highest drying rate was expected. This led to substantial shrinkage of the pellets as drying proceeded and correspondingly more than 6% increase in off-dryer bulk density. For the highest throughput (and bed depth) this shrinkage effect was muted due to slower drying rates, and the overall reduction in mass of pellets due to moisture loss was predominant leading to only a slight decrease in off-dryer bulk density. For the intermediate throughput, the shrinkage and moisture loss effects balanced each other, leading to almost no change in bulk density during drying. The pellet shrinkage effect during drying can also be clearly seen from bulk density trends with respect to drying air temperature. The off-extruder bulk density was obviously not significantly impacted by drying temperature, as the drying is a post-extrusion phenomenon. However, drying temperature had a significant effect on off-dryer bulk density. Drying rates increased with air temperature and so did the degree of pellet shrinkage, leading to greater increase in off-dryer bulk density even though substantially higher moisture loss (kg/kg) was also observed. A similar relationship between shrinkage and drying temperature was reported previously in experiments using a custom built lab-scale batch dryer, with decrease in final diameter of extruded trout feed pellets as temperature progressively increased from 70 to 140 °C [9]. This trend was attributed to higher phase mobility in the pellets when there was a greater difference between product and glass transition temperatures.
Table 2. Bulk density (g/L) of aquatic feed before and after drying as affected by product throughput, dryer air temperature and grain used in formulation.
At similar throughput and drying temperature, the sorghum-based pellets experienced greater shrinkage and increase in bulk density than wheat, due to less transformation of starch during processing in the case of former and consequently lower water binding and water retention potential and higher drying rates, stemming from the differences in endosperm of the two grains [8,37]. The binding force between the internal water and various components of the pellet including a substantial amount of cereal grain can also impact the energy required to dry a unit mass of the product [29]. This is discussed later in the context of drying energy and efficiency.
Pellet durability index data are shown in Table 3. Product throughput, which controlled the bed depth on dryer belts and in turn affected the drying rate, had a significant impact on PDI. The lowest throughput of 82 kg/h led to the least PDI of 82.8%. This was the treatment that corresponded to the smallest bed depth and highest drying rate, which led to the highest degree of pellet shrinkage as discussed above and caused fractures in the matrix and reduction in overall binding and integrity. In contrast, the higher throughputs of 108–121 kg/h led to slower drying rates and better pellet integrity with PDI of 78.8–88.0%. This was contrary to one of the hypotheses of this research. A previous investigation on drying of extruded aquatic feed also reported positive effects of reduced drying rates, achieved via higher air humidity, on pellet durability [17]. Increase in drying air temperature from 93 to 115 °C led to higher drying rates and increased shrinkage as was seen from bulk density data in Table 2. Consequently, a slight decreasing trend was observed in PDI (from 84.3 to 83.6%), which was however not statistically significant. This was again in contrast to one of the proposed hypotheses. It is possible that the negative effect of pellet shrinkage on durability was offset by increased starch transformation at higher temperatures. Differing results were reported in the same study that was described above, with increase in dryer air temperature from 70 °C to 140–150 °C leading to significant improvement in pellet durability and reduction in fines [17]. The much higher pellet size (6–8 mm diameter versus < 3 mm in the current research) and thus lower specific surface area was probably the cause as it led to relatively lower drying rates and thus the impact of starch transformation outweighed the shrinkage effects. However, other reasoning was offered by the authors for linking drying temperature and pellet shrinkage to durability, which neither fit the observations of this research nor considered the role of starch. It can also be seen from Table 3 that under similar drying conditions of throughput (or bed depth) and air temperature, sorghum-based pellets had slightly less PDI than wheat. Although these results are not statistically significant, the weaker matrix of the former due to differences in the endosperm as mentioned above could be a potential reason.
Table 3. Pellet durability index (PDI) of aquatic feed as affected by product throughput, dryer air temperature and grain used in formulation.

