World Environmental and Water Resources Congress 2020: Hydraulics, Waterways, and Water Distribution Systems Analysis

Three-dimensional (3D) computational fluid dynamic (CFD) simulation is gaining popularity in recent years for stream flow modelling. It is necessary when local flow patterns are of interest and/or there exist in-stream structures. Representation of complex terrain, however, is a major obstacle in 3D CFD modelling. Traditionally, the terrain-conforming method is widely used in which terrains are accurately represented by a 3D mesh; i.e., the mesh conforms to the terrain geometry. This method is straightforward in implementation and accurate in resolving the near-terrain flows. A drawback is that such a mesh is difficult to generate when the terrains are complex. The mesh quality may become too poor to maintain solution stability and accuracy. An alternative is the terrain-embedding method with which terrains are embedded in a background mesh. The background mesh may be generated without the requirement of conforming to the terrain so that mesh generation is relatively simple and good mesh quality may be maintained. Special algorithm, however, is needed to take into account the effect of the embedded terrains on the nearby flow. In this study, both CFD methods are adopted to simulate a selected flow case. The objective is to understand the pros and cons of the two methods through a real laboratory case, not merely in theory. The study is to pave a way to simulate sediment transport and scour cases when stream bed is moving.

Relationships to predict the air demand of baffle drop structures are presented herein. The associated data for the methodology were developed from physical model studies. Illustrative calculations and guidance for mitigating undesirable air transport to connected tunnels are also presented.

In 1967 Engelund and Hansen proposed a sediment energy equation based in part on the characteristics of the sediment waveform. In this same monograph, based on the compilation of flume data they showed that the ratio of flow resistance to sediment waveform was a constant. This permitted a substitution of flow resistance for waveform—resulting in the widely used Engelund-Hansen sediment transport equation. The difficulty with the E-H equation is that flow resistance remains a function of waveform, which in most cases cannot be directly measured. In this paper, use of the E-H ratio (dimensionless shear stress to the 5/2 power divided by the dimensionless sediment transport rate) is evaluated further. It is shown that for bedforms in the dune range that the non-grain roughness component of alluvial roughness is a semi-logarithmic function of the E-H ratio. Beyond the dune range it is shown that non-grain roughness is a function of base dune roughness and a shape factor that is a linear function of Froude number. Utilization of the E-H ratio therefore provides for direct estimates of several important alluvial hydraulic properties including: flow resistance, sediment waveform characteristics, and sediment transport. The approach is valid for dune, transitional, and upper regime bedforms up to a Froude number of 0.8.

An experimental study was conducted to investigate effect of adding boulders with different boulder spacing (8D, 6D, and 4D) on mean and turbulent flow characteristics. Measurement points were selected at centerline and side of the flume. Mean flow depth did not vary significantly in different scenarios but mean streamwise velocity decreased at centerline after adding boulders especially at shorter spacing. It appeared flow regime for all of the scenarios is isolated-roughness flow. For centerline points, root-mean squared velocity, turbulent kinetic energy, and turbulence intensity increased after adding boulders and by decreasing boulder spacing however, integral scale parameters decreased after adding boulders and did not change significantly by decreasing boulder spacing. Also, relationships between time-averaged streamwise velocity and normalized turbulence intensity was found for each scenario.

Scour processes downstream of hydraulic structures are some of the more challenging problems for hydraulic engineers because of the potential for scour to cause incidents and failures with catastrophic consequences to society and the surrounding environment. In addition, urbanization and climate changes frequently cause flow capacity deficiencies for various hydraulic structures to safely pass extreme flood events and meet current regulations. In this regard, PK weirs represent an innovative control structure that can be successfully used in new and rehabilitation projects to fulfill this trend of increasing discharge capacity demand. Nevertheless, the effects of an increased discharge on local scour are still under-explored. In order to fulfil this gap of knowledge, a dedicated laboratory model was built and experimental tests were conducted to analyse the scour mechanism downstream of a PK weir structure. One uniform noncohesive granular bed material was tested under different combinations of discharge and tailwater depth. This paper presents a preliminary analysis of the scour morphology downstream of a PK weir, focusing on the effects of different tested variables at equilibrium conditions.

Herein are described methods and results of analysis of impacts on biological resources in the Upper Arroyo Seco that may result from storm flow diversions associated with the implementation of the proposed Pasadena Water and Power (PWP) Arroyo Seco Canyon Project (ASCP). The ASCP would repair, replace, and/or rehabilitate the diversion and intake structures in the Arroyo to maximize the capture of PWP’s surface water rights. Two-dimensional hydraulic models were prepared of the Devil’s Gate Reservoir area for representative wet, average, and dry years for the existing and with-diversion scenarios. Hydraulic modeling indicates that only during dry years do with-diversion conditions have expected changes to maximum reservoir pool depth. Within the stream section of the study area modeling showed little or no measurable change in hydraulic conditions (i.e. depth, velocity, inundation area, and shear) with project diversions. Additional sediment transport modeling and groundwater analysis indicated no impacts from proposed conditions. An ecological modeling study was prepared for the study area. The ecological effects of flow diversions were assessed based on modeled relationships between proposed hydraulic changes and the probability that the resultant physical conditions can support southern willow riparian habitat conditions conducive to occupancy by least Bell’s vireo. A statistical approach, rather than a mechanistic approach, was used to estimate the probability of occurrence of vireo under physical conditions associated with flow diversion scenarios. All hydraulic variables tested showed significant statistical differences (p<0.001) in willow occurrence between predicted future conditions and historic flow characteristics. Given the small sample size of vegetation polygons and the high number of hydraulic observations, it is believed the significance estimates may be overestimated. The models predict a slight decrease in probability of occurrence of willow habitat in dry years in the proposed condition. However, net changes in probability are less than 10%.

