Threat Ensemble Vulnerability Assessment -
Sensor Placement Optimization Tool
(TEVA-SPOT) Graphical User Interface User's Manual
Version 2.3.1
September, 2012
U.S. EPA National Homeland Security Research Center (NHSRC)
EPA/600/R-08/147
-------
Disclaimer
This manual contains specific examples for demonstration purposes in order to easily familiarize
a user with the TEVA-SPOT Graphical User Interface version of the software.
If the user is not familiar with the various components that comprise a water distribution system
and how these are represented in a model, please review the EPANET User's Manual. If the
user needs more information about the sensor placement optimization tool (SPOT), please refer
to the TEVA-SPOT Toolkit Users Manual, which is included with the TEVA-SPOT software.
The U.S. Environmental Protection Agency through the Office of Research and Development
funded, managed, and participated in the research described here under an interagency
agreement. The views expressed in this TEVA-SPOT Tutorial are those of the authors and do not
necessarily reflect the views or policies of the U.S. EPA. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
Work at Argonne National Laboratory was sponsored by the U.S. EPA under an interagency
agreement through U.S. Department of Energy Contract DE-AC02-06CH11357.
Portions of TEVA-SPOT software resulted from work under Contract No. DE-AC02-
06CH11357 between the U.S. Department of Energy (DOE) and the University of Chicago,
Argonne, LLC for the operation of Argonne National Laboratory.
The TEVA-SPOT graphical user interface software must not be redistributed or sold.
-------
Questions or Problems
Questions concerning this document or the TEVA-SPOT software should be addressed to:
Robert Janke
U.S. EPA/NHSRC
Mail Stop NG-16
26 W. Martin Luther King Drive
Cincinnati, OH 45268
(513) 569-7160
i anke.robert@epa.gov
Problems with the TEVA-SPOT software can be handled by emailing Robert Janke at the
above email address
in
-------
Added or Changed Features...
• Incorporation of complete EPANET-MSX capabilities... refer to the EPANET-MSX Users Manual
for details. This resulted in a number of changes that are not readily visible to the user, such as the
significant change to the underlying data storage structure for the EPANET-MSX contaminant
concentration data. To run MSX capabilities within TEVA-SPOT, perform the following steps (in
this order):
o Create a new ensemble
o Load the INP file
o Load the EPANET-MSX file
o Specify the contaminant(s) under Injection Definition, specify SINGLE or MULTIPLE
contaminants.
o For SINGLE contaminant, choose particular species from drop down box and then enter data
as before, i.e., NAME, SPECIES, START and STOP times.
o For MULTIPLE contaminants, Add MASS INJECTION RATE, CONCENTRATION ZERO
THRESHOLD and WATER QUALITY TOLERANCE values for each available
contaminant under the "SPECIES" drop down box.
o Available "SPECIES" are determined by what is designated under [SPECIES] in the .MSX
file.
o Proceed as before....
• Node set definitions, ability to specify ensemble of contamination scenarios using a random number
generator process.
• Output file(s) (C:\TEVA-SPOT-Database\COLLECTION NAME\ENSEMBLE NAMEVHealth
Impacts Analysis) that details number of nodes with population that received a dose at or above the
specified dose levels.
• Infrastructure Impacts Analysis (IIA) module. Output file(s) in (C:\TEVA-SPOT-
Database\COLLECTION NAME\ENSEMBLE NAMEMnfrastructure Impacts Analysis). Tab
delineated text file provides total number of pipe feet contaminated (based on witnessing
contamination) at or above each specified concentration (e.g., mg/L) indicated.
• Ability to run Health Impacts Analysis without dose response functionality.
• User defined model for the timing of tap water ingestion. 0 (12:00 AM) to hour 24.
• Addition of the specification of the "Number of worst case dosage scenarios able to be saved".
Present in the Health Impacts Analysis module, allows the user the ability to analyze detailed
concentration data for a small number of scenarios based on dose levels.
THINGS TO BE CAREFUL ABOUT OR AWARE OF....
Flow Units
All flow units within an EPANET inp file are converted to gallons per minute. For example, cubic meters
per hour flow rate units are converted in TEVA-SPOT to gallons per minute. If the VC impact measure is
in units of gallons, there are 264 gallons per cubic meter, so results will appear quite high if believed to be
cubic meters.
Running simulations with very small dose levels...
With small and very small injection amounts, there are many nodes with very small concentrations which
turn into very small doses, which then get accounted for in the small dose thresholds which can result
large errors. This can be overcome or reduced by specifying a value in "Concentration Zero Threshold
mg/L" box of "Injection Definition" that is 1E-6 or greater.
IV
-------
Using the "Zero Concentration Threshold" in Injection Definition...
The "Concentration Zero Threshold" is applied at each water quality time step, not at the reporting
interval. The difference being that if your water quality time step is less than the reporting interval, the
time-weighted average concentration could be smaller than the zero threshold value! For example, if only
one of the water quality (WQ) steps had a concentration equal to the zero threshold and all other WQ
steps in the hour were either 0 or below the threshold.
VISUAL C for MSX Analyses
MSX-based analyses run much slower than standard (tracer or first order decay) simulations in EPANET.
Using a Visual C compiler can help reduce runtimes. "VC" is specified in the .MSX file for COMPILER.
For 32 bit computers, the visual studio express can be downloaded from the web for free, but for 64 bit
computers the only option is to buy the visual C program, i.e., M.S. visual studio.net. When installing
TEVA-SPOT GUI on a computer that is 64 bit, the only option is to have the full version of visual studio
otherwise the software will assume that no VC compiler is available and the resulting simulations will run
much, much slower.
Removing ensembles that were deleted from showing up in the TEVA-SPOT "Load Ensemble" Window
In order to remove the ensembles that you deleted from showing up in TEVA-SPOT, for instance when
you try to load an ensemble, please follow these steps:
Open TEVA-SPOT and choose "Ensemble Management". You may need to browse to the appropriate
collection.
1. Identify all the ensembles that were deleted or will no longer load. Write down their names.
2. Close TEVA-SPOT
3. Open Windows Explorer or a MY COMPUTER window and browse to the TEVA-SPOT
collection within the TEVA-SPOT-Database where the ensembles were located.
4. Create a blank folder in the directory or the TEVA-SPOT collection where the ensemble was
located for each of these ensembles that you want removed. Be sure to name each folder exactly
as it appears under "Ensemble Management". We are re-creating a folder in order to give TEVA-
SPOT something to delete.
5. Open TEVA-SPOT, choose "Ensemble Management" again and select or highlight those
ensembles that you had previously deleted through Windows Explorer. Once selected, choose
DELETE.
6. The ensembles should now be removed.
-------
Forward
Since its inception in 1970, EPA's mission has been to pursue a cleaner, healthier environment
for the American people. The Agency was assigned the daunting task of repairing the damage
already done to the natural environment and establishing new criteria to guide Americans in
making a cleaner environment a reality. Since 1970, the EPA has worked with federal, state,
tribal, and local partners to advance its mission to protect human health and the environment. In
order to carry out its mission, EPA employs and collaborates with some of the nation's best
scientific minds. EPA prides itself in applying sound science and state-of-the-art techniques and
methods to develop and test innovations that will protect both human health and the
environment.
Under existing laws and recent Homeland Security Presidential Directives, EPA has been called
upon to play a vital role in helping to secure the nation against foreign and domestic enemies.
The National Homeland Security Research Center (NHSRC) was formed in 2002 to conduct
research in support of EPA's role in homeland security. NHSRC research efforts focus on five
areas: water infrastructure protection, threat and consequence assessment, decontamination and
consequence management, response capability enhancement, and homeland security technology
testing and evaluation. EPA is the lead federal agency for drinking water and wastewater
systems and the NHSRC is working to reduce system vulnerabilities, prevent and prepare for
terrorist attacks, minimize public health impacts and infrastructure damage, and enhance
recovery. This Users Manual for the TEVA-SPOT GUI software is published and made
available by EPA's Office of Research and Development to assist the water utility community.
Jonathan Herrmann, Director
National Homeland Security Research Center
Office of Research and Development
U. S. Environmental Protection Agency
VI
-------
License Notice
The U. S. Government is granted for itself and others acting on its behalf a paid-up, non-
exclusive, irrevocable worldwide license in the binary software libraries associated with TEVA-
SPOT to reproduce, prepare derivative works, and perform publicly and display publicly,
including the right to distribute to other Government contractors.
The binary software libraries may be distributed, provided that they are only distributed as part,
and remain a part of the TEVA-SPOT software. Argonne has agreed that all users of TEVA-
SPOT and derivative systems have a perpetual, royalty-free license to use and distribute for free
JeoViewer, DIAS, and MSV Java Utility Library binary software as part of TEVA-SPOT.
However, the source code is not provided as part of this licensing agreement. The existing
executables for Jeo Viewer, DIAS, and MSV Java Utility Library, as well as any improvements
and modifications to these systems, are and will be freely available to the EPA and the users of
TEVA-SPOT as part of this effort.
Portions of TEVA-SPOT are being distributed under the Lesser GNU Public License (LGPL).
The LGPL is described in the LICENSE.lgpl file included with the software. TEVA-SPOT
includes a variety of other software packages with different licenses. The use of this software
should be for research purposes, there are restrictions on commercial usages in some cases:
- EPANET: Lesser GNU Public License
- tevaUtils: Lesser GNU Public License
- grasp: AT&T commercial license
- randomsample, sideconstraints - ATT Software for noncommercial use.
- ufl - Common Public License
vn
-------
Acknowledgements
The National Homeland Security Research Center would like to acknowledge the following
organizations and individuals for their support in the development of the TEVA-SPOT User's
Manual and in the development and testing of the TEVA-SPOT Software.
Office of Research and Development - National Homeland Security Research Center
Robert Janke
Regan Murray
Terra Baranowski-Haxton
Argonne National Laboratory
Thomas Taxon
Science Applications International Corporation
Rakesh Bahadur
William Samuels
Sandia National Laboratories
Jonathan Berry
Erik Boman
William Hart
Lee Ann Riesen
Cynthia Phillips
Jean-Paul Watson
University of Cincinnati
James Uber
American Water Works Association Utility Users Group
Kevin Morley (AWWA)
Vlll
-------
Table of Contents
Disclaimer ii
Questions or Problems iii
Forward iv
License Notice v
Acknowledgements vi
List of Figures and Tables ix
1. Introduction 1
1.1. Data Structure Concepts 1
1.2. Installation 2
2. Getting Started with TEVA-SPOT 3
2.1. Tutorial Example 4
2.2. Starting TEVA-SPOT 5
2.3. Collection 5
2.4. Ensemble 6
2.5. Import EPANET Input File 6
2.6. Population 8
2.7. Execution Control 9
2.8. Ensemble Options 10
3. Ensemble Analysis Mode 13
3.1. Injection Definitions 14
3.2. Node Set Definitions 16
3.3. Node Injections 18
3.4. Scenario Sets 19
3.5. Base Ensemble Execution 21
4. Health Impacts Analysis 22
4.1. Contaminant Name 24
4.2. Dose Calculation Parameters 26
4.3. Population Parameters 27
4.4. Dose Thresholds 28
4.5. Miscellaneous 30
4.6. Health Impacts Analysis Execution 31
5. Sensor Placement Algorithm 32
5.1. Specifying the Solvers 33
5.2. Specifying the Sensor Set Sizes 34
5.3. Specifying the Objectives 36
5.4. Dose Threshold 37
5.5. Specifying the Response Times 38
5.6. Specifying the Constraint Sets 39
5.7. Specifying the Location Categories Sets 40
5.8. Specifying the Costs 41
ix
-------
5.9. Specifying the Detection Limits 41
5.10. Specifying the Selection Method 42
5.11. Sensor Placement Execution 42
6. Maps 43
6.1. Default Map 43
6.2. Maps Produced by Health Impacts Analysis 44
6.2.1. Demand-Based Population Estimate 44
6.2.2. Estimated fatalities by Injection Location 44
6.2.3. Additional Maps 45
6.3. Maps Produced by the Sensor Placement Algorithm 46
6.3.1. Sensor Locations 47
6.3.2. Detected Events (Nodes Only) 48
6.3.3. Detected Events (Links and Nodes) 50
6.3.4. Detected Events (Links Only) 50
6.3.5. Sensor Counts 50
6.4. Charts 51
6.4.1. Average Estimated Fatalities by Time 52
6.4.2. Additional Charts 52
6.5. Tables 54
6.5.1. Health Impacts Analysis - Injection Impact Table 55
6.5.2. Health Impacts Analysis - Additional Tables 56
6.5.3. Sensor Placement Summary 56
6.5.4. Trade-off Analysis Data 56
6.5.5. Impacts and Detect!on Time Tables 57
6.5.6. Sensor Ranking 58
7. Filtering Criteria 59
7.1. Filtering Controls 59
7.2. Manage Button 59
7.3. Creating a Filter Specification 59
8. Regret Analysis Mode 64
8.1. Input Data for Regret Analysis 65
8.2. Mode Menu 66
8.3. Regret Menu 66
8.4. New 66
8.5. Ensemble Menu 67
8.6. Load 67
8.7. Save 68
8.8. Close 68
8.9. Execution Control 68
8.10. Specifying the Parameters for Regret Analyses 69
8.11. Add sensor designs 71
8.12. Selecting Impact Sets 72
8.13. Scheduling the Available Executions 74
9. Regret Analysis Output - Tables 75
10. Regret Analysis Mode: Filtering Criteria 77
11. Ensemble Management: Export Capabilities 78
-------
11.1. Import Ensemble 78
11.2. Export Ensemble 79
11.3. Export Map Data 80
12. Trouble-Shooting 80
12.1 Stop & Restart TEVA-SPOT Services 81
12.2. Log Files 82
12.3. Example Log File 82
APPENDIX A Water Utility Requirements for using EPA's TEVA-SPOT software 84
Utility Network Model 84
Sensor Characteristics 89
Design Basis Threat 91
Performance Measures 92
Utility Response 92
Potential Sensor Locations 93
Population 95
Additional Reading 97
List of Figures
Figure 1. Data flow chart of TEVA-SPOT 1
Figure 2. Flow chart for EPANET Simulations 13
Figure 3. Health Impact Analysis flow Chart 22
Figure 4. Sensor Placement Flow Chart 32
List of Tables
Table 1. Example Input Parameters for TEVA-SPOT Ensemble Mode 4
Table 2: Information and data required to design sensor networks using TEVA-SPOT 84
Table 3: Sensor File for Use in Sensor Placement Module 90
Table 4: Fields of Sensor Data Files for Loading Sensors into Health Impacts Assessment (HIA)
Sensors 90
Table 5: Meaning of the Keywords in Location Categories Files 94
Table 6: Effect of Keyword in LC File, and Treatment of Nodes by SP Algorithm 94
Table 7: Fields of Population Data Files 95
XI
-------
1. Introduction
Threat Ensemble Vulnerability Assessment - Sensor Placement Optimization Tool (TEVA-
SPOT) allows users to define a water network contamination "scenario," simulate the spread of
the contaminant or contaminants throughout the water network, analyze the consequences, and
display the results in a variety of graphical and tabular forms. The ultimate aim of a simulation
and its subsequent analysis is to determine the vulnerability of the distribution system to
contaminant releases and determine the optimal locations to place a set of water quality sensors
in the network to mitigate the impacts of contamination. This Users Manual helps to familiarize
users with the TEVA-SPOT Graphical User Interface (GUI) functionality. TEVA-SPOT
assumes ones familiarity and use with EPANET. Users should first be sure their network model
runs in EPANET prior to running the model in TEVA-SPOT.
1.1. Data Structure Concepts
The TEVA-SPOT user typically creates an ensemble of contaminant release locations in a water
network, simulates contaminant release(s), performs various ensemble analyses on the results,
and then displays the results. The goal of these steps is to determine and characterize certain
effects of the contaminant release, given its inherent uncertainties. For example, without
knowing precisely which nodes represent potential contaminant release locations, one can
determine average and worst case exposures, doses, infections, and estimated fatalities on a
collection of all possible release locations. This section defines and discusses the principal data
structures. These objects are hierarchically organized as shown in Figure 1.
Figure 1. Data Flow Chart of TEVA-SPOT
-------
The user must assign a name to each ensemble, injection definition, node set definition,
collection, regret analysis group, and regret analysis when it is created, and may assign a name to
each scenario set. Node injections and scenarios do not have user-assigned names but can be
considered to have a two-part name consisting of those of the injection definition and node set
definition on which they are based.
Two modes of operation exist for TEVA-SPOT: (1) ensemble analysis mode and (2) regret
analysis mode. Ensemble analysis mode is found on starting TEVA-SPOT. Ensemble analysis
mode is where contaminant vulnerability analyses and sensor network designs are performed.
Regret analysis mode is where sensor network designs can be further evaluated.
Regret analysis is a trade-off analysis that determines how well a particular sensor design meets
an objective(s) other than the one they were intended to achieve. Regret analysis uses the impact
files and sensor designs generated by TEVA-SPOT and designated by the user in order to create
or edit regret analysis parameters.
1.2. Installation
The distribution executable (.exe file) contains the TEVA-SPOT program and Users Manual.
The prerequisite software can be obtained from the FTP site below or from the internet by
searching for the required items below. Prerequisite software includes the Java Development Kit
(JDK) Version 1.6 update 20 and the Python Scripting Language, version 2.6 or later. The JDK
and Python products need to be installed first prior to the installation of TEVA-SPOT. The JDK
and Python versions installed need to be consistent with the computer's capabilities. In other
words, use the 64 bit versions of JAVA and Python if hardware and software on the host
computer is 64 bit compatible. TEVA-SPOT has been tested on Windows XP and Windows 7,
not Windows Vista. Below are the JDK and Python installers needed depending on the computer
hardware and operating system that TEVA-SPOT will be installed on:
• jdk-6u20-windows-586.exe (32 bit version)
• jdk-6u20-windows-x64.exe (64 bit version)
• python-2.6.5.msi (32 bit version)
• python-2.6.5amd64.msi (64 bit version)
After the prerequisites of JAVA and Python are installed, the TEVA-SPOT system is installed
and configured to the c:\Program Files\TEVA-SPOT folder. Output files generated by TEVA-
SPOT are stored in c:\TEVA-SPOT-Database. During installation the drive location can be
changed if desired, e.g., D:\Program Files\TEVA-SPOT.
Updates to TEVA-SPOT can be installed by downloading the update from the EPA TEVA
Research Program website (http://www.epa.gov/nhsrc/water/teva.html). TEVA-SPOT can also
be obtained from the following File Transfer Protocol (ftp) website: ftp://scienceftp.epa.gov/ and
entering the appropriate login and password information which can be obtained via email from
janke.robert@epa.gov. All users who request TEVA-SPOT will be automatically placed on an
email list for notification of updates!
-------
Each TEVA-SPOT installation (.exe) file is tagged with a date of release, e.g., TEVA-
SPOTInstaller-2.3.1-MSX Beta 201100926.exe, signifying a release date of September 26, 2012
(last number in the file name). The initial two numbers in the file name describe the version
number, i.e., 2.3.1, refers to the major release number (2) and the minor release number (3) and
are consistent with the version numbering for the TEVA-SPOT toolkit. The final number in the
file name that describes the version number, i.e., 1, designates a sub-minor release which is
specific to TEVA-SPOT graphical user interface.
Older versions of TEVA-SPOT should be removed first using the following procedure before
installing the updated version:
• Choose STOP TEVA-SPOT Services from START Programs/TEVA-SPOT
• Move the "contaminants.xml file" from the "client" folder under Program Files/TEVA-
SPOT/client to suitable location. During installation of a TEVA-SPOT update the
contaminants.xml file is replaced with a blank version. After installation replace the
blank contaminants.xml file with the saved contaminants.xml file.
• Choose Uninstall TEVA-SPOT from START Programs/TEVA-SPOT
• Navigate to Programs/TEVA-SPOT and Remove the TEVA-SPOT folder and its
contents.
• Install the TEVA-SPOT update by double-clicking on the TEVA-SPOT executable icon.
Updates are installed without lost of older ensemble data residing in the TEVA-SPOT-Database
folder.
2. Getting Started with TEVA-SPOT
This section provides a quick overview of TEVA-SPOT, including input parameters,
management of ensembles and collections, and starting and executing TEVA-SPOT. It is
important to understand that running TEVA-SPOT initiates TEVA-SPOT services that work
independently of the graphical user interface (GUI) client. In other words, closing TEVA-SPOT
after an ensemble or regret analysis has started will not stop the simulations or analyses.
Choosing "Terminate" will stop the TEVA-SPOT services but it may take some time to complete
depending on the EPANET model and the option running at the time "Terminate" is selected.
Choosing "Stop TEVA-SPOT Services" or "Restart TEVA-SPOT Services" will stop TEVA-
SPOT simulations and analyses immediately.
Therefore, if it is desired to stop TEVA-SPOT quickly in order, for example, to change input
parameters, select "Restart TEVA-SPOT Services," from the START Menu, then close the
TEVA-SPOT client, and then reopen TEVA-SPOT and load the appropriate ensemble and
continue as needed.
If you are experiencing problems it usually helps to initiate the Restart TEVA-SPOT Services,
close down TEVA-SPOT, and then restart TEVA-SPOT after the Restart TEVA-SPOT
Services has completed.
-------
2.1. Tutorial Example
Table 1. Example Input Parameters for TEVA-SPOT Ensemble Mode
EPANET
Simulations
Health
Impact
Analysis
TSOto
Impacts
Analysis
Infrastructure
Impacts
Analysis
Sensor
Placement
Input Parameter
Simulation Time
Scenarios (Set of release nodes)
Time of Release (Duration)
Population Estimate
Mass Injection Rate (mg/min)
Contaminant
User designates: Response Time
Delay, Detection Limit
Base filename and then selects which
impact metrics are desired.
Contaminant name
Concentration thresholds
Exposure Model (Water ingestion)
Sensor Set Sizes
Response Delay Times
Detection Limits
Sensor Placement Algorithm
Sensor Placement Objective/Statistic
Parameter Value
24 hour
All non-zero demand nodes
Ihr
Demand Based
1.74E+04 mg/min
(The Mass Injection Rate is always in mg/min)
Enter contaminant data
Response time delay: minutes
Detection limit: mg/L of chemical or toxin or
organisms, cells, or user specified "units" per liter.
Concentration units: mg/L of chemical or toxin or
organisms, cells, or user specified "units" per liter.
Demand Based
1,5,10, 15,20,25
0, 360, 720 minutes
0, 0.01, 1 milligrams per liter
Heuristic
Population Exposed, Mean, or another metric
Table 1 shows input parameters to run TEVA-SPOT in Ensemble Mode. Remember TEVA-
SPOT is based on running an extended period simulation based on the introduction of a
contaminant. Therefore. QUALITY must be set to CHEMICAL in the EPANET input (*.inp)
file.
-------
2.2. Starting TEVA-SPOT
The TEVA-SPOT simulation
can start by going to the Start
Menu and clicking "All
Programs," and navigating to
TEVA-SPOT
This will open the TEVA-
SPOT frame as shown.
H TEVA-SPOT (No Ensemble Loaded)
Ensemble File Mode Edit About
Map [ Charts || Tables]
Map Default
2.3. Collection
The first step is to define the Collection Management.
Navigate to Collection Management under the
Ensemble Menu. This will open a Collection
Management window.
