Content uploaded by Laurent David
Author content
All content in this area was uploaded by Laurent David on Sep 27, 2023
Content may be subject to copyright.
Proceedings of the 40th IAHR World
Congress
21-25 August
2023
Vienna, Austria
DOI number
©2023 IAHR. Used with permission / ISSN-L 2521-7119
Scale effects in a Vertical Double Slot Fishway
Anne-Fleur Lejeune(1), Damien Calluaud(1), Laurent David(1), Gérard Pineau(1), Jean Carl Rousseau(2), Richard
Tessier(2), Sylvain Richard(3), Laurent Dupuis(2), Olivier Mercier(3) and Pierre Sagnes(4)
(1) Institut PPrime, CNRS-Université de Poitiers-ISAE-ENSMA, Pôle R&D Écohydraulique, OFB-IMFT-PPRIME, Poitiers, France,
anne.fleur.lejeune@univ-poitiers.fr; damien.calluaud@univ-poitiers.fr; laurent.david@univ-poitiers.fr; gerard.pineau@univ-poitiers.fr
(2) Institut PPrime, CNRS-Université de Poitiers-ISAE-ENSMA, Poitiers, France,
jean.carl.rousseau@univ-poitiers.fr; richard.tessier@univ-poitiers.fr; laurent.dupuis@univ-poitiers.fr
(3) Office Français de la Biodiversité (OFB), Direction de la police et du permis de chasser, Pôle R&D Écohydraulique, OFB-IMFT-
PPRIME, Toulouse, France,
sylvain.richard@imft.fr
(4) Office Français de la Biodiversité (OFB), Direction de la recherche et de l’appui scientifique, Pôle R&D Écohydraulique, OFB-IMFT-
PPRIME, Toulouse, France,
pierre.sagnes@ofb.gouv.fr; olivier.mercier@ofb.gouv.fr
Abstract
This paper presents results on the scaling effects between a full-scale a Vertical Double Slot Fishway (VDSF)
in Malause, on the Garonne River (France) and its 1:6 model reproduction in the Pprime Institute of the
University of Poitiers (France), which assured the Froude similitude. Acoustic Doppler Velocimetry (ADV)
measurements were performed to access the three components of the velocity on two profiles in each fishway.
Full-scale measurements are exposed. Flow topology, mean flow and unsteady velocity components features
as well as turbulence kinetic energy are evaluated for both scales and compared to one another. Results show
a good agreement between the dynamics of the turbulent flow in both scales, which indorse the use of laboratory
scale models for designing full-scale fishways in the future and in particular VDSFs. The flow and turbulence
features could be estimated from laboratory experiments and extrapolated for an ecohydraulics point of view.
Keywords: Vertical double slot fishway; Ecohydraulics; Scale effects; In-site and laboratory experiments; Turbulent flow;
1. INTRODUCTION
Adopted in 2000, the Water Framework Directive dictates that all European countries must achieve a good
ecological status or potential of their water bodies. This involves, among other things, to insure longitudinal
connectivity, especially in classified rivers. The removal of obstacles may be a solution, but many of them are
useful for human activities and cannot be withdrawn. As an alternative, fishways are often installed to allow fish
passage upstream of obstacles.
Although the general design of the Vertical double slot fishways (VDSFs) is relatively old (Clay, 1995), their
flow characteristics have been the subject of very few studies (Fujihara et al., 2003; Zhang et al., 2023). To our
knowledge, no publications focus on the hydraulics of only VDSFs. The existing VDSFs were designed on the
basis of criteria used for the Vertical single Slot Fishways (VSFs) that were widely used and studied (Rajaratnam
et al., 1986; Marriner et al., 2014; Klein & Oertel, 2016; Ballu et al., 2016, 2017; Zhao et al., 2022).
A study of the turbulent flows occurring inside a VDSF is underway at the Pprime Institute in Poitiers
(France) (Lejeune et al., 2022). It aims at identifying the impact of different parameters (slope, discharge…) on
the flow patterns and at optimising the design criteria for future VDSFs. Experiments are carried out on a scale-
model of an existing French VDSF. However, all the quantities measured on the scale-model cannot be
transposed to full-scale due to scaling effects, specifically concerning the fluctuating quantities. Indeed, scale
effects occur when the force ratios are different between the natural phenomenon and the laboratory controlled-
one. The bigger the scale factor, the further the observations will be from one another. In this work, an estimation
of the scale effects between a full-scale VDSF and its reproduction (six times smaller, using the Froude
similitude) is presented.
