Effect of large scale roughness of block ramps on dissolved oxygen efficiency in water

b

width of flume

C_d

downstream oxygen concentration

C_s

dissolved oxygen saturation concentration

C_u

upstream oxygen concentration

E_f

relative dissolved oxygen efficiency

E₂₀

dissolved oxygen efficiency at T = 20°

f₁

functional symbol

f₂

functional symbol

f₃

functional symbol

h

height of block ramp

k_s

roughness height

L

length of block

Re

reynolds number

S

slope of block ramps

T

temperature

W

width of block ramp

W_e

weber number

X

roughness arrangement

y

approach flow depth at the ramp toe

y_c

critical water depth

y_t

tailwater depth

y₀

flow depth on block ramps

dynamic viscosity

ρ

fluid density

surface tension

Water pollution results when organic and inorganic substances are introduced into water resources. The presence of these contaminants in natural and manmade open channels lead to decrease dissolve oxygen (DO) levels below the necessary threshold for aquatic species. Apart from improving the performance of wastewater treatment plants, the enhancement and the promotion of self-purification of water flow systems can play crucial roles as an environmental repair in aquatic ecosystems. The river self-purification capacity in rivers polluted by oxidizable organic contaminants might be significantly influenced by the rate of transfer of oxygen from the atmosphere in water. Therefore, the augmentation of DO, as an important indicator of water quality, provides adequate circumstances for the self-purification of waters and protects the aquatic environment as well.

Hydraulic structures can enhance the amount of DO in a river system by rapidly changing the flow regimes and creating turbulent conditions in the water flow in a short distance. A number of studies have been conducted to investigate whether gas transfer efficiency is influenced by changing different hydraulic properties and geometrical parameters of hydraulic structures (Mansori Konsestani et al. (2018), Felder et al. (2019) and Esmaeili Varaki et al. (2021a, 2021b)). The general conclusion of all these studies is that utilizing these hydraulic structures can boost the DO level. The major criticism of these structures concerns the change in river morphology. Sustainability and adaptability of hydraulic structures with water body ecosystems have become prominent topics in the past decade, therefore, a better understanding the causes of gas transfer in rivers can open new horizons in the exploration of how to improve oxygen level naturally and sustainably.

Although temperature and light, as the external environmental factors, play important roles in the dissolve oxygen level in a river, a foremost impacting component is river geomorphology (Vannote et al. 1980; Song et al. 2011; Song et al. 2018). Pool-riffle streams (PRS) are one of the natural complex hydrological flow structures in rivers. Unlike the pool which is characterized by deep depth and low velocity, the steep slope and the large gravels of riffle lead to enhancement of air entrainment, which is comparable to oxygen transfer by macro-roughness in cascades or stepped weirs (Chanson 1993).

Kim (2003) studied the hydraulic parameters effects of riparian riffles on oxygen transfer through the air entrainment. Their filed survey was carried out on a river that was well protected from artificial human activities. It revealed that oxygen transfer was directly proportional to the flow velocity, the flow discharge, the Froude number and the size of the riverbed materials. The comparison showed that the average value of oxygen transfer (E₂₀) of the riffles was higher than that of smooth chuts reported by Chanson (1997).

Among manmade hydraulic structures, block ramps (BRs) function similarly to a pool-riffle stream. BRs are known as an environmentally and riverbed-friendly structures because of their low environmental effects and their capacity to allow for small changes in the riverbed. Other advantages of this structure include high stability, the ability to progressively overcome riverbed height variations while achieving regulated energy dissipation and cost-effectiveness. BRs have absorbed more and more attentions due to their advantages that can be used in hydraulic restoration projects as a good alternative to drops and sills structures (Esmaeili Varaki et al. 2021a, 2021b).

Many studies have addressed various aspects of BRs, including hydraulic characteristics across a range of flows, energy dissipation, stability, sediment transport, and ecological performance. Pagliara & Chiavaccini (2006) experimentally investigated the scale effects of roughness conditions for the BRs with a slope range between 1 V:4H and 1 V:12H. They found that for a certain discharge, a larger-scale roughness led to a greater energy dissipation. Pagliara et al. (2008) carried out experiments on physical BRs models to investigate the effect of tail water condition on energy loss over stepped channels with different bed materials. They concluded that the ramp scale roughness and the ramp submergence condition are the parameters influencing the relative energy loss while the influence of the ramp slope was regarded as insignificant. Pagliara & Palermo (2010) analyzed the scour depth downstream of the BRs in the presence of protection structures for a range of tailwater conditions varying from low to very high submergence conditions. They found that the tailwater depth has a significant impact on the scour geometry and simple relationships were presented to estimate the main lengths of the scour hole downstream of the ramp toe.