3.3. Water Removal and Energy Consumption During Drying

Water removal and energy consumption data are shown in Table 4. As throughput increased from 82 to 121 kg/h, the overall rate of moisture removal during the drying and cooling process rose proportionately from 16.1 to 25.2 kg/h while the moisture loss in the pellets was almost unchanged (0.203–0.207 kg-water/kg-raw). As discussed earlier, the aquatic pellets reached their equilibrium moisture content of 5% or lower for all treatments. Thus, the almost linear relationship between moisture removal rate and throughput and little change in pellet moisture loss was expected. In other words, the drying was adequate irrespective of the throughput and the product bed depth. Here it is important to make a distinction between the overall moisture removal rate in the dryer versus the localized drying rate. The latter actually decreased due to increase in throughput and higher bed depth. It was interesting to note, as throughput increased from 82 to 108 kg/h the normalized gas energy consumption (Qg,norm) changed from 27,735 kJ/kg-raw to 21,034 kJ/kg-raw representing a 24% reduction. Correspondingly, the specific energy consumption or the energy required per unit mass of water removed decreased substantially from 136.3 to 102.5 MJ/kg-water (25% reduction). The SEC relates to drying costs and is also an indirect measurement of drying efficiency. While drying efficiency parameters are discussed in Section 3.4, it is clear from these data that, contrary to one of the hypotheses of this study, increase in throughput to a certain extent led to lowering the cost of drying. One important reason is that the gas energy is used to not only heat up the product and the air, but also the dryer itself. The latter is a substantial heat sink and requires a relatively fixed energy input at a given drying temperature, irrespective of the throughput. As throughput further increased from 108 to 121 kg/h, the normalized energy consumption and SEC were relatively constant. Even though the lower energy consumption effect plateaued at higher throughput, increasing the dryer load actually was not detrimental. Thus, it is possible that the dryer was not being used at its maximum potential even at the highest throughput and that parameters such as dryer size, residence time and air velocity, and temperature could be further optimized.
Table 4. Moisture removal metrics, normalized gas energy or energy consumption per kg of raw material processed (Qg,norm) and specific energy consumption per kg of water removed (SEC) as affected by product throughput, dryer air temperature and grain used in formulation.
Raising the drying temperature from 93 to 104 °C led to only slight increases in rate of moisture removal from 15.9 to 16.2 kg/h and pellet moisture loss from 0.201 to 0.205 kg/kg-raw, but very little to no change in total Qg,norm and SEC. The moisture removal rate and pellet moisture loss increased further (17.2 kg/h and 0.217 kg/kg-raw, respectively) as drying temperature increased to 115 °C leading to the lowest final pellet moisture content of 3.5% of all treatments, but Qg,norm increased by 32% and SEC by 25% due to the greater heating requirement. These trends indicated that drying at temperatures higher than 93 °C at best did not improve cost effectiveness and was even detrimental at the highest temperature in the range studied. It should be noted though that dryer loading and other factors can make a difference to this outcome and that the temperature study was conducted at a low throughput of 79 kg/h. The SEC data in this research were comparable in order of magnitude to the range of 50–250 MJ/kg-water reported previously for drying of different agricultural products using convective hot air drying [29,30]. However, the drying temperatures in these studies were relatively very low (40–70 °C). Also, different methods were used to calculate specific energy consumption including normalizing with either mass of products being dried or mass of water removed, accounting also for energy consumption of blowers, and use of inlet and exhaust air enthalpies rather than actual fuel or power used by heaters for thermal energy consumption calculations.
There was no substantial difference in the rate of overall moisture removal and pellet moisture loss between sorghum and wheat-based aquatic feed pellets. However, the sorghum-based feed required 12.6% more normalized gas energy (27,735 versus 24,624 kJ/kg-raw) and had 13.6% higher specific energy consumption (136.3 versus 120.0 MJ/kg-water). This was possibly because of the higher localized drying rate in the case of sorghum-based aquatic feed as mentioned before, which led to greater demand for heating to maintain the air temperature at 104 °C.

3.4. Dryer Efficiency Metrics

Data related to various drying efficiency metrics are shown in Table 5. The efficiency related to energy usage (or energy efficiency, Eenergy) is based on the fraction of theoretical gas energy consumption that was used for heating up the product, dryer air and cooler air, and for evaporation of water. Only a small percentage (11.9–14.3%) of the thermal energy going into the drying process was used up in these operations. The remainder 85.7–88.1% represented the fraction of thermal energy used up in heating the dryer frame, air and product losses, and also the conversion efficiency of gas into thermal energy. No particular trends were observed with respect to throughput, but Eenergy reduced from 13.4 to 14.3% in the case of lower drying temperatures of 93–104 °C to 11.9% for the highest temperature. This was probably due to more thermal energy being used up in heating the dryer frame. In the literature as well, the sensible heat used in heating the drying chamber, heat loss through the dryer walls by conduction from the high-temperature interior to the lower temperature ambient and air loss from leaky dryer structures are the some of the reasons put forward for low drying efficiency [29].
