MAO Xi

State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu 610065, China, E-mail: maowhiteknight@163.com

FU Jing-jing

HydroChina Hua Dong Engineering Corp.9, Hangzhou 310014, China

TUO You-cai, AN Rui-dong, LI Jia

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

(Received December 15, 2011, Revised July 5, 2012)

INFLUENCE OF STRUCTURE ON HYDRAULIC CHARACTERISTICS OF T SHAPE FISHWAY*

MAO Xi

State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu 610065, China, E-mail: maowhiteknight@163.com

FU Jing-jing

HydroChina Hua Dong Engineering Corp.9, Hangzhou 310014, China

TUO You-cai, AN Rui-dong, LI Jia

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

(Received December 15, 2011, Revised July 5, 2012)

While dams cut off the migratory route of fish, and affect their growth and breeding, fishways are built to restore the connectivity of rivers and facilitate the life cycle of these migratory fish. This paper proposes a new kind of fishway called the “T shape fishway”, and studies it by using both numerical simulations and physical experimental methods with the aim to design a fishway in China for migratory fish with swimming and jumping capability much poorer, such as schizothorax davidi and schizothorax prenanti, as compared with migratory fish in other places of the world, like trout and salmon. It is indicated that the range of velocity in each regular pool of the T shape fishway is from 0.42 m/s to 1.22 m/s when the flow discharge Q is 350 L/s, and there are no large vortex zones found there. This study shows that the T shape fishway can meet the requirements of these migratory fish quite well in terms of velocity and flow patterns in China.

fishway, migratory fish, baffle, drop sill, numerical simulation, experiment

Introduction

Fishways, also known as fish-ladders or fishpasses, are hydraulic structures which allow the migration of fish through obstructions in rivers, and are generally studied under three types: the Denil type[1], the pool and weir type[2], and the vertical slot type[3].

The denil fishway, first reported in 1909 and named after the inventor, is rectangular chutes with closely spaced baffles or vanes on the bottom and sides that run along the edge of the chute. In the Denil fishway, the fish must pass through without stopping because there is no resting place for them. Besides, the velocity in the Denil fishway is too high for fish ofmany species to ascend. In other words, the Denil type is only suitable for fish with very good swimming capability[4].

The pool and weir fishway consists of consecutive weirs and pools, includes a series of weirs with or without orifices and notches, and is suitable for fish with good swimming capability. Fish can take a rest in the pool when it passes over a weir at its burst speed, then the next weir and pool, and so on, till it completes the ascent, and the pool and weir fishway can not provide the whole section migration[5].

In the vertical slot fishway, the water flows from one pool to the next through a vertical slot in the baffle, forming a water jet, which causes a central turbulence and thus the energy dissipation, and at the same time forms areas of much lower flow velocity on either side. In each pool of the vertical slot fishway, there are always two large vortex zones at the sides of the slot. The fish may move from one pool to the next without jumping at any depth within the slot,while the presence of low velocity lateral areas allows it to take a rest. However, when the fish try to ascend from these slots, it must use its burst speed because the velocity of the flow is much higher than that in other places, and even, much higher than the fish could manage[6].

In both the pool and weir fishway and the vertical slot fishway, the turbulence level in the pools is extremely high, which plays an important role in the development of the flow. In addition, the turbulence is important for evaluating the biological efficiency of a fishway design, because it would make the fish fatigue, and high turbulence levels would confuse the fish in finding their way up through the pools, as pointed out by previous studies[7]. An effective fishway must attract fish easily and enable them to find out, to enter, to pass through, and to exit safely with a minimum cost in both energy and time. If the velocity and the turbulence energy in the pools are too high, the fish would not be able to swim through the fishway[8]. Therefore, the efficiency of a fishway design is determined by hydrodynamic variables such as the velocity and the turbulence fields in each pool. For these reasons, it is important to decrease appropriately both the velocity and the turbulence fields through the use of an adequate model.

Because the swimming capabilities of the fish species can be divided into three categories as the sustained speed, the prolonged speed and the burst speed based on the speed and the muscle use[9], the physical and hydraulic characteristics of a fishway design may be suitable for some fish species but not for others. In this respect, the pool and weir fishway is suitable for fish with good jumping capability, and the vertical slot fishway is suitable for fish with good sprint capability. Besides, the flow velocity in the fishway must match the swimming capabilities of the fish species for which the system is intended.

