Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Drying Technology

An International Journal

ISSN: 0737-3937 (Print) 1532-2300 (Online) Journal homepage: https://www.tandfonline.com/loi/ldrt20

Energy efficient spray drying by increased feed dry


matter content: investigations on the applicability
of Air-Core-Liquid-Ring atomization on pilot scale

Marc O. Wittner, Heike P. Karbstein & Volker Gaukel

To cite this article: Marc O. Wittner, Heike P. Karbstein & Volker Gaukel (2019): Energy efficient
spray drying by increased feed dry matter content: investigations on the applicability of Air-Core-
Liquid-Ring atomization on pilot scale, Drying Technology, DOI: 10.1080/07373937.2019.1635616

To link to this article: https://doi.org/10.1080/07373937.2019.1635616

Published online: 01 Jul 2019.

Submit your article to this journal

Article views: 25

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=ldrt20
DRYING TECHNOLOGY
https://doi.org/10.1080/07373937.2019.1635616

Energy efficient spray drying by increased feed dry matter content:


investigations on the applicability of Air-Core-Liquid-Ring atomization on
pilot scale
Marc O. Wittner, Heike P. Karbstein, and Volker Gaukel
Karlsruhe Institute of Technology, Institute of Process Engineering in Life Sciences, Chair of Food Process Engineering,
Karlsruhe, Germany

ABSTRACT ARTICLE HISTORY


Spray drying is the main process for production of powdered products from liquid sources. Received 16 October 2018
One way to reduce energy consumption of spray drying processes is preconcentration of Revised 6 June 2019
feed liquids by using more energy efficient techniques, such as membrane reactors or falling Accepted 20 June 2019
film evaporators. As feed viscosity increases with dry matter content, efficient atomization at
KEYWORDS
high viscosities is necessary. In contrast to standard pressure swirl (PS) atomizers, Air-Core- Spray drying; sustainability;
Liquid-Ring atomization may be an option. Experiments on a pilot scale show its high ACLR; high viscosity;
potential for application in spray drying processes as feed dry matter content may be maltodextrin; whey protein
increased up to 50% compared to PS atomization. The theoretical energy savings associated concentrate
with the 50% feed solids increment is about 29%. Further investigations are required
to assess the complexity, the practicality and the potential benefits for industrial-
scale operations.

1. Introduction atomization for all known atomization techniques.[5]


To achieve sufficient drying, all droplets must be
Water removal is very energy consuming in spray dry-
ing processes, especially since internal energy recovery small enough to dry within the residence time inside
is restricted.[1,2] Moreover, the powder throughput of the spray dryer. Incompletely dried particles often
a spray dryer is limited by its specific water evapor- show an enhanced stickiness, leading to process dis-
ation rate at given process conditions. As a result, turbances due to material depositions within the dryer
industrial applications aim to increase the feed dry or to unforeseen changes of powder properties during
matter content to the highest possible value prior to storage.[6] Moreover, the droplet size distribution dir-
the actual spray drying process.[3] According to a ectly influences the size distribution of resulting pow-
model calculation on industrial spray drying by Fox der particles.[7]
et al.,[4] an increase in feed dry matter content by 1% For droplet formation, generally wheel or nozzle
point leads to a decrease in thermal energy consump- type atomizers are used in spray drying processes.[3]
tion of the spray dryer by 3.8% and a decrease of total Concerning nozzle type atomizers, pressure-swirl noz-
energy consumption of 2.5% when the energy for pre- zles are most commonly used in spray drying proc-
concentration in an evaporator is taken into account. esses on industrial scale. However, for pressure swirl
Alternatively, an increase in powder production cap- (PS) nozzles the highest processable viscosity is con-
acity of 4.5% is possible. Although thermal losses may siderably low, compared to pneumatic atomizers.[8] In
lead to reductions of these values in practical applica- practical use, the exact value of the maximum pro-
tions, increases in feed dry matter contents promise cessable viscosity is dependent of the required atom-
reductions in energy consumption during spray drying ization result, atomizer geometry, feed liquid
processes.[3] The major drawback of increased feed properties, and applied liquid pressure.
dry matter contents is that viscosity increases sharply In contrast, external mixing pneumatic atomizers
for many liquids. An increase in viscosity complicates allow atomization of highly viscous liquids as the

CONTACT Marc O. Wittner marc.wittner@kit.edu Institute of Engineering in Life Sciences, Chair of Food Process Engineering, Karlsruhe Institute of
Technology, Kaiserstrasse 12, Karlsruhe, 76131 Germany.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/ldrt.
ß 2019 Taylor & Francis Group, LLC
2 M. O. WITTNER ET AL.

