&EPA
United States
Environmental Protection
Agency
RESEARCH REPORT
Technologies and Techniques for Early Warning
Systems to Monitor and Evaluate Drinking
Water Quality: A State-of-the-Art Review
Office of Research and Development
National Homeland Security
Research Center
-------
-------
EPA/600/R-05/156
Technologies and Techniques for Early Warning Systems to Monitor and
Evaluate Drinking Water Quality: A State-of-the-Art Review
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology
Health and Ecological Criteria Division
August 25, 2005
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
DISCLAIMER
This document has been reviewed in accordance with U. S Environmental Protection Agency policy
and approved for publication and distribution. The research described here was managed under EPA
Contract 68-C-02-009 to ICF Consulting, Inc.
Mention of commercial products, trade names, or services in this document or in the references
and/or endnotes cited in this document does not convey, and should not be interpreted as conveying
official EPA approval, endorsement, or recommendation.
Questions concerning this document or its application should be addressed to:
Jafrul Hasan, Ph.D.
USEPA Headquarters
Office of Science and Technology, Office of Water
1200 Pennsylvania Avenue, NW
Mail Code: 4304T
Washington, DC 20460
Phone: 202-566-1322
Email: hasan.jafrul@epa.gov
August 2005
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
ACKNOWLEDGMENTS
The research described here was funded by the National Homeland Security Research Center, Office
of Research and Development, and managed by the Office of Science and Technology, Office of
Water.
EPA gratefully acknowledges the contributions of the following persons and organizations to the
development of the main Report Technologies and Techniques for Early Warning Systems to
Monitor and Evaluate Drinking Water Quality: A State-of-the-Art Review.
EPA Work Assignment Manager: Jafrul Hasan
ICF Work Assignment Manager: David Goldbloom-Helzner
Co-Work Assignment Manager: Audrey Ichida
ICF staff: Tina Rouse and Mark Gibson
Subject matter experts consulted: Stanley States, Walter Grayman, and Rolf Deininger
Internal (EPA) Reviewers:
Jonathan Herrmann Office of Research and Development/National Homeland Security Research
Center
Irwin Silverstein Office of Water/Water Security Division and Office of Research and
Development/National Homeland Security Research Center
John Hall Office of Research and Development/National Homeland Security Research
Center
Roy Haught Office of Research and Development/National Risk Management Research
Laboratory
Robert Janke Office of Research and Development/National Homeland Security Research
Center
Alan Lindquist Office of Research and Development/National Homeland Security Research
Center
Matthew Magnuson Office of Research and Development/National Homeland Security Research
Center
Regan Murray Office of Research and Development/National Homeland Security Research
Center
Grace Robiou Office of Water/Water Security Division
Cesar Cordero Office of Water/Office of Science and Technology
Jafrul Hasan Office of Water/Office of Science and Technology
August 2005 111
-------
U.S. EPA Office of Water
Early Warning Systems
External (non-EPA) Reviewers:
Ronald J. Baker
Frank Blaha
Erica Brown
Bill Clark
Ricardo DeLeon
Wayne Einfeld
Lee Glascoe
Kevin Morley
My-Linch Nguyen
Irwin Pikus
Connie Schreppel
Alan Roberson
U.S. Geological Survey
American Water Works Association Research Foundation
Association of Metropolitan Water Agencies
Association of Metropolitan Water Agencies
Metropolitan Water District of Southern California
Sandia National Laboratories
Lawrence Livermore National Laboratory
American Water Works Association
American Water Works Association Research Foundation
University of Virginia
Mohawk Valley Water Authority
American Water Works Association
August 2005
IV
-------
U.S. EPA Office of Water Early Warning Systems
TABLE OF CONTENTS
DISCLAIMER i
ACKNOWLEDGMENTS iii
LIST OF EXHIBITS xi
ACRONYMS and ABBREVIATIONS xiii
1. Report Highlights 1
1.1 Desired Characteristics of Integrated Early Warning Systems (EWSs) 2
1.2 Conclusions and Recommendations 3
1.2.1 General Conclusions and Recommendations 3
1.2.2 Specific Conclusions and Recommendations 4
2. Introduction 9
2.1 Concern for the Water Supply 9
2.2 Role of EPA in Water Security and Early Warning Systems 9
2.2.1 National Homeland Security Research Center Research 10
2.2.2 Environmental Technology Verification (ETV) Program 11
2.2.3 Technology Testing and Evaluation Program (TTEP) 11
2.2.4 American Society of Civil Engineers (ASCE) Guidelines for Designing
an Online Contamination Monitoring System 11
2.2.5 Water Utility Users Group 11
2.2.6 Water Security Working Group (WSWG) 12
2.2.7 Water Contaminant Information Tool (WCIT) 12
2.2.8 Distribution System Research Consortium (DSRC) 12
2.2.9 EPA Resources/Guidance 12
2.3 Purpose of this State-of-the-Art Review 13
2.4 Approach for this State-of-the-Art Review 16
2.5 Sources of Information 16
2.5.1 Experts 16
2.5.2 Conferences and Seminars 17
2.5.3 Published Literature 17
2.5.4 Website Resources 17
2.5.5 Workgroup Efforts 17
2.6 Criteria for Selecting Products/Technologies 18
3. Desired Characteristics and Features of Integrated Early Warning Systems 19
3.1 Desired Characteristics for Integrated Early Warning Systems 19
3.1.1 Rapid Response Time 19
3.1.2 Range of Contaminants 20
3.1.3 Automation and Remote Operation 21
3.1.4 Affordable Cost 23
3.1.5 Low Skill and Training 23
August 2005 V
-------
U.S. EPA Office of Water Early Warning Systems
3.1.6 Source of the Contaminant 23
3.1.7 Sensitivity 23
3.1.8 Minimal False-positives/False-negatives 24
3.1.9 Robustness and Continuous Functioning 25
3.1.10 Third Party Verification 25
3.2 Design Features of Integrated Early Warning Systems 25
3.2.1 Evaluate the Need for an Integrated EWS 28
3.2.2 Establish EWS Plan and Coordination 29
3.2.3 Determine Overall Approach to EWS Design 30
3.2.4 Develop Detailed Design of an Integrated Early Warning System 32
4. Features of an EWS Related to Data Acquisition and Analysis; Contaminant Flow;
Sensor Placement; Alerts; Data Security; and Communication, Response, and
Decision Making 37
4.1 Real-Time Data Acquisition and Data Analysis 37
4.1.1 Basics 37
4.1.2 Demonstration Projects, Tests, and Products 40
4.2 Contaminant Flow Predictive Systems 41
4.2.1 State-of-the-Art Systems and Current Research and Development 42
4.3 Sensor Placement 45
4.3.1 Current Research and Development 45
4.3.2 State-of-the-Art Systems 46
4.4 Alert Management Systems 47
4.5 Integrating Water Distribution Modeling and Data Acquisition Systems 47
4.6 Data Security 48
4.7 Communications, Response, and Decision Making 49
5. Multi-Parameter Water Quality Monitors as Candidates for Early Warning
Systems 53
