9.1 Introduction

The fast-moving turbine parts create challenging hydrodynamic conditions for a fish passing through a turbine. Hence, several modelling approaches have been developed to assess the impact on the fish during a turbine passage, which potentially causes injuries. Survival rates can be predicted based on the fish species, the turbine main dimensions and the operating condition.

The most simplified modelling approach is statistical modelling of a set of experimental data e.g. (Larinier and Travade 2002). Such empirical models are based on specific field test data. Therefore, transferability to other application cases is only limited. To provide a more general applicability, a number of models have been developed throughout the decades, which account for the physical conditions and the resulting biological response of the fish. Such models are available for the most common turbine types, i.e. Kaplan and Francis turbines.

Collisions of the fish with the rotating runner blades, so-called strike events, are the most evident source of damage. Thus, most models evaluate the risk of exposure to relevant physical load by calculating the runner blade strike probability. This can be derived by theoretical considerations of a simplified geometric object that passes through the runner blades. Such strike probabilities can facilitate basic assessment of overall damage rates. Different models have been developed for blade strike probability calculation. Most eminent are those of von Raben (1958) and Montén (1985).

In recent years more enhanced methods based on numerical simulations (e.g. Richmond et al. 2014) were developed allowing the analysis of typical physical impact variables as stress, shear and barotrauma of the downstream fish passage. These physical impacts can then be correlated to the biological impact on the fish.

In the following, enhancements of both simple und advanced modelling methods are investigated. Additionally behavioural aspects are considered in the modelling as these influence the risk of potential injuries in a passage event. The modelling methods are applied to the FIThydro Testcases in Bannwil, Guma and Obernach, and analyzed at representative operating conditions.

9.2 Introduction to Testcases

The impact of turbine passage on fish depends not only on the turbine type, which is applied at a certain power plant. Even for the same turbine type, several factors, like size, rotational speed, number of runner blades and operation mode, play an important role. To accurately judge the effect of these factors on turbine fish passage a more detailed evaluation of each particular situation is required. In the following, the focus is on typical run-of river power plants, which mostly are equipped with axial turbines, either with movable blades or as propeller machine. The investigation performed in the FIThydro project focusses on bulb turbines. In Table 9.1 a brief overview on the main parameters of the Testcase power plants is given.

Table 9.1 Main parameter of testcase turbines
Full size table

In the following, the modelling approaches are briefly introduced. Then some exemplary results of the research at the FIThydro sites are presented. More detailed results of both the fish passage modelling and the experimental tests are presented in the FIThydro documentation (Geiger and Stoltz 2019).

9.3 Simple Modelling Methods

Based on the existing simple modelling approaches for strike probabilities, recent work on generalized modelling (Geiger 2018) showed, that model differences originate from different assumptions for fish body alignment to the flow direction close to the turbine. This aspect is related to the fish behaviour during turbine passage, which is still not sufficiently understood. For typical turbine designs and operating conditions, the different blade strike models provide comparable results. The comparison of the simple modelling approaches with CFD based analysis showed that the modelling accuracy depends strongly on hydraulic conditions. Opening angles of runner blades and corresponding flow angle close to the runner blade entrance edge need to be estimated correctly. These values are frequently not directly available. Therefore, several authors provided estimates or empirical formulas. As the design of turbines is based on fundamental laws of physics, design guidelines for hydraulic machines can provide well-estimated values for blade strike modelling. This allows an impact assessment with simple and fast methods respecting the underlying physical relations and engineering procedures. The comparison of this method with 3D-CFD results for the FIThydro Testcases showed a better agreement for parameters like the flow angles. Therefore, it provides more accurate modelling results and a broader applicability, compared to previous empirical models (Geiger et al. 2020a).

In a second step survival rates can be derived based on the runner blade strike probabilities. As it is well known that not every collision of fish with a runner blade results in relevant injury further aspects need to be taken into account. If the relative speed of the fish in relation to the runner is small, the impact of a strike event is less significant. Thus, the biological response of a fish species is playing a role. In the 20th century, these relations were included in the modelling approaches by statistical methods using empirical data of test results at specific hydropower plants (e.g. von Raben (1958), Montén (1985)). However, this approach causes transferability problems, especially for untypical turbine setups or operating conditions.

In the last decades, experimental campaigns were conducted to gain detailed reference information on resulting fish injury in case of physical load, for example in function of blade shape and blade strike velocity (Turnpenny et al. 2000; Amaral et al. 2011). This biological response data can be combined with existing blade strike models to improve the modelling method. As blade strike related mechanical injury is usually the dominant impact source for fish passage at low head run-of-river HPP, this simplification can provide useful information.