A stilling basin, is often optimized in dimensions based on the characteristics of hydraulic jumps formed within the basin. Many types of energy dissipators are used to shorten the jump-length and hence, reducing the length of basin. A study was conducted to investigate the effectiveness of six-legged concrete (SLC) elements as dissipators. Models of these elements were placed on the bed of an experimental flume downstream of a spillway chute in different densities of the SLC elements. The incoming flow had Froude number in the range of 5.3 to 8.1. The results indicate that the greatest reduction of sequent depth ratio and dimensionless length of jump occur when the density of elements is 36%. The average reduction of hydraulic jump length on the roughened bed for densities of 36%, 63%, and 100% is 29% and at 23% and 17%, respectively, in comparison with the smooth bed.

Scour at bridge piers is one of the most studied and challenging problems for hydraulic engineers. Many researchers analyzed such scour phenomenon, focusing on the effect of hydraulic conditions and geometric configurations. In particular, previous analyses focused on the scour evolution and equilibrium configuration under both clear- and live-bed conditions, providing useful design criteria, accounting for the effect of hydraulic conditions and pier geometry. On the contrary, only a few studies deal with the effect of debris accumulation on scour evolution and equilibrium configuration. But, to the best of authors’ knowledge, none of them analyzed the effect of debris accumulation for an isolated pier located at different transversal positions with respect to the channel bank. In order to fill this gap of knowledge, we conducted experimental tests by varying the transverse position of the pier within the channel. Furthermore, we tested different debris accumulations varying the percentage blockage ratio. Experimental results revealed that scour equilibrium configurations are affected by the tested parameters. Therefore, the aim of this paper is to present a preliminary analysis of the scour hole features, highlighting similitudes and differences in terms of scour characteristics caused by the pier location in the presence of debris accumulation.

The introduction of spur dikes into a river flow field can help to improve the stability of a river channel. While the flow velocity directly at the spur dike tip or crest may increase, a large embayment area downstream of the spur dike will form with reduced flow velocities. Due to the increase in flow velocity at the spur dike, a scour hole will form. Many hydraulic structures are tested by building scaled models, which is very costly and lacks versatility to model all factors correctly. The use of two-dimensional numerical models allows for increased efficiency and accuracy of hydraulic modeling. This allows for enhancing relationships between dynamic variables and the estimated scour depth. The length of the spur dike and the flow rate were varied in the experiments. It was found that as the length of a spur dike increases, the depth of the scour also increases. This held a stronger correlation than the increase in flow rate. A relationship was developed between the maximum flow velocity, the upstream flow velocity, and the upstream Froude number to determine the maximum scour depth. This relationship proved to be more accurate than past relationships proposed using data from physical model analysis. The new relationship lowered the percentage-error from 14% to 1% when the predicted scour depths were compared with the measured scour depths. The error was reduced from 7.3% to 1.6% for the long spur dike simulations and from 21.4% to 13.2% for the short spur dike simulations.

The present numerical study attempts to estimate the equivalent Manning’s roughness in combining open channel junction flows. The three-dimensional flow field in a combining open channel junction is simulated using the CFD software, Fluent 17.2. Simulated results for water surface elevation and velocity distribution are validated against the experimental results available in literature for junction flow over smooth bed. The numerical model is used to simulate nine different flow conditions using three different discharge ratios (ratio of main channel discharge to total discharge) and three bed roughness heights. The head loss is calculated considering the energies at three sections (two at upstream of the junction and one at outlet). An equivalent Manning’s roughness is obtained from the head loss. For discharge ratio of 0.75, values of equivalent Manning’s roughness for the channel reach, starting from the location where junction starts and up to the channel outlet, are found to be 0.012, 0.016, and 0.018 for bed roughness height = 0 mm, 2 mm, and 5 mm, respectively. For discharge ratio of 0.25, values of equivalent Manning’s roughness are found to be 0.016, 0.020, and 0.023 for bed roughness height = 0 mm, 2 mm, and 5 mm, respectively. The results show that the equivalent Manning’s roughness in a channel junction flow is a function of the discharge ratio and bed roughness. The effect of bed roughness on water surface profile and bed shear stress is also presented. The present study will be useful for analyzing the flow resistance in channel junction in a simplified way.