Proceedings of the 40th IAHR World Congress
21-25 August 2023, Vienna, Austria
©2023 IAHR. Used with permission / ISSN-L 2521-7119
2. DATA ACQUISITIONS
2.1. Description of experiments and positions of profiles measured
Two sets of similar experiments were conducted: one in a full-scale VDSF, located in Malause on the
Garonne River (France) in July 2022, the other on a 1:6 scale-model of this VDSF in the Pprime Institute,
University of Poitiers (France). Acoustic Doppler Velocimetry (ADV) measurements were carried out during both
campaigns with a SonTek MicroADV 16 MHz. The three-components of the velocity on several profiles with a
50 Hz frequency over a 300 s acquisition time were collected. Raw data were then filtered with a phase-space
filter to remove aberrant values, as recommended for this type of data (Goring & Nikora, 2002). Post-treated
data were used to determine information about turbulence (mean velocity, turbulence kinetic energy, …) and
temporal and spatial flow patterns in VDSFs. For the full-scale experiment, the ADV was immersed inside a
pool and moved into the desired positions using a metal transverse structure that can be seen below (Figure
1).
(a)
(b)
Figure 1. Experimental set-up for (a) the full-scale ADV measurements (Vertical Double Slot Fishway of
Malause, Garonne River, France) (b) the 1:6 scale model ADV measurements (Pprime Institute, University
of Poitiers, France)
The measurements were conducted in the fifth pool from upstream (over seven) in the on-field full-scale
VDSF and in the third pool (over five) in the 1:6 laboratory experiments. The main dimensions of the pools of
the full-scale VDSF are presented below (Figure 2). The slot width is of b = 450 mm, the width and length of
the pools are B = 4 m and L = 5 m, respectively. The fishway is equipped with bottom macro-roughnesses
(truncated cones, density of 10%, using the methodology implemented before (Ballu et al., 2017)). The
diameters of each cone at its base is DMR – b = 200 mm and at its top DMR – h = 150 mm. The fishway slope is of
s = 4%. The water level drop between two successive pools was measured all along the VDSF and the average
value is Δh = 174 mm, and the water depth in the middle of the fifth pool is of h = 2 m. The flow discharge during
the 1:1 experiments has been measured using an Acoustic Doppler Current Profiler (ADCP). For the laboratory
experiments, the flowrate was controlled and measured using a flowmeter. The flowrate was normalised using
a previously introduced expression [1] (Puertas et al., 2004).
Proceedings of the 40th IAHR World Congress
21-25 August 2023, Vienna, Austria
[1]
with Q the flow discharge (m³·s-1), g the standard gravity (9.81 m·s-2) and b the width of the slot (m). The
measured QA is then equal to 7.0.
The profiles studied for this work are the profile C, which is transversal at X = 1065 mm (= 2.37·b) and the
profile E’, which is longitudinal and located at Y = 1950 mm (= 4.33·b). A specific point of the profile C (C5),
located at X = 1065 mm (= 2.37·b) and Y = 1295 mm (= 2.88·b) where the velocity from the slot jet is maximum
was also investigated in order to study the temporal fluctuation of the velocity in this zone. All the elements of
interest are located at half the water depth, hence Z = 1000 mm (=2.22·b). The elements of interest are also
drawn on the figure below (Figure 2).
Figure 2. Dimensions of the full-scale VDSF pools and profiles positions
2.2. Scaling laws
The 1:6 scale laboratory model was established from the dimensions of the VDSF in Malause using the
Froude similitude. This criterion is often used to study flow characteristics in fishways, since it jointly considers
inertia and gravity forces, both important in this type of free-surface flow. In order to maintain the same Froude
number [2] in the 1:1 (full-scale) VDSF and in the 1:6 scale model, the following equations for time (t) [3], mean
or fluctuating velocity (U) [4] and turbulent kinetic energy (k) [5] can be deduced, using n as the geometrical
scale (here n = 6).
[2]
[3]
[4]
[5]
with U the velocity (m·s-1), g the standard gravity (9.81 m·s-2), h the water depth (m).
Proceedings of the 40th IAHR World Congress
21-25 August 2023, Vienna, Austria
©2023 IAHR. Used with permission / ISSN-L 2521-7119
Using those expressions, it was possible to transform the measures obtained in the 1:6 scale model back
to full-scale for further comparisons. The following results will be presented in 1:1 scale (whether native 1:1
scale or corrected to 1:1 scale for scale-model results).