total dry matter contents were performed in the study


presented. Based on previous studies on ACLR atom-
ization performance,[12,23] a viscosity range between
0.03 and 0.39 Pas was chosen for the spray drying
experiments for this study. The corresponding dry
matter range is between 30 and 47% for the model
system used. For atomization, gas pressures of 0.6 and
0.8 MPa are applied. Since an increase in gas pressure
Figure 1. Different types of internal mixing pneumatic atom- enhances spray stability with increasing viscosity,[12]
izers: (a) Y-Jet atomizer, (b) Effervescent Atomizer, and (c) Air- we hypothesize that the stability of subsequent drying
Core-Liquid-Ring atomizer.
processes also increases. Here spray stability is defined
by the fluctuation of the SMD over time, while pro-
kinetic energy of compressed gas is used for droplet
cess stability and product quality of the drying process
disintegration.[9] Another advantage is that liquid flow
are assessed through observation of material deposits
rate and droplet size distribution can be adjusted
within the dryer. Resulting powder characteristics,
independently of each other in a relatively wide range.
such as particle size distribution, water content, water
However, high gas consumption rates of external mix-
activity and bulk density were investigated. In the
ing pneumatic atomizers restrict economic operation
second part of the study, potential energy savings of a
on a larger scale. For ExMP atomizers, ALRs from 1
theoretical application of ACLR atomization in indus-
to 10 are usually applied.[3] As a result, this type of
trial spray drying processes involving high feed dry
atomizer is mostly used in laboratory or pilot scale
matter contents were estimated, based on the above
spray dryers.[3]
mentioned model calculation of Fox et al.[4]
As an alternative, internal mixing atomizers are
reported to atomize high viscous feeds at low gas con-
sumption rates with typically applied ALRs below a 2. Materials and methods
value of 0.2.[10–12] In this group of atomizers, com-
2.1. Model system
pressed gas and feed liquid are brought into contact
in a mixing zone and leave the atomizer through an Aqueous solutions of maltodextrin (CDry MD01958,
exit orifice. For gas injection, several geometries have Overlack GmbH, M€ onchengladbach, Germany) were
been suggested so far (e.g.[5,11,13–16]). In Figure 1, a used as the model system. In order to ensure uniformity
selection of three different types of IMP atomizers is of surface tension for all solutions, a concentration of
shown. These are (a) Y-Jet atomizer; (b) Effervescent 2% powdered whey protein concentrate WPC80
Atomizer (EA); and (c) Air-Core-Liquid-Ring (ACLR) (Sachsenmilch GmbH, Leppersdorf, Germany) was
atomizer. Depending on the injection geometry and added. As a result, total dry matter contents of 30, 40,
process conditions used, different two phase flow pat- 45, and 47% were adjusted in a combination of WPC 80
terns are developed inside the exit orifice.[17] It is and maltodextrin. For feed liquid preparation, the pow-
known that unregulated internal flow patterns, such as der moisture contents (dry basis) of 4.50% (MD) and
those occurring in a Y-Jet atomizer or in an EA at 4.59% (WPC) were taken into account. The values were
low ALR and high liquid viscosity, directly lead to validated repeatedly during storage by means of the
temporal instabilities in the spray produced.[18–20] method described in Section 2.3. The solutions were
Since stable spray droplet sizes are required in spray characterized by viscosity (m), surface tension (r), and
drying processes, the ACLR atomizer was devel- density (q). All measurements were performed in tripli-
oped[21] based on investigations of internal flow pat- cate at a temperature of 25  C. This temperature was
terns of an EA.[18,22] For ACLR atomizers, a high selected as the reference for room temperature,[12] as
temporal stability of the Sauter mean diameter (SMD) the feed was not preheated prior to injection into the
was found in short time measurements of spray drop- spray dryer (see Section 2.2.3). Mean values and relative
let sizes with liquid feed viscosities of up to uncertainty values ur(c) of these parameters, as well as
0.69 Pas.[12] However, the gas pressure applied must the total dry matter composition of the solutions used,
be increased as viscosity levels rise in order to ensure are summarized in Table 1. Viscosity measurements
temporal stability of the SMD.[12] were performed using a rotary rheometer (MCR 101/
To investigate the general applicability of ACLR 301, Anton Paar GmbH, Graz, Austria) with coaxial
atomization in spray drying processes, pilot scale cylinder geometry (CC27). Within the measurement
spray drying experiments with feed liquids of different range of shear rates between 1 and 1000 s1, all
DRYING TECHNOLOGY 3

