An agent-based approach for generating virtual sewer systems

Machine Learning for Detecting Virus Infection Hotspots Via Wastewater-Based Epidemiology: The Case of SARS-CoV-2 RNA

Wastewater-based epidemiology (WBE) has been proven to be a useful tool in monitoring public health-related issues such as drug use, and disease. By sampling wastewater and applying WBE methods, wastewater-detectable pathogens such as viruses can be cheaply and effectively monitored, tracking people who might be missed or under-represented in traditional disease surveillance. There is a gap in current knowledge in combining hydraulic modeling with WBE. Recent literature has also identified a gap in combining machine learning with WBE for the detection of viral outbreaks. In this study, we loosely coupled a physically-based hydraulic model of pathogen introduction and transport with a machine learning model to track and trace the source of a pathogen within a sewer network and to evaluate its usefulness under various conditions. The methodology developed was applied to a hypothetical sewer network for the rapid detection of disease hotspots of the disease caused by the SARS-CoV-2 virus. Results showed that the machine learning model's ability to recognize hotspots is promising, but requires a high time-resolution of monitoring data and is highly sensitive to the sewer system's physical layout and properties such as flow velocity, the pathogen sampling procedure, and the model's boundary conditions. The methodology proposed and developed in this paper opens new possibilities for WBE, suggesting a rapid back-tracing of human-excreted biomarkers based on only sampling at the outlet or other key points, but would require high-frequency, contaminant-specific sensor systems that are not available currently.

Wastewater reuse and energy saving require a more decentralized urban wastewater system? Evidence from multi-objective optimal design at the city scale

Decentralization is recognized as an emerging solution for a more sustainable urban wastewater system (UWS) for the future. However, the debate of centralization vs. decentralization at the system's planning stage remains unresolved, mainly due to the complexity of the system's spatial structure and the multiple design objectives, such as water reuse and energy conservation. This paper presents the Sustainable Urban Wastewater System Generator (SUWStor) as a tool to address this issue. Integrating a graph representation of the system structure and the ant colony algorithm, SUWStor can produce Pareto optimal solutions for system design under three objectives: minimizing the capital cost, minimizing the operational energy consumption, and maximizing the water reuse capacity. The model is used for system design in a 100-square-km new city, the Xiong'an New District in China. Compared to the solution based on human experience, the model can reduce the system's capital cost by 7% and the operational energy in the pipe network by 26%, while maintaining the water reuse capacity at 100%. With this model, the relation between the optimal system layout and the choice over different design objectives can be discussed for any given area. In our case study, the optimal capacity of WWTPs for the lowest-cost solution is 48,000 m³ per day, leading to a total number of WWTPs of 5. As the water reuse level increases to maximum, the optimal capacity reduces to 15,000 m³ per day, where the number of WWTPs is 16. The model is also able to perform significantly better than the locally optimized results, in which only the WWTP locations are fixed at their optimal values. This demonstrates the importance of a global optimization model in designing the integrated UWS.

A Simplified Sanitary Sewer System Generator for Exploratory Modelling at City-Scale

Future climatic, demographic, technological, urban and socio-economic challenges call for more flexible and sustainable wastewater infrastructure systems. Exploratory modelling can help to investigate the consequences of these developments on the infrastructure. In order to explore large numbers of adaptation strategies, we need to re-balance the degree of realism of sewer network and ability to reflect key performance characteristics against the model's parsimony and computational efficiency. We present a spatially explicit algorithm for creating sanitary sewer networks that realistically represent key characteristics of a real system. Basic topographic, demographic and urban characteristics are abstracted into a squared grid of 'Blocks' which are the foundation for the sewer network's topology delineation. We compare three different pipe dimensioning approaches and found a good balance between detail and computational efficiency. With a basic hydraulic performance assessment, we demonstrate that we attain a computationally efficient and high-fidelity wastewater sewer network with adequate hydraulic performance. A spatial resolution of 250 m Block size in combination with a sequential Pipe-by-Pipe (PBP) design algorithm provides a sound trade-off between computational time and fidelity of relevant structural and hydraulic properties for exploratory modelling. We can generate a simplified sewer network (both topology and hydraulic design) in 18 s using PBP, versus 36 min using a highly detailed model or 1 s using a highly abstract model. Moreover, this simplification can cut up to 1/10th to 1/50th the computational time for the hydraulic simulations depending on the routing method implemented. We anticipate our model to be a starting point for sophisticated exploratory modelling into possible infrastructure adaptation measures of topological and loading changes of sewer systems for long-term planning.

