Urban flood modeling with a novel coupling method of surface and sewer hydrodynamic processes.

Drainage modeling that accurately captures urban storm inundation serves as the foundation for flood warning and drainage scheduling. In this paper, we proposed a novel coupling ideology that, by integrating 2D-1D and 1D-2D unidirectional processes, overcomes the drawback of the conventional unidirectional coupling approach that fails to properly represent the rainfall surface catchment dynamics, and provides more coherent hydrological implications compared to the bidirectional coupling concept. This paper first referred to a laboratory experimental case from the literature, applied and analyzed the coupling scheme proposed in this paper and the bidirectional coupling scheme that has been widely studied in recent years, compared the two coupling solutions in terms of the resulting accuracy and applicability, and discussed their respective strengths and weaknesses to validate the reliability of the proposed method. The verified proposed coupling scheme was then applied to the modeling of a real drainage system in a region of Nanjing, China, and the results proved that the coupling mechanism proposed in this study is of practical application value.

Improving the interpretability of deep reinforcement learning in urban drainage system operation.

Deep reinforcement learning (DRL) has been increasingly used as an adaptive and efficient solution for real-time control (RTC) of the urban drainage system (UDS). Despite the promising potential of DRL, it is a black-box model whose control logic and control consequences are difficult to be understood and evaluated. This leads to issues of interpretability and poses risks in practical applications. This study develops an evaluation framework to analyze and improve the interpretability of DRL-based UDS operation. The framework includes three analysis methods: Sobol sensitivity analysis, tree-based surrogate modelling, and conditional probability analysis. It is validated using two different DRL approaches, i.e., deep Q-learning network (DQN) and proximal policy optimization (PPO), which are trained to reduce combined sewer overflow (CSO) discharges and flooding in a real-world UDS. According to the results, the two DRLs have been shown to perform better than a rule-based control system that is currently being used. Sobol sensitivity analysis indicates that DQN is particularly sensitive to the flow of links and rainfall, while PPO is sensitive to all the states. Tree-based surrogate models effectively reveal the control logic behind the DRLs and indicate that PPO is more comprehensible but DQN is more forward-looking. Conditional probability analysis demonstrates the potential control consequences of the DRLs and identifies three situations where the DRLs are ineffective: a) the storage of UDS is fully utilized; b) peak flows have already passed through actuators; c) a substantial amount of water enters one location simultaneously. The proposed evaluation framework enhances the interpretability of DRL in UDS operations, fostering trust and confidence from operators, stakeholders, and regulators.

Enhancing interpretability and generalizability of deep learning-based emulator in three-dimensional lake hydrodynamics using Koopman operator and transfer learning: Demonstrated on the example of lake Zurich.

Three-dimensional lake hydrodynamic model is a powerful tool widely used to assess hydrological condition changes of lake. However, its computational cost becomes problematic when forecasting the state of large lakes or using high-resolution simulation in small-to-medium size lakes. One possible solution is to employ a data-driven emulator, such as a deep learning (DL) based emulator, to replace the original model for fast computing. However, existing DL-based emulators are often black-box and data-dependent models, causing poor interpretability and generalizability in practical applications. In this study, a data-driven emulator is established using deep neural network (DNN) to replace the original model for fast computing of three-dimensional lake hydrodynamics. Then, the Koopman operator and transfer learning (TL) are employed to enhance the interpretability and generalizability of the emulator. Finally, the generalizability of DL-based emulators is comprehensively analyzed through linear regression and correlation analysis. These methods are tested against an existing hydrodynamic model of Lake Zurich (Switzerland) whose data was provided by an open-source web-based platform called Meteolakes/Alplakes. According to the results, (1) The DLEDMD offers better interpretability than DNN because its Koopman operator reveals the linear structure behind the hydrodynamics; (2) The generalization of the DL-based emulators in three-dimensional lake hydrodynamics are influenced by the similarity between the training and testing data; (3) TL effectively improves the generalizability of the DL-based emulators.

Optimal sensor placement for the routine monitoring of urban drainage systems: A re-clustering method.

The construction of an efficient monitoring network is critical for the effective and safe management of urban drainage systems. This study developed a re-clustering methodology that incorporates additional perspectives beyond node similarity to improve the traditional clustering process for optimal sensor placement. Instead of targeting event-specific water quality or hydraulic monitoring, the method integrates the water hydraulic and quality characteristics of nodes in response to the demand for routine monitoring. The implementation of this method first applies model simulation to generate the attribute datasets required for clustering analysis, and then re-clusters the initial clustering result according to the constructed re-clustering potential indices. And the information theory-based evaluation metrics were introduced to quantitatively assess the sensor deployment scheme obtained by amalgamating the two clustering results. Two networks with different drainage systems and sizes were chosen as case studies to illustrate the application of the framework. The results demonstrate that the clustering process enables to expand the information contained in the monitoring network, and that the re-clustering strategy can generate more comprehensive and practical solutions upon this basis.

