Hydraulics and swimming behavior of schizothorax prenanti in vertical slot fishways ＊

Rui-dong An, Jia Li, Wen-min Yi, Xi Mao,

1. State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu 610065,China

2. College of Water Conservancy and Hydropower Engineering, Sichuan Agricultural University, Yaan 625014,China

Abstract: This paper presents a novel experiment for the correlation between the hydraulics and the swimming behavior of Schizothorax prenanti, a rare species of fish in southwest China, in passing a vertical slot fishway. With an acoustic Doppler velocimeter, the velocities of a physical model in the equidimension fishway in the Shaping Ⅱ power station are measured. The hydraulic parameters include the hydraulic head drop, the velocity patterns and the flow rate, and the swimming behavior includes the burst-coast, the sustained swimming and the migratory path, and they are analyzed under the test conditions. The sustained swimming velocity is in the range from 0.65m/s to 1.09m/s. The estimated hydraulic head drop of each pool is calculated and is in the range from 0.08-0.11 m, which means that 2.6% slope is reasonable. For the same slope, the maximum velocity is further reduced from 1.24 m/s to 1.14 m/s by using an L shape baffle structure. The above findings are used as the basis to evaluate the hydraulic performance of the fishways, where the L shape baffle structure is expected to be effective for creating a preferencial flow for the fish. For the main stream of the pool, an “Ω” shape flow pattern is more fish-friendly, which can effectively extend the energy dissipation distance and avoid the bursting through a high velocity zone. This paper provides a useful complementary tool for practical designs.

Key words: Hydraulics, swimming behavior, vertical slot fishways, burst-coast, energy dissipation

Introduction

The aquatic environment concerns us recently with the construction of dams in China[1].The installation of dams and weirs has a potential effect on the populations of rare species including Schizothorax prenanti in southwest China by blocking the route of their migrations[2].The vertical slot fishways are one of the main types that are generally used in the restoration projects in large gradient rivers[3-4]. The fishway consists of a number of pools divided by baffles in a rectangular channel. The energy in the fishway is dissipated mainly through the local head loss near the slot and the frictional head loss in the channel.

In the previous fishway design, the focus is usually on the connection between the maximum velocity and the fish burst swimming velocity, the sustainable swimming ability, the sustainable swimming time and the critical swimming ability[5].However, any such single index cannot adequately reflect the overall swimming behavior of migrations[6-7]. In many cases, the structure of the flow in multi-pools of the fishway must be designed to dissipate the energy in the water in such a manner as to enable the fish to ascend and to sustain the swim without any undue stress[8-9].Through studying sixteen fishway designs, Bermudez et al.[10]found the length of the pool is the main geometric dimension that affects the flow in the fishway. The characteristics of the mean flow and the turbulence in the vertical slot fishways were studied by Tarrade et al.[11], and the results show that the flow pattern always follows one of two topological models depending on the ratio of length/width of the pool.

One effective approach adopted in this study is to track the fish swimming path and establish hydraulic mechanisms by detailed velocity velocity measurements on specially designed physical models. Such mechanisms can also be used to optimize the engineering design to enable the fish to migrate through the passes efficiently. Lennox et al.[12]tracked the behavior of the adult Atlantic salmon in the field under natural conditions through waterfalls. It is shown that 29 of the 39 fish have successfully approached and ascended the pool-and-orifice ladder at Pitlochry Dam. Silva et al.[13]assessed the influence of the turbulent flow on the swimming behaviour of the Iberian barbell, and found the highest rates of the passage and the corresponding shortest times in the experiments conducted with the offset orifices.Correlations ware found to be the highest between the horizontal component of the Reynolds shear stress and the fish transit time, especially for smaller smaller size-individuals[14]. In some experiments, the hydraulic and swimming aspects of fish passes were tested for the Atlantic salmon, showing that a significantly higher proportion of the Atlantic salmon move through the vertical slots than the overflow weirs[15-17].

