Aquacultural Engineering 85 (2019) 106–113

Contents lists available at ScienceDirect

Aquacultural Engineering
journal homepage: www.elsevier.com/locate/aque

Water circulation induced by mechanical aerators in a rectangular vessel for T

shrimp aquaculture
⁎
Tomoaki Itanoa, , Taishi Inagakia, Choji Nakamuraa, Ren Hashimotoa, Naohiro Negoroa,
Jinsuke Hyodob, Syuta Hondaa
a
Department of Pure and Applied Physics, Faculty of System Engineering, Kansai University, Suita, Japan
b
MMA Solutions Co. Ltd.

A R T I C LE I N FO A B S T R A C T

Keywords: A water streamer was designed for the purpose of enhancing cost-efficient circulation between the water surface
Inland aquaculture and bottom of shrimp aquaculture ponds. We took direct measurements of the water current field induced by the
Paddle-wheel aerator designed aerator in a large rectangular reservoir of dimensions (L)50 × (W)19 × (D)1.3 m3 and compared the
Circulation results with those of a standard Taiwanese paddle-wheel aerator. Vertical circulation between the surface and
Water current field
the bottom induced by the paddle wheel aerator was less than that by the designed aerator. Furthermore, the
Dissolved oxygen
paddle-wheel aerator consumed more electric power than the designed aerator. The structures of water current
Turbulent mixing
induced by the two aerators are elucidated, and the advantages and disadvantages of the aerators are discussed.
Auxiliary employment of the designed aerators may contribute to delivery of high DO water throughout the
pond, which is more efficient than a paddle wheel aerator.

1. Introduction resulting in eutrophication. As a result, residential areas surrounding

aquaculture ponds may face environmental pollution. Thus, many
Shrimp farming vs catches. Aquaculture has become an important practical issues regarding the ecological and economical sustainability
source of food and commercial products in the world today. On behalf of shrimp aquaculture still remain unsolved.
of the United Nations Food and Agriculture Organization, Gillet (2008) Aeration. Unlike marine aquaculture in a finite space surrounded by
reported that, “Today, world annual production of shrimp, both capture an open environment, such as rivers or seas, inland aquaculture is op-
and farmed, is about 6 million tons. Currently, over 40% of world erated in an isolated vessel or coastal pond with little connection to the
shrimp production is from farming ⋯.” In general, aquaculture is outer environment. In such an isolated space, the biological condition
classified into marine and inland aquaculture, of which the latter is spontaneously tends to extremely differ from the natural environment.
mainly employed for shrimp aquaculture. In fact, inland aquaculture Thus, the shortage of dissolved oxygen (DO) concentration in water for
(farming) is flourishing, especially in tropical and semi-tropical areas in aquatic creatures to breathe is one of the technical issues regarding the
the Far-east, Southeast Asia, and Latin America. Inland aquaculture has biological water condition in inland aquaculture for shrimp. Intensive
been supporting the recently rising global consumption of shrimp and methods are commonly practiced in inland aquaculture, where the
prawns, instead of wild shrimp captures by means of trawling fisheries, shrimp concentration is extraordinarily dense, compared to the natural
the ratio of which in total production has fallen since 1990. While density. In most cases, additional enforcement of water circulation is
shrimp aquaculture has brought about increased income for developing required in order to increase DO in water, which is called “aeration”.
countries in the abovementioned areas, it has been accompanied by the Improved aeration can ensure commercial and eco-friendly production
deforestation of natural mangroves for the construction of shrimp ponds of shrimp through aquaculture, as well as increase the production ca-
and the salination of groundwater and agricultural land (Murai, 2007). pacity of aquaculture ponds in cases where DO is a vital factor.
Demolishing mangrove forests, which serve as nurseries of post-larval Moreover, an excellent aeration system is expected to induce an op-
shrimp and other creatures, is likely to decrease near-shore fish and timum and sustainable state in aquaculture ponds, such as “aqua-
shrimp catches. Uncontrolled pursuit of commercial profits from shrimp ponics”, which implies a sustainable synthetic food chain in an isolated
aquaculture sometimes lead to inadequate wastewater treatment, aquacultural system involving bacteria, algae, and aquatic plants, as

⁎
Corresponding author.
E-mail address: [email protected] (T. Itano).

https://doi.org/10.1016/j.aquaeng.2019.03.006
Received 28 December 2018; Received in revised form 13 March 2019; Accepted 27 March 2019
Available online 29 March 2019
0144-8609/ © 2019 Elsevier B.V. All rights reserved.
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

well as aquatic animals.