At this stage there are no names in the window.
Create a new Collection by choosing the New button.
This will open New Ensemble Collection window. The
OK button is initially disabled (grayed out), whereas the
Cancel button is always enabled.
(On some computers, the window "New Ensemble
Collection" may not completely open because of screen
resolution. In this case, drag the window until the OK
button is visible). The window can be resized in both the
horizontal and vertical directions.
New...
Load...
Save
Save As,,.
Close
Ctrl+N
Ctrl+0
Ctrl+S
Collection Management,,,
Ensemble Management,,,
ImportEPANET.inpFile,,, Ctrl+I
Population
HIA Sensors
Execution Control Ctrl+E
Available Ensemble Collections
April_201D_Tests [Main]
Ensemble Collection Inform^
Ensembles
Regret Analyses
n^~u
1 / 663.0 bytes
0 / 0,0 bytes
668,0 bytes
Rename J 1
[ Close"
Move J [ Delete ]
D
Click the Ensemble Name text box and type the name
DEMO for the new ensemble.
Caution: Avoid entering a name that contains embedded
blanks, e.g., "SmallTown" is okay, but "Small Town" is
not.
The OK button becomes enabled. Click the OK button.
Close the Collection Management window. The screen
will still read "No Ensemble Loaded."
i New Ensemble Collection
Ensemble Collection Location Main
Ensemble Collection Name | Demol
"OK I I Cancel
-------
2.4. Ensemble
The next step is to create a New Ensemble.
In the Ensemble Menu, click New.
The New Ensemble dialogue box opens.
From the Ensemble Collection pull-down menu,
select Demo [Main].
Click the Ensemble Name text box and type the
name Demo. This is the name of the Ensemble.
Click the OK button. The dialogue box closes.
New,,,
Ctrl+N
Load,..
Save
Save As,..
Close
Ctrl+O
Ctrl+5
Collection Management...
Ensemble Management,,,
Import EPANET , inp File... Ctrl+I
Population
HIA Sensors
Execution Control Ctrl+E
New Ensemble
Ensemble Collection Demo [Main]
Ensemble Name Demo
At the top of the screen, it will show TEVA-SPOT (Demo in Demo [Main]).
= TEVA (Demo in Demo [Main])
Ensemble File Mode Edit
Tables
Map
Q, X i^ Map Default
Filtering
2.5.
Import EPANET Input File
In the Ensemble Menu, click Import
EPANET.inp File
Note: Import EPANET.inp File is only
available as an option when an ensemble is
loaded. EPANET .net files are not supported by
TEVA-SPOT.
An Open dialogue box opens to select the
EPANET input file.
The imported EPANET .inp file MUST
have QUALTIY set to CHEMICAL and
all SOURCES removed. If SOURCES
are present TEVA-SPOT will remove
them with a notification to the user.
[He Mode Edit
New... Ctrl+N
Load... Ctrl+O
Sav« Ctrl+S
Save As...
Close
CofecBon Management..
Ensemble Management..
| Import EPANET.inp Fte... Ctrl+l |
Population >
Htft Sensors M
Execution Control Ctrl+E
-------
Browse to the
directory where your
network model is
located (where the
EPANET input file is
located).
Click the icon for
your network, similar
to the
DemoNetwork.inp at
right. Choosing the
*.inp file places the
name of the file into
File name: text box
Click the Open
button to import your
network. For
demonstration
purposes the
following,
DemoNetwork.inp
file is used.
OOpen
Look in; |ir= input
My Recent
Documents
Desktop
My
Documents
My
Computer
My Network
DemoNetwork. inp!
Places Files of type: EPANET input file (*. inp)
ine aiaiog DOX
closes, and the
desired water
network is imported
as shown.
7
-------
2.6. Population
The Population command is always enabled (never grayed out), although when no ensemble is
loaded into memory, all the subcommands of the Population submenu are disabled (grayed out).
Click Population in
the Ensemble menu
to display the
Population
submenu.
Click Import
Population Data...
in the Population
sub menu. An Open
dialogue box opens.
New..,
Load...
Save
Sav«As...
Close
Qrl+N
Orl+0
Cfrf+S
Collection Management..
Ensembfe Management...
ImpcrtEPAisET .inpflie... Orl+l
Import Population Ddta..
Execution Control
Orl+E
Navigate to the
directory containing
the population data
file to be imported.
Select your
population data file
(e.g., Population.txt).
Choosing this file
places the name of
the file into File
name: text box.
Click the Open
button to import the
selected population
data.
Note: The procedure
to create the
Population.txt file is
discussed in
Appendix A. Also
note that if using a
demand-based
population model no
imported population
file is needed.
ir= input
^ ,J .-'IBB
My Recent
Documents
Desktop
My
Documents
My
Computer
My Network
|s.'DernoNet»ork.inp
!sJDemoNet*ork..NET
Fife name; Population.tot
Places RiBS of type: A|| Fi|BS
-
TEVA-SPOTcontains two types of population models: Demand and
Census. Census refers to the user specifying the number people at
each non-zero demand node and loading the associated data into
TEVA-SPOT via a text file.
-------
2.7. Execution
Control
Click Execution
Control in the
Ensemble menu.
The Execution Panel
is added
Note the TFVA-
OD/^T1 <~,_ J,,1_,,.
brUl modules.
D EPANET
si rnul 3.ti on s
Health
T
impacts
Analysis
(HIA)
D TSO-to-
Impacts
Analysis
D Infrastructure
Impacts
Analysis
D Sensor
Placement
The TSO-to-Impacts
Analysis is the stand-
alone application.
A separate TSO-to-
Impacts Analysis
occurs as the initial
step in Sensor
Placement.
|F*e Mode Edit
New.,. Ctrt+N
Load... Ctrl+O
Save Ctrl+S
Save As,..
dose
Colectfon Management..
Bnsemble Management,.
Import BPANET.ip Fife,,. Ctrl+l
Population >
HIA Sensors *
Execution Control Ctrl+E
i
1 Execution H
.. , -
O D EPANET Simulations Edit tscimaced , ,me Kemammg
*„__, , Estimated Completion Time
Q |_J TSO to Impacts Analysis Edit
| Q Q| Infrastructure Impacts Analysis Edit
'
Q |_| Sensor Placement Edit
-------
2.8. Ensemble
Options
Click Ensemble
Options in the Edit
menu.
The Ensemble
Options box allows
users the ability to
modify key EPANET
parameters associated
with: Time, Water
Quality, Hydraulic,
Simulation, and
Decay.
The Description text
box can be used to
record pertinent
information about the
ensemble (e.g.,
reasons for the
ensemble, network
model information
and dates, essentially
anything that would
explain or document
the results for future
use).
Under Simulation
Parameters, the Use
all servers is checked
as a default.
Uncheckthis if a
large network
(»13,000 nodes) is
being used. This will
result in only one
processor being used
for EPANET and
HIA (if run together
or only EPANET if
run alone) and TSO-
2-Impacts analyses if
El Ensemble Options
Description
test]
Time Parameters—
Duration
Quality Timestep
Hydraulic Timestep
Reporting Interval
Status: OK
163
hours
minutes v
-Simulation Parameters—
0 Use all servers (Uncheck for extremely large networks)
Q Perform Mass Balance Analysis
hours
hours v
Water Quality Parameters
0 Average concentration over reporting interval
WQ Tolerance Value l.OE-6
Hydraulic Parameters
Maximum Trials
CHECKFREQ
MAXCHECK
DAMPLIMIT
40
10
0,0
Unbalanced Condition | Continue w/trials
Additional Trials flO
Number of simultaneous SP executions] Max
Decay Parameters -
Decay Name chemical
rReaction Order—
Bulk |l v|
Wall |l v|
Tank 11 vl
Reaction Coefficients
Bulk 0.0
Wall 0.0
Limiting Concentration 0.0
Wall Coeff. Correlation faO~
Cancel
NOTE: Choosing PERFORM MASS BALANCE
ANALYSIS will increase run times!!
Mass tracking data for each simulation or scenario will appear in two
text files (*.txt) within the TEVA-SPOT-Database with the file names:
(1) massData.txt and (2) massDataWithCalc.txt.
massDa.ta.txt file provides for each scenario: mass injected, mass
removed, mass in tanks, and mass in pipes for each time step of the
simulation.
massDataWithCalc.txt provides for each scenario the same data but
with additional calculations as follows: Mass Injected (I), Mass
Removed (R), Mass In Tanks (T), Mass In Pipes (P), Mass In Pipes (I-
R-T), Mass In Network (I-R), and Mass In Network (T+P)
Each of these text files will have the ensemble name as a prefix to the
name of the text file.
10
-------
the stand-alone
application is used.
Perform Mass
Balance should be
unchecked and used
only by the more
experienced user.
This function
provides for each
scenario contaminant
mass tracking results
in terms of mass
injected, removed,
mass in pipes and
mass in tanks. The
resulting text files
appear within the
TEVA-SPOT-
Database/Ensemble
Name/BaseEnsemble
directory.
The Time,
Hydraulic, and
Decay Parameters
are consistent with
EPANET2.00.12.
For more
information, please
see the EPANET
Users Manual.
Under Decay
Parameters, the
Decay Name
specifies the name of
the contaminant that
is being studied.
Under Water
Quality Parameters,
Water Quality
(WQ) Tolerance
specifies the water
quality tolerance used
Time Parameters
Duration
168
i
Quality Tirnestep 1
Hydraulic Timestep 1
Reporting Interval 1
Status: OK
Water Quality Parameters
0 Average concentration over reporting interval
WQ Tolerance Value 11 ,OE-6
Hydraulic Parameters
Maximum Trials
CHECKFREQ
MAXCHECK
DAMPLIMIT
Unbalanced Condition
Additional Trials
Simulation Parameters
Number of simultaneous SP executions Max
Decay Parameters
Decay Name chemical
H71 Use all servers (Uncheckfor extremely large networks^
| | Perform Mass Balance Analysis I
Bulk
Wall
Tank
-Reac
Bulk
Wall
1 v|
0.0
0,0
Limiting Concentration
Wall Coeff. Correlation
0.0
0.0
Decay Parameters
Decay Name
-ks)
H
n
Keacc
Bulk
Wall
Tank
React
Bulk
Wall
ion uroer
0 v
0 v|
0 v
ion Coefficie
0,0
0.0
Limiting Concentra
Wall Coeff. Correlc
ion 0,0
tion 0.0
OK
Cancel
Note: If Reaction Order and Coefficients are used the results of
contaminant decay will be seen in the Health Impact results, both
baseline (no sensors case) and sensor design cases.
11
-------
in the water quality
simulation. This
value will be
overridden by any
values specified
under an Ensemble
Injection Definition
Note the check box
for Number of
simultaneous SP
executions. SP refers
to sensor placement.
Various options
(Max, 1,2, etc.) will
be available in the
drop down box
depending on the
computer hardware
and memory
resources available
on the host computer.
The options provide
the ability to
minimize the parallel
processing of sensor
network designs to
conserve RAM
memory. If the
model is large (i.e.,
greater than 10,000
nodes) and the
number of feasible
sensor locations is
also large (i.e.,
greater than 10,000),
the user may need to
set this to something
other than "Max," in
order for the sensor
designs to be
completed.
Time Parameters
Duration
Quality Timestep
Hydraulic Timestep
Reporting Interval
Status: OK
168
1
30
30
hours
V
minutes
V
minutes
V
minutes
V
Water Quality Parameters
Average concentration over reporting interval
WQ Tolerance Value
0,01
-Hydraulic Parameters
Maximum Trials
CHECKFREQ
MAXCHECK
DAMPLIMIT
Unbalanced Condition
40
10
0.0
CONTINUE
For more information on Water Quality Tolerance parameter
please refer to the EPANET Users Manual.
Ensemble Options
Description
Time Parameters
Duration
Quality Timestep 1
Hydraulic Timestep 1
Reporting Interval
Status: OK
Simulation Parameters —
0 Use all servers (Uncheck for extremely large networks)
DnarfgrmMi.rfEilimi8Anil.jfHr
Decay Parameters
Water Quality Parameters —
0 Average concentration over reporting interval
WQ Tolerance Value l.OE-6
Hydraulic Parameters
Maximum Trials J4I]
CHECKFREQ J2~
MAXCHECK
DAMPLIMIT
Unbalanced Condition
Additional Trials
10
0.0
Continue w/trials v
10
Number of simultaneous SP executions Max
Decay Name chemical
-Reaction Order
Bulk |l v|
Wall |l v|
Tank 11 vl
Reaction Coefficients -
Bulk 0.0
Wall 0.0
Limiting Concentration
Wall Coeff. Correlation
0.0
0.0
12
-------
3. Ensemble Analysis Mode
This section describes how to use the execution controls in the ensemble analysis mode. Figure 2
shows the hierarchy of the EPANET Simulations.
EPANET
Simulations
lat
Injection
Definition
Node Set
Definitions
Attributes
Non Zero
All Junctions
All Nodes
User Defined
Node
Injections
Enumerated
Pipe Diameter
Scenario
Sets
Figure 2. Flow Chart for EPANET Simulations
13
-------
3.1. Injection Definitions
In the Execution panel, click
the Edit button to the right of
the EPANET Simulations
check box.
This will open the EPANET
Simulations Parameters
dialog box. EPANET
Simulations will be grayed
out until information has been
added.
Available Executions—
Scenarios—
0 EPANET Simulations
Modules—
0 _. Health Impacts Analysis.
0 ISO to Impacts Analysis
Sensor Placement
0 ] Sensor Placement
Status
Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute Terminate
Click the Add button in the Injection Definitions panel of the Edit BaseEnsemble Parameters
dialog box.
H Edit BaseEnsemble Parameters
[ Import,,. ] [ Export,.. ]
nj-Lj-n E mi i->n-
_l ,- r p r
1 Ntfae j&t uerinitions
Non-Zero Demand Junctions (106)
All Junctions (407)
All Nodes (410)
Add |
Delete 1
Add ]
Edit
Delete i
Create » 1
Juac injections
1 Remove 1
| OK | | Cancel ]
Add »
m
-.cerrai 10 _.£ts
ip) Scenario Set
0 Simulations
[ New J
The Injection Definition dialog box opens.
Type in the required information specified in
the Injection Definition window.
Inserting a WQ Tolerance (mg/L) value will
override the WQ Tolerance specified in the
EDIT/ Ensemble Options box; however no
value need be specified.
Inserting a value (e.g., 0.000001) into the
Concentration Zero Threshold (mg/L) box
will result in zero values for all contaminant
concentrations below the indicated value. No
value need be specified.
H Injection Definition
Attributes
Name
Mass Injection Rate (mg/rnin)
Concentration Zero Threshold (mg/L)
WQ Tolerance (mg/L)
Start Time
Stop Time
Description
OK
Cancel
14
-------
Click the OK button to accept the injection
definition.
Only required data entries are: Name,
Mass Injection Rate, and Start and Stop
Times.
Injection Definitf
IName
Mass Injection Rate (mg/min)
Concentration Zero Threshold (mg/L)
WQ Tolerance (mg/L)
0.0
1 Start Time _J| hours v
St°P Time j |[j^ ;
Description
The Injection Definition dialog box closes, and the injection definition is added to the ensemble
and displayed in the Injection Definitions pane of the Edit BaseEnsemble Parameters dialog
box. (Note: 0.833 mg/min was entered as the Mass Injection Rate, the Start Time was 0 hour,
and the Stop Time was 1 hour).
@ Edit BaseEnsemble Parameters
[ Import... ] [ Export,., ]
toxic #1 (0.833 from 0 hours to 1 hour)
rJodt _.et Definitions
Non-Zero Demand Junctions (106)
All Junctions (407)
All Nodes (410)
1
Add ]
Ed*
Add ]
Edit
Delete
1 Create » |
Remove
| OK | | Cane
ei |
| Add » |
m
-•(p3* Scenario Set
0 Simulations
j New |
Import and Export allows the easy incorporation or export of Injection Definitions and Node
Set Definitions into TEVA-SPOT.
Edit BaseEnsemble Parameters
Injection Definitions
Node 5et Definitions
Non-Zero Demand Junctions (106)
All Junctions (407)
Nodes(410)
Add |
Node Injections
Scenario Sets
0 Simulations
| Mew |
15
-------
3.2. Node Set Definitions
Four options are available for Node Set Definitions. Node set defines the set of nodes that are
treated as potential contamination injection sites. The three pre-loaded options are:
• Non-Zero Demand Junctions
• All Junctions
• All Nodes (includes junctions, tanks, and reservoirs)
H Edit BaseEnsemble Parameters
[ Import:... ] [ Export... j
nj&. ion c mi ions
Cont ( 1 7400 , 0 from 0 hours to 1 hour)
won-iero uernana junctions (byj
All Junctions (92)
All Modes (97)
, m ,,
Delete
Add |
Edit
Delete |
ueate »
LJ c n]6C Ii3n=.
Ftemove
| OK | [ Cancel |
Add >:--
rxj
jCcHdl IO j6 a
0 Simulations
Pernova [New]
The fourth option is a user defined node set, which is a smaller sub-set of nodes. Click Add to
create a user defined node set. Two choices are listed Enumerated and PipeDiameter. To edit a
user defined node set, Click Edit to view the associated Node Set dialog window. Only the user
defined node sets can be edited.
Select Enumerated and a Node Set dialog window opens. This option is used to select specific
nodes as the contamination injection sites. Highlight nodes in the Available area and then Click
Add to add them to the Selected area. Click the Add from File button to use a pre-defined list
of nodes. In this Node Set window, the Name of the node is required along with at least one
node listed in the Selected area.
U Node Set
Name |
m j
Available
Junction 10 A
Junction 101
Junction 105
Junction 107
Junction 109
Junction 1 1 1
Junction 113 v
Add »
Remove ^
ot
IS
Selected
-------
Clicking "Add From File" will open a dialogue box:
Look in:
1 • ^k
1 i*|
My Recent
Documents
gr—\
l*J
Q Client y $ L?||!g[3
_|docs
Qjar
£3 tos
,_j modules
Q props
||) contaminants. xml
3] ensjemplate.xml
jj ensemble. xsd
^J modules. xsd
/ RQ rundient.bat
My Document?
My Computer
V
Places
>1 sp.xsd
>] teva-base.^sd
Filenme: [ Open
Ftfes of type; MF9a .«, | Cante!
Select PipeDiameter and another Node Set dialog window opens. This option is used to select
nodes that are connected to pipes of specified diameters. In this Node Set window, the Name of
the node set is required along with either a value in the Minimum or Maximum Diameter text
box.
Checking the box "Only include non-zero demand junctions" will result in only the non-zero
demand nodes associated with the pipe diameter specifications to be identified.
Name
Nodes
Minimum Diameter
Maximum Diameter
Q Only include non-zero demand junctions
17
-------
3.3. Node Injections
Select one Injection Definitions and one Node Set Definitions.
B Edit BaseEnsemble Parameters
[ Import:... ] [ Export,.. "]
Mimrni-IWTIimFlEllifflM
Non-Zero Demand Junctions (59)
41 Junctions (92)
Add ]
Edit |
Delete ]
Edit
Delete |
[ Create » ]
| OK | | Cancel |
| Add » |
m
0 Simulations
"••-. - . [ New ]
Click Create » button.
Edit BaseEnsemble Parameters
Import... Export,,.
Injection Definitions
Add ]
Edit |
Delete ]
-Node Set Definitions
Node Injections-
^on-Zero Demand Junctions (59)
All Junctions (92)
Injection: Cont / Nodes; All Nodes (97)
Scenario Sets-
Add >?
0 Simulations
Remove [New ]
The node injection formed by combining the selected injection definition with the selected node
set definition is created and displayed in the Node Injections pane of the Edit BaseEnsemble
Parameters dialog box.
18
-------
3.4.
Scenario Sets
Click the New button of the Scenario Sets pane in the Edit BaseEnsemble Parameters dialog
box. A scenario set is a folder that defines a set of injection nodes or scenarios.
A new scenario set, named Scenario Set, is created and displayed in the Scenario Sets pane.
Because the newly created scenario set is empty, its folder icon is closed and is not preceded by a
box plus or minus sign.
H Edit BaseEnsemble Parameters
[ Import... ] [ Export... ]
'HWBM-IM-limPllil'l™
toon-Zero Demand Junctions (59)
JAII Junctions (92)
^H
Add ]
Edit |
Delete |
1 Add |
r~iit i
Delete |
^^|
[ Create » ]
^^^^^^^^^^^|
Injection: Cont / Nodes: All Nodes (97)
1 Remove
| OK | | Cancel |
^H
| Add » )
Q Scenario Set
0 Simulations •
Pern | [tew] |
Select the currently loaded ensemble's node injections by clicking on the Node Injections pane
of the Edit BaseEnsemble Parameters dialog box. The selected node injection becomes
highlighted.
Select the currently loaded ensemble's scenario sets by clicking on the Scenario Sets pane of the
Edit BaseEnsemble Parameters dialog box. The selected scenario set becomes highlighted.
The Add » button of the dialog box becomes enabled.
19
-------
Click the Add » button. The scenario formed from the selected node injection is added to the
selected scenario set.
B Edit BaseEnsemble Parameters
( Import:... ] [ Export:... ]
•Cont (17400.0 from 0 hours to 1 hour)
Node Set Definitions —
Non-Zero Demand Junctions (59)
All Junctions (92)
Add ]
Edit |
Delete |
Edit
Delete
fflfflSfifflWfflMiBfflWIHHiBMS^^^^H
[ Remove ]
| OK | | Cancel )
^^J
•
B-ioBBIiBBII
» Injection; Cont/ Nodes: All Nodes (97)
97 Simulations
Remove |[New J
Click OK to close the Edit BaseEnsemble Parameters window.
20
-------
3.5. Base Ensemble
Execution
To schedule the simulation
of the ensemble's attack
scenario, click the
EPANET Simulations
check box.
Click the Execute button
under Status. The Execute
button will only be enabled
when one of the Available
Executions are checked.
Q G EPANET Simulations [ Edit j
Q G Health Impacts Analysis | Edit |
O D TSO to Impacts Analysis | Edit |
Q G Infrastructure Impacts Analysis [ Edit |
p|
Q G Sensor Placement | Edit |
1
[ Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute Terminate
Q G EPANET Simulations | Edit |
1
Q G Health Impacts Analysis | Edit |
Q G TSO to Impacts Analysis | Edit |
Q G Infrastructure Impacts Analysis [ Edit 1
r-.l i_
Q O Sensor Placement | Edit |
Estimated Time Remaining
Estimated Total Time
hSCImaCed Lompletlon 1 ime
Execute , T jrrninate
At the completion of the
execution, the EPANET
Simulations will be
highlighted with a green
checkmark.
Available Executions
Scenarios—
Modules
D EPANET 5J™iatiorii
G Health Impacts Analysis
TSO to Impacts Analysis
Sensor Placement
Estimated Time Remaining 00:00:00:00
Estimated Total Time 00:00:00:14
Estimated Completion Time 07/22/2010 02:12:17 PM
Execute Terminate
Note that if the progress bar and estimated completion time
information are not viewable during execution, close the
Execution Control Panel (above box) and Reopen.
21
-------
4. Health Impacts Analysis
This section describes how to use the execution controls in ensemble analysis mode. Figure 3
shows the hierarchy of the Health Impacts Analysis module.