This paper presents the results of the experimental investigation of a vertical slot fishway of different slopes and baffle structures. The hydraulics in the pool and the slot, the path of the jet and the swimming behavior of are analyzed for more effective and simpler designs of the sustained swimming.

1. Swimming ability and sustained swimming

1.1 Schizothorax prenanti

The successful running of the fishway is mainly determined by two aspects: either the fish can find the entrance or a sustained swimming can be maintained.Both are closely related to fish’s swimming behavior,which are rheotaxis and countercurrent. The rheotaxis is defined by the response flow velocity that reflects the tropism of the current and the direction sensitive of the fish. The contercurrent ability is defined by the burst swimming velocity and the critical swimming velocity for overcoming the velocity in fishways.

Juvenile one-year-old Schizothorax Prenanti is the research objective in this work (with the test sample shown in Fig. 1), which is a unique rare kind of fish in Southwest China and a benthic and cold-water fish living in the junction of the torrent and subcritical flow. In March and April, the reproductive population migrates in short distance (less than 100 km) to spawn in Min River, Dadu River[18]. The exact migration distance is still unclear, which is believed to be less than 100 km. Because of its small size and weak swimming ability, Schizothorax Prenanti is often selected as the indicator of the aquatic ecosystem protection in the hydropower developing river.

Fig. 1 (Color online) Schizothorax prenanti testing sample

1.2 Swimming ability test

The requirement of the sustained swimming for the fish may significantly impact the design of the fishway with respect to its total length and structure,which means not only requiring the maximum velocity in the vertical slot to be less than the burst swimming velocity but also paying more attention on the sustained swimming flow pattern. Because the fish cannot maintain the burst velocity in a whole migration process, the “burst-coast” pattern in the swimming behavior is used by the principle of least effort. The critical swimming velocity vcis defined for describing the sustained migration known as the cruising velocity. Normally, the cruising velocity is given by a range close to vc. The response flow velocity vris defined as the minimum velocity that makes the fish to reflect rheotaxis and successfully find the direction in the migratory routes. Therefore,the range for the fish friendly sustained swimming should be between the response flow velocity and the burst swimming velocitybv.

A swim flume (Loligo Systems SW10200,Denmark) is used to measure the fish’s swimming velocity. The swim flume has a total water volume of 90 L and the dimension of the swimming chamber is 0.7 m×0.2 m×0.2 m (length×width×height). A honeycomb screen is fixed at the upstream of the swimming chamber to reduce the turbulence and ensure the uniform water velocity across the swimming chamber.The water flow is driven by a single propeller powered by an electric motor with a variable frequency drive. The water temperature in the swimming chamber is measured by a mercury thermometer and ranges from 23°C to 25°C, and DO is measured by a dissolved oxygen instrument and ranges from 7.0 mg/L to 8.1 mg/L.

The response flow velocity, the critical swimming velocity and the burst swimming velocity are determined by using the velocity increasing method[19-20].A single fish is put into the flume at a low velocity range (10-3m/s) for 1 h to eliminate the stress during the fish transfer process. The test is conducted from 60% of the estimated vcand the flow velocity isgradually increased in a step of 15% estimated velocity and in an interval of 15 min. The experiment is finished when the fish is too fatigue to continue.

Table 1 Swimming ability test samples, test conditions of prenant’s Schizothoracin

The vccan be expressed as follows

where vpis the highest velocity maintained in the whole interval,iv is the increased flow velocity, tfis the time between the latest velocity and the fatigue point,it is the time step.cv can be expressed as the velocity for sustained swimming over half an hour and the burst-coast behavior for the fish. The results of the swimming ability test are given in Table 1.

The velocities are measured by using a propeller flow meter, andrv andbv are measured with the same method: the velocity increasing method. According to the results,bv is around 0.85-1.53 m/s with the mean velocity of 1.22 m/s.rv is around 0.07-0.13 m/s with the mean velocity of 0.10 m/s.cv is around 0.65-1.09 m/s with the mean velocity of 0.85 m/s.

