Journal of Water Resources Planning and Management

The high percentage of reliability of water resources systems has always been considered as a positive advantage by users. However, in arid and semiarid areas in which the flow rate of the reservoir is highly variable, it makes sense to reduce the system reliability and allocate less water to demand sites in order to avoid critical conditions such as reservoir depletion and to reduce severity of failure in low water months. In this study, a parameterization-simulation-optimization (PSO) based operation model was used whereby the efficiency of two different hedging rules was studied based on one-dimensional and two-dimensional relationships between release, storage, and input flow considering the stochastic conditions of the input flow. The optimal hedging parameters were determined through linking the reservoir simulation model to the multiobjective imperialist competitive algorithm. The model combines stochastic and historical monthly flow data of the Marun River (10,800 months in total) to optimize the system and extract hedging rules for the Marun Dam reservoir in Iran. In order to validate the developed model, the combination of stochastic data and residual historical data (612 months in total) was used. The model results from two different hedging policies were then compared with those from standard operating policies (SOPs). The results indicated that PSO model based on one-dimensional reservoir hedging policy compared to two-dimensional hedging methods and SOP was able to manage needs allocation in dry months and prevent reservoir depletion.

Contaminant intrusion in a water distribution network is an important concern because it can have hazardous consequences for the population. Reacting immediately is crucial to prevent or reduce the further propagation of contamination. In terms of contamination scenario characteristics, optimization is researched extensively as a valuable methodology to provide information. This work presented a procedure preceding the optimization which considerably reduces the search space for a potential contaminant source location. For each suspect node, a simulation is conducted with unrealistically high contaminant concentration injected throughout the whole simulation. If the sensors do not register contamination in a subsequent scenario, then that node can be eliminated as a possible contaminant source. The methodology is applicable for both single and multiple contaminant injection nodes. This approach was investigated in multiple benchmark networks and for different sensor placements in the literature. By coupling the proposed search space reduction method with an optimization approach, a novel efficient methodology for contamination source detection was presented.

Effectively detecting leaks is critical to improving leakage control management. Acoustic detection is one of the main leakage detection methods, and has been widely used in water utilities. Nevertheless, the effectiveness of this method is unsatisfactory in cases with various types of noise. To tackle this problem, this work proposes a leakage spectrogram to represent the features of leakage signals, and developed a time–frequency convolutional neural network (TFCNN) model to identify leakage signals. The performance of the TFCNN model was compared with other classification models (i.e., decision tree, support vector machine, multilayer perceptron, random forest, and extreme gradient boosting) under different signal-to-noise ratio (SNR) conditions. The results showed that the proposed method improves the accuracy and stability of leakage detection. Compared with other classification models, the TFCNN model had the best performance, and its mean accuracy reached 98% under different SNR conditions. Even in the case of $-10\mathrm{dB}$ SNR, the mean detection accuracy reached 90%. In practice, the mean detection accuracy reached 99% for different time–frequency resolutions. The transfer learning–based TFCNN model is a promising method for leakage detection in cases in which there are insufficient data sets.

An integrated design method that complements the increasing uptake of water-saving schemes (WSSs) is presented. WSSs may consist of water-saving fittings, appliances, rainwater-harvesting systems (RWHSs), and water-reuse schemes that provide the link between water distribution systems (WDSs) and sanitary sewers (SSs). The method developed here considers the interaction between WDSs, WSSs, and SSs to provide unified solutions. In this paper, a multiobjective optimization problem is formulated and solved considering three objectives, which are (1) minimization of the total cost incurred by implementing water system interventions, i.e., excluding the associated cost savings; (2) maximization of cost savings, which are benefits conferred by interventions; and (3) minimization of system water demand. The decision variables include conventional water network design interventions (i.e., addition of pipes) and the water-saving equipment at a household level. The main constraints include both WDS and SS hydraulic performances. The optimal trade-off solutions are obtained using nondominated sorting genetic algorithm optimization processes. This method was demonstrated in the realistic subsystem of the Tsholofelo Extension water and SS networks in Gaborone, Botswana. The results show that integrating and optimizing WDS, WSS, and SS designs would improve water security and lead to more sustainable water systems.

Hypothetical and real case studies were combined to explore the feasibility and effectiveness of a surrogate-based cyclic feedback updating approach for groundwater contamination source identification (GCSI) at dense non-aqueous-phase liquid (DNAPL)-contaminated sites. Support vector regression (SVR), kriging, and kernel extreme learning machine (KELM) models were integrated to build a surrogate model of the multiphase flow simulation model with a high computational efficiency. A mixed homotopy-differential evolution (DE) algorithm is presented to solve the optimization model, in which the integrated surrogate model was embedded, to obtain the identification results, and a cyclic feedback updating process was developed to gradually improve the results. Finally, GCSI uncertainty analysis was conducted using the Monte Carlo method. The results showed that the integrated surrogate model accurately approximates the simulation model, with a mean relative error of only 2.56%. The combination of the homotopy algorithm and DE algorithm provided an effective improvement over the traditional heuristic algorithm, and the mean relative error of the identified source characteristics was limited to 3.28%. GCSI accuracy was significantly improved after the application of the cyclic feedback updating method by reducing the mean relative error of the final identification results to 2.14%. In addition, the probability distribution characteristics of the identification results were obtained via uncertainty analysis to provide a comprehensive and reliable reference for decision makers.