Table 1. Physical properties of the investigated solutions at 25  C: viscosity (m), surface tension (r), density (q), and correspond-
ing uncertainty values ur (c).
ctotal (% d.b.) cWPC 80 (% d.b.) cMD (% d.b.) m (Pas) ur(m) (%) r (Nm–1) ur(r) (%) q (kgm–3) ur(q) (%)
30 2 28 0.03 2.33 0.049 1.08 1139 0.07
40 2 37 0.12 7.23 0.049 0.27 1185 0.74
45 2 43 0.22 2.87 0.049 0.57 1211 0.69
47 2 45 0.39 4.61 0.049 1.03 1234 0.06
Concentrations of total dry matter content ctotal, maltodextrin (cMD) and WPC 80 (cWPC 80) are given as dry base values (d.b.).

solutions demonstrated Newtonian behavior. Surface 20 kg h–1 was used. The total height of the drying
tensions and densities were measured using a Wilhelmy chamber was 3 m, including a cylindrical part with a
plate system (DCAT 21, DataPhysics Instruments height of 2 m and a diameter of 1.5 m. For powder
GmbH, Filderstadt, Germany). separation, a cyclone was used that was connected to
the bottom of the drying chamber with a flexible
pipe of 100 mm in diameter. The drying processes
2.2. Process equipment and procedure
were operated in co-current flow at an air inlet tem-
2.2.1. ACLR atomization perature of 180  C and an air flow rate of 500 kgh–1.
In an ACLR atomizer, the compressed atomization The resulting outlet temperature was adjusted to
gas is injected into the feed liquid stream through a 90 ± 3  C by variation of the feed liquid flow in a
capillary closely before the exit orifice. This results in range between 15 ± 2 L h–1. The feed liquid was sup-
a characteristic annular two phase flow pattern gener- plied at room temperature of 23 ± 2  C. All drying
ated inside the exit orifice. To date, no information trials were performed in triplicate.
on the process dependent liquid film thickness is
available. However, current research seeks to estimate
2.3. Powder characterization
these parameters. For further information on operat-
ing principal and geometry, refer to the literature on Powders were characterized by particle size distribu-
ACLR atomization.[8,12,21,23–25] In the atomizer used, tion, moisture content, water activity and bulk density.
the mixing area had a length of 2.4 mm. Diameter and Particle size distributions were measured by laser dif-
length of the exit orifice, as well as diameter of the fraction spectroscopy (Horiba LA950, Retsch
gas capillary, were 1.5 mm each. Technology GmbH, Haan, Germany). In this measure-
ment device, a gas flow with maximum pressure of
2.2.2. Media supplies 0.4 MPa was applied for dispersion of particles in an
For atomization, the model system was supplied by an air filled measurement chamber. Moisture contents
eccentric screw pump (MD 006-12, seepex GmbH, were calculated by weight loss after oven drying at
Bottrop, Germany). The flow rate could be adjusted 105  C to constant mass. The applied drying tempera-
within a range of 5–20 Lh1 and was measured by a ture is a standard temperature for food and dairy
flow meter (VSI 044/16, VSE GmbH, Neuenrade, powders such as WPC80.[26] This temperature is also
Germany). Compressed air was used as the atomiza- well below the onset temperature of decomposition of
tion gas. It was supplied by a compressor (Renner MD.[27] Water activities were measured by a dedicated
RSF-Top 7.5, Renner GmbH, G€ uglingen, Germany) measuring instrument (AquaLAB 4TE, METER Group
that delivered a maximum gas pressure of 0.9 MPa. Inc., Pullman, USA). Bulk density was measured
The gas pressure was adjusted to 0.6 and 0.8 MPa according to DIN-ISO-697, 1984. All measurements
using a pressure regulator. The gas volume flow were performed in triplicate.
through the atomizer’s mixing chamber resulted from
gas pressure and liquid flow rate and was measured 2.4. Calculation of applied atomization energy
by a gas flow meter (ifm SD6000, ifm electronic,
Essen, Germany). Gas flow rates are given at standard The energy consumption of the atomization step is
conditions (273.15 K, 0.1013 MPaabs) and varied calculated by the applied energy density per mass of
between 4.9 and 7.7 Nm3h1. resulting powder Em, powder according to the following
Equation (1). In this calculation, the moisture content
2.2.3. Pilot scale spray drying after drying (c moisture) is included:
A one-stage pilot scale spray dryer (Werco SD20, Hans EV
Em;powder ¼ (1)
G. Werner Industrietechnik GmbH, Reutlingen, qL  ðctotal þ cmoisture Þ
Germany) with maximum water evaporation capacity of
4 M. O. WITTNER ET AL.

