Peter Chalk Centre

University of Exeter

Stocker Road

Exeter

EX4 4QD

Tel: +44 (0)1392 263637

E-mail: CCWI2019@exeter.ac.uk 

17th International Computing & Control for the Water Industry Conference

1st - 4th September 2019
University of Exeter, UK
  • Facebook Social Icon
  • Twitter Social Icon
  • Instagram

Poster session1

Please note the time shown on this page is automatically adjusted by the server according to the time zone set in your computer.

Hall 1

Slobodan Djordjevic

Chair:

to

Development of a performance assessment system to improve water supply service in Iraqi Kurdistan

Kegong Diao

Presenter:

Authors:

Faris Saleem, Katherine Huddersman, Andrew Wright and Kegong Diao

The key problem for most water supply utilities in the developing countries is the total absence of an efficient performance assessment system in terms of performance assessing and monitoring. Kurdistan region of Iraq is no different. Therefore, there was limited data and unreliable information available or standardised information for water stakeholders about the actual situation regarding the utility’s performance situations. This problem made it difficult for the relevant staff to know the actual efficiency of the current and past performance. Thus, to solve this problem, developing an effective performance assessment system is a key driver for water utilities to deliver sustainable water services and to ensure that the utility’s various functions are operating properly. This paper will provide insight into, how water utilities can develop their applicable performance assessment structure, while taking into consideration their local operating environments. So, based on utility’s objectives and priorities, this study will need to define three essential components of the proposed system namely: performance area/assessment criteria, its suitable performance indicators, and its corresponding acceptable performance targets. The major aim of this study is to develop a proper system or framework for assessing the performance of public water services sector in Iraqi Kurdistan, from water utility’s point of view to drives the efficient, effective and sustainable water services provision. The research methodology will adopt a participatory approach for developing an appropriate performance assessment framework and will use questionnaire for quantitative data collection. Then it will use a bench-marking method and the partial performance indicators as a tool to analyse the collected data. In this point, to test the applicability of the developed system, this study will benchmark individual indicator against its established acceptable performance targets to identify performance gaps that need improvement.

to

Pump scheduling optimization for a transmission main in the Umbria region

Marco Cifrodelli

Presenter:

Authors:

Marco Cifrodelli, Renzo Patacca, Silvia Meniconi, Andrea Vitali, Caterina Capponi and Bruno Brunone

The Upper Tiber Valley Transmission Main (UTVTM) is an important tree network that provides drinking water from the Montedoglio artificial lake (Tuscany) to the northern areas of the Umbria region (Central Italy). During the last years the UTVTM has been upgraded and improved to increase the discharge and then to disconnect the existing wells because of the poor quality of the withdrawn water. For these purposes a new water treatment plant and three new pumping stations have been built. The aim of this work is the UTVTM hydraulic, economic and management analysis, in the case of ordinary water consumption and in the ones of water crisis in the main cities of this area (Perugia and Città di Castello). The hydraulic analysis has been carried out by either excluding or including the pumping stations, in order to investigate the optimal scheduling of pumps by ensuring the maximum network reliability at minimum energy consumption. It is worth noting that the operation of pumps imposes significant costs on a pipe network for energy supply and pumps maintenance. For each condition more than three thousands hydraulic simulations have been carried out, by varying the pump patterns and assuming at least a pressure of 10 m in each secondary tank delivery node. The optimization procedure has allowed reducing the forecast operating costs by 9.2 % (water deficit both in Città di Castello and Perugia) and 89.8 % (ordinary conditions). The obtained results confirm that the pump optimizations are essential for an efficient management of water pipe systems.

to

Optimal DMAs for The Hague

Peter van Thienen

Presenter:

Authors:

Karel van Laarhoven, Ina Vertommen and Peter van Thienen

For years, district metered areas (DMAs) have been playing an essential role in pressure regulation and leakage loss reduction in many countries. In the Netherlands, however, DMAs have not been applied as rigorously, possibly due to both height differences and water losses (<6%) generally being low. The Dutch drinking water companies have only recently begun to develop an interest in applying DMAs to learn more about the local behaviour of water flow and demand in the distribution network, and to further reduce non-revenue water. This presents the Dutch utilities with the challenging task of designing a subdivision into DMAs for large swathes of their drinking water distribution networks at once. The distribution networks themselves were never designed with DMAs in mind and are highly meshed, which complicates the task of finding efficient DMA designs that can be realized with a limited number of DMA boundaries. Minimizing the number of boundaries is crucial, however, since installing volume flow meters and/or valves is an expensive and arduous task. Moreover, the Dutch water utilities are reluctant to close pipes on DMA boundaries, which would reduce the number of required flow meters, lest they degrade the security of supply through the network. This makes finding efficient designs for network partition all the more important. Numerical optimization methods can help with this task. Within the joined research program of the Dutch dune water companies, KWR added functionality to their in-house optimization tool Gondwana to employ evolutionary algorithms for DMA design. Here, we report the first implementation of this functionality in practice. KWR worked together with the utility Dunea to design DMAs for Dunea’s largest and most complex distribution network: the city of The Hague.

to

Analysis of sensitivity to time step when modelling a dual function rainwater management system

Vanessa Speight

Presenter:

Authors:

Emily Dalby, Virginia Stovin and Vanessa Speight

Rainwater management systems are becoming increasingly researched and implemented as a way of capturing and storing rainwater for single or dual function (i.e. rainwater harvesting and stormwater management). However, It is argued that the use of a daily time step provides false confidence in a system’s stormwater management capability due to time step sensitivity. Within this research, 2 fundamental algorithms, Yield After Spillage (YAS) and Yield Before Spillage (YBS), are implemented within a 1 demand 1 storage model scenario for a full 1 year period (2007) (Fewkes and Butler, 2000; Jenkins et al., 1978). The research focuses mainly upon the model’s sensitivity to time step with regards to peak flow and system performance in estimating total spillage and yield (Campisano and Modica, 2014). Peak flow is generated from an event period of rainfall and plotted against an estimated greenfield runoff rate of 5L/s/ha. It is shown that the time step used greatly affects the generated spillage result, with peak runoff from a 1-minute time step exceeding that of a daily time step by up to 10 times. The results are consistent with previous research undertaken by Fewkes and Butler (2000). The study suggests that YAS is a stronger algorithm than YBS, with little sensitivity to time step. From the results generated when modeling for dual function capabilities, time step is seen to alter the accuracy of generated results. It is suggested from the research conducted that any future research regarding the dual function capabilities of a rainwater management system models at an hourly time step or below, in a YAS algorithm.

to

Green infrastructures modelling - A new framework for the evaluation of model usability

Mayra Rodriguez

Presenter:

Authors:

Mayra Rodriguez, Guangtao Fu and David Butler

Modelling Green Infrastructures (GI) allows a more efficient application of these practices and provides evidence to better and further planning and policy development. Several reviews on GI models are available, however, models are generally classified based on traditional model typologies such as rainfall and flow routing method used, contaminant range and GI techniques included in the model. Although these typologies are relevant, a classification based on the model usability appears to be more pertinent. This study aims to provide a new framework for the classification of the different models available and a potential structure for model development based on the definition of usability and three key prevailing challenges on GI modelling: uncertainty, complexity and performance assessment.

to

Application and reliability evaluation of k-median model for optimal deployment of leak sensors by field experiment

Youngwook Nam

Presenter:

Authors:

Youngwook Nam, Yasuhiro Arai, Takaharu Kunizane and Akira Koizumi

Water leakage can be a cause of sink holes due to the weakening of the ground, in addition to financial damage caused by the cost loss of purified water for purification. Therefore, it can be said that it is best to know the location and size and take action before large-scale damage from continuous water leakage occurs. The problem is that most of the water pipes are buried underground, so it's difficult to figure out the form and size until they are drilled and confirmed directly. In order to solve this, various leak detection methods have been proposed, such as a method of virtual leakage simulation using EPANET, moving along the inside of a pipe and detecting by special equipment for detecting leakage, imaging a buried pipeline using a spectrometer. The authors proposed and reviewed optimized sensor placement models for efficient leak detection in the previous studies: Arai et al. (2016), Arai et al. (2017), and Nam et al. (2018). In this research, we plan to reduce the number of sensors required for leak detection to half by using the optimal placement model proposed in the above three papers, and to ensure economic efficiency. At the same time, we would like to verify whether half of the sensors can reliably detect leaks in the target area.

to

Smart Water Distribution Network to Every Rural House Hold in Telangana, India

Abhilash Kancharla

Presenter:

Authors:

