Potential of supervised machine learning algorithms for estimating the impact of water efficient scenarios on solids accumulation in sewers.

Understanding the negative effects of widespread implementation of optimal water efficient solutions may have on existing centralised sewer systems is still limited - one of these effects is the accumulation of solids in sewer pipes. Predicting these effects requires setting up and simulating complex detailed hydraulic sewer network models. Often, precise details of the sewer network layout and diurnal patterns of the wastewater flows are not available, limiting the applicability of using model predictions for such phenomena. In this study, the applicability of supervised machine learning (ML) algorithms for the development of a simplified surrogate model to predict solid accumulation in sewer pipes was investigated. A large number of highly variable sewer networks were synthetically generated and used to produce results that can be generalizable within the limitations of the current study. A hydrodynamic sewer model was set up and simulated for each synthetic sewer network and various scenarios in which different water-efficient solutions were considered. Simulation results indicated that the most impacts are expected to occur in the upstream part of the sewer networks, and that with 50% reduction in (waste-)water flows, 3-20% more pipes are expected to accumulate solids. It was further found that ML algorithms can be used to successfully predict locations of solids accumulation in sewer pipes without using hydrodynamic models. A simple tool based on the findings of this study, sparing the need to conduct complex hydraulic simulations, was developed. It allows the user to enter a set of pipe characteristics and the proportion of flow that is reduced due to the implementation of water efficient solutions, and it predicts whether the pipe will accumulate solids or not. The study results and the proposed ML algorithms can support the implementation of optimal water-efficient solutions that will promote designing and managing the water sensitive cities of the future.

Fluctuating potentials in GaAs:Si nanowires: critical reduction of the influence of polytypism on the electronic structure.

In this work, the effects of Si doping in GaAs nanowires (NWs) grown on GaAs (111)B by molecular beam epitaxy with different Si doping levels (nominal free carrier concentrations of 1 × 1016, 8 × 1016, 1 × 1018 and 5 × 1018 cm-3) are deeply investigated using scanning electron microscopy (SEM), transmission electron microscopy (TEM), grazing incidence X-ray diffraction (GID), photoluminescence (PL) and cathadoluminescence (CL). TEM results reveal a mixture of wurtzite (WZ) and zinc-blende (ZB) segments along the NW axis independently of the Si doping levels. GID measurements suggest a slight increase of the ZB fraction with the Si doping. Low temperature PL and CL spectra exhibit sharp lines in the energy range 1.41-1.48 eV, for the samples with lower Si doping levels. However, the emission intensity increases and is accompanied by a clear broadening of the observed lines for the samples with higher Si doping levels. The staggered type-II band alignment only determines the optical properties of the lower doping levels in GaAs:Si NWs. For the higher Si doping levels, the electronic energy level structure of the NWs is determined by electrostatic fluctuating potentials intimately related to the amphoteric behavior of the Si dopant in GaAs. For the heavily doped NWs, the estimated depth of the potential wells is ∼96-117 meV. Our results reveal that the occurrence of the fluctuating potentials is not dependent on the crystalline phase and shows that the limitation imposed by the polytypism can be overcome.

SiGe layer thickness effect on the structural and optical properties of well-organized SiGe/SiO2 multilayers.