Empirical Research. Research and innovations in the last decades
have provided fundamental understanding of aeration that has im-
proved the ability of primitive aeration systems to produce strong
currents with high oxygen transfer efficiencies, which increase aqua-
cultural productivity. However, aquaculture has been inherently asso-
ciated with traditional methodologies, which have been improved
gradually and empirically thus far. Thus modern aeration systems are
still not the focus for achieving optimum conditions, and the im-
portance of improving the aeration system is probably not understood
sufficiently by most aquacultural farmers and equipment manu-
facturers, as pointed out by Lawson (1997). The difficulty is also at-
tributed partly to the fact that the aquacultural aeration system is in-
fluenced by chemical, biological, and meteorological factors with less-
repeatability, which is critical for scientific investigations. Here, we
speculate the process of atmospheric oxygen being dissolved and de-
livered to aquatic creatures at the bottom, from the viewpoint of fluid Fig. 1. Paddle-wheel aerators and centrifugal water stirrers employed in a
dynamics. commercial shrimp pond at Karawang, Indonesia, in September 2015 during
Diffusion and advection. If a water body is in equilibrium with at- the dry season. All the paddle-wheel aerators in this pond (one of them is
mospheric oxygen at the surface in contact with air, there is no net captured at the left bottom corner) cease in the control experiment for the
transfer of oxygen between the atmosphere and water. However, when comparison of aerators.
water is undersaturated with oxygen, oxygen from air will dissolve in
water, and the reverse is true when water is supersaturated with variation of DO measured in aquacultural ponds. We performed com-
oxygen. The atmospheric solubility of oxygen is mainly determined by mercial shrimp aquaculture at Aquaculture Business Center (ABC)
atmospheric temperature, pressure, and gas composition, where de- Karawang, West Java, Indonesia (Fig. 1). Time variation of DO was
crease in temperature and increase in atmospheric pressure cause a taken from a shrimp aquaculture in September 2015 during the dry
corresponding increase in the amount of oxygen dissolvable in a unit season. The DO measurements were taken using optical DO meters
volume of water. The increase of DO is also caused by increase of (RINKO ID, JFE Advantech Co., Ltd, Hyogo, Japan). The optical DO
air–water contact interface and exposure time, which may be artificially meter works on the principle of the dependency of the emission lifetime
controlled by employing mechanical aerators. Mechanical aerators of light excited at the sensing membrane on oxygen concentration
splash a large amount of water near the surface into the air while re- around the membrane. Thus, such an optical DO sensor requires no
leasing air bubbles under water. In this manner, they may produce minimum inflow velocity, which is advantageous compared to common
foamy wave crests with air entrainment and microscopic splashes at the galvanic DO sensors that consume oxygen on the membrane and exhibit
water surface around the aerators, significantly increasing net air–water some dependency on flow current. Shrimp aquaculture was practiced in
contact interface. The instantaneous pressure may overcome the satu- two similar aquacultural ponds (1.2 m average water depth) with dif-
rated vapor pressure to enhance DO locally. Even if water is macro- ferent types of mechanical aerators. DO meters were positioned at a
scopically at rest, the oxygen dissolved at the surface may be diffused depth of 0.1 m from the water bottom at two horizontal locations: one is
towards the bottom by thermal molecular motion. This process may be near the center of the pond, and the other is ashore at 3 m from the
termed “oxygen diffusion” from the surface to the bottom. short side of the pond of 42 m × 122 m. Two distinct mechanical
Here, we note that aquacultural shrimps do not have any swim aerators, paddle wheel (PW) and centrifugal water stirrer (CWS), which
bladder, and they mostly crawl on the pond bottom using their legs and are explained later, were placed at the center of each pond.
tail. Therefore, the environment where they thrive may be considered a As shown in Fig. 2, all the measured DO values increased during the
two-dimensional surface at the bottom and not a three-dimensional daytime and decreased at night. This is because oxygen produced by
volumetric space. Thus, in aquaculture ponds for shrimp, the con- algae photosynthesis during daytime is consumed at night. In general, 2
sumption of oxygen occurs mainly at the bottom, and mechanical mg/L of DO is fatal for living creatures including shrimp and one of the
aerators for aquaculture are required to deliver large amounts of water curves arrived at the fatal value from midnight to early morning. Note
containing high DO concentrations from the surface to the bottom. The that algae has a significant influence on DO value in aquaculture ponds.
delivery can be performed through water circulation between the sur- Such uncontrollable factors that are dependent on each pond probably
face and the bottom and can be largely enhanced by turbulent, swirling reduced the average of DO values in PW pond taken on the second
motions. This process may be termed “oxygen advection” from the measurement in Fig. 2(right) compared to those in CWS in average. It
surface to the bottom in the pond, the timescale of which is fairly should be emphasized that a couple of DO values in the CWS pond do
shorter than the diffusion process. not change significantly during the day compared to the PW pond
Production, consumption, and realization of equilibrium. It should be through the measurements in a couple of days. In the daytime, DO
emphasized that algae have a large influence on the amount of DO in values in CWS are less than those in the PW pond. As shown later, a
aquacultural ponds because they produce oxygen during the daytime paddle-wheel type of aerator works at O(103)∼O(104) W (see Cancino,
and consume oxygen at night. Additionally, they develop volume- 2004), which is over 10 times the running cost of the centrifugal water
trically in the pond and increase DO, but they also suppress vertical stirrer. It appears unclear why the DO value in the pond with circula-
water circulation for oxygen advection. If oxygen production and con- tion induced by a much more cost-effective aerator is higher than that
sumption by algae or shrimps are absent, the distribution of DO prob- induced by the other. This is the principal motivation of the present
ably remains unchanged to a large extent after saturation. Oxygen study.
transfer in aquaculture ponds involves various physical and biological We considered that one of the reasons is a difference in the dis-
processes: dissolution, diffusion, advection, production, and consump- tribution of water currents induced by the aerators, which contributes
tion. To understand the entire oxygen transfer process in aquacultural largely to the delivery of DO throughout the pond. Although there have
ponds, the entire region of aquacultural ponds, including shrimps and been several studies on DO concentration (Moulick et al., 2002;
algae, should be comprehensively investigated. Cancino et al., 2004; Brown et al., 2016), most of them are either
DO measurement in Karawang. Here, we show an example of time phenomenological or macroscopic. To the best of our knowledge,