Exploring the structural factors of resilience in urban drainage systems: a large-scale stochastic computational experiment

The focus of infrastructure design and management has turned from a reliability-based approach to a resilience-based one. Resilience is a system's ability to maintain its function and minimize failure consequences when faced with exceptional conditions. This study carries out a large-scale computational experiment to study how resilience is affected by system's structure in a combined sewer system. We build a stochastic generation model, involving a random sampling of facility locations and a graph-based random walk sampling algorithm to generate various layouts of pipelines. The performance of these virtual systems are assessed in the Storm Water Management Model. We apply statistical techniques on these samples to study the relation between resilience and system structure. Results show that the number of combined sewer overflow (CSO) outfalls is a more important factor of resilience compared to the number of wastewater treatment plants (WWTPs). Some locations are found more preferable for WWTP or CSO outfall placement, while adding WWTPs or outfalls at other locations might even lower the system's resilience. Size of the sub-catchments of the CSO outfalls also affects resilience. Although this effect is statistically significant, the extent is not remarkable compared to other factors. We further study the structural features of the cost-effective systems. The highest achievable resilience level increases as the number of CSO outfalls decreases and so does system's cost. This results from the difference in CSO quantity, therefore this dilemma can be cut off by end-of-pipe storage or treatment which specifically tackles CSO. The conclusion of this study provides an insight into the structural factors of combined sewer systems' resilience and can provide guidance for system's planning.

Multiscale Resilience in Water Distribution and Drainage Systems

Multiscale resilience, i.e., coordinating different scales within a system to jointly cope and mitigate risks on any single scale, is identified as the feature of a complex resilient system. However, in water distribution systems (WDSs) and urban drainage systems (UDSs), the inherent resilience is usually not multiscale resilience. By referring to the larger scale to larger pipes serving both local users and some other users at smaller scales, it can be found that smaller scales are not responsible for providing resilience to cope with failures in larger scales. These are because the main function of traditional water systems is to deliver water from upstream to downstream. This study demonstrates that improving multiscale resilience in WDSs and UDSs needs to allow water to travel reversely in the system via providing extra capacities and/or connections at smaller scales. This hypothesis is verified via case studies on a real world WDS and UDS.

Modelling transitions in urban water systems

Long term planning of urban water infrastructure requires acknowledgement that transitions in the water system are driven by changes in the urban environment, as well as societal dynamics. Inherent to the complexity of these underlying processes is that the dynamics of a system's evolution cannot be explained by linear cause-effect relationships and cannot be predicted under narrow sets of assumptions. Planning therefore needs to consider the functional behaviour and performance of integrated flexible infrastructure systems under a wide range of future conditions. This paper presents the first step towards a new generation of integrated planning tools that take such an exploratory planning approach. The spatially explicit model, denoted DAnCE4Water, integrates urban development patterns, water infrastructure changes and the dynamics of socio-institutional changes. While the individual components of the DAnCE4Water model (i.e. modules for simulation of urban development, societal dynamics and evolution/performance of water infrastructure) have been developed elsewhere, this paper presents their integration into a single model. We explain the modelling framework of DAnCE4Water, its potential utility and its software implementation. The integrated model is validated for the case study of an urban catchment located in Melbourne, Australia.

To connect or not to connect? Modelling the optimal degree of centralisation for wastewater infrastructures

The strong reliance of most utility services on centralised network infrastructures is becoming increasingly challenged by new technological advances in decentralised alternatives. However, not enough effort has been made to develop planning tools designed to address the implications of these new opportunities and to determine the optimal degree of centralisation of these infrastructures. We introduce a planning tool for sustainable network infrastructure planning (SNIP), a two-step techno-economic heuristic modelling approach based on shortest path-finding and hierarchical-agglomerative clustering algorithms to determine the optimal degree of centralisation in the field of wastewater management. This SNIP model optimises the distribution of wastewater treatment plants and the sewer network outlay relative to several cost and sewer-design parameters. Moreover, it allows us to construct alternative optimal wastewater system designs taking into account topography, economies of scale as well as the full size range of wastewater treatment plants. We quantify and confirm that the optimal degree of centralisation decreases with increasing terrain complexity and settlement dispersion while showing that the effect of the latter exceeds that of topography. Case study results for a Swiss community indicate that the calculated optimal degree of centralisation is substantially lower than the current level.