2. Experimental arrangement and hydraulics in vertical slot fishway

2.1 Experimental arrangement

The experiment is carried out at the SKLH (State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, China), where detailed hydraulic conditions can be implemented.The same scale model of the fishway consists of a reinforced concrete structure over 30 m long with a single 2.4 m by 2.0 m or 2.7 m by 2.0 m rectangular pool according the body length of the fish (as shown in Fig. 2). The physical model contains twelve separate pools and a rectangle corner (the domesticated pool). The model has the same size with the prototype,a vertical slot fishway in the second cascade Shaping hydropower station (Shaping Ⅱ) in Dadu River (as shown in Fig. 3). The velocities are measured by using a Nortek Vectrino, a three dimensional acoustic Doppler velocimeter (3D-ADV) in the bottom, the middle and the surface ( z1= 0.1m , z2= 0.6 m,z3= 1.1m), with 1-25 Hz sampling frequency. The water surface elevation is measured by using a laser levelling instrument.

Fig. 2 (Color online) 1:1 physical model of vertical slot fishway

Fig. 3 (Color online) Shaping II dam construction site. Photo taken in 2013 by author, flow direction from left to right,dotted line-plan of the fishway, solid line-dam site

According to the sample test results of the body size in the prototype river section, the width of the vertical slot is 0.3 m. The slopes are set to s1=4.2%,s2=2.6%according to the allowable range in the engineering design, to evaluate the influence of the flow patterns, limited by the practically possible design of Shaping II. For the lower limit ( s2= 2.6%),two different structures (single baffle and L shape, 45°slot angle) of a long baffle are used for reducing the velocity near the slot (as shown in Fig. 4). Table 2 provides a summary of the experiment, and three variables for the arrangement of the experimental study are used with the same water depth under normal operating conditions for the minimum H1=0.5 m and the maximum H1= 1.2 m. Since the model can be used to simulate a part of the fishway project, the water depths of the import and the rest pools with a flat slot are controlled to the same for forming a realistic hydrodynamic condition.

Fig. 4 Plane view of pool structure adopted in experiment

Table 2 Summary of hydraulic test

2.2 Experimental results: Velocity and efficiency

Measurements are made of the flow rate, the hydraulic head loss and the three dimensional mean velocities. According to the conservation of energy,the hydraulic head drop Δh is the elevation difference between the upstream and the downstream of the pool can be expressed by the following equation

wherebv is the designed velocity in the pool, g is the the gravity acceleration and ϕ is the flow rate coefficient which takes a value in the range of 0.85-1.00 in the experiment. Assuming that vb= 1.22 m/s ,the estimated Δh can be calculated to be in the range of 0.08-0.11 m. Δh is believed to be a key factor influencing the maximum velocity in each pool, which is also related to the burst swimming behavior.

The water surface surges in the front of the baffle on both sides and then the water level drops rapidly at the vertical slot. The whole process is repeated for each pool from the upstream to the downstream, with a difference between the left and right banks. The first three pools witness the most dramatic fluctuations of the water level, with the maximum hydraulic head drop in the third pool with the maximum flow velocity.The efficiencies of the different designs and the measurements of Δh are presented in Table 3, which shows that T3 enjoys a better hydraulic condition as compared to T1, T2. While the experimental results of Δh are higher than 0.08 - 0.11 m, which also shows that the L shape baffle plays an active role.