107
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

Fig. 3. A paddle-wheel aerator TA-55H in action; the aerator is held by four

tethers at the center of Katayama civic outdoor pool (September 2016, copied
from Itano et al. (2018)).

devices with either rotating vertical or horizontal shafts with paddles or

screws to circulate water in the pond. Water is broken into droplets for
the purpose of mixing water as well as increasing the DO concentration
of water. The standard evaluation procedure of efficiencies for several
types of aerators proposed by Colt and Orwicz (1991) provides evidence
that, among mechanical aerators, the “paddle-wheel aerator” is rela-
tively efficient at circulating water than other types of aerators.
Paddle-wheel aerators consist of a rotating horizontal shaft, “hub”,
driven by either electric motor or combustion engine, and several
paddles attached on a wheel. The assembly of the hub and paddle is
called a “drum”. Just as a sightseeing side-wheel boat plows a river, the
paddles of the aerator over the pond surface produce a jet-like water
downstream. Fig. 3 shows a commercial small paddle-wheel aerator,
TA-55H, manufactured by Team Aqua Co. Ltd., Taiwan. Similar aera-
tors have been widely adopted in aquacultural ponds to cultivate var-
ious types of fishes or shellfishes (Boyd, 1998), as well as in shrimp
cultivation in Indonesia, Thailand, and Taiwan. In the present study, we
installed the TA-55H aerator in a relatively large rectangular reservoir
and measured the velocity field of the water stream induced by the
Fig. 2. Time variation of DO measured in the aquacultural ponds. DO meters aerator.
were set at two locations, ashore and at the center of a pond; The thick curves in
The plan and side views of TA-55H are shown in Fig. 4. The
the figure represent the paddle-wheel (PW) pond, and dashed curve represents
streamwise and spanwise extents of the aerator are 1635 mm and 1744
the centrifugal water stirrer (CWS) pond. (top) Sep. 21–22, 2015 (bottom) Sep.
22–23, 2015.
mm, respectively. In the side view, the submerged part of the aerator
(shaded, at about 100 mm depth) corresponds to the lower part of the
float, which supports the mechanical and electrical assemblies, such as
previous studies are not based on a fluid mechanical perspective, except motor, reducer, and shaft, with a couple of paddle-wheels on the sur-
for a few numerical studies (Peterson et al., 2000; Pfeiffer et al., 2007; face of the water. The drum can be lowered deeper into the water by
Yamayoshi et al., 2008; Huggins et al., 2004). In the present study, we adjusting the float height to increase water stream while decreasing DO
took direct measurements of the water velocity field induced by two in water, and vice versa. According to Lawson (1997), paddle-wheel
distinct mechanical aerators located in a rectangular reservoir, using an aerators have specific optimum drum depth and rotational speed, but in
electro-magnetic velocimeter. The structures of the water current in- practical use, most aerators operate at a paddle depth of 100–150 mm
duced by the two aerators are elucidated, and the advantages and and drum rotational speed of 80∼90 rpm under an electric power re-
disadvantages of the aerators are discussed. Note that we restricted quirement on the order of 1000 W. In the present study, the motor of
ourselves in the measurement of fluid motion. The measurement of DO TA-55H was driven by alternating current with the three phase 200 V
field, which probably requires the existence of living creatures such as class, which was boosted by a 750 W class general-purpose, frequency-
shrimps at the bottom consuming some amount of oxygen delivered by adjustable inverter (FR-D710W-0.1K, Mitsubishi Electric, Tokyo,
aerators, will be investigated in future works. Japan). The rotation rate of the main shaft could be controlled by
manually adjusting the output frequency of the inverter. Via a reducer,
2. Mechanical aerators the vertical rotation motion of motor shaft was converted to horizontal
rotation motion of the shaft attached to a couple of drums. When the
Paddle-wheel Aerator. According to Lawson (1997), aquacultural lower part of drums were submerged at about 100 mm, the drums ro-
aerators can be generally classified into the three categories: gravity tated at (100 ± 3) rpm. The distribution of water velocity in the pond,
aerators, air diffusion systems, and mechanical aerators. Furthermore, which is not directly relevant to the total volumetric flow rate of the
some hybrids are practically utilized. Gravity aerators employ the water aerator, is required to estimate the volumetric flow rate of circulation
potential energy as in waterfalls for the purpose of dissolving oxygen throughout the pond driven by the aerator (see Brown and Tucker,
into water, whereas air diffusion systems consist of several air com- 2014).
pressors to directly supply air into porous pipes or other devices located Centrifugal Water Stirrer. We designed another mechanical aerator,
at a depth under water. In contrast, “mechanical aerators” include