Parameter values shown as integers in the tables should be entered as integers; those shown with
decimal places should be entered as decimal fixed-point (e.g., 1.307) or decimal floating-point
(e.g., 2.87el3) values.
Hfrdtlh lMpO£l 1
Anatytfi
t •!• _ j_ _ • _ J^_^
\3^f ] "2gjaKn 1 [ Lev? WiDsfcnoaLsl DOK 1 Lrac "j i Dncass \
Cdk;jlatK!n y:''Trr'-.-'; "HT5nc4rt BBVXDTEC R^noivc 1 1 Pro^iOElan 1
PcraTCTCT J J \. J ^i?1nDd J Cdcukrlcri J L FcrnnctDrs J
Total Mossl
Engffstion
Timing *"
(Demand
BasEd ~
ATU5 :m
(b Fixed
Times
Itlpevian
V:*rne Uct!C(> *
(Demand
SlKF.li ~
i Random 1
(Fixed
Volume j
Mean
Yo umr 1
•
Population DOSE f# or Worst- J Average "1 R™*
Mndr ' Thr«hnlri5r 1 case (o Sodv Mass r Resums* •
^^^^^^^^^ rnuw ^£^U dlV^
-•""- Me od
Dflmnnd fttsponrfie 1 N^rm.i IZR Pr-^hit
B^sed ™ Thresholds p ~
Fixed Siqmoid
Volume
Per Capita J LDSDyiDSD
llsxgr.
&r.ta
•
on
Latency J_
Tirnr
Fafaiiby
Time ~
FiLaliLy ]_
Kate
Figure 3. Health Impact Analysis Flow Chart
22
-------
In the Execution panel,
click the Edit button to
the right of the Health
Impacts Analysis
checkbox.
Note: Health Impacts
Analysis will be grayed
out until information is
added.
Available Executions
Scenarios —
Modules
0 Health Impacts Analy
TbO to Impacts Analysis
Sensor Placement
Sensor Placement
Status
Estimated Time Remaining 00:00:00:00
Estimated Total Time 00:00:00:14
Estimated Completion Time 07/22/2010 02:12:17 PM
Execute
Terminate
The Edit Health Impacts
Analysis Parameters
dialog box opens.
Notice the Primary
Species and Set Species
boxes. These appear
when EPANET-MSX is
initiated by loading an
.MSXfile
Primary Species ASb
Dose Calculation Method
Ingestion Timing Model
Ingestion Volume Mode!
Population Model
Estimated Population
Per Capita Usage (GPD)
Ingestion of Tap Wate
ATUS
|V-
V
Random v
Demand-based V
•:=!.: 175000
Calc |210. 19518
-Dose Thresholds
Dose Thresholds
Response Thresholds
0 Calculate Dose-Response
Dose Response Options
Contaminant Defaults
Edit
Edit
Miscellaneous
Number of worst-case fatality scenarios to save |o
Number of worst-case dosage scenarios to save JO
Q Use one server per node
0 Normalize
Average Body Mass (kg) |?D.O
Dose-Response calculation method | Probit vj
LD5Q / ID50 (mg/kg) 41.0
Beta
-Disease Progressio
Latency Time (hrs)
Fatality Time (hrs)
Fatality Rate
|4.34
24
24
0.5
23
-------
4.1. Contaminant Name
The cursor is initially
positioned in the
Contaminant Name text
box. Type in the name of
the contaminated that is
going to be simulated.
The other option is to use
a contaminant database,
which is described below.
If a database of
contaminants has been
setup (XML file format),
the user can click the
Select button next to
Contaminant Defaults
under Dose Response
Options to choose a
particular contaminant.
The Select Contaminant
dialog box opens,
displaying an example
contaminant.
El Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Method
Ingestion Timing Model
Ingestion Volume Model
.
Population Model
Estimated Population
Per Capita Usage (GPD)
rv -TL IJ
Dose Thresholds
Response Thresholds
Number of Worst-case to
Q Use one server per r
Ingestion of Tap Water v
1
S|
Demand-based v |
Calc 0
Calc |o,0
[ Edit |
I Edit 1
Save 0
ode
Contaminant Defaults] Select,., j
-Dose Response Method
n Normalize
Dose-Response calculation method | Probit v |
LD50 / ID50 (organisms) 0,0
Beta 0.0
Latency Time 0
Fatality Time 0
Fatality Rate 0.0
NOTE: The format for the XML file is as follows:
Select a particular
contaminant (e.g.,
Hypothetical TOX # 1)
by clicking on its name.
The name becomes
highlighted, and the OK
button becomes enabled.
Click the OK button.
The dialog box closes and
the name of the selected
contaminant is entered
into the Contaminant
Name textbox.
Contaminants
Hypothetical TOX #1
Cancel
The "contaminants.xml" must be located in the appropriate directory,
e.g., C:\Program Files\TEVA-SPOT\Client directory in order to be
recognized by TEVA-SPOT.
24
-------
Under Dose Response
Options, values
associated with the
contaminant can be
entered.
For demonstration
purposes, the example
contaminant, EXAMPLE
TOX#1, would have the
following values entered
under Dose Response
Method:
Dose-Response
calculation method:
Probit
LD50/ID50: 0.001
Beta: 4.34
Under Disease
Progression Parameters,
the following values
would be entered:
Latency Time: 24
Fatality Time: 24
Fatality Rate: 0.5
If the contaminant is
considered a chemical or
toxin with units of mg,
then Normalize is
selected.
If the contaminant is
considered a biological
with units of organisms or
spores, then Normalize is
not selected.
Average Body Weight
(kg) is shown when the
Normalize box is
checked. Type 70.0 for
Average Body Mass
B Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Method Ingestion of Tap Water v
Ingestion Tinning Model
Ingestion Volume Model v
Population Parameters
Population Model Demand-based
Estimated Population
Calc
_
Per Capita Usage (GPD) Calc [o.O
Dose Thresholds -
Dose Thresholds
Response Thresholds
Miscellaneous
Edit
Number of Worst-case to Save 0
Q| Use one server per node
Select,,.
Dose Response Options
Contaminant Defaults
- Dose Response Method
0 Normalize
Average Body Mass (kg) 70,0
Dose-Response calculation method Probit iV,|
LD50 / ID50 (rng/l)
Beta
0.0010
4.34
-Disease Progression Parameters
Latency Time
Fatality Time
Fatality Rate
24
24
0.5
@ Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters -
Dose Calculation Method
Ingestion of Tap Water v
Ingestion Timing Model
Ingestion Volume Model
Population Parameters
Population Model | Demand-based
Estimated Population
Calc m
Per Capita Usage (GPD) Calc [o
Dose Thresholds
Dose Thresholds
Response Thresholds
Miscellaneous
Number of Worst-case to Save
Q Use one server per node
Contaminant Defaults
Select,,.
0 Normalise
Average Body Mass (kg) 0.0
Dose-Response calculation method Pro
Dit
LD50 / ID50 (mg/l) 0,0
Beta 0.0
Disease Progression Pt
Latency Time 0
Fatality Time 0
Fatality Rate 0,0
Normalize should only be used for contaminants, such as metals
or toxins, which have an ID-50/LD-50 dose expressed in units per
kilogram body weight.
25
-------
4.2. Dose Calculation
Parameters
Only Ingestion of Tap
Water is available in the
current version of TEVA-
SPOT for Dose
Calculation Method.
From the Ingestion
Timing Model pull-down
menu, three options are
available: Demand-
based, ATUS, and 5
Fixed Times
For demonstration
purposes, ATUS is
selected.
From the Ingestion
Volume Model pull-
down menu, two options
are available: Random
and Fixed Volume. If
Fixed Volume is selected
then, Mean Volume
(liters/person/day) is
shown.
For demonstration
purposes, Random is
selected.
Normally, it is
recommended that users
select 5 Fixed Times and
Fixed Volume, with a
value of 1 liter per person
per day. The random-
based models (timing and
volume) will result in
varying results due to
randomness.
El Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Methon Ingestion of Tap Water v
Ingestion Timing Model
Ingestion Volume Model
"
Demand-based
Population Parameters -
Population Model
Estimated Population Calc |o
Per Capita Usage (GPD) Calc [Q.Q
Dose Thresholds
Dose Thresholds
Response Thresholds
Miscellaneous
Number of Worst-case to Save
Q Use one server per node
Select,.,
Dose Response Options
Contaminant Defaults
Dose Response Method
Q] Normalize
Dose-Response calculation method Probit v
LD50 / ID50 (organisms)
Beta
0,0
0,0
Disease Progression Parameters
Latency Time
Fatality Time 0
Fatality Rate 0,0
Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Metho
Ingestion Timing Model
Ingestion Volume Mode
Population Parameters
Population Model
Estimated Population
Per Capita Usage (GPD
d Ingestion of Tap Water v
v
Demand-based
ATUS
5 Fixed Times
Demand-based
Calc |0
) Calc |o.O
TL L IJ
Dose Thresholds
Response Thresholds
1 [ Edit
II Edit
ii
Number of Worst-case to Save 0
fj Use one server per node
Dose Response Options
Contaminant Defaults
Select,,.
0 Normalize
Average Body Mas;
Dose-Response cal
LD50 / ID50 (rng/l)
Beta
(kg)
:ulation method
70.0
Probit v]
0.0010
4.34
Latency Time 24
Fatality Time 24
Fatality Rate 0.5
ATUS = American Time Use Survey. Data collected by the Census
to characterize when people eat and most likely drink water.
26
-------
4.3. Population
Parameters
Under Population
Model, two pull-down
options are available:
Demand-based and
Census-based.
If Demand-based is
selected, then Estimated
Population and Per
Capita Usage (GPD) are
shown If Census-based
is selected, then a
Population file must be
loaded.
For demonstration
purposes, select Demand-
based from the pull-down
menu.
Type in a value of 200.16
for Per Capita Usage
(GPD) gallons per person
per day.
The Estimated
Population box gets
highlighted. Click Calc to
calculate the total
population being served.
Using the Demand-based
population model allows
the user to input the
population served by the
distribution system being
evaluated or by inputting
an estimated per capita
usage factor in units of
GPD to determine the
population for the
distribution system.
H Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Method
Ingestion Timing Model
Ingestion Volume Model
Ingestion of Tap Water v
ATUS
Random
Population Parameters
Population Model Demand-based
Estimated Population | Calc | 0
Per Capita Usage (GPD)
Dose Thresholds—
Dose Thresholds
Response Thresholds
Edit
Edit
Miscellaneous
Number of Worst-case to Save
fj Use one server per node
Select,..
Dose Response Options
Contaminant Defaults
Dose Response Method
0 Normalize
Average Body Mass (kg) |70,0
Dose-Response calculation method Probit v
LD50 / ID50 (mg/l) 0,0010
Beta
4,34
Disease Progression Parameters
Latency Time 24
Fatality Time 24
Fatality Rate 0,5
Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Method
Ingestion Timing Model
Ingestion Volume Model
Ingestion of Tap Water
ATUS
Random
Demand-based
.Population Parameters-
^opulation Model
Estimated Population | [Calc || 0
Per Capita Usage (GPD) 200,16|
pDose Thresholds—
Dose Thresholds
Response Thresholds
Miscellaneous
Edit
Edit
Number of Worst-case to Save 0
fj Use one server per node
Select,,.
Dose Response Options
Contaminant Defaults [
Dose Response Method
0 Normalize
Average Body Mass (kg) |70,0
Dose-Response calculation method Probit v I
LD50 / ID50 (mg/l) 0,0010
Beta
|4.34
Disease Progression Parameters
Latency Time 24
Fatality Time 24
Fatality Rate 0,5
27
-------
4.4. Dose Thresholds
These parameters (Dose
Thresholds and
Response Thresholds)
are optional.
Click the Edit button to
the right of the Dose
Thresholds text box.
The Edit Dose
Thresholds dialog box
opens.
Dose thresholds can be used to evaluate contamination impacts
without specifying a particular contaminant. In this case, the
threshold could be an action or advisory level.
Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Method
Ingestion Tinning Model
Ingestion Volume Model Random
Ingestion of Tap Water v
ATUS
Demand-based
Population Parameters
Population Model
Estimated Population Calc |7876l|
Per Capita Usage (GPD) . aic [200.16
Dose Thresholds —
Dose Thresholds
v|
Edit
Response Thresholds
Edit
Miscellaneous -
Number of Worst-case to Save 0
fj Use one server per node
Select.,.
Dose Response Options
Contaminant Defaults I
Dose Response Method
0 Normalize
Average Body Mass (kg) |70.0
Dose-Response calculation method Probit v
LD50 / ID50 (mg/l) [o.OOlO
Beta
4.34
Disease Progression Parameters
Latency Time
Fatality Time
Fatality Rate
24
24
0.5
Click the Dose
Thresholds text box.
Type in the values for
each of the Dose
Thresholds of interest.
[The user can enter
multiple values.]
For demonstration
purposes, type 0.001 and
press Enter. Again type
0.002 and press Enter.
Click the OK button to
accept the Dose
Thresholds values
Dose Thresholds
0.0010
0.0020
Dose Threshold
Cancel
Note: Values can be entered as decimal or in scientific notation,
e.g.,1.0E-3.
Dose thresholds are described in Davis, M. J., andJanke, R.
(201 Oa). "Patterns in potential impacts associated with
contamination events in water distribution systems." J. Water
Resour. Planning Manage., (16 March 2010),
10.1061/(ASCE) WR. 1943-5452.0000084.
28
-------
The Edit Dose
Thresholds dialog box
closes.
The defined Dose
Thresholds are now
displayed in the window.
El Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Method [ingestion of Tap Water v
Ingestion Timing Model 1ATUS v
Ingestion Volume Model | Random
Population Parameters
Population Model
Estimated Population
Per Capita Usage (GPD) [Calc
Dose Thresholds
Dose Thresholds
Response Thresholds
Miscellaneous -
Number of Worst-case to Save 0
fj Use one server per node
Dose Response Options
Contaminant Defaults
Select,,.
@ Normalize
Average Body Mass (kg) [70.0
Dose-Response calculation method | Probit v |
LD50 / ID50 (mg/l)
Beta
-Disease Progr
Latency Time
Fatality Time
Fatality Rate
sssio
0.0010
4,34
24
24
0.5
Click the Response
Threshold text box.
Type in the values for
each of the Response
Thresholds of interest.
[The user can enter
multiple values.]
For demonstration
purposes, type 0.25 and
then press Enter. Again
type 0.50 and then press
Enter
Click the OK button to
accept the Response
Thresholds values. The
Edit Response
Thresholds dialog box
closes.
Edit Response Thresholds
Response Thresholds
0.25
0.5
Response Threshold
Cancel
29
-------
4.5.Miscellaneous
Inserting an integer
number into the Number
of Worst-case to Save
box results in the
corresponding number of
scenarios being saved.
Each scenario saved
contains contaminant
concentration data for the
numbered scenario. A
"1" will result in the
highest ranked (in terms
of health impacts)
scenario being saved;
whereas "5" will result in
the top 5 highest ranked
scenarios being saved.
For demonstration
purposes, type in a value
of 1 in the box for
Number of Worst-case
to Save.
Use one server per node
is a feature that will
reduce the computational
burden of the HIA (and
EPANET simulations if
run with HIA)
simulations on the host
computer. Checking this
box will result in only one
HIA scenario being run at
a time on the host
computer regardless of
how many processors are
available.
The Number of Worst-case to Save refers to the number of
scenarios for which complete contaminant concentration versus time
data is saved for further analysis. This data is saved in the TEVA-
SPOT-Database folder under the particular Ensemble Name and
within the Health Impacts Analysis subdirectory (TEVA-SPOT-
Database/Collection/Ensemble/Health Impacts Analysis folder).
0 Edit Health Impacts Analysis Parameters
Contaminant Name
Dose Calculation Parameters
Dose Calculation Method
Ingestion Timing Model
Ingestion Volume Model
Ingestion of Tap Water v
ATLIS
Random
(-Population Parameters—
Population Model Demand-based
Estimated Population Calc 73761
Per Capita Usage (GPD) [ Calc ]
Dose Thresholds
Dose Thresholds
0.0010,0.0020
Response Thresholds 0.25,0,5
Dose Response Options
Contaminant Defaults I
-Dose Response Method—
0 Normalize
Average Body Mass (kg)
Dose-Response calculation method
LD50 / ID50 (rng/l) 10,0010
Beta
Select,..
4,34
-Disease Progression Parameters -
Miscellaneous
Number of Worst-case to 5ave 0
fj Use one server per node
Latency Time 24
Fatality Time 24
Fatality Rate 0,5
Cancel
If "1" is used, then the worst case scenario in terms of dose
response based impacts is chosen, "2" would result in the top 2
scenarios being saved.
NOTE: This value should not exceed 10 for any network
greater than 3,000 nodes due to the large storage space
required.
30
-------
4.6. Health Impacts Analysis
Execution
To determine the health impacts
from the EPANET simulations that
ran previously, click the Health
Impacts Analysis check box
Click the Execute button in Status.
Execution
.. , .
& O EPANET Simulations Edit
, ,
Q 0 Health Impacts Analysis [ Edit ]
Q U TSO to Impacts Analysis Edit
^ rM t
O Ei^il:
a
1 1 Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute
Terminate
At the completion of the execution,
the Health Impacts Analysis will be
highlighted with a green check.
Available Executions
Scenarios
Health Impacts Analysis II Edit
j TSO to Impacts Analysi:
Status
Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute
Terminate
31
-------
5. Sensor Placement Algorithm
The Sensor Placement algorithm solves an optimization problem. The algorithm tries to find a
solution to the problem of placing a given number of sensors in the water network to optimize a
stated objective, usually subject to several constraints. Figure 4 below highlights the inputs for
the sensor placement module within TEVA-SPOT. For more detailed description of these
features, please refer to the TEVA-SPOT Toolkit Users Manual.
Figure 4. Sensor Placement Flow Chart
In the Execution panel, click the
Edit button to the right of the
Sensor Placement check box.
Available Executions
Scenarios -
EPANET Simulation.?
- Modules
Health Impacts Analysis
ISO to Impacts Analysis
Sensor Placement
Sensor Placement
Status
Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute I Terminate
32
-------
The Edit Sensor Design
Parameters dialog box opens.
All the parameters in the Sensor
Design Parameters are entered by
clicking the Edit buttons.
i- Edit Sensor Design Parameters
om lirorl
I None defined - Required
Sensor Design Parameters- -
Solvers [None defined - Required
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes)
Constraints
Location Cateogories
Costs
Detection Limits
None aeTinea - no pp goai: none
Irtani lirtarl
None defined - using 0
None defined - Optional
None oeTinea - using
11nraHnnc
|N/A
None defined - using 0
Selection Methods
|N/A
0 Create All Impacts
0 Sensor Designs will be generated
Edit
Edit
Edit
Cancel
5.1.
Specifying the Solvers
Click the Edit button to the right
of the Solvers text box.
The Choose Solvers dialog box
opens.
Click the Add Solver button.
Multiple solvers (GRASP
(Heuristic) and Lagrangian) can
be selected and added. Sensor
designs specific to each solver will
be created. However, run times
will increase by a factor of two
because sensor designs will be
created for each algorithm
selected.
Lagrangian algorithm is generally
used for large or very large
networks which will not run using
the GRASP algorithm due to
computer memory limitations.
Choose Solvers
Solvers
PICO is a true optimal solver, but only available in the toolkit version
33
-------
For demonstration purposes, select
GRASP solver by clicking on the
enabled name in the list of solvers.
Click the OK button to accept the
Solvers parameter, e.g., Heuristic.
GRASP is the default solver. It
will generally provide the better
solution.
The Choose Solvers dialog box
closes.
The figure shows that zero sensor
designs will be generated: this is
because a sufficiently complete set
of Sensor Placement parameters
has not yet been entered.
Note: OK button is disabled.
& Edit Sensor Design Parameters
5.2. Specifying the Sensor Set
Sizes
Click the Edit button on the right
of the Sensor Set Sizes text box.
The Edit Sensor Set Sizes dialog
box opens.
Heuristic (grasp)
u - ai least one iion-zero
roni lirorl
Sensor Design Parameters
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes) JNone defined - using 0
Constraints
Location Cateogories
Costs
Detection Limits
Edit
isrone aennea - using AI
N/A
None defined - using 0
Selection Methods
IN/A
None defined - Required
I None aorrea - no pa goai: none
| None defined - Optional
Edit
Edit
[s/1 Create All Impacts
0 Sensor Designs will be generated
OK
Cancel
34
-------
Click the Sensor Set Size text box.
For demonstration purposes, type
1 and press Enter. Repeat this
process for 5, 10, 15, 20, and 25,
remembering to press Enter after
each number.
When done, click OK. The Edit
Sensor Set Sizes dialog box
closes.
NOTE: Values must be
integers.
Again because a sufficiently
complete set of Sensor Placement
parameters has not yet been
entered, the figure shows that zero
sensor designs will be generated.
Note: OK button is still disabled.
o
i
5
10
15
20
25
Sensor Set Size
Cancel
Edit Sensor Design Parameters
Heuristic (grasp)
0, 1, 5., 10, 15, 20, 25
Sensor Design Parameters
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes) None defined - using 0
| None defined - Required
TOnenernea - no pa goai: nonen
roni lii-url I
] Create All Impacts
0 Sensor Designs will be generated
Cancel
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
None defined - Optional
None aennea - using AH
1 nrj^J"inn«
N/A
None defined - using 0
N/A
( Edit |
[ Edit |
Edit
| Edit |
Edit
35
-------
5. 3. Specifying the Objectives
LxliCK tne JLdit Dutton to tne rignt
of the Objectives text box.
The Edit Objectives dialog box
opens.
Click the Goal pull-down list box.
. . .
ror demonstration purposes, select
Population Exposed as the
desired goal by clicking it. The list
closes and the selected goal is
displayed in the list box.
Select the Mean statistic as the
desired statistic by clicking it. The
list closes and the selected statistic
is displayed in the list box. The
\\^ot*sit PJISIP ^t/iti^tir rpnmrp^i
longer run times.
,. , , ...I 11
Click the Add button and then
click OK.
Note: The other goals and statistics
used to create additional objectives
are described in tne liiVA-brUl
Toolkit Users Manual which is
located at C:\Program
Files\ TEVA-SPOT\ CttenMocs
provided the default option was
chosen for installing TEVA-
SPOT.
I
I
i- Edit Objectives X
Goal Statistic
# of Failed Detections "j | Mean
[ Add ]
[ OK ] [ Cancel ]
1- Edit Objectives X
Mean Population Exposed (pe_mean)
Goal Statistic
j Popu lation Exposed v 1 1 Mean
[ Add |
[ OK ] [ Cancel j
1 Edit Objectives [X
Mean Population Exposed (pe_mean)
j. _j.j ^i
Goal Statistic
Population Exposed ,« Mean
1 Add |
I OK ]| Cancel ]
I
|
J
n
36
-------
The Edit Objectives dialog box
closes and the objectives are now
listed in the Objectives text box.
The updated figure shows that
seven (7) sensor designs will now
be generated. These seven designs
correspond to the seven different
sensor set sizes that were selected.
Note: OK button is now enabled.
This indicates that there is
sufficient information provided to
calculate the sensor locations.
However, the user can specify
additional data as described in the
following sections.
Although the box "Create All
Impacts" is checked - for faster
run times this box should not be
checked Checking the box,
"Create All Impacts" results in
the generation of additional tso-2-
impact files (located in the TEVA-
SPOT-Database) for any objective
not already selected, which can be
used for Regret Analyses.