Importance of anthropogenic climate impact, sampling error and urban development in sewer system design

Urban drainage design relying on observed precipitation series neglects the uncertainties associated with current and indeed future climate variability. Urban drainage design is further affected by the large stochastic variability of precipitation extremes and sampling errors arising from the short observation periods of extreme precipitation. Stochastic downscaling addresses anthropogenic climate impact by allowing relevant precipitation characteristics to be derived from local observations and an ensemble of climate models. This multi-climate model approach seeks to reflect the uncertainties in the data due to structural errors of the climate models. An ensemble of outcomes from stochastic downscaling allows for addressing the sampling uncertainty. These uncertainties are clearly reflected in the precipitation-runoff predictions of three urban drainage systems. They were mostly due to the sampling uncertainty. The contribution of climate model uncertainty was found to be of minor importance. Under the applied greenhouse gas emission scenario (A1B) and within the period 2036-2065, the potential for urban flooding in our Swiss case study is slightly reduced on average compared to the reference period 1981-2010. Scenario planning was applied to consider urban development associated with future socio-economic factors affecting urban drainage. The impact of scenario uncertainty was to a large extent found to be case-specific, thus emphasizing the need for scenario planning in every individual case. The results represent a valuable basis for discussions of new drainage design standards aiming specifically to include considerations of uncertainty.

Exploring critical pathways for urban water management to identify robust strategies under deep uncertainties

Long-term projections for key drivers needed in urban water infrastructure planning such as climate change, population growth, and socio-economic changes are deeply uncertain. Traditional planning approaches heavily rely on these projections, which, if a projection stays unfulfilled, can lead to problematic infrastructure decisions causing high operational costs and/or lock-in effects. New approaches based on exploratory modelling take a fundamentally different view. Aim of these is, to identify an adaptation strategy that performs well under many future scenarios, instead of optimising a strategy for a handful. However, a modelling tool to support strategic planning to test the implication of adaptation strategies under deeply uncertain conditions for urban water management does not exist yet. This paper presents a first step towards a new generation of such strategic planning tools, by combing innovative modelling tools, which coevolve the urban environment and urban water infrastructure under many different future scenarios, with robust decision making. The developed approach is applied to the city of Innsbruck, Austria, which is spatially explicitly evolved 20 years into the future under 1000 scenarios to test the robustness of different adaptation strategies. Key findings of this paper show that: (1) Such an approach can be used to successfully identify parameter ranges of key drivers in which a desired performance criterion is not fulfilled, which is an important indicator for the robustness of an adaptation strategy; and (2) Analysis of the rich dataset gives new insights into the adaptive responses of agents to key drivers in the urban system by modifying a strategy.

Assessing the impact of transitions from centralised to decentralised water solutions on existing infrastructures--integrated city-scale analysis with VIBe

Traditional urban water management relies on central organised infrastructure, the most important being the drainage network and the water distribution network. To meet upcoming challenges such as climate change, the rapid growth and shrinking of cities and water scarcity, water infrastructure needs to be more flexible, adaptable and sustainable (e.g., sustainable urban drainage systems, SUDS; water sensitive urban design, WSUD; low impact development, LID; best management practice, BMP). The common feature of all solutions is the push from a central solution to a decentralised solution in urban water management. This approach opens up a variety of technical and socio-economic issues, but until now, a comprehensive assessment of the impact has not been made. This absence is most likely attributable to the lack of case studies, and the availability of adequate models is usually limited because of the time- and cost-intensive preparation phase. Thus, the results of the analysis are based on a few cases and can hardly be transferred to other boundary conditions. VIBe (Virtual Infrastructure Benchmarking) is a tool for the stochastic generation of urban water systems at the city scale for case study research. With the generated data sets, an integrated city-scale analysis can be performed. With this approach, we are able to draw conclusions regarding the technical effect of the transition from existing central to decentralised urban water systems. In addition, it is shown how virtual data sets can assist with the model building process. A simple model to predict the shear stress performance due to changes in dry weather flow production is developed and tested.