EPA/600/R-19/174 | October 2019
www.epa.gov/homeland-security-research
United States
Environmental Protection
Agency
oEPA
Future of Water Distribution
Modeling and Data Analytics Tools
Office of Research and Development
Homeland Security Research Program

-------
Future of Homeland Security Research
Program's Water Distribution Modeling
and Data Analytics Tools
by
Terranna Haxton, Robert Janke, Regan Murray, and Jonathan Burkhardt
U.S. Environmental Protection Agency
Cincinnati, OH 45238
Walter Grayman
W.M. Grayman Consulting Engineer
Oakland, CA
Hiba Ernst
Retired, U.S. Environmental Protection Agency
Cincinnati, OH 45238
Contract EP-C-12-014
Homeland Security Research Program
Cincinnati, OH 45238

-------
EPA/600/R-19/174
October 2019
Disclaimer
The U.S. Environmental Protection Agency (EPA) through its Office of Research and
Development funded, managed, and collaborated in the research described herein under contract
EP-C-12-014 to Aptim and contract EP-C-17-023 to Versar, Inc. It has been subjected to the
Agency's review and has been approved for publication. Note that approval does not signify that
the contents necessarily reflect the views of the Agency. Any mention of trade names, products,
or services does not imply an endorsement by the U.S. Government or EPA. The EPA does not
endorse any commercial products, services, or enterprises. The contractor role did not include
establishing Agency policy.
A portion of the research described here included interviews, which were done in accordance
with the Paperwork Reduction Act. In addition, the interviews and the panel workshop were
conducted as a peer review effort to obtain comments, suggestions, and new ideas from a broad
and diverse group of reviewers.
11

-------
EPA/600/R-19/174
October 2019
Table of Contents
Disclaimer	ii
Abbreviations	v
Acknowledgements	vi
Executive Summary	vii
1	Introduction	1
2	HSRP's Vision for Situational Awareness Tools	3
3	Water Distribution System Modeling Tools	4
4	Expert Interviews Prior to Workshop	6
5	Future Trends	8
6	Engagement with EPANET Community	11
7	Expert Peer Review Workshop	12
7.1	Panel Charge Questions	12
7.2	Workshop Agenda	13
7.3	Workshop Discussions	13
7.3.1	Future High-Level Directions and Needs	13
7.3.2	Brainstorming Future of the Drinking Water Industry	14
7.3.3	Experts' Suggestions for HSRP's Future Modeling Direction	15
7.4	Experts' Take Away Messages	15
8	Project Conclusions for HSRP Tools	17
8.1	TEVA-SPOT	18
8.2	CANARY	18
8.3	EPANET Multi-Species extension (MSX)	18
8.4	EPANET Real-Time extension (RTX)	19
8.5	Water Security Toolkit (WST)	19
8.6	Water Network Tool for Resilience (WNTR)	19
8.7	Premise Plumbing	19
9	References	21
10	Appendix	25
10.1	Modeling Terms and Definitions	25
10.2	Potential Tool Enhancements	26
10.2.1	Threat Ensemble Vulnerability Assessment - Sensor Placement Optimization Tool
(TEVA-SPOT)	26
10.2.2	CANARY	27
in

-------
EPA/600/R-19/174
October 2019
10.2.3	EPANET Multi-Species extension (MSX)	28
10.2.4	EPANET Real-Time extension (RTX)	28
10.2.5	Water Security Toolkit (WST)	29
10.2.6	Water Network Tool for Resilience (WNTR)	29
10.2.7	Premi se Plumbing	30
10.3 Future of Drinking Water Brainstorming List	31
iv

-------
EPA/600/R-19/174
October 2019
Abbreviations
AWWA	American Water Works Association
CWS	contamination warning system
DBP	disinfection byproduct
EPA	U.S. Environmental Protection Agency
EWRI	Environmental and Water Resources Institute
GIS	geographic information system
HSPD	Homeland Security Presidential Directive
HSRP	Homeland Security Research Program
MSX	EPANET Multi-Species extension
PPD	Presidential Policy Directive
RTX	EPANET Real-time extension
SCADA	supervisory control and data acquisition
TEVA-SPOT	Threat Ensemble Vulnerability Assessment - Sensor Placement
Optimization Tool
WNTR	Water Network Tool for Resilience
WST	Water Security Toolkit
v

-------
EPA/600/R-19/174
October 2019
Acknowledgements
In addition, the contributions provided by the interviewees, workshop panelists, and workshop
participants are acknowledged, including:
•	Steve Allgeier, U.S. Environmental Protection Agency (EPA), Office of Water
•	Dominic Boccelli, formerly University of Cincinnati
•	Steve Buchberger, University of Cincinnati
•	Robert Clark, EPA, Office of Research and Development (retired)
•	James Cooper, Arcadis
•	Jerry Edwards, Bohannan Huston
•	Biju George, DC Water
•	James Goodrich, EPA, Office of Research and Development
•	Ed Hackney, SUEZ
•	John Hall, U.S. EPA, Office of Research and Development
•	Laura Jacobsen, formerly Las Vegas Valley Water District
•	Sri Kamojjala, Las Vegas Valley Water District
•	Sudhir Kshirsagar, Global Quality Corp
•	Kevin Lansey, The University of Arizona
•	Mark LeChevallier, formerly American Water Works Company Inc.
•	Yeongho Lee, Greater Cincinnati Water Works
•	Morris Maslia, formerly Centers for Disease Control and Prevention
•	Sean McKenna, IBM Research
•	Johnathan Moor, Northern Kentucky Water Di stri ct
•	Lindell Ormsbee, University of Kentucky
•	Lew Rossman, formerly EPA, Office of Research and Development
•	Kenneth Rotert, EPA, Office of Water
•	Elad Salomons, water resources and systems consultant
•	William Samuels, Leidos
•	Feng Shang, formerly Innovyze
•	Vanessa Speight, Latis Associates/University of Sheffield
•	Jeff Szabo, EPA, Office of Research and Development
•	Michael Tryby, EPA, Office of Research and Development
•	Jim Uber, CitiLogics
•	Deborah Vacs Renwick, EPA, Office of Water
•	Thomas Walski, Bentley Systems
vi

-------
EPA/600/R-19/174
October 2019
Executive Summary
The U.S. Environmental Protection Agency's (EPA) Homeland Security Research Program's
(HSRP) mission is to focus on the research questions, needs, and response capabilities to help
communities deal with environmental catastrophes. One research objective is improving the
ability of water utilities to prevent, prepare for, respond to and recover from water contamination
incidents that threaten public health. To assist in this objective, HSRP has been developing
research prototype software tools that can help the drinking water industry in analyzing the
security and resilience of water distribution systems to all emergency situations, including both
man-made and natural disasters.
In 2016, HSRP conducted a project to help review the existing water distribution system
modeling program and identify a path forward for their water distribution system modeling and
data analytical tools. The project focused on the state of the science, engagement with the water
sector, and functionality improvements for the tools. To assist in outlining a path forward, the
project involved conducting interviews with the water distribution system community, reviewing
other resources that identified future directions in the drinking water field, and hosting a
workshop with experts to obtain their suggestions for the future. Representatives from the
drinking water sector, including utilities, professional organizations, software vendors,
consultants, academic researchers, and government employees, were included in the interviews
and workshop.
The water distribution modeling and data analytical tools that HSRP and their partners have
developed to help the water community throughout an emergency include Threat Ensemble
Vulnerability Assessment - Sensor Placement Optimization Tool (TEVA-SPOT), CANARY,
EPANET Multi-Species extension (MSX), EPANET Real-Time extension (RTX) libraries,
Water Security Toolkit (WST), and Water Network Tool for Resilience (WNTR). Research is
also being conducted to better simulate and understand water quality within the plumbing system
within buildings. The tools are described in more detail in the report section, Water Distribution
System Modeling Tools.
The first step in the review was a series of interviews with subject matter experts from the
drinking water community (e.g., drinking water utilities, engineering consultants, software
developers/vendors, government, academia) that represented a diverse range of backgrounds,
positions, and perspectives within the water industry and the government. The interviewers
provided their observations on the barriers and obstacles to using water security and resilience
tools, the technical gaps of the tools and/or research needs, and collaboration opportunities
between HSRP and the water community. A more detailed summary is presented in the report
section, Expert Interviews Prior to Workshop. The following are examples of the provided
suggestions:
•	Software tools need to be easy to install and use, since water utilities often do not have
enough staff or time.
•	Water security software tools need to support the day-to-day operations of the utility in
addition to water security.
•	More basic research on the fate and transport of contaminants within the distribution
system is needed.
vii

-------
EPA/600/R-19/174
October 2019
•	A calibrated network model is important to integrate with hydraulic and water quality
monitoring data to provide a better understanding of dynamics within the distribution
system.
•	EPA should connect, talk, and work directly with water utilities to encourage the usage of
HSRP'stools.
The second part of the project involved identifying future trends in the industry. This included a
series of Environmental and Water Resources Institute (EWRI) committee interviews with water
distribution system analysis professionals; a report on trends in water distribution system
analysis by an American Water Works Association (AWW A) committee; and an article on a
vision for what water distribution systems might look like in the year 2050. These resources
identified that aging water infrastructure, infrastructure assessment, real-time modeling and
advanced analytics, and research/innovation are topics of interest in this field. In addition,
recommendations for the vision and future development of EPANET was collected during the
EPANET Visioning Summit held in Reston, Virginia on April 3-4, 2018, since EPANET is an
important component of many of HSRP's water security and resilience modeling tools.
The final part of the project involved hosting a workshop with a panel of experts and other
participants. On April 9-10, 2018, EPA held the Water Security, Response, Resilience Modeling
and Data Analytic Tools Workshop. The purpose of the peer-review workshop was to obtain
additional comments, suggestions, and new ideas for HSRP's water security and resilience tools
to assist in the development of a path-forward strategy for the tools to support water utility
needs. Five drinking water subject matter experts formed the panel, while other water
professional experts participated in the discussions. The workshop provided an opportunity to
share information regarding the state of the science for drinking water system security and
resilience. The workshop included presentations about HSRP's modeling program, discussions
on the drinking water industry's future directions and needs, presentations and discussions of
HSRP's modeling tools, and suggestions for the future of these tools. A few of the suggestions
are listed below, while additional suggestions are provided in the report section, Expert Peer
Review Workshop.
•	Focus on the day-to-day operation of utilities more than on rare incidents like terrorist
acts and natural disasters.
•	Seek better input and feedback from different stakeholders (e.g., large utilities, small
utilities, consultants) throughout the development process as they have different needs.
•	Continue development of fundamental tools that can be adapted by software companies
to provide enhancements and roll out to utilities.
Overall, this project highlighted some commonality within the drinking water community. One
key message was to ensure the accuracy of the network model since it needs to reflect real
operating conditions within the water distribution system. This can often be accomplished by
creating a real-time model to depict the actual system dynamics by linking system hydraulic data
(e.g., flows, pressures, tank levels) with the network model. The review highlighted that real-
time modeling is a direction that the water industry is moving towards. Another key message was
the need for accurate representation of the water quality within the distribution system. This can
be achieved by collecting more experimental data on the fate of different water constituents to
understand how they react with one another, developing more detailed fate/reaction equations,

