Design of Longitudinal Air Bubbler System Inside Ship Lock

Abstract

This study investigated the expected behavior of installing a longitudinal air bubbler system in the St Lambert Lock, Quebec, Canada to bring “warm” water from the depths of the lock chamber and reduce ice buildup on the lock walls during the winter navigation. The bubbler manifold would be installed near the bottom of the lock chamber and extend along the entire wall of the lock chamber. Compressed air would be provided to the manifold and the air would be released through a series of holes (orifices) installed along the length of the manifold. A vertical water current would be induced in the lock chamber by the rising air bubbles. The continuous supply of water to the surface should reduce ice formation along the wall of the chamber. This use of a longitudinal air bubbler system to reduce ice buildup along lock walls during the winter season is a novel use of an air bubbler system in North America in a navigation lock.

1 Introduction

The St Lawrence Seaway Management Corporation (SLSMC) contemplates installing a pilot-level longitudinal air bubbler system in the St. Lambert Lock to bring “warm” water from the depths of the lock chamber and reduce ice buildup during the winter season. Figure 1 below shows the ice buildup on the lock walls which can become significant enough to impede the clear passage of vessels without intervention. Figures 2 and 3 show the current methods of ice removal from the lock walls with a customized ice scraper tug and a long arm excavator.

Fig. 1.

Ice buildup on the lock wall surface

Fig. 2.

Ice removal with a customized ice scraper tug

Fig. 3.

Ice removal with long arm excavator

The bubbler manifold would be installed near the bottom of the lock chamber and extend along the entire south wall of the chamber where ice buildup is most pronounced due to reduced solar exposure. Compressed air would be provided to the manifold and the air would be released through a series of holes (orifices) installed along the length of the manifold. A vertical water current would be induced in the lock chamber by the rising air bubbles. The continuous supply of water to the surface should reduce ice formation along the walls of the chamber.

This use of a longitudinal air bubbler system to reduce ice buildup along lock walls during the winter season is a novel use of an air bubbler system in North America at a navigation lock. In all other cases known to the authors, air bubbler systems have been installed at locks primarily to create horizontal flow velocities at the water surface to move or deflect floating surface ice. Those systems are relatively high flow bubblers that require large air flow rates (see for example Rand 1988; Haehnel 2016). The SLSMC became aware of an application of a longitudinal air bubbler system to create vertical currents in Finland at locks located along the Saimaa Canal.

The proposed longitudinal air bubbler system would not be required to create substantial horizontal flow velocities at the water surface. Rather the goal of the proposed system would be to provide a constant supply of relatively “warm” water to the surface in order to suppress or a least reduce ice formation on the lock wall surface. Here, “warm” refers to the temperature of water that is greater than 0 ℃ with no specification of an actual temperature.

Unfortunately, nothing is known about the vertical temperature structure of the water contained in the St. Lambert Lock during the winter, and little is known about the environmental conditions in the lock at the water surface during the winter season. Therefore it is not possible to determine the required supply of water that must be brought to the surface. This conceptual design will estimate the longitudinal air bubbler performance over a range of practical orifice diameters, orifice spacings, manifold sizes, and air compressor sizes. A best design will then be selected based on the estimated performance, practicality of installation, likelihood of success, and potential for future modification, if required.

2 Background

There are many publications describing the mechanics of air bubble plumes that are relevant to the use of air bubbler systems at navigation structures (See for example, Kobus (1968), Ashton (1974), Tuthill and Stockstill (2005), and Haehnel (2016)). Ashton (1974) specifically addresses the use of air bubblers systems to suppress surface ice. These authors address the use of “high flow” bubblers. The proposed longitudinal air bubbler system that is the focus of this study would be used to suppress ice formation using a “low flow” bubbler. The basis of this approach rests on the change of density of water with temperature. Fresh water is an unusual substance in that its maximum density occurs at 4 ℃ (approximately 39 °F). Because of this, during the winter season, the water in the lock is likely to be stratified with the coldest water at the surface. The air bubbler is used to induce an upward vertical current to move the “warm” water from the depths of the lock chamber to the surface. This continuous supply of warmer water should suppress the formation of ice.

