Abstract
Water reservoirs are barriers to fish migration, making it difficult for fish to access their habitats and reproduction areas. This creates a major environmental problem, which is effectively addressed by the use of Fish Passes (FP), which allows fish to migrate upstream. An FP usually consists of a sloping channel with baffles, which divide the channel into sections - pools. The most common and most suitable for the species found in Greece are the Vertical-Slot Fish passes (VSF).Most existing practical design methods for VSFs use empirical equations and/or suggested values for the main dimensions of the tanks in order to dissipate the energy of the incoming jet, reduce its velocity and drive it smoothly to the next slot. These methods are usually based on measurements, mainly of flow velocities and turbulence kinetic energy dissipation obtained from experimental investigations in physical models. Often, this design approach is not successful, thus making it imperative to develop ratio ...
Water reservoirs are barriers to fish migration, making it difficult for fish to access their habitats and reproduction areas. This creates a major environmental problem, which is effectively addressed by the use of Fish Passes (FP), which allows fish to migrate upstream. An FP usually consists of a sloping channel with baffles, which divide the channel into sections - pools. The most common and most suitable for the species found in Greece are the Vertical-Slot Fish passes (VSF).Most existing practical design methods for VSFs use empirical equations and/or suggested values for the main dimensions of the tanks in order to dissipate the energy of the incoming jet, reduce its velocity and drive it smoothly to the next slot. These methods are usually based on measurements, mainly of flow velocities and turbulence kinetic energy dissipation obtained from experimental investigations in physical models. Often, this design approach is not successful, thus making it imperative to develop rational design methods. One such method is by using an Integrated Mathematical Model (IMM). An IMM consists of a hydrodynamic model and an ecological model. The hydrodynamic model, which calculates the flow field in the VSF, is usually a CFD (Computational Fluid Dynamics) model. The ecological model uses the characteristics of the flow field to calculate the behaviour-movement of the fish, seeking to minimise the energy consumed, i.e. following the low velocity and turbulence regions. CFD models allow optimizing the design of an VSF by performing modifications to the geometric characteristics of the pools. However, for the results of a CFD model to be considered reliable, they must be confirmed by experimental data. In most VSFs research papers, it is usually assumed that the flow is two-dimensional, so it is satisfactorily described in most of the tank, but not in the slot region, where the flow is three-dimensional. For the hydrodynamic simulation of rivers, mainly because of the very large size of the considered section, we use two-dimensional models that solve the Saint-Venant equations and give very good results in a short computational time. Fish behaviour simulation models (FBM) provide the necessary information on fish movement and behaviour in rivers or VSFs. In the FBM, the basic functions of fish are simulated, which, in response to hydrodynamic stimuli from the environment, such as flow velocities and acceleration differences, make the decision to perform specific movements. The area within which a fish perceives the hydrodynamic stimulus is defined by 5 location points around the perimeter of the fish and is called the sensory ovoid. In general, three modes of fish swimming can be distinguished (1) continuous, (2) prolonged and (3) burst. These modes are associated with their respective movements which are (a) active swimming in the direction of the flow, (b) passive swimming and entrainment by the flow, (c) avoidance and changing angle with respect to the flow, and (d) escape and swimming against the flow. The most important FBM available in the literature is the ELAM model. ELAM is a model of the Individual Based Model (IBM) format that simulates fish as independent units, considering differences between them, such as size. It combines the Particle Tracking method with the minimum energy movement principle, i.e. the fish prefers low velocity and turbulent regions with small flow changes. This PhD thesis poses the main research question which is: "How do we rationally design an VSF with an IMM so that it is effective? "To answer the aforementioned research question the following 5 sub-research questions must be addressed: (1) How can we describe with a mathematical model the movement of fish in a VSF? (2) How can we describe with a mathematical model the movement of fish in a river in the area, where a VSF is located? (3) How do we design a VSF so that it allows the safe movement of fish? (4) How do we place a VSFs entrance near a dam to attract fish?(5)Which turbulence model best simulates the flow field in a VSF? The original point of this dissertation, which answers the research questions posed, is the development of an IMM to rationally design a VSF to be effective. The IMM consists of two sub-models (a) an IMM for rivers and (b) an IMM for VSFs. Each IMM shall consist of a hydrodynamic model and a FBM. The hydrodynamic model used in rivers is TELEMAC-2D and the model used in VSFs is FLOW-3D. The FBM for both rivers and VSFs were structured as part of this thesis. The IMM was calibrated and validated with experimental data from the international literature and 2 applications followed. The first application was aimed at the geometric optimization of VSF. A hydrodynamic simulation of the flow in VSF for different geometric characteristics was performed. The simulation results were combined with fish and macroinvertebrate habitat suitability indices and finally the VSF with those geometric characteristics that create conditions suitable for most of the species considered was selected. The second application was the design of a VSF for an irrigation dam with a variable reservoir water level. In this application, the Fish-Pass Tower (FPT) was proposed. The FPT tower consists of 14 rows of VSFs and a total of 99 pools, which have a side opening through which reservoir water enters the VSF. The holes are regulated by sluice gates and their role is to allow water to enter the VSF depending on the reservoir level. The following conclusions can be drawn