Sediment accumulation in combined sewers can induce blockage and odor problems. Among various cleaning methods, using self-cleaning device-generated flushing waves has been thought to be an effective solution. In this study, a series of numerical tests were conducted using CFD software to investigate the cleaning efficiency of deposited sediment particles based on a simplified self-cleaning device. The CFD model was validated by the experimental and numerical results in the literature. The effects of several parameters including the flushing gate height, sediment bed thickness, sediment bed length, and sediment bed position on cleaning efficiency were discussed. A relative accumulative transport rate was defined to analyze the cleaning efficiency. Results showed that the lowest height of the flushing gate had the best effects on sediment removal. The flushing waves generated from the sudden opening of the flushing gate were capable of cleaning sediment deposits in the given initial sediment bed thickness, length, and position. The required time duration for cleaning the sediment deposit completely increased about 6, 3, and 3 times when the sediment bed thickness, sediment bed length, and distance between the flushing gate and sediment bed increased 10, 4, and 7 times, respectively.
Both the flow hydrodynamics and sediment scour models were well validated with the experimental results.
The required time duration for cleaning the sediment deposit increased about 6, 3, and 3 times when the sediment bed thickness, sediment bed length, and distance between the flushing gate and sediment bed increased 10, 4, and 7 times, respectively.
Sedimentation in sewers has been recognized as a severe issue since it is related to both hydraulic problems and environmental pollution in urban drainage. The presence of sediment deposits in storm sewers can cause pipe blockage, a reduction in hydraulic capacity, resulting in low drainage efficiency (Todeschini et al. 2010; Regueiro-Picallo et al. 2018; Yang et al. 2019; Liu et al. 2021). Moreover, the pollutants attached to the sediment particles can be re-released and thus form hydrogen sulfide and methane by actions of chemical and/or biological processes, inducing water quality of receiving water bodies, odor concerns, and pipe corrosion (Fan et al. 2003; Liu et al. 2015, 2021; Chen et al. 2022).
The above issues have been studied extensively in the literature. Recently, Tang et al. (2020) considered the clay content and proposed a method for predicting sediment deposition profile. By measuring bed shear stress and flow velocity, Regueiro-Picallo et al. (2020) addressed the relationship between sewer sediment composition and its erodibility in conditions of biological transformations. As sewer overflow contains pollutants such as methane, pathogenic and fecal organisms, heavy metals (Sambito et al. 2020; Liu et al. 2021; Chen et al. 2022), effective monitoring strategies are receiving attention from a few researchers. Yaroshenko et al. (2020) reviewed the real-time monitoring methods and concluded that microwave spectroscopy and chemical materials integration is a new trend in monitoring. To identify the pollution source position, Sambito & Freni (2021) and Sambito et al. (2022) optimized the sensor location in monitoring pollutants in sewers based on Bayesian optimization approach and hydraulic information, and showed its effectiveness in a field test.
To remove the deposited sediment, self-cleaning devices in sewers are considered to be cost-effective techniques (Campisano et al. 2004; Bong et al. 2013; Safari et al. 2017; Yang et al. 2019; Montes et al. 2021; Safari & Aksoy 2021). By storing a certain amount of water in sewers, the valve of a self-cleaning device is opened when the water level attains a certain value. Then the flushing waves generated from the flushing gate wash sediment deposit in sewers. The sediment particles will continue transporting downstream till caught by the collection facilities or mechanical equipment (Campisano et al. 2019). The mechanism of this process is to increase the bed shear stress so that the sediment can be scoured and transported downstream (Bong et al. 2013; Safari et al. 2015). In recent years, this cost-effective technique has been considered as a preventive and reactive method to deal with sediment deposits.
