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

2C Smart systems

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

Collaborative

Xu Wang

Chair:

to

Cyber-Physical System Security Open Challenges in Smart Water Network

Haitham Hassan Mahmoud

Presenter:

Authors:

Haitham Hassan Mahmoud and Wenyan Wu

Cyber-Physical Systems (CPSs) are the core drivers of the optimal networking of real-time monitoring water systems in the implementation of Water 4.0. This facilitates the intelligent networking of water users to maintain the sustainability of water infrastructure, but this will increase the need for mitigating the risks more than ever using cyber and physical security framework. Cyber-physical system Security (CPSS) has attracted a lot of research attention due to increasing the entry points for the attackers to the system as the result of the upgrading of the water networks into smart ones. There are several examples of cyber-physical system attacks directly reported to current IoT and smart sensor technology being adapted such as; the ransomware attack in the city of Atlanta in March 2018 and Ukraine attack in December 2015. Moreover, there are many other attacks that are expected to occur in water infrastructure such as Compromising remote sites, Hot pivots, Cell-phone WIFI, Stuxnet, etc [1]. Furthermore, a study in 2018 [2] showed that for every sixty seconds the cyber-crime costs more than $1.1 million and impacts more than 1,800 people along with affecting the infrastructure and the service, consequently. Thus, this paper reviews the potential attacks and its prevention approaches at each layer of the Smart Water Network (SWN) based on Internet-of-things (IoT). The attacks and Countermeasure approaches are explained, and the literature has supported this argument. Moreover, the impact of these attacks on SWN is discussed. Furthermore, a recommendation of security procedures for water utilities is proposed.

to

Using Hydrodynamic Modelling to Identify Optimal Locations to Install Local Real Time Control Systems within Urban Drainage Networks

Marco Eulogi

Presenter:

Authors:

Marco Eulogi, Sonja Ostojin, Pete Skipworth, Will Shepherd, James Shucksmith and Alma Schellart

Real-time control systems (RTC) are designed to operate and manage urban drainage networks to reduce CSO spill volumes and/or flood risk. CENTAUR is a cost effective RTC system that utilizes local sensing and autonomous Flow Control Devices (FCDs) inserted into existing infrastructure, to maximise the use of in-sewer storage during rainfall events. Whilst local in-sewer storage capacity can be considered for rapid assessment of potential FCD installation locations, this procedure does not consider the interaction of hydrodynamics and FCD operation. This paper proposes a novel modelling approach to test FCDs performance at different locations during a sewer flooding scenario. A sewer network selected for this study is simulated with EPA Storm Water Management Model, with the FCD operation simulated using a Matlab interface to replicate the system control algorithm. For each potential installation location, available in-sewer storage volume and flood reduction are calculated and ranked. This allows effectiveness of installation at each potential location to be quantified. A total of 375 potential FCD installation locations are identified and evaluated using the hydrodynamic methodology, with storage capacity ranging between 0.1 and 250 m3. The analysis of simulation results indicates that flood mitigation and in-sewer storage capacity within the network are positively correlated (correlation coefficient 0.82). However the achieved flood reduction is also significantly influenced by the hydraulic behaviour of the system. Whilst ranking of in-sewer storage of nodes in the network can be used to pre-screen potential FCD locations, network hydrodynamics should also be considered to determine optimum installation positions. The proposed methodology automatically evaluates such hydrodynamics to identify best locations to install the flow control device, and further can be used to investigate the interaction of multiple FCDs.

to

Event Management and Post Event Response Planning for Smart Water Systems

Eirini Nikoloudi

Presenter:

Authors:

Eirini Nikoloudi, Michele Romano, Zoran Kapelan and Fayyaz Memon

The water industry in the UK and worldwide faces considerable challenges in making effective use of sensor and other data that is collected in water distribution systems in near real-time. This data is still not effectively utilised in a water company’s control room, especially when it comes to identifying a suitable strategy to respond to failure events in near real-time. Relevant academic work has not adequately addressed this challenge mainly due to the focus on specific stages (i.e. isolation, impact assessment or interventions) rather than the overall response process. In this study, an overall response methodology that aims to address the above limitation is proposed. It also suggests improved features comparing to previous related work, such as improved impact assessment based on a number of practical indicators, realistic intervention options as applied in the UK water industry, consideration of both operator proposed and automatically generated response strategies and a standardised way of visualising the impact of every response strategy. The methodology is integrated into a decision support -type tool in order to enable interaction between human and machine and hence facilitate decision-making in the control room. The methodology of the present study is demonstrated on the real network of B-City. The results suggest that the methodology successfully supports the operator in the control room, because it enables comparison between different automatically generated and operator proposed response strategies and/or hybrid response strategies resulting from the human-machine interaction.

to

A Real-time Response Mechanism to Cyber-physical Attacks on Water Distribution Systems

Riccardo Taormina

Presenter:

Authors:

Riccardo Taormina and Stefano Galelli

Digitalization is revolutionizing urban water infrastructure, granting higher level of service and other significant advantages on assets management and maintenance. Unfortunately, the network of intelligent devices used for process monitoring and control offers a vast attack surface that can be exploited by hackers to cause severe harm to both end-users and physical assets. The increased awareness on the potential vulnerability of urban water systems has led to recent efforts by the research community to secure Water Distribution Systems (WDS), which are particularly vulnerable due to their distributed nature. In this work, we advance the state-of-the-art on WDS security by developing a solution for the recovery phase—a necessary step needed to control a WDS once an attack has been detected. More specifically, we contribute an automated mechanism for detecting cyber-physical attacks and easing the recovery process through the real-time control of the WDS actuators (e.g., pumps, valves). The end goal is to minimize the damage to assets and end-users via timely attack detection and response. The proposed framework consists of an attack detection mechanism and a simulation-based optimization scheme. Both elements are based on Deep Learning Neural Networks, which grant fast, yet accurate, representation of the physical processes. The search for the optimal response strategy is carried out by an evolutionary optimization algorithm which can be set to pursue different objectives such as the minimization of water supply shortages or the number of people exposed to a contamination event. The proposed framework is tested on synthetic case studies involving different network maps to evaluate how the proposed approach performs for different types of threats and scales up to large networks. The results obtained on the synthetic cases studies are further discussed to illustrate the applicability of the proposed methodology to real-world settings

to

Valve status detection using smart meter temperature and flow

Jonas Kjeld Kirstein

Presenter:

Authors:

Jonas Kjeld Kirstein, Klavs Høgh, Martin Rygaard and Morten Borup

Up-to-date knowledge about the status and location of valves in water distribution networks is of high priority for many utilities. A successful detection is, however, often limited by the amount of resources available. Within recent years, utilities have installed smart meters at an increasing rate, and the collected data, including e.g. temperature measurements, provide utilities with large amounts of additional information about their systems. As the drinking water is heated up or cooled down through the network, temperature measurements contain valuable information about the water’s path to the consumer. In this study, a hydraulic network model was combined with a heat transfer model of a district metering area in Copenhagen, Denmark. First, temperature data was simulated for all nodes in the network by closing five valves and using real flow and temperature measurements at the inlets. Next, a novel optimization procedure was run to see if the differences in measured and simulated temperatures over the network’s pipes could be used to identify the closed valves. Starting with an assumption of all valves being open, the introduced optimization procedure identified all closed valves successfully solely based on temperature measurements. Moreover, the optimization procedure converged at a significantly faster rate than when applying a genetic algorithm. The ongoing work assesses the effect and uncertainties that appear in a real case study of two district metering areas, in both cases with more than 500 smart meters installed.

to