Abhilash Kancharla, Naveen Naidu and Murari Raja Raja

Water distribution systems (WDS) are essential parts of urban infrastructure systems, as they deliver water from water sources to domestic, commercial, and industrial water users to maintain their daily activities. The design, construction, operation and maintenance of WDS are a huge challenge for engineers around the world. They are one of the most expensive public infrastructures works as they require a high level of capital investment for construction and recurring investment for maintenance. The aim of our research is to supply potable water to every household in sufficient quantity, quality and with required pressure using smart techniques like Geographical Information System and simulation software (EPANET). Pamapur village - A real-life large water distribution network has been designed by using the above smart water techniques. Pamapur is a large village located in kothakota Mandal of wanaparthy district in Telangana state, India.

to

Making water meters truly smart by adding additional features

Mirjam Blokker

Presenter:

Authors:

Mirjam Blokker

Standard water meters are being upgraded and then all of a sudden are said to be “smart meters”. Today the most common upgrade is that towards AMR (automatic meter reading), which basically means that the meter is read from a distance, and potentially has a higher reading frequency. This type of meter would hardly qualify as a smart meter. Water utilities are setting up business cases for upgraded water meters, and especially in the Netherlands the conclusion for residential metering is often that the cost is too high. The reason is that the only cost benefits that are being considered is related to the billing system. We believe that by adding more features to the water meter, it is possible to make the network of water meters truly smart and benefits in other fields will become apparent. The first step to identify the potential is a qualitative assessment of the potential benefits to several stakeholders of many different features that can be added to water meters.

to

Dynamic Clustering for Real-time Water Distribution Systems Management

Mengning Qiu

Presenter:

Authors:

Mengning Qiu and Avi Ostfeld

Water distribution systems (WDSs) are essential infrastructures in every city and town. The sizes and complexities of WDS networks are increasing rapidly as a result of the rapid rate of urbanization worldwide. Consequently, WDS network simulation, a basis in all WDS problems, has become increasingly challenging to compute. Therefore, a simplified representation of the existing WDS model can help the researchers and water utilities to (1) speed up the simulation and optimization process, (2) better understand the WDS network structure, and (3) allow real-time monitoring of the WDS network. One way to create such a simplified representation of a WDS network is by node clustering. To date, WDS networks are often statically divided into a number of clusters that are formed by examining the steady states of a network over the full extended period simulation (EPS). The clusters created this form, although more generalized, can be further simplified for a part of the EPS and later be broken down into smaller clusters. To this end, a dynamic clustering method is introduced in this paper by installing control devices in the system. This dynamic clustering method takes advantages of the real-time hydraulic information of the WDS and the clusters created this way can help the water utilities in terms of district meters partitioning, sensor placements, and contamination detection.

to

ENSEMBLE LEARNING FOR SHORT TERM URBAN WATER CONSUMPTION FORECAST

B Bharanidharan

Presenter:

Authors:

B Bharanidharan, M S Mohan Kumar and Ramachandran Parthasarathy

Efficient water resource planning is highly dependent on the accuracy of the water consumption forecast. Complications in water supply occur due to large variations in usage over a short period of time. Important concern in the practice of consumption forecast is the attention paid to modelling related issues. In this paper, random forest method and techniques such as conditional inference trees (CTREE) and Recursive partitioning (RPART) along with bagging are used for prediction of daily water consumption pattern. Correlation Coefficient (CC), Mean Absolute Percentage Error (MAPE), Normalized Root Mean Square Error (NRMSE) are used as performance measures. A decrease in forecasting error for RPART and CTREE is observed by using bagging techniques. Obtained results show that the RPART Bagging has better results among all tested methods.

to

Analysis of Water Demand Pattern by Household Characteristics using Automatic Meter Reading Data

Taehyeon Kim

Presenter:

Authors:

Taehyeon Kim, Jeewon Seo, Kibum Kim, Jinseok Hyung and Jayong Koo

Recently, in the field of waterworks, along with the 4th industrial revolution, information and communication technology(ICT) has been developed, an effort is made to build up a smart waterworks system to promote the efficient operation and economical maintenance of waterworks facilities, which leads to the research and development of devices(Kim et al., 2013). Especially, smart water meters based on automatic meter reading are introduced in many cities(Gulisano et al., 2014). For efficient water demand management, it is necessary to investigate the internal factors, including the housing type that may affect the water demand of each household, resident population by household, average income and family members and the external factors like weather information and local characteristics. This study conducted a survey through the questionnaire items in the following Table 1 in order to reflect the demographic and social characteristics of H City, the target area by household. This study analyzed the patterns of the use of water according to the characteristics of the users, analyzing the real-time water demand in the meter reading through a smart water meter based on this. In addition, this study classified water demand patterns, using the data of the automatic meter reading of water demand in each household and the factors of the demographic and social characteristics investigated through a survey. This study analyzed the cause for the difference in the characteristics of the classified water demand and calculated the basic unit of the standard water demand by each characteristic. It is expected that the water demand patterns according to the characteristics of the users presented in this study can be utilized as the baseline data for the determination of the design capacity of waterworks facilities and the management of water demand according to the prediction of the water demand in urban planning.