-------
EPA/600/R-19/174
October 2019
and linking this information with water quality data collected in the distribution. The goal would
be providing a more complete understanding of the water quality in the system in real-time.
Collaboration was another focus area identified during the project. Experts emphasized that it
was important to engage with the water community through all phases of research to understand
the problems that the industry is facing, develop solutions for these problems, and transfer the
solutions to the industry. In addition, the need for user-friendly modeling and data analytic tools
that assist with the day-to-day operation of the water distribution system was highlighted.
Specific suggestions for the improvements for each individual tool are provided in the Project
Conclusions section of the report.
IX

-------
EPA/600/R-19/174
October 2019
1 Introduction
National security has been a responsibility of U.S. Environmental Protection Agency's (EPA) for
a long time. Executive Order 12656 (signed November 1988), titled "Assignment of Emergency
Preparedness Responsibilities," directs EPA to take on two responsibilities: (1) develop guidance
on acceptable emergency levels of chemical, nuclear, and biological materials and (2) develop
plans to ensure the availability of potable water after a national security incident. Other
legislative acts have also shaped EPA's role in responding to national security emergencies,
including the Comprehensive Environmental Response, Compensation and Liability Act, the
Emergency Planning and Community Right-to-Know Act, and the Clean Water Act, as well as
others. Shortly after the events of September 11, 2001, the National Strategy for Homeland
Security (OHS, 2002) was developed, which outlined specific objectives for protecting critical
infrastructure, defending against catastrophic threats, and preparing and responding to
emergencies.
EPA's responsibilities for water security were further mandated in the Public Health Security
and Bioterrorism Preparedness and Response Act (Bioterrorism Act) of 2002, and a series of
Homeland Security Presidential Directives (HSPDs). Three HSPDs directly affect EPA's role in
national emergencies, namely HSPDs-5, 7, and 8, which deal with national incident management
system and response, critical infrastructure prioritization and protection, and national
preparedness, respectively. HSPD-9, titled "Defense of United States Agriculture and Food,"
established a national policy to defend the agriculture, food, and water systems against terrorist
attacks, major disasters, and other emergencies. HSPD-9 required EPA to develop and support
intelligence operations and analysis capabilities for the water sectors. Requirements included
support for surveillance and monitoring systems for water quality and new countermeasures for
detecting contaminants in the water supply. More recently, Presidential Policy Directive-21
(PPD-21) on Critical Infrastructure Security and Resilience advanced a national effort to
strengthen and maintain secure, functioning, and resilient critical infrastructure against all
hazards. PPD-21 identifies 16 critical infrastructure sectors and their assigned sector specific
agency. EPA is the assigned lead agency for the water and wastewater systems sector.
As a result of the September 11th terrorist events and the subsequent anthrax attacks in New York
City, Washington D.C., and Florida, EPA established the Homeland Security Research Program
(HSRP) within the Office of Research and Development in 2002. HSRP conducts applied
research by developing systems-based technology solutions to increase the capability of EPA to
achieve its homeland security responsibilities as identified in statues and presidential directives.
HSRP recognized a thorough understanding of the nature of biologicals, chemicals, and
radiological agents, their transport in the environment, and their effects on human health was a
critical research need to help ensure critical infrastructure was secure against threats.
In terms of the water sector, a major focus of HSRP has been the development of research
prototype software tools that can be used by the water industry to analyze water contamination
threats to drinking water distribution systems. The water community recognized the vulnerability
of drinking water distribution systems to contamination threats. The initial software tools
quantified the potential impacts or consequences to public health from contamination releases
within the drinking water distribution system and then used the results to optimally determine
1

-------
EPA/600/R-19/174
October 2019
where best to place sensor monitoring stations to reduce public health impacts or decrease the
detection time of the contamination incident. The consequence and sensor placement
optimization software tools provided key capabilities for a water utility to design and implement
an online water quality monitoring system as a component of a contamination warning system
(CWS) to improve a community's response to a contamination threat. Later, accidental and
natural disasters, such as the Deepwater Horizon oil spill in 2010 and Hurricane Sandy in 2012,
demonstrated the need for communities to be better prepared for and able to recover from all
types of hazards. The "all hazards" approach seeks to provide technical support for dealing with
incidents, regardless of the type of contamination or cause (e.g., intentional or unintentional
physical disruption). The all hazards approach has been expanded to include improved resilience
against man-made and natural disasters. With the broader definition of threats, HSRP expanded
its water security tools to include the capability to analyze a broader range of threats.
In December 2016, HSRP undertook a project to conduct an evaluation and technical review of
their modeling tools and programs with an eye towards assessing and designing their future
research program. The review focused on the state of the science, how to encourage broader use
of water distribution system modeling and data analytical tools by the water sector, how to
improve the technical functionality of the tools, and how to foster partnerships with the
commercial sector. To provide a broad basis of views, the review engaged representatives of the
water sector, including water utilities, professional organizations, software vendors, consultants,
and academic researchers in this field.
The elements of the evaluation and technical review of modeling tools and programs included
the following:
•	Conduct external and internal interviews with a variety of professionals in the drinking
water industry to gather their comments on HSRP's water modeling program and any
research needs associated with improving the operations, water quality, or resilience of
drinking water systems
•	Review published literature to identify potential future directions within the water
distribution system analysis field
•	Identify potential enhancements for HSRP's existing water distribution system modeling
tools to support improved security and resilience of water systems
•	Engage with the EPANET developer and user community to identify overlapping areas of
interest since EPANET provides the hydraulic and water quality simulation engines for
HSRP's modeling and simulation software tools
•	Convene an expert panel to provide additional input to build on the findings from the
external and internal interviews
•	Summarize the results of the evaluation and technical review
•	Develop a path forward strategy for HSRP's water modeling research to continue the
advancement and delivery of its tools and methodologies to help improve the security and
resilience of the water community
The primary goal for performing the evaluation and technical review, as well as developing this
report, is to foster communication and collaborations within EPA and the wider water
community to develop a path forward for HSRP's water security and resilience tools.
2

-------
EPA/600/R-19/174
October 2019
2 HSRP's Vision for Situational Awareness Tools
To identify research gaps and to better understand potential security issues and concerns of the
U.S. water community, EPA engaged with numerous water experts, stakeholders from
government, industry, and academia. From these engagements, EPA developed the Water
Security Research and Technical Support Action Plan to summarize the needs and define
research projects to improve the security of the nation's water and wastewater systems (U.S.
EPA, 2004). The needs were organized across the following seven recommended focus areas:
•	Protect drinking water systems from physical and cyber threats
•	Identify drinking water threats, contaminants, and threat scenarios
•	Improve analytical methodologies and monitoring systems for drinking water
•	Contain, treat, decontaminate, and dispose of contaminated water and materials
•	Plan for contingencies and address infrastructure interdependences
•	Target impacts on human health and inform the public about risks
•	Protect wastewater treatment and collection systems
The American Water Works Association (AWW A) collaborated with HSRP to form the Water
Utility Users Group (Roberson and Morley, 2005; Morley et. al., 2007; Janke et al., 2011). This
group was comprised of utilities interested in developing monitoring systems to detect
contamination threats, and they partnered with HSRP in their development of prototype software
tools to help design, implement, and evaluate online water quality monitoring systems as a
component of a CWS. The partnership helped to ensure that HSRP was focused and grounded in
improving the security of water distribution systems by developing relevant tools and
methodologies that could be used by the water community (Morley et al., 2007).
Through these engagements, HSRP and the wider water community recognized that to achieve a
timely, effective response to contamination threats, a real-time understanding of water system
operations across the distribution system during an incident was needed. This capability is called
"situational awareness." As a major water contamination incident or significant emergency
unfolds, hydraulic and water quality conditions in the distribution system can change quickly and
dramatically. Therefore, it is important to accurately understand in real-time the changing
hydraulic and water quality behavior of the water distribution system during such emergency
events. This understanding or situational awareness requires a continuous collection, assessment
and assimilation of critical, real-time data and information.
Because of the complexity of water distribution systems, network models are commonly used to
predict flow directions and water quality at different locations and times. Tools using these
network models were widely recognized as needed functionality to support situational awareness
and response efforts. As such, HSRP developed a plan to develop tools built on top of hydraulic
modeling capabilities that could be combined to provide a real-time, situational awareness of
drinking water distribution system behavior. The situational awareness tools would support more
effective and rapid decision making, and other specific tools would help in planning for incidents
ahead of time (e.g., where to locate sensors), collecting and interpreting data during an incident
3

-------
EPA/600/R-19/174
October 2019
(e.g., analyzing sensor data) or in response to incidents after they had occurred (e.g., where to
flush contaminants from the system).
The 2013 Roadmap to a Secure and Resilient Water Sector (CIPAC Water Sector Strategic
Priorities Working Group, 2013) identified and prioritized activities needed to improve the
security and resilience of the drinking water and wastewater infrastructure. One of the top
priority activities listed in the roadmap is to "support the development and deployment of tools,
training, and other assistance to enhance preparedness and resiliency" of water infrastructure
systems, including the work being developed by HSRP. A revised roadmap, Roadmap to a
Secure and Resilient Water and Wastewater Sector (Water and Wastewater Sector Strategic
Roadmap Work Group, 2017), provided updates on previous activities and identified priority
activity areas for the next five years. These roadmaps helped to extend HSRP's original plans
which were focused on contamination incidents to more broadly address resilience to all types of
disasters.
3 Water Distribution System Modeling Tools
Since its inception in 2002, HSRP's vision for the water distribution system modeling program
was to develop methodologies and tools for the nation's drinking water systems to help make
them safer and more resilient. HSRP researchers and their partners have developed systems
analysis methodologies including mathematical modeling, simulation, optimization, data
analytics, and software engineering that can be applied to drinking water systems. Through these
efforts, HSRP has created research prototype software tools to help the water community
throughout the continuum of an emergencies. These software tools were developed based on the
paradigm that water system security, resilience, and response capabilities are dependent on: (1)
identifying threats and vulnerabilities, (2) designing, developing, and implementing engineering
systems that can quickly detect such threats, and (3) developing and deploying tools and
methodologies that can be used to help minimize the response and recover times following a
water system attack and/or disruption. These prototype software tools include the Threat
Ensemble Vulnerability Assessment - Sensor Placement Optimization Tool (TEVA-SPOT),
CANARY, the EPANET Multi-Species extension (MSX), the EPANET Real-Time extension
(RTX) libraries, the Water Security Toolkit (WST), and the Water Network Tool for Resilience
(WNTR). Additionally, efforts are underway to better understand water quality after water leaves
the distribution system and enters the premise plumbing systems of the community, such as
residences, businesses, schools, and hospitals. The Appendix provides some terms and
definitions associated with water distribution system modeling. A short description of each
software tool is provided below.
TEVA-SPOT enables users to evaluate threats and consequences from a contaminant release into
a drinking water distribution system. Using the results from a site-specific consequence
assessment, TEVA-SPOT allows the user to investigate locations within the distribution system
where best to place real-time monitoring or sampling stations to help mitigate contamination risk
to public health and to support the development of an emergency response plan in the event of a
contamination incident. TEVA-SPOT also allows users to evaluate existing site-specific
monitoring plans to determine their effectiveness in addressing a wide range of user defined
4