A large number of researchers have studied the influencing factors of flushing efficiency in sewer cleaning. The factors can be mainly classified into flushing hydraulics, pipe geometry, and sediment properties (Montes et al. 2021). These three factors include parameters, such as stored water head (Guo et al. 2004), number of flushes (Bong et al. 2013), flushing duration (Campisano et al. 2004), flow velocity (Sun et al. 2022), bed shear stress (Yang et al. 2019), backwater (Jin et al. 2016), pipe slope (Zhang et al. 2011), pipe diameter (Safari et al. 2018), pipe cross-section shape (Safari & Aksoy 2021), surface roughness (Knight & Sterling 2000), sediment cohesiveness (Regueiro-Picallo et al. 2018; Tang et al. 2020), sediment thickness and width (Campisano et al. 2019; Montes et al. 2021), sediment particles median grain size (Zhang et al. 2011), gradings, porosity, and density (Campisano et al. 2019; Sun et al. 2022). By studying these factors, researchers aimed to address flushing efficiencies in sewer cleaning and to propose simple dimensionless equations for the design and operations of self-cleaning devices.
In analyzing the effects of sediment properties, the deposited sediment height and length are essential parameters in flushing efficiency. Several studies have focused on changes in deposit bed height to quantify the removal efficiency. Bong et al. (2013) conducted a series of experiments on the efficiency of flushing and found the minimum number of flushes increased by 1.5 times as the sediment bed deposit thickness doubled. Campisano et al. (2019) presented that a thinner sediment layer reduced the rate of sediment available for erosion and transport, thus resulting in a decrease in the total volume removed from the channel. Safari & Aksoy (2021) found that the rectangular cross-section channel was the most advantageous shape in flushing efficiency among the trapezoidal, circular, rectangular, U-shaped, and V-bottom-shaped cross-section channels in a scouring model. Liu et al. (2021) performed the sediment scouring and transportation process at different locations of a pipeline, finding that the scouring rate at the front section decreased with the increasing sediment thickness.
However, the above studies were at rapidly varied flow rates, in which conditions the stored volume of water in the self-cleaning device was considered small. In practice, when the valve of the self-cleaning device is opened by the asymmetric pressure induced by the accumulated water, a volume of stored water upstream of the device can be found, as the pipe slope is usually about 0.15% (Shahsavari et al. 2017). That is to say, the volume of stored water in the self-cleaning device is far larger than in those experimental tests conducted in rapidly varied flow rates. Moreover, the position of a self-cleaning device needs to be specified to take effective actions in cleaning the sediment deposit. Therefore, effects of flushing waves on sediment cleaning in the relatively constant water level tank conditions should be studied.
This paper studies the sediment transportation induced by the flushing waves, which are generated from a simplified self-cleaning device. Variable factors of flushing gate installation height, sediment bed height and length, distances between the valve and sediment bed are analyzed by discussing their flushing efficiencies on cleaning sediment deposit. The temporal bed profiles characteristics and flushing time durations required for sediment cleaning completely are obtained. The results will provide useful references for designing the self-cleaning devices, as well as hydraulic information in evaluating or monitoring pollutants in sewers.
To simulate the sediment scouring and deposition processes, a CFD model of Flow-3D (User manual 2013) is applied. The model can track the free water surface by the volume of fluid (VOF) method. It utilizes the fractional area volume obstacle representation to adapt to the grids. These methods increase computing efficiency significantly by eliminating the additional cells in capturing complex geometric regions and tracking the free water surface of these regions.
Sediment transport model
Flume model and boundary condition
The numerical model is shown in Figure 1(c). Specified pressure boundary conditions with a fixed height of the water head that equaled the initial water depth in the tank were applied to the inlet boundary while the outflow boundary was set to the out boundary. The top side of the channel was a free water surface and was specified as the atmospheric pressure. Wall boundary conditions were assigned to the bottom and the two sides along the flume. The standard wall functions were employed in the boundary conditions used for velocity and turbulence at the wall. When the simulation started, water in the tank was released and would generate flushing waves propagating downstream, and thus scoured the sediment bed.