to

Pilot Implementation of a Hydraulic State Estimator for the Strategic Supply Network of Madrid (Spain)

Sarai Díaz

Presenter:

Authors:

Javier González, Sarai Díaz and Javier Campos

The water 4.0 transformation of water supply systems demands signal interpretation in order to convert all data provided by telemetry systems into real information about the network state. Hydraulic State Estimation (HSE) can be used to determine the most likely state of a supply system (in terms of pressures and flows) at any location within the network. This approach enables to gather all the information about the functioning of the system, including its physical definition, its flow governing equations, the associated consumption patterns and data of the telemetry system, considering different metrology and signal transmission errors. In this work, a pilot implementation of HSE is developed over a confined area of the strategic transport network of Madrid (Spain), 800 km long, with 126 DMAs and including 12 pumping stations, 16 water tanks and 120 pressure reducing valves. Such a complex network is useful to prove the approach implementation capacity. Results show the potential of HSE not only to estimate the hydraulic state of the system and its associated uncertainty, but also to detect anomalies (such as bursts, undeclared changes in the network topology, changes in control elements or sensor failures) and identify the most suitable positions for locating additional metering devices.

to

Comparison of SSPs and VSPs in terms of Generating Transient Flow in WDNs

Nazli Mehzad

Presenter:

Authors:

Nazli Mehzad, Keyvan Asghari and Mohammad Reza Chamani

Water Distribution Networks (WDNs) are one of the important infrastructures in development of urban areas. These systems must deliver enough water to consumers in due time with appropriate quality. A prosperous performance of WDNs depends on the proper functioning of the components including pumping stations. There are two types of pumps, namely, Single Speed Pumps (SSPs) and Variable Speed Pumps (VSPs). In previous studies, the advantages of using VSPs in pumping stations to increase the hydraulic reliability and decrease the excess pressures as well as the energy costs have been clarified in steady state flow. Frequent and rapid changes of pump status lead to rapid changes of flow conditions from one steady flow state to another which is called transient flow and influence the performance of WDNs. Fluctuations of pressure in transient flow are often neglected in the scheduling of pumping stations in WDNs. In this paper, the performance of WDN is evaluated in terms of generating transient flow by means of SSPs and VSPs. In this respect, the transient code proposed by Larock et al. (1999) is promoted to evaluate VSPs besides SSPs. In a simple WDN the use of VSPs can reduce the pressure's fluctuations compared to SSPs. Moreover, the pressure's fluctuations is very sensitive to the time when the ultimate speed of the pumps is reaching which is evaluated in this paper.

to

Using a full-scale facility to investigate discoloration in drinking water systems: Challenges and solutions after the first year of experiments at Queen’s University

Artur Sass Braga

Presenter:

Authors:

Artur Sass Braga, Rain Saulnier, Alexandria Cushing and Yves Filion

The first experiments in the Drinking Water Distribution Laboratory (DWDL) at Queen’s University (Canada) began in 2017 and the aim of these initial tests was to investigate water quality deterioration in drinking water distribution systems. The full-scale distribution laboratory facility was inspired by a similar large-scale laboratory developed by the Pennine Water Group at Sheffield University (UK), which can mimic the behaviour of a drinking water distribution system by circulating water between reservoirs and pipe loops, using similar pipe materials, diameters, flowrates and pressure levels of real networks. This study presents several challenges faced in the first year of experiments in the facility, and a number of innovative solutions to improve the quality of the experimental data. The laboratory has a great potential to contributes for discoloration phenomena understanding, but further research is needed to improve current methods and capabilities.

to

The Impacts of Spatially Variable Demand Patterns on Water Distribution System Design and Operation

Kegong Diao

Presenter:

Authors:

Kegong Diao, Robert Sitzenfrei and Wolfgang Rauch

Resilient water distribution systems (WDSs) need to minimize the level of service failure in terms of magnitude and duration over its design life when subject to exceptional conditions. This requires WDS design to consider scenarios as close as possible to real conditions of the WDS to avoid any unexpected level of service failure in future operation (e.g., insufficient pressure, much higher operational cost, water quality issues, etc.). Thus, this research aims at exploring the impacts of design flow scenarios (i.e., spatial-variant demand patterns) on WDS design and operation. WDSs are traditionally designed by using a uniform demand pattern for the whole system. Nevertheless, in reality, the patterns are highly related to the number of consumers, service areas, and the duration of peak flows. Thus, WDS are comprised of distribution blocks (communities) organized in a hierarchical structure. As each community may be significantly different from the others in scale and water use, the WDSs have spatially variable demand patterns. Hence, there may be considerable variability of real flow patterns for different parts of the system. Consequently, the system operation may not reach the expected performance determined during the design stage, since all corresponding facilities are commonly tailor-made to serve the design flow scenario instead of the real situation. To quantify the impacts, WDSs’ performances under both uniform and spatial distributed patterns are compared based on case studies. The corresponding impacts on system performances are then quantified based on three major metrics; i.e., capital cost, energy cost, and water quality. This study exemplifies that designing a WDS using spatial distributed demand patterns might result in decreased life-cycle cost (i.e., lower capital cost and nearly the same pump operating cost) and longer water ages. The outcomes of this study provide valuable information regarding design and operation of WDSs; e.g., assisting the optimal design.

to

Evaluating RADAR-based Precipitation Measurements – an Application of the Python-Library “Wradlib”

Junqi Mao

Presenter:

Authors:

Junqi Mao

Flash flood is an emergency at the hour timescale between a rainfall and the onset of flooding. Researchers assess flash floods using hydrological models driven by RADAR-based precipitation data. One of the major problems with these models is the practice of using biased or incorrect raw data, because inputting data without post-calculation miscalculates the simulation results. This practice is attributable to the shortage of dampening correction methods. We intended to offset this gap by introducing a new correction method, which applies and further extends a python toolbox called “wradlib”. We start by calibrating the original German Weather Service (DWD) reflectance RADAR product RY. Then, these calibrated results are compared with the reference YW dataset in the HiOS (Hinweiskarten Oberflächenabfluss & Sturzflut) project. This comparison aims to find reasons behind the differences in the precipitation inputs and thus provide more detailed information for three catchments affected by flash floods: Simbach am Inn, Stöckach and Kulmbach. As a result, a more precise precipitation dataset for these catchments is built for further research, together with a python tool acting on the creation of qualified input data. Discrepancies between RY and YW datasets are examined through time, location, cause, and the expected magnitude for further catchments. Subsequently, a mature evaluation system is developed that not only deals with the uncorrected RADAR data but also hopes to serve discrepancy analysis in the future.

to

Integration of blockchain and model predictive control for the operation of water systems

Úrsula Flores-Gama

Presenter:

Authors:

Úrsula Flores-Gama and José Agustín Breña-Naranjo

The opportunities that the innovations brought by the new digital technologies to the water sector such as internet of things, blockchain and artificial intelligence are particularly promising for the emerging economies. Costs and inefficiency of centralized water systems can be excessive, in consequence, emerging economies like the Mexican, will choose to develop and manage hydric systems in a non-conventional grid. Dynamic and data driven systems can help to integrate and optimize, pumps, valves and smart sensors; devices can communicate individually between them or to the cellphone of a user, send information instantly, share it in the cloud or with users, making it easier the development and management of these new systems [5]. The new technologies help to the acquisition, distribution, processing and analysis of the information related to the system variables, turning it more accessible and comprehensible to the citizens and the stakeholders. On the other hand, we have the optimization models for water systems, this are probably the most used tools to solve the planning and management problems in the local and regional level [1]. There are many different optimization algorithms that can and has been applied to the problems related with the management and operation of hydric systems [3]. Among the more common, linear and nonlinear programming [2], dynamic [1]. or predictive control [4], this last one has not been yet explored in the sector in Mexico making it an innovating tool for the country. Also, the use of the emerging blockchain technology has not been explored in our country in the hydric resource field, and globally there are few applications developed using this technology in this field. Therefore, this work explores the integration of blockchain technology with an optimization method, such as the predictive control model, to achieve a better operation of water systems.