Here, Ev is the volumetric energy density applied to total energy consumption of a hypothetical industrial
the liquid flow. Density and dry matter content of the powder production process is calculated by adding
liquid medium are qL and ctotal, respectively. together the energy consumptions of liquid concentra-
According to St€ahle et al.,[8] EV is calculated for tion, atomization and drying. Of course, these assump-
ACLR atomization as follows: tions must be adapted carefully whenever we aim to
pG;abs estimate the energy consumption of a specific process.
pL  QL þ MG  R  T  ln pG;0
Ev ¼ (2)
QL
3. Results and discussion
The energy delivered by the liquid flow is calculated
by multiplying the relative liquid pressure pL and the 3.1. Influence of feed dry matter content on
volume flow of the liquid QL. The energy calculation of process stability and powder characteristics
the gas flow is based on isothermal gas compression. Spray drying processes are highly dependent on the
This calculation includes the mass flow of the gas MG, combination of atomization and drying conditions.
temperature T, ideal gas constant R, absolute gas pres- When droplets are insufficiently atomized for the given
sure pG,abs and atmospheric gas pressure pG,0. drying conditions, the residence time in the spray dryer
According to St€ahle et al.,[8] we can assume an equilib- is too short to convert the largest droplets into dry par-
rium of pressure between gas and liquid in the mixing ticles. Incompletely dried particles tend to stick to the
chamber of the ACLR atomizer, as the pressure inside dryer walls, especially at the bottom of the drying
the mixing chamber is dominated by the pressure loss chamber. As a consequence, severe material deposits
over the exit orifice. Accordingly, pL is replaced by rela- can be created over time. Thus for a general evaluation
tive gas pressure pG in further calculations. In PS atom- of process stability, the development of material depos-
ization, the atomization energy is completely delivered its at the bottom of the drying chamber was visually
by the liquid flow. As a result, the following equation assessed after one hour of operation. In Figure 2, typ-
can be used for calculating EV in PS atomization:[8] ical depictions illustrate the dependency of applied gas
P pL  QL pressure pgas and feed dry matter content ctotal. The
Ev ¼ ¼ ¼ pL (3) hole, visible in the center of the images, represents the
QL QL
exit of the drying chamber. Here dried particles and
Here P is the power applied by the used pump, pL humidified drying air are supposed to leave the drying
is the relative liquid pressure and QL is the liquid flow area in the direction of the cyclone for powder separ-
rate. According to this equation, EV equals the liquid ation. The exit hole and attached piping have a diam-
pressure pL in this case. eter of 100 mm. In case of ctotal ¼ 47% and pgas ¼ 0.6
no stable operation was possible over the required
2.5. Theoretical energy consumption of operation time of one hour. Therefore, no image is
concentration and spray drying presented for this process condition in Figure 2.
From the given images, we see that material depos-
Energy consumption for liquid concentration and dry- its qualitatively increase as dry matter content
ing in industrial scale spray dryers is estimated based increases and gas pressure decreases. These tendencies
on the literature data cited in the study presented here. were present in all repetitions performed. Moreover,
Although there are several more advanced models for it was observed that material deposits generated at
prediction of dryer specific energy consumption avail- ctotal ¼ 40% (pgas ¼ 0.6 and 0.8 MPa), as well as at
able in the literature (e.g.[28–31]), the assumptions made ctotal ¼ 45% and pgas ¼ 0.8 MPa appeared powdery
here are kept simple, as the intention is to provide an and dry. In contrast, deposits generated at ctotal ¼
idea of potential energy savings through the application 45% and pgas ¼ 0.6, as well as at ctotal ¼ 47% and
of ACLR atomization in industrial spray drying proc- 0.8 MPa appeared very moist and sticky. These charac-
esses. According to Fox et al.,[4] the following assump- teristics can be explained by a decrease of spray stability
tions are made: For liquid concentration, a single stage as dry matter contents increase or gas pressure
falling film evaporator with thermal vapor recompres- decreases.[12] Here, the residence time in the drying
sion is used. Here 0.5 kg of steam per kg of water evap- chamber is apparently not sufficient to dry the largest
oration are needed. The energy consumption of the generated spray droplets. Resulting large particles
spray dryer is 2 kg of steam per kg of water evaporation. remain wet and stick to the dryer wall. Once this adhe-
For production of 1 kg of steam an energy equivalent of sion process starts, a sticky layer of material develops
2.25 MJ is required. Based on these considerations, the that holds back an increasing amount of particles. As
DRYING TECHNOLOGY 5