Small streams and local waterways tend to flood on a more frequent basis than larger rivers. The flood plains associated with these small streams and waterways provide storage for excess stormwater runoff for a range of storm events from brief, intense summer showers to multiday storm events. These flood plains are responsible for lessening the impact of regular storm events on the surrounding properties. The loss of flood plains can contribute significantly to the degradation of small streams and local waterways as well as significantly impacting the surrounding properties. This study examines the historic loss of flood plains in a small watershed. The results of the study showed that there has been a substantial loss of flood plain area in the upper Turtle Creek watershed from historic development practices. The development practices not only included residential subdivisions and commercial sites but other land disturbances such as highway construction, railroad construction, and farming. This information can serve as a basis to develop future guidelines to preserve the remaining flood plains and to aid in the restoration and mitigation from the historic loss of flood plain areas.

Immersed boundary (IB) method has been extensively researched and used due to its advantage of simulating flow over complex and moving objects in rivers. However, for high Reynolds number flows, it is still a challenge to produce correct and smooth wall shear stress that is essential for sediment transport. This challenge comes from the non-uniform wall distance of IB cell center and the discontinuous wall functions in boundary layers. Specifically, when the background grid is fixed and the immersed boundary geometry is arbitrary, the distance from IB cell center to immersed boundary is non-uniform. Also, the distance to its image point, which is generally proportional to the distance of IB cell center, has the same non-uniformity and the image point may be located in the log-law layer or laminar layer. However, in many existing IB method, there are two discontinuous wall functions in these two boundary layers. As a result, the shear velocity may be estimated with image points and different wall functions, which gives a non-smooth wall shear distribution. Moreover, some of these distances are extremely small, resulting in numerical instability and even divergence. In this work, an improved treatment is proposed where the image point distance is set to be proportional to the size of IB cell, which produces a more uniform and stable distribution of the distance. Consequently, all image points are able to be in the log-law layer and follow the same wall function to provide smooth wall shear stress. This IB method has been implemented in the U² RANS code. Cases in 1D and 2D are shown.

In geophysical flows, the presence of sediment in suspension and transport can play an important role in modifying the turbulent properties of the carrier fluid. Much research has focused on the effects of relatively small diameter (less than 1 mm) quartz-density (2.65 g/cm³) particles on turbulent flow. Asian carp is a wide-spread invasive aquatic species in the U.S., causing severe ecological problems in rivers and lakes. Unlike sediment, Asian carp eggs are semi-buoyant particles (~1.05 g/cm³ when initially spawned and ~1.00 g/cm³ in the post-water-hardening period) whose diameter stabilizes to approximately 5 mm. This paper examines how turbulent flow is affected by the presence of particles serving as surrogates for Asian carp eggs as a function of turbulence intensity. Experiments were conducted in a mixing box with the oscillating grid placed near the bottom boundary, and 2D PIV was used to quantify the turbulent characteristics of the carrier fluid. Five paired experiments with and without Asian carp egg surrogate particles were conducted. Results show that under different grid oscillation frequencies (2 to 6 Hz), the mean kinetic energy of the carrier fluid decreased slightly in the presence of the particles, but the turbulent kinetic energy of the fluid did not change appreciably. This suggests strongly that Asian carp eggs in suspension do not modify turbulence intensity of the carrier fluid. These experimental results provide important insight into the entrainment, transport, and deposition of Asian carp eggs, which can inform models to predict the future spread of this invasive species.

Since Rhie and Chow’s momentum interpolation was proposed, it has been widely used on collocated meshes. Improvements and correction methods for the momentum interpolation have been developed as well to resolve its dependency problem on the relaxation parameter, on the time step, and instability problem for flows with discontinuities. In this study, the momentum interpolation is assumed not only at each edge between two neighboring cells but also around each edge and its associated neighboring edges. The difference between the momentum interpolation at the edge and around the edge, defined as the so-called interpolation correction, is used to correct the velocity and flux across the edge. According to the demonstrations of examples, in general the proposed interpolation correction method is able to accelerate the convergence process, especially on meshes with non-uniformities and irregularities.

In order to protect the city of New Orleans from the Mississippi River flooding, the Bonnet Carré Spillway (BCS) was constructed from 1929 to 1936 to divert flood water from the river into Lake Pontchartrain and then into the Gulf of Mexico. However, a BCS opening event causes a lot of environmental problems in the lake. This paper presents the application of a two-dimensional numerical model (CCHE2D), developed at the National Center for Computational Hydroscience and Engineering, for simulating the flow circulations, sediment transport, and salinity distribution in Lake Pontchartrain during the Bonnet Carré Spillway (BCS) is opened for flood release. When the spillway is opened for flood release, large amount of fresh water and sediment discharged from the Mississippi River into Lake Pontchartrain, and then into the Gulf of Mexico, causing significant increase of lake sediment concentration and decrease of lake salinity. The aquatic environment and fish habitat in the lake are greatly affected by the opening events. CCHE2D model was applied to simulate the flow, sediment, and salinity in Lake Pontchartrain during the recent BCS flood releases in 1997, 2008, and 2011. The simulated results were compared with field measured data and satellite imageries, and good agreements were obtained. The flow patterns, sediment concentrations, and salinities in the lake due to the three flood release events were discussed. In addition, the salinity recovery processes in the lake were also simulated. The simulation results provides useful information to evaluate the environmental impacts of the flood events on lake ecosystems.