to

Dynamics of Indonesia's Water System: A Modelling Approach

Korinus Nixon

Presenter:

Authors:

Korinus Nixon Waimbo, Dragan Savic and Fayyaz Ali Memon

Despite having abundant global freshwater, Indonesia will eventually face scarcity if the water is not properly managed. This is likely to happen because of the population growth, economic development and urban expansion. We have carried out a modelling exercise using system dynamics modelling technique aimed at reproducing the behaviours of the global water system in Indonesia considering the socio-economic factors and the interactions with energy and food sectors. The water demand in Indonesia is structurally influenced by the national income per capita, population, electricity production, and wetland area. Over the period of 2000 to 2015, national water demand increases from 114 billion m3 to about 210 billion m3 (i.e., 85% growth). Out of the total water withdrawn, about 85% is used for agricultural activities, 10% is used for domestic, and 5% is used for industry. Our model shows that water intensity in the domestic sector increases gradually overtime (i.e., from 65.2 m3/capita to 67.5 m3/capita over the period from 2000 to 2015). The industrial water intensity, on the other hand, is decreasing at a faster rate of about 4% per year on average (i.e., from 79 m3/MWh to 41.5 m3/MWh for the year 2000 to 2015). The average income elasticity of domestic water demand is about 0.05 percent while the average income elasticity of industrial water demand is about 66 percent. At the national level, Indonesia has no water scarcity issue, however, there might not be the case at the sub-regional level (e.g, by major islands). Therefore, assessing the water sector at the sub-national level is part of our future work.

to

Using decision trees to create a blockage likelihood model based on preceding rainfall

Sabrina Draude

Presenter:

Authors:

Sabrina Draude, Edward Keedwell, Emma Harris, Zoran Kapelan and Rebecca Hiscock

Blockage reduction is an ongoing challenge for water and sewerage companies (WaSC) in the UK. Fewer blockages mean increased customer satisfaction, lower environmental impact and reduced costs for WaSC. In this study, a blockage likelihood model was created based on asset and historical incident and rainfall ident data from a DCWW wastewater network. The methodology was applied to two different highly populated regions in the South East of Wales (Cardiff and the Merthyr Valleys). The two areas vary significantly in ground slope with the Merthyr Valleys located at the bottom of a steep hilly catchment and Cardiff area is fairly flat. The purpose of this paper is to describe how decision tree analysis was performed on the two different locations in order to determine the relationship between rainfall and sewer blockages across these two catchments. Decision trees have previously been used by Bailey et al [1] to create a predictive blockage likelihood model but without considering past rainfall as a potential explanatory factor. The study has successfully used decision tree modelling on a large dataset to prove that the preceding rainfall before a blockage affects the likelihood of blockage formation. The results show that in Cardiff the two preceding days before a blockage incident significantly effect whether a blockage will form. However in the Valleys the seven days before the incident were more likely to affect whether a blockage would form. Follow on work will incorporate blockage likelihood based on preceding rainfall to create a prioritised scheduling model for proactive maintenance on DCWW network dependent on the predicted rainfall.

to

Real-Time Flood Forecasting in a Large Catchment Using a Hybrid of Physically Based and Data-Driven Methods

Laura Wignall

Presenter:

Authors:

Laura Wignall, Slobodan Djordjevic and Albert Chen

Hydroinformatics tools such as flood models are used within flood management to assess risk, and mitigate against damages and loss, however real-time forecasting is considered underutilised. Urban catchments in developed regions have access to the data and resources required for a real-time flood forecasting system, however there are catchments which are considered flood prone, where it is not currently considered possible. This study presents a methodology for the implementation of a real-time flood forecasting system, using the Gandak River catchment in Nepal and Northern India as an initial case study. The region is classified as flood prone, and has one of the highest rural population densities in India with many people living on and around the floodplains. Currently the area experiences flooding caused by rainfall from South-West and North-East monsoons, as well as water released from the barrage located at the Indo-Nepal border. Freely available data sources have been used due to the lack of measured data available. The Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) and Landsat imagery produced by the U.S. Geological Survey were used to establish a 1D hydrodynamic model of the river network. In place of a physically based hydrological model, a data driven Artificial Neural Network model has been used to simulate rainfall runoff in the catchment using rainfall obtained from the National Aeronautics and Space Administration (NASA) Tropical Rainfall Measuring Mission (TRMM) Precipitation dataset. The study explores the potential for real-time flood forecasting within catchments were it may not yet be considered feasible, through the use of freely available data and data driven techniques.