Table 5. Drying efficiency metrics as affected by product throughput, dryer air temperature and grain used in formulation. Eenergy = energy efficiency; Eevap = evaporation efficiency; and Ezero-loss = theoretical throughput calculation-based zero heat loss efficiency.
Dryer efficiency based on theoretical energy required to evaporate water based on its latent heat of evaporation (or evaporation efficiency, Eevap) was very low in general (1.51–2.24%), which indicated that only a miniscule fraction of the thermal energy input was used for the actual drying or removal of water, while the rest was used up for other operations linked to heating up air streams, product and the dryer frame, and various losses including those linked to the conversion efficiency of gas to thermal energy. Eevap increased from 1.60 to 2.24% with increase in throughput from 82 to 121 kg/h, but decreased from 1.84 to 1.51% with increase in drying temperature from 93 to 115 °C. This mirrored the trends observed for the specific energy consumption that were discussed previously. Therefore, higher throughput led to better evaporation efficiency, but increasing the drying temperature did not have benefits. The exhaust air from the drying operation had a relative humidity of 5.7% for the treatment with a drying temperature of 93 °C and a relative humidity of 5.2% for the treatment with throughput of 121 kg/h. Clearly, the exhaust air was far from being spent even at the lowest drying temperature or at the highest throughput and had sufficient capacity for more moisture removal. The dryer lacked air recirculation control based on humidity sensors. The continuous air intake and exhaust were wasteful, and it is evident that the dryer had the capacity to handle even higher throughputs. The drying efficiency based on maximum theoretical throughput, assuming no losses and no heat absorbed by the dryer frame (Ezero-loss), ranged from 2.40 to 3.56%, and the trends were similar to those discussed for Eevap.
In general, the various metrics discussed here pointed to lower dryer efficiency for sorghum-based aquatic feed as compared to wheat-based feed. This trend was similar to that of SEC. It is a result of higher gas consumption for the former due to differences in starch transformation and faster localized evaporation rate as discussed previously.
Specific energy consumption, which was described in Section 3.3, is a commonly used metric in research related to dryer performance [29,30,32]. However, SEC by itself is not a measure of efficiency. The data on energy efficiency is a unique contribution of this study as it is rarely reported. Evaporation efficiency of convective hot air drying of chamomile flowers was reported to range from 2 to 7% [30]. The efficiency increased as dryer temperatures increased from 40° to 60 °C. This was in contrast with the current results, which indicated the lowest evaporation efficiency at the highest drying temperature of 115 °C. The chamomile study was conducted in a batch dryer and the total drying time decreased with increasing temperature thus influencing the results, whereas in conveyor-belt drying the residence time is a function of belt speed. Moreover, the efficiency of chamomile drying was based on theoretical calculation of energy required to heat air and not the actual fuel consumption for heating as in the current study. The former approach ignores various heat losses and heat sinks, leading to higher evaporation efficiency as compared to this research. This shortcoming and also the large difference in drying temperature ranges that were investigated in the two studies also possibly contributed to the divergence in trends.
Estimation of efficiency allows benchmarking of drying performance and its improvement. As was discussed, the dryer was operated below capacity for the range of throughputs studied (~80–120 kg/h). Increase in throughput was shown to improve the drying efficiency at a constant temperature. Unlike the observations in this study, increasing the drying temperatures can also lead to improvement in efficiency and reduction in specific energy consumption if used in combination with even higher throughputs [32]. Another way to enhance drying efficiency would be to reduce the energy demand for heating air through exhaust air recirculation and heat recovery [29,32]. The low RH of exhaust air in this study as discussed above further affirms the feasibility of this approach. While the duct work and other aspects of the dryer were designed for air recirculation, humidity sensor-based controls would make it more effective along with optimal spacing and placement of the air inlet, exhaust and heat source that would allow effective routing of a portion of the exhaust air to the heat source. It is clear from this study that characterizing of efficiency of the pellet drying process provides further insights into better equipment design and processing conditions than modeling studies focused on drying kinetics alone that are more common [9,38,39,40].

