With river water quality deterioration in recent years, an increasing number of river water quality control studies have been conducted. Among relevant methods, aeration and vegetation planting are effective techniques. The combination of aeration and vegetation can improve the purification effect on the water quality. Based on flume experiments, the mass transfer coefficient of dissolved oxygen in rivers with floating vegetation patches of different diameters under hydrodynamics was studied. Large-diameter floating vegetation can effectively reduce the breaking of bubbles and increase the mass transfer coefficient of dissolved oxygen in rivers. According to mechanism analysis, a model of the oxygen mass transfer coefficient in floating vegetated channels was proposed, and a favorable simulation effect was obtained. This type of research could provide a theoretical basis for selecting and arranging vegetation in aeration floating vegetated channels.

  • Explored the effect of different floating vegetation diameters on oxygen mass transfer.

  • Provided a formula for the oxygen mass transfer coefficient related to the diameter of floating vegetation.

With improvements in human production and living standards, increasingly more pollutants are discharged into rivers, and water pollution in rivers is becoming increasingly serious (Chinyama et al. 2016; Li et al. 2021). Since the beginning of the new century, scholars have directed more attention toward addressing water quality and environmental problems of rivers (Takić et al. 2017; Jiang et al. 2019). As the carrier of pollutants in rivers, vegetation plays an important role in the river environment (O'Briain et al. 2018; Ielpi et al. 2022). The main types of vegetation in river channels include emergent vegetation, submerged vegetation and floating vegetation. Floating vegetation (water lilies, pondweed, American lotus, Eichhornia crassipes, duckweed, and common bladderwort) is often applied in rivers because of its high aesthetic value (Rooney et al. 2013; Han et al. 2018). Moreover, aeration can increase river water reoxygenation and improve the pollutant retention efficiency (Olgac et al. 1976; Chen et al. 2019). Therefore, it is important to study the dissolved oxygen (DO) mass transfer coefficient produced by aeration in floating vegetation channels.

In aeration, bubbles are injected into water through an aeration device at the bottom of the river. In the process of bubbles rising toward the water surface and breaking, DO is transmitted through the interface between bubbles and water. The transmission efficiency has been extensively determined and studied. It is generally considered that the mass transfer coefficient of DO is mainly related to the aeration rate, bubble size and water velocity (Burris et al. 2002; Zoheidi et al. 2017). Fayolle et al. (2010) demonstrated that when the velocity varied between 0 and 0.42 m/s, the total DO transfer coefficient increased by 29%. Gillot et al. (2000) showed that when the velocity reached 0.44 m/s, the oxygen transfer efficiency of the aerator increased by 38%, and the bubble diameter decreased by 24% compared with that in static water. Zoheidi et al. (2017) investigated the relationship between the bubble size distribution and aeration rate. To determine the bubble size distribution during bubble plug transition flow in a narrow rectangular channel, Zhang et al. (2022a) proposed two sets of bubble fitting methods for transition flow.

After vegetation is added, the hydrodynamic conditions of the river channel become more complex (Pu 2023), and the influence on DO mass transfer increases. Vegetation exerts a notable blocking effect on water flow, which greatly disturbs vegetation (Pu et al. 2019; Ikani et al. 2023, 2024). In emergent and submerged vegetation channels, the water flowing through vegetation areas can cause considerable disturbances (Nepf 1999, 2012), which can obviously increase the mass transfer efficiency of DO. John et al. (2024) reported that the effect of vegetation on the turbulent burst cycle is mostly obvious up to approximately two-thirds of the vegetation height. Bai et al. (2023) obtained the oxygen transfer coefficient under rigid vegetation flow and found that vegetation could effectively increase the oxygen transfer coefficient. Tseng & Tinoco (2020) examined the relationship between the mean flow velocity and turbulent kinetic energy production in submerged and emergent vegetated rivers and developed a modified surface renewal model using turbulent kinetic energy production as an indicator of the DO transfer efficiency. In contrast to those produced by submerged and emergent vegetation, the hydrodynamic conditions produced by floating vegetation obviously differ (Figure 1). Floating vegetation can be regarded as the upper boundary of the river, and the vertical velocities at the middle of the river are high and low in the top and bottom layers, respectively (Bai et al. 2020; Bai & Sun 2022). Second, floating vegetation can cover the water surface. Bubbles do not break when they encounter floating vegetation but will break when they are transported to areas not covered by floating vegetation, which reduces the breaking efficiency of bubbles on the water surface. This obviously causes an increase in the time of bubble formation in water and helps to increase the mass transfer efficiency of DO.
Figure 1