The potential global warming impact on the extreme and mean streamflow of the Upper Blue Nile (UBN) River Basin was projected for the 2050s and 2080s by three hydrological models [Nedbør-Afstrømnings Model (NAM), Variable Infiltration Capacity (VIC) model, and Watflood model] driven by Representative Concentration Pathway (RCP) 4.5 and 8.5 climate change scenarios of four general circulation models (GCMs) dynamically downscaled by a regional climate model, Weather Research and Forecasting (WRF). The three hydrological models were able to capture the flow dynamics accurately in both calibration and validation periods. The mean daily maximum (minimum) temperature of UBN are projected to increase by about 1.35°C–2.38°C (1.72°C–2.74°C) under the RCP4.5 climate scenario and 2.22°C–4.47°C (2.5°C–5.1°C) under RCP8.5 climate scenario in future periods with reference to 1976–2005 as the base period. However, changes projected for the mean annual precipitation vary widely, ranging from $-10.3\%$ to 19.4%. The projected increase in evapotranspiration and increase or decrease in precipitation will result in a projected increase or decrease in streamflow of UBN. Overall, the median of mean annual streamflow of UBN is projected to decrease by 7.6% within a range of $-19.7\%$ to $+17.7\%$ in the 2050s and by 12.7% within a range of $-26.8\%$ to $+31.6\%$ in the 2080s. The ensemble mean of annual maxima (minima) of high return periods are projected by the three hydrologic models to be larger (smaller) in the 2050s and 2080s, respectively. On the whole, uncertainties due to the structure of hydrologic models, especially for low flow projections, are likely larger than uncertainties from GCMs and climate change scenarios. The results suggest that UBN is likely to experience more frequent and severe hydrologic extremes (flooding and droughts) in the future. Preliminary adaptive measures have been presented to mitigate the possible impact of droughts on UBN.

Reliable forecasts of water demand that account for factors that drive demand are imperative to understanding future urban water needs. The effects of meteorological dynamics and sociocultural settings are expressed weakly in many published municipal water demand models, limiting their utility for high-accuracy urban water demand modeling. To fill this gap, this paper presents an empirical daily urban water demand model based on a 365-day trailing average per capita demand that incorporates functions and factors for meteorological, seasonal, policy, and cultural driving forces. A nonlinear iterative regression model of daily water demand was calibrated and validated with historical data (2005–2015) for El Paso, Texas, a major urban area in the American southwest which had a consistent water conservation policy during the study period. The model includes daily temperature and precipitation response functions (which modify demand by as much as $\pm 20\%$ relative to the annual average), as well as factors that capture effects of month of the year, day of the week, and special holidays (which modify demand within $\pm 15\%$ relative to the annual average). For the validation period (2011–2015), the model performed well, with a coefficient of determination ($R^2$) of 0.95, a Nash–Sutcliff efficiency of 0.94, a mean absolute-value relative error of 4.38%, a relative standard error of estimate of 5.82%, a relative RMS error of 5.71%, and a mean absolute-value peak-day error of 2.78%. The use of these site-specific demand variables and response curves facilitates parsimonious urban water demand forecast modeling for regional water security.

An accurate prediction of future water consumption is necessary to create a satisfactory design for a water distribution system. In this study, two new hybrid approaches are proposed for accurately predicting future hourly and monthly water demands. The first approach is based on the hybridization of ensemble empirical mode decomposition (EEMD) and difference pattern sequence forecasting (DPSF), and the second is based on the hybridization of EEMD with DPSF and autoregressive integrated moving average (ARIMA). Historical hourly water consumption datasets of southeastern Spain and monthly datasets of Nagpur, India are used for assessing the performance of the proposed approaches. The performance of the EEMD-DPSF approach is checked using the root mean square error (RMSE), mean absolute error (MAE), and mean percentage absolute error (MAPE). Further, the results are compared with those obtained using PSF, ARIMA, DPSF, their hybrid models, and various other ANN models. The proposed EEMD-DPSF method is found to perform significantly better than the other state-of-the-art methods in terms of prediction accuracy without compromising time and memory complexities. The comparison between the two proposed models demonstrates that the EEMD-DPSF approach provides better results, whereas the EEMD-DPSF-ARIMA approach requires shorter computational time.