108
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

Fig. 6. (upper) Top view of the centrifugal water stirrer, CWS, designed for the
purpose of shrimp cultivation and water stirring in a shallow, wide pond. The
original idea of this stirrer is taken from a Japanese patent provided by Inaba
(2005). (lower) Side view of a paddle of the water stirrer, in the form of an
inverted L. The submerged part of the paddle is shaded (revised from Inagaki
Fig. 4. Plan and side views of a commercial small paddle-wheel aerator, TA- et al., 2017).
55H, manufactured by Team Aqua Co. Ltd., Taiwan (revised from Itano et al.,
2018). Gray region indicates that the side view corresponds to water in the converter of reduction gears and pulleys with a ratio of 1:311, the shaft
equilibrium state with the aerator floating on the water. The aerator on the rotates at T = 4.3 rpm under water resistance. The paddles rotate in the
pond surface produces a jet-like water stream downwards with remarkable clockwise direction in the top view of the figure. Three extra floats and
splashing, releasing air bubbles into the water. The amount of water corre-
the main body (not illustrated in the figure) support the motor weight
sponding to the volume of the surface layer under the shaft passed by a paddle
as well as the paddles and shaft over the water surface, and sustain the
is represented as ΔV m3, which is approximated to 2.0 × 10−3 m3.
motor and the shaft in the vertical direction such that the three paddles
are almost horizontal and submerged at a depth of 160 mm.
Primarily, the paddles rotating at the surface induce azimuthal
water circulation around the shaft on the horizontal plane at the sur-
face. If the motor operate continuously, the layer of the horizontally
circulating water gradually expands in the horizontal direction and
thickens in the depth direction, and finally water at the bottom of the
reservoir starts circulating in the bottom plane, which may form an
Ekman layer at the bottom of the reservoir. Moreover, the rotation of
the paddles push water away from the surface towards the radial di-
rection due to centrifugal force partly attributable to the sweptback
angle of the beams against the radial direction. Owing to these effects,
water under the CWS is lifted up from the bottom, and the paddle ro-
tation produces secondary circulation between the surface and the
bottom in the reservoir. It should be noted that such circulation be-
tween the surface and the bottom in the reservoir is not expected for
Fig. 5. A centrifugal water stirrer “CWS” installed at the center of Katayama paddle-wheel aerators prevalent in aquaculture. The optimal volu-
civic outdoor pool (September 2017). metric flux induced by the CWS working under a constant electric
power of 25 W is estimated to be 920 m3/h based on the original patent.
“centrifugal water stirrer” (Fig. 5). The prototype of CWS is inspired by
a Japanese patent provided by Inaba (2005) and practical centrifugal 3. Measurements
surface aerators (cf. Cancino et al., 2004), which is not widely adopted
in aquacultural ponds for the time being. Fig. 6 shows the top and side Reservoir. The present measurement of the water velocity field in-
views of the CWS. A stainless steel paddle in the shape of an inverted L duced by the aforementioned mechanical aerators was carried out at
(L 545 mm× W 200 mm), shown in the side view, is mounted on a Katayama civic pool of Suita city, Osaka, Japan. The measurements
horizontal beam (1260 mm). The beam is attached to a vertical shaft were taken on several calm days in the autumns of 2016 and 2017,
with a sweptback angle of 7.6 ° via a center cylindrical float (ϕ 386 mm during which public swimming was on hold. The Japan Meteorological
× H 200 mm), such that the distance between the shaft and the tip of Agency officially provides wind speed measured every 10 min at an
paddle is 1190 mm. The center cylindrical float rotates together with observatory station in Toyonaka city adjacent to Suita city. The wind
the shaft and the three paddles. Connected to an electric motor speed during the measurement periods was less than 4.1 m/s at max-
(3IK15A-CW2E, 1200 rpm, Oriental Motor Co., Ltd.) via a torque imum. This civic pool possesses an indoor reservoir and three outdoor
reservoirs, among which the largest is an outdoor reservoir of about 0.1