to

Hydropower Potential from Water-Energy Storage in Water Distribution Systems

Fayzul Pasha

Presenter:

Authors:

Fayzul Pasha, Matthew Weathers and Brennan Smith

An essential question in urban water distribution systems (WDS) is whether to minimize the excess energy in the system or recover the excess energy as hydropower at some locations in the network after ensuring the minimum pressure requirements. For a given network considering its topographical variations and system layouts, the later may provide better option. Optimizing pump schedules to operate a water distribution system at minimum pressure requirement alone cannot ensure an energy efficient water distribution system, especially for a WDS that has a centrally located pump station and significant topographical variations. Demand uncertainty imposes additional constraints to that end. 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. The possibility of storing excess energy and recovering it as a hydropower needs to be tested. The theory of pumped storage where water is stored as potential energy to generate hydropower at the peak hours can be applied to quantify the hydropower potential from a water distribution system. The installation of pumped storage, which can be defined as water-energy storage can distribute the energy across the water distribution system efficiently to provide required pressure more uniformly minimizing the excess energy. Investigating how the number of water-energy storage affects hydropower potential from a given a water distribution system with its pump station size and location, system layout and topographic conditions, system demand and pressure requirements is the main focus of this study. The topographic conditions, system demand, and structural integrity of the system may limit the number of water-energy storage for hydropower generation purpose. Specifically, the objective of this study is to maximize hydropower generation from a given water distribution systems considering both design of the water-energy storage and operation of the water distribution systems.

to

ENHANCING FLOOD RISK ANALYSIS AT THE RIVER BASIN SCALE, BY COMPARING HYDRAULIC AND GEOMORPHOLOGIC APPROACH

Alireza Faridhosseini

Presenter:

Authors:

Alireza Faridhosseini, Faridani, Madany, Gibson and Raziyeh Farmani

Due to the uncertainty concerning the location of flow paths on active alluvial fans, alluvial fan floods could be more dangerous than riverine floods. This study was conducted on alluvial fan located in northeast of Iran using a combination of the hydraulic portion of the FLO-2D model, the CADDIES-2D model, and the new Geomorphic Flood Index (GFI), which can delineate the flood exposure areas and water depth at the same time. The results of each approach were obtained, which they produced three different results of depth (m), velocity (ms-1), and delineating floodplain area within the selected area. Thereafter, the results of hydraulic models and geomorphic flood method were evaluated and compared. The goal of this research was to introduce a simple but effective solution to estimate the flood hazards.

to

Real-time optimisation of a water supply network using model predictive control

Nilki Aluthge Dona

Presenter:

Authors:

Nilki Aluthge Dona, Aonghus Mc Nabola, Biswajit Basu and Himanshu Nagpal

This paper presents the application of linear model predictive control (MPC) combined with a network simplification strategy for the real-time control of a water supply network. Pumping water through a variable speed pump in a water distribution system can be considered as one of the most energy consuming activities given the fact that pumping is switched on for many hours of the day. Therefore minimizing the pump speed while also meeting demand and pressure constraints within a network can be one of the solutions to minimize the operational cost of water supply. Conducting this exercise in real-time is a challenging problem due to its non-linear and continuous nature (pump speed variation). The main aim of this paper is to control the pump speed while adhering to the demand of the network and simultaneously trying to control the level of water available in a storage tank. The proposed method can provide an optimal pumping strategy using linear model predictive control which is a control method used to control a water distribution network in real-time. Using this method the controller adheres to the aggregated demand of the network while controlling the pump speed and thus removing the need for non-linear programming techniques to solve non-linear flow-head equations. The EPANET-MATLAB Toolkit is used for simulation of the water network and it has the capability to solve the hydraulics within MATLAB. A total of 66% cost savings was observed using this methodology when it was applied to a small case-study network in rural Ireland. Real-time control of water networks, therefore, offers significant potential to reduce costs and environmental impacts over conventional methods of offline network optimization.

to

Development and Validity Assessment of Automatic Leakage Detection Application

Asala Mahajna

Presenter:

Authors:

Asala Mahajna, Sam van der Zwan and Rodolfo Alvarado-Montero

Assessing the validity of data assimilation approach coupled with optimization algorithm for automatic and timely detection of leakage hotspots in water distribution systems

to