Journal of Hydraulic Engineering

Urban water distribution systems (WDSs) are often skeletonized to enable efficient system analysis and management. While different methods are available to account for transient dynamics within the skeletonization processes, they often ignore the potential impacts induced by nodal demand allocations. This paper proposes a transient-based method to skeletonize pipes in series with internal demands, where the optimal demand allocation strategy is determined by a minimization approach associated with a probabilistic evaluation method. In addition, this paper makes the first attempt to investigate the impacts of different nodal demand allocation strategies on reproducing transient dynamics within the skeletonization process. The proposed method is demonstrated for a hypothetical transmission system and a realistic WDS. Results show that the impact of demand allocations on transient dynamics is positively correlated with demand values and that the proposed skeletonization method overall outperforms the traditional methods in capturing the transient dynamics of the original WDS, especially for nodes with relatively large demands.

Unimpeded waterway connectivity is a requirement for all freshwater fish. While box culverts are considered the most effective design in terms of upstream fish passage, circular culverts are very common. Detailed hydrodynamic measurements were undertaken under controlled conditions in a near-full-scale smooth pipe culvert operating at less-than-design flows. Two configurations were tested: a smooth semicircular channel and a circular channel equipped with a small streamwise rib placed asymmetrically. For all investigated flow conditions, the channel flow was subcritical and corresponded to less-than-design conditions. Detailed measurements showed high velocities through the entire cross-section, with no obvious low-velocity region along the smooth wetted perimeter. The presence of an asymmetrical streamwise rib induced the formation of a small well-defined low-velocity zone (LVZ) in the vicinity of the rib. The flow resistance was slightly larger than that in a rectangular channel, for identical boundary roughness and flow conditions. The streamwise rib had a limited impact on the flow resistance, although large transverse gradient in skin friction shear stress, conducive of secondary currents, were observed. While the low-velocity zone size relevant to the upstream passage of small fish was smaller in the smooth circular channel than in a comparable rectangular channel, the introduction of the asymmetrical streamwise rib might create preferential swimming paths for small-bodied fish and juveniles of larger fish. The present physical results may serve as a validation data set for future computational fluid dynamics (CFD) modeling, to assist with the development of more efficient designs.

This paper presents results obtained from a detailed analysis of the experimental data of three hydraulic jumps with low Froude numbers (1.5–1.9) in a narrow flume. With the increasing Froude number, the three generated jumps exhibit an initial increasing then decreasing undular behavior. The turbulent velocity fields are measured with a high spatial and temporal resolution using the particle image velocimetry (PIV) technique. Based on careful evaluations of the valid velocity vectors from the PIV measurements, it is possible to determine the free surface profile of the undular jump. The undular flow patterns are visualized by velocity redistribution between successive crests and troughs, which is further interpreted in terms of the pressure variation according to the curvatures of the undular streamlines. An examination of the momentum flux shows that the velocity and pressure contributions are basically comparable within the subcritical flow region, and the total momentum is generally constant throughout the entire jump except at the transition region immediately downstream of the jump toe. The distributions of the turbulent kinetic energy (TKE) and TKE budget terms are analyzed at characteristic longitudinal locations. The results reveal that the significant turbulent quantities, which are presumably related to the undular flow motion and surface oscillation, are mainly concentrated around the surface. The decomposed turbulent intensities present an anisotropic feature near the jump toe that gradually becomes isotropic with downstream diffusion. The detailed analysis of the isolated contributions of production, advection, diffusion, and dissipation provides new insights into the TKE budget balance around the undular free surface region, which may serve as a reference for future computational model studies.

An experimental nonintrusive methodology is proposed to estimate the spatially averaged erosion rate of fine sediment from a gravel bed. The experiments performed in a laboratory flume show a progressive slowdown of the erosion rate as the level of fine sediment becomes deeper within the gravel matrix, until a maximum depth of erosion is reached. Two original relations are proposed for the maximum cleanout depth and the erosion rate based on a dimensional analysis applied to the experimental results. The proposed erosion rate relation modifies the original Van Rijn formula for a uniform bed, introducing a damping function below the gravel crest. Both the evolution of the erosion rate with depth and the maximum depth of erosion can be defined as simple functions of the general characteristics of the flow and the fine and coarse fractions of the sediment. The proposed approach will lead to improved estimates of the conditions under which fine sediments that have infilled gravel beds are re-entrained. This will help inform strategies aimed at restoring degraded river systems and mitigating the undesired side effects of activities such as sediment flushing which can result in colmation.

The open channel flow community has a history of using a set of universal exponential expressions to describe the vertical variation of turbulence intensity in open channel flows. All of these relationships contain an empirical coefficient that varies with the component direction ($D_u=2.3$, $D_v=1.27$, $D_w=1.63$). Recent laboratory experiments have demonstrated that for smooth and rough bed open channel flows, these constants are not universal but vary with the Reynolds number, bed composition, and in the presence of streamwise counter-rotating vortices and secondary flows. A robust method for the remote prediction of $D_u$, and hence the vertical variation of the streamwise turbulence intensity is developed here for a wide range of flow conditions. The method uses the free-surface turbulence anisotropy as a proxy for the redistribution of vertical velocity fluctuations into surface-parallel components, enabling a correlation between the free-surface turbulence intensity and $D_u$. Implications for the remote prediction of bed composition, sediment transport, and bed shear stress are discussed.