-------
EPA/600/R-19/174
October 2019
threats through the software (U.S. EPA, 2010a; 2016). TEVA-SPOT's results are dependent on
the accuracy and the specific utility operation's snapshot (e.g., maximum, minimum, or average
daily water demands) of the network model used in the analysis. TEVA-SPOT includes a
graphical user interface, but a significant learning curve for understanding and effectively using
TEVA-SPOT is still necessary.
CANARY is an event detection software that water utilities can use to analyze sensor data, such
as chlorine, pH, oxidation reduction potential, total organic carbon, and conductivity, from
typical online water quality monitoring sensors and report when a period of anomalous water
quality is detected in the distribution system. The software uses several statistical analysis
algorithms to evaluate real-time water quality sensor data that is polled every 2 to 15 minutes and
to identify outliers from an established baseline. CANARY was designed to be able to interact
with a supervisory control and data acquisition (SCADA) database, making it able to work with
sensors of any type or brand (U.S. EPA, 2010b; 2012; 2013; 2014a). CANARY can be used with
any sensor to analyze any time-series data. CANARY does not have a graphical user interface.
The multi-species extension to EPANET, MSX, allows users to write their own chemical or
contaminant fate routines (i.e., reaction rate equations) and to incorporate multiple species (e.g.,
interacting chemicals) in EPANET's water quality simulation. MSX can be used to simulate the
fate and transport of biological and/or chemical contaminants, as well as chemicals associated
with disinfection, such as chlorine, chloramine, and disinfection byproducts (DBPs), in a system
(Shang et al., 2008; 2011). MSX was released as a stand-alone, command line application,
without a user interface. MSX is not easy to use and requires the development of reaction kinetic
equations for any species that are to be simulated. Additionally, accurate MSX modeling relies
on an understanding of the water quality and character of the water distribution system (e.g.,
water constituents, extent of biofilms, disinfectants, pipe material) that is difficult to obtain.
RTX is different from the other prototype research tools since it is a set of C++ software libraries
that software developers can use to build real-time hydraulic and water quality modeling
applications. These applications can help improve water distribution system modeling, planning,
and operations. The RTX libraries extend the capabilities of EPANET by integrating the network
model with SCADA data to better reflect daily operational activities as well as operational
changes (Janke et al., 2011; Uber et al., 2014; U.S. EPA, 2015). One application is, RTX:LINK,
which provides utility managers and operators the opportunity to remotely view their streaming
real-time SCADA monitoring data and associated analytics. In addition, the RTX libraries were
developed to support HSRP's other modeling and simulation tools (e.g., TEVA-SPOT, WST,
WNTR) by improving the accuracy of the network models and providing the software
infrastructure for real-time response. Deploying RTX software tools to implement real-time
modeling within a water utility's operations will require a significant effort and possibly
resources. Water utilities will require software development capabilities for the deployment and
application of the RTX software tools.
WST is a suite of software codes (i.e., set of separate, command line executable programs) built
upon the EPANET platform that can be used to assist water utilities in trying to understand how
best to respond to a contamination incident in a water distribution system. WST includes
hydraulic and water quality modeling software and optimization methodologies to investigate
5

-------
EPA/600/R-19/174
October 2019
possible: (1) locations in the system in which the contamination was introduced, (2) locations to
flush contaminated water from the distribution system, (3) locations in the system to inject
chlorine to inactivate contaminants, and (4) locations in the system to take grab samples to help
identify the extent of contamination (U.S. EPA, 2014c). WST does not include a graphical user
interface and requires a significant learning curve, but it does not require software development
capabilities. WST results are dependent on the specific snapshot network model used in the
analysis. WST does not include linkage to real-time models, which would be needed for utility
deployment during an emergency.
WNTR is a Python software package that allows the user to design, simulate, and analyze the
resilience of water distribution systems given internal or external threats or incidents. The
software includes the capability to: (1) create and modify water network models; (2) assign
fragility and survival curves to network components; (3) simulate hydraulics and water quality to
model disruptive incidents (e.g., power outages, earthquakes, pipe breaks, and contamination
incidents) as well as response and repair strategies; and (4) analyze results and generate graphics
to evaluate resilience using a wide range of metrics. WNTR extends the capabilities of EPANET
by allowing for pressure dependent demand simulations, simulating pipe leaks/breaks of
different sizes, changing aspects of the network and/or operations during a simulation (i.e.,
stop/restart), and enhancing the hydraulic solver to be more robust during extreme failure
scenarios (Klise et al., 2017a; 2017b). WNTR is similar to WST in terms of the lack of a
graphical user interface and difficulty to use, since it requires some knowledge and experience
with Python. Similar to TEVA-SPOT and WST, the results of WNTR are dependent on the
specific operation's snapshot of the network model that was used in the analysis.
Premise distribution system modeling represents a new research area within the EPA. The
premise distribution system modeling research is focused on extending the capabilities of
EPANET to study issues such as lead and Legionella (the bacterial causative agent of
Legionnaires' disease and Pontiac fever) contamination as well as water age, a factor in water
quality degradation, in building premise plumbing systems (e.g., in homes, apartments) (Samuels
et al., 2010; Burkhardt et al., 2018). This effort has focused on two main areas: the addition of
dispersion modeling to EPANET and a Python-based script tool for scenario management. The
Python-based tool includes an agent-based demand generation and analysis component that is
used as part of a Monte Carlo based approach to better understand the impact of different factors
(e.g., usage patterns, plumbing layout) on exposure within premise plumbing systems. Additional
functionality is also being developed to understand flushing efficacy within premise plumbing
systems based on different flushing approaches.
4 Expert Interviews Prior to Workshop
A key element of the review process was the collection of ideas and viewpoints through a series
of interviews with subject matter experts representing a broad range of interests and perspectives
(e.g., drinking water utilities, engineering consultants, software developers/vendors, government,
academia). Nine interviews were conducted with external experts and an additional eight
interviews with internal experts (EPA staff and some of their working groups). The internal and
external interviewees represented a diverse range of backgrounds, positions, and perspectives
within the water industry and the government.
6

-------
EPA/600/R-19/174
October 2019
The interview questions were tailored to the specific experience and background of the
interviewee while structured to gain insight across the following topics: (1) barriers and obstacles
preventing water utilities from using the HSRP's water security and resilience tools, (2) technical
gaps in the tools or the water security program, and ideas for advancing the tools or the program,
and (3) strategies to improve collaboration between the program and water utilities and
commercial partners. A sample of the questions asked are provided below.
•	What are the barriers/obstacles to using HSRP's software tools (or tools produced by
other researchers or vendors)?
•	What are the technical gaps in HSRP's water distribution system modeling research
program? What improvements can be made to HSRP's software tools, methodologies, or
algorithms to address the most pressing needs of the water sector?
•	How can HSRP more effectively collaborate with the water sector (i.e., water utilities)
and private sector and industry partners?
The interviewees provided a wide variety of responses to the questions. Some of the key
messages from the interviews are summarized below in categories of barriers, research needs and
capabilities, and collaboration opportunities.
BARRIERS
•	Currently there are no regulations or requirements that mandate utilities to use HSRP's
water security and resilience tools. Similarly, utilities are not required to develop a
network model that is an accurate representation of the water distribution system and its
behavior.
•	Many utilities do not have the necessary SCADA and asset management systems to
adequately support the development and maintenance of a network model or real-time
model. Additionally, monitoring equipment is often not adequately calibrated nor
reliable. This is especially the case for small utilities.
•	At many water utilities, organizational silos exist between engineering, water quality, and
operations, which prevents them from effectively communicating and working together.
Engineering departments often develop and use the network model for only planning
purposes. Operations departments manage distribution system operations but generally
without the use or need for the network model.
•	Water utilities often do not have enough staff or time to use modeling software.
•	Modeling results are often not trusted to assist with making operational decisions due to a
lack of trust in the quality and accuracy of the network model to represent operations.
•	Water security software tools need to support the day-to-day operations of the utility and
not just remain focused on security related concerns.
•	Most water systems have not incorporated complex tools like TEVA-SPOT, WST, or
WNTR into their risk analysis of operations. Many utilities have not addressed
distribution system vulnerability or even water quality modeling.
•	Water utilities need software to be easy to install and operate. Good documentation
(technical and user guides and tutorials) must also be made available.
7

-------
EPA/600/R-19/174
October 2019
NEEDED RESEARCH AND CAPABILITIES
•	Foundational research and practical utility case studies demonstrating the application and
use of the water security and resilience tools are needed.
•	More basic research on the fate and transport of contaminants in the distribution system is
needed. The water community is still using basic methods even for predicting the fate of
chlorine and DBPs in the distribution system.
•	Importance of well-calibrated network models cannot be overstated. Water quality
modeling is dependent on a well-calibrated hydraulic model.
•	Integration of hydraulic and water quality SCAD A data with network models, real-time
analytics, and real-time modeling (including smart water technologies) are important
advances and technologies to provide a better understanding of what is really happening
in the distribution system. However, the selection and adoption of the most appropriate
technologies is a difficult task for most water systems.
•	Water security and resilience tools need to be enhanced to support normal operations and,
therefore, warrant the efforts required for water utilities to implement the software
programs into their operations. For example, TEVA-SPOT and WST could be modified
to examine and optimize the placement of continuous, online monitors to support routine
water quality monitoring and help meet regulatory requirements.
COLLABORATIONS WITH WATER COMMUNITY
•	Presence at national and international conferences is important.
•	Connecting, talking, and working directly with water utilities is important.
•	Publishing practical utility case studies demonstrating the water security and resilience
tools and methods would be useful.
•	Open source software and community software development projects can be very
effective, but they need support and guidance from EPA.
•	Water utilities need technologies that provide a return on investment. Utility case studies
that quantitatively demonstrate the benefits and cost savings from the use of the water
security and resilience tools would be useful.
5 Future Trends
The 2015 Water and Wastewater Systems Sector-Specific Plan (U.S. DHS and U.S. EPA, 2015)
provided the latest blueprint identifying the sector's goals and objectives to help ensure future
security and resilience of the sector. The plan outlined the complex governance, management,
and institutions that support and help to maintain the nation's water systems. The drinking water
community (and the wastewater community) is vast and complex, representing a multitude of
stakeholder segments and federal, state, and local government entities.
While the plan identified the sector's goals (e.g., "Goal 1: Sustain protection of public health and
the environment") and objectives to meet each goal (e.g., "Objective 1: Encourage integration of
both physical and cyber security concepts into daily business operations at utilities to foster a
8

