Mitigation Strategies to Reduce Detrimental Effects of Air Bubbles in Shaft Spillways: A Literature Review

Spillways are water structures with or without gates utilized to drain water from a reservoir [1]. Spillways are considered the most significant and major structures in the dam's life. The components of spillways comprise three main parts: the inlet component, the transport component, and the outlet component [2]. The competency and well-done performance of these structures demand an accurate and credible design. Spillways can be found in different forms: Free Over-Fall, Side Channel Spillway, Chute Spillway, Shaft Spillway, Siphon Spillway, Cascade Spillway, and Tunnel Spillway. The most popular type is the Morning-Glory spillways, which are standing structures inside the reservoir of the dam that play a significant role in the dam's performance [3]. The Morning Glory spillway can be named as a Shaft spillway, which is a kind of drain flood from the reservoir of the dam. This kind is built in narrow valleys where it is hard to construct another type of spillway, and also in ground or rock fill embankments where the spillway should be built at a distance from the dam shaft to minimize the hazard of exceeding. Despite all the benefits of this kind, caution in design should be taken to avoid any constitution of vortices, cavitation, and air bubbles in the inlet of the spillway [4]. The presence of air pockets in hydraulic systems can cause various problems, such as decreased strength, disruption of flow, reduced cross-section, and damage to conduit and pipeline materials. It can also alter the characteristics of the fluid and raise ecological concerns at the discharge point [5].

• Air pockets minimize the efficient cross-section of the pipe, which leads to a reduction in the hydraulic ability of the pipe.
• The fluid properties, specifically density and elasticity, are altered.
• The presence of air pockets can alter both the structure of flow turbulence and the shear force exerted on the walls.
• When a pipe is sealed, the transition from partially filled to filled flow can result in structural vibrations and surges in the flow.
• The presence of air in a system can generate flow disturbances, such as blow-out or blow-back, which can result in vibration, structural damage, and instability of the water surface.

This paper addresses the mechanism of air bubble formation in closed pipes, drop shaft spillways, and Morning Glory spillways. Also, this paper discusses the negative effects of air bubble formation that occur during the flow process and the damage that occurs because of this phenomenon and also provides a comprehensive literature review of strategies that have been used academically and practically to mitigate the formation of air pockets and air bubbles.

This article has used a methodical study of the literature process. Since "Google scholar" and "Research Gate" index an extensive range of hydraulic and water-related engineering and literature in the form of journals, conferences, and book chapters, we chose these two databases to analyze text mining literature that has been issued in air pockets phenomena in hydraulic structures. Additionally, these scientific websites include a number of alternatives for users to conduct advanced literature searches. As a result, we pulled the research papers for this study from the "Research Gate" and "Google Scholar" databases. First, the study papers were retrieved from the January 2024 database indicated. We selected the search terms "air bubbles," "air pockets," "shaft spillways," "slug flow," "blowbacks," and "Geyser" in order to retrieve the research papers. For integrating the search phrase, we used the Boolean "AND" and "OR" operators in the search field. For the first download of published research material, we used a search on "Article Title" and "Article Keywords." From the query result, we extracted 174 research publications in the first phase. After limiting our study to publications published in journals exclusively in the second phase, the total number of articles was down to 133. The number of research publications dropped to 92 in the third phase when we narrowed our focus to the Hydraulic and Water Structures field. In order to weed out unsuitable publications, we reviewed the abstracts, titles, and keywords of the chosen papers in the last stage. Papers that did not pertain to the goal of the systematic review process—namely, the effects of air pockets on hydraulic structures and strategies for mitigating them—were omitted. In the end, 56 research articles that were perfect for the pertinent subject were received. The procedure of selecting articles for review and the total number of papers (N) in each phase are depicted in Figure 1.

1.png

Figure 1. Research framework

The study's statistical chart, which shows the number of papers published between 2010 and 2023 on the relevant topic, highlights the dearth of scholarly studies that address the effects of air pockets and corrective measures to mitigate their hazards in hydraulic structures. The remaining papers reviewed were published prior to 2010, as shown in Figure 2.

2.png

Figure 2. Papers reviewed between 2010 and 2023

Figure 3 shows that most academic studies and research focused on air management by providing mechanical solutions to remove air bubbles and air pockets from hydraulic installations, followed by research dealing with incidents related to both slug flow and blowback, while a smaller number of studies focused on the issue of geyser. There are significant limitations in addressing the issue of cavitation in pipes and spillways.

3.png

Figure 3. Research topics that reviewed

Air rises and adheres to the top of a horizontal conduit when a water-air combination enters. Air pockets form when air bubbles congregate. Geysers form when water flows down a canal under varied hydraulic conditions. The condition that must be met is the preservation of air within the flowing water in the channel. The slope of a straight conduit with a homogeneous cross-section and surrounding pressure and flow velocity affect air pocket movement [11, 15, 16]. Due to increased flow discharge, pockets shift closer to the downstream end and are pushed out into the tailwater. They show backlash by moving against the current.

Vasconcelos and Wright [17] mentioned that air pockets can be created at placements such as drop shafts and air vents where the air outlet can be closed using water over the conduit crown, and air pockets can be developed. Pothof and Clemens [18] stated that air–water flow is an unfavorable case in several systems for the transition of fluid that results in air pockets, in which air in tunnels may get entrapped and form air pockets that can negatively impact the system.

A possible explanation for the production of air pockets in closed sewers, which is not necessarily connected to fast-filling circumstances, is the result of air being trapped in the flow produced by the sudden drop in drop shafts, as explained by Chanson [19]. Even small air bubbles in falling water might be transferred downstream by sewage flow. These bubbles will cluster at the top of filled sewer pipes and produce air pockets [17].

