Agathoklis Agathokleous

Risk-based asset management of water piping networks using neurofuzzy systems

Abstract The efficient and organized management of public utility networks is of paramount importance to a network’s viability and reliable functioning. One of the key components of a suitable network management strategy is the utilization of integrated risk analysis and asset management decision-support systems (DSS) that incorporate both the scientific aspects of risk-of-failure analysis for the network components but also the financial and socio-political parameters that are associated with the networks in study. The study reported on presents a neurofuzzy decision-support system for performing multi-factored risk-of-failure analysis and asset management related to urban water distribution networks. The study is based on two datasets (one from New York City and the other from the city of Limassol, Cyprus), analytical and numerical methods, and artificial intelligence techniques (artificial neural networks and fuzzy logic) that capture the underlying knowledge and transform the patterns of the network’s behavior into a knowledge-repository and a DSS. Among the findings reported on, is a methodology to assess the risk of failure in a network, the factors affecting the reliability of pipe segments, and a neurofuzzy approach to breakage-data analysis, stratification and maintenance prioritization. Pipe-breakage history, pipe material, pipe age, and pipe diameter are shown to be significant risk factors in urban water distribution networks.

Entropy-Based Sensor Placement Optimization for Waterloss Detection in Water Distribution Networks

The work presented herein addresses the problem of sensor placement optimization in urban water distribution networks by use of an entropy-based approach, for the purpose of efficient and economically viable waterloss incident detection. The proposed method is applicable to longitudinal rather than spatial sensing, thus to devices such as acoustic, pressure, or flow sensors acting on pipe segments. The method utilizes the maximality, subadditivity and equivocation properties of entropy, coupled with a statistical definition of the probability of sensing within a pipe segment, to assign an entropy metric to each pipe segment and subsequently optimize the location of sensors in the network based on maximizing the total entropy in the network. The method proposed is a greedy-search heuristic.

Vulnerability of Urban Water Distribution Networks under Intermittent Water Supply Operations

The work presented herein investigates the effects of intermittent water supply (IWS) on the condition and breakage rate of urban water distribution piping networks (WDN), by studying the change in the rate of occurrence of failures before, during and after IWS periods, using statistical and survival analyses. The analyses, based on a seven-year dataset (2003–2010) from a major urban center of about 300,000 residents, take into account information related with breakage incidents and with operating system parameters, as well as external factors and vulnerability assessments of the network’s key components. The results show an increase in the number of waterloss incidents during and immediately after the periods during which IWS practices were implemented, and they reinforce the belief that IWS practices negatively affect the vulnerability of WDNs.

Topological Robustness and Vulnerability Assessment of Water Distribution Networks

The reliability of a water distribution network (WDN) is a function of several time-invariant and time-dependent factors affecting its components and connectivity, most important of which have been shown to be the network’s topology, its operating pressure, the type of key components (such as the diameter, length, material and age of water pipes) and the network’s historical performance (such as the number of previously observed failures in the network). In terms of network topology, this attribute even though generally thought as time-invariant it actually is time-dependent, as the paths in a water distribution network change over time based on the hydraulics in the network (water demand and water pressure/flow alter the way water flows in the piping network). The work described herein examines the time-dependent nature of a WDN topology and by means of a betweenness centrality index (BC) method demonstrates the effect of topology on the network’s vulnerability / reliability. The importance of the betweenness centrality index is demonstrated by use of a case-study water distribution network operated under both normal and abnormal conditions. The proposed method is also coupled with spatial mapping to indicate areas of concern in the network, and with a decision support system to assist in prioritizing actions to improve on the network’s robustness and resilience.

Waterloss Detection in Water Distribution Networks using Wavelet Change-Point Detection

The work presented herein addresses the automatic detection of water losses in water distribution networks (WDN), through the dynamic analysis of the time series related to water consumption within the network and the use of a wavelet change-point detection classifier for identifying anomalies in the consumption patterns. The wavelet change-point method utilizes the continuous wavelet transform (CWT) of time-series (signals) to analyze how the frequency content of a signal changes over time. In the case of water distribution networks the time-series relates to streaming water consumption data from automatic meter reading (AMR) devices, at either the individual consumers’ level or at an aggregated district meter area (DMA) level. The wavelet change-point detection method analyzes the provided time-series to acquire inherent knowledge on water consumption under normal conditions at household or area-wide levels, to then make inferences about water consumption under abnormal conditions. The method is demonstrated on several abnormal WDN operating conditions and anomaly detection cases.