The sediment bed was located at the channel bottom starting at x = 2–4.5 m downstream of the flushing gate. The length of the deposition varied from 3 to 4.5 m depending on the front side and the sediment bed length. The sand bed thickness (HS) was 20–200 mm. Uniform non-cohesive sediment with a medium diameter of 0.5 mm, density of 2,650 kg/m3 and a porosity of 0.65 were used. The sediment angle of repose was 37°. The critical Shields number was 0.0304.
A total of 23 numerical tests were designed, as shown in Table 1. Tests G1–G3 were designed to study the effects of flushing gate height on scouring the sediment deposit while tests HS1–HS7, tests L1–L6, and tests LG1–LG7 were designed to study the effects of sediment bed height, sediment bed length and sediment bed position, respectively. The parameter of td is the time duration that is required for sediment particles total removal.
|Test no. .||HG (m) .||HS (m) .||LS (m) .||LG (m) .||6- LS- LG .||td (s) .|
|Test no. .||HG (m) .||HS (m) .||LS (m) .||LG (m) .||6- LS- LG .||td (s) .|
RESULTS AND DISCUSSION
Validation of the model
The experimental and numerical tests results from Campisano et al. (2004) were used to verify the model. The experiment test was performed by producing flushing waves on cleaning sediment deposit. The rectangular channel was 3.9-m long, 0.15-m wide, and 0.35-m deep with a slope of 0.145%. A 1.3-m long tank was settled upstream of the sediment bed. The distance between the tank and the sand head was 0.3 m. The sand had an almost uniform size in the range of 0.425–0.600 mm. The density of sediment was 2,830 kg/m3. Two initial water levels of 0.1 m (Flushing test A) and 0.13 m (Flushing test B) were tested. No sediment particles were in test A while a 1-m long and 0.03-m high sediment bed was settled in test B.
Flushing gate height
Sediment bed thickness
Sediment bed length
Sediment bed position
To study the flushing efficiency of different distances between the flushing gate and the sediment bed, i.e., sediment bed position, distances ranging from LG = 0.5‒3.5 m in tests LG1‒LG7, as listed in Table 1, were tested. The flushing gate height (HG), sediment bed thickness (HS) and sediment bed length (LS) were kept at 0.2, 0.12, and 1.5 m, respectively, in which LS was a moderate value in tests L1‒L6.
In this paper, the CFD model of Flow-3D, which was validated by experimental and numerical data conducted by Campisano et al. (2004), was employed to study the scouring effects induced by the flushing waves based on the concept of the self-cleaning method. Influences of flushing gate height, initial sediment bed thickness and length, and sediment bed position on cleaning sediment deposit were individually investigated. The main conclusions are in the following:
The sand bed profiles variations were in a very similar trend in each test, i.e., a higher sand bed height was induced by the sudden erosion at the sediment bed head, then backward accompanied by a gradually narrow crest.
From the perspective of flushing time duration, the lower the flushing gate height was, the higher the removal efficiency. The efficiency of flushing waves in cleaning the sediment deposition decreased with the increasing bed thickness while the longer the sediment bed length was the larger the flushing time duration.
The required time duration for totally removing the sediment deposit increased about 6, 3, and 3 times when the sediment bed thickness, sediment bed lenth, and distance between the flushing gate and sediment bed increased 10, 4, and 7 times, respectively.
The flushing waves generated from the gate performed well in cleaning the sediment. All the sediment deposit could be scoured and transported downstream with a distance ranging from 1.0 to 4.5.
The authors acknowledge the support of the National Key R&D Program of China (No. 2022YFC3203200) and the Ningbo Science Foundation (Grant No. 2021J096).
H.F. conceptualized and investigated the study, prepared the methodology, did data curation, and wrote the original draft. S.D. investigated the study, prepared the methodology, supervised, wrote, reviewed, and edited the article. D.Z.Z. conceptualized the study, wrote, reviewed, and edited the article.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.