The U.S. Army Corps of Engineers (USACE) has recently completed a numerical modeling study of hydrodynamics, waves, and salinity for the proposed improvements to the Matagorda Ship Channel (MSC), Texas. The numerical modeling study investigates the impact to currents, waves, and salinity in Matagorda Bay from the proposed ship channel dimensions which are approximately 3 m (10 ft) deeper and 61 m (200 ft) wider. The modeling includes two historical hurricanes and three freshwater inflow conditions for high, medium, and low flow stages with representation of the present and future conditions of coastal boundary water surface elevation. Model results show the proposed ship channel dimensions will slightly increase flow efficiency and current magnitude in the bay. Water levels in the bay tend to increase slightly with proposed improvements and greater river inflow. The water level change in the bay is less sensitive in future condition with higher water level input than the present condition. The difference of salinity variations with and without channel improvements is small in Matagorda Bay, ordinarily less than 2 ppt. In the upper bay, the model minimum salinity becomes slightly greater with the proposed improvements. Higher salinity occurs more often in Matagorda Bay in future condition with higher water level than the present water level. The effect of the proposed ship channel dimensions to water surface elevations, current velocities, and salinities is overall small as compared to the present channel condition.

Traditionally, reduced-scale physical models have been used to evaluate the performance and risk assessment for water control structures. This approach is reliable; however, it is expensive and time-consuming. As an alternative, this paper presents application of computational fluid dynamics (CFD) modeling on the design optimization and performance evaluation of low head gated spillways in South Florida. Two cases studies are demonstrated, (a) risk assessment of various scenarios for erosion control downstream of S5AS Spillway under extreme flow conditions, and (b) optimization of the replacement design of S49 Spillway. In both cases, CFD simulation results of different design, construction, and operation scenarios are used to optimize the pressure field, velocity distribution, and potential scouring impact. This study illustrates that CFD, as a suitable tool, can help accelerate the process of design optimization and refurbishment of a large number of new and existing structures in a cost-effective manner.

Resistance partitioning is a standard method to divide total resistance into linear additive components. Nevertheless, in mountain rivers the consideration of linearity is questionable. This article analyses the resistance partitioning on five reaches with typical mountain river morphologies (three cascades and two step-pools) in the Quinoas River in southern Ecuador through the calibration of two equations: the calibrated Keulegan and another nondimensional hydraulic geometry equation calibrated with resistance measured in a plain bed. Each studied reach was divided into three to five cross sections to take geometric and hydraulic data for six months for a wide range of flows magnitudes. The calibrated Keulegan equation provides the best results based on the values presented in the literature, the calibrated nondimensional hydraulic geometry equation produces different values probably due to the presence of vegetation in the banks of the plain-bed reach.

Humankind have long built dams to provide resilience against droughts and floods. From the plains of the agricultural regions of Wadi Rajul of Jordon in the fourth millennium BCE to the modern United States, dams have served as vital pieces of economic infrastructure. The design functions provided by any dam are subject to degradation as sediment flows through the waterways and fills the storage behind the dam. Sustainable sediment management in reservoirs seeks to preserve long-term capacity by incorporating appropriate sediment management features into designs of new dams, implementing soil and water conservation practices in the watershed, stabilizing streambanks and waterways, and potentially modifying existing structures to retard the rate of sediment accumulation. The intent is to achieve a long-term sediment balance that preserves useful storage. This paper outlines the scope of this historic problem and how sedimentation continues to adversely affect modern dam infrastructure with a focus on small dams and waterways. Also discussed are techniques used to address the sedimentation issues as well as remedial actions. This presentation includes an overview of the variety of analyses and design techniques employed by the United States Department of Agriculture’s Natural Resource Conservation Service (USDA–NRCS) in the planning, design, and maintenance of watershed or river/stream projects. This presentation discusses the scope of the problem and provides an overview of the methods used to model and predict sedimentation.

Sediment transport has implications for activities such as fishing, flood control, scour countermeasures, and dredging through altered flow depths and sediment transport, bank erosion, and bridge scour. To estimate the changes in sediment transport to river discharge, water surface slope, and tailwater depth (to simulate sea-level rise) were altered in an existing sediment transport model. It was found that: (1) in uniform flow upstream of sea-level rise effects, sediment transport is sensitive to discharge; (2) in non-uniform flow affected by sea-level rise, sediment transport is sensitive to water surface slope and discharge; (3) the discharge value to restore the sediment transport rate existing before sea-level rise is proportional to the water surface slope ratio to the 4^th power; and (4) the discharge value to restore the bed sediment size existing before sea-level rise is proportional to the water surface slope ratio to the 3/4^th power. These relationships can be used for future application and planning purposes.

The purpose of this study is to model long-term shoreline changes on the Delaware coast near the Indian River Inlet (IRI) by using the U.S. Army Corps of Engineers shoreline evolution model, GenCade. By utilizing shoreline survey data and bathymetrical data of the inlet, this site-specific model was validated by hindcasting a 12-year-long historical shoreline evolution. The corresponding coastal processes simulated in the model include inlet sediment transport, offshore waves, longshore sediment transport, and coastal protection practices such as sand bypassing operation and beach nourishment. The evolution of shoals and bars around the inlet was simulated by using the inlet reservoir model (IRM). Model simulation skill was evaluated in terms of statistical errors by comparing simulated shoreline changes with observation data. The effect of sand bypassing operation is examined by comparing simulated shoreline changes with and without the bypassing.

