A case study of a pool and weir fishway modeled with OpenFOAM and FLOW-3D

https://doi.org/10.1016/j.ecoleng.2017.01.042Get rights and content

Abstract

OpenFOAM's interFoam solver is validated for modeling the complex three dimensional flow field of a large pool and weir fishway. Numerical results from OpenFOAM and the commercial solver FLOW-3D are compared and validated with in situ experimental velocity, turbulence and water surface profiles. OpenFOAM and FLOW-3D produced similar flow fields. The models accurately predicted velocity magnitudes throughout the pool and the barrier velocity immediately downstream of the notch. The models over-predicted the width and depth of the jet issuing from the notch and significantly under-predicted turbulent kinetic energy levels within it. In both models a large roller occupies a generous volume of the sampled pool, the presence of which is not supported by experimental measurements. The findings demonstrate that interFoam coupled with the standard k  ϵ turbulence model is able to estimate barrier velocities, flow rates, turbulence magnitudes and depths to accuracies acceptable for fishway design evaluation.

Introduction

In recent years, computational fluid dynamics (CFD) has seen increased use in fish passage and related free surface ecohydraulic engineering applications. Numerous researchers have employed 2D and 3D models to gain insights on the flow fields of a variety of fishway types (Cea et al., 2007, Barton et al., 2009, Chorda et al., 2010). Many such studies used commercial CFD codes to predict important fish passage design criteria. Commercial codes are popular for their user-friendly working environment, accuracy and thoroughly validated code. Yet, these advantages often come with substantial licensing fees, scalability limitations and restricted access to source code. OpenFOAM's two phase solver, interFoam, has achieved success modeling complex flows such as hydraulic jumps, sewage collection systems, free surface flows around submerged hydrofoils and determining wave forces on coastal bridge decks (Hayatdavoodi et al., 2014, Seiffert et al., 2014, Bayon-Barrachina and Lopez-Jimenez, 2015a, Bayon-Barrachina and Lopez-Jimenez, 2015b, Prasad et al., 2015). Results from these studies demonstrate the potential of OpenFOAM to predict fishway design parameters of interest such as barrier velocities, pool depths, turbulence metrics and evaluating flow patterns for availability of velocity refuge for migrating fish. Released under the general public license (GNU), OpenFOAM is free and open-source. Liberated from licensing restrictions, parallel scalability is limited only by available hardware, allowing practitioners to more efficiently allocate computational resources to resolve CFD problems.

The choice of using either a 2D or 3D model is usually based on the dimensionality of the expected flow field of the fishway. 3D CFD can require considerably longer simulation times than 2D or 1D CFD methods, yet can provide enhanced details of the flow field not available from lower dimension models. For vertical slot fishways, studies have reported accurate results using 2D models (Cea et al., 2007, Chorda et al., 2010, Bombač et al., 2015) this success is attributed to the characteristic 2D flow defining vertical-slot fishways (Wu et al., 1999, Tarrade et al., 2011). 2D depth-averaged models of nature-like fishways, characterized by numerous surface protruding roughness elements, were shown to acceptably model depths, velocities and turbulence levels (Tran et al., 2016). Tran et al. (2016), nevertheless proposed that 2D simulations of nature-like fishway be performed as precursors to identify promising designs before performing 3D simulations. However, in certain fishways, where flow is expected to be highly 3D in character, practitioners may have no choice other than perform a 3D model.

In addition to resolving the lateral and longitudinal components of the flow, fully 3D models resolve the vertical flow component responsible for many complexities inherent to fishway designs. Many researches have successfully applied 3D models to vertical-slot fishways (Khan, 2006, Barton et al., 2009, Fu et al., 2013, Marriner et al., 2014, Marriner et al., 2016). Notably, Marriner et al. (2014) showed considerable agreement between 3D CFD depth and velocity results and experimental data in a vertical-slot fishway turning pool. Marriner et al. (2016) used 3D CFD to evaluate variations in energy expenditures for salmon along potential trajectories within a number of vertical-slot fishway design iterations. The apparent success of 3D modeling of vertical-slot fishways highlights its potential for evaluating pool and weir fishway designs.

Within a typical pool of a pool and weir fishway operating under a plunging flow regime, the flow over the notch in the upstream weir plunges into the pool and surges to the surface after impacting the backside of the downstream weir (Ead et al., 2004). The plunging portion of the flow, or jet, is a critical design feature since its velocity determines the barrier velocity for migrating fish. 3D modeling has the potential to resolve this complex 3D flow feature which is largely responsible for initiating the dominant flow patterns within the pool. However, to the authors’ knowledge, no studies in the published literature have employed 3D CFD to model pool and weir fishways. Since relatively few studies have investigated the use of OpenFOAM for free surface engineering problems and none to the authors’ knowledge have been on technical fishway, a study of the interFoam solver for application to pool and weir fishways would prove insightful.