A convenient and stable graph-based pressure estimation methodology for water distribution networks: Development and field validation.

Accurate estimation of unknown nodal pressures (nodal heads) is necessary for efficient operation and management of water distribution networks (WDNs), but existing methods such as hydraulic simulation and data interpolation can hardly reconcile estimation accuracy with model construction and maintenance costs. Recent developments in graph signal processing (GSP) techniques provide us with new tools to utilize information in WDN hydraulics and available measurements. In a pilot study, a graph-based head reconstruction (GHR) method was proposed, which used GSP to reconstruct the spatially slow-varying parts of nodal heads from a limited number of field measurements to approximate original heads. GHR has illustrated the effectiveness and ease of implementation of GSP-based methods. However, due to the ill-conditioning reconstruction process and inherent uncertainties, GHR may show unstable results with large errors if pressure meters are not installed at specific optimized locations, which limits its applicability. To solve this problem and discover a stable and convenient method that can support a wider range of applications, a graph-based head reconstruction method with improved stability (GHR-S) is proposed. GHR-S utilizes a rough estimation of unknown pressures as pseudo measurements, which provide additional constraints and avoid the occurrence of unreasonable results during the reconstruction process. A middle-sized network with synthetic data illustrates the stability, convenience, and accuracy of GHR-S with arbitrary meter locations and uncalibrated model parameters. GHR-S is also applied to a large real-life network with field measurements, and successfully estimates the unknown pressures of 83,000 nodes with only 58 measurements, showing its effectiveness in practical engineering.

Maintaining the long-term accuracy of water distribution models with data assimilation methods: A comparative study.

The upgrading of water supply services is calling for more accurate and adaptive numerical models to give insight into actual water distribution systems (WDSs), which underlines the importance of carefully calibrating model parameters. Due to unavoidable uncertainties in the calibration process such as measurement errors, errors in model parameters assumed to be known, and local-optimum of calibration algorithms, calibrated parameters could still contain non-negligible latent errors, and the calibrated model may not able to maintain its long-term accuracy when operating conditions change. To solve this problem, there is growing interest in adopting data assimilation (DA) methods to utilize more comprehensive information in long-term measurements to reduce the impact of uncertainties and maintain the accuracy and stability of calibrated models. In this study, two traditional calibration methods and four DA methods were tested and compared in two WDSs with different structures, which aims to form a general understanding of the behavior and applicability of different methods. The calibration results show that DA methods perform better than traditional methods and are more robust to different types of uncertainties, which provide an effective way to maintain the long-term accuracy of WDS models to enable better management of WDSs. Ensemble-based DA methods such as Particle Filter (PF) and Inferential-Measurement Kalman Filter (IMKF) performed well in the real-life system. They avoid linear approximation and can better estimate the impact of uncertainties to assimilate accurate correction information of the parameters. Gradient-based DA methods such as Extended Kalman Filter (EKF) and Variational Bayesian Adaptive Kalman Filter (VBAKF) have lower computational demand, but they are found to be less robust when dealing with large system uncertainties and nonlinearities.

Water Quality Prediction Based on Multi-Task Learning.

Water pollution seriously endangers people's lives and restricts the sustainable development of the economy. Water quality prediction is essential for early warning and prevention of water pollution. However, the nonlinear characteristics of water quality data make it challenging to accurately predicted by traditional methods. Recently, the methods based on deep learning can better deal with nonlinear characteristics, which improves the prediction performance. Still, they rarely consider the relationship between multiple prediction indicators of water quality. The relationship between multiple indicators is crucial for the prediction because they can provide more associated auxiliary information. To this end, we propose a prediction method based on exploring the correlation of water quality multi-indicator prediction tasks in this paper. We explore four sharing structures for the multi-indicator prediction to train the deep neural network models for constructing the highly complex nonlinear characteristics of water quality data. Experiments on the datasets of more than 120 water quality monitoring sites in China show that the proposed models outperform the state-of-the-art baselines.

Bridging hydraulics and graph signal processing: A new perspective to estimate water distribution network pressures.