5.1 Descriptions of Various Multi-Parameter Water Quality Monitors 53
5.2 Efforts to Determine Performance of Multi-Parameter Water Quality
Monitors and Establish Water Quality Baselines 56
5.2.1 Evaluate Sensor Performance 57
5.2.2 Investigate Baseline for Water Quality 57
5.2.3 Verify Sensor Performance 57
5.3 Efforts to Provide Red Flag EWS and to Identify Specific Contaminants via
Signature Using Multi-Parameter Water Quality Monitors 57
5.3.1 Sensor Response to Contaminants 58
5.3.2 Multi-Parameter Response to Chemical or Biological Agent Simulants 58
5.3.3 Signature Development 58
5.3.4 Further Testing of Signature Concept 58
5.3.5 Multi-Parameter Response to Actual Chemical and Biological Agents 59
6. Technologies that Detect Chemical Contaminants for Early Warning Systems 61
6.1 General Introduction to Assays and Sensors 61
6.2 Available Technology 61
6.2.1 Detection of Arsenic 61
August 2005 VI
-------
U.S. EPA Office of Water Early Warning Systems
6.2.2 Detection of Cyanide 62
6.2.3 Gas Chromatography 63
6.2.4 Enzyme-Based Detection 64
6.2.5 Biosensors 65
6.3 Potentially Adaptable Technology 69
6.3.1 Enzyme-Based Detection 69
6.3.2 Organism-Based Biosensors 70
6.3.3 Infrared Spectroscopy 72
6.3.4 X-ray Fluorescence 73
6.3.5 Ion Mobility Spectroscopy 73
6.3.6 Microchips for Portable Chemical Sensors 74
6.3.7 Microchip Surface Acoustic Wave (SAW) Technology 75
6.3.8 Microchip Chemiresistors 76
6.4 Emerging Technology 76
6.4.1 Organism-Based Biosensors 76
6.4.2 Eukaryotic Cell-Based Biosensors 77
6.4.3 Fiber Optic Cable-Based Sensors 77
6.4.4 Ion Mobility Spectroscopy (IMS) 78
6.4.5 Surface Acoustic Wave (SAW) Technology 78
6.4.6 Raman Spectroscopy 80
7. Technologies that Detect Microbial Contaminants for Early Warning Systems 81
7.1 General Introduction to Assays and Sensors 81
7.2 Available Technologies 81
7.2.1 Immunoassays 81
7.2.2 Detection of Bacterial - ATP 84
7.2.3 Flow Cytometry- and Micro-Flow-Based Technology 85
7.2.4 Bioparticulate Monitors - Light Scattering Technology 86
7.3 Potentially Adaptable Technology 88
7.3.1 Fiber Optic-Based Biosensor 88
7.3.2 Dye-Loaded Microspheres 89
7.3.3 Detection of ATP 89
7.3.4 Cell-Based Biosensor 90
7.3.5 Polymerase Chain Reaction 90
7.3.6 Bio-Optoelectronic Sensor Systems (BOSS) 94
7.3.7 Surface Plasmon Resonance (SPR) 94
7.3.8 Electrochemiluminescence (ECL) 95
7.4 Emerging Technology 96
7.4.1 Lateral Flow Assay 96
7.4.2 Labels 97
7.4.3 Magnetic Beads 97
7.4.4 Flow-Through Columns 98
7.4.5 Raman Spectroscopy 98
7.4.6 Microelectrode Arrays 99
7.4.7 Microarray of Gel-Immobilized Compounds 99
7.4.8 Magnetic Microbeads 99
7.4.9 DNA Microarrays 100
August 2005 Vll
-------
U.S. EPA Office of Water Early Warning Systems
7.4.10 Micro-Cantilever System 100
7.4.11 Photoluminescent Biochips 100
7.4.12 Polymer Microbeads - Taste Chip 101
7.4.13 Molecularly Imprinted Polymers 101
7.4.14 Magnetoelastic Sensors 102
7.5 Concentration Methods 102
7.5.1 Hollow Fiber Ultrafiltration 102
7.5.2 Hydroxyapatite Whole Cell Capture 103
7.5.3 Lectin and Carbohydrate Affinity 103
8. Technologies that Detect Radiological Contaminants for Early Warning Systems .. 105
8.1 General Introduction to Methods of Detection 105
8.2 Available Technology 106
8.3 Potentially Adaptable Technology 107
8.4 Emerging Technology 108
9. Technical Evaluation of Early Warning Systems Ill
9.1 Approach for Technical Evaluation Ill
9.1.1 Verification Studies Ill
9.1.2 Degree of Government Involvement, Support, and Development for
Technology 112
9.1.3 Field Experience and Case Studies 112
9.1.4 Other Studies 112
9.1.5 Expert Opinions 113
9.2 Evaluation of Various Operational Features of EWS 113
9.2.1 Issues and Gaps 113
9.2.2 Conclusions and Recommendations 114
9.3 Evaluation of Multi-Parameter Water Quality Monitors 115
9.3.1 Issue and Gaps 117
9.3.2 Conclusion and Recommendations 122
9.4 Evaluation of Chemical Sensors 124
9.4.1 Issues and Gaps 132
9.4.2 Conclusion and Recommendations 133
9.5 Evaluation of Microbial Sensors 135
9.5.1 Issues and Gaps 142
9.5.2 Conclusion and Recommendations 143
9.6 Evaluation of Radiological Sensors 144
9.6.1 Issues and Gaps 145
9.6.2 Conclusions and Recommendations 147
10. Conclusions and Recommendations 149
10.1 General Conclusions and Recommendations 149
10.2 Specific Conclusions and Recommendations 151
10.2.1 Data Acquisition and Analysis 151
10.2.2 Flow Modeling 152
10.2.3 Sensor Placement 153
10.2.4 Alert Management 153
August 2005 Vlll
-------
U.S. EPA Office of Water Early Warning Systems
10.2.5 Decision Making and Response 154
10.2.6 Multi-Parameter Water Quality Technologies 154
10.2.7 Detection of Chemical Contaminants 155
10.2.8 Detection of Microbial Contaminants 156
10.2.9 Detection of Radiological Contaminants 158
REFERENCES 159
APPENDIX A A-l
APPENDIX B B-l
APPENDIX C C-l
APPENDIX D D-l
ENDNOTES E-l
August 2005 IX
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
LIST OF EXHIBITS
Exhibit 3-1. Drinking Water Contaminant Classes and Examples 22
Exhibit 3-2. Design Features of an Integrated Early Warning System 27
Exhibit 3-3. Conceptual Process for Designing an Early Warning System 28
Exhibit 3-4. Sample EWS Approach using First-Stage and Second-Stage Monitors 31
Exhibit 4-1. Examples of Water Distribution System Modeling Software 42
Exhibit 7-1. Lateral Flow Assay 83
Exhibit 9-1. Evaluation of Water Quality Parameter Monitors 118
Exhibit 9-2. Water Quality Monitors Comparison with Desired EWS Characteristics 120
Exhibit 9-3. Water Quality Monitors as EWS 123
Exhibit 9-4. Chemical Sensors Comparison with Desired EWS Characteristics 125
Exhibit 9-6. Microbial Sensors Comparison with Desired EWS Characteristics 136
Exhibit 9-7. Microbial Sensor Technologies and Techniques 144
Exhibit 9-8. Radiological Sensors Comparison with Desired EWS Characteristics 146
Exhibit 9-9. Radiation Sensor Technologies 147
August 2005 XI
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water
Early Warning Systems
A
AFD
AK
AMS
AOAC
APDS
ASCE
ASTM
ASV
ATP
ATR
AWWA
AwwaRF
BADD
BARC
BCIP
BEADS
BOSS
BTA
CAD
CBR
CBRTA
CBS
CBW
CCD
CDC
CFD
cfu
Ci
CIS
COD
cpm
CRADA
CWS
DARPA
DHS
DNA
DO
DOD
DOE
DSRC
DSS
ECBC
BCD
ECL
ACRONYMS and ABBREVIATIONS
Angstrom
Automated Food Device
Adenylate Kinase
Advanced Monitoring Systems Center
AOAC International (historically Association of Official Analytical Chemists)
Autonomous Pathogen Detection System
American Society of Civil Engineers
American Society for Testing and Materials
Anodic Stripping Voltammetry
Adenosine Triphosphate
Attenuated Total Reflection
American Water Works Association
American Water Works Association Research Foundation
Bio Warfare Agent Detection Devices
Bead ARray Counter
5-Bromo-4-Chloro-3-Indolyl Phosphate Disodium Salt Hydrate
Biodetection Enabling Analyte Delivery System
Bio-optoelectronic Sensor Systems
Bio Threat Alert®
Computer-Aided Drafting