from the calculations of the OMM in rivers.1.The hydrodynamic model can satisfactorily simulate the flow field in a river using the Standard k - epsilon (SKE) turbulence model as proposed in the literature. The Root Mean Square Error (RMSE) correlation index of the flow velocities calculated by the model with respect to the measured ones ranges from 0.03 to 0.054, with an optimum value of zero, while the Index of Agreement (IA) ranges from 0.96 to 0.99, with an optimum value of one.2.The FBM can estimate with reasonable accuracy the percentage of fish that will follow a straight or circular trajectory. More importantly, however, it can estimate the percentage of fish that will exit the model from a given exit. This can be generalized by considering the model output to be the input of a VSF. Given that the hydrodynamic model can be easily modified to simulate different fish attraction currents from a VSF (different outlet flows), then we can use the fish behaviour model to determine under which conditions the largest proportion will move towards the VSF entrance. 3.Accurate simulation of an experimental trajectory is not entirely feasible, because the model introduces randomness in the calculation of its variables, which alter the simulated trajectories. However, if a sufficient number of fish are simulated, from an experimental starting position, most of the calculated trajectories will follow the direction of the measured trajectory. 4. IMM can be used as a tool to draw useful conclusions about habitat suitability by utilizing the experimental data and the calculated flow field. The following conclusions can be drawn from the IMM calculations in VSFs. 1. The flow field in a VSF is quite complex. Often, a three-dimensional CFD model is necessary to calculate the flow field, especially to accurately describe the flow in the regions of the slot. 2. The RNG KE turbulence model and the Volume of Fluid (VOF) method can give accurate results for determining the free surface of the flow with relatively low computational cost. 3. The 3D model calculations showed (1) good agreement with experiments and 2D calculations of the calculated mean flow velocities (RMSE = 0.01 -0. 03), (2) better performance in determining the flow surface in the VSF, which is attributed to the VOF method, and (3) better calculation of turbulence characteristics than the 2D models, which is due to the RNG KE turbulence model (which overcomes the problem of overestimation of turbulence in the SKE model) and the fact that the 3D model takes into account the 3D characteristics of the flow in the VSF. 4.The 3D hydrodynamic calculations have shown that the usual assumptions in VSFs, that (a) the flow is 2D and (b) the simulation of 3-5 tanks is sufficient to achieve satisfactory results, are not always valid, and the above assumptions need to be checked depending on the internal structure and the number of pools of the VSF. 5. The fish behaviour model can estimate with reasonable accuracy the proportion of fish that will follow a particular path upstream of the VSF. The FBM cannot simulate the resting time of fish within the VSF, but it does indicate that if fish do not pass directly from the downstream to the upstream pool, then the time of "aimless" movement within the tank will be in some area, which is considered a resting area. The main research question posed in this thesis was addressed by the following five sub-research questions, which are briefly answered. 1. How can we describe with a mathematical model the movement of fish in a VSF? Often it may be necessary to use a three-dimensional mathematical model and in certain cases a two-dimensional model may be sufficient. By using the mathematical model, it is possible to accurately determine depths and flow velocities and to calculate turbulence characteristics with a reasonably good accuracy. These are the flow characteristics that influence the movement of fish within the VSF. Therefore, by correctly calculating them and considering that fish receive these hydrodynamic stimuli and follow the rules of (a) low energy consumption and (b) turbulence avoidance, we can describe the movement of fish within the VSF. The mathematical model must also consider habitat suitability indices, so that an initial geometric optimization of the FP can be made in terms of its use by more species. 2. How can we describe with a mathematical model the movement of fish in a river in the area, where a VSF is located? Rivers are more complex than VSFs in terms of describing the FBM, whereas the case is different for the hydrodynamic model. It is important to emphasize the greater complexity of the FBM in rivers because fish are already acclimatized to the hydrodynamic characteristics of rivers. Therefore, it is important to have observations for the design fish so that the model can obtain appropriate values for its parameters.3.How do we design a VSF so that it allows the safe movement of fish? To properly design a VSF that allows for the safe movement of fish, we must use a mathematical model to (1) create a flow field appropriate of the fish of interest, (2) provide low velocity resting areas to reduce the required upstream energy in a VSF, and (3) avoid flow depths less than the minimum acceptable depth for fish. 4.How do we place a VSFs entrance near a dam to attract fish? Optimal placement of a VSF is achieved by the following steps: (1) finding the habitat suitability index of the river for the species of interest to place the VSF in an appropriate location; (2) optimizing the VSF using ecological criteria; (3) applying the IMM to the river area including the optimized VSF and (4) optimizing the hydrodynamic characteristics of the VSF inlet and outlet to maximize the proportion of fish moving toward the VSF. This answer is in good agreement with most of the literature.5.Which turbulence model best simulates the flow field in a VSF? The calculation of turbulence characteristics affects the flow field and fish-pass calculations. Thus, this calculation needs to be as accurate as possible, but at the same time effective in terms of computational cost, especially in 3D models. The use of the RNG KE model can achieve satisfactory accuracy with a relatively short computational time for the cases of flow in a VSF. Simpler models (Mixing Length, SKE, Constant viscosity) cannot predict the flow field satisfactorily, while more complex models (Large Eddy Simulation) greatly increase the computational time and complexity of the results.
show more
