1. Introduction

In recent years, numerical models of flow in water bodies are widely used to provide an accurate description of flow velocity distribution. Although all the hydrodynamic processes going on in an aquatic system are difficult to describe, some authors have reviewed the main mechanisms, from large scale [1–5] to small scale and mixing processes [6, 7] or both [8, 9]. Three-dimensional numerical models of flow calculations in natural water bodies have been successfully applied for rivers with single complex flow features [10–14] although there is a little experience with modeling of flow in large dam reservoirs [15]. Studies of flow through curved or straight, open channels of simple cross-sectional shape (rectangular or trapezoidal) have been reported by Demuren [16] and Meselhe [17]. Both authors presented a 3-D

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited. © 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

numerical model, for the calculation of turbulent flow in meandering channels. The works provided a hydrodynamic basis for the study of the mechanisms for the formation of river meanders. The authors employed a finite volume numerical method to solve the full Reynoldsaveraged Navier-Stokes (RANS) equations in conjunction with the standard k-ɛ model. The case exhibits some of the features encountered in real rivers, including longitudinal curvature and varying bed topography. Demuren's results were in reasonable overall agreement with the mean velocity measurements of Almquist and Holley [18]. Meselhe focused on proposing a simplified approach for calculating the water surface elevation as a part of the overall solution procedure.

In a research conducted by Khosronejad [19], a 3-D CFD model was applied to predict the flow hydrodynamics around power intakes within the Dez dam reservoir (Iran). The study was also devoted to a qualitative analysis of sediment transport at the area around the intakes. For incorporating the effects of turbulent flow, the k-ω model was implemented. The finite volume method was used to discretize the RANS equations. The 3-D single-phase model ran for a limited 320 m long reservoir section. The results show that the flow velocity has a maximum value of 1–2 m/s near power intakes and decreases with distance. In addition, the turbulent intensity increases in the area near intake entrance resulting in increasing bed shear stress near intakes. Khosronejad did not present a test of either coarser or finer mesh resolution; moreover,

Three-Dimensional CFD Simulations of Hydrodynamics for the Lowland Dam Reservoir

http://dx.doi.org/10.5772/intechopen.80377

41

Sulejow reservoir is situated on the middle reach of the Pilica river, which is left-hand side tributary of Vistula river, central Poland (Figure 1). One of the biggest artificial reservoirs in

the model was not verified in the field measurements.

Figure 1. Localization of the Sulejow dam reservoir.

2. Study area

All previous studies with both straight and curved channels had adopted a rigid-lid assumption for modeling the free surface. The rigid-lid approximation is a commonly used simplification in the study of density-stratified fluids mainly in oceanography. Assumes, that the displacements of the surface are negligible compared with interface displacements. In a research conducted by Sinha [11], a 3-D model of flow through a 4-km stretch of the Columbia river, downstream of the Wanapum Dam, was developed and validated. Authors succeeded in modeling of flow in rapidly varying bed topography and the presence of islands. In the study three-pronged strategy—composed of the field measurements, the laboratory experiments, and the numerical model—was undertaken. The model solves the (RANS) equations closed with the standard k-ɛ turbulence model.

In today's development level of hardware and software, increasingly complex, engineering tasks can be managed, and, thus, three-dimensional hydrodynamic modeling of dam reservoirs became a potential tool for determining more accurate picture of the dependencies, which prevail in the aquatic ecosystem. Whereas in most studies one- and two-dimensional description of flows may be sufficient, in some special cases, three-dimensional approach is needed to determining, for instance, flow patterns in bends or in the vicinity of hydraulic structures (dams and weirs). These issues are evidently three-dimensional, and spatial character directly affects pollutant transport processes.

Hydrodynamic processes determine the movement of suspended and dissolved matter; heat transfer; intensity of circulation inside the ecosystem; the speed of contaminating processes; and the self-purification of the reservoirs, ultimately provide conditions of the ecosystem function.

Kennedy et al. [15] applied a 3-D model to estimate the hydraulic residence time (HRT) for the Thomas Basin (part of the Wachusett Reservoir in central Massachusetts). The basin was modeled using the FLUENT software package with particles used to track travel time in steady-state conditions. A tetrahedral mesh was used with accurate description of basin bathymetry. The model solved the transient Reynolds equations for turbulent flow with standard k-ɛ closure. Modeling was performed to simulate flow pattern during a period when conditions were isothermal and windless. HRT was estimated to 3–4 days which is about half of the HRT that would be expected based on the theoretical mean residence time. The results of the calculations show that the presence of a primary flow path, large-scale eddies, and stagnation zones contributed to the faster travel time. Reductions in inflow rates produce increased residence times and significant changes in flow patterns. However, the authors did not provide any information about model verification.

In a research conducted by Khosronejad [19], a 3-D CFD model was applied to predict the flow hydrodynamics around power intakes within the Dez dam reservoir (Iran). The study was also devoted to a qualitative analysis of sediment transport at the area around the intakes. For incorporating the effects of turbulent flow, the k-ω model was implemented. The finite volume method was used to discretize the RANS equations. The 3-D single-phase model ran for a limited 320 m long reservoir section. The results show that the flow velocity has a maximum value of 1–2 m/s near power intakes and decreases with distance. In addition, the turbulent intensity increases in the area near intake entrance resulting in increasing bed shear stress near intakes. Khosronejad did not present a test of either coarser or finer mesh resolution; moreover, the model was not verified in the field measurements.