The rivers of Taiwan are rushing because of the steep terrain. Loose soils are also tended to cause severe scours during typhoon period. The toe scour problem, especially in river bends and hydraulic structures, are common in Taiwan. In most cases of the broken bank, the concave side of bank is suffered by the severe oblique impact of the flood. To study the problem, the study conducted a down scaled experiment, which in the similarity to the cross section 27–34 of Fengshan River of Hsinchu County, to measure the maximum scour depth and patterns for studying the mechanism of erosion. The results of general bend scour experiment (without spur) showed that the riverbed resulted in a small amount of accumulation on the concave side of bank before the bend. The scouring was mainly occurred on the convex side of band. The flow rate showed the cross sections on the downstream of the concave side of bank has a faster velocity as the compared to the other side. The flow velocity in the downstream section of the concave side of bank was slowed down after the spurs were set. The results also showed that the reduction rates of scour volume were 7.97% and 4.65% on the 10-yr and 20-yr return period of discharge respectively. The maximum scouring depth was occurred at the front of spur after it was built. In addition to the maximum scour depth front of the spurs, the scouring along the downstream was founded getting away from the bank. The study indicated that building spurs can effectively ease the front erosion of spur and downstream scouring.

The pressure surge prediction in a closed conduit partially pressurized transient flow has been investigated following suddenly blocking the pipe downstream end. In a shock-fitting approach, the rigid column model as well as the Saint-Venant equations set and the ideal gas law for air pocket variations were used for the computations. The aim was to examine the performance of the shock-fitting approach and possible effects of the explicit and implicit time integration schemes on the quality of the solutions. It was found that the shock-fitting approach can improve the quality of the solutions compared to the rigid column model provided that implicit schemes are used for time integrations. The main reason is due to more dissipative property of implicit schemes, compared to explicit schemes, by which the extra kinetic energy associated with the approximation of the models can be dissipated.

Urbanization leads to increase in impervious area and eventually increasing vulnerability to urban flooding. Deer Creek in the city of Brentwood, Missouri, has been experiencing major significant high flows in the past and has flooded the neighboring communities several times. A major step in mitigating the flood is to evaluate the flooding extent. The current research utilizes the Hydrologic Engineering Center River Analysis System (HEC-RAS) and hydrodynamic modelling software—MIKE 21 to study the extent of floodplain in the urbanized area. A comparative study between the models to evaluate the modelling advantages of each model is performed. Moreover, the study focuses on two-dimensional (2D) modelling capabilities of each model as 2D models are more useful in urban areas with more numbers of discrete features. The tradeoff between the accuracy and resources is evaluated while comparing the 2D floodplain modelling capabilities of each models. Channel morphology and flow characteristics for different return period will be incorporated while modeling and the USGS gage data will be utilized for calibration of the models. The models will be calibrated for peak stages along the Deer Creek, with downstream boundary condition of normal depth and the unsteady flow simulations will be performed using the relevant flow hydrographs. The streamflow properties simulated will be derived from the measurements of USGS gaging site near Brentwood, MO.

The Green Bay of Lake Michigan is a major freshwater estuary and an example of degraded terrestrial ecosystems due to intensive and unmanaged anthropogenic activities. Decades of soil erosion and unregulated industrial activities in the watershed led to excessive transport of contaminated and nutrient-rich sediments into Green Bay. Considerable sediment loading triggered deteriorated biogeochemical conditions in the bay such as hypoxia, harmful algal blooms, eutrophication, and degraded water quality. Restoration programs were developed, consisting of important research and field activities. Previous research has shown that managing water quality and ecological conditions in the lower Green Bay system depends on understanding the sediment transport regimes, resuspension and sedimentation rates, and exchange with Lake Michigan. Due to the lack of a reliable and robust sediment transport model for Green Bay, some of those issues have remained unanswered, and water quality studies have encountered uncertainty in the design of restoration plans. In this study, we used a sediment transport model to investigate in detail the sediment dynamics in Green Bay. The model is built upon the finite volume community ocean model (FVCOM) platform, includes the entire Lake Michigan, and focuses on Green Bay with a finer grid compared to that in the open lake areas. The current sediment transport model will play a crucial role in the identification of sedimentation zones and major sediment transport regimes in the bay.

Identifying geologic and geotechnical related potential failure modes for dam safety risk assessments (backward erosion piping, suffusion, suffosion, etc.) rely on characterizing inherent variabilities and uncertainties within the constructed embankment and geological foundation. Subsurface physical and geochemical reactions from changing hydrologic loads on the dam will affect foundation and/or embankment soil performance over extended periods of time, necessitating dam risk analysis on a regular basis. The geochemical and isotopic compositions of water from groundwater, reservoir, piezometers, spillway, and seepage sites were comprehensively investigated at a North Texas dam. This survey utilized isotopic fractionation processes of deuterium and oxygen-18 to provide quantifiable “fingerprints” of water molecules from each water source. Combined with additional geochemical analysis of the groundwater, where suspected hydraulic pathways were tested against the possible cations/anions contributed by mineral dissolution in the soil, the source and degree of mixing of dam seepage could be approximated. Baseline isotopic signatures were established for groundwater (δD -25.6‰, δ¹⁸ O -4.5‰) along with reservoir isotopic ranges (δD -9.10 to -20.00 ‰ and δ¹⁸ O -3.37 to -1.00 ‰). By using a binary mixing model, these values were compared to the isotopic signatures from piezometers and suspected seeps to find subsurface hydraulic connections. At the site, a connection between a seep and the reservoir was found. Isotopically, the two samples were located on the same trendline with similar fractionations (seep δD -18.3‰, δ¹⁸ O -3.04‰; reservoir δD -20.0‰, δ¹⁸ O -3.37‰). Comparing the two samples geochemically a high amount of sulfate contribution (from 31 mg/L to 626 mg/L, a 2,019% increase) lead to the discovery of a potential seepage path through gypsum bearing embankment clay construction fill.