4. Conclusions

Hot air conveyor-belt drying of extruded aquatic feed was investigated, as impacted by different independently controlled variables. Both dryer efficiency and also product quality parameters were evaluated, which has not been done previously in an integrated fashion for agricultural products including aquatic feed. Higher throughput led to better pellet durability, decreased specific energy consumption and improved drying efficiency. On the other hand, increasing the drying temperature led to poorer drying efficiency, without any improvement in pellet quality. Strategies for improvement of drying efficiency were discussed, including using the dryer at a higher and more optimal capacity. A better sensor-based control system was also proposed for partial recirculation of air, to allow it to be exhausted closer to its saturation point, improve heat recovery and increase dryer efficiency. The composition of aquatic feed pellets was also shown to have an impact, with the sorghum-based feed having lower drying efficiency than wheat-based feed. This was attributed to endosperm related differences between the two grains and binding of water but can be addressed by optimizing the drying conditions including the throughput. The study highlights that effective drying is not merely a process of water removal but a sophisticated control problem that requires a nuanced understanding of material science, thermodynamics, and process engineering. In the future, other key aspects of drying of aquatic feed should be addressed as well as including moisture heterogeneity and over-drying, both of which can impact pellet quality as well as drying efficiency.

Author Contributions

Conceptualization, T.G. and S.A.; methodology, T.G., E.W.M. and S.A.; formal analysis, T.G. and S.A.; investigation, T.G.; resources, S.A.; data curation, T.G. and S.A.; writing—original draft preparation, T.G.; writing—review and editing, T.G., E.W.M. and S.A.; supervision, S.A.; project administration, S.A.; funding acquisition, S.A. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors would like to thank Anders Haubjerg and Keith Erdley at Wenger Manufacturing, Inc. for their valuable insights and suggestions on technical aspects of industrial dryers.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. FAO. The State of World Fisheries and Aquaculture 2022; Food and Agriculture Organization: Rome, Italy, 2022. [Google Scholar] [CrossRef] [Scilit]
  2. FAO. The State of World Fisheries and Aquaculture 2024; Food and Agriculture Organization: Rome, Italy, 2024. [Google Scholar] [CrossRef] [Scilit]
  3. Partos, L. FAO: Fish feed costs to remain high. SeafoodSource. 2 September 2010. Available online: https://www.seafoodsource.com/news/aquaculture/fao-fish-feed-costs-to-remain-high (accessed on 15 April 2024).
  4. Alltech. Agri-food outlook 2023: Excerpts. Global Feed Survey Archives. Poultry TRENDS. 15 March 2023. Available online: https://www.poultrytrends.in/tag/global-feed-survey/ (accessed on 15 April 2024).
  5. Alavi, S.; Ambrose, K. Particulate flow and agglomeration in food extrusion. In Production, Handling and Characterization of Particulate Materials; Merkus, H.G., Meesters, G.M.H., Eds.; Springer International Publishing: Cham, Switzerland, 2015. [Google Scholar]
  6. Lazou, A. Food extrusion: An advanced process for innovation and novel product development. Crit. Rev. Food Sci. Nutr. 2022, 64, 4532–4560. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  7. Ali, S.S.A.; Ambasankar, K.; Nandakumar, S. Effect of Varying Levels of Moisture Content on the Extrusion Properties of Asian Seabass (Lates calcarifer) Feed in Twin Screw Extruder. Poult. Fish. Wildl. Sci. 2021, 9, 225. [Google Scholar]