In this work, we report on the production of regular (SiGe/SiO2)20 multilayer structures by conventional RF-magnetron sputtering, at 350 °C. Transmission electron microscopy, scanning transmission electron microscopy, raman spectroscopy, and x-ray reflectometry measurements revealed that annealing at a temperature of 1000 °C leads to the formation of SiGe nanocrystals between SiO2 thin layers with good multilayer stability. Reducing the nominal SiGe layer thickness (t SiGe) from 3.5-2 nm results in a transition from continuous SiGe crystalline layer (t SiGe ∼ 3.5 nm) to layers consisting of isolated nanocrystals (t SiGe ∼ 2 nm). Namely, in the latter case, the presence of SiGe nanocrystals ∼3-8 nm in size, is observed. Spectroscopic ellipsometry was applied to determine the evolution of the onset in the effective optical absorption, as well as the dielectric function, in SiGe multilayers as a function of the SiGe thickness. A clear blue-shift in the optical absorption is observed for t SiGe ∼ 2 nm multilayer, as a consequence of the presence of isolated nanocrystals. Furthermore, the observed near infrared values of n = 2.8 and k = 1.5 are lower than those of bulk SiGe compounds, suggesting the presence of electronic confinement effects in the nanocrystals. The low temperature (70 K) photoluminescence measurements performed on annealed SiGe/SiO2 nanostructures show an emission band located between 0.7-0.9 eV associated with the development of interface states between the formed nanocrystals and surrounding amorphous matrix.

Influence of the layer thickness in plasmonic gold nanoparticles produced by thermal evaporation.

Metallic nanoparticles (NPs) have received recently considerable interest of photonic and photovoltaic communities. In this work, we report the optoelectronic properties of gold NPs (Au-NPs) obtained by depositing very thin gold layers on glass substrates through thermal evaporation electron-beam assisted process. The effect of mass thickness of the layer was evaluated. The polycrystalline Au-NPs, with grain sizes of 14 and 19 nm tend to be elongated in one direction as the mass thickness increase. A 2 nm layer deposited at 250°C led to the formation of Au-NPs with 10-20 nm average size, obtained by SEM images, while for a 5 nm layer the wide size elongates from 25 to 150 nm with a mean at 75 nm. In the near infrared region was observed an absorption enhancement of amorphous silicon films deposited onto the Au-NPs layers with a corresponding increase in the PL peak for the same wavelength region.

Prioritization of rehabilitation interventions for urban water assets using multiple criteria decision-aid methods.

The aim of this paper is to compare sorting and ranking methods for prioritization of rehabilitation interventions of sewers, taking into account risk, performance and cost. For that purpose multiple criteria decision-aid (MCDA) methods such as ELECTRE TRI for sorting and ELECTRE III for ranking are applied in a real case-study and the results obtained are compared. The case study is a small sanitary sewer system from a Portuguese utility located in the metropolitan area of Lisbon. The problem to investigate is the prioritization of the sewer candidates for rehabilitation. The decision maker (a panel group of specialists) has chosen five assessment measures: water level and maximum flow velocity (hydraulic performance indices), sewer importance and failure repair cost (collapse-related consequences of failure) and the risk of collapse. The results show that the outcomes from ELECTRE III are easier to understand than those from ELECTRE TRI method. Two different sets of weights were used, and the sorting and ranking results from both methods were found to be sensitive to them. ELECTRE TRI method is not straightforward as it involves technical parameters that are difficult to define, such as reference profiles and cut levels.

Real-time forecasting urban drainage models: full or simplified networks?

Lead time between rainfall prediction results and flood prediction results obtained by hydraulic simulations is one of the crucial factors in the implementation of real-time flood forecasting systems. Therefore, hydraulic simulation times must be as short as possible, with sufficient spatial and temporal flood distribution modelling accuracy. One of the ways to reduce the time required to run hydraulic model simulations is increasing computational speed by simplifying the model networks. This simplification can be conducted by removing and changing some secondary elements using network simplification techniques. The emphasis of this paper is to assess how the level of urban drainage network simplification influences the computational time and overall simulation results' accuracy. The models used in this paper comprise a sewer network and an overland flow drainage system in both 1D/1D and 1D/2D approaches. The 1D/1D model is used as the reference model to generate several models with different levels of simplifications. The results presented in this paper suggest that the 1D/2D models are not yet suitable to be used in real-time flood prediction applications due to long simulation time, while on the other hand, the simplified 1D/1D models show that considerable reductions in simulation time can be achieved without compromising simulation results (flow and water depth) accuracy.