As mentioned above, a longitudinal air bubbler system in Finland has been applied in locks located along the Saimaa Canal. (Private communication from Tero Sikiö, Head of Unit, Finnish Transport Infrastructure Agency, Inland Waterways) The manifold is located about 3 ft above the bottom of the lock and attached directly to the wall of the lock chamber. The manifold diameter is 20 mm (3/4″ in.), the orifice diameter is quite small, only 0.7 mm (0.02756 in.) and the orifice spacing is 2.0 m.

3 Air Bubbler System Design

The air bubbler system is required to provide sufficient vertical upward water flow to effectively suppress ice formation along the length of the lock wall. Given that the water temperature structure of the lock is not known it is not possible to calculate the ice suppression based on heat transfer calculations. Rather a range of possible designs was determined based on the estimated performance, practicality of installation, likelihood of success, and potential for future modification, if required. The of the longitudinal bubbler system for the St Lambert Lock was investigated using the computer program BUB300 (USACE 2006). The input data for BUB300 include: manifold length and diameter, supply line length and diameter, orifice diameter and spacing, nominal compressor pressure, and submergence depth.

3.1 Bubbler System Dimensions

The relevant system dimensions used in the analysis are described here.

Manifold Length:

The required length of the manifold is 768 ft to provide coverage along the entire length of the south wall of the lock.

Supply Line Length and Diameter:

The air supply line length was estimated to be 100 ft and the supply line diameter was assumed to match the manifold diameters, discussed below.

Manifold Submergence Depth:

The submergence depth varied between 25 ft and 50 ft depending on the operation of the lock and the upstream and downstream water levels (Fig. 4). This compares to the maximum possible depth of the lock, from the bottom of the lock to the top of the wall, which is 60 ft.

Sizing Manifold Diameter to Provide Uniform Flow Out of All Orifices.

The first step in the manifold design is to compare the orifice sizes and manifold size to ensure that the flow rates out of the orifices are as nearly uniform as possible along the length of the manifold. A commonly used rule of thumb is that the flow rates out of the orifices will be nearly uniform if the sum of the area of all the orifices is less than 25% of the cross-sectional area of the manifold. Flow rates through each orifice will be calculated directly with BUB300 and the actual uniformity of a proposed system geometry can be assessed for various input flow rates.

Fig. 4.

Estimated Submergence depth of manifold in feet for three winter seasons

A manifold diameter of 3″ was selected as the best choice to provide practicality of installation, likelihood of success, and potential for future modification. The number of orifices required is equal to the total length of the manifold, 768 ft., divided by the orifice spacing. Table 1 shows the required manifold diameter, calculated using the 25% area criteria discussed above, for various orifice spacings (4, 8, 12, 16 foot) and various orifice diameters (0.7 mm, 1.0 mm, and 1/16″, 1/8″, 3/16″, 1/4″, 5/16″, 3/8″, 7/16″, 1/2″). The orifice diameters of 0.7 mm and 1.0 mm are included to match the orifice diameters used in the Finnish locks. A 3″ diameter manifold will provide sufficient flow capacity to maintain uniform flow across the manifold only for the cases shaded in yellow in Table 1. The detailed analysis of the 3″ manifold was limited to these cases.

3.2 Bubbler System Performance

The manifold and supply line dimensions used in the analysis are shown in Table 2. The range of manifold depths, orifice spacings, and orifice diameters analyzed are listed in Table 3. The largest orifice analyzed was 3/16″ (4.76 mm) because it was found that uniform flow could not be maintained out of all orifices at a larger orifice diameter and 8′ orifice spacing. BUB300 estimated the required air discharge through each of the orifices based on the air pressure supplied by the compressor. The sum of the individual orifice flow rates represents the total airflow to operate the bubbler at the specified pressure. The characteristics of the vertical plume created by the air bubbles were analyzed using the description provided by Ashton (1974) using the calculated air flow rate from each orifice and the depth of submergence of the manifold.