  8. Graff, T.; Davis, D.A.; Alavi, S. Impact of raw material particle size on processing, physical quality and in-vivo performance of grain sorghum and wheat-based extruded feed for tilapia Oreochromis niloticus. Animals 2026, 16, 858. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  9. Haubjerg, A.F. Ensuring Technical Product Quality in the Energy Efficient Hot Air Drying of Extruded Fish Feed. Ph.D. Thesis, Center for Energy Infomatics, Maersk McKinney Moller Institute, University of Southern Denmark, Odense, Denmark, 2016. [Google Scholar]
  10. Adedeji, A.E.; Olalusi, A.P.; Olayanju, T.M.A.; Erinle, O.C. Effect of drying parameters on the physical properties of extruded fish feed. IOP Conf. Ser. Earth Environ. Sci. 2020, 445, 012011. [Google Scholar] [CrossRef] [Scilit]
  11. Haubjerg, A.F.; Simonsen, B.; Jørgensen, B.N.; Veje, C.T. Ensuring technical product quality in energy efficient hot air drying of extruded fish feed: Definition of an industrial research project. In Proceedings of the 18th International Drying Symposium (IDS 2012), Xiamen, China, 11–15 November 2012. [Google Scholar]
  12. Cnossen, A.G.; Siebenmorgen, T.J.; Yang, W.; Bautista, R.C. An Application of Glass Transition Temperature to Explain Rice Kernel Fissure Occurrence during the Drying Process. Dry. Technol. 2001, 19, 1661–1682. [Google Scholar] [CrossRef] [Scilit]
  13. Dai, J.-W.; Rao, J.-Q.; Wang, D.; Xie, L.; Xiao, H.-W.; Liu, Y.-H.; Gao, Z.-J. Process-Based Drying Temperature and Humidity Integration Control Enhances Drying Kinetics of Apricot Halves. Dry. Technol. 2015, 33, 365–376. [Google Scholar] [CrossRef] [Scilit]
  14. Mercier, S. Contribution of external resistance to mass transfer during pasta drying. Dry. Technol. 2019, 37, 59–68. [Google Scholar] [CrossRef] [Scilit]
  15. Kasapis, S. Glass Transition Phenomena in Dehydrated Model Systems and Foods: A Review. Dry. Technol. 2005, 23, 731–757. [Google Scholar] [CrossRef] [Scilit]
  16. Li, R.; Lin, D.; Roos, Y.H.; Miao, S. Glass Transition, Structural Relaxation and Stability of Spray-Dried Amorphous Food Solids: A Review. Dry. Technol. 2019, 37, 287–300. [Google Scholar] [CrossRef] [Scilit]
  17. Haubjerg, A.F.; Veje, C.T. An experimental approach towards increasing mechanical durability of extruded fish feed in the drying process. Dry. Technol. 2019, 37, 1418–1426. [Google Scholar] [CrossRef] [Scilit]
  18. Kaddour, O. Factors affecting the floating aquatic feed pellets quality. Misr J. Agric. Eng. 2018, 35, 1019–1038. [Google Scholar] [CrossRef] [Scilit]
  19. Ogunnaike, F.; Olalusi, A. Modelling kinetics of extruded fish feeds in a continuous belt dryer. Turk. J. Agric. Eng. Res. 2021, 2, 289–297. [Google Scholar] [CrossRef]
  20. Oduntan, O.B.; Oluwayemi, B.J. Optimization of a clay-slate fluidized bed dryer for production of fish feed. Malays. J. Sustain. Agric. 2021, 5, 104–110. [Google Scholar]
  21. Ogunnaike, F. Responce surface of drying parameters on some physical properties related to floatability of extruded fish feeds. Turk. J. Agric. Eng. Res. 2023, 4, 178–190. [Google Scholar] [CrossRef] [Scilit]
  22. Zhang, Q.; Huang, M.; Wang, J.; Zhang, P.; Chen, B.; Shi, Z.; Zhang, J. Investigation on airflow distribution under different feed thickness combined CFD modeling and experimental verification. Dry. Technol. 2021, 39, 306–323. [Google Scholar] [CrossRef] [Scilit]
  23. Zhang, Q.; Liu, X.; Chen, Y.; Zhang, P.; Durgawati, D.; Sutar, P.P.; Gong, J. CFD modeling and experimental verification for the uniformity of air velocity in an industrial dryer. Dry. Technol. 2025, 43, 260–275. [Google Scholar] [CrossRef] [Scilit]
  24. Riaz, M.N. Extruders in Food Applications; CRC Press: Boca Raton, FL, USA, 2000. [Google Scholar]