-------
EPA/600/R-19/174
October 2019
security culture"), it did not specify the technology solutions and business practices that would
be needed to address them. Thus, the water community (i.e., water utilities, industry, consultants,
and academics) should develop them. This section examines the future trends in technologies and
business practices that could help shape the drinking water community into the future and help
meet the goals and objectives outlined in the plan.
One recent trend is the establishment of multiple water clusters, which are regional groups of
private and/or public organizations focused on innovative water technologies. Examples of the
water clusters that focused on drinking water technologies included Colorado Water Innovation
Cluster, Confluence, North East Water Innovation Network, and WaterStart. The technologies
emphasized by these clusters help address water efficiency, water reuse, water-energy-food
nexus, drinking water treatment, water infrastructure, monitoring and modeling, and smart water
(Wood et al., 2018). An initiative that is helping the water industry look towards the future is the
Utility of the Future Today recognition program that started in 2016. The goal of Utility of the
Future Today is help water utilities become more efficient, productive, and sustainable in their
operations. Utilities are recognized for implementing innovative programs and technologies that
help make them more resilient to challenges in their communities. The recognition categories
include water reuse, biosolids reuse, nutrient reduction and materials recovery, partnering and
engagement, energy generation and recovery, energy efficiency, and watershed protection (WEF,
2019).
EPA has also outlined ideas to help advance technology innovation in the water industry. In
"Promoting Technology Innovation for Clean and Safe Water: Water Technology Innovation
Blueprint, Version 2," water scarcity, water quality, aging infrastructure, climate change, and
water accessibility are identified as water challenges affecting our world today and into the
future. To help address these challenges, technologies are needed to conserve and recover
energy, recover nutrients, improve and green infrastructure, conserve and reuse water, enhance
water monitoring techniques, increase resilience of water infrastructure to climate change
impacts, reduce water usage by energy industry, and improve water quality in watersheds (U.S.
EPA, 2014b). Bluefield Research has also identified the top nine trends in 2019 for the water
industry. These trends included smart asset management technologies for leak detection, pressure
management, and workflow management, solutions to address infrastructure leakage, smart
meters that combine water and energy monitoring, decentralized water treatment to reuse water,
and real-time control solutions and water quality monitoring to increase resilience of systems to
climate change impacts. Bluefield also noticed more interest in the water industry for artificial
intelligence and machine-learning solutions as well as software-as-a-service and cloud-based
offerings (Tisdale, 2018).
To identify the likely future trends specifically in drinking water distribution system analysis,
operation, and design, three additional sources of information were examined. These sources
included: a series of interviews available on YouTube with professionals in the field of water
distribution system analysis conducted by an Environmental and Water Resources Institute
(EWRI) committee (EWRI, 2019); an AWWA committee report on trends in the field of water
distribution system analysis (AWWA, 2014); and an article on a vision for what water
distribution systems might look like in the year 2050 (Grayman et al., 2012).
9

-------
EPA/600/R-19/174
October 2019
Starting in 2015, the EWRI History of Water Distribution System Analysis task committee
conducted approximately 50 interviews with professionals who have made significant
contributions in the water distribution system analysis field (EWRI, 2019). The interviewees
were asked various questions about their background and experiences and, most importantly for
this review, about their perceptions of future needs and trends in this field. The interviewees
identified a wide range of topics. The categories discussed the most frequently during the
interviews included water infrastructure assessment, real-time modeling, advanced analytics,
water loss/leakage, research/innovations, monitoring, water quality, economics, and
optimization. Other areas that were mentioned include water and cyber security, energy analysis,
valve management, uncertainty, mega cities, and intermittent water supplies. Several
interviewees emphasized the need for good engineering judgement in addressing issues.
The AWWA Engineering Modeling Applications Committee has regularly conducted surveys on
the state of water distribution modeling. In a 2014 article, they summarized their latest findings
in terms of trends in water distribution system modeling (AWWA, 2014):
•	Utilities will continue to integrate network models with geographic information system
(GIS).
•	Investment in modeling increases, with a capital improvement program will be the top
driver for growth.
•	Operations show the greatest increase in anticipated use of the network model.
•	Real-time data will influence direction and use of network models.
The committee summarized the current use of modeling as follows. "The current level of a water
utility's interest and activity in using hydraulic models is high, and these models can be a critical
tool for addressing the major challenges that many utility's encounter with the planning, design,
and operations of their water distribution systems. With more advanced data systems such as GIS
and asset management becoming standard tool sets within water utilities, the hydraulic model has
evolved in terms of the level of detail being modeled, integration with these systems, and the use
of more advanced hydraulic model applications. With these advancements, there is also added
complexity, with hydraulic model calibration shown as the most challenging aspect of hydraulic
modeling tasks" (AWWA, 2014).
In a chapter titled "Water Distribution System in 2050" in the EWRI published book, "Toward a
Sustainable Water Future - Visions for 2050," future trends identified for water distribution
systems included investigating different types of distribution designs, having more control and
monitoring within the distribution system, and managing the assets of system more effectively
and efficiently (Grayman et al., 2012). In addition, issues facing distribution systems today and
in the future were:
•	Degraded infrastructure
•	Climate change, resource depletion and increasing demands for water
•	Using highly treated water to satisfy lower quality needs
•	Providing adequate standby fire-flow needs
•	Preventing contamination in distribution systems
•	Reducing energy requirements
•	Providing centralized systems in developing regions
•	Reducing vulnerability to terrorism
10

-------
EPA/600/R-19/174
October 2019
The water industry is encountering numerous challenges, such as water scarcity, water quality,
aging infrastructure, and climate change. The most commonly identified capabilities to address
these challenges include real-time analytics, enhanced modeling and monitoring (e.g., real-time
modeling), integrated asset management systems, and smart water hardware and communication
technologies. These challenges and capabilities can be used to inform research directions and
technology development that HSRP and others pursue in the future to make drinking water
systems more resilient.
6 Engagement with EPANET Community
EPANET provides the hydraulic and water quality simulation engines for the HSRP's modeling
and simulation software tools. Therefore, an important step in the review of these tools was to
engage with the EPANET community on the future vision and development direction of
EPANET. EWRI in association with EPA, the National Center for Infrastructure Modeling and
Management, and the broad user and open-source software development communities convened
an EPANET Visioning Summit in Reston, Virginia on April 3-4, 2018 (Grayman and Travers,
2018). A more detailed summary of the Summit is provided in Murray et al. (2018). The mission
of the Summit was to develop a shared vision for the future development of EPANET. Thirty-
five participants including representatives from the Summit sponsors, commercial software
companies, engineering consultants, water utilities, academia, and professional organizations
attended. The Summit addressed the three general questions listed below:
•	What is the appropriate structure and style for future EPANET development?
•	What additional functionality is needed in EPANET and other water distribution system
modeling software?
•	How can the various members of the EPANET community work together to best move
EPANET forward in the future?
The participants had general agreement surrounding the continued development of EPANET as
an open-source project, with strong community contributions and based upon a permissive
license. While the general direction of the open source development of EPANET was considered
positive, the participants wanted more specific details about EPANET's future development. It
was anticipated that continued discussion among the represented groups (e.g., commercial
software companies, engineering consultants, water utilities, academia, professional
organizations, and government) would further shape the overall development to ensure that the
diverse EPANET community can work together effectively to further the advancement of
EPANET.
The participants also identified additional functionality that is needed to improve the capabilities
of EPANET and support other water distribution system modeling programs. A total of 44
specific areas of needed functionality enhancements were identified. These were categorized into
four major themes: user experience, development needs, hydraulics, and water quality. Some of
these functionalities could become future research pursuits for HSRP's water distribution system
11

-------
EPA/600/R-19/174
October 2019
modeling program. A few of the functionalities are summarized below. EPANET should include
the ability to:
•	Stop and restart simulations after a change to the simulation parameters (e.g., demands,
pipe status) of the network model
•	Directly integrate extensions like RTX, MSX, and others into one product
•	Run the hydraulic and water quality engines in step with each other rather than
sequentially
•	Run the network model if a portion of the network is disconnected
•	Incorporate real-time data streams into the modeling capabilities to support operations
•	Isolate portions of the network by closing selected valves within the network model
•	Simulate pressure driven demands
•	Update leakage equations to be more realistic
•	Include prototype multi-species water quality reaction/fate equations that are of interest
to water utilities (e.g., chlorine decay, DBP formation and decay, chloramine decay/fate,
temperature modeling, and multiple source tracing)
•	Simulate dispersion under laminar flow in the network for dead-ends and premise
plumbing systems
7 Expert Peer Review Workshop
The three preceding steps (expert interviews, review, and EPANET community engagement)
provided a wide overview of the needs in water distribution system modeling and analysis. In
order to dig deeper into the subject and to address the specific goals of the review, the final step
in information collection was the expert peer review workshop. The workshop on HSRP's Water
Security, Response, Resilience Modeling and Data Analytic Tools, was held at the EPA's
campus in Cincinnati, Ohio on April 9-10, 2018. The purpose of the workshop was to: (1) review
HSRP's water security and resilience modeling goals; (2) obtain additional comments,
suggestions, and new ideas through face-to-face discussions on the status of HSRP's water
security and resilience tools; and (3) develop a path-forward strategy to improve these tools and
increase their usage by water utilities and support their needs. Five drinking water subject matter
experts with in-depth knowledge and experience in the field of water distribution systems
modeling, operation, or analysis, and at least 10 years of experience were selected to form the
panel. They were selected by Versar, Inc., an independent EPA contractor, through a peer review
selection process. Additionally, other water professional experts from the water industry sector
(e.g., water utilities, professional organizations, software vendors, and consultants), academia,
and EPA were in attendance and participated in the discussions. The workshop provided an
opportunity to share information regarding the state of the science for water system security and
resilience, with the focus on drinking water distribution systems.
7.1 Panel Charge Questions
In preparation for the workshop, the peer review panelists were provided background material
including short descriptions of each of the HSRP's water security and resilience tools along with
a preliminary set of suggested enhancements prepared by EPA staff (see Appendix for
enhancement suggestions). Using this information, the panelists were asked to provide responses
12