Diagram of bubbles and the flow velocity in floating vegetation channels.

Figure 1

Diagram of bubbles and the flow velocity in floating vegetation channels.

Close modal

There have been many studies on the effect of aeration on DO mass transfer in river channels (Gillot et al. 2005; Fayolle et al. 2007), but studies on the DO mass transfer coefficient in floating vegetation river channels have not been conducted. The impact of floating vegetation on bubble transport and the mechanisms of oxygen transfer must still be studied. Based on flume experiments, the oxygen mass transfer efficiency was investigated in this paper under different floating vegetation diameters and hydrodynamic conditions. The main research purposes were (1) to explore the influences of floating vegetation on the distribution and migration of air bubbles, (2) to examine the influence of floating vegetation on the DO mass transfer coefficient, and (3) to propose a prediction equation for the oxygen mass transfer coefficient in floating vegetation channels. This study is helpful for optimizing vegetation planting in floating vegetation flumes and improving water purification.

Experimental setup

The experiment was conducted in a circulating water tank (40 cm × 50 cm × 1,000 cm), and the water level was adjusted through the tail gate (Figure 2). The floating vegetation used in the experiment comprised artificial bionic lotus leaves with diameters of 40, 20 and 10 cm. Lotus leaves with different diameters were arranged in the flume (Figure 3). An aeration device was installed at the bottom of the flume, and a camera and DO detection equipment were arranged above the flume. The aeration rate and discharge of flow were used as control variables to divide the experimental groups. Before each treatment, cobalt chloride hexahydrate (CoCl2·6H2O) was added to water as a catalyst and mixed with sodium sulfite (Na2SO3), which was used as an oxygen depletion agent. Once the DO concentration decreased to its minimum (near zero), the reaeration process occurred (Stenstrom et al. 2006). The experimental scheme is summarized in Table 1.
Table 1

Parameters of the different treatments

Diameter of vegetation d (cm)Air flow rate A (N m3/h)Discharge Q (L/s)SlopeWater depth h (cm)Average velocity Um (cm/s)Bubble diameter (cm)
10 10.29 0.0005 16.82 15.30 3.11 2,875 
12.55 0.0005 19.30 16.26 3.14 3,506 
14.22 0.0005 20.31 17.50 3.15 3,972 
20 10.54 0.0005 17.35 15.19 3.07 2,944 
12.42 0.0005 18.11 17.14 3.22 3,469 
14.61 0.0005 20.53 17.79 3.04 4,081 
40 10.18 0.0005 16.23 15.68 3.08 2,844 
12.31 0.0005 18.23 16.88 3.06 3,439 
14.34 0.0005 20.68 17.34 3.16 4,006 
Diameter of vegetation d (cm)Air flow rate A (N m3/h)Discharge Q (L/s)SlopeWater depth h (cm)Average velocity Um (cm/s)Bubble diameter (cm)
10 10.29 0.0005 16.82 15.30 3.11 2,875 
12.55 0.0005 19.30 16.26 3.14 3,506 
14.22 0.0005 20.31 17.50 3.15 3,972 
20 10.54 0.0005 17.35 15.19 3.07 2,944 
12.42 0.0005 18.11 17.14 3.22 3,469 
14.61 0.0005 20.53 17.79 3.04 4,081 
40 10.18 0.0005 16.23 15.68 3.08 2,844 
12.31 0.0005 18.23 16.88 3.06 3,439 
14.34 0.0005 20.68 17.34 3.16 4,006 

Note: .