109
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

Fig. 7. Satellite photograph of the largest outdoor reservoir in Katayama civic

outdoor pool. The dimensions of the reservoir are (L) 50 × (W)19 × (D)1.3
[m3].

ha extent ((L)50 m × (W)19 m; 1.3 m average water depth) with

several tiny ducts for sluice and drainage. In the presence of the main Fig. 8. Raw data of the time series of horizontal components of water velocity
building of the indoor reservoir and a few private residential apart- taken at the point (x, y, z) = (0.0m, 3.0m, 0.1m) for 600 s. (upper) Water flow
ments surrounding the outdoor reservoir, the background atmospheric induced by the paddle-wheel aerator on the October 25, 2017. (lower) Water
wind is weak, variational, and non-directional. Therefore, its effect on flow induced by the CWS on September 14, 2017.
the water stream can be considered to be negligible, which was con-
firmed by measurements of the water stream under calm wind with no from the current vector measured at several grid points in the pool. The
aerator. The bottom of the reservoir was designed with a slight incline measurements were made using two electro-magnetic water current
towards the center from the two short sides for the purpose of drainage. meters (VM-801L and VMT2-50-08PS, Kenek Co., LTD., Tokyo, Japan).
The maximum depth of water is 1.4 m at the center, and 1.2 m at the The probe of this current meter is oval shaped, scaled by about 10 mm.
sides. In addition, there are narrow ramps of 0.4 m height along the two Normal components of the water velocity perpendicular to the major
long sides at the bottom. The reservoir comprises a concrete foundation axis of the oval were continuously obtained as analog voltage outputs
and walls coated with waterproof paint and has an exactly rectangular via amplification. We fixed the oval major axis of the probe towards the
shape from the top view (Fig. 7). The center of the reservoir on the z direction so as to obtain x and y components of water velocity of the
surface was taken as the origin of our Cartesian coordinate system, and downstream behind the paddle-wheel aerator and around the CWS. It is
the coordinates x and y were taken as the short and long sides, re- guaranteed by the manufacturer that the current meter provides an
spectively, and z is from the surface to the bottom. accuracy within 2% of the full scale selected from the ranges of 100,
Tethering aerators. Although the sluice and drainage system were 250, and 500 mm/s. Thus, the current meter can distinguish water
stopped at the beginning of September, aquatic bacteria, insects, and velocity up to 2 mm/s at least, which is an advantage in identifying
algae did not flourish until the end of September because of chlorine, subtle water streams induced at a large distance from the aerator.
which was used to prevent them from proliferating during the public Average and dispersion of water velocity. For each measurement at a
swimming service. Tethered to the four corners of the reservoir, the two grid position, a time series of water velocity every second over 600
mechanical aerators, TA-55H and CWS, were installed one by one (TA- seconds was recorded using digital data logger MR8875 (Hioki E.E.
55H was fixed as its jet downstream was directed along y direction) at Corp., Nagano, Japan). Fig. 8 shows the time series of horizontal
the origin. The measurements were conducted in the configuration components of water velocity taken at the grid point closest to the two
where aerators are as isolated from the reflection effect of reservoir aerators. The variance of the time series obtained from the paddle-
walls as possible. This is because our interest is on the difference in flow wheel aerator is apparently larger than that from the CWS. This can be
structure induced by the two distinct aerators in an idealized un- attributed to the fact that the water current induced by CWS is more
bounded pond from the view point of fluid dynamics (less wind and less stable. In the figure, both the time series possess several characteristic
boundary effect), and we will discuss how the difference contributes to frequencies. While the fastest variations common at all the measure-
the shrimp aquaculture. The drums of TA-55H were submerged by ment points could be attributed either to the rotational frequency of the
about 100 mm, and their rotational speed was manually maintained drums and the paddles, the slow variation peculiar to the measurement
between 97∼103 times per minute on average, by controlling the at the grid points near the aerator is attributed to a certain unsteady
output frequency of the inverter. The three paddles of CWS were sub- large-scale turbulent motion around the aerator, which will be in-
merged by about 150 mm in depth, and their rotation rate was main- vestigated in future works. The average ūi and the standard deviation
tained at 4.3 rpm under water resistance. σ(ui) of water velocity at each grid point is calculated from the raw data,
Electro-magnetic water current meter. Velocity fields were constructed as follows:

110
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

Fig. 10. Velocity field of the water current induced by CWS on the y–z plane.
Measurements were taken at three different depths, z = 0.1 m, 0.6 m, 1.1 m.