-------
EPA/600/R-19/174
October 2019
for the five charge questions. For the first three questions, responses were requested in terms of
general comments as well as specific comments on each tool individually. General comments
were requested for the last two charge questions. A preliminary report summarizing the
panelists' responses was assembled and were provided to the panelists prior to the workshop.
The charge questions are listed below.
1.	How could HSRP's software tools or systems modeling and data analytics research be
used to support the current and future needs of the water sector? What modifications
could be made to these tools to encourage their use in applications beyond drinking water
distribution systems?
2.	What improvements can be made to HSRP's software tools, methodologies, or
algorithms to address the most pressing needs of the water sector and provide multiple
benefits? How can HSRP increase the usability and applicability of their tools to meet the
needs of the water sector?
3.	What other innovative methods are available from the research community that should be
evaluated by our systems modeling research program or incorporated into HSRP
products? Please describe any such methods and why they should be incorporated.
4.	How can systems modeling and data analytics techniques be used in new or different
ways to advance water sector research priorities? What new research areas can be
identified?
5.	With the potential of reduction in resources, how should HSRP focus their efforts on
these tools to provide the most benefit to the water sector?
7.2	Workshop Agenda
The 1 '/2 day workshop was composed of the following steps:
•	Introductions and overview presentations on HSRP's research and modeling program
•	Description of the objectives of the review program and structure of the workshop
•	Discussion of the future high-level directions and needs of the drinking water industry
•	Separate discussions on each of the seven HSRP tools and research areas
•	Brainstorming on where the field will go in the next 25 years
•	Final suggestions for the future direction of HSRP's water security, response, resilience
modeling and data analytic tools
7.3	Workshop Discussions
Highlights of the discussions during the workshop are presented below organized by the
following topics:
•	Future high-level directions and needs (See Section 7.3.1.)
•	Brainstorming on next 25 years (See Section 7.3.2.)
•	Experts' suggestions on HSRP's future modeling direction (See Section 7.3.3.)
7.3.1 Future High-Level Directions and Needs
The first discussion topic during the workshop was on future high-level directions and needs,
including a summary of the charge question comments submitted by the panelists. The following
list highlights future high-level directions and needs relevant to tool development that were
identified. Other topics discussed during the workshop included sensors and pressure
management.
13

-------
EPA/600/R-19/174
October 2019
•	Engage utilities and other end users
o EPA needs to create a plan for the development of their tools, specifically one that
includes engaging with customers to understand their needs,
o Utilities are no longer just an end-user. A recent shift has occurred in which utilities
participate in the innovation process. EPA needs to recognize this and bring the
utilities in sooner to the development of tools.
•	Integrate water security tools into day-to-day tools used by utilities
o If a tool is designed for water security only, it might not be useful to the end-user
when it is needed.
o Utilities might not be able to justify the costs for tools unless the perceived benefits
outweigh the costs. Utilities tend to be more concerned about items such as pipe
break assessments, water loss, energy efficiency, pressure management, and age of
system, rather than security,
o Utilities are driven by regulatory requirements (such as for water quality), resources,
and customer expectations, and by imminent health and safety issues. Cost benefit
comes in play for compliance strategies and non-regulatory initiatives. Tools that
address multiple challenges are more useful to the water industry.
•	Look to the future
o For any tool created now, there will be a ramp-up period of up to 10 years before it is
fully used by the water industry. EPA needs to include this aspect in their tool design,
o Forward thinking is challenging, and most utilities have a better understanding of past
and current problems.
o The industry is moving towards consolidated management through regional networks
or virtual networks. As such, individual utilities might not have to fully understand all
the details of modeling,
o The future workforce will be more knowledgeable about information technology and
these types of tools will be more useful to them. Institutional knowledge will be lost
with expected workforce retirements,
o Utilities should move toward real time operations systems like the electric industry.
Tools exist today for the water sector, but they are separate, disconnected system that
are not connected.
•	Consider approaches to increase the use of tools
o Analyses could be conducted to compare utility use of HSRP tools with
violations/compliance. If results show that use of tools correlates with fewer
violations, smaller systems might be more likely to use the tools,
o A risk assessment scoring system could be developed so that communities could rate
themselves in terms of risk. This will encourage lower rated communities to adopt
tools that will help them respond to risks,
o Simplified models could be explored for small systems, since they do not have
programmers/modelers on staff and rely mainly on vendors/consultants. This could be
accomplished by conducting sensitivity analysis that could be applied to small
systems.
7.3.2 Brainstorming Future of the Drinking Water Industry
Another portion of the workshop included brainstorming the future of the drinking water
industry. A variety of technologies or changes in water that might influence the field over the
14

-------
EPA/600/R-19/174
October 2019
next 25 years were identified. Some of these were advancements in infrastructure materials (e.g.,
smart pipes) and computer resources (e.g., artificial intelligence, cloud computing); changes in
water administration (e.g., point-of-use treatment, concept of net zero water and conservation);
and shifts in the customer expectations (e.g., real-time communication on water usage and water
quality). A detailed list of the brainstorming topics is provided in the Appendix.
7.3.3 Experts' Suggestions for HSRP's Future Modeling Direction
The recommendations of the panel in terms of the water distribution systems modeling and data
analytics research program are summarized below.
•	Investigate new approaches and technologies for the tools, such as smart phone or tablet-
based apps, cloud-based systems, data analytics, and artificial intelligence.
•	Focus on the day-to-day operation of utilities more than on rare incidents like terrorist
acts.
•	Seek more input and feedback from different stakeholders (e.g., large utilities, small
utilities, consultants) along the development process as they have different needs.
•	Build connections with water utilities, associations (AWW A, Association of State
Drinking Water Administrators) and the Water Research Foundation.
•	Adapt tools to anticipate potential future changes in EPA regulations.
•	Consider the workforce of the future.
•	Develop more simplified tools (e.g., real-time, sensor placement) that are plug-and-play
and easier to use.
•	Conduct field studies to validate the theories and assumptions included in the modeling
tools (e.g., complete mixing at pipe junctions, fate models).
•	Increase usability and applicability of tools by:
o Identifying target audience and performing a needs assessment
o Focusing EPA research in their areas of strength and expand efforts to leverage other
government and private sector capabilities for other areas
o Developing a software sustainability plan that includes all players
o Making sure that software core (EPANET) is stable and usable
•	Update and regularly maintain tools. New tools require additional resources.
•	Discontinue several areas of development because commercial providers can incorporate
HSRP's developments into mainstream programs.
•	Look at the electrical power industry approach as a vision for water utilities - using real
time analytics and using links to asset management and to other data for immediate
response.
7.4 Experts'Take Away Messages
After the workshop the experts were asked to summarize their recommendations for the future
path forward for HSRP tools. This included providing five take away messages that they wanted
to share with EPA. Summaries of the experts' take away messages are categorized below into the
following three topic areas: tools, collaboration/coordination, and marketing/strategy/vision. As
the experts represented various viewpoints (e.g., academia, consulting engineer, utility), the
messages could be conflicting.
TOOLS
15

-------
EPA/600/R-19/174
October 2019
•	Continue supporting the updates and improvements to EPANET to ensure that it is stable
and useable as it is a very important tool for utilities and the water industry. Additional
features could include uncertainty and/or calibration modules.
•	Develop and modify tools in steps, initially focusing on including capabilities that could
be of most value to the water utilities, e.g., assist with automated valve isolation and
simplified real-time modeling.
•	Integrate tools into one product rather than separate codes/modules.
•	Tools should be modified and geared towards direct application by the end users (utility
staff and consultants) and incorporate the latest advances in user interfaces.
•	Modify and update tools for direct application by the end users (e.g., utility staff and
consultants), which primarily rely upon modern graphical user interfaces. In small
drinking water utilities, staff is often limited to just an operator that is mainly familiar
with plug-and-play tools rather than an engineer who understands code and compiling
libraries.
•	Provide maintenance and support of these tools to ensure continued and expanded use by
utilities.
•	Support premise plumbing research and model development, since it could assist with
understanding Legionella, lead, and other public health and design questions within
homes and other buildings.
COLLABORATION/COORDINATION
•	Collaborate and engage with utilities in the planning, development, and adoption of the
tools, since utilities are central to advancements in the water sector.
•	Work through existing programs (like the Partnership for Safe Water) to extend
participation in the tool development.
•	Increase coordination with organizations, such as AWWA and Water Research
Foundation, and states to understand the daily challenges at the average utility, such as
main breaks and DBP compliance, and identify how the tools can be used to support
these challenges in addition to planning for system security.
•	Develop a sustainability plan for the tools that includes all the players.
•	Explore the creation of a panel that includes utility staff for actively guiding the tool
development from initial phases through beta testing.
MARKETING/STRATEGY/VISION
•	Encourage adoption of software tools for utilities that could most benefit from the tools
but are not aware or trained on how to use them. This serves the core mission of
supporting water systems to prepare and recover from disasters, and more importantly, to
protect public health and the environment.
•	Obtain better data on tool usage and examine any correlation between compliance
violations for utilities that do and do not use tools. More simply demonstrate how the use
of network models might provide a business case for determining the level of future
support for tools.
•	Gain wider acceptance and usage of the tools by improving them to address pressing
utility needs and focusing on day-to-day operations of a utility, such as main breaks and
regulatory needs.
16

-------
EPA/600/R-19/174
October 2019
•	Develop a full-scale strategic plan/vision for the future of these tools that includes
addressing contamination threats and system resilience.
•	Promote technology transfer of the tools to the water modeling industry.
•	Market directly to utilities, particularly progressive ones, by applying the tools to real
systems and presenting the results at AWWA conferences.
•	Focus on the areas that EPA does well (e.g., EPANET, fate and transport modeling) and
leverage others for their expertise (e.g., user interfaces).
•	Identify the target audiences for the research/tools and perform a needs assessment.
8 Project Conclusions for HSRP Tools
Overall, this project highlighted some commonality between the different perspectives (e.g.,
academia, government, engineering practitioner, utility) within the drinking water community. A
major theme was the accuracy of the network model itself. All parts of the review highlighted
that the underlying network model needs to accurately reflect the water distribution system and
its operation. A more accurate representation is obtained by collecting data (e.g., flows,
pressures, tank levels) more frequently within the system and then linking it to the network
model to create a real-time model that depicts the actual system dynamics. In addition to the
hydraulics of the system, a more accurate representation of the water quality in the system is
needed. One step towards this could involve more experimental data on the fate of different
water constituents to understand how they react with one another. These data would provide
more detailed reaction equations, which could be linked to water quality data collected in the
distribution system. This second step would provide a more complete understanding of the water
quality in the system in real-time.
Another focus area identified during the project was collaboration. The majority of the subject
matter experts emphasized that it was important to engage with the water community through all
phases of research and development to understand the problems that water utilities are facing,
develop solutions for these problems, and transfer the solutions to the commercial industry for
refinement, maintenance, and dissemination. An additional highlight of the project was that
modeling and data analytic tools need to assist utilities with the day-to-day operation of the water
distribution system rather than the relatively rare occurrences related to security or contamination
incidents. Tools focused on the day-to-day operation would provide utilities with a better
understanding of the system dynamics, which would be beneficial during emergency situations.
The discussions also highlighted that most of the water community needs user-friendly tools that
are easy to understand and implement as water utilities have many competing priorities and
limited staff and resources.
In terms of HSRP's future steps in upgrading and developing water distribution modeling and
data analytics tools, this project has illustrated that it is a complex, multi-dimensional task. It
should be influenced by the specific criteria to be used in prioritizing future actions, by the
overall available resources, and by the development costs and time associated with each potential
action. The following discussion identifies ideas for the future development of HSRP's water
distribution modeling and data analytics tools.
17