The patterns of water consumption are important data and, if well compiled, can be important to sizing and operation rules in a water distribution system (WDS). The aim of this study is to compare the water consumption levels before and after the implementation of a smart metering program in the city of Brasília, Brazil. The system was conceived in a residential block set that is part of region of Brasília called Jardins Mangueirais. The district meter area (DMA) presents 420 residential units and the intention of the smart metering was to provide a real time monitoring and processing system for water consumption at individual user level. Thereby, to register unit’s consumption, the Environmental Sanitation Company of Brasília (CAESB) installed 424 telemetry-equipped water meters: one for each unit, other three to register the consumption of the common areas and another one to identify the total water inflow to the DMA tank. In addition, an infrastructure for radio transmission was installed and a tool for data availability via web was developed. In order to determine the impacts of the smart metering system in consumption patterns, this study presents the evolution of consumption for the main water meter as well as for the individual units and for the common area water meters. The results suggest that there is stability in the water consumption when the main meter and the common area meters are analyzed for a short period, but there is a reduction when comparing with a control group. It is also possible to conclude that the impact of the smart metering system in the reduction of water consumption was higher for those units that presented lower volume of water consumed.

In water distribution networks (WDNs), hydraulic transients reveal information about the state of the WDN and its responses to operational changes in valves and pumps as well as pipe leaks and bursts. Thus, transient simulation, as a conventional approach to model and predict these transient behaviors, has shown increasing promises in various applications, such as WDN design, fault detection, sensor placement, and risk assessment. To facilitate the integration of transient simulation in these case-specific applications, this work contributes the transient simulation in water networks (TSNet), a new open-source Python package. TSNet includes the capacities to: (1) read EPANET INP files, (2) establish initial conditions, (3) perform transient simulations in various conditions, including valve and pump operations, leaks, and bursts, and (4) obtain and visualize results. TSNet not only provides users with a freely accessible software, rendering easiness for interaction, but also presents a platform where users and developers can further develop, improve, and extend the existing transient model. This paper includes the discussion on TSNet modeling framework, capabilities, an example application, and resources to help users to familiarize and interact with TSNet. TSNet can be downloaded from the GitHub repository, and the package documentation is available from Read the Docs. The GitHub repository contains links to software documentation, example codes and networks, and contacting information for reporting bugs and questions.

When assessing the potential effects of climate change on water resources, it is important to consider water demand. In order to do this, the effects of climatic variables on water demand must first be determined. This study utilizes historical daily water demand and climate data for the Boston metropolitan area in order to analyze the effects of temperature and precipitation on water demand and develop a multiple linear regression water demand model. This study breaks annual demand into two parts: (1) a winter base demand, unaffected by climate, and (2) a summer demand, which is modeled solely as a function of climatic variables. The coupled model II predicts summer water demand as a function of climatic variables, and results of the model can be used to detect trends in water demand, which can be used to hypothesize what future water demand may look like based on climate projections for the Boston area.

Due to rapid urbanization, the existing water distribution networks need to be expanded by adding sub-networks for the newly developed areas. The stochastic nature of demands requires several simulations of the full network for optimum design and cost minimization. The size of the simulation problem thus becomes complex and computationally expensive. The reduced models can be handy for designing and optimizing such networks. This paper describes the usage of equivalent electrical circuit theory to reduce a large water distribution network for focussed analysis of a sub-network. Two methodologies are proposed to obtain the equivalent network using linear and nonlinear forms of the Thevenin theorem. Unlike other network reduction methods, the reduced equivalent networks derived from these methodologies have only two elements, such as a reservoir and an equivalent pipe for the existing network. In the first method, the nonlinear pipe network is replaced and implemented with its analogous linear electrical network in a circuit simulator for finding open circuit voltage and short circuit current at the desired node and branch. The equivalent pipe network parameters are estimated using these values. In the second method, the equivalent network parameters are extracted by ffitting the driving point headloss characteristics at the desired node with a suitable headloss formula. Applicability and comparison of the proposed network reduction methodologies are demonstrated on a realistic water distribution network connected with a sub-network. The simulated results of the reduced networks are compared with the solutions of the original nonlinear network. The reduced network obtained from the second method is found to yield more accurate results than the first method when the driving point headloss characteristic curve is fitted accurately. The reduced networks can be solved in much less computation cost than the original network. Therefore, the proposed network reduction methodologies are beneficial for analyzing a focused part, i.e. sub-network, of a large pipe network with stochastic demands in much less computational effort.