Figure 2

Structure of the flume.

Figure 2

Structure of the flume.

Close modal
Figure 3

Arrangement of the floating vegetation.

Figure 3

Arrangement of the floating vegetation.

Close modal

Measured oxygen mass transfer coefficient

The bubble diameter was derived from images obtained by the high-speed camera (Canon PowerShot G6) located on the right side of the flume (Figure 2). Light was provided by a 600-W laser lamp installed on the left side of the flume. Calculation of the Sauter diameter requires determining the size of a minimum of 100 bubbles. The equivalent diameter can be defined as follows (Fayolle et al. 2007):
(1)
where and are the major and minor axes, respectively, of ellipsoidal bubble i (m) and can be obtained from bubble images (Figure 4).
Figure 4

Measurement methods for the bubble diameter and : (a) Measured area of the bubble diameter, (b) measurement methods for and , and (c) curve fitting of the measured values.

Figure 4

Measurement methods for the bubble diameter and : (a) Measured area of the bubble diameter, (b) measurement methods for and , and (c) curve fitting of the measured values.

Close modal
The Sauter diameter () required to determine the gas‒liquid interface area can be calculated as follows:
(2)
where is the equivalent diameter of bubble i and N is the total number of bubbles.
The DO concentration was measured at three points using a DO probe (YSI 57), covering six measuring points arranged along the water flow direction (5 cm below the water surface and 1, 3 and 5 m away from the diffuser) (shown in Figure 2). The sampling frequency was set to 5 s. The oxygen transfer coefficient can be expressed under standard conditions (20 °C and 1,013 hPa) as follows:
(3)
where is the oxygen transfer coefficient at temperature T (°C), which can be obtained by curve fitting (Figure 4), and (h−1) is the oxygen transfer coefficient at 20 °C.
Figure 5

Blocking effect of floating vegetation on bubbles.

Figure 5

Blocking effect of floating vegetation on bubbles.

Close modal

Simulated oxygen mass transfer coefficient

The oxygen transfer coefficient can be predicted as follows (Fayolle et al. 2007):
(4)
where is the diffusion coefficient of oxygen in water at 20 °C = 1.89 × 10−9m2/s (Terashima et al. 2016), is the relative velocity of the gas phase (m/s), is the bubble diameter, and is the air volume fraction.
As shown in Figure 5, at time from the aeration device to the water surface, the bubbles in vegetation-covered areas cannot be directly broken but must be moved to areas without vegetation coverage to break. Assuming that aerated bubbles are evenly distributed across the channel liquid surface, the number of bubbles moving from the inside to the outside of the vegetation area during this time can be derived from the shaded part in Figure 5. Here, l is the maximum distance across which bubbles can move out of the vegetation area during period .
(5)
(6)
(7)
Then, Equation (4) becomes the following:
(8)

Model accuracy calculation

The coefficient of determination (R2) and mean absolute error () were used to assess the simulation effect between the simulated and measured data. and R2 can be determined as follows:
(9)
(10)
(11)
(12)
(13)
where N denotes the number of lateral measuring points; X and Y denote the calculated and measured values, respectively; SSE is the sum of squares of the error; and SST is the total sum of squares.
The relationship between the DO mass transfer coefficient and the average velocity is shown in Figure 6. The DO mass transfer coefficient exhibited a significant relationship with the vegetation diameter. The larger the vegetation diameter is, the higher the corresponding DO mass transfer coefficient. The mass transfer coefficient of DO decreased with increasing average velocity. This is inconsistent with previous studies, which showed that the mass transfer coefficient of oxygen increased with increasing average velocity (Burris et al. 2002; Vermande et al. 2007). The reason for this may be that the average velocity variation range of previous studies is large (0–4 cm/s) (Vermande et al. 2007), and the flow turbulence caused by velocity leads to an obvious increase in the DO transfer coefficient. However, our velocity variation range was relatively small (15.19–17.79 cm/s), and the change caused by turbulence was relatively small. ranged from 2,844 to 4,081 in our research, but it ranged from 0 to 53,637 in Vermande et al. (2007). A greater velocity could cause more bubbles to overflow out of the vegetation area, leading to a decrease in the bubble content in the water body and a subsequent decrease in the mass transfer coefficient of DO.
Figure 6

Relationship between the average flow rate and oxygen mass transfer coefficient.