According to the statistics, the swimming and jumping capability of the commonly migratory fish in China, such as schizothorax davidi, schizothorax prenanti[10], are much poorer than the migratory fish in other places of the world, like trout, salmon[9,11]. Incomplete statistics show that over 40 fishways in China were built based on the successful experiences from other countries. These fishways perform not very well[12]. The reason is possible due to the difference of the swimming and the jumping capabilities between the fish in China and that in other places of the world. When these fishways were designed, this difference had not been taken into a full consideration.

In our previous study[13], a new fishway, with three-levels of drop sill and a groove in each pool was put forward. According to the experimental studies, it was concluded that the drop sill was effective for energy dissipation, so the drop sill is used again in this study.

In this study, another new fishway named the “T shape fishway” is proposed based both on numerical simulations and experiments. The numerical simulation is a much easier and cheaper way to test and design a better fishway structure, as compared with experiment. Hence, in this study, numerical simulations are used to design and optimize the new fishway, and the experimental results are used to validate the numerical model. Once the numerical model is validated by the experimental data, it is used to analyze the flow pattern in more detail in the fishway pools, as well as to evaluate the hydraulic characteristics of the fishway design. The agreement between the numerical simulations and experiments confirms that the T shape fishway can meet the migratory requirements of the fish with poor swimming and jumping capability in China, in terms of the velocity and the flow pattern.

Fig.1 The T shape fishway

1. Numerical simulations

The T shape fishway has a T shape baffle in each pool. The prototype consists of a rectangular channel with sloping floor of 4 m in width. It is divided into six regular pools, with each pool of 5 m in length. Two transition pools are designed, one at the beginning and the other at the end of the channel, with flat floor of the same width of 4 m, as that of each regular pool, and a length of 10 m, twice as that of each regular pool. The flow discharge in the prototype is assumed to be Q=350L/s , and the slope S=2.6%. The linear ratio between the prototype and the model is 2:1. For the numerical and physical models as shown in Fig.1, the width of each regular pool (inclu-ding two transition pools) is =2mB. The length of each regular pool is =2.5mL, and that of each transition pool is =5mL＇. For this fishway, these lateral baffles are vertical to the X axis and the central baffle is parallel to the X axis, forming the inlet and the outlet (from the second regular pool to the sixth, the frontal pool?s outlet is the posterior pool?s inlet). The width of the inlet and the outlet is =0.4mb. The manger of the T shape baffle is Lm=0.4m in length and Bm=0.1m in width, and the central baffle is Lc=0.8m in length and Bc=0.2m in width. For all mentioned baffles, the height is H=1.2m. There is also a symmetric drop sill in each regular pool, Ld=0.6m in Y axis direction, Bd=0.1m in X axis direction, and Hd=0.4m in Z axis direction.

The Computational Fluid Dynamics (CFD) software is used in the numerical simulations.

1.1 Governing equations

These governing equations include the continuity equation, the momentum equation, and the kεequation[14]. In this study, the water density ρ is assumed to be a constant, ρ=103kg/m3.

The continuity equation can be written as

The momentum equation can be written as

The kε- equation can be expressed as

where

u, v, w represent the velocity components in X, Y and Z directions, p is the time averaged pressure, μ andtμ are the molecular viscosity coefficient and the turbulence eddy viscous coefficient of the water, g is the acceleration of gravity in Z direction, k is the turbulence kinetic energy, and ε is the dissipation rate of the turbulence kinetic energy, kσ,εσ are the Prandtl numbers corresponding to k, ε, respectively. C1ε, C2εand Cμare empirical constants: σk=1.0, σε=1.3, C1ε=1.44, C2ε= 1.92, Cμ=0.09[14,15]. Gkis the generation term of the turbulence kinetic energy k, and it is caused by the mean velocity gradient.