7.1 Slug flow caused by air bubbles

The occurrence of slug flow, characterized by the formation of air bubbles, is widely regarded as a detrimental phenomenon. Multiple studies have been conducted to forecast its happening and devise strategies to alleviate its adverse consequences. Slug flow refers to a fluid input system that is distinguished by the existence of liquid slugs alternating with sizable air bubbles. The formation and entrapment of these air bubbles occur during the slug input event, resulting in flow pauses and disturbances in pressure and velocity. Mandhane et al. [29] identified five distinct inflow types for slug flow: stratified, wavy, plug, slug, and bubbly conduit. The UC Multiphase Pipes Flowing Data Bank contains 5,935 flow pattern observations that are checked against the various flow pattern maps shown in Figure 8 for two-phase gas-liquid flow in horizontal pipes.

8.png

Figure 8. Suggested flow pattern map [29]

It is demonstrated that the novel flow regime correlation, which builds upon the research conducted by Aziz and Gorier [30], agrees with the data more closely than the remaining correlations that were examined. Included is a FORTRAN application for this correlation.

It's also noteworthy that not a single one of these maps was able to accurately predict the regime of distributed bubble movement. Actually, there is some evidence to suggest that the data—at least the flow pattern observation—may not be entirely accurate. Plotting all 5,935 points on VSG, VSL coordinates reveals that a significant portion of the observations dubbed "dispersed bubble" are located in the area with high gas rates and low liquid rates.

In a study conducted by Ansari and Arzandi [31], the researchers examined the effects of ribs on the flow characteristics in slug flow under atmospheric pressure. They experimented on the adiabatic two-phase air-water flow in horizontal, ribbed, smooth ducts shown in Figure 9.

9.png

Figure 9. Experimental work [31]

• In a laboratory setting and at about atmospheric pressure, the study was carried out in a rectangular duct. The three placements of the rectangular ribs were the top wall (air side), bottom wall (water side), and both bottom and top walls (water and air sides). The ribs measured 1, 2, and 4 mm in height. The flow regime's area was influenced by the ribs, but neither its type nor shape was altered. The findings led to the following conclusions:
• The study was able to distinguish between several flow systems and their limits. It was revealed that the frontiers of several systems are affected by the rib position. At low velocities, the lack of dynamic stability was noted.
• The flow phases in the ribbed duct were categorized as stratified, wavy, plug, and slug flow because they resembled those in the smooth duct.
• The ribs shifted the transition lines towards lower gas velocities and altered the flow diagrams in comparison to the smooth duct.
• The borderline that separates stratified and slug flow regimes stayed constant in length while the ribs were situated on the top wall, and it tended to decrease in gas velocity as the rib height rose.
• The borderline separating the stratified and slug flow regimes inclined towards lower gas velocity when the ribs were situated on the bottom wall. Moreover, the line's length shrank and ended at e = 4 mm.

This study made a valuable addition to the understanding of how ribs affect the slug flow regime in conduits. Nevertheless, further research is required to ensure the validity of the experimental model. Additionally, more research can look into the impact of ribs at depths greater than 4 meters.

The impact of viscosity on the stratified-to-slug transitions in horizontal conduits between 2.5 cm and 9.5 cm diameters was investigated by Andritsos et al. [32]. With glycerine-water solutions with viscosities of 1, 20, and 100 cP, transitions were seen. At the inlet, the two phases were mixed using a straightforward T-junction entry. The liquid and gas were delivered into the 9.53 cm pipe at two distinct points.

Two parallel-wire conductance sensors that extended vertically throughout the whole pipe cross-section were used to determine the depth of the liquid layer running down the pipe's bottom, or hL, in one test section. The time average of the readings from the conductance probes was used to represent the liquid height in studies involving small-wavelength waves. In a second test segment, the liquid height fluctuation along the pipe circumference was measured using pairs of short parallel 0.51 mm chromel wires spaced 45 degrees apart. The two sections have been divided by 10.2 cm in the 2.52 cm pipe and 26.7 cm in the 9.53 cm pipe. The viscosity increases but does not affect the air-water flow. According to experimental findings, there is a stabilization in this kind of plot when liquid viscosity increases when it comes to the commencement of slugs.

There are certain limitations to the study that warrant further investigation. The mechanism investigated by Lin and Hanratty [33] appears to be supported at first glance by the observed influence of liquid viscosity and the close match of the experimental findings for air-water with the prediction. A more thorough analysis of the data using extremely viscous liquids, however, revealed that the influence of pipe diameter differed from what was anticipated.

Unknown nonlinear processes are responsible for the large-wavelength disturbances' emergence from the normal wave train. The mechanism underlying the initiation of slugs in liquids with viscosities similar to water is difficult to observe because, under a wide range of flow conditions, waves (caused by variations in gas-phase pressure in phase with the wave slope) cover the interface of the stratified flow before slugs appear. This study provided information about the impact of viscosity on slug flow; more research is needed to lessen the influence of viscosity on the slug-flow regime.

Matsubara et al. [34] explored air-water flow systems in a micro-channel to establish their boundaries. The water phase had speeds from 0.033 to 4.94 m/s, whereas the air phase had speeds from 0.025 to 66.3 m/s. Previous conduit inflow research focused only on inflow technique and narrow pipe pressure distribution. In their study, Aghebatie and Hosseini [25] conducted a computational analysis of the slug flux phenomenon across a range of ratios of air intake speed to water intake velocity, spanning from 1.1 to 34. This work uses CFD techniques to numerically explore two-phase flows using the InterFoam solver in OpenFOAM, an open-source software. The majority of earlier troubling research on slug flow was conducted in tiny microchannels where the air bubbles' expansion was barely noticeable. On the other hand, this study looked into systems that experience significant pressure decreases and thus have sudden increases in air phase volume. A culvert's intake could be used to induce the slug flow phenomenon by entering water and air at different speeds, with the air to water velocity ratio ranging from 1.1 to 34. First, the temporal and spatial fluctuations of velocity and pressure along the culvert were the main focus of this investigation. The non-dimensional characteristics impacting the slug flow phenomenon are then isolated and examined using dimensional analysis. The position and form of ribs are among the various tactics for lessening the damaging effects of slug flow that are assessed. Using the InterFoam model, the velocity and pressure fields in a slug flow along a conduit are examined. The flow pattern and turbulence effects are modeled using a two-phase approach with the SST turbulence and InterFoam solver model. The Volume of Fluid (VOF) model is used to monitor the air-water interface. The more practical turbulence and mesh size model are chosen by means of verification testing. Analysis is done on the pressure and velocity fields as well as the presence of slug flow. In a conveyance system, strategies to stop the slug flow and lessen its damaging consequences are examined by using the ribs, their shapes, and their positions. The researchers verified the influence of the ribs on the slug incident. The assessment of slug flow damage mitigation strategies is conducted by evaluating pressures, velocity, and disturbances. According to the study, ribs decrease pipe pressure and velocity by a factor of 4 to 6 for V_air/V_water values less than 10. The presence of stepped ribs in the stream bed effectively decreased the population of slugs. This research may reduce the incidence of slug-flow in horizontal conduits. To study the rib efficiency in these structures, it is advantageous to use these techniques in hydraulic structures that have more intricate shapes than conduits like spillway structures.