Chemical, Biological, and Radiological
Chemical, Biological, and Radiological Technology Alliance
Case-Based Systems
Chemical Biological Warfare
Charge-Coupled Device
U.S. Centers for Disease Control and Prevention
Computational Fluid Dynamics
Colony Forming Unit
Curies
Customer Information Systems
Chemical Oxygen Demand
Counts Per Minute
Cooperative Research and Development Agreement
Contamination Warning System
Defense Advanced Research Projects Agency
Department of Homeland Security
Deoxyribonucleic Acid
Dissolved Oxygen
Department of Defense
Department of Energy
Distribution System Research Consortium
Distribution System Simulator
Edgewood Chemical Biological Center
Electrolytic Conductivity Detector
Electrochemiluminescence
August 2005
Xlll
-------
U.S. EPA Office of Water
Early Warning Systems
EDS
ELFA
ELISA
ELOD
EMPACT
EOC
EPA
EPS
ETV
EWS
FBI
FDA
FID
FPW
FT-IR
GC
GC-MS
GE
GIS
GMR
HA
HANAA
HRP
HSPD
I-CORE
ICS
ICWater
IDSE
ILSI
IMS
INL
ISAC
ISE
JBAIDS
LAN
LEMS
LIMS
LLNL
LRAD
LRN
MAGIChip™
MALS
MALLS
MCL
MEMS
MIP
MIT
Event Detection Software
Enzyme-Linked Fluorescent Immunoassay
Enzyme-linked Immunosorbent Assay
Estimated Limit of Detection
Environmental Monitoring for Public Access and Community Tracking
Emergency Operations Center
U.S. Environmental Protection Agency
Extended-Period Simulation
Environmental Technology Verification
Early Warning System
Federal Bureau of Investigation
U.S. Food and Drug Administration
Flame lonization Detector
Flexural Plate Wave
Fourier Transform Infrared
Gas Chromatography
Gas Chromatograph-Mass Spectrometer
Genomic Equivalent
Geographic Information Systems
Giant Magnetoresistive
Hydroxyapatite
Hand-Held Nucleic Acid Analyzer
Horseradish Peroxidase
Homeland Security Presidential Directive
Integrated Cooling/Heating Optics Reaction
Incident Command System
Incident Commanders Water Modeling Tool
Initial Distribution System Evaluation
International Life Sciences Institute
Ion Mobility Spectroscopy
Idaho National Laboratory
Information Sharing and Analysis Center
Ion Selective Electrode
Joint Biological Agent Identification and Diagnostic System
Local Area Network
Liquid Effluent Monitoring System
Laboratory Information Management System
Lawrence Livermore National Laboratory
Long Range Alpha Detection
Laboratory Response Network
Micro Array of Gel-Immobilized Compounds
Multi-Angle Light Scattering
Multi-Angle Laser Light Scattering
Maximum Contaminant Level
Micro-Electro-Mechanical System
Molecularly Imprinted Polymers
Massachusetts Institute of Technology
August 2005
XIV
-------
U.S. EPA Office of Water
Early Warning Systems
MOEMS Micro Optical Electro Mechanical Systems
MS Mass Spectroscopy
MW Molecular Weight
Nal Sodium Iodide
NALOD Nucleic Acid Limit of Detection
NASA National Aeronautics and Space Administration
NDWAC National Drinking Water Advisory Council
NHSRC National Homeland Security Research Center
NNI National Nanotechnology Initiative
NRMRL National Risk Management Research Laboratory
NSF NSF International (historically National Sanitation Foundation)
NTA National Technology Alliance
OGWDW Office of Ground Water and Drinking Water/U. S. EPA
OHS Office of Homeland Security/U. S. EPA
OLM On-Line Liquid Monitoring System
ORD Office of Research and Development/U. S. EPA
ORNL Oak Ridge National Laboratory
ORP Oxidation-Reduction Potential
OW Office of Water/U.S. EPA
PCR Polymerase Chain Reaction
PDD Presidential Decision Directive
PEC Photosynthetic Enzyme Complex
pfu Plaque Forming Unit
pfu-e Plaque Forming Unit Equivalent
PID Photo lonization Detector
PNNL Pacific Northwest National Laboratories
ppb Parts Per Billion
ppm Parts Per Million
ppt Parts Per Trillion
psi Pounds Per Square Inch
QA/QC Quality Assurance and Quality Control
QLFA Quantitative Lateral Flow Assay
R&D Research and Development
RAD ACS Radiological Assessment Display and Control Software
RAPID Ruggedized Advanced Pathogen Identification Device
RBS Rule-Based Systems
RLU Relative Light Unit
RNA Ribonucleic Acid
ROC Receiver Operating Characteristic
SAIC Science Applications International Corporation
SAW Surface Acoustic Wave
SBIR Small Business Innovation Research
SCADA Supervisory Control and Data Acquisition
SDWA Safe Drinking Water Act
SERS Surface-Enhanced Raman Scattering
SIA Sequential Injection Analysis
SMART™ Sensitive Membrane Antigen Rapid Test
August 2005
XV
-------
U.S. EPA Office of Water
Early Warning Systems
SMP Submitochondrial Particles
SNL Sandia National Laboratories
SOPs Standard Operating Procedures
SPCE Surface Plasmon-Coupled Emission
SPME Solid Phase Microextraction
SPR Surface Plasmon Resonance
SSL Secure Socket Layer
TAM Thermo Alpha Monitor
TCD Thermal Conductivity Detector
TCR Total Coliform Rule
T&E Test and Evaluation
TEVA Threat Ensemble Vulnerability Assessment
TIGER Triangulation Identification Genetic Evaluation of Risks
T&O Taste and Odor
TOC Total Organic Carbon
TRA Technology Readiness Assessment
TTEP Technology Testing and Evaluation Program
UC Ultrafiltration Concentration
UHF Ultra High Frequency
UPT Upconverting Phosphor Technology™
URL Uniform Resource Locator (also known as website address)
USACEHR U.S. Army Center for Environmental Health Research
USAMRIID U.S. Army Medical Research Institute of Infectious Diseases
USGS U.S. Geological Survey
UV Ultraviolet Light
VARA Vulnerability and Risk Assessments
VHP Very High Frequency
VOCs Volatile Organic Compounds
WaterlSAC Water Information Sharing and Analysis Center
WATERS Water Assessment Technology Evaluation Research and Security
WCIT Water Contaminant Information Tool
WDM Water Distribution Monitoring
WERF Water Environment Research Foundation
WISE-SC Water Infrastructure Security Enhancements-Standards Committee
WLA Water Laboratory Alliance
WQS Water Quality System
WS-CWS Water Sentinel Contamination Warning System
WSD Water Security Division
WSTB Water Science and Technology Board
WSWG Water Security Working Group
WUERM Water Utility Emergency Response Manager
August 2005
XVI
-------
U.S. EPA Office of Water Early Warning Systems
1. Report Highlights
Terrorist attacks have heightened concern about intentional threats to the U.S. water system, whether
from physical destruction, computer interference, or chemical, microbial, or radioactive
contamination. Such intentional contamination events can have a profound impact on public health
and confidence in the nation's water infrastructure. An Early Warning System (EWS)a can be an
important tool to avoid or mitigate the impacts of an intentional contamination event in time to allow
an effective local response that reduces or eliminates adverse impacts (ILSI, 1999). An integrated
EWS includes sensors to detect the contaminant; systems to transmit, compile, and analyze data;
links for communication and notification; and protocols for decision making and emergency
response.