Figure 5 shows the velocity contours of the hydraulic test at Z = 1.2 m, which is the water depth preferred by the fish. The jet from the vertical slot plunges into the pool by 45°, to form a clockwise circulation on the right side of the jet flow. The “Ω”shape flow pattern in the pool with a clear interface for the mainstream and the circulation is found in T3.The maximum velocity occurs at the end of the water level drop zone, around 0.1 m downstream the slot.For the design T1, the maximum velocity is almost 1.55 m/s ( x =0.6 m, y =0.7 m, at the surface,downstream the slot) and the average velocity of themainstream is 1.2 m/s, also with an anti-clockwise circulation at the corner. For the design T2, the slope is 2.6%, which effectively reduces the jet flow and the mainstream velocity, with the maximum velocity of around 1.24 m/s ( x =0.6 m, y =0.8 m, Z =1.1 m,downstream the slot). Because the body size is larger than the area of the unfavorable flow pattern, it seems that it can be accepted by the fish. Despite the fact that the burst swimmingbv exceeds the standard for T2,the area of higher velocity and turbulence is acceptable by Schizothorax Prenanti. For T3, the maximum velocity is further reduced to 1.14 m/s for the L shape baffle structure.

Table 3 Experimental results of hydraulic head loss. Total testing length is 14.4 m, 6 pools

Fig. 5 (Color online) Velocity contours on surface of the pool obtained by measurements. z = 1.2 m, the maximum velocity is defined as 1.2 m/s for comparison

The unit water power dissipation E can be expressed as in the following equation, which also provides a criterion for the fish-friendly structure.

where ρ is the water density, Q is the flow discharge and V is the volume of each pool.

Table 4 shows the comparison result of the design for the maximum velocity and its position, the unit water power dissipation and the hydraulic characteristic analysis. Because of the excessive maximum velocity and the turbulence, with the approximate size of the recirculation and the velocity distribution, the sustained swimming condition is also not satisfied for T1.

It must be noted that although the lower slope is more conductive for reducing the mainstream velocity,the difference between T1, T2 with the same baffle design is not clear, as shown in Figs. 5(a), 5(b). No obvious evidence is seen in the swimming behavior difference either. For that reason, a comparison of the swimming behavior between T1, T3 is made. The design of the L shape baffle provides a better energy dissipation efficiency.

The total area of the unit pool ( ZT) of T1=T2,T3 are 1.010 m2, 1.017 m2. The area, where the velocity ＜0.1m/s, equal to vR, and the backflow and the eddy will often occur, can be defined as the disoriented zone, ZD. The area, where the velocity is from 0.10 m/s to 0.85 m/s, can be defined as the preferencial flow zone, ZP, which is suitable for the sustainable swimming by the burst-coast. The convergence area of the high and low flows, where the velocity is from 0.85 m/s to 1.22 m/s, can be defined as the burst flow zone, ZB. The area, where the velocity ＞ 1.2 2 m/s , which means that the fish must

3. Swimming behavior of sustained swimming

3.1 Swimming behaviour

Although a lower slope is more conductive for the fish migration as proved by experiment, the total length of the fishway could not extend unlimited with a definite water level for both upstream and downstream of the dam. An extended fishway length also poses difficulty for the sustained swimming, and the best solution is to optimize the structure of each pool for reducing the mainstream velocity with a constant slope. The swimming behavior is the decisive factor for judging the hydraulic characteristics for the sustained swimming. Therefore, the swimming behavior experiment combined with the hydraulic test is carried out for such purpose.frequently wag its tail or stop migrating, can be defined as the high flow zone, ZH. A dimensionless number n can be defined as the proportion between the fish-friendly area and the total area for each pool.A summary of swimming behavior measurements isshown in Table 5. Parallel tests are carried out for T1,T3, and a valid test is defined as one when 8 pools can successfully be passed. The chosen path is recorded by a Video Camera on the top of the model.