The vertical plume carries a volume of water to the surface. The vertical water discharge carried by the plume increases with depth of submergence of the manifold due to flow entrainment and the plume expands as it rises. An important point is the length of the lock wall protected by the plume reaching the surface.

Protection requires that the plume expands enough to cover the spacing between the orifices. The ratio of the plume width at the surface to orifice spacing was calculated to provide insight into portion of the wall that would be reached by the plume. It must be noted that the equations of Ashton assume that the bubbler orifices are located far from any vertical wall. In the case of the St Lambert Lock application, the bubbler will be located immediately next to the lock wall. It is not thought that this will make a significant difference in the results.

Table 1. Required manifold diameters for 768 ft. length, with various orifice spacings (4, 8, 12, 16 foot) and various orifice diameters. A 3″ diameter manifold will provide sufficient flow capacity for cases shaded in yellow.

Table 2. Overall dimensions of Manifold and Supply Line

Several trial geometries were evaluated to determine the optimal combination of orifice spacing and orifice diameters for the 3-in. manifold. Three scenarios were evaluated:

• Scenario 1. This scenario investigated the impact of changing the orifice diameter. The orifice spacing was set at 8 ft and the submergence depth at 60 ft. The manifold diameter was 4 in. in this case which limited the effects of pressure drop in the manifold due to friction. Orifice diameters from the first five rows of Table 1 were evaluated.

• Scenario 2. This scenario investigated the impact of changing the orifice spacing. The orifice diameter was set at 1/16 in. (1.59 mm) and the submergence depth at 60 ft. The manifold diameter was 3 in. Spacing of orifices included 4, 8, 12 and 16 ft.

• Scenario 3. This scenario investigated the impact of changing manifold submergence depth. The orifice diameter was set at 1/16 in. (1.59 mm) and the orifice spacing at 8 ft. The manifold diameter was 3 in. Performance at depths of 20, 30, 40 and 50 ft were evaluated.

4 Results

Scenario 1: Varying orifice diameter. This scenario investigated the impact of changing the orifice diameter.

The air discharge per orifice is shown in Fig. 5. It should be noted that the air discharge across all orifices along the manifold was very uniform for all cases. The difference from completely uniform was only a few percentages for the largest orifice case.

The air discharge out of an orifice is proportional to the square of the orifice diameter. It can be seen in Fig. 5 that the air discharge per orifice increases rapidly for larger sized orifices. The total air supply required for all the 96 orifices in the manifold is shown in Fig. 6. Again, it can be seen that the total air supply required increases rapidly for larger sized orifices.

The ratio of surface spread of plume to orifice spacing is shown in Fig. 7. An orifice diameter of 1/16 in. or larger should provide good coverage of the lock wall for an orifice spacing of 8 ft (ratio is ~1 or larger). An orifice diameter of 1/16-in was selected to proceed with because it is relatively small which will help limit the airflow rate per orifice and therefore the vertical velocity. The size is a bit larger than the 0.7 mm holes used in the Finnish canals. The slightly larger size will also help to reduce the potential for biofouling.

Fig. 5.

Air discharge per orifice for a range of orifice diameters (Depth 60′, orifice spacing 8′, manifold diameter 4″))

Fig. 6.

Total air supply required for a range of orifice diameters (Depth 60′, orifice spacing 8′, manifold diameter 4″))

Fig. 7.

Ratio of surface spread of plume to orifice spacing for a range of orifice diameters (Depth 60′, orifice spacing 8′, and manifold diameter 4″))

Scenario 2: Varying orifice spacing.

This scenario investigated the impact of changing the orifice spacing. The orifice diameter was set at 1/16 in., the manifold diameter at 3 in. and the submergence depth at 60 ft.