-------
EPA/600/R-19/174
October 2019
Needed future tasks could be as follows:
•	Apply existing or modified tools to case studies
•	Perform basic research on the processes associated with the tools
•	Improve the interface and usability of the tools
•	Expand the capabilities of existing tools
•	Develop "best practices" for application of tools
•	Inform/educate the modeling community about modeling advances
•	Champion the future development and application of specific modeling technologies
HSRP has been involved in all these categories of tasks in the past. Based on suggestions from
the review process, specific emphasis and tasks for each of the seven areas have been identified
and described below.
8.1	TEVA-SPOT
Since its inception over 15 years ago, TEVA-SPOT has been the sensor placement tool most
used for water security related issues. The research community has continued to develop
alternative optimization routines for determining best sensor locations, but these routines
generally have not been incorporated into tools or have not provided user friendly interfaces.
However, the use of TEVA-SPOT in recent years has declined since the emphasis on security-
only related monitoring has decreased. A potential future direction suggested by the review was
expanding the optimal sensor placement tool to design monitoring systems used for routine and
regulatory water quality monitoring (in addition to the security-oriented goals) within the water
distribution system. The information from these optimally located monitoring locations would be
used to improve the day-to-day operation of water distribution systems. TEVA-SPOT has the
potential to help reduce the capital, operational, and maintenance costs of sensor deployments for
monitoring the distribution system.
8.2	CANARY
CANARY was developed to analyze water quality data collected at online monitoring locations
to alert water utilities of abnormal conditions within the distribution system. The review process
highlighted opportunities for CANARY to be used beyond its original design purpose. HSRP
presented other CANARY applications, such as detecting illicit events in rivers. Based upon the
review, future opportunities likely exist with the continued and broader application of CANARY
to time-series data. However, resources would be needed to develop a more user-friendly
interface. As the water industry continues to collect more data, methods to analyze and evaluate
the data are needed to assist in the identification of any issues with the water being monitored.
8.3	EPANET Multi-Species extension (MSX)
MSX was developed over a decade ago by EPA. It has been directly incorporated into most of
the widely used commercial water distribution system software packages. However, both the
EPANET version and commercial versions of MSX have found only limited use outside of the
research environment. To expand its use in the water industry, EPA could conduct basic research
to develop a library of chemical fate routines for the most widely modeled parameters (e.g.,
chlorine and DBPs). With more accurate representations of the reaction dynamics of water
18

-------
EPA/600/R-19/174
October 2019
constituents, a water utility would have a better understanding of how changes in the treatment
process and/or source water would affect the water quality within the distribution system.
8.4	EPANET Real-Time extension (RTX)
The review process affirmed that real-time modeling and use of real time data in the operations
of water systems was at the top of the list for areas of future modeling and analysis of water
distribution systems. It also showed that the work that HSRP has done in the development of
RTX has been in the forefront in this field and that it has significantly spurred development and
applications in this area. However, the review also suggested that commercial entities (software
developers, vendors, consultants) are better positioned to develop and apply future software in
this field and that HSRP should take a secondary, support role with real-time modeling. Future
EPA research directions in this area could be assisting utility partner(s) with the integration and
use of real-time technologies to improve routine, day-to-day operations and response to
emergency situations. The goal of this work would be to advance adoption of real-time modeling
and analytics into the daily operations of drinking water utilities.
8.5	Water Security Toolkit (WST)
Following a contamination incident within a distribution system, a utility implements response
actions to mitigate the effects. WST was developed to assist with the evaluation of these
responses. The review process suggested that the focus of HSRP tools should be more aligned
with day-to-day utility operations (e.g., flushing programs) rather than with security incidents. In
addition, the review process found that WST would not be useful during a real emergency due to
time constraints required to run the analysis. Future directions for WST would require the
linkage with a real-time water distribution model to evaluate the effectiveness of different
response strategies available (e.g., source identification, flushing, chlorine disinfection,
sampling) in real-time.
8.6	Water Network Tool for Resilience (WNTR)
Resilience has become an important concept within the water industry and other infrastructure
systems. WNTR was viewed favorably within the review process as a tool that could be used to
support future research. Since WNTR is a relatively new tool, additional case study applications
could be pursued to understand the consequence and resilience response results it provides (e.g.,
effects of different operations' scenarios). A better understanding of WNTR results and findings
is needed to help ensure the production of useful analysis and information, and support analysis
of critical importance to utilities (e.g., aging infrastructure, natural disasters). In addition, to
make the tool more user-friendly, WNTR functionality could be incorporated into future versions
of EPANET using the Python-based, new user interface. These future directions could assist
drinking water utilities of all sizes to use WNTR to gain a better understanding of how
infrastructure (e.g., pipe, tank, or pump) failures could affect operations and the delivery of water
to customers.
8.7	Premise Plumbing
Premise plumbing has become a significant issue of interest in the water distribution field due to
water quality and security concerns. Currently it is an active area of research driven by several
contamination incidents that have occurred following wildfires (e.g., Camp Fire in Paradise,
19

-------
EPA/600/R-19/174
October 2019
California), distribution system contamination (e.g., 2014 Elk River [West Virginia] chemical
spill), and water treatment modifications (e.g., Flint [Michigan] water crisis, Pittsburgh
[Pennsylvania] water crisis). Thus, the reviewers identified premise plumbing as a research area
to pursue. One future task that EPA could undertake in this area would be convening a workshop
on the status, the needs, and the potential future directions of water quality research and
modeling in premise plumbing systems. This would be similar to when EPA sponsored a
workshop on water quality modeling in distribution systems in 1991 that kickstarted
development in that area. A better understanding of the fate of water constituents would help to
improve water quality in premise plumbing systems.
20

-------
EPA/600/R-19/174
October 2019
9 References
AWWA (American Water Works Association), 2014, "Water distribution System Survey
Subcommittee, Engineering Modeling Applications Committee. Committee Report: Trends in
water distribution system modeling," J. AWWA, vol. 106, no. 10, pp. 51-59, October 2014.
Burkhardt, J., Murray, R., Woo, H., and Mason, J., 2018, "Modeling drinking water lead
exposure from premise plumbing," presented at EWRI World Environmental & Water Resources
Congress 2018, Minneapolis, MN, June 3-7, 2018.
CIPAC (Critical Infrastructure Partnership Advisory Council) Water Sector Strategic Priorities
Working Group, 2013, "Roadmap to a Secure and Resilient Water Sector," Available (accessed
16 August 2019): https://www.asdwa.org/wp-content/uploads/2016/07/2013-Roadmap-to-a-
Secure-Resilient-Water-Sector.pdf
EWRI (Environmental and Water Resources Institute), 2019, "Water distribution systems
analysis history project," YouTube, (Accessed 19 April 2019):
https://www.voutube.com/chaimel/UCL4EjqZ2.wKqep2-iYfBdK2e
Grayman, W.M., LeChevaillier, M.W., and Walski, T., 2012, "Water distribution systems in
2050," Chapter 26 in: Toward a sustainable water future: visions for 2050, W.M. Grayman, D.P.
Loucks, and L. Saito, eds. American Society of Civil Engineers, Reston, VA.
Grayman, W. and Travers, R., 2018. "EWRI Convenes Summits on Future EPANET and
SWMM Development." EWRI Currents, vol. 20, no. 2, Available (accessed 19 August 2019):
https://issuu.com/asce-ewri/docs/spring 2018 currents - issue
Janke, R., Morley, K., Uber, J.G., and Haxton, T., 2011, "Real-time modeling for water
distribution system operation: integrating security developed technologies with normal
operations," presented at AWWA Water Sec. Conf. and Dist. Systems Symposium, Nashville,
TN, Sept. 11-14, 2011.
Klise, K., Bynum, M., Moriarty, D., and Murray, R., 2017a, "A software framework for
assessing the resilience of drinking water systems to disasters with an example earthquake case
study," Env. Model & Software, vol. 95, pp. 420-431, Sept. 2017.
Klise, K.A., Moriarty, D., Bynum, M.L., Murray, R., Burkhardt, J., and Haxton, T.M., 2017b,
"Water network tool for resilience (WNTR) user manual," Washington, DC, USA, EPA/600/R-
17/264, Available (accessed 19 August 2019):
https://cfpub.epa.gov/si/si public record report.cfm?dirEntry] 793
Morley, K., Janke, R., Murry, R., and Fox, K., 2007, "Drinking water contamination warning
systems: water utilities driving water security research," J. AWWA, vol. 99, no. 6, pp. 40-46,
June 2007.
21

-------
EPA/600/R-19/174
October 2019
Murray, R., Grayman, W., Parsons, B., Whitten, B., Boccelli, D., Cleveland, T., Ostfeld, A.,
Strasser, A., and Rowney, C., 2018, "Results from the EWRI summit on the future of EPANET,"
presented at 1st International WDSA/CCWI2018 Joint Conference, Kingston, Ontario, Canada -
July 23-25, 2018.
OHS (White House Office of Homeland Security), 2002, "National strategy for homeland
security (July 2002)," Available (accessed 19 August 2019):
https://www.dhs.eov/piiblication/first-national-strategY-homeland-securitv
Roberson, J. A. and Morley, K.M., 2005, "Contamination warning systems for water: an
approach for providing actionable information to decision-makers," AWWA, Denver, CO.
Samuels, W.B., Bahadur, R., Grayman, W.M., and Borja, R.P., 2010, "Guidelines for the
prevention of contamination of drinking water in buildings and large venues," Part 3. Key
Application Areas, 2. Water In: Wiley Handbook of Science and Technology for Homeland
Security, Ed., J.G. Voeller. John Wiley & Sons, Inc., Hoboken, NJ, Available (accessed 22
August 2019): https://doi.org/10.1002/9780470087923.hhs691
Shang, F., Uber, J.G., Rossman, L.A., 2008, "Modeling reaction and transport of multiple
species in water distribution systems," Environ. Sci. & Technol, vol. 42, no. 3, pp. 808-814, Feb.
2008.
Shang, F., Uber, J.G., and Rossman, L., 2011, "EPANET multi-species extension user's manual
and software," Washington, DC, USA, EPA/600/S-07/021, Available (accessed 19 August
2019): https://cfpub.epa.gov/si/si public record report.cfm?dirEntryI 188
Tisdale, R., 2018, "The changing water landscape, nine trends to watch in 2019" WaterWorld,
Dec. 2, 2018, Available (accessed 19 August 2019):
https://www.waterworld.com/municipal/water-utilitv-management/asset-
management/arti cl e 058/the-changing-water-landscape
Uber, J.G., Hatchett, S., Hooper, S., Boccelli, D., Woo, H., and Janke, R., 2014, "Water utility
case study of real-time network hydraulic and water quality modeling using EPANET-RTX
libraries," Washington, DC, USA, EPA/600/R-14/350, 2014, Available (accessed 19 August
2019): https://cfpub.epa.gov/si/si public icecal	report.cfm?dirEntryId=288792
U.S. Department of Homeland Security (U.S. DHS) and U.S. EPA, 2015, "Water and
Wastewater Systems Sector-Specific Plan," Available (accessed 16 August 2019):
https://www.dhs.gov/sites/default/files/pubHcations/nipp-ssp-water-2015-508.pdf
U.S. Environmental Protection Agency (U.S. EPA), 2004, "Water security research and technical
support action plan," Washington, DC, USA, EPA/600/R-04/053, Available (accessed 16 August
2019): https://nepis.epa.gov/Exe/ZvPIJRL.cgi?Dockey=Pl0049GF.txt
U.S. EPA, 2010a, "Sensor network design for drinking water contamination warning systems: a
compendium of research results and case studies using the TEVA-SPOT software," Washington,
22