Experimental data are presented by Gualtieri and Chanson [35] from measurements made in a rectangular horizontal flume with slightly developed inflow conditions, as shown in Figure 10.

10.png

Figure 10. Test setup [35]

For inflow Froude number Fr₁, the vertical ranges of the void fraction and the air bubble count rate were noted in the range of 5.2 to 14.3. Near the jump toe, a rapid detrainment phase was seen, but over longer distances, the air diffusion layer's structure was clearly visible.

These fresh data were contrasted with older data, which were often gathered at lower Froude values. The comparison showed that the greatest void percent C_max grows with the raising Fr₁ at a specified distance from the leap toe. The highest possible void fraction and bubble count rate's vertical placements matched those of earlier research. The researchers discovered air bubble proportions that were at their maximum, which aligned with findings from earlier studies. Furthermore, the higher boundary of the air dispersion layer increases linearly from the leap's start.

The research spillway and drop shaft slug flow has concentrated on pipe sections, steady-state circumstances, and 2D numerical simulations. Prior studies ignored the effect of inflow mode on spillway slug flow velocity. Previous studies ignored spillway slug flow management. Aghebatie and Hosseini [26] studied Alborz morning glory slug inflow using 3D numerical simulation. Slug flow was reduced in tunnels with different slopes and diameters using various methods. These approaches are evaluated by measuring spillway flow hydrodynamic parameters including pressure and velocity. Aghebatie and Hosseini [26] studied tunnel slug generation, growth, and hydraulic parameter changes. Open FOAM was used to simulate the Albornz spillway in 3D. Figure 11 shows the Alborz dam spillway numerically.

11.png

Figure 11. Alborntz Glory spillway 3D numerical simulation [26]

Multiple discharges were simulated numerically. Both modes show symmetry as the inflow ascends to the peak and then lowers into a vertical column. The flow is then accelerated along the column, effectively trapping air bubbles within the liquid. The experimental and numerical discharges were compared in the reservoir for six different water levels. The distribution of pressure along the x-axis was analyzed and depicted in Figure 12 for three discharges in the numerical model. There was a strong agreement between the two methods, as shown in Figure 12.

12.png

Figure 12. Investigation of the numerical model [26]

The text refers to a study conducted by Aghebatie and Hosseini [26], which examines Figure 12 of a numerical model.

The presence of many influx zones within the tunnel is determined by elements such as the movement of air, the speed of air and water flow, and the shape of the tunnel. Small discharges in the tunnel create a fluctuating water surface, with the height of the waves increasing in proportion to the discharge rate. When water waves encounter the apex of a tunnel at specific flow rates, their velocity decreases, resulting in disturbances and vibrations throughout the tunnel, perhaps leading to incidents of obstruction. Choking results in a decrease in the tunnel's ability to release fluid, causing the control section to change from a crest to an orifice. During the process of water transitioning from a wavy surface to being filled, many air masses become trapped and form within the tunnel, resulting in pressure disturbances. Figure 13 illustrates a pipe that is blocked at the vent hole and is currently undergoing slug flow mode. This mode is distinguished by a series of liquid slugs alternating with sizable air bubbles. The presence of this unfavorable flow system causes disturbances in both pressure and velocity within the tunnel, leading to increased strain on the tunnel or conduit. Additionally, the liquid creates tiny bubbles that mix to form a tube-wide layer. Curvature angle, diameter, slope, and water discharge affect tunnel slug flow. Aghebatie and Hosseini [26] used several methods to prevent slug flow and promote air evacuation, improving the Morning Glory spillway's efficiency.

13.png

Figure 13. Slug flow constitution in slope tunnel [26]

Aghebatie and Hosseini [26] showed that the performance of the Morning Glory spillway can be improved when the diameter of the tunnel is larger than the height of the water above the crest. Also, this study was discussed extensively as a result of its significance in addressing slug flow incidents using the developed 3D CFD technique, which is considered the most advanced model to study any complex problem in general and two-phase flow in drop shafts specifically, comparing to limitations in other methods to address issues related to slug flow incidents. This study could simulate accurately the slug flow incident in Morning Glory spillways. Moreover, this study could address several parameters and their impact on slug flow, and due to its significance, more efforts should be adopted this technique in this field to address more extensive parameters.

7.2 Air management strategies

Previous studies have mentioned various measures to manage air in conduits and spillways for the mitigation of air bubble formation to prevent the detrimental effects of air bubbles that are formed in the Morning Glory spillway due to air entrainment at the intakes of conduits during mobilization. In general, the strategies of air management are classified as follows: air banning, taking into account hydraulically eliminating air bubbles from spillways, and applying of air-emission tools, like air valves or air vents.