The air discharge per orifice is shown in Fig. 8. It can be seen that the orifice spacing does not have a significant impact on the discharge per orifice over the range of compressor pressures. This is due to the relatively minor pressure drops through the 3 in. manifold.

The total air supply required is shown in Fig. 9. At a given pressure, the total air supply required is linearly proportional to the number of orifices. The ratio of surface spread of plume to orifice spacing is shown in Fig. 10. The 4 foot spacing provides extra coverage with a ratio greater than 2.0. The 12 foot and 16 foot spacing provide poor coverage with ratios of 0.6 or less. The 8 foot spacing has a ratio just over 0.8 for most pressures evaluated, meaning that at least 80% of the wall would be covered by the bubble plume at the surface.

Fig. 8.

Air discharge per orifice for a range of orifice spacings (Depth 50′, orifice diameter 1/16″, manifold diameter 3″))

Fig. 9.

Total air supply required for a range of orifice spacings (Depth 50′, orifice diameter 1/16″, manifold diameter 3″)

Fig. 10.

Ratio of surface spread of plume to orifice spacing for a range of orifice spacings (Depth 50′, orifice diameter 1/16″, manifold diameter 3″)

Scenario 3: Varying the manifold submergence depth this scenario investigated the impact of changing manifold submergence depth.

The air discharge per orifice is shown in Fig. 11. At lower compressor pressures the discharge per orifice increases as the submergence depth decreases. However, at higher pressures, the increase in orifice discharge increases less as the submergence decreases. When the ratio of the manifold pressure to the outside pressure is 0.528 or less, the orifice discharge does not change at all with decreasing submergence. At this point the velocity from the manifold is choked, that is, the orifice discharge velocity is no longer sensitive to the pressure outside the manifold and depends only on the pressure inside the manifold. Choked flow is a common occurrence with compressible flows.

The total air supply required is shown in Fig. 12. Similar to the previous figure, the impact of choked flow is evident on the total air supply required.

The ratio of surface spread of plume to orifice spacing is shown in Fig. 13. The spread of the plume is linearly proportional to the depth of submergence. Therefore less submergence means that the plume has less opportunity to spread. It can be seen that manifolds with less submergence do not protect the entire lock wall.

Fig. 11.

Air discharge per orifice for a range of submergence depths (Orifice spacing 8′, orifice diameter 1/16″, manifold diameter 3″)

Fig. 12.

Total air supply required for a range of submergence depths (Orifice spacing 8′, orifice diameter 1/16″, manifold diameter 3″)

Fig. 13.

Ratio of surface spread of plume to orifice spacing for a range of submergence depths (Orifice spacing 8′, orifice diameter 1/16″, manifold diameter 3″)

5 Discussion

The performance of a conceptual longitudinal air bubbler system for the St Lambert Lock was estimated based on BUB300 (USACE 2006) and the approach of Ashton (1978). The air flow through each orifice, the total air supply required for all the orifice, and the fraction of the lock wall protected by the bubbler plume was estimated for a range of orifice diameters, orifice spacing, and manifold submergence depths.

The air flow out of each orifice is strongly influenced by the orifice diameter. In general, for a given pressure difference across the orifice, the air flow out of the orifice is proportional to the square of the velocity. It is not thought likely that a large air flow is required to suppress ice formation along the lock wall. Therefore a relatively small orifice will likely provide sufficient air flow for this task. The use of a relatively small orifice is supported by orifice sizes used for longitudinal bubblers at locks located in Finland along the Saimaa Canal. (Private communication from Tero Sikiö), where orifice diameters of 0.07 mm are used. This size is much smaller than any orifices used at navigation locks in the US (however, bubblers at locks in the US are generally used to create large horizontal currents at the water surface to move surface ice.) This study focused on an orifice diameter of 1.59 mm (1/16″). This is larger than the Finish orifices but smaller than any currently in use at navigation locks. There are potential practical problems with very small orifice sizes – especially blockage of the orifice caused by marine debris, rust, and other small objects.