The basic relations between hydraulic boundary conditions, runner blade strike probabilities and biological response enable a simple assessment of fish damage rates during turbine passage as shown in Fig. 9.1. Comparison with literature references shows typical accordance of predicted mortality rates within a magnitude of about 10 % of the experimentally observed values (Geiger et al. 2020a). The unresolved accuracy of experimental data remains an issue in this context, as well as the influence of the actual fish behaviour during turbine passage.

Fig. 9.1
figure 1

Flow chart of the main steps and aspects of damage rate modelling

Full size image

The most common assumption for modelling purpose is that a fish passes through the turbine at mid-radius; the fish body is aligned with the flow and has the identical velocity, without active swimming speed. A comparison of model results with experimental observation of damage rates suggests that this assumption provides typically suitable accordance. The generalized modelling also allows for correct modelling of different passage locations, orientations and speed.

The simple modelling methods enable a fast and inexpensive quantification of damage rate magnitudes. They do not require special hardware, software or detailed information, for example about the turbine geometry. Only basic turbine and power plant parameters like runner diameter, number of runner blades, rotational speed, discharge and head are required. In order to improve fish passage by tailor-made runner blade designs for enhanced fish passage, simple modelling approaches are not advisable. The exposure to hydrodynamic forces and the impact on the fish during the turbine passage can only be roughly estimated. The comparison within the FIThydro project also showed systematic deviations of such simplified modelling approaches from the detailed 3D CFD evaluations (Geiger and Stoltz 2019).

9.4 Advanced Modelling Methods Using CFD

To capture detailed information on hydrodynamic conditions during turbine passage the numerical modelling is most accurate. A widely used approach is to perform steady state CFD modelling and to analyse the simulation results following streamlines and extracting the physical relevant information concerning strike, pressure, turbulence and shear along them. Besides this rather simple procedure, other approaches using transient CFD are also available, which use i.e. particle tracks to evaluate the strike risk. As these models are rather complex and time-consuming to apply, it is not suitable as standard procedure during the design process of a water turbine. Hence, the focus of the research work in the FIThydro project was to identify a process suitable for industrial applications.

The evaluation of the fish passage assumes that the pathway of a fish follows a streamline through the turbine. These streamlines are generated with a stationary CFD simulation and are then post-processed with a tool applying the biological performance assessment (BioPA) developed by Pacific Northwest National Laboratory (PNNL) (Richmond et al. 2014). The information for the stressor exposure of pressure, strike, shear and turbulence can be extracted directly from the streamlines. For strike, the velocity vectors close to the blade entrance edge are used in order to calculate the strike probability and the impact velocity a fish experiences when colliding with the blade. For the other stressors an exposure probability is derived based on a large number of streamlines. As presented in Fig. 9.2, the injury risk can then be derived by combining the physical information with dose response data of respective fish species. A score is then integrated over the product of exposure probability and exposure mortality of the fish. The value is high when the risk of passage injury is low. It is understood that for now, the score does not represent an absolute passage-survival estimate. However, it offers a systematic way to evaluate trade-offs associated with various hydraulic solutions. An optimization of the hydraulic shape of the turbine can then be performed in order to reduce the risk of an injury during turbine passage (DeBolt et al. 2015).

Fig. 9.2
figure 2

Example of BioPA evaluation with probability distribution and biological dose-response of stressor variable as typically applied (see also Richmond et al. 2014)

Full size image

As a part of the FIThydro project, investigations at the laboratory Testcase in Obernach were performed. The impact of turbine passage on brown trout of various fish lengths was investigated for the Kaplan Bulb-Turbine at different operating modes (Geiger et al. 2020b).

Figure 9.3 presents the results of the analysed full load point, including a variation of fish length. The results underline the importance of reasonable definition of the relevant fish length, depending on life stages of fish, fish species or trash rack clearances. The field test assessment of the fish passage by the team of the Technical University of Munich could be confirmed by the CFD based evaluation. The identified injuries at site were mainly caused by strike events, no clear indication for barotrauma and shear related injuries could be found. This was also confirmed by the CFD-based modelling method; only the shear stress influence seems to be over predicted. As presented in Fig. 9.3 the strike modelling also exceeds the values of the experimental results. Eye catching is the fact that a continuous shift in the results seems to be present.

Fig. 9.3
figure 3

Results of CFD modelling in comparison to field test for the rated operating point with H = 2.5 m and Q = 1.5 m3/s at the Obernach Testcase

Full size image

Looking into existing studies we came across the results of experimental tests in the Oak Ridge Laboratory (Bevelhimer et al. 2017), which indicate that tail strikes rarely lead to an injury of the fish. This corresponds well to the principal definition of fish body region of a trout in which the tail region is approximately 1/3 of the body. Accordingly adapting the effective fish length by a factor of 2/3, improves the agreement of the strike modelling to the experiments significantly and leads to an excellent match of the datasets. Additionally, other effects like fish orientation also partly contribute to this effect.