Figure 6

Relationship between the average flow rate and oxygen mass transfer coefficient.

Close modal
A comparison between the simulation results of the DO mass transfer coefficient and the measured results is shown in Figure 7. The simulation effect of the model is satisfactory. The MAE was 0.2258, and the R2 value was 0.8551. However, most measured values of the mass transfer coefficient of DO were lower than the simulated values and were not evenly distributed on either side of 1:1 line. The main reason for this finding may be that in the prediction process, we assumed that bubbles would break only when moving out of the vegetation area with increasing flow velocity. However, since the vegetation and water surface are not completely bonded, bubbles may also break when entering the spaces between the vegetation bottom and the water surface. At this time, we assumed that the remaining 20% of the bubbles would break between the vegetation bottom and the water surface, and we revised Equation (5). The revised equation is provided below. After revision, the accuracy improved, with an R2 value of 0.96 (Figure 8). The fitting degree between the different vegetation treatments and the water surface varied. If other blades are used, the coefficient must be corrected again.
(14)
where is the fracture coefficient of the bubbles.
Figure 7

Comparison of the measured and simulated oxygen mass transfer coefficients.

Figure 7

Comparison of the measured and simulated oxygen mass transfer coefficients.

Close modal
Figure 8

Comparison of the measured and simulated oxygen mass transfer coefficients with the fracture coefficient.

Figure 8

Comparison of the measured and simulated oxygen mass transfer coefficients with the fracture coefficient.

Close modal

Floating vegetation, as common riverway vegetation, profoundly impacts hydrodynamic forces and water quality (Li et al. 2021; Rudi et al. 2021). After adding aeration measures, sufficient oxygen will accelerate the retention of pollutants by vegetation and bacteria (Zhang et al. 2022b). The results of this paper showed that floating vegetation is helpful for improving the oxygen mass transfer efficiency in aeration channels, but this was determined by comparing the results with Equation (5). When l is greater than the vegetation diameter d, bubbles will theoretically overflow out of the vegetation area, so the vegetation slightly affects the mass transfer coefficient of DO. Therefore, a reasonable arrangement of aeration devices in line with floating vegetation channels could help to improve the oxygen mass transfer efficiency in rivers.

The change in the oxygen mass transfer coefficient caused by rigid vegetation varies, as rigid vegetation generates a mechanical diffusion coefficient and a turbulent diffusion coefficient formed by eddy currents between vegetation units. Floating vegetation does not yield these two diffusion coefficients, so its impact on DO diffusion is relatively limited compared to that of rigid vegetation.

In real-world cases, floating vegetation can impede the transfer of oxygen at the air‒water interface. Moreover, areas where floating vegetation naturally occurs often experience very low water velocities. These locations are naturally preferred by vegetation, suggesting that they are likely already populated by other forms of natural vegetation (i.e., submerged and emerged). The density of vegetation in these areas may hinder light penetration and photosynthesis, consequently leading to oxygen depletion and increased community respiration. These complexities highlight the challenges associated with the presence of real vegetation. The application of the experimental results is limited by the assumption that the primary source of oxygen in the channel is the aerator positioned at the bottom rather than oxygen diffusion across the air‒water interface.

The distribution of floating vegetation in reality is not as regular as that in this paper, and the effect of irregular floating vegetation on the oxygen mass transfer coefficient must be studied further. Notably, the change in the oxygen mass transfer coefficient impacts the pollutant degradation efficiency in floating vegetation channels, which should be examined further. This paper is based mainly on indoor experiments. To improve the results of this paper, additional outdoor validation experiments should be performed.