In order to investigate the hydraulic characteristics of air in pipes with small diameters—25.4 mm and 38.1 mm—May et al. [36] carried out laboratory studies. The objective of the study was to examine the relationship between the essential velocity and pipe's slope and diameter, as well as the relationship between the velocity of the air pocket and the head loss it causes. Notable conclusions from this inquiry were as follows:

Although critical velocity is usually represented by the dimensionless critical Froude's number, this confuses the relationship between velocity and diameter. It was discovered that a bigger-diameter pipe requires a larger clearance velocity when critical velocity is applied (Figure 14).

14.png

Figure 14. Critical velocity versus conduit inclination [36]

Physically speaking, this makes sense if one assumes that, even with the length held constant, the larger tube would possess a higher volume air pocket. Since the buoyant force in the bigger diameter pipe could be higher than in the smaller pipe in this scenario, moving the pocket would need a higher inertial drag force and, thus, a higher velocity.

As shown in Figure 14, tests conducted in the 38.1 mm diameter pipe across air pocket lengths of 38, 152, and 279 cm reveal that the critical number for Froude was constant with pocket length. This conclusion is unexpected at first because a larger air pocket capacity would be expected to have a larger buoyant force, requiring a larger inertial drag force for moving the pocket. Often, at critical velocity, a pocket of air wouldn't necessarily pass through the structure intact but would instead gradually break into distinct fragments that would then proceed downstream. This theory is supported by data. This would continue until every pocket was removed, often in a short amount of time.

The link between critical velocity and pipe diameter is not indicated in any of the other investigations. Based on two sizes, the study's findings indicate that the bigger diameter pipe needs a higher clearing velocity, as shown in Figure 14. Additional confirmation of this conclusion across a larger range of sizes is required. Without a doubt, surface tension—which this study did not take into account—will become more significant as the diameter drops and eventually take control.

It is arguable that head loss from air in pipes significantly affects a system's hydraulic integrity. This head loss reduces the energy available to drive flow when air builds up, leading to subpar performance or total blockage. The general hypothesis that the vertical height of the pocket can be used to approximate air pocket head loss was validated by laboratory tests.

Less than 10 percent of the mean flow velocity was discovered to be the velocity of an air pocket traveling downstream in the direction of the flow. The inertial drag force is the main force driving the pocket, while buoyancy and friction act as barriers. Because of the increased buoyant force, the pocket velocity falls as pipe slope increases.

The critical velocity for a pipe with a diameter of 25.4 mm and 38.1 mm differs measurably, however, the diameter has no bearing on the speed of an air pocket.

Automated air release valve installation at pipeline's crucial locations is a popular and usually efficient way to remove air through a system. Placing air release valves has several drawbacks, including increased initial and ongoing operating costs, which can be substantial for emerging nations' communities, system complexity, and the need for sporadic maintenance. The fact that such valves are usually unavailable, even in big cities, as well as rural ones, is another issue.

The constitution of the air vortex considers important hydraulic problems that occur at the water conduit inlet which leads to the constitution of air vortex. If there is insufficient submergence at the entrance, such as in the event of a vortex, there is a possibility of a significant amount of air entering the pipes. The crucial submergence must be established at the exact moment that air entrapment begins to avoid inlet entrapment. The critical ratio is affected by the intake pipe diameter, water velocity in the inlet and conduit, and input basin bottom permeability [37, 38].

The phenomenon of air bubble transmission in channels with a downward slope is regarded as an intricate issue. Air bubbles will ascend when the buoyant force surpasses the gravitational force acting in the opposite direction. However, if dominance was established through the application of a pulling force, it may be deduced that pockets of air would be transported in the same direction as the flow [27, 39, 40]. Nevertheless, the complex array of issues related to two-phase flow has restricted the majority of theoretical inquiries on this topic. Therefore, most studies investigating the movement of air bubbles in water quality rely on experimental analyses. Wisner et al. [41] identified two possible techniques for eliminating air bubbles from spillways: sweeping, which entails achieving a critical water velocity to completely remove air bubbles, and generation and drag, which involves utilizing a hydraulic jump to displace the air and eliminate large air bubbles at the end.

Multiple studies have primarily concentrated on conducting experimental experiments to investigate air elimination. These investigations have emphasized the crucial task of determining the optimal flow velocity needed to eliminate air from spills. Extensive research has been conducted on the process by which air pockets transition, resulting in the determination of the critical water velocity at which the transition of air pockets occurs downstream. These criteria can identify and fix spillway air entrainment. The critical velocity is correlated with numerous characteristics, including the Froude number, Reynolds number, surface tension, and spillway tilt, according to dimensional analysis. In spillways 175-200 mm or greater, viscosity and surface tension are less noticeable [18, 42]. Further research is needed to demonstrate the influence of surface tension and viscosity in shaft spillways.

Escarameia [22] evaluated the feasibility of hydraulically clearing accumulated air in pipes by conducting experimental studies for the motion of air pockets in a 150 mm pipeline diameter with varying slopes (0 to 1:200). The test results were used to create a prediction equation that calculated the critical velocity needed to eliminate air from a water-carrying conduit. These results supported certain earlier conclusions and added new, crucial information, including: an equation for estimating the critical mean pipe velocity for air pocket movement was calculated from the experiments, demonstrating the dependency of the critical velocity on the air pocket and slope size. In upward slopes, no flow is required to move the air pocket, and no effect of pipe material is anticipated for pipes with a diameter larger than 20 mm. However, in real applications, the possibility of long air pockets being retained at joints should not be discounted. A large air pocket will require a critical velocity of about 0.8 m/s to be expelled; steeper slopes would necessitate greater critical velocities, of the order of 1.2 m/s for a 1:200 slope pipe. Even though the current study was carried out in pipes with diameters of the order of meters, further research is required to address any potential scaling effects of translating laboratory inquiry results to pipes with diameters on the scale of meters.

To expand on Kent's findings, we suggested the following relationship for the clearance flow number, which accounts for slight downward sloping angles less than 15º:

$F_c=\frac{V_c}{\sqrt{g D}}=0.61+0.56 * \sqrt{\sin \theta}$

These results are subject to the same constraints as those listed in Kent's paragraph. Additionally, a definition for the air-water discharge proportion was derived [22], and it was confirmed at pipe angles up to θ = 16.7º.