'i- Edit Sensor Design Parameters
sensor Design rarameiers
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes)
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
Heuristic (grasp) [ Edit
0, 1, 5, 10, 15, 20, 25
pe_mean
None aerinea - no pa goai: none
rani rircirl
None defined - using 0
None defined - Optional
rjune aennua - using All
1 Qcations
N/A
None defined - using D
N/A
[ Edit |
I Edit I
[ Edit ]
[ Edit ]
[ Edit |
[ Edit ]
Edit
[ Edit
Edit
| 0 Create All Impacts
7 Sensor Designs will be generated |
Cancel
Using Regret Analysis, the user can evaluate sensor
network design(s) developed for a particular objective
against other objectives if "Create All Impacts" is
selected.
5.4. Dose Threshold
Click the Edit button to the right
of the Dose Thresholds text box.
The Edit Dose Thresholds dialog
box opens. If no Dose
Thresholds are entered, no sensor
designs for dose thresholds will be
created. Population Dosed (PD)
must be selected to produce
sensor network designs based on
dose thresholds.
37
-------
Click the Dose Threshold textbox.
Type a value of 0.001, and then
press the Enter key.
Again type a value of 0.002 and
press the Enter key.
Click OK button to accept the
changes to the Dose Thresholds
parameter.
5.5. Specifying the Response
Times
Click the Edit button to the right
of the Response Times (minutes)
text box.
The Edit Response Times dialog
box opens. If no Response Times
are entered, the default is zero.
i- Edit Dose Thresholds
-Dose Thresholds
0,0010
0,0020
Dose Threshold
Click the Response Time
(minutes) text box
Type the desired value 0 and press
the Enter key.
Repeat this process for 360 and
720, remembering to press the
Enter key after each entry.
Click OK button to accept the
changes to the Response Times
parameter.
38
-------
The figure shows that 21 sensor
designs will be generated (one for
each combination of the seven
sensor set sizes and three response
times).
Note here that sensor network
designs based on dose thresholds
will not be created - because PD
was not included as an objective.
If PD was included there would be
a total of 63 sensor designs!
Edit Sensor Design Parameters
Sensor Design Parameters
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes)
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
Heuristic (grasp)
0, 1, 5, 10, 15, 20, 25
pe_mean
Edit
0.0010, 0.0020
0, 360, 720
None defined - Optional
Hone aerlnea - using f\n
InrjHnnc
N/A
Edit
None defined - using 0
N/A
fc/| Create All Impacts
21 Sensor Designs will be generated
Cancel
5.6. Specifying the Constraint
Sets
Click the Edit button to the right
of the Constraints text box.
The Edit Constraints dialog box
opens.
Click on OK button as this is
Optional.
Constraints can be specified in
the same manner as objectives.
Constraints restrict any sensor
design to a solution
described by the constraint.
39
-------
5.7. Specifying the Location
Categories Sets
Click the Edit button to the right
of the Location Categories text
box.
The Edit Location Categories
dialog box opens. Click the Add
"ALL" locations button
Selecting "New Location
Categories" button brings up a
dialogue box where the user can
browse to a sensor file which
defines feasible, infeasible, fixed,
and unfixed sensor locations. See
Appendix A, Table 3 for
formatting of the sensor table.
Edit Location Categories
Location Categories -
Add "All1
locations
1 New Location Categories
Cancel
Note that after clicking "New
Location Categories" the box
expands and default options are
available for specifying feasible
sensor locations for the algorithm
selected to choose from.
The default options are: All
Locations, Non-Zero Demand
Nodes, All Junctions, Enumerated
(User Selected), Pipe Diameter,
and SP (Sensor Placement) format.
The SP format provides for
uploading a text file specifying
sensor locations.
Select an appropriate option and
Click the OK button to accept the
changes to the Location
Categories.
Edit Location Categories
Location Categories
All Locations
| New Location Categories
Cancel
0 Edit Sensor Design Parameters
u bensor Designs win De gei
Cancel
Non-Zero Demand Nodes
Ail Junctions
Enumerated
Pipe Diameter
SP Format
40
-------
The figure shows that 21 sensor
designs will be generated (one for
each combination of the seven
sensor set sizes and three response
times).
Sensor Design Parameters
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes)
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
Heuristic (grasp)
0, 1, 5, 10, 15, 20, 25
pe_mean
None defined - no pd goal: none required
0, 360, 720
None defined - Optional
None defined - using All Locations
N/A
None defined - using 0
N/A
Edit
1 *» |
[ Edit
| Edit i
[ Edit
[ Edit
Edit
| Edit
[ Edit
Q Create All Impacts
21 Sensor Designs will be generated
Cancel
5.8. Specifying the Costs
This feature is currently not available.
5.9. Specifying the Detection
Limits
Click the Edit button to the right
of the Detection Limits text box.
The Edit Detection Limits dialog
box opens. If no detection limit is
entered, the default is zero.
Type 0.0 in the Detection Limit
textbox and click Enter.
Repeat this process for 0.1 and 1.0,
remembering to press the Enter
key after each entry.
Click the OK button to accept the
changes to the Detection Limits.
The Edit Detection Limits dialog
box closes, and the changes
become effective.
41
-------
5. 1 0. Specifying the Selection Method
This feature is currently not available.
The figure shows that 63
sensor designs will be
generated (one for each
combination of the seven
sensor set sizes and three
response times, and
three detection limits).
5.11. Sensor Placement
Execution
To determine the sensor
placements, click the
Sensor Placement
check box.
Click the Execute
button. Note that run
times can be quite long.
At the completion of the
execution, Sensor
Placement will be
highlighted with a green
check.
• S Edit Sensor Design Parameters |_X|B
Sensor Design Paramet
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minut
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
Heuristic (grasp) | Edit ]
Q, 1, 5, 10, 15, 20, 25 [ Edit |
pejnean Edit
wreoannea-nopagownane [ ^ j
ss) D, 360, 720 [ Edit ]
None defined - Optional [ Edit ]
None defined - using All Locations [ Edit |
N/A Edit
0,0,0,1,1,0 | Edit ]
[N/A E*
[] Create All Impacts
63 Sensor Designs will be generated
| OK | | Cancel ]
• Execution ••§
., -| U.I IT f
[
& G EPANET Simulations
Mnrli ilp'-
^^ Q Health Impacts Analysis
Q ISO to Impacts Analysis
-Sensor Placement
O 0 l?n?.9.LEJ5c.?.m.?.Q?i
t » 1 L. 1- ""
1 ^jj£ 1 Estimated Time Remaining
' ' Estimated Total Time
Estimated Completion Time
.- * Execute Terminate
Frith I
[ Edit )
Edit j
® QBaseEnsemble
[ Edit~|
Q Ki:i.ri n ipar t Analysis- Edit
® D Health Impacts Analysis | Edit |
.
® 0 Sensor Placement | Edit |
^^^^^^^^^^•-
Estimated Time Remaining OD:00:OQtOD
Esti mated Tota ITime 00:00: 00 : 32
Estimated Completion Time 04/30/2006 01:05:28 PM
1 Execute 1
42
-------
6. Maps
Output maps are produced in ensemble analysis mode by performing a health impacts analysis
and by running the sensor placement algorithm.
6.1.
Jr |bfe Be r
My [jam lab
a,:\A M*
Default Map
to* Edt
Be
Default v
Health Impacts Analysis]. Demand-based Population Estimate
Health impacts Analyse]: Dosage by [opciwn Lowflon
Heylih impact Analysis . Dosage by Inpcfon Locaibon - Logarithmic
HMlth InpXW Anjlyis : F*UIHi« br InjKtcn LOCJbXi - tog*rthn«
Health Impacts Analysis : Fdblttles by Peceplur Location • iDDIh pertentile
Health Impacts Anatyse J Fatalities by Receptor Locaicn - IDOtti per cenbte • Logarithmic *
Htorhgr MT..J, |MFMrng
The water network
map of the currently
loaded ensemble is
identified as Default
in the pull-down list
of maps.
43
-------
6.2. Maps Produced by Health Impacts Analysis
In the pull-down list of maps, the names of all the maps produced by performing the health
impacts analysis start with the prefix [Health Impacts Analysis].
6.2.1. Demand-
Based Population
Estimate
Choose [Health
Impact Analysis]:
Population Estimate.
The Demand-based
Population Estimate
map depicts the
population estimated
at each node based
on the demand at the
node.
Demand-based Population Estimate
6.2.2. Estimated
fatalities by Injection
Location
Choose [Health Impact
Analysis]: Estimated
Fatalities by Injection
Location.
The Estimated Fatalities
by Injection Location
map depicts the total
estimated fatalities
attributable to each
possible injection. The
data are depicted at the
single node where the
injection occurs.
'
j
Erewi* Fte KwK E*
l-B) oure racfe
'^X& MSPJ(*8ltJ.^*^iF*l«Jt(^ta^«^ *|HI»m)l Mr*g! j^W-5
Fatalities by Injection Location
I \ >v^$?
^
...
*
- .
£r
Poetei
-M:.
Note that this map indicates "Fatalities by Injection Location," a
software update now titles all such maps with "estimated" before
facilities.
44
-------
6.2.3. Additi onal Map s
TEVA-SPOT also produce the following maps which can be displayed by choosing the
appropriate function:
Estimated Fatalities by Injection Location — Logarithmic: Depicts the same data as the
Estimated Fatalities by Injection Location map, but with the colors distributed on a logarithmic,
rather than a linear, scale.
Dosage by Injection Location: Depicts the total ingested dosage attributable to each possible
injection. The data are depicted at the single node where the injection occurs.
Response by Injection Location: Depicts the total response (number of individuals who become
diseased) attributable to each possible injection. The data are depicted at the single node where
the injection occurs.
Estimated Fatalities by Receptor Location — 10th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 10th percentile of
total estimated fatalities.
Estimated Fatalities by Receptor Location — 25th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 25th percentile of
total estimated fatalities. It also identifies that injection or injection combination and the total
estimated fatalities attributable to it.
Estimated Fatalities by Receptor Location — 50th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 50th percentile of
total estimated fatalities. It also identifies that injection or injection combination and the total
estimated fatalities attributable to it.
Estimated Fatalities by Receptor Location — 75th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 75th percentile of
total estimated fatalities. It also identifies that injection or injection combination and the total
estimated fatalities attributable to it.
Estimated Fatalities by Receptor Location — 90th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 90th percentile of
total estimated fatalities.
Estimated fatalities by Receptor Location — 100th Percentile: Depicts the estimated fatalities at
each node attributable to the injection or injection combination ranked at the 100th percentile of
total estimated fatalities (i.e., to the most lethal scenario). It also identifies that injection or
injection combination and the total estimated fatalities attributable to it.
45
-------
All of the maps mentioned above have associated logarithmic maps which depict the same data
but with the colors distributed on a logarithmic, rather than a linear, scale.
6.3. Maps Produced by the Sensor Placement Algorithm
Ft* N*xfe Edit
SOheu-Btic-MS010-Cepe_me*i-RT720-LCAtl Locatens-CLO.l-SMstrf]; Detected Ewnts (links cnty)
SOheLristic-llS31D-oepe_riBafr'P.T720-LCAllLocaOQrtsOBpe_me*vRT720-i.CAfl i.ocatO's-CU-SMsttl): Detected gventj 0r*$ only)
SOheuristtc-NS010-Cepe_m3an-R;T720-LCAll Lrcatters-Ctl-SMstdl; Detected Evenfe (nodes only)
jTKan-RT720-LGM! Locattors-CLl-SMstety Sensor Locations
: sensor Counts
The map naming convention follows a specific pattern. The first part of the name refers to the
solver chosen for the optimization. The rest of the name refers to the number of sensors, the
objective of the optimization, the response time, the location categories, the detection limit, and
finally the selection method. Selection method has no specific meaning now, but is a place holder
for future development.
For example, SOheuristic-NS025-OBpe_mean-RT720-LCAllLocations-DLO-SMstd refers to the
sensor design that was created using the following sensor design parameters: SO refers to solver =
heuristic, NS or number of sensors = 25, OB or objective = mean estimated population exposed,
RT or response time = 720 minutes, LC or location category = all locations, DL or detection limit
= 0 and SM or selection method = standard.
All of the sensor designs created will have the following maps associated with them: Sensor
Locations, Detected Events (nodes only), Detected Events (links and nodes), and Detected Events
(links only). Each map is described in more detail in the following sections.
46
-------
6.3.1. Sensor Locations
The locations of the sensors are marked by stars. Three categories (Existing, Selected, and
Ignored) of sensor locations are assigned different colors, according to the legend in the map.
Right clicking on the legend of a particular sensor removes or adds that type from or to the map.
Examine the map of sensor locations for 25 sensors at a response time of 720 minutes.
From the Map pull-down list, select the map entitled [SOheuristic-NS025-OBpe_mean-RT720-
LCAllLocations-DLO-SMstd]: Sensor Locations. The map below will be displayed with the 25
sensors being identified with green stars.
Fie Mode bit
'•^ V Sf Mip
Sensor Locations: SOheuristic-NS025-OBpe_rnean-RT720-LCAII Loca lions- DLO-SMstd (Heuristic (grasp))
TiriR;: 720inlnule^
Objectives: MRnn Population Fxposed
Constraints: None
f»™»»
#««^
jr
47
-------
6.3.2. Detected Events (Nodes Only)
Now examine the map of the detected
events for nodes only associated with 5
sensors and a response time of 360
minutes.
From the Map pull-down menu, select
the map entitled [SOheuristic-NSOOS-
OBpe_mean-RT360-LCAllLocations-
DLO-SMstd]: Detected Events (nodes
only).
Although it is difficult to see at the
whole-network scale, this map also
depicts the 5 sensors placed by the
Sensor Placement algorithm, each
colored with a different color
identified in the legend.
To see the sensors, zoom in to a part of
the map.
To zoom, right click on the map to get
the Map Pop Up Window as shown.
From Zoom Scaled, select Zoom In 2X
Pg Down.
Copy Location
Copy Bounding Box
Copy Map (Ctrl-C)
Undo Zoom (Ctrl-Z)
Redo Zoom (Ctrl-Y)
Zoom Out 2X (PgUp)
Zoom Scaled
Zoom Named
Zoom Entire DB (Home)
Redraw Map (5pace)
Abort Map Redraw
Print... (Ctrl-P)
Save Map As Image
Save Window As Image
48
-------
This zoomed map is then displayed.
Note: To remove the zoom, right click
on the map and select Undo Zoom.
49
-------
6.3.3. Detected Events (Links and
Nodes)
Now examine the map of the detected
events for links and nodes associated
with 5 sensors and a response time of
360 minutes.
From the Map pull-down menu, select
the map entitled [SOheuristic-NSOOS-
OBpe_mean-RT360-LCAllLocations-
DLO-SMstd]: Detected Events (links
and nodes).
6.3.4. Detected Events (Links Only)
Now examine the map of the detected
events for links only associated with 5
sensors and a response time of 360
minutes.
From the Map pull-down menu, select
the map entitled [SOheuristic-NSOOS-
OBpe_mean-RT360-LCAllLocations-
DLO-SMstd]: Detected Events (links
only).
6.3.5. Sensor Counts
Select [SensorPlacementSummary]:
Sensor Counts.
By coloring its nodes appropriately,
the Sensor Counts map shows the
number of sensor designs that placed a
sensor at the node. For all sensor
network designs created, this map
shows which locations were selected
most often.
50
-------
6.4. Charts
Charts are produced in ensemble analysis mode by performing a health impacts analysis and are
not produced by running the sensor placement algorithm. No chart has been designated as the
default chart.
By clicking on the Charts tab, a Chart pull-down list will appear with the names of all the charts
produced by performing a health impacts analysis starting with the prefix [Health Impacts
Analysis]. These charts present health impacts either as a time series (i.e., a function of time), as
a cumulative distribution function, or by scenario (i.e., injection location or injection
combination, in rank order). For this tutorial only one chart will be discussed.
Ffe Mode Eft
ft&te »!ati»t WtutoiPbt
[Health impacts iral-'sis]: H#Kt Fstaty Note by kerarlo
Irrpjcte Anafysis]: hfecHore, Oseja, and Fatate bj Time • IOC* Psrcertlle Case
mfKtw, Owss, and FauntB by T«« • m Wcwtife Cut
[Haiti ItrpsS Ma^s]: Infections, D«sg, at Fatjltes by Taw • SJi P?fcentife Caa
[HH*-I ifrests iralysis]: Infectms, OseiH, and Fatate by T.me • Mlh Percenfjle CSB
51
-------
6.4.1. Average Estimated
Fatalities by Time
Select [Health Impacts
Analysis]: Average Estimated
Fatalities by Time.
This chart depicts estimated
fatalities averaged over all the
possible injections or injection
combinations defined by the
contaminant release scenario,
as a function of time in hours.
6.4.2. Additional Charts
TEVA-SPOT also produces the following charts which can be displayed by choosing the
appropriate function:
Total Estimated Fatalities by Scenario: Depicts the estimated fatalities resulting from each
injection or injection combination, sorted in order of increasing lethality. The curve in this chart
is obtained by sorting the rows of the Ensemble Summary table in increasing order of the data in
the Number of Estimated Fatalities column, then plotting the data in that column.
Maximum Concentration by Scenario: Depicts the maximum concentration attributable to
each injection or injection combination, sorted in order of increasing concentration. The curve in
this chart is obtained by sorting the rows of the Ensemble Summary table in increasing order of
the data in the Max Concentration column, then plotting the data in that column.
52
-------
Maximum Individual Dose by Scenario: Depicts the maximum individual dose attributable to
each injection or injection combination, sorted in order of increasing dose. The curve in this
chart is obtained by sorting the rows of the Ensemble Summary table in increasing order of the
data in the Max Individual Dose column, then plotting the data in that column.
Total Infected Population by Scenario: Depicts the number of infected individuals attributable
to each injection or injection combination, sorted in order of increasing infections. The curve in
this chart is obtained by sorting the rows of the Ensemble Summary table in increasing order of
the data in the Number of Infected column, then plotting the data in that column.
Highest Fatality Nodes by Scenario: Depicts the minimum number of nodes whose populations
have to be combined to yield 90% of the estimated fatalities attributable to the injection or
injection combination, sorted by increasing effect. The curve in this chart is obtained by sorting
the rows of the Ensemble Summary table in increasing order of the data in the Nodes for 90%
Estimated Fatalities column, then plotting the data in that column.
Estimated Fatalities Cumulative Distribution Plot: Depicts the cumulative probability
distribution of the estimated fatalities attributable to each injection or injection combination,
sorted by increasing lethality. The curve in this chart is obtained by sorting the rows of the
Ensemble Summary table in increasing order of the data in the Number of Estimated Fatalities
column, then plotting essentially the row number (or, more precisely, the fractional position of
the row within the table) as a function of the row's datum in that column.
Estimated Infections, Disease, and Fatalities by Time — 10th Percentile Case: Depicts, for
the injection or injection combination yielding estimated fatalities ranked at the 10th percentile
level, the growth and decay over time of the estimated number of infected (i.e., individuals
exposed but not yet diseased), the estimated number of diseased (i.e., individuals showing
symptoms but not yet deceased), and the estimated number of fatalities. It also identifies the
injection or injection combination yielding the 10th percentile case and the number of estimated
fatalities in that case.
The same chart can be displayed for the 25th, 50th, 75th, 90th, and 100th Percentile Cases.
53
-------
6.5. Tables
By clicking on the Tables tab, a Tables pull-down list will appear with the names of all the tables
produced by executing the health impacts analysis and sensor placement.
No table has been designated as the default table. Note that all tables can be output to Microsoft
Excel by choosing the Table pull down menu when on the "Table" tab is selected. Alternatively,
data from tables can be copied (select row, hit control-A, then Control C) and pasted (control V)
into a spreadsheet or other document.
rMj. fifc-. -J».
riff MOOfi
Map '-.nsfs
Tstfe Default
[Health Jnqpacts. .arH^K]: Ensemble Percentites Table*
[Mwlth Jmpgcfe Analyse]: Fjftalflies fiWepSSr Hjfe
[K*ealth Irnpacts Analyse]: Inaction Impacls Table
[Health Impact* Analysis]:
[SB«orPliCenr*rrt£uiiriniJr>]: All
[SensorflacemerrtSumnwrir]; Sensor Flacetnent Summafy
54
-------
6.5.1. Health Impacts Analysis - Injection Impact Table
The Injection Impact Table shows, for each possible injection or injection combination (i.e.,
scenario):
• the maximum concentration attributable to the scenario,
• the maximum individual dose attributable to the scenario,
• the number of infected individuals attributable to the scenario,
• the number of estimated fatalities attributable to the scenario,
• the minimum number of nodes whose populations have to be combined to yield 90% of the
estimated fatalities attributable to the scenario, and
• the number of nodes with estimated fatalities attributable to the scenario.
* TEVA (Demo in Demo [Main])
Ensemble File Mode Edit
Map 'Charts] Tables
Table H!Smiil8IBI8aBBEBHSB18ll8HBIill!iBiBSB!I^^^^^^^^^^^^^^^^^^^^^^^^M
Injectio...
UNO1
UNO1000
UNC-100
UNO 1001
JUNG- 1002
UNO 1003
UNC-1005
UNC-1004
UNO 1006
UNO 1007
UNC-1008
UNC-1009
UNC-101
UNO 1010
UNO 1011
UNC-1012
UNC-1013
UNC-1014
ViNO 1016
UNO 1015
UNC-1017
UNOigia
UNC-102
JUNO 1019
UNO 1020
UNC-1021
3UNC-1023
UNO 1024
UNO 1025
UNC-1026
UNC-1027
UNO 1028
UNO 1029
UNO 103
UNO1030
UNC-1031
UNO1032
UNO 1033
UNO 1034
UNC-1035
UNC-1036
UNO1037
UNO 1038
UNO 1039
JUNC-1040
UNC-1041
UN01042
UNC-1044
UNC-1043
UNC-1046
UNC-1045
UNC-1047
Max Concentration
0
274.243
951.332
706.959
159.102
335.12
478.941
658.404
481.581
310.575
3.144
17.775
261.196
391.281
175.014
1,401.841
408.611
4.729
513.76
4,575.001
17.615
563.662
374.852
738.753
0
17.78
1,202.154
0
0
0
586.72
0
658.808
496.67
0
632.692
190.021
159.482
790.328
46.098
2,595.901
267.399
43.45
18.256
0
20.264
70.717
20.03
101.42
28.882
49,733
296.389
Max Individual Dose
0
46.304
83.667
105.423
27.477
18.508
52.999
9.19
61.072
7.542
0.386
2.306
14.49
42.609
24.196
28.611
37.599
0.526
10.631
810.468
0.829
42.307
17.518
109.497
0
0.882
8.6
0
0
0
106.122
0
71.427
19.345
0
7.115
23,893
31.241
107.277
2.818
25.164
11.497
5.612
1.401
0
0.455
0.455
0.777
5.699
1.07
7,105
17.474
Number of Infected
0
7,425.141
434.991
498.738
14,086.864
12,816.36
3,577.147
9,338.417
3,472.211
10,694.095
21,054.131
18,622.463
172
13,947.548
11,685.573
86
12,425.067
20,837,473
9,924.442
19
19,897,23
3,268.71
417
1,655.354
0
19,458.252
77
0
0
0
2,444.479
0
2,253.641
86
0
6,709.184
2,783.34
3,184.594
57
8,597.085
1,963.842
2,673,193
8,596.776
18,855.621
0
19,313.893
18,801.719
11,279.837
9,088.63
18,295.936
1,439,399
2,196.097
Number of Fatalities
0
7,425.141
434.991
498,738
14,086.864
12,816.36
3,577.147
9,338,417
3,472.211
10,694.095
21,054.131
18,622.463
172
13,947.548
11,635.573
86
12,425.067
20,837.473
9,924.442
19
19,897.23
3,268.71
417
1,655.354
0
19,458.252
77
0
0
0
2,444,479
0
2,253.641
86
0
6,709.184
2,783.34
3,184.594
57
8,597.085
1,963,842
2,673.193
8,596.776
18,855.621
0
19,313.893
18,801.719
11,279.837
9,088.63
18,295.936
1,439.399
2,196.097
^^^^^^^^^^^^M
Summary Table
Nodes for 90% Fatalities
0
64
7
11
97
93
52
75
50
84
139
137
5
79
89
2
94
141
79
1
143
49
6
24
0
140
2
0
0
0
43
0
28
3
0
64
33
29
2
41
25
31
41
145
0
127
123
61
44
142
24
16
Nodes with Fatalities
0
287
8
19
320
307
295
297
291
297
411
414
7
386
326
2
421
423
308
1
420
287
9
259
0
410
2
0
0
0
280
0
302
4
0
281
339
160
2
337
259
335
335
478
0
512
507
355
338
498
44
43
55
-------
6.5.2. Health Impacts Analysis - Additional Tables
• Estimated fatalities Receptor Nodes Table - The Estimated fatalities Receptor Nodes table
shows, for each node, the number of estimated fatalities at that node attributable to six specific
injections or injection combinations (i.e., scenarios): the ones whose consequent total estimated
fatalities are ranked at the 10th, 25th, 50th, 75th, 90th, and 100th percentiles (the last of these six
is the most lethal scenario).