Figure 2. Material deposits at the bottom of the drying chamber, in dependency of feed dry matter content (ctotal ¼ 40, 45, and
47%) and atomization gas pressure (pgas ¼ 0.6 and 0.8 MPa).

well, smaller particles that are already dried can adhere the dryer, the moisture content of the powder produced
to this sticky layer. As a consequence, a small number from ctotal ¼ 47% at 0.8 MPa is unexpectedly low, com-
of large droplets can begin a kind of chain reaction, pared to the value of the powder produced from ctotal ¼
leading to fast development of severe deposits inside 45% at 0.6 MPa. Since wet and sticky deposits were
the dryer. According to results on ACLR atomization found for both processes, similar moisture contents for
performance,[12,23] the described increase in mean both processing conditions were expected. The lower
droplet size and distribution width due to increasing moisture content of samples at ctotal ¼ 47% may result
viscosity can be compensated by an increase in specific from potentially lower specific humidity of the inlet dry-
atomization energy. This can be done by increasing the ing air or increased outlet temperatures. This should be
gas pressure at a constant liquid flow rate.[12] investigated in further studies.
Consequently, production of non-sticky particles is Water activity resembles the tendencies found from
possible for dry matter contents of up to 45% under moisture content measurements. However, the differ-
the applied drying conditions when gas pressure is ence in water activities at ctotal ¼ 45% is not statistic-
increased from 0.6 to 0.8 MPa. Compared to the refer- ally significant. As water activity and final moisture
ence point of PS atomization at a dry matter content content are directly connected to the drying condi-
of 30% in the model system used, an increase by 15 tions, it can be assumed that lower values can be
percentage points can be achieved using ACLR atom- reached in an industrial process with a longer drying
ization, even on a pilot scale. time. For bulk density, no significant differences were
In addition to process stability, the powders pro- observed between the processes (0.38 – 0.41 kgL1).
duced were characterized by particle size distribution, Moreover, no correlations between processing condi-
moisture content, water activity and bulk density. In tions and particle size distributions were found. For
Figure 3, values of moisture content (A), water activity all processes x50,3 values of 93 ± 14 mm were measured.
(B), and bulk density (C) are presented depending on It must be considered that particle size distributions
feed dry matter content and applied gas pressure. are fractionated by the process, because droplets larger
An increase in feed dry matter content from 40 to than the process specific threshold value for dryable
45% significantly increases the moisture content of the particles are held back in deposits in the drying cham-
powder produced from 5.7 to 8.4% at pgas ¼ 0.8 MPa. A ber. Nevertheless, high x90,3-values of up to 777 mm
further increase of ctotal to 47% results in a moisture were found in combination with high uncertainty val-
content of 8.9%; however, the increase is not statistically ues of up to 48% in repeated measurements of the
significant, compared to ctotal ¼ 45%. In case of ctotal ¼ same sample. It is most likely that occasional large
40 and 45%, a decrease in applied gas pressure to values of x90,3 are based on the development of
0.6 MPa results in a significant increase of moisture con- agglomerates during powder transport, separation or
tent to 7.5 and 9.8%, respectively. Taking into account storage. This assumption is supported by spray
the above-described observations of material deposits in droplet size measurements under similar conditions
6 M. O. WITTNER ET AL.

Figure 3. Moisture content xwb (A), water activity aw, (B) and bulk density qb, and (C) of produced powders in dependency of
feed dry matter content (ctotal ¼ 40, 45, and 47%) and applied gas pressure (pgas ¼ 0.6 and 0.8 MPa).

Figure 4. Energy consumption of process steps evaporation, atomization and spray drying, as well as of the complete process in
dependency of the feed dry matter content supplied to the spray dryer.

(ctotal ¼ 47%, pgas ¼0.8 MPa, QL ¼ 20 Lh1), where efficient than PS atomization.[8] Nevertheless, energy
x90,3 did not exceed values of 270 ± 25 mm.[23] consumptions of ACLR atomization is only 4–7 times
higher than PS atomization. This is well below the
3.2. Potential energy savings by applying ACLR level of external mixing pneumatic atomizers, which
atomization in industrial spray drying processes are reported to consume up to 20 times the energy of
PS atomizers.[6] Based on the trials performed we can
Like all pneumatic atomization techniques, ACLR
conclude that the application of ACLR atomization
atomization has been reported to be less energy
DRYING TECHNOLOGY 7