The goal of this EWS document is to review the state-of-the-art technologies and techniques for
integrated EWSs for drinking water infrastructure, particularly for finished water supplies and
distribution systems. The report summarizes and evaluates current and emerging EWS technologies
for identifying general categories of chemical, microbial, and radiological contaminants. It also
identifies future directions, technical issues, and research gaps. Information was gathered from a
variety of sources, including company information, government information, verification studies,
field case studies, and expert opinions.
The basis of this project is outlined in the Water Security Research and Technical Support Action
Plan, Section 3.Be,1 which recommends testing and evaluation of drinking water and other EWSs,
focused on distribution systems. This study also supports the implementation of Homeland Security
Presidential Directive (HSPD)-92 even though the study was started before HSPD-9 was issued.
HSPD-9 directs the U.S. Environmental Protection Agency (EPA) to "develop robust,
comprehensive, and fully coordinated surveillance and monitoring systems, ... that provide early
detection and awareness of disease, pest, or poisonous agents." To focus on the most promising
products and technologies in this relatively new and rapidly progressing area, criteria were
developed in this study to select technologies and products that are available now, are potentially
adaptable to EWSs for drinking water systems, or have the potential to emerge as EWSs in the
future.
Early Warning System or EWS is a term used throughout this document. The term is derived from the
use of EWSs in monitoring source water. Detectors have been used to identify changes in source water
quality upstream of a water intake (e.g., in rivers or streams for chemical spills). In its truest sense, an
EWS would provide for an early warning of a contaminant prior to human exposures with public health
impacts (e.g., "detect to warn"). Such may not be the case in the first generation of contaminant warning
in drinking water distribution systems. An alternative term of art has evolved based on threat recognition,
that being a "contamination warning system." A contamination warning system involves the active
deployment and use of monitoring technologies/strategies and enhanced surveillance activities to collect,
integrate, analyze, and communicate information to provide a timely warning of potential water
contamination incidents and initiate response actions to minimize public health and economic impacts.
Based on currently available technologies, a contamination warning system is weighted toward "detect
to treat." However, with advances in technology and the ability to detect specific contaminants or
specific contaminant categories in near real-time, it is anticipated that a contamination warning system
will move toward "detect to warn."
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
In response to HSPD-9, EPA is working with technical experts and stakeholders in the drinking
water community to design a robust and comprehensive drinking water system monitoring program
that would provide early indication of an attack and minimize public health consequences. An
outcome of these efforts is WaterSentinel, which is a proposed demonstration project where EPA,
in partnership with select utilities and laboratories, would design, deploy, and evaluate a model
contamination warning system for drinking water security. A more detailed description of how the
thinking behind the WaterSentinel concept has evolved is included in Appendix A.
This document is a state-of-the-art review and reflects knowledge and information gathered through
May 2005. The fast pace of technology and technique development and the implementation of
EWSs for drinking water indicate that this document should be viewed as a snapshot in time. It will
be useful to utilities as they plan activities that provide early warning against potential threats and
attacks to their water systems. As the state-of-the-art advances, periodic updates of this document
will be necessary.
1.1 Desired Characteristics of Integrated Early Warning Systems (EWSs)
An EWS is an integrated system for monitoring, analyzing, interpreting, and communicating
monitoring data, which can then be used to make decisions that are protective of public health and
minimize unnecessary concern and inconvenience to the public. To become a widely used,
effective, and reliable part of a water distribution security and quality monitoring system, an ideal
integrated EWS should demonstrate a number of characteristics, such as the following:
• provide a rapid response
• include a sufficiently wide range of potential contaminants that can be detected
• exhibit a significant degree of automation, including automatic sample archiving
• allows acquisition, maintenance, and upgrades at an affordable cost
• require low skill and training
• identify the source of the contaminant and allows accurate prediction of the location and
concentration downstream of the detection point
• demonstrate sufficient sensitivity to detect contaminants
• permit minimal false-positives/false-negatives
• exhibit robustness and ruggedness in continually operating in a water environment
• allow remote operation and adjustment
• function continuously
• allow for third party testing, evaluation, and verification
When autility considers the development of an integrated EWS, especially for a distribution system,
it should go through a structured process to determine the need for and approach to using an EWS.
Some experts in the field of EWS design for drinking water distribution systems advocate a "tiered"
approach with two stages. The first stage uses continuous real-time sensors that can provide a
generic warning or trigger an alarm when a contaminant is detected in the water. Examples of this
first stage include common multi-parameter online sensors typically used for monitoring water
quality (e.g., pH, conductivity, chlorine residual, etc.). A warning would trigger a second stage
using more specific and sensitive technology to confirm and identify the contaminant (ILSI, 1999;
Hasan et al., 2004). A second-stage technology might be located in the field or could be brought
to the site as a field portable unit. Another design could include first stage warnings in conjunction
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
with data from customer complaints and public health surveillance as triggers for confirmatory
testing. Although real-time continuous monitoring may be the ultimate long-term goal, there may
be intermediate EWS architectures that would be more effective until real-time monitors become
available.
1.2 Conclusions and Recommendations
The conclusions and recommendations are based on a scientific and technical evaluation of EWS
technologies. An expert review of qualitative, semi-qualitative, and quantitative information was
conducted using various sources including verification studies, government research, and expert
opinions.
The following are general conclusions and recommendations from this review of state-of-the-art
technologies and techniques for EWSs. These are followed by specific conclusions and
recommendations organized by the features of the EWS: data acquisition and analysis; flow
modeling; sensor placement; alert management; decision making and response; multi-parameter
water quality technologies; and detection of chemical, microbial, and radiological contaminants.
The recommendations include a list of near term and long-term knowledge and research gaps.
1.2.1 General Conclusions and Recommendations
Viable integrated EWSs that meet the desired characteristics and can be routinely used are several
years away. Some individual components are available currently; however, others need further
development. Designs of EWSs for water distribution systems are largely theoretical or in
preliminary stages. The systems that are in place do not have all the features of an integrated EWS,
as described by this report. Most sensor and EWS components have not been third party tested or
verified, and the types of contaminants and levels of exposure have not been well defined to support
selection of sensor technologies. Utilities will need verification and demonstration studies to
evaluate various manufacturer claims.
Short-Term Research Needs
• An in-depth review of EWS design and implementation should be conducted.
• Methods for vulnerability assessments should be adapted to focus on contamination
scenarios.
• Research on international efforts on EWS is needed.
• Grab-sampling protocols and analysis technologies should be developed quickly.
Long-Term Research Needs
• Survey case studies and analyses should be performed on monitors/sensors/detectors used
by utilities.
• EWS components need to undergo performance testing.
• Potential contaminants list should be reviewed on an ongoing basis.
• Concentrations that must be detected by a sensor should be reviewed on an ongoing basis.
• The fate and transport (including exposure levels, doses, and detectible concentrations) of
contaminants, especially toxic byproducts, should be examined.
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
• Results from laboratory research by various agencies on EWSs should be replicated and the
conclusions of EWS studies should be shared among government agencies and water utility
stakeholders.
1.2.2 Specific Conclusions and Recommendations
• Data Acquisition and Analysis
Data collection by Supervisory Control and Data Acquisition (SCAD A) or other automated systems
is essential to handle the large volume of data from online sensors in an EWS. Much of the required
data acquisition software and hardware already exists. However, software for security (e.g.,
encryption) of SCAD A systems for EWSs is still under development and needs verification, but can
probably be addressed by utilities simultaneously with general security issues.