• Ensemble Percentiles Table - The Ensemble Percentiles table shows the maximum
concentration, maximum individual dose, number of infected individuals, and number of estimated
fatalities attributable to the injection or injection combination at the 10th, 25th, 50th, 75th, 90th,
and 100th percentile levels, as well as the average of those quantities over all the injections or
injection combinations.
• Population - The Population Table shows the population at each node. If the
Population Model HIA parameter is set to From Scenario, the map shows the population data
imported into the currently loaded ensemble; if it is set to Demand-based, the map shows the
population deduced from the demand for water at each node of the water network.
• Tables Produced by the Sensor Placement Algorithm - In the output products pull-down list
on the control bar, the names of all the tables produced by performing a sensor design start with
the prefix [SensorPlacementSummary]. These tables present either summary data resulting from
the sensor design or its trade-off analyses.
6.5.3. Sensor Placement Summary
From the Table pull down menu, select Sensor Placement Summary table.
Fte M:* cat
* r:w*i-*T=«HCAl! I watms-OtQ-riMsW Harki*; fraxi)
-- HT^O-LC* ! ..I 4 GU V. - •
Sensor Placement Summary
out CbJeCTM* Himunaaa MB En
6.5.4. Trade-off Analysis Data
This table shows the name of the sensor design, the solver that produced the design, the number of
sensors, the objective, the abbreviated form of the objective, the response time in minutes, the
constraint, the location categories, the cost, the detection limit, the selection method, and the
impact value corresponding to the objective named in the 4th and 5th columns.
Trade-off Analysis Data
56
-------
6.5.5. Impacts and Detection Time Tables
An "Impacts and Detection Time" table appears for each sensor network design. The description
for each sensor network design appears in the table pull down menu followed by the name of the
table, i.e., Impacts and Detection Time.
Each row provides by Injection Location (node name) the sensor that detected the event (if
detected), detection time when detected (length of the simulation time if undetected, time in
seconds), and the impact associated with the event after detection. Units of the impacts are
defined by the objective used for the sensor network design.
• H TEVA-SPOT (Update_Users_Manua[_2.1 .4 in Release_Jan-09_Tests [Main]) •
Ensemble File
I Map i Charts j
Table [Big!
Table Mode Edit About
Tables
Iffi^B^TiniHSIiffiWT^HErn^ESHSilffiroCTHBIBSHntBITO^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^B v Filtering [ Manage ] [
Impacts and Detection Time: SOnew_grasp-NS010-OBpe_mean-RTOOO-SMstd
10 Sensors
Response Time: 0 minutes
Objectives: Mean Estimated Population Exposed
Constraints: None
Injection Detected by Detection Time Impacts
JUNC-2570
JUNC-159
JUNC-2037
JUNC-2420
JUNC-2810
JUNC-2703
JUNC-564
JUNC-1806
JUNC-124
JUNC-124
JUNC-1248
JUNC-1248
JUNC-124
JUNC-1248
JUNC-1248
JUNC-2059
JUNC-136 JUNC-124
JUNC-2148 JUNC-2059
JUNC-1062
JUNC-1S97
JUNC-1S54
JUNC-1132
UNC-1170
UNC-697
JLIMC-631
UNC-624
JUNC-1963
JLINC-2300
JUNC-2542
JUNC-673
JUNC-1222
JUNC-1355
UNC-1207
JUNC-1248
JUNC-1248
JUNC-1248
JUNC-1270
JUNC-1270
JUHC-613
JUHC-613
JUHC-613
JUNC-124
JUNC-124
JUNC-124
JUNC-657
JUNC-1270
JUNC-1270
JUNC-1270
120
360
300
240
180
360
360
420
540
300
660
660
720
360
180
60
360
360
1560
1620
2400
420
240
600
300
0
0
0
0
0
0
0
2,246
2,652
0
82
195
1,976
0
0
0
0
0
85
1,223
2,803
843
0
1,005
0
57
-------
6.5.6. Sensor Ranking
A "Sensor Ranking" table appears for each sensor network design. The description for each
sensor network design appears in the table pull down menu followed by the name of the table, in
this case Sensor Ranking. Each row provides the prioritized sensor location, starting with the
most effective sensor location to reduce impacts and ending with the least effective for the given
sensor network design.
The column headings consist of sensor (node ID), cumulative impact, incremental impact
reduction (percentage), and cumulative impact reduction (percentage). The units of impact are
specified by the objective selected for the sensor network design. The first row provides the base,
no sensors impact. Impact is mean impact.
, TEVA-SPOT (Update_Users_Manual_2.1.4 in Release_Jan-09_Tests [Main])
Ensemble File Table Mode Edit About
Hap Charts Tables
Sensor Ranking: SOnew_grasp-NS010-OBpe_mean-RTOOO-SMstd
10 Sensors
Response Time: 0 minutes
Objectives: Mean Estimated Population Exposed
Constraints: None
Sensor
No Sensors
JUNC-362
JUNC-1248
JUNC-657
JUNC-1270
MIC-124
JUNC-2499
JUNC-480
JUNC-468
JUNC-2059
JUNC-613
Cumulative Impact
65,651,273
5,231.542
1,763.728
1,011.885
702.339
395.647
311.852
235.256
186.587
139.397
116,838
Incremental Reduction
0,0%
92.0%
66.3%
42.6%
30.6%
43.7%
21.2%
24.6%
20,7%
25.3%
16,2%
Cumulative Reduction
0,0%
92.0%
97.3%
98.5%
98.9%
99.4%
99.5%
99.6%
99.7%
99,8%
99,8%
58
-------
7. Filtering Criteria
Output products (i.e., maps, charts, and tables) are displayed in TEVA-SPOT's main window.
Without filtering, the universe of available choices presented to the user in the output products
pull-down list box on the control bar is often very large. By reducing the number of choices, the
filtering controls make it easier to focus on those that meet particular criteria.
7.1.
Filtering Controls
The Filtering Controls are located
in the right half of the control bar
and include the Manage button
and filter specification pull down
list box.
Filterirb Manage No Filtering
7.2. Manage Button
Click the Manage button.
The Display Options Filter
Management dialog box opens.
Click the ADD button in the
Display Options Filter
Management dialog box.
7.3. Creating a Filter
Specification
The Edit Filter Settings dialog
box opens.
The cursor is initially positioned
in the Name text box, which is
initially empty.
^ Display Options Filter Man.
Filters
Edit Delete
Analysis: Health Impacts Analysis
LJ Data Type
Logarithmic
Map Type
3H iPercentile
At anytime, the user can click the plus button + to the right of the Analysis pull-down list box to
fully expand the filtering tree, or adjacent minus button - to fully collapse it.
59
-------
Click the plus button + to the
right of the Analysis pull-down
list box to fully expand the
filtering tree.
Then move and resize the dialog
box until the whole filtering tree
is visible.
Because none of the attribute
check boxes are checked initially,
the value checkboxes are all
initially disabled.
Analysis: Health Impacts Analysis
D Safe Iffii
^| Dosage
| | Fatalities
| | Population
||Response
Q-| I Logarithmic
(~|No
Yes
Q- O Map Type
_| Injection Location
Receptor Location
D Percentile
100th
[JlOth
At all times, the "Results: items found" list box displays the output products selected by the
current tree of filtering criteria, and the number of those products is indicated.
Initially the tree selects no output products and the current count is 0.
When attribute or value nodes are checked or unchecked in the filtering tree, the contents of the
Results list box are recomputed and redisplayed, and the count is updated.
60
-------
Click the Name textbox and type
the name DcnioF liter.
Click the Analysis pull-down list
box in the Filter Criteria pane and
select the Sensor Placement
analysis type. A new filtering tree
is displayed.
1
'
I Edit Filter Settings >
Name DernoFilter |
Q:|J- , ^. ;4- rj
~Hitef criteria ~
Analysis: Sensor Placement V|B0
@Q] Detection Limit
| j-j Q
& O Links Included
No
Yes
S •• [HI Location Categories
| All Locations
6- G Map Type
Sensor Counts
! | Sensor Detected Events
' ] Sensor Locations
0--Q Number of Sensors
pi 10
|=15
Q25
Q-- d Objective
Mean Population Exposed (pe_mean)
&•• Q] Response Time
PJ360 minutes
I ]720 minutes
0 D Selection Method
^ .Normal
^••••n Solver
; | ) Heuristic (grasp)
Results: 0 items found
OK Cancel
]
H
61
-------
Within the filtering tree, click
check boxes
Map Type [Sensor Locator] and
Response Time [360 minutes]
The OK button becomes enabled
whenever the count becomes non-
zero.
Click the OK button to accept the
newly created filter specification
The Edit Filter Settings dialog
box closes and the filter
specifications are created.
The Display Options Filter
Management dialog box is
redisplayed to include the
complete name of the new filter
specification in the Filters list.
i- Edit Filter Settings
Name DernoFilter
-Filter Criteria
Analysis: Sensor Placement
V
6- P Detection Limit
• j lO.O
•P Links Included
1 res
Q Location Categories
0 Map Type
| Q Sensor Counts
; Q Sensor Detected Events
• M Sensor Locations
8- 1 _ | Number ot Sensors
--nao
- P Objective
Mean Population Exposed (pe_mean)
y @ Response Time
0 leFmjnutei
~l 720 minutes
|_| Selection Method
' | | Normal
•P Solver
'—[HI Heuristic (grasp)
Results: 7 items found
[SOheuristic-NSOOu-OBpe_mean-RT360-LCAII Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS001-OBpe_mean-RT360-LCAll Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS005-OBpe_mean-RT360-LCAII Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS010-OBpe_mean-RT360-LCAII Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS015-OBpe_mean-RT360-LCAII Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS020-OBpe_rnean-RT360-LCAll Locations-DLO-SMstd]: Sensor Locations
[SOheuristic-NS025-OBpe_mean-RT360-LCAII Locations-DLO-SMstd]: Sensor Locations
Display Options Filter Man.
Filters
Add
Edit Delete
62
-------
Use the Edit button to change a Filter Specification and use the Delete button to delete a Filter
Specification.
Current Filter Selections Pull Down List Box
Under the Map Pull Down menu, only seven maps are available for display after the filtering.
63
-------
8. Regret Analysis Mode
Regret Analysis provides a means to select the best sensor design among many. Multiple
ensembles are created, e.g., different contaminants, release times, mass release rates, response
delays, detection limits, and with each of these ensembles sensor designs are created. The
impact files and sensor designs are loaded into the Regret Analysis and Evaluate Sensor (see
TEVA-SPOT Toolkit) is used to compute the matrix of impacts. Every sensor design chosen is
evaluated against every impact. The performance measures (Preferred Overall Design and
Preferred Design to Minimize Maximum Deviation) are used to compute the difference between
the best design (which is the design created for the particular impact) and all the other designs.
There are no parameters to enter like there is in the Ensemble Analysis Mode.
The Preferred Overall Design and the Preferred Design to Minimize Maximum Deviation
provide two indices of a sensor network design's overall performance. The Preferred Overall
Design is determined by minimizing the square root of the sum of the squared deviations for a
sensor design from the best case sensor design across all scenarios. The Preferred Design to
Minimize Maximum Deviation is determined by minimizing the maximum deviation from the
best case for a particular sensor design, across all scenarios.
It is important to understand the Regret Analysis prior to execution. Regret Analysis serves as a
means to select a preferred sensor network design among many developed. Regret Analysis can
be used to select a preferred sensor network design among different contaminants, release times
or mass release rates (setup through different ensembles) and different response times and
different detection limits. A single set of sensors should be selected, e.g., 10 sensors, in
performing a Regret Analysis. Sensor designs developed using Population Dosed (PD) with
Dose Thresholds can now be used in the Regret Analysis, they could not be used with some
earlier versions TEVA-SPOT.
Using Regret Analysis to determine a preferred sensor network design among different
objectives should be done with caution.
64
-------
8.1. Input Data for Regret Analysis
Setup and run Ensemble Analyses for two different mass injection rates using the data below.
The ensembles will be distinguished by the two mass release rates below, please name the
ensembles "AprilSO" and "low," to represent the higher mass loading and lower mass loading
ensembles, respectively. Save each ensemble to the Collection named Tutorial. Run all three
modules in each ensemble, i.e., EPANET Simulations, Health Impact Analysis, and Sensor
Placement.
EPANET Simulations
Two different ensembles:
Mass Injection Rate:
1.74E+04 mg/min for 1 hr
4.34E+03 mg/min for 1 hr
Injection Definition
Attributes
Name
Mass Injection Rate
Start Time
Stop Time
Description
Contaminant
l,74e+04
Ei
hours
hours
V
Tutorial Ensemb
Health Impact Analysis
All parameters remain the same.
i Edit Health Impacts Analysis Parameters
Contaminant Name EXAMPLE AGENT #1
Dose Calculation Method Total Mass
T t'T''Mrlln Hh H V
ingestion i iming i loaei uemana oasea
Ingestion Volume Model Demand-based v
Mean Volume | 2,
Population Model Demand-based
Per Capita Usage] 0.139
Dose Thresholds —
Dose Thresholds 0.0010,0.0020 |[ Edit
Response Thresholds 0.25 [ Edit
Miscellaneous- —
Number of Worst-case to Save 1
Contaminant Defaults Select...
Average Body Mass 70
1^1 Normalizel
Dose-Response caiculation method Probit
LD50/ID50 0.02
Beta 4.88
[-Disease Progression Parameters —
Latency Time 1
Fatality Time 8
Fatality Rate 1.0
65
-------
Sensor Placement
All parameters remain the same.
8.2. Mode Menu
The Mode menu is used to toggle
back and forth between the two
processing modes: Regret Analysis
Mode and Ensemble Analysis Mode.
Clicking Mode in the menu bar
displays the Mode menu.
8. 3. Regret Menu
Clicking Regret in the menu bar
displays the Regret menu. Three
menus (Regret, Ensemble, and Mode)
are anchored on the menu bar in
regret analysis mode.
8.4. New
Click New. . . in the Regret menu.
The New Regret Analysis Name
dialogue box opens.
Click the Regret Analysis Name text
box and type the name Demo
1 i Edit Sensor Design Parameters |X|I
Solvers
Sensor Set Sizes
Objectives
Dose Thresholds
Response Times (minutes)
Constraints
Location Cateogories
Costs
Detection Limits
Selection Methods
Heuristic (grasp)
0, 1, 5, 10, 15, 20, 25
pe_mean
iduny denned - no pci 19031: none
f earn lirnrl
360, 720
None defined - Optional
All Locations
N/A
[ Edit ]
( Edit 1
[ Edit 1
[ Edit ]
[ Edit ]
1 Edit ]
1 Edit 1
Edit
0.0 |[ Edit ]
N/A Edit
0 Create All Impacts
14 Sensor Designs will be generated
L
OK | Cancel
teiigiliP'''"'
-------
The OK button becomes enabled.
0 New Regret Analysis Name
Regret Ana lys is Co I lection Tutor ia I [Ma in 1
Regret Analysis Name |Demo|
OK
Cancel
8.5. Ensemble Menu
Clicking Ensemble in the menu bar
displays the Ensemble menu. Click
Load Ensembles... in the Ensemble
menu. Opens Ensembles dialogue.
TEVA Reeret Analvsls (Demoln Tutorial [Main])
(Mods
Load Ensembles,.
Click the list box and select a
collection name: Tutorial.
Ensemble AprilSO is for high mass
loading contaminant.
Ensemble low is for low contaminant.
Select both Ensembles from load list
box to load the then click their names
while holding the Ctrl key down.
The OK button becomes enabled.
Click Okay.
Select One or More Ensembles to Load
Ensemble Collection Information —
Size 420.0 Mb
Last Saved Date 04/30/200,.,
Last Ensemble Execution 04/30/200..,
Last HIA Execution 04/30/200...
Last TSO2Impact Execution Never
Last Sensor P lace rnent Ex,,. 04/30/200.,,
8.6. Load
Click Load... in the Regret menu,
The Regret Analyses dialogue box
opens.
HTEVA Regret Analysis (Demo in Tutorial [Main])
Ctrl+N
Ctrl+O
Orl+S
Save As...
Ctoa
CoHecUon Hnajemer*...
Reyet Analysis Managenwit-..
Execution Ctrtrd CTLF
67
-------
Select Demo and Click the OK
button.
Regret Analyses
Regret Analysis Collection
Tutorial [Main]
Tutorial
8.7.
Save
The Save command is only enabled when a regret analysis group is in memory. Click Save in the
Regret menu., or press Ctrl+S. There is no response from the system after a successful save.
~&&Close
Clicking Close in the Regret menu will close the currently loaded regret analysis group, i.e., it
will no longer be loaded. Its name, and that of the containing collection, are removed from the
name stripe and replaced by the legend No Regret Analysis Loaded. If the Execution panel is
being displayed, it is removed from the display. Do Not Close Now.
8.9.
Execution Control
Click Execution Control in the
Regret menu. The Execution panel is
added. To close, Click the Close
button 3 in the upper right-hand
corner of the panel to remove it. Do
Not Close Now.
In the Execution panel, click the Edit
button to the right of the Regret
Analyses check box.
The Regret Analysis Management
dialogue box opens. (See next
screen).
Available Executions
•Regret Analysis
I I Regret Analyse >
Status
IB
Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute i Terminate
68
-------
8.10. Specifying the Parameters
for Regret Analyses
Click the Add button. It will open the
Regret Analysis Parameters box
which is displayed in the pane below.
When the Regret Analysis Parameters
box (below) opens it may be
necessary to resize the box using the
mouse cursor in order to view the
parameters, e.g. OK box and Name.
0 Regret Analysis ... |x
Regret Analyses
0 total sensor evaluations will be executed,
Close
Display of Regret Analysis Parameters
Q Regret Analysis Parameters X
Name
_ „
Filter Settings: Select All |
a 0 Ensemble
\ 0 (ApriBO Tutorial [Main])
: 0 (low Tutorial [Main])
- 0 Number of Sensors
i 08
05
- 010
1 015
i 020
' 025
- 00
' 0360
; 0720
S IV! Detection Limit
: 00.0
( 00.1
1-01.0
Available Selected
ApriBO: SOheuristic-NSOOO-OBpe_mean-RTOOO-LCAII Locations-DLO-SMstd A
ApriBO : soheuristic-NSOOO-OBpe_mean-RTOOO-LCAII Locations-DLO. 1-SMsU
ApriBO : SOheuristic-NSOOO-OBpe_mean-RTOOO-LCAII Locations-DLl-SMsW
ApriBO : SOheuristic-NSOOO-OBpe_mean-RT360-LCAII Locations-DLO-SMstd
ApriBO : SOheuristic-NSOuu-OBpe_mean-RT360-LCAII Locations-DLO. 1-SMsta1
ApriBO: SOheuristic-NSOOO-OBpe_rnean-RT360-LCAII Locations-DLl-SMstd
UpriBO : SOheuristic-NSOOO-OBpe_mean-RT720-LCAII Locations-DLO-SMstd
ApriBO : soheuristic-NSOOO-OBpe_mean-RT720-LCAII Locations-DLO. 1-SMstd
UpriBO : SOheuristic-NSOOO-OBpe_mean-RT720-LCAII Locations-DLl-SMstd
;ApriBO : SOheuristic-NS001-OBpe_mean-RTOOO-LCAII Locations-DLO-SMsta1
ApriBO : SOheuristic-NS001-OBpe_mean-RTOOO-LCAII Locations-DLO. 1-SMsU
ApriBO : SOheuristic-NSOul-OBpe_rnean-RTOOO-LCAII Locations-DLl-SMstd
ApriBO : SOheuristic-NS001-OBpe_rnean-RT360-LCAII Locations-DLO-SMstd
UpriBO : SOheuristic-NS001-OBpe_mean-RT360-LCAII Locations-DLO. 1-SMstd
ApriBO : SOheuristic-NS001-OBpe_mean-RT360-LCAII Locations-DLl-SMstel
ApriBO : SOheuristic-NS001-OBpe_mean-RT720-LCAII Locations-DLO-SMstd
ApriBO : SOheuristic-NS001-OBpe_rnean-RT720-LCAII Locations-DLO. 1-SMsto1
ApriBO : SOheuristic-NS001-OBpe_mean-RT720-LCAII Locations-DLl-SMstd
ApriBO: SOheuristic-NS005-OBpe_mean-RTOOO-LCAII Locations-DLO-SMstd w
«
Filter Settings: (Select All]
i- 0 (ApriBO Tutorial [Main])
0 (low Tutorial [Main])
E 00
i- 0 * of Failed Detections
; 0 Estimated Population Exposed
: 0 Estimated Population Killed
1 0 Extent of Contamination
: 0 Mass Consumed
M Population Over Dose Threshold
' 0 Time To Detection
; 0 Volume Consumed
Available Sebcted
ApriBO: tso2Impact Analysis_RT_0000_DL_0.0 [pd, pe, me, vc, to, ec, nfd, pk] A
ApriBO: tso2Impact Analysis_RT_0000_DL_0.1 [pd, pe, me, vc, W, ec, nfd, pk] ~~
ApriBO: teo2Impact Analysis_RT_0000_DL_1.0 [pd, pe, me, vc, td, ec, nfd, pk]
ApriBO: teo21mpact Analysis_RT_0360_DL_0.0 [pd, pe, me, vc, to, ec, nfd, pk]
ApriBO: teo2!mpact Analysis_RT_0360_DL_0.1 [pd, pe, me, vc, to, ec, nfd, pk]
ApriBO: tso2!mpact Analysis_RT_0360_DL_1.0 [pd, pe, me, vc, td, ec, nfd, pk]
ApriBO: teo2Impact Analysis_RT_0720_DL_0.0 [pd, pe, me, vc, to, ec, nfd, pk]
ApriBO: tso2Impact Analysis_RT_0720_DL_0.1 [pd, pe, me, vc, td, ec, nftl, pk]
ApriBO: teo2Impact Analysis_RT_0720_DL_1.0 [pd, pe, me, vc, to1, ec, nfd, pk]
low: teo2Impact Analysis_RT_0000_DL_0.0 [pd, pe, me, vc, td, ec, nfd, pk]
low: teo2Impact Analysis_RT_0000_DL_0.1 [pd, pe, me, vc, td, ec, nfd, pk]
low: tso2Impact Analysis_RTJ]OOOJ3L_1.0 [pd, pe, me, vc, td, ec, nfd, pk]
low: tso2Impact Analysis_RT_0360_DL_0.0 [pd, pe, me, vc, td, ec, nfd, pk] „
Q mean Q median Q var Q tee Q worst
0 sensor evaluations will be computed.