Based on flume experiments and theoretical analysis, in this paper, the influence of vegetation characteristics on the oxygen mass transfer coefficient in floating vegetation channels was studied, and the following conclusions were obtained:

  • (1) The average flow velocity in the floating vegetation channel exerts a relatively limited impact on the oxygen mass transfer coefficient. With increasing flow velocity, the DO mass transfer coefficient slightly decreases.

  • (2) According to the diameter of floating vegetation, an improved oxygen mass transfer coefficient equation is established, and the simulation effect of the model is satisfactory. According to theoretical analysis, the simulation effect can be further improved after adding the breaking coefficient.

  • (3) When l is greater than the vegetation diameter d, floating vegetation exerts a negligible effect on aeration, which also provides a theoretical reference for selecting and arranging floating vegetation in floating vegetation channels.

  • (4) The effect of irregular floating vegetation on the oxygen mass transfer coefficient should be studied further, and additional outdoor validation experiments should be conducted.

Y.B. analyzed and interpreted the data and was a major contributor in writing the manuscript.

The paper was supported by National Science Foundation for Young Scientists of China (Grant No. 42207099); Scientific Research Foundation of Zhejiang University of Water Resources and Electric Power (xky2022004); Key Technology Research and Development Program of Zhejiang (No. 2021C03019); Zhejiang Provincial Natural Science Foundation of China (LZJWD22E090001); key technology development and application demonstration of comprehensive management and resource utilization of cyanobacteria in Taihu Lake Basin (Key R&D funds of Zhejiang Province: 2021C03196).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Bai
Y.
,
Xia
Y.
,
Geng
N.
,
Qi
Y.
,
Huang
D.
,
Zhao
Y.
,
Huang
L.
,
Shen
D.
,
Sun
G.
,
Xu
C.
&
Hua
E.
2023
Research on oxygen transfer in an aerated flow with emergent vegetation
.
Journal of Hydrology
617
,
128935
.
Burris
V. L.
,
McGinnis
D. F.
&
Little
J. C.
2002
Predicting oxygen transfer and water flow rate in airlift aerators
.
Water Research
36
(
18
),
4605
4615
.
Chen
J.
,
He
Y.
,
Wang
J.
,
Huang
M.
&
Guo
C.
2019
Dynamics of nitrogen transformation and bacterial community with different aeration depths in malodorous river
.
World Journal of Microbiology and Biotechnology
35
(
12
),
1
12
.
Chinyama
A.
,
Ncube
R.
&
Ela
W.
2016
Critical pollution levels in Umguza River, Zimbabwe
.
Physics and Chemistry of the Earth, Parts a/b/c
93
,
76
83
.
Fayolle
Y.
,
Cockx
A.
,
Gillot
S.
,
Roustan
M.
&
Héduit
A.
2007
Oxygen transfer prediction in aeration tanks using CFD
.
Chemical Engineering Science
62
(
24
),
7163
7171
.
Fayolle
Y.
,
Gillot
S.
,
Cockx
A.
,
Bensimhon
L.
,
Roustan
M.
&
Heduit
A.
2010
In situ characterization of local hydrodynamic parameters in closed-loop aeration tanks
.
Chemical Engineering Journal
158
(
2
),
207
212
.
Han
L.
,
Zeng
Y.
,
Chen
L.
&
Li
M.
2018
Modeling streamwise velocity and boundary shear stress of vegetation-covered flow