The low spatial density of monitored nodal pressures (nodal heads) has already become a bottleneck restricting the development of smart technologies for water distribution networks (WDNs). Inferring unknown nodal heads through available WDN information is an effective way to bypass data limitations, but an accurate and easy-to-implement method is still absent. For general WDNs, the spatial distribution of nodal heads is approximately 'smooth' as there are few dramatic head changes. If heads can be divided into components with different spatial varying speeds, then they can be approximated by a few slow varying components. On this basis, a graph-based head reconstruction (GHR) method is proposed, which employs graph signal processing technologies to reconstruct the slow varying parts to estimate unknown nodal heads. Four metrics are proposed to bridge WDN hydraulics and signal processing to quantify the similarity of adjacent nodal heads, which enhance the smoothness of heads over the graph, and thus increase estimation accuracy. GHR was tested with different parameter settings and compared with other head estimation methods. Results showed that GHR has less restrictive parameter requirements compared with hydraulic simulation, and outperforms traditional data interpolation methods with better accuracy. At a larger looped network under potential model uncertainties and measurement errors, GHR still accurately estimated the heads for more than 10,000 unknown nodes, achieving a mean absolute error of 0.13 m using only 100 pressure meters. Thus the proposed method provides an efficient, robust, and convenient way to estimate unknown nodal heads in WDNs.

Contamination source identification in water distribution networks using convolutional neural network.

Contamination source identification (CSI) is significant for water quality security and social stability when a contamination intrusion event occurs in water distribution systems (WDSs). However, in research, this is an extremely challenging task for many reasons, such as limited number of water quality sensors and their limitations in detecting contaminants. Hence, some researchers have introduced consumers' complaint information as an alternative of sensors for CSI. But the problem with this approach is that the uncertainty of complaint delay time has a great impact on the identification accuracy. To address this issue, this study constructed complaint matrices to present the spatiotemporal characteristics of consumer complaints in an intrusion event and proposed a new methodology employing convolution neural network (CNN)-a deep learning algorithm-for the purpose of pattern recognition. CNN aimed to explore the inherent characteristics of complaint patterns corresponding to different contaminant intrusion nodes and to improve the performance of identifying the contamination source based on consumer complaint information. Two case studies illustrated methodology effectiveness in WDSs of various scales, even with the high uncertainties of complaint delay time. The comparison between CNN and a back-propagation artificial neural network algorithm demonstrates that the former framework possesses stronger robustness and higher accuracy for CSI.

Deep learning identifies accurate burst locations in water distribution networks.

Pipe bursts in water distribution networks lead to considerable water loss and pose risks of bacteria and pollutant contamination. Pipe burst localisation methods help water service providers repair the burst pipes and restore water supply timely and efficiently. Although methods have been reported on burst detection and localisation, there is a lack of studies on accurate localisation of a burst within a potential district by accessible meters. To address this, a novel Burst Location Identification Framework by Fully-linear DenseNet (BLIFF) is proposed. In this framework, additional pressure meters are placed at limited, optimised places for a short period (minutes to hours) to monitor system behaviour after the burst. The fully-linear DenseNet (FL-DenseNet) newly developed in this study modifies the state-of-the-art deep learning algorithm to effectively extract features in the limited pressure signals for accurate burst localisation. BLIFF was tested on a benchmark network with different parameter settings, which showed that accurate burst localisation results can be achieved even with high model uncertainties. The framework was also applied to a real-life network, in which 57 of the total 58 synthetic bursts in the potential burst district were correctly located when the top five most possible pipes are considered and among them, 37 were successfully located when considering only the top one. Only one failed because of the very small pipe diameter and remote location. Comparisons with DenseNet and the traditional fully linear neural network demonstrate that the framework can effectively narrow the potential burst district to one or several pipes with good robustness and applicability. Codes are available at https://github.com/wizard1203/waternn.

A simple but robust convergence trajectory controlled method for pressure driven analysis in water distribution system.

This paper presents a novel convergence trajectory controlled method to perform pressure driven analysis (PDA) in water distribution systems (WDSs). The proposed method makes forcibly the convergence error decrease continuously, which is fundamentally different from the traditional uncontrolled convergence process, thereby ensuring a robust convergence behavior for hydraulic analysis with PDA in WDS. In addition, two Relaxation Factor section strategies are developed to control the convergence trajectory towards the desired downtrends. The novel methodology is implemented based on EPANET3.0 by modifying the source code which is available in GitHub (https://github.com/OpenWaterAnalytics/epanet-dev). Firstly, the improved code was validated extensively with a benchmark WDS under rigorous boundary conditions. Subsequently, four challenging different size WDSs are also tested in terms of the effectiveness and efficiency. The results illustrate that the proposed method is able to enable the convergence of PDA to be more stable and more robust, even under some extreme abnormal boundary conditions.