  25. Monti, M.; Gibson, M.; Loureiro, B.A.; Sá, F.C.; Putarov, T.C.; Villaverde, C.; Alavi, S.; Carciofi, A.C. Influence of dietary fiber on macrostructure and processing traits of extruded dog foods. Anim. Feed. Sci. Technol. 2016, 220, 93–102. [Google Scholar] [CrossRef] [Scilit]
  26. Baller, M.A.; Pacheco, P.D.G.; Vitta-Takahashi, A.; Putarov, T.C.; Vasconcellos, R.S.; Carciofi, A.C. Effects of thermal energy on extrusion characteristics, digestibility and palatability of a dry pet food for cats. J. Anim. Physiol. Anim. Nutr. 2021, 105, 76–90. [Google Scholar] [CrossRef] [Scilit]
  27. Chen, Y.; Graff, T.; Cairns, A.C.; Griffin, R.; Siliveru, K.; Pezzali, J.G.; Alavi, S. Use of Insect Meals in Dry Expanded Dog Food: Impact of Composition and Particulate Flow Characteristics on Extrusion Process and Kibble Properties. Processes 2025, 13, 2083. [Google Scholar] [CrossRef] [Scilit]
  28. Akgün, M.; Kontaş, E. The Effect of the Periodic Drying Method on the Drying Time of Hazelnuts and Energy Utilization. Foods 2024, 13, 901. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  29. Nwakuba, N.R.; Asoegwu, S.N.; Nwaigwe, K.N. Energy consumption of agricultural dryers: An overview. Agric. Eng. Int. CIGR J. 2016, 18, 119–132. [Google Scholar]
  30. Motevali, A.; Minaei, S.; Banakar, A.; Ghobadian, B.; Khoshtaghaza, M.H. Comparison of energy parameters in various dryers. Energy Convers. Manag. 2014, 87, 711–725. [Google Scholar] [CrossRef] [Scilit]
  31. Mouhoubi, K.; Boulekbache-Makhlouf, L.; Guendouze-Bouchefa, N.; Freidja, M.L.; Romero, A.; Madani, K. Modelling of drying kinetics and comparison of two processes: Forced convection drying and microwave drying of celery leaves (Apium graveolens L.). The Annals of the University Dunarea de Jos of Galati. Fascicle VI-Food Technol. 2019, 43, 48–69. [Google Scholar] [CrossRef] [Scilit]
  32. Ziegler, T.; Jubaer, H.; Teodorov, T. Bottlenecks in continuous hops drying with conveyor-belt dryer. Dry. Technol. 2022, 40, 2598–2616. [Google Scholar] [CrossRef] [Scilit]
  33. Graff, T.M. Extruded Aquatic Feed—Impact of Ingredients and Processing on Physical Quality of Feed, Animal Performance and Environmental Footprint. Doctoral Dissertation, Kansas State University, Manhattan, KS, USA, 2025. [Google Scholar]
  34. AACC International. Approved Methods of Analysis, 11th ed.; AACC International: St. Paul, MN, USA, 2010. [Google Scholar]
  35. Tornado. Free Online Psychrometric Calculator. 2025. Available online: http://www.hvac-calculator.net/index.php?v=2 (accessed on 31 October 2025).
  36. AFIA. Feed Manufacturing Technology V; Schofield, E.K., Ed.; American Feed Industry Association: Arlington, VA, USA, 2005. [Google Scholar]
  37. Graff, T.; Ngo, T.; Davis, D.A.; Alavi, S. Evaluation of Different Levels of Process Energy in Sorghum and Wheat-Based Diets on the Growth Performance of Pacific White Shrimp, Litopenaeus vannamei. Aquac. Nutr. 2024, 2024, 1985325. [Google Scholar] [CrossRef] [Scilit]
  38. Lambert, C.; Cartailler, J.; Rouchouse, S.; Almeida, G.; Courtois, F. Characterization and modeling of cooling and drying of pellets for animal feed. Dry. Technol. 2018, 36, 255–266. [Google Scholar] [CrossRef] [Scilit]
  39. Wang, W.; Chen, J.; Jin, N.; Wang, H.; Wang, L.; Wu, J. Thin-layer drying model, drying rate, and effective water diffusion coefficient of pelleted feed. Int. J. Chem. Eng. 2024, 2024, 7092556. [Google Scholar] [CrossRef] [Scilit]
  40. Wang, W.; Wu, J.; Zou, F.; Wang, H.; Wang, L. Numerical simulation of deep bed cooling drying process of pellet feed based on non-equilibrium model. Appl. Sci. 2025, 15, 9445. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.