1.2 The Volume Of Fluid (VOF) model

The major problem is solved with the fixed grid of finite elements, with consideration of the overfall geometry as well as the number of pools and the discharge. The multiphase model in terms of the CFD is introduced for the solution of the free surface flow. The VOF is a commonly used model to simulate the motion interface.

The basic procedure of the VOF is to construct the shape of the phase interface by solving the independent momentum equation and considering the flow volume fraction of each element mesh. The Ffunction is used to construct and trace the free surface. In the VOF model, the interface between two phases is traced by solving the continuous equation of the volume fraction in one phase or more, and the equation can be written as

In Eq.(7), F is the volume function for the water phase and the air phase. In each element mesh, the sum of water and air fractions is 1. When =0F, it means that the air phase occupies the whole element. When =1F, it means that the water phase occupies the whole element. When 01F＜＜, it means that the water phase and the air phase share the element together.

1.3 Solution method and boundary conditions

The Pressure Implicit with Splitting Operators (PISO) method is used in this study. The “velocity inlet” is taken as the inlet boundary condition, and the value of the velocity is the average velocity taken from earlier experimental work[13]. The values of k and ε can be taken as

In Eq.(8),iu andih are the velocity and the depth of the inlet, respectively. The hex shaped grid is used in the whole model, and the time step is set as 0.002 s.

1.4 Mesh

The flow field in the T shape fishway is computed using a block-structured numerical mesh which contains an inlet pool, six active pools and an outlet pool.

Fig.2 Detail of the selected numerical mesh (top view)

The mesh size is fairly uniform over each entire pool, with a slightly higher mesh density in the slot region, as shown in Fig.2.

Table 1 Characteristics of the numerical meshes

Fig.3 Velocity distribution in =0.4z

Fig.4 Spatial distribution of the turbulent kinetic energy k and the dissipation rate ε obtained by the numerical model (Z=0.2m)

In order to obtain a mesh independent solution, a mesh convergence analysis was carried out on the design, and three block-structured meshes with different spatia l reso lutions w ere tested, na me ly, M -1, M-2 and M-3.Thenumberofelementsineachmeshwas47 108, 94 245 and 211 243, as shown in Table 1. No significant differences were found in the results obtained with the two finer meshes (meshes M-2 and M-3), as shown in Fig.3.

Fig.5 Velocity distribution in different layers of T shape fishway from the bottom

Fig.6 Schematic diagram of hydraulic loop in the T shape fishway

Thus, an average element size of 5.032×10-4m3(M-2) is adopted, with a slightly higher mesh density in the slot region, as shown in Fig.2.

1.5 Turbulence

Excessive turbulence will make it difficult for fish to orientate themselves correctly. The turbulence kinetic energy, k, and the dissipation rate of the turbulence kinetic energy, ε, can be taken as two indicators of the turbulence in the pools.

The spatial distribution of the turbulent kinetic energy, and its dissipation rate, computed with the k-ε model, are shown in Fig.4 (Z=0.2m).

The results shown in Fig.4 confirm that for both the turbulence kinetic energy k and the dissipation rate of the turbulence kinetic energy ε, their spatial distributions are related with the flow pattern, developed in the pools, although the maximum values are always observed in the pool inlets. Figure 4 also indicates that both k and ε are always larger at the narrow-bands (like the places A-D shown in Fig.1(a)) than in other places, in other words, most energy in the fishway is always consumed at the narrow-bands, as one of the advantages of the slot and drop sill.

1.6 Results and discussions of the numerical simulations

Figure 5 shows the velocity distribution nephogram in different layers of the T shape fishway. In each regular pool, the maximum velocity of around 1.23 m/s appears in the inlet zone and at the sides of the T shape baffle, because the inlet is much narrower than other parts and a high speed jet is formed. The range of the velocity is from 0.9 m/s to 1.15 m/s. Thevelocity is much lower in the front of the T shape baffle, because the flow is blocked up by the T lateral baffle. The range of the velocity in the mainstream is from 0.4 m/s to 1.25 m/s. The height of these drop sills is 0.4 m. Behind the drop sill, the velocity in the layer Z=0.8m is much higher than that in other two layers, Z=0.4m and Z=0.2m, as shown in Fig.5. In the downstream transition pool, the inlet slot forms a high speed jet, and the velocity at the inlet zone is much higher than in other regular pools, owing to the disappearance of the T shape baffle.