$\frac{Q_g}{Q_w}=\frac{F_g}{F_w}=0.0025\left(F_{r 1}-1\right)^{1.8}$

Escarameia [22] took into consideration uncertainties associated with air entrainment in conduits and incorporated a safety factor into these equations. However, Escarameia [22] identified differences in the curves indicated by several research studies, such as whether the curves were convex or concave. Multiple studies have performed experiments in a channel with a consistent diameter and steep inclinations. Nevertheless, they have offered a diverse array of applications for their equations that establish a connection between the crucial velocity of flow and the pipe's diameter, inclination, and gravity.

Nevertheless, the trustworthiness of D remains uncertain because there is not enough empirical evidence to support it. Moreover, the establishment of these databases and standards is derived from studies that especially concentrate on particular cases. To fully comprehend the greater range of applications, it is important to have a thorough grasp of airflow, which includes taking into account the impacts of different scales.

An equation developed by Escarameia [22] can be employed to compute the volumes of air bubbles within the specified range of n (0.0002 to 2). Escarameia [22] also examines flow systems that transition from bubble flow to plug flow with low airflow rates. However, it has been observed that lower air-flow values lead to lower cleaning flow values [18]. Therefore, it is essential to take into account these scale effects in order to effectively extrapolate experimental findings.

7.3 Hydraulic jump technique

The procedure of air entrainment using hydraulic jumps entails the technique of removing air bubbles from spillways with a downward inclination. This suggests that the hydraulic jump approach induces the fragmentation of substantial air pockets into smaller air bubbles, resulting in the phenomenon of air entrainment. If the flow velocity exceeds the critical velocity for air bubble removal, these tiny bubbles will be transferred downstream. Multiple empirical studies have examined hydraulic jumps in restricted pipes caused by an abrupt bottom slope change after a gradual transition. Figure 15 shows these phenomena at siphons, after a control gate, and in the presence of an air pocket contained at a higher elevation or down a downward slope during pressurized inflow. Furthermore, a hydraulic jump occurs when there is a transition from supercritical flow to subcritical flow. This hydraulic phenomenon causes significant local disturbance, resulting in the dissipation of energy and air entrainment. The entrapped air bubbles might be carried downstream or moved upstream due to their buoyancy, which depends on the flow's transmission capability.

15.png

Figure 15. (a) Change in lower inclination; (b) siphon; (c) sluice gate; (d) air pocket; (e) a profile presents hydraulic jump in an inclines profile with downstream pressurized flow [43]

Inflow conditions at the inlet and exit dictate the hydraulic jump location. Wang and Chanson [44] confirmed that physically replicating hydraulic jumps in well-defined flow conditions is accurate. The researchers measured the hydraulic jump roller void ratio, bubble generation, and interface velocity. Air trapped in the system was also measured. Air transition through hydraulic jumps has been studied in horizontal channels with circular cross-sections [45] and conduits with downward and upward inclined circular cross-sections. An air pocket will maintain its integrity and unchanged form while the water discharge is augmented downstream. However, when air is removed from the system, the air bubble shrinks dramatically in a sloped pipe, raising the hydraulic surge. Nevertheless, the spillway entrance air pocket remains constant [22]. Air bubbles and hydraulic jumps are considered when designing air movement regulations. In order to investigate the behavior of air pockets and bubbles downstream of a hydraulic jump situated at the end of a sizable air pocket, Gonzalez et al. [46] suggested a linear equation. To validate the equation's practical use, both theoretical and experimental research was conducted. The formula was developed using research conducted by researchers who came after Kalinske and Bliss [45] as well as by those who did not. The suggested linear relationship is:

$Q^2 w / g D^5=S$

Undefined water flow level, Q_wwater flow rate [m³/s], D pipe diameter [m], g acceleration resulting from gravity [m/s²], and S pipe slope [-] are the terms that make up the expression Q²w /gD⁵.

In order to analyze the air bubbles and pockets, Q²w /gD⁵ is compared to the pipeline system's downward-sloping pipe sections. Air bubbles and pockets go downstream through the pipe when Q²w /gD⁵ is higher than the pipe slope. The air pockets and bubbles will go upstream when it is smaller than S. The air behaved as predicted by the linear equation, which was confirmed by measurements and observations made in an experimental device. The experimental study was conducted using a physical model of 101.6 mm diameter acrylic pipes. It is necessary to look at this equation for other pipe diameters.

Aydin and Ulu [47] conducted an inquiry to analyze the aeration of the flow through the aerator holes located on the hood, in order to prevent damage caused by cavitation in a high-head siphon-shaft spillway. A 3D computational fluid dynamics (CFD) methodology, utilizing the finite-volume method, was utilized to solve the Reynolds-averaged Navier-Stokes (RANS) equations. The work focused on analyzing the flow behavior of incompressible viscous and turbulent fluids in full-scale two-phase numerical models. The Volume of Fluid (VOF) approach was employed to model the motion of two phases (water and air) by allocating a volume fraction to each fluid across the whole solution domain. A thorough investigation was performed using reliable CFD models to assess the efficiency of the aerators that were mounted on the siphon hood. These aerators aim to decrease the sub-pressure within the siphon shaft and minimize the occurrence of cavitation. Three distinct aeration diameters were taken into account for three separate working heads. The analysis results indicate that as the pressure in the siphon with the water level lowers, the air entrainment rate increases, while the discharge performance experiences a minor decline. Given that siphon pressures diminish when water levels are low (operating heads), the likelihood of cavitation also increases. In this context, an elevation in air entrainment will effectively mitigate the risk of cavitation. Furthermore, it has been noted that the size of the aeration diameters has a substantial impact on the process of air entrainment, even when the water level remains constant. Excess air entering the siphon can interrupt the flow. Therefore, selecting the right aeration diameter is critical. This study also showed that well-formulated numerical models simplify hydraulic structure design.