-------
EPA/600/R-19/174
October 2019
DC, USA, EPA/600/R-09/141, Available (accessed 16 August 2019):
https://cfpub.epa.gov/si/si public record iv port.cfm?dirEntrvL I - * N J
U.S. EPA, 2010b, "Water quality event detection systems for drinking water contamination
warning systems: development, testing, and application of CANARY," Washington, DC, USA,
EPA/600/R-010/036, Available (accessed 16 August 2019):
https://cfpub.epa.gov/si/si public record report.cfm?dirEntry]
U.S. EPA, 2012, "CANARY user's manual (version 4.3.2)," Washington, DC, USA,
EPA/600/R-08/040B, Available (accessed 16 August 2019):
https://cfpub.epa.gov/si/si public record report.cfm?dirEntrvld _ J ^r
U.S. EPA, 2013, "CANARY training tutorials," Washington, DC, USA, EPA/600/R-13/201,
Available (accessed 19 August 2019):
https://cfpub.epa.gov/si/si public rec< icport.cfm?dirEntryId=261777
U.S. EPA, 2014a, "Configuring online monitoring event detection systems," Washington, DC,
USA, EPA 600/R-14/254, Available (accessed 19 August 2019):
https://cfpub.epa.gov/si/si public record report.cfm?dirEntryId=287299
U.S. EPA, 2014b, "Promoting technology innovation for clean and safe water: water technology
innovation blueprint, version 2," Washington, DC, USA, EPA 820-R-14-006, Available
(accessed 19 August 2019): https://www.epa.gov/sites/production/files/2014-
04/documents/clean water blueprint flmal.pdf
U.S. EPA, 2014c, "Water security toolkit user manual: version 1.3," Washington, DC, USA,
EPA/600/R-14/338, Available (accessed 19 August 2019):
https://cfpub.epa.gov/si/si public record report.cfm?dirEntryId=288918
U.S. EPA, 2015, "Enhancements to the EPANET-RTX (real-time analytics) software libraries -
2015," Washington, DC, USA, EPA/600/S-14/441, Available (accessed 19 August 2019):
https ://cfpub. epa. gov/ si/si public record report. cfm?dirEntrvId=3 09673
U.S. EPA, 2016, "Threat Ensemble Vulnerability Assessment - Sensor Placement Optimization
Tool (TEVA-SPOT) graphical user interface user's manual," Washington, DC, USA,
EPA/600/R-13/014, Available (accessed 19 August 2019):
https://cfpub.epa.gov/si/si public record ivport.cfm?dirEntryId=257684
Water and Wastewater Sector Strategic Roadmap Work Group, 2017, "Roadmap to a secure and
resilient water and wastewater sector," Available (accessed 19 August 2019):
https://www.waterisac.org/sites/default/files/public/	Water Sector Roadmap FIN
WEF (Water Environment Federation), 2019, "Utility of the future," Available (accessed 19
August 2019): https://www.wef.org/utilitv-of-the-fiitiire/
23

-------
EPA/600/R-19/174
October 2019
Wood, A.R., Harten, T., and Gutierrez, S.C., 2018 "Approaches to identifying the emerging
innovative water technology industry in the United States," Journal AWWA, vol. 110, no. 5, pp.
11-21, May 2018.
24

-------
EPA/600/R-19/174
October 2019
10 Appendix
The Appendix provides definitions of the terms used for modeling distribution system, and more
detailed information on the potential tool enhancements that were provided to the panelists
before the 2018 Water Security, Response, Resilience Modeling and Data Analytic Tools
Workshop. In addition, the detailed list of brainstorming topics discussed during the workshop is
also presented.
10.1 Modeling Terms and Definitions
The drinking water distribution system infrastructure is the principle asset of a community water
system. Drinking water distribution systems are designed to deliver water from a source (e.g.,
treatment facility) to the customer at the required quantity, quality, and pressure to meet their
needs. EPANET is a software application developed by the U.S. Environmental Protection
Agency (EPA) to simulate the hydraulics, operations, and the fate and transport of water
constituents in drinking water distribution systems. The most popular commercial water
distribution system modeling software packages were built using the EPANET solvers for
hydraulics and water quality.
In the context of water distribution systems, the term network model is used to describe the
operation of the water distribution system (e.g., times of day when valves open/close, tank levels
to trigger pump operations, times of day when pumps turn on/off) and the characteristics of its
infrastructure components (e.g., pipes, pumps, valves, tanks, reservoirs, service connections).
The term is often used to describe the input file format required for EPANET. The network
models for drinking water distribution systems can vary greatly in the level of detail used to
describe the water system (e.g., number and size of pipes included in the network model). Often
a less detailed network model has a greater level of skeletonization (i.e., only transmission pipes
are represented and not service connections). Additionally, network models generally describe
the water distribution system infrastructure from the point just after treatment (e.g., finished
water reservoir or clearwell) to an aggregated set of customers service connections. Most
network models do not include every customer's service connection individually. The number of
customers aggregated can vary greatly, with greater aggregation leading to more skeletonization
of the network model.
The network model is usually developed to represent a specific snapshot of the water utility's
operation (e.g., maximum, minimum, or average daily water demands). These network models
are not connected with real-time sensor data, and, therefore, they might not accurately represent
the distribution system behavior during an emergency incident (e.g., pipe break or
contamination). A real-time model represents a network model that is continuously fused with
real-time data, in sufficient quantity and quality, to more accurately predict the current
conditions of the water distribution system operations. With a historical record of operations data
(e.g., a SCADA [supervisory control and data acquisition] historian record of operations data for
a period of one year or longer), a real-time model could also be capable of examining past
operational practices and incidents (e.g., pipe break) to analyze how different practices could be
implemented to minimize the consequences or to improve the response times. In addition, the
25

-------
EPA/600/R-19/174
October 2019
ability to accurately predict customer demands, flows, pressures, and other system parameters is
critical for determining how best to respond during an emergency.
The primary developer of the water distribution system network model is the community water
utility or their consultant. From the utility's perspective, their responsibility for the piping
infrastructure stops at the customer's curb stop, while the customer service line and interior
plumbing are the homeowner's responsibility. Thus, utility network models do not include the
customer's piping and infrastructure, which is also known as the premise plumbing or premise
distribution system. The design, attributes, and operational practices (i.e., customer's behavior
and use of water using appliances) describing the premise distribution system is incredibly
varied.
10.2 Potential Tool Enhancements
EPA's Homeland Security Research Program (HSRP) prepared a list of potential future upgrade
tasks associated with each of the seven existing HSRP water distribution system modeling tools.
These proposed tasks are delineated below.
10.2.1 Threat Ensemble Vulnerability Assessment - Sensor Placement Optimization Tool
(TEVA-SPOT)
•	Evaluate lessons learned from past applications of the TEVA-SPOT. TEVA-SPOT has
been used to determine the optimal locations for monitors in a few dozen water systems.
In each case, a detailed analysis was performed that integrated the software with an in-
depth understanding of the water system. The method recognized that each water system
was a unique entity with different characteristics and operations that would result in a
different set of locations for monitors. The purpose of this task is to use the knowledge
gained from these past applications of TEVA-SPOT to try to develop some generalized
rules/guidelines for determining the best locations for monitors.
•	Conduct a utility case study application for emergency preparedness. This work area
would include performing a utility threat, consequence, and security improvement case
study to demonstrate the capabilities within TEVA-SPOT to assist utilities in their
consequence assessment and emergency preparedness efforts. This would be a full
immersion utility case study whereupon HSRP representatives would work closely with
the partnering utility representatives to apply the TEVA-SPOT methodology to their
water system and then work with water utility representatives or their consultants to
understand and implement the results for improved monitoring and security.
•	Examine methods for extending TEVA-SPOT capabilities to include routine water
quality monitoring. TEVA-SPOT was originally designed as a mechanism to investigate
the vulnerability of a specific water utility to a wide variety of contamination incidents
and to help design a real-time monitoring system to quickly detect contaminants in the
distribution system. The purpose of this work area would be to examine how the TEVA-
SPOT capabilities could be modified and extended to include the design of monitoring
systems used for routine and regulatory water quality monitoring (in addition to the
security-oriented goals).
•	Investigate an alternate monitoring paradigm involving continuous monitors. This work
area would involve partnering with a drinking water utility to use TEVA-SPOT to
identify a small number of optimized sampling locations that could be used for more
26

-------
EPA/600/R-19/174
October 2019
frequent monitoring of routine water quality parameters. The TEVA-SPOT approach of
selecting sampling locations for routine monitoring would be compared to the current
approach of identifying sampling locations. Typically, water utilities sample from many
locations with single or only a few grab samples. For large utilities, driving around
collecting samples is difficult due to traffic issues and logistical concerns. A dedicated
continuous monitoring station could be considerably less costly. This work would
evaluate the technical (i.e., demonstrate improved public health protectiveness) and
business cases for optimized sampling and continuous monitoring.
•	Study multiple simultaneous contaminant threat scenarios to water systems. The purpose
of this work area would be to investigate the consequences associated with multiple,
simultaneous contamination injections. Better understanding of and quantification of
consequences associated with such scenarios is needed to not only better identify the
magnitude of consequences that water utilities could actually face but the ease at which
multipronged attacks could be perpetrated. A better understanding of such threats and
their consequences will help our partners and stakeholders to be better prepared and to be
able to assist and respond to such threats.
10.2.2 CANARY
•	Continue exploratory testing and case studies. The purpose of this work area would be to
further explore the capabilities of the current version of CANARY in different
applications with different types of sensor data. As sensors become cheaper and more
abundant, the need for tools to perform robust analysis is increasing. Having a tool like
CANARY that is capable of providing anomaly detection on a wide variety of sensors is
also increasingly valuable. Continued research in new and different areas would assist the
community by helping them determine if CANARY is appropriate for their application
and what type of benefits can be achieved from monitoring.
•	Expand CANARY'S usability. The purpose of this work area would be to improve
usability and reduce the learning curve related to using CANARY. The focus would be to
expand the capabilities of CANARY to include the following features: (1) introduce easy
set up feature, or attempt to provide a more seamless interaction with SCAD A/data
sources (i.e., smart features, or wizards); (2) provide more visual information during
analysis (i.e., live graphs); and (3) integrate an auto parameterization tool into the user
interface that would streamline the process of integrating data from a new
source/application.
•	Expand CANARY'S functionality to include additional algorithms. The purpose of this
work area would be to expand functionality to include some additional simple analytics
and modify current analytics to be more customizable to improve overall functionality.
This work could look at adding some simple algorithms (e.g., mean or median based
approaches) and multiple simultaneous algorithms, or automatic confirmatory analysis, to
help reduce false alarms (e.g., comparison of multiple algorithms to provide a best two-
out-of-three type analysis). This could also include expanding how different signals are
analyzed or allow for two different sets of parameters to be applied to different sets of
signals in the data stream (i.e., turbidity has one set, while conductivity has another).
•	Expand CANARY'S functionality to include advanced algorithms. The purpose of this
work area would be to identify, test, and add advanced machine learning algorithms (e.g.,
artificial intelligence approaches) to CANARY. Some approaches have been reported in
27