This leads to another special focus within the FIThydro project, the modelling of fish behavioural effects. The research goal was to evaluate the significance of these influences on the impact a fish experiences during turbine passage.

These effects could be studied for the test site Bannwil in Switzerland and the Guma power plant in Spain. In a first phase, tests with BDS sensors were performed for both power plants. The sensors were injected at the intake of the power plant and passed neutrally buoyant with the flow through the turbine recording pressure data. This data was then compared with a standard BioPA analysis of the same operating conditions. The results showed a good agreement for the assessment of the nadir pressures. In the second phase, the modelling was extended to cover the effect of behavioural aspects.

One interesting consideration is the passage location of a fish. Three passage locations as illustrated in Fig. 9.4 at inside, middle and outside having the same area were used to analyse several operating conditions.

Fig. 9.4
figure 4

Variation of passage location coloured by strike survival ratio from low (blue) to high (red)

Full size image

The results in Fig. 9.5 show a representative operating point for the Bannwil power plant. It can be seen that the passage location influences the survival ratio based on the different stressors. The results show that the passage location has a significant influence on survival. A passage in the middle of the blade is favourable regarding pressure and shear influences and a passage close to the hub is favourable especially regarding strike, due to the low impact velocities.

Fig. 9.5
figure 5

Variation of passage location at an operating point close to the optimum at the Bannwil testcase

Full size image

Besides the passage location, also the fish orientation and potential swimming speed influence the strike rate. Based on the assumption that a fish swims against the main flow direction and maintains his swimming depth the basic strike formula is extended as presented in Fig. 9.6 (Geiger 2018).

Fig. 9.6
figure 6

Velocity triangle and strike formula including fish velocity and orientation (Geiger 2018)

Full size image

Typical flow velocities during turbine passage are low in the intake region, but as soon as the flow approaches the guide vanes the flow accelerates and velocities increase rapidly above more than 2 m/s, which is typically the velocity most fish can swim against for a certain amount of time. Faster velocities are only possible for short sprints. Like this most fish species will barely be able to withstand the flow conditions and have only limited capabilities to control or influence turbine passage trajectories in close vicinity of the turbine. Nevertheless, a principal study is performed, and the results presented in Fig. 9.7 indicate that a fish actively swimming against the main flow increases the passage time through the turbine and therefore the strike probability. At the same time the impact velocities increase slightly. Therefore, the overall risk of a strike injury rises.

Fig. 9.7
figure 7

Influence of fish swimming speed and fish orientation on survival rate at full load operation at the Guma power plant

Full size image

Regarding the orientation of the fish in relation to the blade a derivation of the von Raben correlation (von Raben 1958) is used, which assumes that the fish is oriented with the flow. Variations of the orientation as applied in Fig. 9.7 are presented in relation to the absolute flow direction, which is indicated with 0°. A range from a radial to an axial orientation is considered, whereas an angular deviation of more than 45° to the main flow is highly unlikely. Analysing the results, the standard angle of 0° is a rather conservative approach, with a maximum tolerance of 1–2% expected for the fish passage assessment. Benchmarking the different influencing factors, fish length and passage location have a much higher impact as the orientation on the survival rate modelling. In general, the importance of any impact also depends on turbine size and speed, as well as the operating conditions. Site-specific considerations are advisable.

9.5 Summary and Outlook

During the FIThydro project simple and more complex modelling methods were applied. Depending on the purpose and the stage of the development of a power plant, the right method needs to be chosen. Especially in early stages of a project simple methods are sufficient to gain a general overview, however it is recommended to include influencing factors in the analyses to avoid a blurry picture of the situation.

It is important to represent the hydraulic boundaries as accurate as possible by applying relevant operating conditions. In addition, relevant fish species need to be identified to judge the biological sensitivity correctly and to set the right focus for mitigation measures. More advanced CFD based methods enable an enhanced turbine design to improve fish passage conditions significantly. Projects like the Ice Harbor power plant at the Snake River (Foust et al. 2013) in the US show that it is already possible to apply these modelling methods successfully in the turbine design process.

In the future, a good and accurate modelling in combination with the application of sensors might avoid the need of life fish tests. An Assessment of possible impacts on the ecology of river reaches is already feasible in early stages of a hydropower project. Accordingly, if needed mitigation measures to minimize the influences can be designed.