Table 4 Characteristic analysis of hydraulics

Table 5 Comparison of hydraulics and corresponding swimming behavior of sustained swimmingZT3=1.017 m2

Table 5 Comparison of hydraulics and corresponding swimming behavior of sustained swimmingZT3=1.017 m2

Design Z D/m2 Z P/m2 Z B/m2 Z H/m2 Z B / Z T Z P / Z T Path/m S bw ei hm a vm i i on rg Disoriented Preferences Burst Failure wF a r ge q gu i ne gn t tl ay i l fiS shu-s ft ra ii en ne dd ly Burst-coast Hydraulic v ＜ vr v r ＜ v ＜vc v c ＜ v ＜v b v ＞ vb Main stream n 0.3-0.7 m/s T1 0.002 0.575 0361 0.072 0.36 0.57 5.2 T2 0.002 0.641 0.284 0.084 0.28 0.63 None T3 0.024 0.667 0.326 0 0.32 0.66 4.4

Fig. 6 (Color online) Swimming behavior experiments

3.2 Path chosen by Schizothorax Prenanti

To approach the passes involves a movement of two basic paths in T2, T3 as shown in Figs. 6(a), 6(b),although other paths are also possible. According to the swimming behavior of Schizothorax Prenanti, the burst-coast mode is usually adopted for sustainable swimming in a long distance fishway[15]. The fish moves along the jet flow almost to the side of the pool by frequently wagging tail from the maximum velocity area, then swims along the convergence path of the mainstream and makes a turn rapidly at the end of the baffle. For the Path B in T3, the fish moves directly along the midline line and crosses the mainstream then turns to the outlet by 45°.

The whole process for the Path B is shorter than that for the Path A (as shown in Fig. 7), and the fish swims quite smoothly without rapidly turning and frequently wagging its tail. The velocity for each path is similar in the range of 0.7-0.8 m/s, but is decreased to 0.3 m/s by the end of the flume, which means that ZPcan be confirmed as the fish-friendly sustained swimming condition. By the principle of least effort for the fish swimming, the L shape baffle with a gradient of 2.6% is a more reasonable structure than the normal type for the sustainable fish-friendly swimming.

Fig. 7 Velocity contours and path tracks

4. Conclusions

This paper presents the results of an experimental study of the hydraulics of three designs of vertical slot fishways with different slopes and baffle structures in a model of the same size. The fishway model consists of a reinforced concrete structure of over 30 m long with a single 2.4 m by 2.0 m or 2.7 m by 2.0 m rectangular pool. Meanwhile, the swimming behavior for the sustained swimming of Schizothorax prenantiwas is also analyzed to obtain some quantitative information about the correlation between the velocity pattern, the energy dissipation efficiency and the migratory path in the pool.

The sustained swimming velocity can be determined in the range from 0.65 m/s to 1.09 m/s. The estimated hydraulic head drop of each pool can be obtained as 0.08-0.11 m, which means that 2.6% slope is more reasonable. For the same slope, the maximum velocity is further reduced from 1.24 m/s to 1.14 m/s by using the L shape baffle structure. The above findings can be used as the basis to evaluate the hydraulic performance of the fishways, and the L shape baffle structure is expected to be effective in creating a preferencial flow for the fish. For the main stream of the pool, the “Ω” shape flow pattern behaves more fish-friendly, which can effectively extend the energy dissipation distance and avoid the bursting through the high velocity zone.

Apart from this, the recirculation and the eddies may cause the disorientation for the fish, and the Reynolds shear stress might also be an important factor, which is not considered in this paper[10-11].Compared with the previous study, an evaluation method is provided by combining the hydraulics and the swimming behaviour to have a more fish-friendly design. To approach the passes, the movement can take place in two basic paths. By adopting the burst-coast mode for the sustainable swimming in a long distance, the L shape baffle with a gradient of 2.6%is a more reasonable structure for the sustainable fish-friendly swimming.

The great advantage of the vertical slot fishway is that the fish can swim through the slot at any desired depth. This is particularly important for other species and the frequent scheduling of the water level.Although the swimming and hydraulic aspects are essential for a good design of a specific fishway, all kinds of damages might be caused by an obstacle,which can never be fully compensated.