-------
EPA/600/R-19/174
October 2019
the literature, but these methods would require a significant amount of testing and
validation before including them in CANARY. Generally, artificial intelligence
approaches would require a re-parameterization for each type of signal set (i.e., one set of
parameters would not work in two different applications), which would require support
from some type of automatic parameterization tool.
•	Build a framework for integration of CANARY and real-time modeling. The purpose of
this work area would be to study the potential integration of CANARY analysis as part of
a real-time model/monitoring (RTX) framework to better leverage the capabilities of both
platforms. The first step would be to investigate the use of CANARY in monitoring real-
time modeling efforts. This could potentially reduce false alarms from CANARY and
provide additional information to the real-time modeling effort related to model accuracy.
Integration with a real-time model would likely require some type of automatic time
shifting algorithm to account for temporal variations between the model and the sensor
data.
10.2.3	EPANET Multi-Species extension (MSX)
•	Incorporate MSX into EPANET user interface. As MSX is a command-line program, it is
not currently connected with the EPANET user interface. To enhance viewing of the
MSX simulation results, an EPANET user interface version with MSX would need to be
developed and released.
•	Modify the MSX code and process to accommodate/support a library of MSX fate
models. The development of a specific MSX fate model requires both domain knowledge
on the specific water quality processes being modeled, and some level of
experience/knowledge of the MSX software. Modifications to the MSX code to access
and utilize the library of fate models in EPANET and development of additional fate
models could expand the use of this software.
•	Implement the full-scale MSX library. A mechanism for managing and supporting the
MSX library could be developed to include available fate models and methods to update
the models as needed. Additionally, fate models and parameter sets could be expanded as
needed. HSRP could serve the dual role of developing the new fate models and/or
evaluating fate models developed by outside groups.
10.2.4	EPANET Real-Time extension (RTX)
•	Extend RTX software and documentation. This potential work area would evaluate and
implement RTX's prototype demand forecasting tool into an RTX library. In addition,
this work could investigate the inclusion of RTX libraries within newer versions of
EPANET development to provide a better interface and to support real-time modeling.
The RTX libraries and RTX:LINK software need improved documentation to help
further the development and deployment of the technologies at water utilities.
•	Develop an RTX virtual test application. This work area would create an RTX software
application that uses a utility distribution system model, generates synthetic SCADA
data, and demonstrates threat, consequence, and response strategies (water security and
resilience tools) in a real-time framework to potential utility partners.
•	Conduct additional real-time modeling case study applications. In partnership with
additional drinking water utilities, this work area would examine the application of RTX-
based tools. It would involve designing and deploying cloud-based real-time data
28

-------
EPA/600/R-19/174
October 2019
analytics (RTX:LINK) and real-time modeling at partnering utility(ies). The purpose of
this work would be to assist utility partner(s) with the integration and use of these
technologies to improve routine, day-to-day operations and response to emergency
situations. The goal of this work area would be to advance adoption of real-time
modeling and analytics into the daily operations of drinking water utilities.
10.2.5	Water Security Toolkit (WST)
•	Test and exercise the current WST package. WST would be further tested and applied
under a wide range of situations, primarily using small network models found in the
literature. These applications will serve the purpose of exercising the software and
identifying any issues, and providing additional understanding of the characteristics of
responding to a contamination incident.
•	Conduct one or more large-scale case study applications. The primary focus of this work
would be the use of WST on one or two critical case study applications. These case
studies could use either the re-creation of recent actual contamination incidents (e.g.,
West Virginia or Corpus Christi, Texas) or the creation of realistic, hypothetical
contamination incidents. These re-creations would serve as a verification/validation that
the WST components are functioning properly, and thus, provide greater confidence that
it could be used for future incidents.
•	Evaluate WST application for routine utility operations. Water utilities perform a variety
of daily activities to ensure that water of high quality is delivered to consumers. This
work would evaluate the application of WST to routine utility operations, such as
flushing and sampling. These applications could use WST to evaluate and design
effective flushing and sampling programs.
•	Develop a response best practices report. This potential work area could help to identify
common response action trends by completing case-study applications with a range of
water utility systems. These trends could be related to pre-defining locations to flush
and/or take samples, to evaluating how effective response actions are without knowing
the source of the contamination, and to exploring the combination of different response
actions.
•	Examine the simulated usage of WST in conjunction with a real-time model. A real-time
network model could be linked with WST to provide a near real-time response tool. The
process could be simulated by using a network model derived from a real-time network
model linked with SCADA data as input to WST. This could then be used to compare
how the response actions differ if the response was occurring in real-time.
•	Modify WST to include additional functionality and interface improvements. Future
development of additional functionality in WST could be a future work area. This could
include building a more user-friendly interface, modifying specific response actions, and
adding new response action options.
10.2.6	Water Network Tool for Resilience (WNTR)
•	Conduct additional case study applications. A series of case study applications of the
WNTR software could be completed in order to validate WNTR, ensure the results
produce useful information to water utilities, and determine the path forward for the
software and documentation. This work could be done in collaboration with EPA regions,
29

-------
EPA/600/R-19/174
October 2019
states, water utilities, consultants, nonprofits, or other interested organizations. The
following scenarios could be pursued: hurricane, flooding, wildfire, tornado, and cyber.
•	Enhance WNTR to include prioritizing infrastructure repair and replacement. In support
of the Administration's goals to invest in infrastructure upgrades, WNTR could be
enhanced to support analysis and prioritization of drinking water infrastructure repair and
replacement. WNTR would be modified to incorporate failure probabilities due to aging
infrastructure, material types, and other factors. Scripts could be written to optimize the
order of repairs based on a variety of criteria. Costs of repairs could be included as well
as metrics to measure the benefits of improvements. In addition, methods could be
developed to allow for the comparison of different repair and replacement strategies.
Finally, methods could be extended to incorporate resilience analysis into the
prioritization process (i.e., make decisions about repair and replacement not just based on
age of materials and failure rates but also on benefit provided for improved security and
resilience to disasters).
•	Transfer WNTR to consultants, water utilities, EPA regions, and states. Using
information collected from case study applications of WNTR, guidance on the best
practices for applying WNTR to assess resilience of water systems to disasters could be
developed. This could also include a series of demo scripts and outputs, and a series of
training sessions at conferences or via webinars. The training would be intended for
consultants who might apply WNTR in support of contracts with water utilities, states,
and EPA regions.
•	Integrate WNTR with EPANET. To make the capabilities of WNTR more widely
accessible, this work would be directed towards integrating WNTR capabilities within the
new EPANET user interface (through the Python plug in) as well as future versions of
EPANET. A beta version of EPANET 2.2 includes some of the capabilities of WNTR -
pressure dependent demands and pipe leak models. This work could extend this new
EPANET version to incorporate some of the other advances included in WNTR.
Alternately, the use of WNTR through the new EPANET user interface Python plug-in
could be demonstrated to show others how to use WNTR through EPANET.
10.2.7 Premise Plumbing
•	Develop risk assessment tool for lead in homes. This work would include enhancements
to EPANET to incorporate dispersion, validation of the EPANET model with laboratory
data, and development of a risk assessment framework in a Python based tool that builds
on the models. This work would continue the 30+ week validation study in a real-scale
home plumbing system simulator at EPA, validate results, and develop a Python-based
demand simulator.
•	Develop a risk assessment tool for Legionella in large buildings. This work would
investigate the use of EPANET for modeling Legionella in buildings, such as hospitals. A
long-term study of Legionella within the home plumbing system simulator has also been
underway, and data from that can be leveraged to support the modeling effort. In
addition, EPA researchers have been working with several hospitals to mitigate risks and
add treatment, and at least one hospital is willing to share the data needed to develop an
EPANET model of their drinking water infrastructure.
•	Develop home and building "calculators." This potential work area would investigate and
develop premise plumbing "calculators" that run EPANET under the hood and provide
30

-------
EPA/600/R-19/174
October 2019
information useful to homeowners or building operators, similar to EPA's National
Stormwater Calculator which uses Storm Water Management Model. These calculators
could provide information about expected lead concentrations in water given homeowner
information, or expected benefits from flushing prior to drinking, or using filters, or other
mitigation techniques. In addition, operators of large buildings could use the tool to
minimize Legionella or other risks.
•	Promote the scientific study of water quality in premise plumbing. Premise plumbing has
become an important issue in the water distribution field due to water quality and security
concerns. In this work area, EPA would take an active role in studying and promoting the
scientific investigation of premise plumbing with special emphasis on modeling premise
plumbing.
10.3 Future of Drinking Water Brainstorming List
•	Artificial intelligence
•	Intelligent water networks: cyber communication networks, smart pipes
•	Cloud computing
•	Point-of-use treatment
•	Real-time response
•	Prescriptive analytics
•	Predictive rather than reactive
•	Net zero water
•	Conservation
•	Changing workforce
•	Changing behavior patterns
•	More engaged consumer/customer expectations
•	Paradigm shift in the way we look at innovation
•	Prioritization for replacement of aging infrastructure
•	New infrastructure materials
•	Valuation of water - use of bottled water
•	Message on water quality of water supply
•	Fuel cells as water supply
•	Drivers for change
•	More technology
•	Drinking water vs fire protection water, changes in firefighting techniques
•	Future design criteria
•	Dual systems
•	Improvements in hydrants, sensors, meters
•	Proliferation of sensors
•	Full tracking and accounting of water
•	Integrated water and waste water systems
•	Calibrated models, full system models
•	Decentralized treatment
•	Increase in cyber security
•	Redundant control systems
31

-------
vvEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300

-------