Urban pluvial flooding in Jakarta: applying state-of-the-art technology in a data scarce environment.

Available data relating to major pluvial flooding events in Jakarta, Indonesia were used to investigate the suitability of two different levels of sophistication in urban modelling tools for modelling these events. InfoWorks CS v9.0 was employed to build 1D and 1D/2D models of a 541 ha area of inner city Ciliwung River catchment which has a history of being particularly badly affected by flooding during heavy rainfall events. The study demonstrated that a 1D model was sufficient to simulate the flood extent of a major event using the limited data available. While the 1D/2D model also performed well, more data and time would have been required to match the 1D model's simulation of flood extent. Much more detailed data would have been required to produce reliable results in the 1D/2D model and to enable any kind of verification or calibration of the two models beyond visual comparison with crude flood extent maps.

Sensitivity analysis of surface runoff generation in urban flood forecasting.

Reliable flood forecasting requires hydraulic models capable to estimate pluvial flooding fast enough in order to enable successful operational responses. Increased computational speed can be achieved by using a 1D/1D model, since 2D models are too computationally demanding. Further changes can be made by simplifying 1D network models, removing and by changing some secondary elements. The Urban Water Research Group (UWRG) of Imperial College London developed a tool that automatically analyses, quantifies and generates 1D overland flow network. The overland flow network features (ponds and flow pathways) generated by this methodology are dependent on the number of sewer network manholes and sewer inlets, as some of the overland flow pathways start at manholes (or sewer inlets) locations. Thus, if a simplified version of the sewer network has less manholes (or sewer inlets) than the original one, the overland flow network will be consequently different. This paper compares different overland flow networks generated with different levels of sewer network skeletonisation. Sensitivity analysis is carried out in one catchment area in Coimbra, Portugal, in order to evaluate overland flow network characteristics.

The influence of digital elevation model resolution on overland flow networks for modelling urban pluvial flooding.

This paper presents the developments towards the next generation of overland flow modelling of urban pluvial flooding. Using a detailed analysis of the Digital Elevation Model (DEM) the developed GIS tools can automatically generate surface drainage networks which consist of temporary ponds (floodable areas) and flow paths and link them with the underground network through inlets. For different commercially-available Rainfall-Runoff simulation models, the tool will generate the overland flow network needed to model the surface runoff and pluvial flooding accurately. In this paper the emphasis is placed on a sensitivity analysis of ponds and preferential overland flow paths creation. Different DEMs for three areas were considered in order to compare the results obtained. The DEMs considered were generated using different acquisition techniques and hence represent terrain with varying levels of resolution and accuracy. The results show that DEMs can be used to generate surface flow networks reliably. As expected, the quality of the surface network generated is highly dependent on the quality and resolution of the DEMs and successful representation of buildings and streets.

Contribution of Geographic Information Systems and location models to planning of wastewater systems.

This paper presents the contributions of Geographic Information Systems (GIS) and location models towards planning regional wastewater systems (sewers and wastewater treatment plants) serving small agglomerations, i.e. agglomerations with less than 2,000 inhabitants. The main goal was to develop a decision support tool for tracing and locating regional wastewater systems. The main results of the model are expressed in terms of number, capacity and location of Wastewater Treatment Plants (WWTP) and the length of main sewers. The decision process concerning the location and capacity of wastewater systems has a number of parameters that can be optimized. These parameters include the total sewer length and number, capacity and location of WWTP. The optimization of parameters should lead to the minimization of construction and operation costs of the integrated system. Location models have been considered as tools for decision support, mainly when a geo-referenced database can be used. In these cases, the GIS may represent an important role for the analysis of data and results especially in the preliminary stage of planning and design. After selecting the spatial location model and the heuristics, two greedy algorithms were implemented in Visual Basic for Applications on the ArcGIS software environment. To illustrate the application of these algorithms a case study was developed, in a rural area located in the central part of Portugal.