A water distribution system can be viewed as a graph consisting of nodes and links. Those characterize the consumers and pipes and other network elements such as pipe junctions, valves, sources, tanks, pumps, and reservoirs. In addition to supplying the consumers required consumptions at required pressures, sustaining resiliency, the ability to detect contaminants, sensor leakage minimization, and others, are warranted. Amongst other methods, one of the approaches to accomplish part of these objectives (e.g., leakage control), is clustering (or district metering areas formation). Clustering algorithms vary with network size and objectives and are continuously evolving into practical techniques. The goal of this study is to introduce a new dynamic (i.e., time-varying) clustering methodology for trading-off system objectives such as cost, resiliency, and water quality (quantified through water age). The method is demonstrated on a small illustrative example application and a more complex water distribution system.

The purpose of this study is to develop a graph theoretic approach to evaluate segment vulnerability and importance (or criticality), so that isolation valves may be more strategically placed and key isolation valves can be identified. The method quantifies reachability from sources to segments for given segment isolation scenarios, allowing identification of critical and vulnerable segments. An important feature of this work is that it distinguishes between true sources (wells, treatment plants) and ephemeral sources (storage tanks) which have limited capacity. The capability to visually identify key segments is another advance presented in this paper. Preliminary results further indicate that the method captures the decrease in segment vulnerability as the number of valves increases in a network.

In recent years, automated systems of supervisory control and data acquisition (SCADA) have been widely used in water distribution networks, which have made the distribution system vulnerable to cyber-attacks. One of the main limitations of the previously proposed cyberattack detection models is that tampering the SCADA’s pressure readings would not be reflected on the estimated nodal demands. As a result, cyber-attacks caused by such manipulations would not be detected. To address this challenge, this paper incorporates pressure readings in the detection algorithm as a constraint and develops a mixed integer nonlinear programming (MINLP) formulation to estimate the nodal demands of a network, as well as its flow rates and nodal pressure readings. These values are estimated using the SCADA readings for link flows, pump’s status, tank’s water level, and total demand of a district metering area. Nodal pressure heads are calculated using Bernoulli’s and Hazen-Williams equations and pump’s head equation.

A shale oil or gas field can have hundreds of wells. Each well needs a substantial amount of water in a relatively short period for hydraulic fracture with the water usage demand varying as a step function. These wells can be scheduled for development over a long period. Thus, a water system, routing fresh and produced water for hydraulic fracture of wells, is different from a conventional water distribution system since the fresh/produced water system has significant variations in demand as a function of time and location. This distinction can result in challenges to the hydraulic modeling and design of such systems. In this paper, an optimization methodology is presented to model and design a typical fresh/produced water system. First, all produced water is routed towards the active demand wells in the system. Since multiple flow paths may exist from a produced well to demand well, a flow path needs to be identified that will minimize the hydraulic head at each produced well. Second, a few freshwater sources from among all potential sources are selected by minimizing friction losses along the flow paths from the freshwater sources to the active demand wells, constrained by the freshwater source limits. Third, booster pumps are located and sized to minimize energy consumption and ensure there are no MAOP violations or no column separations in the entire system. A hydraulic model was developed for a hypothetical fresh/produced water system. The system was successfully optimized to use all produced water available before utilizing freshwater. The developed optimization methodology was validated by using the hypothetical system under specified operations. Utilizing a fresh/produced water system for hydraulic fracture is relatively new to the industry and has created the need for a unique optimization methodology to model and design such a system. This methodology was developed considering the significant variations in water usage demand hydraulic fracture creates.

Millions of people around the world rely on intermittent water supply systems (IWS) where water supply is limited to specific hours during the day or to specific days in the week. During physical disruptions of the water network, the variables that affect consumer decisions are dynamic, and consumer behavior is affected by the attributes (in particular, length, and intensity) of the disruption. This study analyzes the short-term decision-making process of residential water consumers in a city in the Middle East. In this city, consumers have long adapted to the intermittency of water supply by installing household water storage with various capacities. However, the volumes of the stored water are often not sufficient to supply the regular household demand for the duration of the disruption, and consumers have to take actions regarding their water supply and/or water consumption. Based on the results of a survey of more than 250 households, a set of binary probit models were developed to model consumers’ decisions related to the timing of their responses to the disruption, their willingness to pay for faster delivery of water using tankers, and their willingness to pay to avoid waiting in-line at the tankers’ location. The results of the analysis show how variables such as household characteristics, tankers’ prices and waiting times, consumers’ knowledge of their households’ water situation, prior experience with disruptions, and socioeconomic parameters affect the consumers’ decisions when the piped IWS is disrupted.