The partial-flip bucket is successfully applied to a spillway outlet design with an approximately $50\mathrm{m}/\mathrm{s}$ flow velocity in Jinping-I Dam in Sichuan, China. The specific design of a part-flipped area on a floor results in a long and narrow jet of high-velocity flow released from the flip bucket. The reductions in the pressure on the side wall and the flow discharge impact on the riverbed downstream confirm that the energy dissipation issue is effectively solved in high-head dams. However, detailed information on the transverse movement of flow in the partial-flip bucket is lacking. The present study uses numerical simulations to analyze interior flow structures, including pressure and transverse velocity distributions. The effect of transverse pressure differences on the water flow contraction is investigated. Transverse pressure differences between the sidewall and interior water are accompanied by sudden pressure release in the unflipped area; differences are more pronounced around the incipient cross sections of the unflipped area and decrease as the water flows over the downstream flip bucket. This pressure difference results in transverse velocities, leading to the transverse contraction of the water flow. Compared to the stalling conditions of the traditional deflector flip bucket, the partial-flip bucket can extend the practical applicability of flow discharge operations.

We perform an analysis of the response of a reach of the Minnesota River, US to changes in sediment loading, including sand and mud. The reach in question extends 160 km downchannel from Mankato to the confluence with the Mississippi River. To develop a morphodynamic model of this reach, we consider a one-dimensional formulation of coupled flow, sediment transport, and channel bed/floodplain morphodynamics and derive model inputs from field parameters where possible. We show that the output of wash load (mud) is about four times that of bed material (sand) and that changes in bed material input have little effect on sediment output over 600 years. That is, sand input is mostly sequestered in the study reach. However, changes in wash load input have a near-immediate effect on sediment output. Thus, reducing the input of wash load would have a greater impact on sediment delivery to the Mississippi River than reducing the input of bed material load.

The authors developed a Hydrologic Engineering Center–River Analysis System (HEC–RAS) model to serve as the key component of a new, first-of-its-kind, short-term operational flow forecasting system for the Niagara River. The Niagara River transports a continental-scale flow (with an annual mean of roughly $\mathrm{6,300}{\mathrm{m}}^3/\mathrm{s}$) that supports the economy of both the United States and Canada through hydropower generation, tourism, and other activities. The river also serves as a link connecting the two most downstream lakes (Lakes Erie and Ontario) in the largest system of lakes on Earth. Despite its significance, the authors know of no federally operated, short-term forecasting system for the Niagara River. Hydropower facilities management and other water resources management activities on the river have historically relied on an array of experimental, in-house, or proprietary models to simulate and forecast Niagara River flows. The study presented here fills this gap in large-scale hydraulic modeling and engineering science by calibrating a HEC–RAS model for the Upper Niagara River and customizing it to meet the operational requirements of the National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) Northeast River Forecasting Center (NERFC). The skill of the new forecasting system, which was recently deployed in its operational environment at the NERFC, will depend in large part on the accuracy of meteorological boundary conditions. The authors envision a more comprehensive assessment of the system’s forecasting skill and other potential future model improvements as an area for future research.

Gas transfer in dam spillways often leads to supersaturation of total dissolved gases (TDGs) that can cause fish mortality. Quantifying TDG associated with this process is crucial in the development of operating alternatives to minimize environmental risks to downstream aquatic habitats. In this study, the transfer of TDG in a spillway was evaluated through field observations at the Seven Mile Dam on the Pend d’Oreille River in British Columbia, Canada. The dam degasses high-TDG water due to a ski-jump design where aerated flows are generated from a flip bucket at the end of the spillway chute. A simplified mathematical formulation, incorporating physical processes related to air entrainment, bubble characteristics, and mass transfer across free surface and bubbles, was tested and verified for prototype flows to partition gas transfer in the spillway face, free jet, and plunge pool supported by extensive field measurements. Due to gas exchange dominated by bubble-mediated transfer, substantial degassing of TDG was observed during spill operations, with degassing in the free jet being considerably higher compared to the spillway face and plunge pool. The gas transfer efficiency was high when pre-aeration (air entrainment on the spillway face) occurred. Practical relationships were proposed to estimate degassing in the jet and assess overall gas transfer efficiency. Results from this study can help inform water management decisions during periods of elevated TDG, particularly in cascading hydropower systems.

Overshot water wheels are hydropower converters generally employed for head differences up to 6 m and maximum flow rates of $150-200\mathrm{L}/\mathrm{s}$ per meter width. The maximum hydraulic efficiency (80%–85%) is constant for rotational speeds below the critical speed, whereas the efficiency linearly decreases at higher rotational speeds due to the increase of water losses at the inflow. To improve the efficiency when the rotational speed is above the critical speed, an improved geometric design was investigated by implementing a theoretical model validated using experimental results. The new geometry consists of a circular wall around the periphery of overshot water wheels. The wall redirects into the buckets the water flow that is lost at the inflow, improving the efficiency up to 1.5 times at high rotational speeds.

The pickup probability of bedload sediment can be defined as the channel width fraction that is only associated with entrained particles. With this definition, the dimensionless bedload transport rate can be formulated as the product of the pickup probability and the conditionally averaged Einstein number. The bedload formula so obtained appears to be the simplest in comparison with the previous bedload functions derived with different considerations. Its predictions agree well with the published experimental data for the condition of plane-bed of uniform sediment grains.