Iron stability on the inner wall of prepared polyethylene drinking pipe: Effects of multi-water quality factors.

Iron is currently one of the main contaminants of drinking water. The inner walls of drinking pipes can cause iron to release in water chemistry, which alters the water quality, including its chloride, sulfate, bicarbonate, pH, and humic acid (HA) levels. Hence, the goal of this research was to improve our understanding of the multi-water quality factors affecting iron release in polyethylene pipes. An array of bench-scale experiments were conducted exposing model water with different concentrations of chloride, sulfate, bicarbonate, HA, and different pH levels to prepared polyethylene pipes following the response surface methodology. The single role of HA during iron release is also evaluated by changing its concentration. A comprehensive study revealed that regression models could be used to describe the relationship between the five water quality parameters and iron release. The coefficients of determination were 0.890 and 0.870 for the fitting equations of total and soluble iron concentrations in water, respectively. In the presence of HA, the concentration of iron in water increased more rapidly than that for the other four factors (chloride, sulfate, bicarbonate, and pH). In addition, the Visual MINTEQ results suggest that a lower HA concentration tended to increase the degree of saturation of iron solids. In turn, this limits iron release and considerably increases the iron concentration in water.

A practical multi-objective optimization sectorization method for water distribution network.

In recent years, water leakage problems in water distribution networks (WDN) have been attracting more attention, with an emphasis on energy and water resources. As one of the measures used for flow monitoring and leakage control, water network sectorization is a research hotspot in China. This paper, which begins with an introduction of present sectorization methods, proposes a multi-objective optimization sectorization method based on a comprehensive consideration of the hydraulics, water quality and economy. This method is based on the non-dominated sorting genetic algorithm (NSGA-II) which is a heuristic algorithm for multi-objective optimization to obtain the optimal schemes. In addition, human experience is also considered in the optimization process. In a case study, this method proves to be efficient in producing good results with little impact on the hydraulics and water quality of the WDN, and the results obtained are acceptable for multiple objectives. Therefore, this method provides references for the transformation of future water distribution network sectorization.

Equilibrium and kinetic modeling of iron adsorption and the effect by chloride, sulfate, and hydroxyl: evaluation of PVC-U drinking pipes.

The update of pipeline was quick over the last few years and the plastic pipes were widely used in the drinking water distribution systems (DWDSs), especially in the small-diameter pipes. In this study, the iron adsorptive characteristics and the affecting factors in unplasticized poly(vinyl chloride) (PVC-U) pipe were investigated. Results showed that the average amount of iron in the 10-year-old PVC-U pipe's interior surface was 2.80 wt% which was almost 187 times larger than that in a new one. Goethite (α-FeOOH) and magnetite (Fe3O4) were the major iron compounds in the scales which covered on the old pipes' interior surface and showed loose and porous images under a scanning electron microscope. Moreover, the influence of the iron concentration on the adsorption amount and rate was discussed. The adsorption amount was significantly influenced by iron concentration, but similar adsorption rate was discovered. Notably, iron was quantitatively adsorbed by PVC-U pipe during the experimental period in accordance with the pseudo second order kinetic model. Meanwhile, regression model and response surface methodology were used to analyze the regular of iron adsorption in different concentrations of chloride (Cl-), sulfate (SO42-), and hydroxyl (OH-). It can be concluded that Cl- and OH- showed the strong ability of iron adsorption which were larger than SO42-.

Experimental study using the dilution incubation method to assess water biostability.

Water biostability is of particular concern to water supply as a major limiting factor for heterotrophic bacterial growth in water distribution systems. This study focused on bacterial growth dynamics in the series dilution of water samples with TOC (total organic carbon) values determined beforehand. The results showed that the specific growth rate of Pseudomonas fluorescens P17 varied dramatically and irregularly with TOC value when TOC concentrations were low enough during the initial periods of incubation under given conditions. According to this relationship between bacterial growth rate and TOC, a dilution incubation method was designed for the study of water biostability. With the method under a given condition, a turning-point TOC value was found at a relatively fixed point in the curve between bacterial growth rate and TOC of water sample, and the variation of growth rate had different characteristics below the turning-point TOC value relative to that over this value. A turning-point TOC value similarly existed in all experiments not only with tap water, but also with acetate and mixed solutions. And in the dilution incubation method study, the affections were analyzed by condition factors such as inoculum amount, incubation time and nature of the organic carbon source. In very low organic carbon water environments, the variation characteristics of bacterial growth rate will be useful to further understand the meaning of water biostability.