Table 2. Energy densities EV of feed solution and Em powder of theoretically can be achieved. This value is composed
produced powder (cmoisture ¼ 5%) an correlated uncertainty of a reduction of the energy consumption of the spray
values ur(c) in dependency of feed dry matter content ctotal.
drying step from 9.0 to 5.3 MJ kg powder–1 and an
ctotal (% d.b.) EV (kJL–1) Em powder (kJkg–1) ur(c) (%)
increase in energy consumption of the preconcentra-
30 70 174 4.1
40 78 148 7.6 tion from 0 to 1.1 MJ kg powder–1. The difference in
45 86 144 11.1 energy consumption of the atomization step is consid-
47 86 135 2.7
erably small, as it decreases from 0.17 to 0.14 MJ kg
–1
powder (see Table 2). The decrease in total energy
offers the possibility of increasing ctotal from 30% for consumption equals a reduction of approximately 2%
PS atomization to 45%, even in pilot scale spray dry- (0.17 MJ kg powder–1) per percentage point of increase
ing. This increase in dry matter content equals an in feed dry matter content. This value is well in the
increase in viscosity from 0.03 to 0.22 Pas. In order range of the energy consumption of ACLR atomiza-
to evaluate the potential energy savings of an indus- tion. Consequently, application of ACLR atomization
trial spray drying process while using ACLR atomiza- can lower the total energy consumption of spray dry-
tion for increased feed dry matter contents, Figure 4 ing processes from the first percentage point of
shows the specific energy consumption of the process increase in feed dry matter content over the specific
steps involved. These steps include concentration, threshold value of PS atomization for the high viscous
atomization and drying. Theoretical energy consump- model system.
tion of concentration and drying on an industrial
scale are taken from the literature[4] (see Section 2.5.).
4. Conclusion
For estimation of the energy consumption of ACLR
atomization, volumetric energy density Ev and Energy In our study, the applicability of ACLR atomization
density per mass of produced powder Em powder are in spray drying of highly concentrated feeds was
calculated based on the pilot scale trials performed investigated on a pilot scale. In the spray dryer, a
(see Section 2.4). Results are shown in Table 2. For maltodextrin based solution with a maximum viscos-
calculation of Em powder, a moisture content of cmoisture ity of 0.22 Pas (feed dry matter content of 45%) was
¼ 5% is assumed for the powders produced. For processed. An increase in gas pressure from 0.6 to
ACLR atomization, a gas pressure of 0.8 MPa is 0.8 MPa led to a clear reduction of material deposits
assumed in all calculations. We make these assump- in the dryer. This is most likely based on an
tions because a conservative approximation of energy increased spray stability. In the hypothetical indus-
savings is desired and lower gas pressures or higher trial spray drying process we evaluated, an increase
moisture contents would result in lower values of in dry matter content from 30 to 45% led to possible
Em powder. energy savings of 29%. This value consists of a the-
Finally, all process steps are added to create a total
oretical reduction of the specific energy consumption
energy consumption of the drying process. A feed
of the drying step by 3.7 MJkgpowder1, while the
with ctotal ¼ 30% is considered as the starting point of
energy consumption of the assumed concentration
concentration, since this value is considered as the
step is 1.1 MJkgpowder1. The reduction in energy
highest possible concentration for PS atomization in
consumption of the atomization step
the high viscous model system.
(–0.03 MJkgpowder1) is considerably small.
It is obvious that energy consumption of evapor-
ation increases with increasing ctotal, as more water Moreover, it was shown that the application of
has to be removed from the feed concentrate before ACLR atomization can theoretically lower the total
entering the spray dryer. In consequence, energy con- energy consumption of spray drying processes from
sumption of spray drying decreases with increasing the first percent point of increase in feed dry matter
ctotal. As more energy can be recovered in the evapor- content above the threshold value of PS atomization.
ator than in the spray dryer, the absolute gradient is Moisture content and water activity should be
higher in the case of the spray dryer. The energy con- decreased to meet industrial demands. For the same
sumption of atomization is very low compared to reason, longer drying times in industrial processes
dehydration (concentration and drying). would be beneficial. Therefore, further investigations
Comparing total energy consumptions at ctotal ¼ on the scale-up potential of ACLR atomizers are
30% (A) and at ctotal ¼ 45% (B), a reduction in energy needed for successful implementation in industrial
consumption of 29% (9.2to 6.5 MJ kgpowder–1) spray drying processes.
8 M. O. WITTNER ET AL.