The study conducted by Li et al. [48] investigated the flow patterns and air requirements of a hydraulic jump in a closed pipe with different depths of the submerged outlet. Previous research utilized the measurement of outlet depth to establish flow patterns including submerged exits. The experiment utilized a horizontal Perspex pipe measuring 2 meters in length and 0.20 meters in diameter (D). It also involved a control gate with a variable opening (a) and a water tank located downstream of the pipe. The water tank was equipped with an adjustable gate to vary the depth of discharge.

To evaluate air needs, Liu et al. [49] observed bubble motion and size distribution. As shown in the photo, the bubble was ellipsoidal, with the major axis horizontal and the minor vertical. A horizontal third axis perpendicular to the image plane was identical to the primary axis. Liu et al. [49] assumed this because the horizontal axes differed by less than 10%. The study found that free-surface flow without a hydraulic jump has a greater estimated relative air demand (β) than other flow regimes. The Froude number, denoted as F₀, has an impact on the relative air demand and hydraulic jumps in free-surface flow at the vena contracta. A hydraulic leap occurs when a pressurized pipe undergoes a shift from high-velocity flow to low-velocity flow. This transition has the potential to produce either an immature or mature surface wave. The air demand, relative to the surrounding air, varies between 3% and 14% and decreases as the length of the roller increases until it achieves its maximum capacity within the pipe. In the scenario of a hydraulic leap where only a part of it is submerged, the amount of air required is significantly reduced to less than 1%. The air demand was measured in a prototype low-level outflow at the Hugh Keenleyside (HLK) Dam in British Columbia, Canada. The outlet showcased a half-submerged hydraulic jump. The observed air demand corresponded to the prediction equation established from experimental observations. Furthermore, several nozzle diameters were tested on the upper part of the air vent to assess the influence of air delivery on the hydraulic jump. The air supply was restricted, resulting in a decrease in pressure below atmospheric levels. The sealing of the air vent led to a substantial increase in the magnitude of this sub-atmospheric pressure. The hydraulic jump moved in the opposite direction of the flow as the air demand decreased, caused by the lower air pressure within the enclosed conduit. The study revealed a negative association between air supplementation and complications in the test. Thus, this investigation necessitates the validation of the obtained outcomes.

7.4 Mitigation strategies blowback event

The blowback action of the bottom overflow of the Berg River Dam in South Africa was studied using numerical and physical model studies by Bosman et al. [28]. Behind the emergency gate, in the conduit, is an air vent. On June 12, 2008, blowback observations were recorded during the emergency gate's commissioning test shutdown. A possible option is to use numerical simulation techniques (RNG $k-\varepsilon$ & VOF) to address the problem of excessive airflow. A 1:14.066 scale was used in the construction of the model.

Using a large enough scale is essential when implementing the Froude law in order to reduce the scale effects that result from not adhering to the Weber and Reynolds laws.

Eight distinct points inside the model were used to gather static pressure data throughout the physical model tests. Furthermore, the airshaft's highest point was used to measure both the velocity and direction of the air.

There was a striking similarity between the flow patterns in the physical model and the computational fluid dynamics (CFD) model. It should be mentioned, nevertheless, that the CFD model's discharge rate was discovered to be larger than the physical model's.

When the pressure results from the CFD model were compared to those from the physical model, it was found that the latter had significantly underestimated the amount of energy lost. As a result, the airshaft expulsions seen during the commissioning tests could not be explained by the CFD model. Therefore, in order to accurately investigate and comprehend this phenomenon and discover strategies to reduce it, a deeper analysis of the rebound occurrence using a CFD model is required.

Marks et al. [50] asserted that the primary purpose of the Waller Creek Tunnel Project was to efficiently discharge significant quantities of accumulated air that is produced at the main inflow. Poorly ventilated systems can create explosive situations due to the accumulation of air, which also hinders the tunnel's ability to discharge. Air blowback is a process that can cause explosions when confined air is transformed into sizable air bubbles, resulting in slugs that accumulate substantial air pockets. The geometry of the pipe and/or an uneven hydrodynamic equilibrium prevent these air pockets from escaping unless an unexpected force is applied. As a consequence, an aerial detonation occurs, usually involving the release of pollutants into the atmosphere, which poses risks and causes harm to workers. The Waller Creek Tunnel Project utilized entrapped air management techniques, as outlined in US Department of Interior Engineering Monograph No. 41, to address the issue at hand. These techniques were employed to evaluate the dynamics of two-phase flow and specifically determine the water discharge rate in pipes at which entrapped air begins to flow in a manner that opposes the water flow, leading to blowback. The utilization of a physical hydraulic scale simulation proved advantageous in assessing the intake and designs of tunnel hydraulics, as well as qualitatively evaluating the potential for air entrainment. However, the ability of physical models to accurately measure and evaluate the flow characteristics of trapped air dynamics is limited due to the inability to measure air pockets at the same scale as the model. While EM41 (1980) may not be able to directly measure the magnitude of blowback, it may still be used to measure the airflow at the inlet upstream, which can indicate the formation of air pockets that are a precursor to blowback. In EM41 (1980), the concept of a design foundation is presented as a means to prevent blowback. Additionally, it offers design principles for the safe implementation of air venting systems. Webby [51] proposed a number of measures to effectively reduce blowback accidents. These processes include regular monitoring of the inlet screens and, in the blockage condition, conducting cleaning.

• The screen cleaning process can be done by shutting down the power plant and momentarily opening the dam gate to try and remove debris from the screens. This method is effective when there is restricted access to the inlet structure.
• The "self-burping" procedure to reduce plant load to about several minutes whenever the power plant has been operated at 100 MW or more for at least 12 hours should be strictly adhered to (and this measure is only available at night when energy consumption is minimal).