Short-Term Research Needs
• Standardized methods/guidance for data analysis and interpretation are needed.
• Large-scale data storage and manipulation techniques are needed.
Long-Term Research Needs
• SCADA data security programs should be developed to link existing utility efforts with
security characteristics necessary for EWSs.
• Flow Modeling
Predicting the movement and flow of contaminants in a water distribution system is important not
only to prepare for a potential intentional contamination event, but also to improve the effectiveness
of the EWS design. Distribution system modeling in general, and contaminant flow predictive
systems in particular, are developing rapidly. Current contaminant flow models can also integrate
data from geographic information systems. Calibration has been increasingly used in distribution
systems. Consumption-of-water models are also being occasionally incorporated. Utility efforts
to validate and develop predictive flow models would meet the dual purposes of general planning
(expansion, upgrades, repairs, maintenance) and testing intentional contamination scenarios. These
models also assist in the general management of water quality in the distribution system. Although
there are currently no established calibration criteria in the U.S., static and dynamic calibration
methods exist (EPA, 2005). Also, a committee of the American Water Works Association (AWWA)
did propose a set of possible calibration guidelines in 1999 (ECAC, 1999). These possible
calibration guidelines should serve as a catalyst or starting point to move forward on developing
accepted calibration guidelines or standards. Utilities would need further education and guidance
to make use of such predictive flow models, including a better understanding on how they are
calibrated. EPA's Threat Ensemble Vulnerability Assessment (TEVA) program incorporates flow
modeling (e.g., EPANET) in a probabilistic framework to evaluate contamination events.
Short-Term Research Need
• Improved contaminant flow models are needed.
Long-Term Research Need
• Flow models need to be verified and then used to improve EWS design.
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
• Sensor Placement
Because of both budgetary and technical constraints, utilities can make only a modest initial
investment in sensors within their distribution system, and therefore often want to determine the
most appropriate locations. Without resorting to sophisticated experimental optimization
techniques, the limited numbers of utilities that are beginning to design EWSs and put sensors in
place, usually determine sensor placement based first on logistical constraints (e.g., available power
source, access to communications) and second on larger pipes that serve the most customers.
Current research combining flow models and sensor technology is beginning to be developed, but
such models must be verified before difficult and costly decisions are made by the utilities.
Short-Term Research Need
• Hardware and materials to protect remote sensors are needed.
Long-Term Research Need
• Research into sensor placement parameters is recommended.
• Alert Management
Alert management systems typically consist of two general areas: (1) establishing parameters for
alert triggers and (2) reducing false alarms. Any anomalies in the comparison of sensor data to the
baseline trigger alert the operator. Establishing reliable baseline data is key especially when water
quality fluctuates. Alert management systems usually rely on strict data validation protocols or
specialized software to reduce false alarms. Several companies are working on alert management,
but are in preliminary stages of research, with often proprietary trigger algorithms.
Long-Term Research Need
• Alert management approaches/technologies should be examined and sensitivity to false-
positives and false-negatives should be quantified.
• Decision Making and Response
The process linking the analysis of contamination data with decision making and response is
outlined in EPA's Response Protocol Toolbox; however, additional tools to effectively implement
the process are needed for water utilities. Tools to assist decision making and response, such as the
Water Contaminant Information Tool (WCIT), are being developed and will help fill a current void.
Long-Term Research Need
• Technology to support and implement decision making and response strategies are needed.
• Multi-Parameter Water Quality Technologies
There is ongoing research on the use of the multi-parameter water quality monitors as part of an
EWS for distribution systems. Multi-parameter water quality technology includes readily available
water quality sensors that when analyzed together may be able to identify a physical or chemical
change in water quality. Such a change may suggest that a contaminant has been accidentally or
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
intentionally added. Standard water quality parameters include chloride, specific conductance,
turbidity, free chlorine, oxidation-reduction potential (ORP), pH, dissolved oxygen (DO),
temperature, and sometimes total organic carbon (TOC). Parameters that appear, from EPA's
preliminary tests, to be useful for monitoring in distribution systems include chloride (ion selective
electrode, ISE), specific conductance (electrode), turbidity (nephelometric), free chlorine, and ORP
(EPA, 2004). TOC is also helpful, but may be too expensive to be widely used. The multi-
parameter approach has not been sufficiently evaluated to recommend widespread use. For example,
no tests have been performed on chloraminated systems. Also, there is concern about false-
positives. However, full-scale testing by the U.S. Geological Survey (USGS), EPA, and a
participating water utility company over the next 12 to 18 months may help shed light on false-
positives and whether a system can work with the fluctuations of normal water quality.
Several equipment manufacturing companies are attempting to identify contaminants or classes of
contaminants through water quality parameter signatures. It is difficult to independently validate
or replicate the activities of these companies because their methods and algorithms are proprietary,
which indicates that caution should be used regarding these methods at this juncture. Also, the
examination of water quality parameters in detecting and identifying contaminants is still being
evaluated by EPA, USGS, the Army, and other organizations. There has yet to be a field-scale test
of an EWS with these water-parameter components.
Short-Term Research Needs
• Verified baseline data to calibrate EWS alarm triggers are needed.
• Contaminant specific signatures are needed.
• Validation of event detection algorithms is needed.
Long-Term Research Needs
• Costs and benefits of using TOC sensors should be determined. More affordable and reliable
TOC sensors should be developed.
• Detection of Chemical Contaminants
Portable field kits and devices are available for conducting analyses of grab samples for onsite
detection of many possible chemical contaminants. In contrast to portable field technology, online
detection technologies for specific chemical contaminants are not reasonably available or are not
cost-effective. In the next several years, the field should show further development in terms of cost-
effective and reliable devices. A few new technologies (e.g., microchip-based technology) could
revolutionize the chemical detection field for drinking water. Disinfectant residuals normally
present in treated drinking water (in the U.S.) may present a problem for many of the technologies
(e.g., biomonitors) that detect toxins. A few technologies are mature enough to be recommended
as good candidates for EPA Environmental Technology Verification (ETV) studies.
Short Research Need
• The impact and removal of disinfectant residuals on accuracy of detection should be
examined.
Long-Term Research Needs
• Reliable field kits should be developed.
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
• Existing state-of-the-art detection technologies should be adapted for use in EWSs.
• Detection of Microbial Contaminants
Development of an online microbial detection technology appears to be years away. The light-
scattering methodology shows some promise, but most methods are not suited for continuous online
monitoring or for differentiating among microbes. However, to confirm the presence, viability, and
concentration of a contaminant using a portable field unit with grab samples, there are several
potentially adaptable methods from which to choose, including immunoassay, polymerase chain
reaction (PCR), and adenosine triphosphate (ATP). These methods have not yet been exploited to
their full potential, so will likely continue to be incorporated into new monitoring devices and
systems. Grab sampling could include sampling at scheduled intervals or taking composite
samples (e.g., collecting small volumes of sample continuously over time). In any sampling, the
microbial integrity must be ensured. For drinking water, the challenge for most methods is the issue
of concentrating the sample. A few methods of concentration show promise including the hollow
fiber and the micropump. In general, concentration does not seem to be an insurmountable obstacle
for some methods. One recommended approach was to screen a sample with a generic detector
(e.g., multi-parameter probe or perhaps light scatter), then use an immunoassay device in tandem
with another method for identification. ATP detection kits are promising for detecting microbial
contamination, but current products have not been verified for treated drinking water. In the future,
microchips have great potential for advancing online measurement of contaminants, but presently
the field is not sufficiently mature to provide devices that would meet the needs of drinking water
utilities.
Short-Term Research Needs
• Extraction and concentration technologies need improvement.