A 3 in. manifold provides very uniform flow through all the orifices – within about 3% variation for an 8 foot orifice spacing at 40 PSI pressure. A 3 in. manifold can also provide uniform distribution even if the orifice size is doubled to 1/8″. This provides flexibility if it is found that a larger orifice is required from field testing.

A range of orifice spacings were investigated. It was found that a spacing of 8 foot or less is best for providing protection along the entire length of the lock wall. Spacings greater than 8 foot do not protect the entire wall length because the vertical plumes do not spread out enough to provide flow to the entire distance between the orifices. This is especially true at small submergences (less than 40 ft of water depth). As discussed above, the equations used to estimate the bubbler spreading assume that the bubbler orifices are located far from any vertical wall. In the case of the St Lambert Lock application, the bubbler will be located immediately next to the lock wall. It is not clear if this would make a significant difference in the result. An orifice spacing of 8 ft is reasonable first try. If it is found that a closer spacing is required to provide complete surface coverage, the 3 in. manifold could provide uniform distribution to orifices with 1/8 in. diameter and 4 foot spacing.

Compressor pressures ranging from 30 psi to 70 psi were investigated. In general, the results at lower pressures were quite satisfactory and there was not any significant increase in air flow rate per orifice or in providing protection along the entire length of the lock wall. Note that pressure losses due to couplings, bends, flow valves, and other fixtures in the air supply line and manifold were not included in the calculations. There should be an allowance for extra capacity in the compressor to account for these losses.

6 Field Testing

The SLSMC constructed the pilot low flow air system at the St. Lambert Lock in 2020 using the findings of this analysis. The final dimensions of the system include a 3-in diameter manifold with 1/16-in diameter orifices spaced at 7-ft. The manifold was fastened to the lock wall in a position where it is less prone to effects from sedimentation, impacts from vessels and does not obstruct filling/emptying ports. An example of the manifold fastening method is shown in Fig. 14. System tests performed showed that the system provided continuous wall coverage and moderate flow up the wall but without a visible fountain, or excess disturbance of the surface which could cause splashing and incidental ice formation (Fig. 15).

The system was shown to be effective, and also found to have an incidental benefit from the gentle horizontal current which pushed any floating ice towards the center of the lock where it was easier to flush with vessel traffic. The 1/16 in. orifice diameter provided sufficient airflow such that the need to potentially increase the orifice diameter was determined to be unnecessary. This allowed a modification of the design to use a smaller diameter manifold of 2-in. while still providing uniform flow along the system length.

This design, with 2-in. diameter manifold, was used for the installation two additional bubblers (one on each wall) in the Upper Beauharnois Lock 4 (Fig. 16).

Fig. 14.

Example of manifold attachment to lock wall at Upper Beauharnois

Fig. 15.

Pilot longitudinal bubbler system in operation at St. Lambert Lock. Note bubble plumes visible along the lock wall.

Fig. 16.

Installation of one of a pair of longitudinal bubblers at the Upper Beauharnois Lock 4 in Quebec, CA. Note the installation location above the filling ports along the bottom of the lock.

7 Conclusions

This study investigated the expected behavior of installing a longitudinal air bubbler system in the St Lambert Lock, Quebec, Canada to bring “warm” water from the depths of the lock chamber and reduce ice buildup on the lock walls during the winter navigation. The specialized software BUB300 written by engineers at the US Army Corps of Engineers Cold Regions Research and Engineering Laboratory was used to analyze potential system designs and find an optimal geometry for the subject location. Based on this analysis the SLSMC installed a pilot system was found to perform very well. A minor refinement to the design, a reduction of the manifold diameter from 3 to 2 in. provided some cost savings for future installs while still maintaining the performance capability of the pilot system. These systems are an important tool which provide timely and effective mitigation of ice buildup on lock walls, reducing the need for labor-intensive and potentially hazardous ice removal techniques.