Acknowledgments Application in Spray Drying Processes of Highly


Concentrated Feeds. Chem. Eng. Process. 2018, 128,
The authors further express their thanks to Holger 96–102. DOI: 10.1016/j.cep.2018.04.017.
Schubert, Fabian Wendelberger and Andrea Butterbrodt for [13] Sovani, S. D.; Sojka, P. E.; Lefebvre, A. H.
experimental support. Effervescent Atomization. Prog. Energy Combust. Sci.
2001, 27, 483–521. DOI: 10.1016/S0360-
1285(00)00029-0.
Funding [14] Mlkvik, M.; St€ahle, P.; Gaukel, V.; Zaremba, M.;
The IGF Project AiF 18299 N of the FEI was supported via Jedelsky, J.; Jıcha, M.; Dancova, P.; Vıt, T.
AiF within the program for promoting the Industrial Performance of Twin-Fluid Atomizers for
Collective Research (IGF) of the German Ministry of Atomization of Viscous Solutions. EPJ Web Conf.
Economic Affairs and Energy (BMWi), based on a reso- 2015, 92, 2052.
lution of the German Parliament. [15] Jedelsky, J.; Jıcha, M. Spray Characteristics and
Liquid Distribution of Multi-Hole Effervescent
Atomisers for Industrial Burners. Appl. Therm. Eng.
References 2016, 96, 286–296. DOI: 10.1016/j.applthermaleng.
2015.11.079.
[1] Baker, C. G. J.; McKenzie, K. A. Energy
[16] Zaremba, M.; Weiß, L.; Maly, M.; Wensing, M.;
Consumption of Industrial Spray Dryers. Drying
Jedelsky, J.; Jıcha, M. Low-Pressure Twin-Fluid
Technol. 2005, 23, 365–386. DOI: 10.1081/DRT-
Atomization: Effect of Mixing Process on Spray
200047665.
Formation. Int. J. Multiphase Flow 2017, 89,
[2] Kemp, I. C. Reducing Dryer Energy Use by Process
277–289. DOI: 10.1016/j.ijmultiphaseflow.2016.10.
Integration and Pinch Analysis. Drying Technol.
2005, 23, 2089–2104. DOI: 10.1080/ 015.
[17] Kim, J. Y.; Lee, S. Y. Dependence of Spraying
07373930500210572.
[3] Masters, K. Spray Drying in Practice; Performance on the Internal Flow Pattern in
SprayDryConsult International: Charlottenlund, Effervescent Atomizers. Atomization. Spray. 2001,
Denmark, 2002. 11, 735–756.
[4] Fox, M.; Akkerman, C.; Straatsman, H.; Jong, P. d. [18] St€ahle, P.; Gaukel, V.; Schuchmann, H. P. Influence
Energy Reduction by High Dry Matter of Feed Viscosity on the Two-Phase Flow inside the
Concentration and Drying. New Food 2010, 6, Exit Orifice of an Effervescent Atomizer and on
60–62. Resulting Spray Characteristics. Food Res. Int. 2015,
[5] Omer, K.; Ashgriz, N. Spray Nozzles. In Handbook 77, 55–62. DOI: 10.1016/j.foodres.2015.04.047.
of Atomization and Sprays; Ashgriz, N., Ed.; [19] Avulapati, M. M.; Ravikrishna, R. V. An
Springer: New York, NY, 2011; pp 497–597. Experimental Study on Effervescent Atomization of
[6] Mujumdar, A.S., Ed. Handbook of Industrial Drying, Bio-Oil Fuels. Atomization. Spray. 2012, 22,
4th ed.; CRC Press: Boca Raton, FL, 2015. 663–685. DOI: 10.1615/AtomizSpr.2012006482.
[7] Schuck, P.; Mejean, S.; Dolivet, A.; Beaucher, E.; [20] Mlkvik, M.; Zaremba, M.; Jedelsky, J.; Wigley, G.
Famelart, M. H. Pump Amperage: A New Method Investigation of the spray generated by a Y-jet atom-
for Monitoring Viscosity of Dairy Concentrates izer. Presented at the Proceedings of 20th
before Spray Drying. Lait 2005, 85, 361–367. DOI: International Conference Engineering Mechanics,
10.1051/lait:2005014. Brno, CZ, 2014, pp 404–407.
[8] St€ahle, P.; Schuchmann, H. P.; Gaukel, V. [21] St€ahle, P.; Gaukel, V.; Schuchmann, H. P.
Performance and Efficiency of Pressure-Swirl and Comparison of an Effervescent Nozzle and a
Twin-Fluid Nozzles Spraying Food Liquids with Proposed Air-Core-Liquid-Ring (ACLR) Nozzle for
Varying Viscosity. J. Food Process. Eng. 2017, 40, Atomization of Viscous Food Liquids at Low Air
12317 Consumption. J. Food Process. Eng. 2017, 40, 12268.
[9] Lefebvre, A. H. Atomization and Sprays; [22] St€ahle, P.; Gaukel, V.; Schuchmann, H. P.
Hemisphere: Philadelphia, PA, 1989. Investigation on the Applicability of the Effervescent
[10] Jedelsky, J.; Jicha, M. Unsteadiness in Effervescent Atomizer in Spray Drying of Foods: Influence of
Sprays: A New Evaluation Method and the Influence Liquid Viscosity on Nozzle Internal Two-Phase Flow
of Operational Conditions. Atomization. Spray. 2008, and Spray Characteristics. J. Food Process. Eng. 2015,
18, 49–83. DOI: 10.1615/AtomizSpr.v18.i1.20. 38, 474–487. DOI: 10.1111/jfpe.12178.
[11] Mlkvik, M.; St€ahle, P.; Schuchmann, H. P.; Gaukel, [23] Wittner, M. O.; Karbstein, H. P.; Gaukel, V.
V.; Jedelsky, J.; Jicha, M. Twin-Fluid Atomization of Pneumatic Atomization: Beam Steering Correction
Viscous Liquids: The Effect of Atomizer in Laser Diffraction Measurements of Spray Droplet
Construction on Breakup Process, Spray Stability Size Distributions. Appl. Sci. 2018, 8, 1738. DOI: 10.
and Droplet Size. Int. J. Multiphase Flow 2015, 77, 3390/app8101738.
19–31. DOI: 10.1016/j.ijmultiphaseflow.2015.06.010. [24] Kleinhans, A.; Georgieva, K.; Wagner, M.; Gaukel,
[12] Wittner, M. O.; Karbstein, H. P.; Gaukel, V. Spray V.; Schuchmann, H. P. On the Characterization of
Performance and Steadiness of an Effervescent Spray Unsteadiness and Its Influence on Oil Drop
Atomizer and an Air-Core-Liquid-Ring Atomizer for Breakup during Effervescent Atomization. Chem.
DRYING TECHNOLOGY 9