In this work, a full scale in situ pool and weir fishway is studied both experimentally with 3-component acoustic Doppler velocimetry and numerically using OpenFOAM's interFoam solver and the commercial solver FLOW-3D. The objectives are to (1) compare OpenFOAM with FLOW-3D, (2) perform a Grid Convergence Index (GCI) analysis on both OpenFOAM and FLOW-3D, (3) validate OpenFOAM with experimental velocity, turbulence and water surface level data, and (4) discuss the strengths and weaknesses of the modeling approaches used for the analysis of pool and weir fishways. The findings and modeling methods proposed in this study will be of interest to practitioners applying CFD to assess fishway designs.

Section snippets

The fishway

The fishway under study is an aging pool and weir fishway designed to provide upstream passage of American shad Alosa sapidissima. The fishway is situated adjacent to a large hydroelectric production facility in Eastern Canada. The elevation drop between the reservoir and tail-race water surface levels is 7.8 m. The fishway consists of concrete floors and walls and 32 wooden weirs spaced between 2.5 and 5 m apart. Fig. 1a depicts a plan view of the modeled geometry, aerial front view (Fig. 1b)

Numerical models

When turbulence is simulated for Newtonian flows, both interFoam and FLOW-3D solve the continuity (Eq. (1)) and the unsteady Reynolds-averaged Navier–Stokes equations governing fluid motion (Eq. (2)) using a finite volume method (FLOW-3D, 2012, Moukalled et al., 2015). In Eq. (2), p indicates pressure, v denotes the velocity field, time by t, fb indicates the body forces acting on the control volume and τ denotes shear stresses. Over-bars indicate time averaging and ′ the instantaneous

Study regions for spatial convergence and model comparisons

The spatial convergence and comparative studies of the models were performed in two regions within pool 16. Region 1 (red in Fig. 4) was chosen since it encompasses the high velocity jet issuing from the weir and accompanying high shear zones, whereas region 2 (blue) was chosen as a representative central region of the pool. Zones of high shear show the greatest sensitivity to grid refinement (Hardy et al., 2003, Biron et al., 2007). Both regions consist of 1638 points separated by 0.10 m within

Model comparison and spatial convergence

Earlier it was shown that the solution on the finest mesh of FLOW-3D (F3D-m1) diverged substantially from those of the coarser meshes (F3D-m2, F3D-m3) and the solutions of OpenFOAM. Most notably, the expected recirculatory region under the jet issuing through the notch is not present in the F3D-m1 solution. Experimental work by Ead et al. (2004) has characterized the hydraulics and confirmed the presence of the recirculatory region under the jet in a similar pool and weir fishway operating

Conclusions

The results of this study demonstrate that OpenFOAM's interFoam solver performs well at reproducing measured velocity, turbulence and water surface levels within a pool and weir fishway. A grid convergence study is performed (grid convergence index) and indicates excellent spatial convergence of the solver. Predicted water surface elevations were an average 7% higher than measured profiles along the highly turbulent and air entrained south wall of the pool. Along the tranquil north wall, CFD

Acknowledgements

The authors extend their gratitude to the Natural Science and Engineering Research Council of Canada for providing funding to pursue this research. We are also appreciative of both Maryse Page, Eric Mainville and Jean Caumartin at Hydro-Québec who provided invaluable assistance throughout the course of the project and to Nicolas Simard for his technical assistance with the construction of the ADV traverse.

References (44)

  • A. Bayon-Barrachina et al.

    Numerical analysis and validation of south Valencia sewage collection system diversion

    E-Proceedings of the 36th IAHR World Congress, vol. 17 (no. 4)

    (2015)
  • A. Bayon-Barrachina et al.

    Numerical analysis of hydraulic jumps using OpenFOAM

    J. Hydroinformatics

    (2015)
  • P.M. Biron et al.

    Assessing different methods of generating a three-dimensional numerical model mesh for a complex stream bed topography

    Int. J. Comput. Fluid Dyn.

    (2007)
  • T. Castro-Santos

    Optimal swim speeds for traversing velocity barriers: an analysis of volitional high-speed swimming behavior of migratory fishes

    J. Exp. Biol.

    (2005)
  • T. Castro-Santos

    Modeling the effect of varying swim speeds on fish passage through velocity barriers

    Trans. Am. Fish. Soc.

    (2006)
  • L. Cea et al.

    Application of several depth-averaged turbulence models to simulate flow in vertical slot fishways

    J. Hydraul. Eng.

    (2007)
  • I.B. Celik et al.

    Procedure for estimation and reporting of uncertainty due to discretization in CFD applications

    J. Fluids Eng. Trans. ASME

    (2008)
  • J. Chorda et al.

    Two-dimensional free surface flow numerical model for vertical slot fishways

    J. Hydraul. Res.

    (2010)
  • A. Cotel et al.

    Do brown trout choose locations with reduced turbulence?

    Trans. Am. Fish. Soc.

    (2006)
  • S.S. Deshpande et al.

    Evaluating the performance of the two-phase flow solver interFoam

    Comput. Sci. Discov.

    (2012)
  • S. Ead et al.

    Flow regimes and structure in pool and weir fishways

    J. Env. Eng. Sci.

    (2004)
  • E. Enders et al.

    The effect of turbulence on the cost of swimming for juvenile Atlantic salmon (Salmo salar)

    Can. J. Fish. Aquat. Sci.

    (2003)
  • Cited by (0)

    View full text