• Methods to distinguish concentrated interferants from target contaminants are needed.
• Further development of field-stable reagents is needed.
• ATP detection products should be third party evaluated for EWS application.
Long-Term Research Needs
• Antibodies for unique epitopes (present on threat agents) that show less cross-reactivity
should be developed.
• Research on additional approaches and technologies that can detect emerging, evolving, and
engineered microbes is needed.
• Detection of Radiological Contaminants
There are demonstrated technologies for examining radiation in wastewater, but the transfer or
adaptation to drinking water has not yet taken place. Only a few products claim applicability to
water, some on a grab-sample basis. More products are being developed by a few vendors, but it
is still unclear whether the threat merits use of these expensive products on a real-time basis. The
few items that are commercially available should be verified either by EPA or by a national
laboratory that specializes in radiation. All of the radiological monitoring devices mentioned in this
study usually require specialized expertise for installation, set-up, and routine calibration—even if
they are labeled maintenance-free. Thus, early warning for radiation detection is not currently
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
available for finished water or distribution systems, and the market forces for such development may
not be strong.
Short-Term Research Need
• Beta- and gamma-radioactivity detectors should be developed and verified for drinking water
monitoring applications.
Long-Term Research Needs
• Low-cost, online radioactivity monitors are needed.
• Monitors should be developed that are specifically intended for water distribution systems.
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
2. Introduction
This chapter first presents background on the general concern for the water supply including the
threat of intentional contamination of the water supply and the potential benefits of an EWS. It then
describes the EPA's role in water security and EPA's efforts on EWS. Further, it outlines the
objective of this study, which is to review the current and emerging technologies and techniques for
integrated EWSs for drinking water infrastructure, particularly for finished water supplies and
distribution systems. The chapter identifies current capabilities, future directions, technical issues,
and research gaps. Finally, it presents the approach used to conduct the review.
2.1 Concern for the Water Supply
Terrorist attacks have heightened concern about intentional threats to the U.S. water system, whether
from physical destruction, computer interference, or chemical, microbial, or radioactive
contamination. Such intentional contamination events can have a profound impact on public health
and confidence in the nation's water infrastructure. An EWS can be an important tool to avoid or
mitigate the impacts of an intentional contamination event, reliably identifying a high impact
contamination event in source or finished drinking water in time to allow an effective local response
that reduces or eliminates adverse impacts (ILSI, 1999).
2.2 Role of EPA in Water Security and Early Warning Systems
EPA has a lead role in protecting the water supply and specifically to support efforts to develop
EWSs. This role is outlined in several regulations, national strategies, and presidential directives,
including the following:
• Presidential Decision Directive (PDD) 63, signed May 22,1998, designates EPA as the lead
for security of the national water infrastructure.
• The National Strategy for Homeland Security (July, 2002),3 designates EPA as responsible
for protecting our national water supply.
• The Public Health Security and Bioterrorism Preparedness and Response Act of 2002 (the
Bioterrorism Act), requires that community water systems serving > 3,300 people conduct
vulnerability assessments and prepare emergency response plans. Also, the Act charges
EPA with reviewing current and future methods to prevent, detect, and respond to the
intentional introduction of chemical, biological or radiological contaminants into community
water systems.
• HSPD-7 on Critical Infrastructure Identification, Prioritization, and Protection (HSPD-7),
signed December 17, 2003, also reinforces EPA's role as the sector specific agency for the
water infrastructure.
• HSPD-9, signed on February 4, 2004,4 instructs the federal agencies responsible for
agriculture, food, and water security to "develop robust, comprehensive, and fully
coordinated surveillance and monitoring systems ... that provide early detection and
awareness of disease, pest, or poisonous agents." This study, although initiated before
implementation of HSPD-9, supports the efforts of HSPD-9.
In response to HSPD-9, EPA is working with technical experts and stakeholders in the drinking
water community to design a robust and comprehensive drinking water system monitoring program
August 2005
-------
U.S. EPA Office of Water Early Warning Systems
that would provide early indication of an attack and minimize public health consequences. An
outcome built upon existing efforts is Water Sentinel, which is a demonstration project proposed in
the Fiscal Year 2006 budget where EPA, in partnership with select utilities and laboratories, would
design, deploy, and evaluate a model contamination warning system for drinking water security.
A contamination warning system involves the active deployment and use of monitoring
technologies/strategies and enhanced surveillance activities to collect, integrate, analyze, and
communicate information to provide a timely warning of potential water contamination incidents
and initiate response actions to minimize public health and economic impacts.
Although EPA is continually refining its conceptual design for the program, WaterSentinel would
adopt a four-fold approach to detecting contamination involving, first, the monitoring of water
quality parameters; second, the direct monitoring and laboratory analysis of high priority chemical,
biological, and radiological contaminants; third, the integration of water system data with existing
public health surveillance systems; and fourth, active surveillance of customer complaints. In
addition to other critical sources of information, such as intelligence threat analysis and reports from
local law enforcement, WaterSentinel would harness and leverage an array of data streams in
supporting a robust contamination warning system. A more detailed description of how the thinking
behind the WaterSentinel concept has evolved is included in Appendix A.
Since 2001, EPA has formed the Water Security Division (WSD) in the Office of Water (OW) and
the Office of Homeland Security (OHS). EPA has published the Strategic Plan for Homeland
Security. To spearhead the research efforts on homeland security, EPA formed the National
Homeland Security Research Center (NHSRC) in the Office of Research and Development (ORD5).
The Water Security Research and Technical Support Action Plan (Action Plan), prepared by
NHSRC and WSD, emphasizes the need for improving analytical monitoring and detection of
biological, chemical, and radiological threats in drinking water systems as part of securing drinking
water supplies and systems. The NHSRC has been working with the Department of Homeland
Security (DHS), other government agencies, national laboratories, water stakeholders, and the utility
industry to coordinate and conduct research on various issues including EWSs (Appendix B). EPA
has initiated many efforts related to EWSs; they are included below.
2.2.1 National Homeland Security Research Center Research
EPA/ORD' s National Risk Management Research Laboratory (NRMRL) has multiple above-ground
distribution system simulator (DSS) units, located at the Test and Evaluation (T&E) Facility in
Cincinnati, Ohio. The Water Assessment Technology Evaluation Research and Security (WATERS)
Center is located at the T&E Facility and has access to NRMRL's DSS units. EPA has six different
DSS units in operation at the facility. The DSS units are designed and fabricated to evaluate and
understand the dynamics that influence water quality within water distribution infrastructure systems
in the U.S. and abroad. All DSS units are designed and fabricated above ground to permit easy
access to the entire pipe distribution system. The DSS units consist of six individual pipe-loop
circulating units, three single-pass dead-end units, and two decontamination research loops. EPA
is currently conducting research at the WATERS Center to evaluate various sensor and monitoring
technologies, distribution system modeling, disinfection and decontamination, and data acquisition
systems. The sensors and monitoring technologies are currently being evaluated to see how they
respond to accidental and intentional contamination events and threats in a distribution system
(Haught and Goodrich, EPA, personal communication).
August 2005 10
-------
U.S. EPA Office of Water Early Warning Systems
2.2.2 Environmental Technology Verification (ETV) Program
The ETV Program operates as a public/private partnership through agreements between EPA and
private testing and evaluation organizations. The goal of ETV is to provide credible performance
data for commercial-ready environmental technologies to speed their implementation for the benefit
of vendors, purchasers, permitters, and the public. Of the six ETV centers, three are working in
water security relevant areas: ETV Advanced Monitoring Systems Center (AMS - Battelle), ETV
Drinking Water Systems Center (NSF International), and ETV Water Quality Protection Center
(NSF International6). Some testing areas formerly performed by ETV are being transferred to
NHSRC's Homeland Security Technology Testing and Evaluation Program7 (TTEP).