Chlorination is applied as a primary and/or secondary disinfectant to avoid outbreaks of waterborne-diseases in drinking water supplies. Besides the desired effect of inactivating pathogens, chlorine also reacts with natural organic matter and precursors in raw water to form various disinfection by-products (DBPs), such as trihalomethanes (THMs). Spray aeration is a post-treatment method commonly employed in distribution systems to remove THMs from the water. It is known that to effectively remove THMs in relatively enclosed systems, proper ventilation is required to evacuate the contaminated air. Using a pilot-scale tank, this study assessed the effect of tank design parameters on the removal of the individual THM species. The parameters included air flow rate, blower angle, and vent location. In addition, model simulations were conducted with computational fluid dynamics software to observe the air stream path in the headspace as the aforementioned parameters were modified. It was observed that THM removal varied under different conditions. For example, locating the vent near the spray aeration nozzle (center of the tank) resulted in slightly lower removal compared to other vent locations. Additionally, blower angle induced <5% difference in total THM removals. Higher air flow rates increased THM removal up to ~70%; dibromochloromethane was the most receptive to increased air flow rate, followed by bromoform, dibromochloromethane, bromodichloromethane, and chloroform. These findings indicate that optimizing parameters associated with the blower and vents can enhance THM removal in storage tanks employing spray aeration processes. Therefore, air flow rate and vent placement should be carefully considered during construction and retrofitting of storage tanks with spray aeration systems.

This research work focused on the analysis of hydraulic transients induced by a pump trip off in a sewer pumping system located in Lisbon, Portugal. Field experiments have been carried out to collect transient pressure and steady-state flow rate data in the pumping station. The influence of dynamic effects on the system behaviour and on the numerical modelling has been investigated. Pressure transients were simulated considering the elastic water hammer model by means of the traditional method of characteristics as well as incorporating unsteady friction and gaseous cavitation by using the discrete gas cavity model. The boundary conditions of the hydraulic model were a pump with inertial properties at the upstream end and a free discharge to the atmosphere at the downstream end. Immediately downstream the pump, a check valve is installed and its closure was determined. Good agreement has been obtained between measured and computed pressure head histories if gaseous cavitation was taken into account in the hydraulic solver.

Saltwater intrusion in fresh water aquifers due to excessive groundwater pumping and sea-level rise has become a major concern mostly in coastal cities around the world. Subsequently, salt can be drawn to water distribution systems and can cause health problems if the amount is excessive. To document the time and location of unsafe levels of salt, system network modeling was performed here for a realistic looped water distribution system having 958 pipes, 874 junctions, one tank, and one ground water source. It was found that (1) the salt reaches the upstream portions quickly, (2) salt spread is sensitive to the conditions of outer junctions, and (3) salt level oscillates due to oscillating user demand. In addition, the optimal location of a flushing pond was examined as an alternative to conventional system flushing, where the flushing pond will capture the maximum amount of contaminated water possible. It was found that a flushing pond can reduce the amount of contaminated water discharged to the environment and the pond holding capacity is sensitive to its location.

The current study intends to evaluate the impact of using time series models as priors on the demand estimation process. Two different priors were compared: an informationless prior and a novel approach for prior generation based on the explicit propagation of the estimated demand uncertainty through autoregressive (AR) models. The proposed novel AR-type model can be understood as a natural extrapolation of the classic AR model that considers the estimated demands as random variable inputs instead of point estimates. The proposed models were tested utilizing a realistic case study containing one week of observed measurements. In addition, since sensor failures are commonly observed in real systems, the robustness of the adopted methods was evaluated under different levels of missing flow measurements. The results indicate that the uncertain AR prior was clearly superior to the informationless prior for both failure and no failure situations resulting in consistently lower errors and much smaller confidence intervals without significant losses in terms of reliability.

Minimizing or recovering excess energy is an important question for energy management in water distribution system (WDS). For a given network and topographical variations with system layouts, the recovering excess energy may provide better option. Comprehensive understanding of excess energy, which can be defined as the residual energy after the minimum pressure requirement met, is the basis of quantifying hydropower potential. Water-energy storage (i.e., pumped storage) which stores excess energy can maximize hydropower potential and can regulate energy distribution in system more uniformly and effectively. The topographic conditions, system demand, and capital cost may limit the number of water-energy storage to be installed. For a given system, how does the number of storage affect hydropower potential is the main focus of this study. Specifically, the objective of this study is to maximize hydropower generation considering three design parameters of water-energy storage: base elevation, maximum height, and diameter. A reference network, Net2 with slight modifications is used in this study. Results show that the average total hydropower generation increases with the increase of number of water-energy storage installed in the system (1,407 kWh for 1T and 1,469 kWh for 6T conditions). However, the rate of increase of hydropower generation reduces as the number of storages increase. More pipe flow and larger energy outflow from storage are observed under higher storage configurations (60 m³/s and 2,147 kWh under 1T and 109 m³/s and 2,577 kWh under 6T conditions) that show the potential use of WDS for energy storage.

Loops in water distribution networks (WDN) provide redundancy. There are surrogate measures of redundancy, quantifying the percentage of actual loops to maximum potential loops for a given network. However, it has been noted in the literature for over three decades that not all loops are equal or contribute to redundancy in the same way. There is a gap in differentiating between WDN layouts with the same percentage of loops, bypassing key network properties. This study applies a method previously used for biological network analysis, to capture more detail on WDN loops. Specifically, this method decomposes the original graph of the WDN to a tree graph, a network model representing loop connectivity. Analyzing such a network model is promising for enhancing the quantification of WDN redundancy and robustness. Measures of loop connectivity are then compared with hydraulic simulation data for benchmark WDN(s).