Tanzania water supply systems (WSSs) are facing resilience problems, compromising the quality and coverage of water supply services during floods. The expected rise in flood frequency and magnitude due to climate change and rapid urbanization will result in more frequent and severe impacts to the systems, prompting the need to strengthen resilience to potential impacts and minimize recovery times. This study adopted a five-stage process—literature review, preassessment, pretesting, a three-round Delphi survey, and tool evaluation—to develop a qualitative technical resilience assessment tool (QTRAT). Thematic analysis and standard descriptive statistical analysis were applied during the study. A qualitative tool with 4 principles and 10 indicators is proposed with a final semiquantitative technical index (FSTRI), where system maintenance, future expansion capability, renewal plans, and spare parts and equipment availability are the leading indicators. Evaluation results suggest that Tanzania urban WSSs have moderate resilience and are in particular need of improved design guidelines encompassing resilience aspects. The tool developed in this study contributes to informed decision-making about priority planning and budgeting of water organizations to improve the urban WSSs’ technical resilience.

This study introduces an approach to plan future water rates to achieve water distribution system (WDS) sustainability triple-top line (TTL) targets. The WDS components are modeled by connecting the EPANET hydraulic model with multiple interconnected subsystem models in a hybrid-system dynamics configuration. The approach is demonstrated with the hypothetical network, U-City, to optimally set the user fee to maximize a TTL sustainability index ($SI$). Overall, three demonstrations were performed to test the model: (1) identifying the influence of water price elasticity (WPE); (2) $SI$ sensitivity to a water price adjustment in rate and time; and (3) optimal planning of a water price adjustment strategy. The first demonstration illustrated that neglecting WPE leads to an overestimation of revenue. The second demonstration confirmed that aggressive water pricing does not lead to increased $SI$. Finally, the third demonstration showed that frequent and lower water price increases are more favorable for a higher WPE to create a sustainable system. In summary, the proposed approach can provide a useful way to analyze future water rates to maximize the sustainability of a WDS.

In this paper, the transient behavior of a single-branch pipe system (Y-system) is analyzed by means of numerical experiments in which a wide range of branch characteristics (i.e., size, location, and operating conditions) is investigated. In the executed systematic analysis, focused on the role of minor branches, the pressure signals of the Y-system have been compared with the ones of a single pipe, assumed as a reference, by means of the values of the determination coefficient, $R^2$. The provided explicit relationships between $R^2$ and the characteristics of the branch clearly show that the actual role of the branch is determined by the combination of the characteristics more than by any single one. Such relationships are a reliable tool for the system skeletonization, i.e., they allow evaluating in which conditions a given branch can be neglected because it does not affect significantly the transient behavior of the system. The reliability and practical implications of the proposed methodology are discussed by considering a large supply system with 10 minor branches as a case study.

Water supply management is challenging due to diverse stakeholders, conflicting objectives, increasing water demand, increasing regulation complexity, changing climate, and infrastructure constraints. To manage residual risk for water supply systems, utilities often apply water shortage mitigation plans to navigate through rare but extreme water shortage conditions before investing in an expensive new supply infrastructure. Some supply source augmentation strategies can be enabled through temporary reduction of regulatory constraints, e.g., groundwater production permits, that are protective of natural systems. Understanding the potential impact of pumping excursions on the groundwater level is important before implementing such management strategy. This study examines the resilience of water levels in the surficial aquifer system to temporarily increased groundwater pumping above a permit limit in the Tampa Bay region, located in west central Florida. Two major modeling tools were used, including the calibrated Integrated Northern Tampa Bay (INTB) model that dynamically couples hydrologic simulation of uplands and water bodies with groundwater and Unit Response Matrix (URM), which estimates groundwater level changes at monitoring wells due to pumpage changes at production wells. Both recovery time and maximum groundwater level changes are investigated. Results reveal that the median recovery time to preexcursion conditions for surficial aquifer in the study area is within 1 year for nearly all investigated scenarios. This indicates groundwater production excursions can serve as an attractive measure to mitigate severe water shortage conditions with no long-term adverse impacts to natural systems. Implication of this is a potential deferment of significant investment in infrastructure in the study area with the use of an effective water shortage mitigation plan to manage residual risk of the water supply system.

In water distribution networks (WDNs), pipe failure is the most frequent type of equipment malfunction, resulting in water supply suspension. To determine the water suspension area and to properly plan restoration work, the accurate identification of isolated subsystems with respect to valve locations is critical. The water suspension area can be divided into intended and unintended isolations, depending on the valve layout. This technical note discusses the conventional method of searching the valve-controlled isolation areas and specifically examines the weakness of the unintended isolation (UI) search algorithm that uses a node-by-node adjacency matrix. In this study, we propose a new search algorithm using a link-by-link adjacency matrix to identify the UI segments. Demonstrations on branch and loop networks show that the new approach correctly identifies the UI in various operating conditions. The proposed method is expected to replace the conventional method and can be applied in various valve-controlled segment analyses, such as optimal valve placement, real-time valve control for dynamic WDN operation, and postaccident pipe recovery planning.