Extensive laboratory studies were carried out to investigate the portion of flow passing through surface of the porous block ramp by Pagliara & Lotti (2009). They found that permeable base material had the greatest effect on increasing relative energy dissipation. In other words, the percent of flow rate through the base ramp material increased the relative energy dissipation along a ramp with different bed slopes. Weitbrecht et al. (2017) conducted experimental tests to determine the parameters that provoke the failure of the unstructured block ramp for various bed slopes. As a result, they developed a model for the determination of equilibrium slope on the unstructured BRs in relation to the dimensionless discharge.

The literature review shows that very few studies of DO transfer in water flow over block ramps have been investigated. The most relevant of these is a research undertaken by Rajwa-Kuligiewicz et al. (2020). Their field-based research addressed the impacts of BR morphology and spatial variability of hydraulic parameters (Reynolds number, velocity, Froude number, and shear stress) on DO distribution in real riverine conditions. Their measurements were taken twice at low and high flow discharges conditions. They reported that the discharge and the submergence condition of the rapid ramp are important factors on water temperature and DO level at the BR structure.

Although many studies have investigated energy dissipation over the BRs and scour process downstream of a BR, little research has been conducted to assess the effects of geometrical and hydraulic parameters on DO efficiency of BRs. Therefore, the current study focused on flow patterns and the dissolved oxygen efficiency associated with BRs with large-scale roughness in both free and ramp hydraulic jump conditions; the results were compared with the smooth reference configurations.

The experimental tests were carried out using the laboratory flume (8.5 m long, 0.88 m wide, and 1.0 m deep) of the hydraulics modelling laboratory of the University of Guilan, Iran (Figure 1). A centrifuge pump was used to supply a flow rate of up to 70 l/s. A motor speed controller was used to adjust the electromotor of the pump, enabling the flow discharge to be quickly and accurately adjusted. The discharge was measured by an ultrasonic flowmeter Adakflow model with an accuracy of ±99%. An adjustable gate at the end of the flume allows the operation of the required downstream water level.

Runs were carried out for BRs made of wood and plastic with three different slopes (1 V:3H, 1 V:5H and 1 V:7H). To estimate the impact of surface roughness of the BRs on DO efficiency, two granular materials were selected based on the proposed ratio y_c/k_s, recommended by Pagliara & Chiavaccini (2006). Intermediate and large-scale roughness conditions were obtained by gluing the large-scale (y_c/k_s< 2.5) and medium-scale (2.56.6) granular materials on a wooden plate. The tested arrangement patterns of surface roughness included the structured Stripe arrangement (Figure 2(a) and 2(e)), the structured Stripe staggered arrangement (Figure 2(b) and 2(f)), the staggered arrangement (Figure 2(c) and 2(g)), and the interlocked arrangement (Figure 2(d) and 2(g)). Main features of the tested BR configurations and the important characteristics of surface roughness materials are presented in Table 1.

DO measurements upstream and downstream of block ramps was carried out by using two DO-meter devices Hach HQ30D model. Devices were aligned in the flow direction 0.7 m upstream from the starting point of the ramps and 2.5 m downstream from the toe, where the air bubbles vanished entirely. The device could measure temperature as well. Consequently, both the water temperature and oxygen concentration were recorded at the same time. Drinking water was used for all tests to prevent the effect of chemicals and pollutants from origin water. Since the difference in the upstream and downstream concentrations is proportional to the concentration gradient, the deoxygenating process is an important step in the preparation of water for quantifying the influence of studied geometric and hydraulic parameters on DO value. Therefore, Na₂SO₃ solution with a concentration of 70 (mg/l) was used for chemical deoxygenation. For this purpose, a storage tank of Na₂SO₃ solution was installed at inlet of flume, i.e. 3 m upstream of the BRs. The Na₂SO₃ solution was injected by a multi injector that was inserted along the width of the flume and at mid-depth of flow. The deoxygenation reagents were added to water until the upstream DO reached 1.8 (mg/l), approximately. After disconnecting the injection, measurements were continuously recorded upstream and downstream of the block ramps until reaching the initial saturation level of DO.