figure, we may conclude that the aerator located at the origin provides
a horizontal and almost unidirectional jet stream downwards
throughout the reservoir under consideration. Thus, a fluid element
travels along a horizontal path in a trajectory resembling the Arabic
numeral “8”. In fact, from the data obtained at several grid points at
x = 0 and 6 m < y < 12 m, a downward jet-like flow, uy, is inter-
polated as approximately uy =−0.014y + 0.26. Moreover, if we as-
sume the incompressibility of the water current and the mirror sym-
∂u ∂u
metry on the x = 0 plane, then ∂zz = − ∂xx = 0.014 /s. Thus we obtain
uz = 1.4 mm/s on the y axis as a rough estimation on the vertical
component of water flow in the downstream.
Velocity field by CWS. The mean water velocity fields induced by
CWS, (u¯ x , u¯ y ) , taken at the horizontal planes at z = 0.1 m and z = 1.1 m
are mapped in Fig. 11. The error bars at the tip of each vector represent
those in Fig. 9. We note that, compared with the error bars in Fig. 9, the
fluctuation of water current induced by CWS is relatively small even
Fig. 9. Velocity field of the water current induced by TA-55H. Measurements
were taken at two horizontal planes, (upper) z = 0.1 m and (lower) z = 1.1 m. near the aerator. The small fluctuations at all the grid points suggest
that the current is relatively steady and that the water current structure
estimated from the mean velocity measured is robust.
1 T
u¯ i (x , y, z ) = T
∫0 ui (t , x , y, z )dt
1 T
σ (ui (x , y, z )) = T
∫0 (ui (t , x , y, z ) − u¯ i (x , y, z ) ) dt 5. Discussion

where i (i = x, y, z) is the component index and T is the total mea- Note that the velocity fields at the surface and bottom planes in
surement time. Fig. 11 are remarkably different, which is in contrast to Fig. 9. The
azimuthal component of the induced velocity fields are clockwise in
4. Velocity field both cases. Therefore, the horizontal circulation common in Fig. 11
may be attributed to shear stress between the surface and bottom
Velocity Field by TA-55H. The velocity field, (u¯ x , u¯ y ) , in Fig. 9 is the planes. In contrast, the radial component of water current on the sur-
mean water current induced by TA-55H. The measurements were taken face and bottom planes are positive and negative, respectively. Taking
at several grid points at the two horizontal planes, (shallow) z = 0.1 m into account that the velocity field is incompressible, we can deduce the
and (deep) z = 1.1 m downstream of the aerator. The average and
variance were obtained from sequential raw data taken for T = 600 s
per measurement at a grid point. The x and y error bars centered at the
tip of vectors represent a fluctuation of the water current estimated
from the standard deviation, ± σ(ui(x, y, z)) for i = x, y.
As shown in the figure, TA-55H induces a unidirectional jet stream
downwards along the y axis from the origin. The magnitude of the
water current is more than 200 mm/s at the measurement grid closest
to the aerator. The jet stream is dispersed and reduced downstream
away from the aerator. Under the restriction of the reservoir wall,
certain return currents were observed within 4 m from the side walls
located at x = ± 9, which would probably be weaker in a larger re-
servoir, such as a commercial aquacultural pond.
It should be noted that the water current at the bottom of the re-
servoir (z = 1.1) is similar to that at the surface (z = 0.1). This simply
suggests that the water current at the bottom is induced via shear stress
attributed to the surface current, which is driven primarily by the
aerator. Accordingly, this implies that the mean flow does not largely
contribute to the vertical circulation expected for the purpose of mixing
high DO water between the surface and the bottom of the reservoir.
Such mixing in the vertical direction is dependent on unsteady turbu-
lent flow inherent to the jet stream, which is reflected in large fluc-
tuations near the origin, as shown in the figure.
velocity field by CWS on the vertical cross section. We additionally
measured the mean water velocity field on the y-z plane, (u¯ y , u¯ z ) , which
is mapped in Fig. 10. The error bars spanned at the tip of each vector Fig. 11. Velocity field of the water current induced by CWS. Measurements
represent the current fluctuation observed in the time series. From the were taken at two horizontal planes, (upper) z = 0.1 m and (lower) z = 1.1 m.