69
-------
Initially all boxes are checked. Uncheck all sensor set sizes except 10 in the Sensor Design Filter
Settings
April*):GOharsac WSQ1Q Qepe_mean a TOE -Ca!l Lccs-xr* CUJ
- 000tKt**i limit
Mn.o
go.i
iff to
tow- OC*wl5»c NCQ10 CCpe rrean RTOCO LCAll LocatBns CU CMMd
tow SCtwSW NGC1D Cepe mea^ RT36O LCAfl Locations 1X0 GMitrf
tow-GC*WS«: NGG10 CCpe mear>RT360LC4ll LOCattons CtO.l GMstt)
*CW-2C*wu-i3t>cWMlOCepe fnean RTD6O U:AIP wcaltcm Cti CMstf
fcw: SC*Wurtetlc NCQ1Q «tpe_frean RT720 LCAll Locattons CtO C^tt
KWIl Lacatotu-U-iU-bMstd
rmtuU Set
[fi-t.t Ail
* of TaflEd Detecflwra
ropula«cfi Esposed
PGpufeOeo Killed
! extent of ContamfiaOoo
£• PcpulaDco Uver Loss Threshold
'"volume
AprUDC: Hojirnpact Anaiysss PT O72) [X 1.0 IpA pe. me.
: -iiiLi'1-::." -;••.:,:.-.::. RT 0000 DL 0.0 [pd. pe. me, •«:,
: tso2!mp3d: Analysis RT 0000 DL 0.1 [pd. pe. me. vc,
: Go2J»ip3ct Analysts RT GOOD OL 1.0 [pd. pe. me. «c.
tew: tsoafppact Analyst RT ODffl DL 0.0 [pd. pe, me. uc.
In the Impact Set Filter Setting uncheck all except Estimated Population Exposed
• Regret Analysts Parameters
•HWSMtrge: | aHect all j A«.»Ltil
lfW.1) .^"'
: NSOU) -Depe.mwn-RTOOO LCAll L«aOB*lXO.:
(»«. (Ion Tt
AprKB: Giltai-stt NC010 C6p»,rr«n PT35O LC*H Locanera-DLl ^
n P7000 LCAtiLeeatMfw-CiOGMsai
00
M3GO
tow: SCfteircK t
n P-T720 LCAll LocatxwCU), l-
Pi Estimated P
BCsnmaWl PopuUtvn • itod
Ml), 'ill' •
Over Dow Threshold
**rOO: ttoanpact Am(yBUlTjOOflO_CLJXO [pd, pt, me, *c, fi K. nH, pfc] «
AprCO' Boampia Ar»Hwa_RTJJOOOJXJ31 [p4 pt, »nc, vc,«, scr rtw, pi:J
AurO3: bO3n«Ct Ani*RWj*T,OaiDJCl.l-0 tpd. pO. me. vc. W. «.. nfd. pkj
[pd, po. me, wr. t^ tx. nfO, pV]
[pd. pe «. vc. tf « nH. pf]
: tsoZlnpiCt ArwIyMv,^ t .OKOjd ,OO [pd, pV. rnc, «. tl K. r.H. pV ]
JtprCO: «a31n*«tflAa)^*_RTjJ7»j*jM [p4 p*, iw, *. H «. nftt ptj
•prhPO: soampact AnaJywi_RT J3730 JX_I.O [pd; pe, me, vc, tf, ee. nfd. pi:]
ow:tso2!trp«;tAoalr»t»^T.OOI».Dt 0.0 [pd. (w, rr^ w, U,«,r*J,pk]
t uvitr&tojt* JOOODJX_D.I [pd, pp. UK. «, id, «r, rftt [*]
cm : tawtmpact Analrtts.RT JMOO.OU J.P [pa, pe, trc. »C, td, *, r«, pV ]
HOHmpactAAa^»J?T_CO60jx_0.0[pO. pe. me. vc. (0. «.«J. pyj
IS03tnpJCtA(Ulrfft_ftT,03e0..n. 0,1 Ipd, p«, me, *c, U. «., »*d. pk)
.i.o [pa, p«*, ™*r, «, M, «, «*n (*i
Impaca
0 sensor erabatxnt * 'I be compued.
70
-------
8.11. Add sensor designs
Multi-select one or more sensor designs in the Available list box of the Sensor Designs pane by
clicking them (with the Shift and/or Ctrl keys held down, if necessary). The selected sensor
designs become highlighted, and the right transfer button » I becomes enabled.
Click the right transfer button. The highlighted sensor designs in the Available list box of the
Sensor Designs pane are transferred to the Selected list box, and the button becomes disabled.
[Remove sensor designs.] Multi-select one or more sensor designs in the Selected list box of the
Sensor Designs pane by clicking them (with the Shift and/or Ctrl keys held down, if necessary).
The selected sensor designs become highlighted, and the left transfer button « I becomes
enabled.
Click the left transfer button. The highlighted sensor designs in the Selected list box of the
Sensor Designs pane are transferred to the Available list box, and the button becomes disabled.
H Regret Analysis Paramet
u
r> (low lufiTul [MJ«;|)
Sa
0720
r [yfOvWctionumi
B: QOhBiriSlt-NS01(M^5ajT**i'RT360 LC
Jp-iWi 90Teu-iHie'N»10'Oat*.,rtwn PTBO LC
o 10 cepe_roe»i PTOOO LCAHLocatcra 1X0 SM«fl
B LafcaH*n^XO'5M*td
km-. ^**iir,li-«^li;M*pi- rrti.ri-*miIH! Alllii,W.pk]
vc, tf, sc, n«, pk]
U, ec nfd, f* I
U.«, rrtd, c» 1
«.«, nftt pfcj
B,«. flffl, pH]
U. «., rrh], f*.J
M,«:, nM, [*]
"s:
Qwian QHT Qw
71
-------
8.12. Selecting Impact Sets
[Add impact sets.] Multi-select one or more impact sets in the Available list box of the Impact
Sets pane by clicking them (with the Shift and/or Ctrl keys held down, if necessary). The
selected impact sets become highlighted, and the right transfer button * I becomes enabled.
Click the right transfer button. The highlighted impact sets in the Available list box of the
Impact Sets pane are transferred to the Selected list box, and the button becomes disabled.
[Remove impact sets.] Multi-select one or more impact sets in the Selected list box of the
Impact Sets pane by clicking them (with the Shift and/or Ctrl keys held down, if necessary). The
selected impact sets become highlighted, and the left transfer button * I becomes enabled.
0 Regret Analysis Parameters
SEnsor D&sigi-i;
-Impact Sets
0 sensor evaluation? wii! be computed.
Filter Settings : | Select All
B 0 Ensemble
0 (AprilSO Tutorial [Main])
0 (low Tutorial [Main])
1 '-' @ Number of Sensors
••DO
DI
i D =
j - ' 10
: ; DIS
D20
D25
3! [i/j Response Time
1 00
0360
0 720
B 0 Detection Limit
; 0o.o
i 00.1
0 1.0
Available Selected
»
4<
ApriGO ; SOheLiristic-NS010-OBpe_maan-RTOOO-LCAII Locations-DLO-SMstei
ApriGO: SOrieuristic-NS010-OBpB_mean-RTOOO-LCAIILocations-DL0.1-SMsttl
ApriGO: SOhauristic-NS010-OBpa_maan-RTOOO-LCAII Locations-DLl-SMstd
AprilSQ : SOheuristic-NS01D-OBpe_mean-RT360-LCAII Locations-DLO-SMstd
4pril30:SOheuristic-NS01Q-OBpE_mean-RT360-LCAIILocations-DL0.1-SMstel
ApriGO: SOheuristic-NS010-OBpB_rnaan-RT360-LCAII Locations-DLl-SMstd
ApriGO: SOrieuristic-NS010-OBpa_maan-RT720-LCAII Locations-DLO-SMstd
ApriGO: SOheuristic-NSOlO-OBpe mean-RT720-LCAII Locations-DLO. 1-SMstri
AprilSO : SOheuristic-NS01D-OBpe_mean-RT720-LCAII Locations-DLl-SMstd
low: SOheuristic-NS010-OBpe_rnean-RTOOO-LCAII Locations-DLO-SMstd
low:SOheuristic-NS010-OBpe_ni8an-RTOOO-LCAIILocations-DL0.1-SMsttl
low: SOheuristic-NS010-OBps_ni8an-RTOOO-LCAII Locations-DLl-SMstd
low: SOheuristt-NS010-OBpe_maan-RT360-LCAII Locations-DLO-SMstd
low: SOheuristic-NS01D-OBpe_mean-RT360-LCAIILocations-DL0.1-SMstd
low: SOheuristic-NS010-OBpe_rnean-RT360-LCAII Locations-DLl-SMstd
low: SOheuristic-NS010-OBpB_maan-RT720-LCAII Locations-DLO-SMstd
low:SOIieuristic-NS010-OBp8_niBan-RT720-LCAIILocations-DL0.1-SMstd
low: SOheuristic-NS01D-OBpB_maan-RT720-LCAII Locations-DLl-SMstd
Filter Settings; LSBtBCtAl1]
S hn-^nible
| r D (AprlfSO Tutorial [Main])
| L ] (low Tutorial [Main])
&• P^pon-if? Time
i [•••• j o
1 | 360
£s3- 5 Impact
I- fj # of Failed Detections
Estimated Population Exposed
ent of Contamination
Mass Cori'i.u triPij
Population .Over Dose Threshold
Available
Q nfd O pe O rnc | |pk Q vc | |ec
'
>n
Selected
AprilSO; tso2Impact Analysis_RT_0000_DL_0,0 [pd, pe, me, vc, td, ec, nfd, pk] A
AprilSO: tso2Impact Analysis RT 0000 DL 0.1 [pd, pe, mr, vc, td, ec, nfd, pk] •
AprilSO: tso2Impact Anafysfe_RT_OODOJX_l.a [pd, pe, me, vc, td, ec, nfd, pk]
AprilSO: tso2Impact Analysis _RT_0360_DL_0,Q [pd, pe, me, vc, td, ec, nfd, pk]
AprilSO: tso2Impact Anaiysis_RT_0360_DL_0,l [pd, pe, me, vc, td, ec, nfd, pk]
AprilSO; tso2Impact Ana!ysis_RT_0360_DL_l,0 [pd, pe, me, vc, td, ec, nfd, pk]
AprilSO: tso2Impact Analysis RT 0720 DL O.Q [pd, pe, mr, vc, td, ec, nfd, pk]
AprilSO: tso2Impact Ana!ysis_RT_0720_DL_0.1 [pd, pe, me, vc, td, ec, nfd, pk]
AprilSO; tso2Impact Analysis_RT_0720_DL_l,0 [pd, pe, me, vc, td, ec, nfd, pk]
ow; tso2ImpaetAnalysis_RT_OQOO_DL_0.0 [pd, pe, me, vc, td, ec, nfd, pk]
ow; tso2Impact Analysts _RT_OQDO_DL_i.a [pd, pe, me, vc, td, ec, nfd, pk]
aw: tsD2ImpactAnalysis_RT_Q360_DL_0.0 [pd, pe, me, vc, td, ec, nfd, pk]
ow; tso2Impact Analysis _RT_0360_DL_0,1 [pd, pe, me, vc, td, ec, nfd, pk]
ow; tso2ImpactAnalysis_RT_Q360_DL_l,Q [pd, pe, rnc, vc, td, ec, nfd, pk] v
,-, • •
ad Qjmeari Q median Qvar Qtce | | worst
72
-------
Select pe (population exposed) under Impacts and Mean for Statistics
B Regret Analysis Parar
- p;Mjn*»f ft-j
DM
HS
UC«^
'?,... •• .-.il
-
AfXil-0: bi^npa(Ar.dtoii_*1JMMJX_Li.U [pi. pi. it.,
ApritlO; WfitCTi*«.lAI«h*BJ*.»J»»JJL_a S Jmi, m, fit,
V'W>: brtfllrrp*! *nahf«»_H IJMWJM.J-P ifnt, pi, Wj
OlW.pl. J.Olptt. pi. m
ivif mfMrr 4r.ilr:« 3~ 07^ .r» n n jpO, pn. irr
s~ i^OO Ci 0.1 jpd. po. m
-
t^l-imt-Arj^^u'_Uajl_U._LM!pif. pf. mt
tso2L-ipi!jH«hrlQ_R;_(lHIJ3L_3.0(pU. (*. tit.
•i IKaWp*tA™ffMj!tjM(rtJ!L_D,Qferf, pn, mt.
I1-.5G H f. S IPO. pp. T
fl H ;.n :pel. p-. 'i
ct, irfil, I*]
«. "W, [*]
H. "hl.plj
-..f-'.
-..[• •
4f*3
flf. nlii. plf]
.>- '•,-. r< ;
rrc npk n* n* n«
| v-n^jr]"'
This dialogue box contains the name of the
Regret Analyses (Demo), multiple Regret
Analyses can be setup and run at the same time
This dialogue box also displays the total
number of sensor evaluations in parentheses.
Click Close
O Regret Analysis
-Rpnrpt Analvsps-
Demo(342 sensor evaluations)
Edit
Delete
342 total sensor evaluations will be executed
73
-------
8.13. Scheduling the Available Executions
Click the Regret Analyses check box and a
green check mark appears.
TEVA-SPOT enables the Execute button (in
the Status pane). If that button remains
disabled, it means that the user must enter
missing execution parameters (i.e., create one
or more regret analysis parameter sets) before
the executions can be started.
Available Executions
-Regret Analysis
O| M Beget Analyse;!
Status
Not Running
Estimated Time Remaining
Estimated Total Time
Estimated Completion Time
Execute
Terminate
Click the Execute button in the Execution
Control Panel.
At the top of the pane is an indicator showing
the status of the current execution. Before the
Execute button is clicked, this indicator shows
a status of Not Running.
If execution terminates normally, the indicator
changes to Done.
Regret Analyses should generally run quickly.
Available Executions
Analysis-
0 Regret Analyses
Estimated Time Remaining 00:00:00:00
Estimated Total Time 00:00:02:09
Estimated Completion Time 05/15/2008 03:20:35 PM
74
-------
9. Regret Analysis Output - Tables
Unlike the output tables produced in ensemble analysis mode, those produced in Regret Analysis
Mode cannot be sorted by clicking a column head. In the output products pull-down list on the
control bar, the names of all the regret tables start with a prefix of the form [
_] = [Demo pe_mean]. Here, is the name given to a regret analysis
parameter set (of the regret analysis group whose execution produced the regret tables) when the
parameter set was created, and and are respectively one of the impacts and
one of the statistics whose check boxes were checked in that parameter set.
For the Demo example, this is the regret analysis which evaluated all of the optimal sensor
network designs from the "AprilSO" and "low" ensembles of threats for 10 locations and
determines which sensor network design(s) performs best.
It is helpful to export the table of data to a spreadsheet. Here is how export the data manually:
• Click on an individual cell which highlights an entire row.
• Next, holding the control key of your keyboard press the letter "A".
• Next, holding the control key of your keyboard, press the letter "C".
• Open a spreadsheet (Microsoft EXCEL) and press control-V and paste the Regret
Analysis table below into an empty spreadsheet. Conversely, the Regret Analysis table
can be exported to Excel via the export command under the Table menu.
The table of regret data can also be exported to EXCEL automatically. While under the Tables
Tab, select Table/Export to EXCEL and follow the instructions.
The table is composed of two sections of data:
• Top portion of data table corresponding to the first set of impact files evaluated are the
impacts data output from the sensor network designs evaluations. In the Demo example,
these data are Population Exposed.
• Lower portion of data table corresponding to the second set of impact files evaluated are
the impacts data represented as a difference from the no sensors baseline case. For
instance, notice column D (first row below no sensors) has a value of 353.179. This
corresponds to the no-sensors case for the indicated ensemble. Column E, the first sensor
design moving from left to right, same row shows an impact value of 230.839. The
corresponding value to the column E value of 230.839 in the lower portion of data table
has a value of 0.346397. The value of 0.34697 was found by subtracting the quantity of
(230.839 divided by 353.179) from 1. In other words, the data in the lower portion of the
table represents the percent difference for each design from the baseline, no-sensors case.
Across the top of the spreadsheet (columns) are the sensor network designs corresponding to
each set of data. In the spreadsheet highlight the row which is identified by "ENS". ENS is
short for Ensemble. Format the highlighted row for wrap-around text. This will highlight the
sensor network design names.
In rows down the spreadsheet are the impact files for which the sensor designs are evaluated
against. The data in the table are the impact values, i.e., population exposed health impact
75
-------
values. At the bottom of the spreadsheet are the performance metrics. The far right hand
column is labeled "BEST". For each row, BEST represents the lowest value (e.g., number of
people exposed) for the sensor network design evaluated given the impact measure defined in the
row of the spreadsheet. At the bottom of the BEST column are the performance metrics, i.e.,
lowest values, determined from all the sensor network designs. The preferred designs are color-
coded. The Preferred Overall Design(s) is highlighted green and the Preferred Design(s) to
Minimize Maximum Deviation is highlighted blue.
Ultimately, utility personnel should use the information gained from TEVA-SPOT to inform the
decision for placing sensors in the distribution system.
HTEVA Regret Analysis (Demo in Tutorial [Main])
RagretTibiei
Table ; [Demo pe.tnewn]: Regret *mhrtt)
v r.-ur
ngpMinag. ] waniwrg
Regret Analysis for Mean Estimated Population Exposed
A 6 C D
rqpact Bnwnjji
ENS
' V
ftT
i.c
1 n.
iesponw THTW OMKttyv . , *» Sac
C f
10 10
'- .Q
All L'XF.ffF. All LOC
C C 0. 1
Normal rfarma
IVt
C
10
1°
MOTH All LQcaiid
1.0
r-i-.r-.-r-, -,'
H 1 J K
..Tuttrial A... Tutorial A , Tutcrial fl... TuicriaU
10 10 10 10
,360 360 ,360 720
0-0 0.1 _1.0 O;.
ttyrna iKfi'ia' Mrmal ttcririal
L M
... tuErialj*i. .. Ti/ofalj1
LH LO
720 7ffl
0 1 ' L 0
Normal Normal
H 0
...-ruJEriii trw Tutorial to
10
0
ns All t.oc»
vrJ
tonal
i""
(Of a All LOcatlcr
0 1
'riVi-ii*.
f
j:
1,0
No- mar
9
'o.o
formal
ft £
jw Tutorial lowTui
"io '10
nf. /ill LOC-3t.«jns All
0,: :.D
I^Xr.Ml VA
T
crial tew-TuiB
LO
720
ctatera ^11 L
o:
n -
C'-2157404 0.23051?; Q.145G225 0.16C0252S 0.16929650 0.12507534 0.13431420'0.1429KK-i C lJ"='~r-;
0,0915i-446 OOS467378 0.10991Q73 0.106*917 0.1055272 0.tJ?050741 OO97&332 0,096097 Q,OC1379it
p,06917456 O.0652S062 0.0872920: 0.0923610...0.0905029? 0.076423 0.0791666.. 0079166S . C.0759534
.i3.QS74608fl 0-059963Q7 0.07239672 O.D937337. .0-006P719 Q.071&2&57 0-C'7-^157C.. OC7-?5345 &022495ei
C.C291JJ':. 0-0287247B 0.0325754...0,03307103 0.0340591 O.O404J«0 . OG40O0797 003946";'. C 02^2722
.'O.0217906 00219152. 00252959 O.D2795«. 0.0295600 .0.0336316 O03K053.. O 0349199. b.0087972.
00679*62. C0670670...p.0625400
006241564 00607258...&0615314 .
OO415603 O 0405270.. b.0204676
OOK09043 003111>1 00169030
0020t0865 Q0195393 OOL052591
0015435S...OOW5995 00055520
OOW6958...fl.OW2963. 0006*605
I 0,26132637 0.2^7H26S 0.2246H1
£ 0.19:77525 0,15775206 0.148M141;
:• 0,J*J&i;"bT* 0..^.5;~-: O..X150S33 O.L3QS
76
-------
10. Regret Analysis Mode: Filtering Criteria
For basic filtering functionality, refer to section 7.
Click the Manage button.
The Display Options Filter Management dialog box
opens.
Name Attribute - To select the Regret Analysis output
tables yielded by regret analyses associated with a
regret analysis parameter set having a particular name,
check the Name attribute and the value identifying the
desired name.
Impact Attribute - To select the Regret Analysis
output tables yielded by regret analyses associated with
a regret analysis parameter set specifying a particular
impact, check the Impact attribute and the desired
value identifying the desired impact.
Statistic Attribute - To select the Regret Analysis
output tables yielded by regret analyses associated with
a regret analysis parameter set specifying a particular
statistic, check the Statistic attribute the value
identifying the desired statistic.
0Display Options ...
Filters
Edit I Delete
H Edit Filter Settings X
Name
Filter Criteria
Analysis: Regret Analysis
&••["! Impact
I '•••• i _! Estimated Population Exposed
3 [""[Name
I '*•••• QH Demo
$••• P Statistic
I L- O Mean
•-'•• UK Table Type
'-••• fj Analysis
Results: 0 items found
Cancel
Filtering Regret Analysis Output Tables - Each Regret Analysis output table possesses a set of
attributes that characterize the table, and each attribute that a regret table possesses assumes
exactly one value. Filter specifications designed to select Regret Analysis output tables contain
attribute nodes for all the attributes possessed by the existing Regret Analysis output tables and
value nodes for all the values that those attributes assume in the existing tables.
Regret Analysis Value - To select the Regret Analysis output tables that are regret tables (tables
of regret analysis results), check the Table Type attribute and the Regret Analysis value.
77
-------
11. Ensemble Management: Export Capabilities
The Ensemble Management dialogue box provides the ability to Import, Export, Export Map
Data, Rename, Move, and Delete ensembles.
11.1. Import
Ensemble
Import Ensemble
provides the ability to
import an ensemble
created by another
user and on a different
computer.
H Ensemble Management
Ensemble Collection
testl [Main]
Ensembles
test-demo-1
Ensemble Collection Information
Read Only
Size
Last Saved Date
Last Ensemble Execution
Last HIA Execution
Last TS02Impact Execution
Last Sensor Placement Execution
False
1.0Mb
10/20/2009 01:32PM
10/20/2009 01:32PM
10/20/2009 01:32PM
Never
10/20/2009 01:32PM
Clicking "Import"
brings up a dialogue
box to navigate to the
location of the
ensemble. The Select
button allows the user
to browse to the
location where the
ensemble is located.