In this study, the effect of different rough materials on air entrainment in water and on related DO level was tested. A range of roughness elements including two different particle sizes and four different roughness arrangement are included in the present study. All models were tested under different flow conditions with flow rates varying between 0.025 m³/s to 0.065 m³/s. Hydraulic jumps downstream of steep slope ramps were controlled by a gate. The free hydraulic jumps formed in a cross section where the flow direction was practically horizontal, and the ramp hydraulic jumps were regulated by the gate to locate the upstream of the hydraulic jump in one-third from the ramp toe. A total of 486 experimental runs have been carried out to investigate the impact of flow conditions, ramp slopes and roughness characteristics on DO efficiency.

Although the aim of this study was on air entrainment mechanisms in BRs, observations were undertaken with both smooth and rough BRs to demonstrate the impact of the ramp roughness on the flow regimes. The visual observations showed that the classification of flow regimes was identical for all rough ramp configurations except configurations e and f (Figure 2). The influence of flow separation between spaces of roughness elements on the surface of the BR in flow resistance increased by reduction of flow discharge and the local submergence (Weitbrecht et al. 2017). The local vortexes intensity in the front part of roughness elements was dependent on the surface roughness height. For a constant discharge, the strongest vortex interaction was for configurations with the coarse particles. The transition from roller vortexes to undular vortexes occurred by increasing the flow discharge. A further increase in discharge, consequently, abatement of the resistance of the particles against the flow, the floating vortexes led to noticeable water-surface fluctuation.

The configurations e and f behave like the step-pool systems. The nappe and skimming flow regimes were recognized as the dominant flow regimes over these configurations. For low discharges, the supercritical flow was observed over the roughness elements. The presence of bed particles induced great resistance to the flow due to dissipation of turbulent energy by grains. In the configurations e and f, the space between two consecutive series of surface materials was not enough for generating the hydraulic jumps. Therefore, no hydraulic jump was observed in pools. With increasing the flow discharge, the skimming flow regime with a plane water surface emerged. For high water depth over the ramps, the flow skimmed over the tops of the surface materials and a pseudo-boundary formed.

The air-water flows on a smooth ramp exhibited typical supercritical flow pattern on a surface with steep slope. A boundary layer was generated on the solid surface and the flow path was not sufficient to lead the outer edge of the boundary layer reach the free surface. Therefore, the non-aerated flow with smooth and glassy free surface was observed. Different mechanisms were detected to contribute to air entrainment in flow over the rough BRs. Experimental observations suggested two main sources may include: (1) air entrainment due to the turbulence in a free surface boundary layer shear flow created by surface roughness (2) singular aeration at the hydraulic jump occurring downstream where the flow jet impacts into the downstream pool (Figure 4). According to the macroroughness conditions, three main flow regimes can be observed over the ramps. Small air pockets which were caused by free surface undulations were formed over the upstream part of the ramp. With the increase of turbulent intensity, the aeration was enhanced and promoted the formation and the development of air bubbles by free surface undulations and strong interaction with the bed surface. In this flow condition, the oriented vortices were generated over rough bed. Further increase in the flow rate led to the appearance of skimming flow condition with high turbulence and strong momentum transfer. Therefore, large numbers of small air bubbles were transferred into the water and the interfacial area appears white.

The transition from subcritical to supercritical regimes was accomplished via an internal hydraulic jump at the toe of the ramp. Due to the highly turbulent flow condition and bubbly two-phase flow environment, hydraulic jumps behave as a self-aeration mechanism. The turbulent flow in a hydraulic jump is related to the Froude number of jump and the submergence condition. The highly aerated turbulent flows with large numbers of groups of small air bubbles were developed downstream the ramp due to the free hydraulic jump. However, for the same discharge, the streamwise velocity and the turbulent cavity decayed in the submerged jump. In the submerged hydraulic jumps, the bubbles were bigger in size but much smaller in number than in air–water mixture caused by free hydraulic jumps. The impacts of tail water depth on the air bubble entrainment in hydraulic jumps were shown in Figure 5.

The two equations were developed for free and ramp hydraulic jumpcondition, respectively. In Figure 6, the changes in aeration efficiency versus h/y_c ratio are shown for smooth ramps with slope of 1:3.