Disaster Resilience of Water Distribution Networks

Water distribution networks are infrastructure threatened by natural or man-made disasters. WDNs are primary lifelines that should be able to provide their full or partial services immediately after a disaster. Designing a WDN taking into consideration its resilience and robustness is a must today. Moreover, WDNs are part of a citys infrastructure system and thus are interconnected to all essential lifeline networks. Recent studies of failures highlight the increased vulnerability of lifelines due to their interconnectivity and the need to consider mutually dependent network properties in designing resilient systems.

Chapter 3 – Vulnerability Assessment of Water Distribution Networks Under Abnormal Operating Conditions and Nonseismic Loads – The Case of Intermittent Water Supply (IWS)

The chapter presents, as a reference to the increased levels of vulnerability stemming from abnormal operating conditions, the case of intermittent water supply (IWS) and of related mathematical models. The chapter starts with a basic statistical and proceeds with a more detailed analysis, by use of the survival, proportional hazard rate, and regression tree methods. The analysis covers both normal and abnormal operating WDN conditions, varying levels of data complexity, and various WDN component classes. The goal is to demonstrate how historical records can be processed with analytical and numerical models, to identify underlying data patterns, and to eventually assess the corresponding risk of failure for each network element.

Hydraulic Vulnerability Assessment of Water Distribution Networks

The chapter presents a methodology for the seismic and hydraulic assessment of the reliability of urban water distribution networks (WDN) based on general seismic assessment standards, as per the American Lifelines Alliance (ALA) guidelines, localized historical records of critical risk-of-failure metrics pertaining to the specific WDN under assessment, and hydraulic simulations using adapted EPANET models. The proposed methodology is applicable to WDN under either normal or abnormal operating conditions (such as intermittent water supply), and the assessment of reliability incorporates data of past nonseismic damage, the vulnerabilities of the network components against seismic loading, hydraulic modeling, and the topology of a WDN. The network reliability is subsequently assessed using Graph Theory, while the system reliability is calculated using Monte Carlo simulation coupled with a hydraulic analysis identifying various methodology options for hydraulic quantity evaluation and hence hydraulic network performance. The methodology proposed is demonstrated on a real-scale district metered area (DMA) in a city-wide WDN.

Chapter 8 – From Historical and Seismic Performance to City-Wide Risk Maps

Following the discussion on the vulnerability of water distribution networks stemming from normal and abnormal operating conditions, under both nonseismic and seismic loads, the chapter discusses spatio-temporal aspects in vulnerability analysis. Spatial analysis (in the form of vulnerability “heatmaps”) is used to demonstrate the effects of nonseismic historical performance of WDNs on their seismic vulnerability and to discuss the various implications of network connectivity on WDN vulnerability. Utilized in the proposed spatio-temporal analysis are the American Lifelines Alliance (ALA) guidelines for the seismic vulnerability assessment of water distribution networks and the proposed friendly amendments to them (provided in Chapter 5 ), incorporating historical performance data and survival analysis to localize the probability of failure per WDN component.

Vulnerability Assessment of Water Distribution Networks Under Normal (Continuous Water Supply, CWS) Operating Conditions and Nonseismic Loads

Abstract The chapter discusses the basic concepts of vulnerability assessment, from WDN component to network assessment, and provides a short introduction to survival analysis. Survival analysis is then utilized in the development of vulnerability curves for several WDN components (such as water mains and house connections) and various classes (such as various pipe materials, diameters, and number of previously observed breaks). The analysis focuses on WDN vulnerability under normal operating conditions, and it leads itself to vulnerability considerations for WDNs under abnormal operating conditions (discussed in Chapter 3 ).