Ensembles are
imported and exported
as a compressed ZIP
file.
test-dem
Ensemble
Read Oi
)-l
E Import scenario | >
File | Seieci:'."."|
Import | Cancel [
I
-\
1
Size
Last Saved Date
Last Ensemble Execution
Last HIA Execution
Last TS02Impact Execution
Last Sensor Placement Execution
1.0Mb
10/20/2009 01 ;32PM
10/20/2009 01:32PM
10/20/2009 01 ;32PM
Mever
10/20/2009 01:32PM
78
-------
11.2. Export
Ensemble
Clicking EXPORT
opens the dialogue
box as shown.
Choices include:
D Include ISO
D Include HIA
Results
D Include
tso2Impact
results
D Include
tso2Impact
(SP) results
D Include Sensor
Placement
results
If no boxes are
checked, then just the
ensemble definition
will be exported.
Checking the "Include
TSO" box results in
the export of the
entire ensemble of
results, including the
binary contaminant
concentration data
fromEPANET. The
resulting ZIP file may
be very large!
Checking the "Include
HIA Results" will
provide an export of
only the HIA results.
Checking additional
will include identified
items.
E Ensemble Management
Ensemble Collection
Netl
Netl
Netl
Netl
Netl
Netl
Netl
n Include TSO (1 ensemble, 18,8 Mb)
n Include HIA results (1 ensemble, 1,3 Mb)
Include tsoZIrnpact results
G Include tsoZImpact (SP) results (1 ensemble, 86,4 kb)
Q Include Sensor Placement results (1 ensemble, 536.7 kb)
Ense
Rea
Size
Las
Last Ensemble Execution
Last HIA Execution
Last T502Impact Execution
Last Sensor Placement Execution
04/23/2009 11:53 AM
05/26/2009 02:57 PM
Never
09/26/2009 09:23 PM
Export Map Data
Rename
B Export scenario: Net1_ROO_D24h_S168_popD_... fX] |
File
Op
|| Select... |
Q Include T5O (1 ensemble, 18.8 Mb)
n Include HIA results (1 ensemble., 1.3 Mb)
Include tsoZImpact results
G Include tsoZImpact (SP) results (1 ensemble, 86.4 kb)
Q Include Sensor Placement results (1 ensemble, 536.7 kb)
Export
Cancel
79
-------
11.3. Export Map
Data
Exporting Map Data
provides easy
incorporation into
ArcGIS. Results are
saved in the dBaselV
format. All results are
exportable except the
sensor placement
results dealing with
detection events on
links. These maps
will require topology
information, e.g.,
ESRI shapefile which
is currently not
available.
Map Export data are
exported as ZIP Files.
Select allows the user
to select the export
destination for the ZIP
file
tnsenriDieg
S
test-derm
Ensemble
Read Oi
Size
Last Sa\
Last En<
1 ^ch HTi
Last TS<
H Export Map Data: Testnet3 X
File | Select... j
rv !_•
Q| Health Impacts Maps
Q| Sensor Placement
Data Format
G dBase IV
^j ESRI Shapefile
Export Cancel
.
Import Export
Export Map Data
Rename Move Delete
[ Close ]
For the dbf files which appear in the ZIP file, "Inj" refers to the
injection scenarios data. For example, Inj_fatalities.dbf provides the
number of estimated fatalities by injection scenario (node ID) along
with x and y coordinates data for each contaminant injection. "Rec"
refers to receptors. "Dot" refers to dose-over-threshold data which is
followed in the file name by the respective threshold.
12. Trouble-Shooting
TEVA-SPOT GUI generates numerous log files to help the user identify
problems. However, to view the log files a live or real-time editor is
needed. Windows notepad is a live editor but is not compatible with the
log files output by TEVA-SPOT, i.e., everything looks jumbled. An
open source program called emacs can be downloaded from the link
below. Installing emacs on the desktop will allow a drag and drop
approach for viewing each of the log files.
• http://www.xemacs.org/Download/win32/index.htmltfInnoSetup-
Stable-Download
The log files are located in the following directory: C:\Program
Files\TEVA-SPOT\tevaLogs. Within the "tevaLogs" folder are
additional folders with the logs separated by each TEVA-SPOT process.
The TEVA-SPOT processes include: (1) AnalysisServer, (2)
DBServer, (3) ExecutionController, (4) Server, and (5) ServerBroker
80
-------
12.1 Stop & Restart
TEVA-SPOT Services
Under the START
Menu/TEVA-SPOT
are programs for
managing TEVA-
SPOT executions.
-' Restart TEVA-5POT Services
-' Start TEVA-5POT Services
;i Stop TEVA-SPOT Services
T| TEVA-SPOT
(^ Uninstall TEVA-SPOT
Users Guide
^j Zip Log Files
If TEVA-SPOT seems held up or frozen for an extended period of time,
RESTART TEVA-SPOT Services can be selected. "Restart" or
"Stop/Start" TEVA-SPOT Services will stop all simulations and ready
TEVA-SPOT for re-execution of simulations and analyses.
Prior to choosing "Restart" or "Stop/Start" TEVA-SPOT services
ensure that TEVA-SPOT simulations and analyses (executions) are
indeed stopped and not just running slowly due to a difficult sensor
placement problem. It is easiest to check if TEVA-SPOT is running by
examining the Windows Task Manager (right click Windows Start
Menu bar and choose Task Manager). Examine Task Manager to
determine if any TEVA-SPOT Services are running (you may need to
check box "show processes from all users). Processes for TEVA-SPOT
include:
• java.exe
• tevaAnalysisServer.exe
• tevaBroker.exe
• tevaDBServer.exe
• tevaExecCtlr.exe
• tevaRMIRegistry.exe
• tevaServer.exe (multiples of these depending the number of
Server folders under c:/Program Files/TEVA-SPOT/Server
• randomsample.exe
81
-------
12.2. Log Files
For the more
experienced user, log
files can be examined
to determine the
possible problem.
The execution log
files appear within the
"Execution-
Controller" folder.
Similar files are
available for the other
TEVA-SPOT
processes.
Exec utionContro Her
File Edit View Favorites Tools Help
Qfiack - Q
Ci'iPrograrriFileslTEVA-spiJT'itevaLngsiF^rijtionController
l|] ExecutionController_2010-07-06, memory, log
IE ExecutionController_2010-Q7-06.log
||1 Exeojtioneontroller_2010-07-06.5tat5.log
ExecutimController_2C10-Q7-06,stats,detail,log
f|] ExecutionController_2010-07-06.execstats.log
l|] ExecutionController_2Q10-u7-Q6,assign,log
Size Type
456 KB Text Document
7 KB Text Document
1 KB Text Document
1 KB Text Document
1 KB Text Document
2 KB Tent Document
Date Modified
7/6/2010 12:46 PM
7/6/2010 8:45 AM
7/6/2010 3:45 AM
7/6/2010 7:36 AM
7/6/2010 7:36 AM
7/6/2010 7:36 AM
12.3.
File
Example Log
Examine the
Execution-Controller
log files first for
problems. Below are
descriptions of
selected log files.
Look for "Severe Exceptions" in the ExecutionController.. .log" file,
except ignore the following error: "Exception occurred while retrieving
DB Server: java.rmi.NotBoundException: TEVADatabaseServer"
Example ExecutionController... stats, log file: EPANET/HIA simulation. The three "..." will
display the date for the log file.
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
(07/06/201015:28:
07.981):
.-07.981):
.-07.997):
.-07.997):
08.497):
08.559):
09.059):
.-09.106):
: 09.544):
:09.591):
executionName
executionName
11 8330.0(0 hours-
12 8330.0 (0 hours-
13 8330.0(0 hours-
21 8330.0(0 hours-
22 8330.0(0 hours-
23 8330.0 (0 hours-
31 8330.0(0 hours-
32 8330.0 (0 hours-
execution
execution
1 hour)
1 hour)
1 hour)
1 hour)
1 hour)
1 hour)
1 hour)
1 hour)
Tune
Time
4.172
4.172
0.5
0.546
0.546
0.531
0.469
0.469
successful completion
successful completion
Server name
Server name
true COMPUTEKNAME[204.47.178.206]-1
true COMPUTEKNAME[204.47.178.206]-2
true COMPUTERNAME[204.47.178.206]-1
true COMPUTERNAME[204.47.178.206]-2
true COMPUTERNAME[204.47.178.206]-1
true COMPUTEKNAME[204.47.178.206]-2
true COMPUTEKNAME[204.47.178.206]-1
true COMPUTEKNAME[204.47.178.206]-2
Time is execution time in seconds. "True" indicates that the simulation executed correctly to
completion. A "False" here indicates a problem. "ComputerName" will be replaced by the
name of the computer that TEVA-SPOT is running on. Under "ExecutionName" the first
number refers to the EPANET model node ID, in this case "11".
82
-------
Example ExecutionController. . . stats, log file: TSO-2-Impact simulation. ModuleRunner
refers to each TSO-2-Impact analysis.
(07/06/201015:28:14.966): executionName execution Time
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
INFO:
(07/06/2010 15:28:14.966):
(07/06/2010 15:28:14.966):
(07/06/2010 15:28:15.012):
(07/06/2010 15:28:15.028):
(07/06/2010 15:28:15.044):
(07/06/2010 15:28:15.059):
(07/06/2010 15:28:15.075):
(07/06/2010 15:28:15.075):
ModuleRunnerO
ModuleRunner 1
ModuleRunner4
ModuleRunner3
ModuleRunner2
ModuleRunner6
ModuleRunnerS
ModuleRunner?
0.266
0.25
0.031
0.062
0.032
0.031
0.016
0.016
successful completion Server name
true
true
true
true
true
true
true
true
COMPUTERNAME[204.47.178.206]-1
COMPUTERNAME[204.47.178.206]-2
COMPUTERNAME[204.47.178.206]-2
COMPUTERNAME[204.47.178.206]-1
COMPUTERNAME[204.47.178.206]-2
COMPUTERNAME[204.47.178.206]-1
COMPUTERNAME[204.47.178.206]-2
COMPUTERNAME[204.47.178.206]-1
Example ExecutionController.. .stats.log file: Sensor network design analysis
INFO: (07/06/2010 15:28:41.043):
INFO: (07/06/2010 15:28:41.043):
INFO: (07/06/2010 15:28:45.012):
INFO: (07/06/2010 15:28:48.762):
INFO: (07/06/2010 15:28:52.668):
INFO: (07/06/2010 15:28:56.309):
INFO: (07/06/2010 15:28:59.981):
INFO: (07/06/2010 15:29:03.652):
INFO: (07/06/2010 15:29:07.262):
INFO: (07/06/2010 15:29:09.762):
INFO: (07/06/2010 15:29:12.215):
executionName
execution Time successful completion Server name
SOgrasp-NS010-OBpe_mean-RTOOO-SMstd23.702 true
SOgrasp-NSO 10-OBpd_mean_00-RTOOO-SMstd 3.953
SOgrasp-NS005-OBpe_mean-RTOOO-SMstd3.734 true
SOgrasp-NS005-OBpd_mean_00-RTOOO-SMstd 3.89
SOgrasp-NS004-OBpe_mean-RTOOO-SMstd3.625 true
SOgrasp-NS004-OBpd_mean_00-RTOOO-SMstd 3.672
SOgrasp-NS001-OBpe_mean-RTOOO-SMstd3.656 true
SOgrasp-NSOO 1 -OBpd_mean_00-RTOOO-SMstd 3.61
SOgrasp-NSOOO-OBpe_mean-RTOOO-SMstd2.485 true
SOgrasp-NSOOO-OBpd_mean_00-RTOOO-SMstd 2.453
#[204.47.178.206]-2
true #[204.47.178.206]-2
#[204.47.178.206]-2
true #[204.47.178.206]-2
#[204.47.178.206]-2
true #[204.47.178.206]-2
#[204.47.178.206]-2
true #[204.47.178.206]-2
#[204.47.178.206]-2
true #[204.47.178.206]-
Here "SO" refers to the sensor placement solver. "GRASP" refers to the heuristic algorithm.
"NS" is number of sensors. "OB" refers to the objective for sensor placement, e.g., "pd" is
population over dose threshold while "pe" is population exposed. "RT" is response delay time.
If a Detection Limit was specified it would be included in the name. Completion time is in
seconds. Note that "#" was used to replace "ComputerName" to save space in this manual.
The final step of the ensemble execution will result in the following entry in the
ExecutionController... stats.log file:
INFO: (07/06/2010 15:29:16.246):
INFO: (07/06/2010 15:29:16.246):
executionName execution Time successful completion Server name
SensorPlacementSummary 3.406 true #[204.47.178.206]-2
This entry indicates the time taken to write the sensor placement summary results to the
graphical user interface.
83
-------
APPENDIX A
Water Utility Requirements for using EPA's TEVA-SPOT software
The Threat Ensemble Vulnerability Assessment Sensor Placement Optimization Tool, TEVA-
SPOT, is a software product and decision-making process developed by EPA's TEVA Research
program to assist in determining the best location for sensors in a distribution system. TEVA-
SPOT software has been applied by EPA staff using models and data provided by utilities. In
some cases, significant improvements in the models have been made in order to bring them up to
the standards required by the TEVA-SPOT software. In order to streamline the application of
TEVA-SPOT, this appendix was developed to describe the requirements and steps that utilities
must follow to use TEVA-SPOT.
Table 2 summarizes the data and information required for a water utility to use TEVA-SPOT;
each component is described in more detail in the text. In addition to having an appropriate
utility network model, utilities will need to make decisions about the nature of the Contamination
Warning System they are designing and the types of security incidents that they would like to
detect.
Table 2: Information and data required to design sensor networks using TEVA-SPOT
INFORMATION AND DATA
NEEDED FOR SENSOR
PLACEMENT
Utility Network Model
Sensor Characteristics
Design Basis Threat
Performance Measures
Utility Response
Potential Sensor Locations
DESCRIPTION
The model (e.g. EPANET input file) should be up-to-date, capable
of simulating operations for a 7-10 day period, and calibrated with
field data.
Type of sensors or sampling program, detection limits, and (if
applicable) event detection system
Data describing type of event that the utility would like to be able
to detect: specific contaminants, behavior of adversary, and
customer behavior
Utility specific critical performance criteria, such as time to
detection, number of illnesses, etc.
Plan for response to a positive sensor reading, including total time
required for the utility to limit further public exposure.
List of all feasible locations for placing sensors, including
associated model node/junction.
Utility Network Model
The TEVA-SPOT software relies upon an EPANET hydraulic model of the network as the
mechanism for calculating the impacts resulting from contamination incidents. Therefore, an
acceptable model of the distribution system is needed in order to effectively design the sensor
system. The following sub-sections describe the various issues/characteristics of an acceptable
hydraulic model for use within TEVA-SPOT.
EPANET Software Requirement - TEVA-SPOT uses EPANET, a public domain water
distribution system modeling software package. In order to utilize TEVA-SPOT, existing
84
-------
network models that were not developed in EPAKET must be converted to and be demonstrated
to properly operate within EPANET (Version 2.00.10). Most commercial software packages
utilize the basic EPANET calculation engine and contain a conversion tool for creating an
EPANET input file from the files native to the commercial package. The user may encounter
two potential types of problems when they attempt to make the conversion: (1) some commercial
packages support component representations that are not directly compatible with EPANET such
as representation of variable speed pumps. The representation of these components may need to
be modified in order to operate properly under EPANET; (2) additionally, the conversion
programs developed by the commercial software vendors may also introduce some unintended
representations within EPANET that may require manual correction. Following conversion, the
output from the original model should be compared with the EPANET model output to ensure
that the model results are the same or acceptably close (see section on Model Testing).
Extended Period Simulation - TEVA-SPOT uses both the hydraulic and water quality
modeling portions of EPANET. In order to support the water quality modeling, the model must
be an extended period simulation (EPS) that represents the system operation over a period of
several days. Typically a model that uses rules to control operations (e.g., turn pump A on when
the water level in tank B drops to a specified level) are more resilient and amenable to long
duration runs than are those that use controls based solely on time clocks. Model output should
be examined to ensure that tank water levels are not systematically increasing or decreasing over
the course of the run since that will lead to unsustainable situations.
The required length of simulation depends on the size and operation of the specific water system.
However, in general, the length of the simulation should reflect the longest travel times from a
source to customer nodes. This can be calculated by running the EPANET model for water age
and determining the higher water age areas. In determining the required simulation length, small
dead-ends (especially those with zero demand nodes) can be ignored. Typically a run length of 7
to 10 days is required for TEVA-SPOT though shorter periods may suffice for smaller systems
and longer run times required for larger or more complex systems.
Seasonal Models - In most cases, water security incidents can take place at any time of the day
or any season of the year. As a result, sensor systems should be designed to operate during one
or more representative periods in the water system. It should be noted that this differs
significantly from the normal design criteria for a water system where pipes are sized to
accommodate water usage during peak seasons or during unusual events such as fires. In many
cases, the only available models are representative of these extreme cases and generally,
modifications to such models should be made to reflect a broader time period prior to use with
TEVA-SPOT. Specific guidance on selecting models is provided below:
• Optimal situation: the utility has multiple models representing average conditions
throughout the year, typical higher demand case (e.g., average summer day) and typical lower
demand case (e.g., average winter day).
• Minimal situation: the utility has a single model representing relatively average
conditions throughout the year.
• Situations to avoid: the utility has a single model representing an extreme case (e.g.,
maximum day model).
85
-------
• Exceptions: (1) If a sensor system is being designed to primarily monitor a water system
during a specific event such as a major annual festival, then one of the models should reflect
conditions during that event; and (2) If the water system experiences little variation in water
demand and water system operation over the course of the year, then a single representative
model would suffice.
Model Detail - A sufficient amount of detail should be represented in the model for use within
TEVA-SPOT. This does not mean that an all-pipes model is required nor does it mean that a
model that only represents transmission lines would suffice. At a minimum, all parts of the
water system that are considered critical from a security standpoint should be included in the
model, even if they are on the periphery of the system. The following guidance drawn from the
Initial Distribution System Evaluation (IDSE) Guidance Manual of the Final Stage 2
Disinfectants and Disinfection Byproducts Rule provides a reasonable lower limit for the level
of detail required for a TEVA-SPOT analysis (USEPA, 2006).
Most distribution system models do not include every pipe in a distribution system. Typically,
small pipes near the periphery of the system and other pipes that affect relatively few customers
are excluded to a greater or lesser extent depending on the intended use of the model. This
process is called skeletonization. Models including only transmission networks (e.g. pipes larger
than 12 inches in diameter only) are highly skeletonized while models including many smaller
diameter distribution mains (e.g. 4 to 6 inches in diameter) are less skeletonized. In general,
water moves quickly through larger transmission piping and slower through the smaller
distribution mains. Therefore, the simulation of water age or water quality requires that the
smaller mains be included in the model to fully capture the residence time and potential water
quality degradation between the treatment plant and the customer. Minimum requirements for
physical system modeling data for the IDSE process are listed below.
• At least 50 percent of total pipe length in the distribution system.
• At least 75 percent of the pipe volume in the distribution system.
• All 12-inch diameter and larger pipes.
• All 8-inch diameter and larger pipes that connect pressure zones, mixing zones from
different sources, storage facilities, major demand areas, pumps, and control valves, or are
known or expected to be significant conveyors of water.
• All 6-inch diameter and larger pipes that connect remote areas of a distribution
system to the main portion of the system or are known or expected to be significant conveyors of
water.
• All storage facilities, with controls or settings applied to govern the open/closed
status of the facility that reflect standard operations.
• All active pump stations, with realistic controls or settings applied to govern their
on/off status that reflect standard operations.
• All active control valves or other system features that could significantly affect the
flow of water through the distribution system (e.g., interconnections with other systems, pressure
reducing valves between pressure zones).
Model Demands - The movement of water through a distribution system is largely driven by
water demands (consumption) throughout the system. During higher demand periods, flows and
velocities generally increase and vice versa. Demands are usually represented within a model as
86
-------
average or typical demands at most nodes and then (a) global or regional demand multipliers are
applied to all nodes to represent periods of higher or lower demand, and (b) temporal demand
patterns are applied to define how the demands vary over the course of a day. In some models,
demands within a large area have been aggregated and assigned to a central node. In building a
model for use with TEVA-SPOT, rather than aggregating the demands and assigning them to
only a few nodes, each demand should be assigned to the node that is nearest to the actual point
of use. Both EPANET and most commercial software products allow the user to assign multiple
demands to a node with different demands assigned to different diurnal patterns. For example,
part of the demand at a node could represent residential demand and utilize a pattern
representative of residential demand. Another portion of the demand at the same node may
represent commercial usage and be assigned to a representative commercial diurnal water use
pattern. TEVA-SPOT supports either a single demand or multiple demands assigned to nodes.
Model Calibration/Validation - Calibration is the process of adjusting model parameters so that
predicted model outputs generally reflect the actual behavior of the system. Validation is the
next step after calibration, in which the calibrated model is compared to independent data sets
(i.e., data that was not used in the calibration phase) in order to ensure that the model is valid
over wider range of conditions. There are no formal standards in the water industry concerning
how closely the model results need to match field results nor is there formal guidance on the
amount of field data that must be collected. Calibration methods that are frequently used include
roughness (c-factor) tests, hydrant flow tests, tracer tests and matching model results over
extended periods for pressure, flow and tank water levels to field data collected from SCADA
systems or special purpose data collection efforts.
The IDSE Guidance Manual stipulates the following minimum criteria in order to demonstrate
calibration. "The model must be calibrated in extended period simulation for at least a 24-hour
period. Because storage facilities have such a significant impact upon water age and reliability
of water age predictions throughout the distribution system, you must compare and evaluate the
model predictions versus the actual water levels of all storage facilities in the system to meet
calibration requirements." For TEVA-SPOT application, the water utility should calibrate the
model to a level that they are confident that the model adequately reflects the actual behavior of
the water system being represented by the model. Some general guidelines for
calibration/validation are shown below:
• If the model has been in operation actively and for several years and has been applied
successfully in a variety of extended period simulation situations, then further substantial
calibration may not be necessary. However, even in this case, it is prudent to demonstrate the
validity of the model by comparing the model results to field measurements such as time-varying
tank water levels and/or field pressure measurements.
• If the model has been used primarily for steady state applications, then further
calibration/validation emphasizing extended period simulation is needed.
• If the model has been recently developed and not undergone significant application, then
a formal calibration/validation process is needed.
Model Tanks - Tank mixing models: Most water distribution system models use a "complete
mix" tank representation that assumes that tanks are completely and instantaneously mixed.
EPANET (and most commercial modeling software models) allow for alternative mixing models
such as last in-first out (LIFO), first in-first out (FIFO), and compartment models. If a utility has
87
-------
not previously performed water quality modeling, they may not have determined the most
appropriate tank mixing model for each tank. Since the tank mixing model can affect
contaminant modeling and thus the sensor placement decisions, tank mixing models should be
specified in the EPANET model input files.
Model Testing - The final step in preparing the model for use in TEVA-SPOT is to put the
model through a series of candidate tests. Following is a list of potential tests that should be
considered.