3. SCALE EFFECTS ESTIMATION
The topology, mean flow and fluctuations occurring inside the 1:6 scale VDSF were presented previously
(Lejeune et al., 2022). Here, the velocity magnitude and the turbulent kinetic energy along the two profiles
studied are presented below (Figure 3 and Figure 4). The same topology was observed in both scales, and the
results are really close to one another. This is a really promising result, since it includes data from field
experiments, where hydraulic conditions were less controlled than in laboratory experiments. In profile C, the
two jets coming from the slots created a rise of the mean velocity on the velocity profile. Outside of the area of
the jets, the velocity was much lower in both configurations. The two jets mixed together in the middle of the
pool in both scales, as can be seen in profile E’, as the mean velocity increased significatively before decreasing
downstream of the middle of the pool. The same patterns of turbulent kinetic energy could be observed in both
scales. The fluctuations were highest in the middle of profile C, with two smaller pikes due to the wavering of
the jets. In profile E’, the highest values of turbulent kinetic energy were located in the area close to the central
baffle of the previous pool and they then decreased along the upstream-downstream direction, except near the
middle of the central baffle in both scales.
The 1:6 scale-model overestimated the velocity slightly, in high-velocity areas (such as the two jets on
profile C). The scale effects seemed rather limited on mean velocities. Also, the scale-model underestimated
slightly the turbulent kinetic energy along the profile E’, while it gave similar results to full-scale along the profile
C. The fluctuations seemed more dependent to scale effects than the mean velocity, although the results are
still quite close. Several reasons can explain those small differences. The choice was made to design the scale
model based on the Froude similitude; however, the Reynolds number was different, since both criteria cannot
be maintained in the scaling. The Reynolds number -using the slot width for the reference dimension and the
theoretical velocity occurring inside a slot based on water level drop between pools- of the full-scale is equal to
Re – full-scale = 831000 whereas the Reynolds number of scale-model is equal to Re – scale-model = 61000 (more than
thirteen times lower than Re – full-scale). Furthermore, the ADV-probe used to collect the velocities was the same
in the two measurement campaigns and in proportion to the pool volumes, both the volume intrusion by the
probe and the volume of measurements had a bigger impact on the 1:6 scale-model than on the full-scale VDSF.
The velocity fluctuations occurring inside the full-scale VDSF had larger amplitudes as can be seen below on
the specific point, C5 (Figure 5). The fluctuating velocity on the streamwise-direction was slightly more
contained in the 1:6 scale-model than in full-scale. Nevertheless, the frequency of the two were rather similar,
meaning that the unsteady flow is not affected by scale in this configuration.
(a)
(b)
Figure 3. Velocity magnitude for the full-scale VDSF (in red) and the 1:6 scale-model (in blue) for (a) profile
C and (b) profile E’. The discharge QA [1] was equal to 7.0 and the slope s was equal to 4%.
Proceedings of the 40th IAHR World Congress
21-25 August 2023, Vienna, Austria
(a)
(b)
Figure 4. Turbulent kinetic energy for the full-scale VDSF (in red) and the 1:6 scale-model (in blue) for (a)
profile C and (b) profile E'. The discharge QA [1] was equal to 7.0 and the slope s was equal to 4%
(a)
(b)
Figure 5. Time evolution of the u' fluctuating velocity in 1:1 scale at point C5 for (a) the 1:6 scale-model and
(b) the full-scale VDSF.
Proceedings of the 40th IAHR World Congress
21-25 August 2023, Vienna, Austria
©2023 IAHR. Used with permission / ISSN-L 2521-7119
4. CONCLUSIONS
In order to understand hydraulics inside fishways and draw conclusions on the impact of different
parameters on flow patterns, the use of scaled models in laboratories is widely applied. However, the impacts
of scaling are less known and are often a source of uncertainties when designing full-scale fishways. This study
compared results from a full-scale Vertical Double Slot Fishway (VDSF) in Malause on the Garonne River
(France) and its 1:6 scale-model reproduction (Pprime Institute, University of Poitiers, France). It appeared that
the flow topologies of both scales were really similar, that the 1:6 scale-model is an appropriate way to
understand and estimate the flow occurring inside a VDSF. There are some differences between the quantitative
values, but they are truly minor considering the experiments undertook. Those differences could come from
several well-known sources, like the difference in Reynolds number due to Froude similitude or the inflow
conditions that were difficult to adjust and control to reproduce exactly in both VDSFs. The difference could be
also come from measurement artefact. Indeed, both experiment campaigns were conducted with the same
probe, meaning that the ratio of the intrusion and integration volumes to the volume of the pool were different
in the two scales. Despite these uncertainties, the results show good agreement and allow us to rely on the 1:6
measurements to provide valuable information and criteria for the design of VDSFs, from an ecohydraulic
perspective.