2.2.3 Technology Testing and Evaluation Program (TTEP)
The TTEP was developed by NHSRC in 2004 to rigorously test technologies against a wide range
of performance characteristics. TTEP's mission is to service the needs of water utility operators,
building and facility managers, emergency responders, consequence managers, and regulators by
providing reliable performance information from a trusted source. The technology categories of
interest include detection, monitoring, treatment, decontamination, computer modeling, and design
tools for protecting water and wastewater infrastructures and the outdoor environment. When
possible, technologies will be tested for their ability to detect chemical, biological, radiological
(CBR), and warfare agents. As an outgrowth of the ETV, many aspects of TTEP are similar to ETV.
However, TTEP does not offer verification statements or endorsements to the vendors and it may
test technologies with or without the involvement of the vendor.
2.2.4 American Society of Civil Engineers (ASCE) Guidelines for Designing an Online
Contamination Monitoring System
Under a cooperative agreement with EPA, the American Society of Civil Engineers/Water
Infrastructure Security Enhancements-Standards Committee (ASCE/WISE-SC) has developed
interim voluntary security guidance that covers the design of online contamination monitoring
systems to detect intentional contamination events (Phase I of the project8). Phase II of the project
was to develop appropriate training materials for use in the water sector. Phase III of the project is
to develop community consensus voluntary best practice standards for designing an EWS.
2.2.5 Water Utility Users Group
The AWWA has convened a group (formerly known as the Water Contaminant Detection Working
Group) of water utilities to offer their opinions and experiences in contaminant detection. This
group has met on several occasions in the past year (2004 to 2005) and representatives from EPA
have been invited to participate in those meetings. Of particular interest to the Users Group is
EPA's TEVA Research Program for Drinking Water Distribution System Security. The TEVA
Research Program is developing software tools, methodologies, and strategies for assessing the
public health consequences of chemical and biological attacks on drinking water, and for designing
and evaluating mitigation and response strategies. Partnering with a small group of AWWA water
utilities, the TEVA Research Program will use network models and water quality data gathered from
the utilities to validate and improve TEVA. An interdisciplinary team of researchers from EPA,
several Department of Energy (DOE) National Laboratories, and universities are collaborating to
August 2005 11
-------
U.S. EPA Office of Water Early Warning Systems
develop these tools. Ultimately, the products of the TEVA Program will be useful for designing
EWSs specific to an individual utility. This information will be aimed at evaluating strategies for
sensor locations in drinking water distribution systems, and for optimizing response and recovery
activities in the aftermath of a contamination event (Murray et al., 2004).
2,2,6 Water Security Working Group (WSWG)
The National Drinking Water Advisory Council (NDWAC) is comprised of members of the general
public, state and local agencies, and private groups concerned with safe drinking water. The
NDWAC advises EPA on the Agency's programs related to drinking water. NDWAC has several
working groups that make recommendations to the full council, which in turn advises EPA on
individual regulations, guidances, and policy matters. One group, the WSWG, has been charged by
EPA to (1) identify, compile, and characterize active and effective security practices and policies
for drinking water and wastewater utilities and provide an approach for considering and adopting
these practices and policies at a utility level; (2) consider mechanisms to provide recognition and
incentives that facilitate a broad and receptive response within the water sector to implement these
active and effective security practices and policies, and make recommendations as appropriate; and
(3) consider mechanisms to measure the extent of implementation of these active and effective
security practices and policies, identify the impediments to their implementation, and make
recommendations as appropriate. The WSWG began meeting in July 2004, and delivered a draft
report on these issues to the NDWAC in June 2005. In June 2005, the NDWAC unanimously
approved and adopted the WSWG's findings unchanged and elevated these findings to the status of
recommendations to the EPA.
2.2.7 Water Contaminant Information Tool (WCIT)
EPA's 2004 Homeland Security Strategy calls for the deployment of WCIT to provide easy access
to key information on priority contaminants, and to develop components of the Tool, including data
on treatability and toxicity levels. In this program, detailed information is to be made available to
utilities during a confirmed contamination event when the utility calls EPA's National Response
Center. The information is then faxed to the utility. The Strategy recommends that the WCIT
should be revised periodically, as new information becomes available.
2.2.8 Distribution System Research Consortium (DSRC)
Formed in June 2003, the DSRC is led by the EPA's NHSRC. The DSRC includes federal
employees from a number of agencies with experience in water infrastructure, the water industry,
EPA Program Offices and Regions, and other interested organizations. The DSRC provides a forum
for information exchange on a diverse set of water distribution system security topics. Topics have
included EWS research (e.g., sensors, field studies, sensor placement) and treatment and
decontamination research.
2.2.9 EPA Resources/Guidance
EPA/O W/Office of Ground Water and Drinking Water (OGWDW) maintains a Water Infrastructure
Security website that includes various guides including the Response Protocol Toolbox (e.g., Site
Characterization and Sampling Guide, Analytical Guide), a Security Product Guide,9 Standardized
August 2005 12
-------
U.S. EPA Office of Water Early Warning Systems
Analytical Methods for Use During Homeland Security Events (on NHSRC website), and a list of
Sensors for Monitoring CBR Contamination.
2.3 Purpose of this State-of-the-Art Review
EWSs are needed for drinking water distribution systems because contamination of drinking water,
especially in distribution systems, is an important concern. Research and development of
technologies applicable to EWSs are rapidly progressing, making up-to-date reviews of the EWS
field challenging.
The basis of this project is outlined in the Water Security Research and Technical Support Action
Plan, Section 3.3e,10 which recommends testing and evaluation of drinking water EWSs and, if
applicable, EWSs from other sectors amenable to application in the water environment, focused on
distribution systems. More specifically, Action Plan Section 3.3e has the following four sequential,
yet interdependent tasks:
1. Conduct a survey to gain improved understanding of EWSs that could be employed in
protecting finished water supplies and distribution systems;
2. Perform pilot-scale testing and evaluation of EWSs that could be used by water utilities to
give an early warning of a contaminant threat or accidental contaminant event;
3. Perform field-scale testing and evaluation of EWSs that could be used by water utilities to
give an early warning of a contaminant threat or accidental contaminant event; and
4. Prepare a handbook on the application of EWSs for drinking water supply and system
protection.
This report is intended to fulfill Task 1 of the Action Plan. It includes a comprehensive review of
current and emerging EWS technologies and techniques, and evaluates their state-of-the-art so as
to identify future directions, research gaps, and technical issues. An EWS for finished water
presents a special set of challenges, compared with an EWS for source water. Finished water
includes various treatment chemicals, including chlorine residual. The finished water also flows
through miles of distribution system pipes, which makes the placement of EWS components
difficult. The other three tasks would be addressed by programs such as WaterSentinel. All four
tasks above are being addressed as part of an overall integrated approach being managed by
EPA/NHSRC. Additional information on technologies available for ground water monitoring can
be found in A Review of Emerging Sensor Technologies for Facilitating Long-Term Ground Water
Monitoring of Volatile Organic Compounds (EPA, 2003).
Information is needed on how specific contaminants (microbial, chemical, radiological) affect the
water quality parameters measured by some currently used online monitoring systems, particularly
with regard to which current technologies will best detect a contamination event. A number of
monitoring technologies and products are available that could potentially serve as an EWS, and a
number of suppliers of conventional monitoring systems have begun to advertise them as water
security monitoring systems. However, in most cases, the performance of these systems has not
been fully or independently evaluated. Without basic performance information (such as detection
limits, sensitivity, selectivity, rates of false-positives and false-negatives), it will be difficult to
interpret monitoring results and derive the information necessary to make appropriate public health
decisions.