1. If the model was developed and applied using a software package other than EPANET,
then following its conversion to EPANET, the original model and the new EPANET model
should be run in parallel under EPS and the results compared. The models should give virtually
the same results or very similar results. Comparisons should include tank water levels and flows
in major pipes, pumps and valves over the entire time period of the simulation. If there are
significant differences between the two models, then the EPANET model should be modified to
better reflect the original model or differences should be explained and justified.
2. The EPANET model should be run over an extended period (typically 1 to 2 weeks) to
test for sustainability. In a sustainable model, tank water levels cycle over a reasonable range
and do not display any systematic drops or increases. Thus, the range of calculated minimum
and maximum water levels in all tanks should be approximately the same in the last few days of
the simulation as they were in the first few days. Typically, a sustainable model will display
results that are in dynamic equilibrium in which temporal tank water level and flow patterns will
repeat themselves on a daily (or other periodic) basis.
3. If the water system has multiple sources, then the source tracing feature in EPANET
should be used to test the movement of water from each source. In most multiple source
systems, operators generally have a good idea as to how far the water from each source travels.
The model results should be shown to the knowledgeable operators to ensure that the model is
operating in a manner that is compatible with their understanding of the system.
4. The model should be run for a lengthy period (1 to 2 weeks) using the water age option in
EPANET in order to determine travel times. Since the water age in tanks is not usually known
before modeling, a best guess (not zero hours) should be used to set an initial water age for each
tank. Then after the long run of the model, a graph of calculated water age should be examined
for each tank to ensure that it has reached a dynamic equilibrium and is still not increasing or
decreasing. If the water age is still systematically increasing or decreasing, then the plot of age
for each tank should be visually extrapolated to estimate an approximate final age and that value
should be reinserted in the model as an initial age, and the model rerun for the extended period.
The model output of water age should then be investigated for reasonableness, e.g., are there
areas where water age seems unreasonably high? This exercise will also help to define a
reasonable upper limit for the duration of the model run to be used in the TEVA-SPOT
application.
Following these test runs, any identified modifications should be made in the model to ensure
that the model will operate properly under TEVA-SPOT. Many utilities will not be able to make
all of the above modifications to their network model. In that case, TEVA-SPOT can still be
applied; however the overall accuracy of the results will be questionable and should only be
considered applicable to the system as described by the model.
88
-------
Sensor Characteristics
The TEVA-SPOT analysis requires some assumptions about the detection characteristics of the
sensors. In particular, the sensor type, detection limit, and accuracy need to be specified. For
example, the analysis can specify a contaminant-specific detection limit that reflects the ability
of the water quality sensors to detect the contaminant. Alternatively, the analysis can assume
perfect sensors that are capable of detecting all non-zero concentrations of contaminants with
100% reliability. The latter assumption, though not realistic, provides an upper bound on
realistic sensor performance. If the utility has some idea as to the type of sensor that they are
planning on using and its likely detection performance, then that information should be provided
in a sensor data file. If that information is not available, then some typical default values will be
used in the TEVA-SPOT analysis.
In order to quantify detection limits for water quality sensors, the utility must indicate the type of
water quality sensor being used, as well as the disinfection method used in the system.
Generally, water quality sensors are more sensitive to contaminant introduction with chlorine
disinfection than with chloramine. As a result, contaminant detection limits may need to be
increased in the design of a sensor network for a chloraminated system.
Ongoing pilot studies for EPA's Water Security Initiative use a platform of water quality
sensors, including free chlorine residual, total organic carbon (TOC), pH, conductivity, oxidation
reduction potential (ORP), and turbidity (USEPA, 2005b). The correlation between contaminant
concentration and the change in these water quality parameters can be estimated from
experimental data, such as pipe loop studies (Hall et al., 2007; USEPA, 2005c). Of these
parameters, chlorine residual and TOC seem to be most likely to detect a wide range of
contaminants.
Detection limits for water quality sensors can be defined in terms of the concentration that would
change one or more water quality parameters enough that the change would be detected by an
event detection system (for example, Cook et al., 2005; McKenna, Wilson and Klise, 2006) or a
water utility operator. A utility operator may be able to recognize a possible contamination
incident if a change in water quality is significant and rapid. For example, if the chlorine
residual decreased by 1 mg/L, the conductivity increased by 150 jiSm/cm, or TOC increased by
1 mg/L.
A sensor data file is a text file that can be created with any text editor. TEVA-SPOT requires
two types of sensor files depending on the application. One type of sensor file is used in the
Sensor Placement module under Location Categories. This sensor file is described in Table 3.
The second type of sensor file is used under the command HIA Sensors-Import HIA Sensor
Data. This type of sensor file is described in Table 4. Each line of the sensor file for loading
HIA sensors into TEVA-SPOT consists of nine text fields separated by one or more spaces or
tabs. The contents of the fields are defined in Table 4. The HIA Sensors feature is used to load
a single sensor network design into TEVA-SPOT in order to produce a complete set of maps,
charts, and tables delineating the reduced impacts given the sensor network design. Each sensor
file can be named as the user chooses, with a txt extension.
89
-------
Table 3: Sensor File for Use in Sensor Placement Module
Field
#Name
Description
Literal string.
feasible, infeasible, fixed, and unfixed
The name of a node in the water network.
For example, the sensor placement file should appear as follows (notice there no spaces in the
title name).
#Name: Demo_2_sensors_fixed_location_design
infeasible ALL
fixed Junction-100 Junction-101
Note that after the keywords "infeasible and fixed" there is a space or tab. Similarly, there is a
space or tab between node identifiers. Each line of the must begin with a keyword.
Table 4: Fields of Sensor Data Files for Loading Sensors into Health Impacts Assessment
(HIA) Sensors
Field
SAMPLELOC
Description
Literal string.
The name of a node in the water network.
A string defining the type of the sensor located at the node given by .
Valid values for are given below.
A string defining the type of the sampling performed by the sensor located at the
node given by . Valid values for are given below.
The minimum concentration detectable by the sensor.
The volume, in liters, sampled by the sensor (ignored for sensors having
= REALTIME).
The sensor's sampling frequency in hours.
The elapsed time, in hours, from the start of the simulation at which the sensor
takes its first sample.
The sensor's lag time, in hours, from detection to reporting. This is the same as
Response Time Delay.
Valid values for the are as follows:
EXISTING, which indicates that a sensor (not placed by the Sensor Placement Algorithm)
already exists at the node given by ;
SELECTED, which indicates that the Sensor Placement Algorithm considered the node given by
and, in fact, recommended placing a sensor there;
POTENTIAL, which indicates that the Sensor Placement Algorithm considered the node given
by but did not recommend placing a sensor there; and
IGNORED, which indicates that the Sensor Placement Algorithm did not consider the node given
by .
Valid values for the are as follows:
90
-------
REALTIME, which indicates that the sampling performed at the node given by is
sampled instantaneously (i.e., in real time). There is no linkage to the volume of water that
passes the node of the sensor location when the sample type is REALTIME (i.e., the
is ignored).
COMPOSITE, which indicates that the sampling performed at the node given by is a
collection of samples (i.e., composite sample). The volume of composited water that passes the
node of the sensor location is specified by the . This sample type is useful for
detecting biological organisms that may only exist in very small concentrations.
FILTERED, which indicates that the sampling performed at the node given by is a set
of filtered samples. The volume of filtered water that passes the node of the sensor location is
specified by the . This sample type is useful for detecting biological organisms
that may only exist in very small concentrations.
The sensor file for loading into HIA Sensors must be a text file with each sensor location
(nodelD) and its associated identifiers specified on a single line in the text file. For example, the
text file should appear as follows (with the keywords replaced as appropriate):
SAMPLELOC 0 24 1 24
Design Basis Threat
A design basis threat identifies the type of threat that a water utility seeks to protect against when
designing a CWS. In general, a CWS is designed to protect against contamination threats;
however, there are a large number of potentially harmful contaminants and a myriad of ways in
which a contaminant can be introduced into a distribution system. Some utilities may wish to
design a system that can detect not only high impact incidents, but also low impact incidents that
may be caused by accidental backflow or cross-connections. It is critical for a water utility to
agree upon the most appropriate design basis threat before completing the sensor network design.
Contamination incidents are specified by the type of contaminant(s), the quantity of contaminant,
the location(s) at which the contaminant is introduced into the water distribution system, the time
of day of the introduction, and the duration of the contaminant introduction. Given that it is
difficult to predict the behavior of adversaries, it is unlikely that a water utility will know, with
any reasonable level of certainty, the specific contamination threats that they may face. The
TEVA-SPOT approach assumes that most of these parameter values cannot be known precisely
prior to an incident; therefore, the modeling process must take this uncertainty into account.
As an example, probabilities are assigned to each location in a distribution system indicating the
likelihood that the contaminant would be introduced at that location. The default assumption is
that each location is equally likely to be selected by an adversary (each has the same probability
assigned to it). A large number of contamination incidents (an ensemble of incidents) are then
simulated and sensor network designs are selected based on how well they perform for the entire
set of incidents. Based on their completed vulnerability assessment and other security related
studies, a utility may have some knowledge or preconceived ideas that will assist in refining
these assumptions. Some specific questions to consider are listed below:
91
-------
• Are there certain locations that should be assigned higher probabilities? Should all nodes
be considered as possible introduction sites or only nodes with consumer demand?
• Should utility infrastructure sites, such as storage tanks, be considered as potential
contamination entry locations?
• Are there specific contaminants of concern to this utility?
Performance Measures
The TEVA-SPOT software utilizes an optimization model that selects the best sensor design in
order to meet a specific set of performance measures. These measures are also sometimes
referred to as objectives or metrics. There are several performance measures that are currently
supported by TEVA-SPOT including:
• the number of individuals exposed to a contaminant
• the percentage of incidents detected
• the time of detection
• the length of pipe contaminated
• the contaminated mass consumed
• the contaminated volume consumed
With each of these performance measures, several statistics can be used to design the sensor
network. The following statistics are currently supported by TEVA-SPOT:
• the minimum impact over all contamination events
• the mean impact over all contamination events
• the worst impact over all contamination events
• the lower quartile: 25% of contamination incidents have an impact value less than this
quartile
• the median: 50% of contamination incidents have an impact value less than this quartile
• the upper quartile: 75% of contamination incidents have an impact value less than this
quartile
• the value at risk (VaR): 100*beta% of contamination incidents have an impact value less than
this value, where beta is a user-defined parameter
• the tailed-conditioned expectation (TCE): the mean value of the impacts that are greater than
or equal to VaR
Although it requires more time and input from the user, TEVA-SPOT can also consider multiple
objectives in its analysis. If the water utility has any preferences in the area of performance
measures, they should specify which of the above measures should be used and the relative
importance (weight) to be assigned to each measure. If there are other quantitative measures that
they wish to be considered, these should be specified.
Utility Response
The TEVA-SPOT analysis uses the concept of "response times" in the analysis of the
effectiveness of a sensor system. Response time is an aggregate measure of the total time
between the indication of a contamination incident (e.g., detection of an unusual event by a
92
-------
sensor system) and full implementation of an effective response action such that there are no
more consequences. The following response activities are likely following sensor detection
(USEPA, 2004; Bristow and Brumbelow, 2006):
• Credibility determination: integrating data to improve confidence in detection; for
example, by looking for additional sensor detections or detection by a different monitoring
strategy, and checking sensor maintenance records.
• Verification of contaminant presence: collection of water samples in the field, field tests
and/or laboratory analysis to screen for potential contaminants.
• Public warning: communication of public health notices to prevent further exposure to
contaminated water.
• Utility mitigation: implementing appropriate utility actions to reduce likelihood of further
exposure, such as isolation of contaminated water in the distribution system or other hydraulic
control options.
• Medical mitigation: public health actions to reduce the impacts of exposure, such as
providing medical treatment and/or vaccination.
Past analyses have shown that the benefits of a contaminant warning system (CWS) are very
dependent on a rapid response time. Typically, the TEVA-SPOT analysis assesses a range of
response times between 0 and 24 hours. A zero hour response time is obviously infeasible but is
usually analyzed in TEVA-SPOT as the best-case scenario and thus the maximum benefits that
can be associated with a CWS. Water utilities should assess their own emergency response
procedures and their acceptable risk tolerance in terms of false negative and false positive
responses in order to define a range of response times to be used in the TEVA-SPOT analysis.
Potential Sensor Locations
TEVA-SPOT searches potential sensor locations to determine those set of locations (nodes) that
will result in the optimal performance measure for a particular number of sensors. Utilities can
choose to consider all nodes as potential sensor locations or to limit the search to a subset of
specified nodes.
The primary physical requirements for locating sensors at a particular location are accessibility,
electricity, physical security, data transmission capability, sewage drains, and temperatures
within the manufacturer specified range for the instrumentation (ASCE, 2004). Accessibility is
required for installation and maintenance of the sensor stations. Electricity is necessary to power
sensors, automated sampling devices, and computerized equipment. Physical security protects
the sensors from natural elements and vandalism or theft. Data transmission is needed to
transmit sensor signals back to a centralized SCADA database, and can be accomplished through
a variety of solutions including digital cellular, private radio, land-line, or fiber-optic cable.
Sewage drains are required to dispose of water and reagents from some sensors. Temperature
controls may be needed to avoid freezing or heat damage.
Most drinking water utilities can identify many locations satisfying the above requirements, such
as pumping stations, tanks, valve stations, or other utility-owned infrastructure. Many additional
locations may meet the above requirements for sensor locations or could be easily and
inexpensively adapted. Other utility services, such as sewage systems, own sites that likely meet
93
-------
most of the requirements for sensor locations (e.g., collection stations, wastewater treatment
facilities, etc.). In addition, many publicly-owned sites could be easily adapted, such as fire and
police stations, schools, city and/or county buildings, etc. Finally, many consumer service
connections would also meet many of the requirements for sensor placement, although there may
be difficulties in securing access to private homes or businesses. Nevertheless, the benefit of
using these locations may be worth the added cost. Compliance monitoring locations may also
be feasible sites.
The longer the list of potential sensor sites, the more likely one is to design a high-performing
CWS. Typically, the TEVA-SPOT analysis will consider two or three subsets of nodes. For
example, the set of all nodes in the model, the set of all publicly-owned facilities, and the set of
all water utility owned facilities. The utility should develop a list of the (EPANET) node
numbers associated with the facilities that they would like to have considered as potential sensor
locations. Multiple lists indicating different subsets (such as water utility owned, publicly
owned, etc.) may also be specified.
The user can designate sensors as being either eligible for placement by the algorithm, ineligible
for placement, or required to be placed; the latter category signifies sensors that already exist. A
location categories file is a text file that can be created with any text editor. The file can be
named as the user chooses; TEVA does not require the filename to have a particular extension.
Each line of the file consists of a keyword followed either by the literal ALL or by a blank-
separated list of ids of nodes in the water network. If ALL is specified, the keyword applies to all
nodes in the network; otherwise it applies only to the nodes whose ids are given. The ALL
keyword must be in upper case, as shown, otherwise it is taken as the id of a specific node.
There are four keywords (feasible, infeasible, fixed, and unfixed). At all times
during the reading and processing of a location categories file, each node of the network has a
feasibility property (specified by the feasible or infeasible keyword) and a forceability
property (specified by the fixed or unfixed keyword). The meaning of the keywords is
given in Table 5.
Table 5: Meaning of the Keywords in Location Categories Files
Property
Feasibility
Forceability
Keyword
feasible
infeasible
fixed
unfixed
Meaning
Node is permitted to have a sensor
Node is not permitted to have a sensor
Node is required to have a sensor
Node is not required to have a sensor
Initially, all nodes have the properties feasible and unfixed. The lines of the file are
processed sequentially. Each line causes the specified nodes (or all nodes, if ALL is specified) to
acquire values for both properties, as shown in Table 6. The table also shows how the SP
algorithm treats the specified nodes if the values of their properties are not subsequently
changed.
Table 6: Effect of Keyword in LC File, and Treatment of Nodes by SP Algorithm
94
-------
Specified
Keyword
feasible
unfixed
inf easible
fixed
Values Conferred on Specified Nodes
Feasibility Property
feasible
•/
S
inf easible
>/
Forceability Property
fixed
S
unfixed
•/
S
Eligibility for Sensor
Placement When Values
Are Final
Eligible
Ineligible
Forced
Population
In TEVA-SPOT analyses, the most commonly used performance measure for sensor placement
is the number of persons made ill following exposure to contaminated drinking water. In order
to estimate health impacts, information is needed on the population at each node. There are a
variety of methods that can be used to estimate nodal population.
• The simplest method is to assume a constant relationship between demand assigned to a
node and the population for that node. This method is most appropriate for a homogeneous,
largely residential area. If a water demand-based population method is to be used, the total
population calculated by the model using a per capita water usage rate needs to be verified with
the population served considering billing records.
• Alternatively, if the number of service connections and types of service connections (i.e.
residential, commercial, industrial, etc.) are known for each node, then this information can be
used to estimate population.
• A third alternative involves independent determination of population based on census
data. If the use of a census-based population is desired, the population associated with each non-
zero demand node of the model needs to be determined using the census data and GIS software.
The resulting total population needs to be verified with the population served by the water
system.
If the second or third method is used to estimate nodal population, a population data file needs to
be created with any text editor. The file can be named as the user chooses; TEVA-SPOT does
not require the filename to have a particular extension. Each line of the file consists of two text
fields separated by one or more blanks. The contents of the fields are defined in Table 7 below.
Table 7: Fields of Population Data Files
Field
Description
The name of a node in the water network; the node should be a
junction, tank, or reservoir (i.e., a point-like component of the
network).
The population at the node given by .
Note: Be sure there are no additional hard returns at the beginning or end of the text file. Also,
be sure the file name does not have any spaces in it.
For more information about applying the TEVA-SPOT methodology, see Murray et al. (2007) or
visit the EPA website http://www.epa.gov/nhsrc/water/teva.html.
95
-------
APPENDIX B: EXAMPLE MSX FILE
[TITLE]
Hypothetical biological contaminant with a 3-component broth Model
[OPTIONS]
AREAJJNITS M2
RATEJJNITS SEC
SOLVER ROS2
COMPILER VC ; alternative is none
TIMESTEP 60
ATOL0.0010
RTOL0.0010
[SPECIES]
BULK BC MG
BULK CL MG
BULK T IMG
BULK T2 MG
BULK T3 MG
[COEFFICIENTS]
CONSTANT kl 300
CONSTANT k2 20
CONSTANT k3 0.01
CONSTANT rl 0.05
CONSTANT r2 0.01
CONSTANT r3 0.01
CONSTANT r4 1000.0
CONSTANT k4 86E-6
[PIPES]
RATE BC -k4*r4*BC*CL
RATE CL -kl*Tl*CL-k2*T2*CL-k3*T3*CL-k4*BC*CL
RATETl-kl*rl*Tl*CL
RATE T2 -k2*r2*T2*CL
RATE T3 -k3*r3*T3*CL
[SOURCES]
SETPOINT { SPECIFY NODE FOR Cl- BOOSTER} CL 1.0
SETPOINT { SPECIFY NODE FOR Cl- BOOSTER} CL 1.0
SETPOINT { SPECIFY NODE FOR Cl- BOOSTER} CL 1.0
[QUALITY]
GLOBAL CL 1.0
[REPORT]
NODES ALL
SPECIES BC YES 5
SPECIES CL YES 5
SPECIES Tl YES 5
SPECIES T2 YES 5
SPECIES T3 YES 5
PAGESIZE 0
96
-------
Additional Reading
American Society of Civil Engineers (ASCE), 2004. Interim Voluntary Guidelines for Designing
an Online Contaminant Monitoring System.
Berry, J. et al, 2005. Sensor Placement in Municipal Water Networks. Jour. Wat. Res. Plan.
Manag. 131 (3): 237-243.
Berry, J. et al, 2006. Sensor Placement in Municipal Networks with Temporal Integer
Programming Models. Jour. Wat. Res. Plan. Manag. 132(4): 218-224.
Bristow, E. and Brumbelow, K., 2006. Delay between sensing and response in water
contamination events. Jour. Infrastructure Systems, p. 87-95.
Cook, J. et al, 2005. Decision Support System for Water Distribution System Monitoring for
Homeland Security. Proceedings of the AWWA Water Security Congress, Oklahoma City.
Davis, M. J., and Janke, R. (2007a). "Development of a Probabilistic Timing Model for the
Ingestion of Tap Water," J. Water Resour. Planning Manage., 135(5), 397-405.
Davis, M. J., and Janke, R. (2008). "Importance of exposure model in estimating impacts when a
water distribution system is contaminated." J. Water Resour. Planning Manage., 134(5), 449-
456.
Davis, M. J., and Janke, R. (2010a). "Patterns in potential impacts associated with contamination
events in water distribution systems." J. Water Resour. Planning Manage., (16 March 2010),
10.1061/(ASCE)WR. 1943-5452.0000084.
Davis, M. J., and Janke, R. (2010b). "Assessing Potential Impacts Associated with
Contamination Events in Water Distribution Systems: A Sensitivity Analysis," United States
Environmental Protection Agency, draft report, EPA/600/R-10/061. DRAFT NOT YET
CITABLE.
Hall, J. et al, 2007. On-line Water Quality Parameters as Indicators of Distribution System
Contamination. Jour. AWWA, 99:1:66-77.
McKenna, S. A., Wilson, M., and Klise, K. A. (2008). "Detecting changes in water quality data."
Journal American Water Works Association, 100(1), 74-85.
97
-------
Murray, R., Janke, R., Hart, W. E., Berry, J. W., Taxon, T., and Uber, J. (2008b). "Sensor
network design of contamination warning systems: a decision framework." Journal American
Water Works Association, 100(11), 97-109.
Murray, R., Haxton, T., Janke, R., Hart, W. E., Berry, J. W., and Phillips, C. A. (2010). Sensor
Network Design for Contamination Warning Systems: A Compendium of Research Results and
Case Studies Using the TEVA-SPOT Software. U. S. Environmental Protection Agency, Office
of Research and Development, National Homeland Security Research Center, Cincinnati OH.
EPA/600/R-09/141.
Skadsen, J., Janke, R., Grayman, W., Samuels, W., Tenbroek, M, Steglitz, B., and Bahl, S. 2008.
"Distribution system on-line monitoring for detecting contamination and water quality changes."
Jour.AWWA, 100:7:81-94.
U. S. Environmental Protection Agency, 2004. Response Protocol Toolbox: Planning for and
Responding to Drinking Water Contamination Threats and Incidents.
http://www.epa.gov/safewater/watersecuri ty/pubs/rptb_response_guidelines.pdf
U. S. Environmental Protection Agency, 2005a. Review of State-of-the-Art Early Warning
Systems for Drinking Water.
U. S. Environmental Protection Agency, 2005b. WaterSentinel System Architecture.
http://epa.gov/watersecuri ty/pubs/watersentinel_system_architecture.pdf
U. S. Environmental Protection Agency, 2005c. WaterSentinel Online Water Quality Monitoring
as an Indicator of Contamination.
http://epa.gov/watersecuri ty/pubs/watersentinel_wq_monitoring.pdf
U. S. Environmental Protection Agency, 2006. Initial Distribution System Evaluation Guidance
Manual for the Final Stage 2 Disinfectants and Disinfection Byproducts Rule. EPA 815-B-06-
002. http://www.epa.gov/safewater/disinfection/stage2/pdfs/guide idse full.pdf
98
------- |