2. Physical experiments

In order to compare the results between numerical simulations and physical experiments, a physical model is built with the same size and linear ratio as the numerical model.

All experiments were performed in the State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu, China. The schematic diagram of the experiment device is shown in Fig.6.

Fig.7 Comparison between (a) numerical simulation and (b) physical experiment

2.1 Flow pattern

The comparison between numerical simulations and physical experiments in Fig.7 indicates that the agreement is good in terms of the flow pattern.

The flow process in the T shape fishway can be described as follows: the water is blocked up by the T shape baffle after entering the inlet of the regular pool, to form a backwater zone, then it flows downstream along the two sides of the T shape baffle. Because of the obstruction effect of the drop sill, the water is blocked up again before it reaches the drop sill. A low velocity zone is formed behind the drop sill (from Z=0m to Z=0.4m), then the water flows toward the next regular pool, and this flow process is repeated. The energy dissipation in each pool is scattered in many parts, and there is no large vortex zone, as shown in Fig.7.

2.2 Velocity distribution in regular pool

The experimental work of Wu et al.[16]shows that the flow pattern in each pool is approximately the same for the most regular pools of the fishway. This experimental work focuses on the third regular pool. The screw propeller velocity apparatus, made by Nanjing Hydraulic Research Institute, was used to measure the velocity.

Sixteen measuring points, No. 1-No. 16, are arranged mainly in the mainstream, the measuring points No. 17-No. 19 and No. 20-No. 22 are arranged in two small range vortex zones, respectively, as shown in Fig.8. Table 2 lists the results of these measuring points.

Fig.8 Positions of twenty-two measuring points for the velocity from the bottom

Figure 9 shows the vector diagram in different layers of the T shape fishway (obtained by the numerical simulation).

Table 2 indicates that the maximum velocity appears at Point 2 (Z1=0.2m , V1,max=1.22m/s ), Point 3 (Z2=0.4m, V2,max=1.22m/s ) and Point 1(Z3=0.8m , V3,max=1.21m/s ). The distributions of the velocity at Z1=0.2m and at Z2=0.4m are almost the same in terms of their magnitude. At Z3=0.8m, the velocities at Points 9-16 are very different from those at Z1and Z2because of the drop sill. Table 2 also shows the range of the velocity in each layer: 0.42 m/s-1.22 m/s at Z1=0.2m , 0.44 m/s-1.22 m/s at Z2=0.4m , and 0.52 m/s-1.21 m/s at Z3=0.8m . The results have validated the numerical results, as shown in Fig.5. The comparation between Fig.8 and Fig.9 shows that the flow field of the numerical simulation agrees well with that of the physical experiment.

Table 2 Velocity results obtained by the experiment

3. Disscusion

The turbulent hydrodynamic flows developed in the T shape fishway are mainly characterized by a jet at the inlet zone which is responsible for the most part of energy dissipation, alongside with the obstruction effect of the drop sill. In this study, the new fishway is proposed, which is suitable for the actual situation of the swimming capability of fishes in China, in terms of the velocity and the flow pattern, and the structure and hydraulic characteristics are studied.

Fig.9 Vector diagram in different layers of the T shape fishway

Statistics show that the swimming and jumping capability of the commonly migratory fish in China, such as schizothorax davidi and schizothorax prenanti, are much poorer than that of the migratory fish in other places of the world, like trout, salmon. The maximum velocity that schizothorax davidi and schizothorax prenanti could manage is about 1.74 m/s (the burst speed), while the minimum velocity they could accept is about 0.45 m/s, and the most suitable velocity for them is in a range from 0.8 m/s to 1.2 m/s[10]. The distribution of velocity (0.42 m/s-1.22 m/s) as shown in Table 2 indicates that the T shape fishway can meet the requirements of schizothorax davidi and schizothorax prenanti perfectly in terms of the velocity. Meanwhile, the range of velocity in two vortexzones as shown in Fig.8 is much smaller, compared with that in the large scale vortex zone of the vertical slot fishway as mentioned above, and is beneficial for fish to pass in terms of the flow pattern.