Eng. Process. 2016, 104, 212–218. DOI: 10.1016/j.cep. [29] Schuck, P.; Jeantet, R.; Tanguy, G.; Mejean, S.; Gac,
2016.03.011. A.; Lefebvre, T.; Labussiere, E.; Martineau, C. Energy
[25] Kleinhans, A.; Hornfischer, B.; Gaukel, V.; Consumption in the Processing of Dairy and Feed
Schuchmann, H. P. Influence of Viscosity Ratio and Powders by Evaporation and Drying. Drying
Initial Oil Drop Size on the Oil Drop Breakup dur- Technol. 2015, 33, 176–184. DOI: 10.1080/07373937.
ing Effervescent Atomization. Chem. Eng. Process. 2014.942913.
2016, 109, 149–157. DOI: 10.1016/j.cep.2016.09.006. [30] Silva, CRd.; Martins, E.; Silveira, A. C. P.; Sime~ao,
[26] Schuck, P.; Jeantet, R.; Dolivet, A. Analytical M.; Mendes, A. L.; Perrone, I. T.; Schuck, P.;
Methods for Food and Dairy Powders; Wiley-
Carvalho, A. F. d. Thermodynamic Characterization
Blackwell: Hoboken, NJ, 2012.
of Single-Stage Spray Dryers: Mass and Energy
[27] Saavedra-Leos, Z.; Leyva-Porras, C.; Araujo-Dıaz,
S. B.; Toxqui-Teran, A.; Borras-Enrıquez, A. J. Balances for Milk Drying. Drying Technol. 2017, 35,
Technological Application of Maltodextrins accord- 1791–1798. DOI: 10.1080/07373937.2016.1275675.
ing to the Degree of Polymerization. Molecules 2015, [31] Domınguez-Ni~ no, A.; Cant u-Lozano, D.; Ragazzo-
20, 21067–21081. DOI: 10.3390/molecules201219746. Sanchez, J. A.; Andrade-Gonzalez, I.; Luna-Solano,
[28] Bimbenet, J.-J.; Schuck, P.; Roignant, M.; Brule, G.; G. Energy Requirements and Production Cost of the
Mejean, S. Heat Balance of a Multistage Spray-Dryer: Spray Drying Process of Cheese Whey. Drying
principles and Example of Application. Lait 2002, Technol. 2018, 36, 597–608. DOI: 10.1080/07373937.
82, 541–551. DOI: 10.1051/lait:2002031. 2017.1350863.

You might also like