Although these procedures do not provide complete protection against further blowback accidents, they can help mitigate the risks of occurrence. The researcher also recommended introducing more comprehensive protection against blowback accidents, such as the construction of an air trap and vent column directly downstream of the curvature at the bottom of the vertical drop shaft. Improvements to the air trap and vent shaft for the power plant inlet can be supported by a physical hydraulic model.

Chan [52] carried out a study comprising several tests on a discharge system, including inflow in a drop shaft. The work is based on introducing an air pocket into the tunnel and measuring its movements in the drop shaft with a high-speed camera and pressure sensors. The heuristic study shows that when the drop shaft inflow exceeds a diameter-dependent threshold, compression stresses increase, causing the air pocket to implode and release air-water flow. Although this study provided an explanation of the air blowback mechanism, further investigation is necessary to identify the factors that either promote or decrease air blowback. Yang et al. [14] used a three-dimensional CFD model to simulate two-phase fluid flow through bottom outlets beneath a Swedish embankment dam. The outlet has a bulkhead gate, a gate shaft, a horizontal conduit, and an exit. This study used prototypes to measure air entrainment and addressed the blowback incidence in the shaft's bottom exit, which provided useful practical solutions. The bottom exit has not been widely used in recent decades, mostly due to safety concerns about the entrainment of air and flow variations. There's even the possibility of floods and scouring downstream roads. An alternating current (AC) motor and a wire rope hoist are used to operate the gate. The hoisting capacity limits how slowly the gate can be operated, often requiring several prolonged pauses.

In autumn 2018, the latest field testing was conducted. At $Z_1 \approx+346.00 \mathrm{~m}$, the reservoir's water level was 3.00 m below the FRWL. The tailwater level, $Z_2 \approx+330.00 \mathrm{~m}$, was submerged, along with the conduit and its outlet. The primary goal was to record and assess air blowouts at the conduit exit as well as flow pulsations in the shaft. The gate hoisting mechanism was tested, and the dynamic water pressures acting on the conduit were also observed.

Let Q (m³/s) and ℎ (m) represent the water discharge and gate-opening height, respectively. Initially, the bulkhead gate was opened and maintained at ℎ=0.30 m or approximately 10 m³/s. At the tunnel's exit, there were intermittent blowouts of a distinct mixture of water and air. The blowout events lasted 8–10 s, and the frequency was determined to be T=30–35 s. Its sequence during the blowouts is depicted in Figure 3. Although it was not tested during the tests, the blowout height was thought to reach as high as 4.0–5.0 meters. The event was most likely caused by air that was entrained in the gate shaft and entered the horizontal conduit—which lacked a de-aeration structure—into the space. This was in line with earlier findings for this gate opening.

Subsequently, the gate was opened to ℎ=0.40 m. Q≈14 m³/s was the expected discharge rate. It was noted that this opening caused greater, more pronounced blowout surges and more entrained air to be drawn into the conduit from the shaft. It appeared that the blowouts became nearly continuous and had significantly longer durations at ℎ=0.50 m and Q≈16 m³/s. The subsequent surges that followed became significantly larger, although they occurred less frequently. The gate was then raised to Q≈25 m³/s and h=0.80 m. The blowouts were less frequent below h≈0.30 m. There were no significant blowouts, and the air was discharged almost continuously with a few minor variations. At this point, the field testing was halted. Additional discharges were restricted due to safety concerns over the downstream channel's discharge capacity and gate operation. The study used CFD modeling to overcome the limitation in illustrating the level of air entrainment with geysers at larger gate openings because there were no records available regarding its operation at larger gate openings.

Additionally, the study aims to validate the accuracy of computational fluid dynamics (CFD) modeling in reproducing the hydraulic phenomena observed in real-world tests. It also seeks to identify significant flow scenarios based on gate openings, estimate the volume of trapped air that is challenging to measure using alternative methods, establish a basis for implementing countermeasures and structural adjustments, and serve as a reference for understanding air entrainment in similar outlet configurations. Field observations involve testing multiple consecutive gate openings. The outflow exhibits distinct geysers in the moving water, characterized by a gradually increasing height ranging from 0.30 to 0.50 meters. The air discharge at an opening of 0.80 m remains rather consistent. Simulation focuses on gate openings ranging from less than 1 meter to larger than half of the total opening. The findings indicate that the air within the shaft enters the conduit and, as a result of fragmentation and merging, creates pockets of air that move along the tunnel's ceiling in coordination with the flow, until being released into the tailwater. Apertures smaller than 1 meter can exhibit either periodic blowouts with a regular pattern or a continuous release of air to mimic the movement of air and water. The concordance between field observations and computational simulations of geyser properties, such as their frequency, is substantial. When the apertures are of considerable size, the gate is completely immersed in water, and the discharge occurs without the introduction of any air or the occurrence of blowouts. Further research is required to address the reflux phenomenon, which can be illustrated by the setup of the control of flow channel cross-section and velocity. This study was restricted to examining one parameter, which is the ideal opening height of the gate for minimizing flow.

Yang and Yang [53] examined the malfunctions in deep tunnel discharge systems caused by occurrences that induce the geyser phenomenon. This phenomenon is characterized by the significant impact loads generated when high-pressure entrained air masses are released, leading to potential damage to shafts and spillways. The study utilized an atmospheric scale model to validate the relationship between the occurrence of a geyser and its influencing factors, such as water depth, intake pressure, and intake size. The findings revealed that the disruptions of water and air pressure in a baffle-drop shaft, which lead to the geyser phenomena, are caused either by a high-pressure release of trapped air or by local disturbances in air pressure that result in the swift movement of the combination of air and water.