4. Conclusion

In this study, the experiments were carried out to validate the results of the numerical simulations. The results shown in Fig.5, Fig.8, and Table 2 show that the numerical results are in agreement with the experimental results. The distributions of the velocity in different layers are explicitly obtained through numerical simulations and shown in Fig.5 and Table 2. Besides, Fig.1(b) shows that the T shape fishway can provide the whole section migration, unlike the pool and weir fishway. The agreement between the experimental and numerical results confirms that the T shape fishway can meet the migratory requirements of these fish with poor swimming capability and jumping capability in China, in terms of the velocity and the flow pattern. However, there are still many important issues such as life habits and behaviors of the fish that need further studies.

Acknowledgment

The authors would like to thank the State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu, without whose help the development of experimental work essential to this research would not have been possible.

[1] COUTANT C. C., WHITNEY R. R. Fish behavior in relation to passage through hydropower turbines: A review[J]. Transactions of the American Fisheries Society, 2000, 129(2): 351-380.

[2] EAD S. A., KATOPODIS C. and SIKOLA G. J. et al. Flow regimes and structure in pool and weir fishways[J]. Journal of Environmental Engineering and Science, 2004, 3(5): 379-390.

[3] MAO Xi, TUO You-cai and AN Rui-dong et al. Influence of structure on hydraulic characteristics of fishway[J]. Journal of Sichuan University (Engineering Science Edition), 2012, 44(5): 13-18(in Chinese).

[4] MICHEL L. Dams and fish migration[C]. World Commission on Dams Forum. Cape Town, South Africa, 2000.

[5] KIM J. H. Hydraulic characteristics by weir type in a pool-weir fishway[J]. Ecological Engineering, 2001, 16(3): 425-433.

[6] TEIJEIRO T., PUERTAS J. and PENA L. et al. Evaluating vertical-slot fishway designs in terms of fish swimming capabilities[J]. Ecological Engineering, 2006, 27(1): 37-48.

[7] KATOPODIS C. Developing a toolkit for fish passage, ecological flow management and fish habitat works[J]. Journal of Hydraulic Research, 2005, 43(5): 451-467.

[8] CARL R. S. Developing fish passage and protection at hydropower dams[J]. Applied Animal Behavior Science, 2007, 104(3): 295-325.

[9] BUDY P., THIEDE. G. P. and BOUWES N. et al. Evidence linking delayed mortality of Snake River salmon to their earlier hydrosystem experience[J]. North American Journal of Fisheries Management, 2002, 22(1): 35-51.

[10] LI Ming-de. Chinese fish list (Part III)[J]. Tianjin Aquatic, 2002, 28(2): 20-23(in Chinese).

[11] JOHNSON G. E., EBBERTS B. D. and DAUBLE D. D. et al. Effects of jet entry at high-flow outfalls on juvenile Pacific salmon[J]. North American Journal of Fisheries Management, 2003, 23(2): 441-449.

[12] WANG Xing-yong, GUO Jun. Brief review on research and construction of fishways in China and abroad[J]. Journal of China Institute of Water Resources and Hydropower Research, 2005, 3(3): 222-228(in Chinese).

[13] MAO Xi, LI Jia and YI Wen-min et al. Study on optimizing the structure of the fishway[J]. Journal of Sichuan University (Engineering Science Edition), 2011, 43(S1): 54-59(in Chinese).

[14] WANG Fu-jun. Computational fluid dynamics analysis: The principle and application of CFD software[M]. Beijing: Tsinghua University Press, 2004(in Chinese).

[15] AN Rui-dong, LI Jia. Charecteristic analysis of the plunging of turbidity currents[J]. Journal of Hydrodynamics, 2010, 22(2): 274-282.

[16] WU S., RAJARATMAN N. and KATOPODIS C. Structure of flow in vertical slot fishways[J]. Journal of Hydraulic Engineering, ASCE, 1999, 125(4): 351-360.

* Project supported by the National University Student Innovation Program of China (Grant No. 091061021).

Biography: MAO Xi (1987-), Male, Ph. D. Candidate

TUO You-cai,

E-mail: tyxl200496@163.com