.
Ecological Indicators
92
,
379
387
.
Ielpi
A.
,
Lapôtre
M. G.
,
Gibling
M. R.
&
Boyce
C. K.
2022
The impact of vegetation on meandering rivers
.
Nature Reviews Earth & Environment
3
(
3
),
165
178
.
Ikani
N.
,
Pu
J. H.
,
Taha
T.
,
Hanmaiahgarib
P. R.
&
Penna
N.
2023
Bursting phenomenon created by bridge piers group in open channel flow
.
Environmental Fluid Mechanics
23
(
1
),
125
140
.
Ikani
N.
,
Pu
J. H.
,
Zang
S.
,
Al-Qadami
E. H. H.
&
Razi
A.
2024
Detailed turbulent structures investigation around piers group induced flow
.
Experimental Thermal and Fluid Science
152
,
111112
.
Jiang
C.
,
Yin
L.
,
Li
Z.
,
Wen
X.
,
Luo
X.
,
Hu
S.
,
Yang
H.
,
Long
Y.
,
Deng
B.
,
Huang
L.
&
Liu
Y.
2019
Microplastic pollution in the rivers of the Tibet Plateau
.
Environmental Pollution
249
,
91
98
.
John
C. K.
,
Pu
J. H.
,
Guo
Y.
,
Hanmaiahgari
P. R.
&
Pandey
M.
2024
Flow turbulence presented by different vegetation spacing sizes within a submerged vegetation patch
.
Journal of Hydrodynamics
35
(
6
),
1131
1145
.
Li
A.
,
Yuan
Q.
,
Strokal
M.
,
Kroeze
C.
,
Ma
L.
&
Liu
Y.
2021
Equality in river pollution control in China
.
Science of The Total Environment
777
,
146105
.
Nepf
H. M.
2012
Hydrodynamics of vegetated channels
.
Journal of Hydraulic Research
50
(
3
),
262
279
.
O'Briain
R.
,
Shephard
S.
&
Coghlan
B.
2018
A river vegetation quality metric in the eco-hydromorphology philosophy
.
River Research and Applications
34
(
3
),
207
217
.
Olgac
N. M.
,
Longman
R. W.
&
Cooper
C. A.
1976
Optimal control of artificial aeration in river networks
.
ISA Transactions
15
(
4
),
359
371
.
Pu
J. H.
,
Hussain
A.
,
Guo
Y. K.
,
Vardakastanis
N.
,
Hanmaiahgari
P. R.
&
Lam
D.
2019
Submerged flexible vegetation impact on open channel flow velocity distribution: An analytical modelling study on drag and friction
.
Water Science and Engineering
12
(
2
),
121
128
.
Rudi
G.
,
Belaud
G.
,
Troiano
S.
,
Bailly
J. S.
&
Vinatier
F.
2021
Vegetation cover at the water surface best explains seed retention in open channels
.
Ecohydrology
14
(
2
),
e2263
.
Stenstrom
M. K.
,
Leu
S. Y. B.
&
Jiang
P.
2006
Theory to practice: Oxygen transfer and the new ASCE standard
.
Water Environment Federation
2006
(
7
), pp.
4838
4852
.
Takić
L.
,
Mladenović-Ranisavljević
I.
,
Vasović
D.
&
Đorđević
L.
2017
The assessment of the Danube River water pollution in Serbia
.
Water, Air, & Soil Pollution
228
(
10
),
1
9
.
Vermande
S.
,
Simpson
K.
,
Essemiani
K.
,
Fonade
C.
&
Meinhold
J.
2007
Impact of agitation and aeration on hydraulics and oxygen transfer in an aeration ditch: Local and global measurements
.
Chemical Engineering Science
62
(
9
),
2545
2555
.
Zhang
J.
,
Wang
W.
,
Li
Z.
,
Wang
H.
&
Geng
Y.
2022a
Modeling of velocity and shear stress profiles in the ecological channel with floating vegetation
.
Environmental Science and Pollution Research
30
(
3
),
6506
6516
.
Zhang
M.
,
Mi
T.
,
Deng
J.
,
He
Q.
,
Gu
Z.
,
Zhang
L.
,
Ding
S.
&
Pan
L.
2022b
Bubble size distribution for bubbly-to-slug transition flow in narrow rectangular channel
.
Nuclear Engineering and Design
391
,
111725
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).