To extrapolate criteria design for future fishways, the search of corrective factors will be pursued to improve
slightly the laboratory findings when switched back to 1:1 scale. It will focus on minimising the difference
between full-scale and laboratory turbulence and unsteady measurements, using least square approach.
5. ACKNOWLEDGEMENTS
This work was funded by the French Biodiversity Agency (OFB). The authors acknowledge the financial
support of the CPER-FEDER of the Nouvelle-Aquitaine region for the environmental hydrodynamic platform
(pHE). The authors would also like to acknowledge the French Electricity Company (EDF) for allowing the use
of their fishway in by-pass of the Malause dam, on the Garonne River and MIGADO association for the access
to the video-counting room.
6. REFERENCES
Ballu, A., Pineau, G., Calluaud, D., & David, L. (2016). Influence of the presence of sills on the behavior of
brown trouts (Salmo trutta) in an experimental vertical slot fishway. International Conference on
Engineering and Ecohydrology for Fish Passage, 2016.
Ballu, A., Pineau, G., Calluaud, D., & David, L. (2017). Characterization of the flow in a Vertical Slot Fishway
with macro-roughnesses using unsteady (URANS and LES) simulations. 37th IAHR World Congress.
Clay, C. H. (1995). Design of fishways and other fish facilities - 2nd edition. In Design of Fishways and Other
Fish Facilities. https://doi.org/10.1201/9781315141046
Fujihara, M., Fukushima, T., & Tachibana, K. (2003). Numerical Investigations of Flow in Vertical Single- and
Double-slot Fishways. Transactions of Irrigation, Drainage and Reclamation Engineering, 223, 79–88.
Goring, D. G., & Nikora, V. I. (2002). Despiking Acoustic Doppler Velocimeter Data. Journal of Hydraulic
Engineering, 128(1), 117–126. https://doi.org/10.1061/(asce)0733-9429(2002)128:1(117)
Klein, J., & Oertel, M. (2016). Vertical Slot Fishway : Evaluation of numerical model quality. International
Junior Researcher and Engineer Workshop on Hydraulic Structures. https://doi.org/10.15142/T3R599
Lejeune, A.-F., Calluaud, D., Pineau, G., David, L., Sagnes, P., & Richard, S. (2022). Experimental study of
turbulent flow in a vertical double slot fishway: influence of flow discharge and slope. 14th ISE 2022
Proceedings.
Marriner, B. A., Baki, A. B. M., Zhu, D. Z., Thiem, J. D., Cooke, S. J., & Katopodis, C. (2014). Field and
numerical assessment of turning pool hydraulics in a vertical slot fishway. Ecological Engineering, 63,
88–101. https://doi.org/10.1016/j.ecoleng.2013.12.010
Puertas, J., Pena, L., & Teijeiro, T. (2004). Experimental Approach to the Hydraulics of Vertical Slot Fishways.
Journal of Hydraulic Engineering, 130(1), 10–23. https://doi.org/10.1061/(asce)0733-
9429(2004)130:1(10)
Rajaratnam, N., der Vinne, G. Van, & Katopodis, C. (1986). Hydraulics of Vertical Slot Fishways. Journal of
Hydraulic Engineering, 112(10), 909–927. https://doi.org/10.1061/(asce)0733-9429(1986)112:10(909)
Zhang, D., Bian, X., Shi, X., Deng, J., & Liu, Y. (2023). Design of a bilateralsymmetric multislot fishway and
its comparison with vertical slot fishway in terms of hydraulic properties. River Research and
Applications, March. https://doi.org/10.1002/rra.4125
Zhao, H., Xu, Y., Lu, Y., Lu, S., Dai, J., & Meng, D. (2022). Numerical Study of Vertical Slot Fishway Flow with
Supplementary Cylinders. Water, 14(11), 1–16. https://doi.org/10.3390/w14111772