111
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

profile of water stream of a paddle-wheel aerator conducted by Ya-

mayoshi et al. (see Fig.6 in Yamayoshi et al., 2008, the streamwise
current in an aquacultural pond with 1 m depth did not penetrate below
0.4 m depth from the surface). In reality, around the origin of our re-
servoir, the magnitude of the shear at the bottom of CWS is larger than
that of the present paddle-wheel aerator, which can be estimated from
the comparison between Fig. 11(lower) and Fig. 9(lower). Thus, paddle-
wheel aerators may be more advantageous than CWSs, if the important
factor with respect to performance in sludge transportation is not only
vertical shear but also the time-dependence of the water current.
In the aforementioned commercial shrimp aquaculture at the ABC
Center of Karawang, Indonesia, it has been observed that shrimps
gather immediately under CWS during early morning hours. This sug-
gests that water circulation between the surface and the bottom re-
Fig. 12. The expected spiral trajectories of fluid parcels under CWS. The radial
solves the DO shortage under the CWS. Additionally, a circle at the
component of water current fields on the surface and bottom planes implies the
upward flow under CWS and the circulation between the surface and bottom bottom with a radius of 7 m centered at the origin had a horizontal
throughout the reservoir. water current of more than 15 mm/s. An optimal arrangement of
aerators within a rectangular pond was discussed previously. For ex-
ample, Peterson et al. (2001) performed a numerical simulation and
existence of upward flow under CWS, that is, circulation between the
recommended a diagonal or parallel arrangement of paddle-wheel
surface and bottom of the reservoir. The expected spiral trajectories of
aerators to sustain a large horizontal circulation in a rectangular pond.
fluid elements under CWS is illustrated in Fig. 12. A passive tracer
Adopting the value of the radius as an indicator, one would optimally
travels along one of the spiral trajectories in a cylindrical region around
allocate the positive–negative rotating CWSs alternatively at distances
the z axis with a radius. From the vector field in Fig. 11, the radial
of 14 m in a large pond.
component of ur at z = 0.1 can be interpolated as ur =−0.01x + 0.1
Electric power. Finally, we should take into account the electric
m/s at y = 0 (2 m < x < 6 m) and as ur =−0.005y + 0.1 m/s at x = 0
power consumed by both aerators. In the present measurements, TA-
(4 m < y < 12 m) approximately. If we assume the incompressibility
55H and CWS worked at 630 W and 40 W, respectively. Thus, the
of the water current and the velocity field to be axisymmetric around
∂u 1 ∂ paddle-wheel aerator consumed more than 10 times the running cost of
the z axis, ∂zz = − r ∂r (rur ) ≈ 0.01 ∼ 0.02 /s. The vertical component of
the centrifugal-force aerator. The kinetic energy of water current in the
water current, uz, vanishes at z = 0. Therefore, we can estimate the reservoir is, however, not largely different between them. The large
vertical component of water flow, uz, as approximately 1 mm/s ∼2 difference between them can be attributed to the fact that most of the
mm/s in the cylindrical shell region within radius 6 m < r < 12 m. In energy provided to TA-55H may be consumed in enhancing DO near the
this cylindrical shell region, there is down-flow from the surface by at aerator, but CWS consumes electric energy to generate water current
least uz > 1 mm/s and oxygen dissolved naturally at the surface from kinetic energy. From the previous measurement by Inagaki et al.
the atmosphere may be directly supplied to the reservoir bottom. The (2017), the upward current under the CWS was estimated as 550 m3/h.
vertical water current in the cylindrical shell region of the CWS is According to Brown et al. (2016), the flow rate of the CWS is less than
comparable to that of the paddle-wheel aerator, 1.4 mm/s, deduced in that of paddle-wheel aerators, but water discharge per unit power input
the last section. It should also be noted that the area of the cylindrical of the CWS is fairly efficient because of its lower energy consumption
shell region is over 400 m2, which covers the corresponding area of TA- and of water exchange directly transferring between the surface and
55H. bottom.
Intermittency and sludge transportation. Based on measurements
conducted in the present study, one would consider the paddle wheel 6. Conclusion
aerator, providing mainly a horizontal stream in average, ineffective
compared to CWS. However, returning to the raw data of water current In order to establish an optimal aeration system to maintain the DO
in Fig. 8, we see that the dispersion of the velocity in the downstream of value required for biological activity in water, we evaluated the water
the paddle-wheel aerator is relatively large compared to that of CWS. flows in a rectangular reservoir induced by a standard commercial
From the viewpoint of fluid dynamics, a jet stream in an open channel is Taiwanese paddle-wheel aerator and a self-designed centrifugal water
generally perceived to be unstable. As reflected in Fig. 9, the time stirrer. The reservoir had a maximum depth of about 1.4 m, and the
variation in the water current induced by a paddle-wheel aerator is not measurements were taken using an electro-magnetic water current
negligible. Vertical mixing induced by the paddle-wheel aerator may be meter. Water velocity was found to be horizontally unidirectional for
largely dependent on certain time-dependent fluid motions such that the paddle wheel aerator, and vertical circulation between the surface
the bulk water intermittently either dives or surfaces in the vertical and the bottom induced by the paddle wheel aerator was smaller than
direction, which was cancelled in time-averaging of the raw data. that by the CWS aerator. The generated jet stream loses its kinetic en-
It has been known that paddle wheel aerators have multiple func- ergy as a function of the water viscosity, and its contribution to the high
tions; one is to enhance oxygen injection to the bottom, and the other is DO transfer may be restricted only to the vicinity of the aerator
to transport sludge to a part of the pond (Yamayoshi et al., 2008). Such downstream. The mean electric power consumed to operate the aera-
sludge transportation enlarges an active region for shrimps and helps tors was distinctly different, and most of the energy provided to the
farmers to remove a collected pile of sedimentation from the pond. It paddle wheel aerator was probably consumed in the enhancement of
may be thought that the function of sludge transportation is caused by DO near the aerator, or otherwise simply consumed in generating in-
the vertical shear due to a strong horizontal stream created by a paddle- tense splashing and loud noise. Employment of CWS aerators as aux-
wheel aerator, as if a stream creates a deeper channel. However, the iliary equipment may contribute to the delivery of high DO water
magnitude of the shear at the bottom created by the paddle-wheel throughout a pond that could be greater than that produced by a paddle
aerator is not so large as that would be expected from the strong water wheel aerator. The fundamental assertion of the present study is that
current at the surface. In Fig. 10 the mean water current of the paddle- not only the efficiency to dissolve oxygen, but also the difference of
wheel aerator at z = 0.1 does not penetrate to the depth z ≤ 0.6, which flow structure, should be accounted for in an evaluation between
was also previously confirmed in direct measurement of a vertical aerators for aquacultural purposes. The measurement of the DO field in