The definition of a spillway geyser event has been updated by the analysis of recorded occurrences, firsthand observations, and the identification of potential risks. Regression analysis was employed to find a correlation between the greatest height of a geyser and its assessment. The link between explosion strength, water depth, air pressure, and air volume during geyser phenomena was studied to propose a dominant formula for a baffle-drop shaft geyser. Comparing and validating these two models revealed that they can reliably estimate geyser height and occurrence. The data also showed that increasing the barrier's resistance and stiffness at the shaft base prevents baffle structural failure. Yang and Yang [53] found several things, but there are limitations. Specifically, the model test of the geyser was conducted under standard atmospheric pressure conditions, which restricts the ability to accurately replicate the air pressure in real-world scenarios. Furthermore, another limitation is the measurement and assessment of the fluid dynamics of water entrainment in this physical prototype. Therefore, it is crucial to create standard mathematical models and conduct experiments at low atmospheric pressure to study the characteristics of drop shaft geysers. Moreover, considering the impact of regular discharge operation on the hydrodynamic load of the barrier is also necessary. Although this study provides valuable insights into assessing geyser height, it fails to incorporate approaches aimed at minimizing or preventing geyser occurrences. Additionally, it does not establish a connection between evaluating geyser incidents and their causes.

This paper aims to comprehensively examine all relevant findings pertaining to the process underlying the production of air bubbles and air pockets in spillways and pipelines. This research comprehensively examined all studies pertaining to the adverse outcomes and incidents arising from the occurrence of air pockets in hydraulic systems. Furthermore, it presented the tactics and techniques employed to alleviate or eradicate air pockets and air bubbles. The paper's conclusions can be summarized as follows:

1. The literature review conducted concluded that most of the literature showed that the negative consequences resulting from air sinuses are blowback, geyser, head loss, cavitation, and slug flow.
2. Several studies introduce beneficial contributions to explain the mechanism of occurrence of blowback incidents in hydraulic structures due to air-water flow. However, there is a lack of predictive or analytical solutions to predict or evaluate the critical pressure required to cause damage as a result of blowback. Also, there is a need to address the parameters that positively or negatively affect blowback phenomena. Several parameters can be investigated, such as conduit cross-sections, vent holes, structure geometry, slope, in addition to the intensity of velocity and fluid flow.
3. Slug flow incidents occur due to slugs-air bubbles mixture that results in disability to upload the trapped air in the conduit, resulting in challenges in separating slugs and air bubbles and consequently, slug flow takes place in conduits and causes damage.
4. Through the literature review, it was found that there is a weakness in predicting slug flow in pipes due to incorrect data and recordings. It was also found that there is a limitation in studying the effect of viscosity on slug flow. The literature also revealed the lack of experimental studies dealing with the prediction of slug flow and the locations of its formation. However, recent academic studies using CFD techniques could address this issue, in which CFD using numerical studies could simulate complex structures, like morning glory spillways, and predict slug flow in these structures, in addition to addressing significant parameters such as velocity and fluid pressure.
5. Recent studies could introduce a reliable relationship to evaluate geyser height and its occurrence by correlating its influencing factors, such as water depth, intake pressure, and intake size. The relation between explosion strength, water depth, air pressure, and air volume during geyser phenomena was also studied, but there are restrictions due to the ability to accurately replicate the air pressure in real-world scenarios. Furthermore, another limitation is the measurement and assessment of the fluid dynamics of water entrainment in physical prototypes. Therefore, it is crucial to create standard mathematical models and conduct experiments at low atmospheric pressure to study the characteristics of drop shaft geysers. Furthermore, considering the impact of the regular discharge operation on the hydrodynamic load of the barrier. Although this study provides valuable insights into assessing geyser height, it fails to incorporate approaches aimed at minimizing or preventing geyser occurrences. Additionally, it does not establish a connection between evaluating geyser incidents and their causes.
6. Regarding problems related to head loss, studies failed to find formulas to assess head loss in these structures. Although the studies addressed several factors affecting reducing head loss, such as the length of the air pocket and the angle of the conduit with respect to the horizontal axis, the studies have limitations to predict head loss accurately or conclude formulas in this regard. So, more research should be conducted using experimental studies, practical applications, and numerical simulations to accurately address and determine extensive factors' effects on head loss. Pothof et al. [24] offered an actual practical application to evaluate head loss addressing several conditions in addition to several parameters, and the study showed that the head loss caused by an air pocket has a reciprocal relationship with the inflow number and Weber number of the pipe and a direct link with surface tension. The outcomes obtained can benefit further investigations and investigate more parameters that can influence head loss caused by air pockets. Also, the study neglected the influence of turbulence on the dynamic surface tension and thereby should be investigated further.
7. Several studies introduce remedial strategies to mitigate these consequences caused by air pockets. An air vent is strategically positioned downstream of the emergency gate in the pipeline to mitigate air blowback.
8. The study showed the limitation in the computational fluid dynamics (CFD) model to evaluate blowout in hydraulic structures.
9. Several processes have been suggested to mitigate blowout: comprise the build of an air trap and vent column directly downstream of the curvature at the lower of the vertical drop shaft. Air trap and vent shaft for the power plant inlet can be improved by supporting a physical hydraulic model. Although these strategies can be efficient to reduce air blowout, the literature suggested optimizing the height of the gate openings to reduce the blowout, and therefore there is a need to conduct other investigations to address the phenomenon of reflux, represented by the introduction of the channel inclination angle, flow control, speed, and channel cross-section.
10. Researchers showed the efficiency of the ribs on the flow characteristics in slug flow under atmospheric pressure in horizontal smooth and ribbed ducts, adopting several parameters: rib location, rib height, and rib shape.
11. Also, cavitation mitigation strategies have been discussed, comprising the use of aerator holes located on the hood, the size of the aeration diameters.
12. The Computational Fluid Dynamics (CFD) is a suitable method to investigate the head versus discharge and cavitation index for various inlet shapes of morning glory spillways in comparison to physical models.
13. The strategies of air management are classified as follows: air banning, and the application of air-emission tools, like air valves or air vents. The installation of automatic air release valves at critical points in a pipeline is a common, and frequently effective, method for removing air from a system. Sweeping, which entails achieving a critical water velocity to completely remove air bubbles, and generation and drag, which involves utilizing a hydraulic jump to displace the air and eliminate large air bubbles at the end.