The behavior of the flow over BRs with macro-roughness (y_c/k_s < 2.5) and with small-scale roughness (2.5 < y_c/k_s < 6.6) on aeration and corresponding dissolved oxygen efficiency was investigated in this study. The effect of the varying roughness on aeration efficiency of block ramps with the slopes of 1:3 and 1:7 is illustrated in Figure 7. As observed for all BRs, the DO efficiency increases with an increase in dimension of bed materials. A comparison between Figure 6(a) and 6(b) shows the roughness effect was more pronounced in block ramps with 1:3 slope than that of the BRs with 1:7 slope. This is often explained as due to the breaking of large eddies into the smaller eddies by blocks (Tritico & Hotchkiss 2005). As the slope of BR increases, the flow interaction between eddies and surface roughness materials increases. Therefore, the breaking of larger eddies into smaller eddies increases the magnitude of DO efficiency along the block ramp length.

Figure 8 and Table 2 show the relative efficiency of DO (E_f) of the BRs tested with the slopes of 1:3, 1:5 and 1:7 (V: H) at 2 mg/l upstream DO concentration for different hydraulic jump conditions. A comparison of results indicates that the S₁R₂A₁ configuration in macro-roughness condition (y_c/k_s < 2.5) and under a free hydraulic jump condition has the highest DO efficiency. The S₁R₂A₁ configuration with an average E_f value of 1.45 increased the aeration efficiency by 47%. However, for the BRs with the slopes of 1:5 and 1:7, the S₂R₂A₃ and the S₃R₂A₂ configurations had better performance when compared with the other geometries that were tested. Under the macro-roughness condition (y_c/k_s < 2.5) and the free hydraulic jump condition, the aeration efficiency (E₂₀) of the S₂R₂A₃ and the S₃R₂A₂ configurations increased by 38 and 32%, respectively. The results indicate that, for all configurations, the aeration efficiency increases with increasing the slope of the ramps. It could be due to the breaking of larger eddies by the interaction of wake generated by roughness lying on the bed of the BRs. Further study is needed to investigate the relation between the turbulence generated by surface roughness and the further production of turbulence in form of wake.

Since the upstream concentration of DO for the S₁R₀ configuration with free hydraulic jump became larger (C_u = 2, 3 and 4 mg/l), there was a decreasing value of E_f to 0.88, 0.54 and 0.32, respectively. However, in case of ramp hydraulic jump corresponding E_f values were 0.59, 0.36 and 0.21. For the S₁R₁A₁, S₁R₁A_2, S₁R₁A₃ and S₁R₁A₄ configurations, which had the best performance, under the free hydraulic jump, the average values of E_f for the C_u = 2, 3,4 and 5 mg/l were 1.43, 0.88, 0.5 and 0.29, respectively. Under ramp hydraulic jump, corresponding average E_f value in the S₁R₁A₁, S₁R₁A_2, S₁R₁A₃ and S₁R₁A₄ configurations measured 1.1, 0.7, 0.38 and 0.2. A comparison of results shows that, in all configurations, the aeration efficiency decreases with increasing the upstream concentration of DO. In Table 3, a comparison of results between the performance of the most efficient configurations and the smooth BRs with slopes of 1:3, 1:5 and 1:7 was shown.

Four nonlinear regression equations, Equations (9)–(12), provide good accuracy for estimating the efficiency of dissolved oxygen in the water flow on block ramps (see Figures 9 and 10).

BRs are characterized by a supercritical flow condition and energy dissipation due to the steep slope of ramp and macroroughness materials of the ramp surface. This study demonstrated the aeration processes over block ramps and corresponding DO concentration. All the experiments were conducted for clear-water conditions and in the presence of both free and ramp hydraulic jump positions. Three ramp slopes 1:3, 1:5 and 1:7, eight different uniform ramp bed materials were tested for a wide range of flow discharge. It was proven that the hydraulic jump position has a strong effect on the DO efficiency. It was observed that change hydraulic jump positions, from free to ramp, causes a reduction in aeration efficiency, as the bubble penetration depth was significantly reduced by downstream pool. The effect of roughness elements size on DO is not negligible. In the presence of large roughness elements (y_c/k_s < 2.5), the DO efficiency increases by forming high amount of eddies and turbulence. Empirical equations was derived to express the DO efficiency of BR based on dimensionless parameters. A good correlation was obtained between experimental data and the results in the proposed model.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

This article does not contain any studies with human participants or animals performed by any of the authors.

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

The authors declare there is no conflict.