112
T. Itano, et al. Aquacultural Engineering 85 (2019) 106–113

the reservoir will be conducted in future work. Colt, J., Orwicz, C., 1991. Aeration in intensive culture. Aquacult. Water Qual. 3,
198–269.
Gillet, R., 2008. Global study of shrimp fisheries. FAO fisheries technical paper 475.
Acknowledgement Huggins, D.L., Piedrahita, R.H., Rumsey, T., 2004. Analysis of sediment transport mod-
eling using computational fluid dynamics (cfd) for aquaculture raceways. Aquacult.
We wish to appreciate Mr Hosokawa and Mr Izu for their efforts to Eng. 31, 277–293.
Inaba, H., 2005. Floating type energy saving water cleaning apparatus. Japanese Patent
promote the research at the primary stage. This research is a product of (3360075).
research financially supported by the Kansai University Fund for Inagaki, T., Nakamura, C., Itano, T., Sugihara-Seki, M., Hyodo, J., 2017. Water circulation
Supporting Young Scholars 2016–2017, and ORDIST group fund 2018. induced by a centrifugal water streamer in rectangular reservoirs. Transactions of the
Japan Society of Mechanical Engineers 83.
The authors acknowledge Suita City Hall for kindly providing Katayama Itano, T., Inagaki, T., Nakamura, C., Hashimoto, R., Negoro, N., Honda, S., Hyodo, J.,
civic outdoor pool as an experimental reservoir. Finally, we would like 2018. Water circulation driven by a paddle-wheel aerator in rectangular reservoir.
to thank Editage for English language editing. Sci. Technol. Rep. Kansai University 60, 15–22.
Lawson, T.B., 1997. Fundamentals of Aquacultural Engineering. CBS Publishers, New
Delhi.
References Moulick, S., Mal, B., Bandyopadhyay, S., 2002. Prediction of aeration performance of
paddle wheel aerators. Aquacult. Eng. 25, 237–271.
Boyd, C., 1998. Pond water aeration systems. Aquacult. Eng. 18, 9–40. Murai, Y., 2007. Ebi to Nihonjin II.: kurashi no naka no gurobaru-ka (in Japanese).
Brown, T.W., Tucker, C.S., 2014. Pumping performance of a modifed commercial pad- Iwanami Shoten.
dlewheel aerator for split-pond aquaculture systems. N. Am. J. Aquaculture 76, Peterson, E.L., Harris, J.A., Wadhwa, L.C., 2000. Cfd modelling pond dynamic processes.
72–78. Aquacult. Eng. 23, 61–93.
Brown, T.W., Tucker, C.S., Rutland, B.L., 2016. Performance evaluation of four different Peterson, E.L., Wadhwa, L.C., Harris, J.A., 2001. Arrangement of aerators in an intensive
methods for circulating water in commercial-scale, split-pond aquaculture systems. shrimp growout pond having a rectangular shape. Aquacult. Eng. 25, 51–65.
Aquacult. Eng. 70, 33–41. Pfeiffer, T., Lawson, T., Church, G., 2007. Engineering considerations for water circula-
Cancino, B., 2004. Design of high efficiency surface aerators: Part 2. rating of surface tion in crawfish ponds with paddlewheel aerators. Aquacult. Eng. 36, 239–249.
aerator rotors. Aquacult. Eng. 31, 99–115. Yamayoshi, N., Kitazawa, D., Konno, A., Chiba, K., 2008. Numerical simulation of the
Cancino, B., Roth, P., Reuß, M., 2004. Design of high efficiency surface aerators: Part 1. current field in the small cultivation pond with the effect of paddle wheel. Seisan
development of new rotors for surface aerators. Aquacult. Eng. 31, 83–98. Kenkyu 60, 55–58.

113