August 2005 13
-------
U.S. EPA Office of Water Early Warning Systems
Implementation of a minimally evaluated monitoring technology or EWS may result in a false sense
of security, since there is no assurance that it is capable of meeting EWS requirements. It could also
result in false-positive alarms that undermine the effectiveness of any monitoring program, or false-
negatives that would discredit an EWS. A significant amount of research is underway to adapt
existing and develop new technologies that may be suitable as EWSs for water systems.
As promising technologies continue to be developed and brought into the commercial market, there
is a need for a mechanism to evaluate and verify EWS performance that includes field evaluation
and testing sites. Ideally, such testing should be conducted according to a standard protocol by an
independent third party, and the subject technology should be evaluated against standardized
methods. This would provide water utilities with the data necessary to make informed decisions
regarding the implementation of a specific technology in an EWS. EPA's TTEP can provide this
independent testing. The TTEP process follows strict quality assurance procedures to evaluate
technology performance. Stakeholders are involved in identifying and selecting technologies for
testing, and in developing testing plans and reviewing evaluation reports.11
Design of an EWS for finished water in the drinking water distribution system is not simple, as there
are many issues and water system characteristics that need to be considered. The issues that should
be addressed in the effective design of an EWS include planning and communication, characterizing
the system, determining the target contaminants for the EWS, selecting an appropriate EWS
technology, establishing appropriate alarm levels and monitoring frequency, employing hydraulic
models to optimize the number and placement of sensors, selecting parameters to be monitored, and
conducting data management and analysis.
Three recent projects complement the scope of this report. In the first project, the ASCE-WISE-SC
white papers and guidance, mentioned earlier in this report, seek to provide specific guidance to
utilities about the design and implementation of an online contaminant monitoring system (ASCE,
2004). In contrast, this study reviews the state-of-the-art EWS technologies and techniques and
provides recommendations (e.g., research needs) to further the development of an integrated EWS.
In the second project, the National Technology Alliance (NTA), through the Chemical, Biological,
and Radiological Technology Alliance (CBRTA) released a report entitled Water Monitoring
Equipment for Toxic Contaminants Technology Assessment (Black & Veatch, 2004). The NTA
leverages commercial investment in technology to meet U.S. security and defense needs. The NTA
report focuses on monitoring technologies for both source and finished water. In contrast, this study
focuses on the components of an integrated EWS (e.g., monitoring, data acquisition, flow modeling),
and specifically addresses finished water monitoring for distribution systems. An article by the
AWWA entitled "Contamination Warning Systems for Water: An Approach for Providing
Actionable Information to Decision-Makers" provides an easy to read summary of monitoring
technologies, monitoring locations, data transmission, alarms, and response (Roberson and Morley,
2005).
The purpose of this study was to provide a current state-of-the-art review (i.e., snapshot in time).
In the process, it also identifies areas of research needed in the short-term and long-term to develop
workable EWSs. This study, however, was not able to address all of the issues associated with
EWSs. The following are limitations in the scope of the study. This study does not discuss in detail
the specific types of drinking water contaminants, and instead focuses on three categories of
contaminants: chemical, microbial, and radiological. These categories are useful on a broad level
August 2005 14
-------
U.S. EPA Office of Water Early Warning Systems
to evaluate monitoring technologies. Specific information, such as the concentrations at which a
contaminant is detectible or effective in impacting human health, is important in evaluating a
specific instrument or set of instruments. However, much of the detailed information and research
efforts on the types and concentrations of contaminants is either not publically available or is
undergoing collection and review. Detection limits for the different technologies are provided when
available from the vendor, government agency, or testing/verification entity. Limits of detection
may not be provided for emerging technology because of the newness of the technology.
The design of an EWS is a complex issue, and the design and development of EWSs for water
distribution systems are still at an early stage and continue to evolve. This study summarizes the
basic design and features of an EWS, but does not cover the topic in detail. For example, the study
does not provide detailed criteria for selecting instruments, locating instruments, setting alarm
levels, and integrating other independent data streams such as public health surveillance and
consumer complaints monitoring. This document is not a guidance document, but instead a state-of-
the-art review. As mentioned above, the ASCE has provided some guidance, including a ranking
of the desired characteristics of the EWS (Carlson et al., 2004). Reviewers of early drafts of this
study suggested that a more detailed treatment of design and implementation of an EWS is needed.
To accomplish this, another multi-stage project would be necessary to further develop an adequate
treatment of design possibilities and issues. Such a project should engage the water sector
stakeholders, including utilities, equipment designers and manufacturers, researchers, and policy
officials.
There are many technologies and products covered in this report. EPA does not recommend or
endorse the specific products mentioned in this report. In addition, much of the information
provided on specific technologies are from vendors and have not been independently verified unless
otherwise indicated. Efforts were made to use more than one source to verify the information;
however, EPA is not responsible for errors in company-provided information. The status of the
technology (e.g., pilot, concept), as well as estimates for field implementation are based on vendor
or government authorities, as indicated. It is particularly difficult to accurately predict when
emerging technologies will evolve into reliable and marketable products.
This project's objective is to conduct a comprehensive review of the state-of-the-art technologies
and techniques for integrated EWSs for drinking water infrastructure, particularly for finished water
supplies and distribution systems. It seeks to identify the status of current and emerging EWS
technologies and future directions, research gaps, and technical issues. Chapter 1 provides report
highlights. Chapter 2 introduces the concept of an integrated EWS and discusses the approach used
to conduct this state-of-the-art review. Chapter 3 presents a description of desired characteristics
and component features that define EWSs. Chapter 4 addresses overall design and operation of
EWSs, including data management and analysis, predictive flow modeling, sensor placement, alarm
management, data security, and response communications. Chapter 5 presents the concept of using
general water quality parameters as indicators for contamination events. Details of efforts being
undertaken to develop multi-parameter water quality signatures for specific contaminants are also
presented in Chapter 5. Chapters 6, 7, and 8 cover chemical, microbial, and radiological detection
methods and technologies, respectively. These chapters are organized in a similar format, with a
general description of a technology followed by specific example products that are available or
could be adapted for drinking water, followed by emerging technologies that are more in the
developmental stage. Chapter 9 addresses the evaluation of the technologies discussed in Chapters
August 2005 15
-------
U.S. EPA Office of Water Early Warning Systems
5 through 8. Chapter 10 explains the conclusions and recommendations from the evaluation of EWS
state-of-the-art. Tables are presented as Exhibits and are labeled by chapter number followed in
chronological number as they appear in the text.
2.4 Approach for this State-of-the-Art Review
This review document on E WSs covers technologies and techniques that may be applicable for many
different water monitoring needs, as well as specific EWS needs. The stepwise approach for
reviewing and evaluating EWS components includes several forms of information gathering and
critical review. Information was gathered from published literature, conferences, seminars,
workgroups, and consultation with experts in the field. The review and evaluation protocol
consisted of the following six steps:
Step 1: Identify experts on contamination issues for distributions systems to assist with the
technical writing of this background document, with emphasis on intentional
contamination events.
Step 2: Summarize characteristics and features for EWSs and the justifications for those criteria
and features.
Step 3: Develop an inventory, and descriptions of function and status, of rapid detection
technologies that might be appropriate components of an EWS. Methods are described in
general terms, then specific products are presented. Technologies include water quality
parameter monitors, and detectors for chemical, biological, or radioactive contaminants.
Step 4: Develop an inventory and report on the status of integrated EWS design and operation.
General water quality monitors and specific detection assays combined with data relay,
analysis, and display technologies are all part of EWS design. Where these technologies
are located, how they interact, and how they are made secure and reliable are also integral
parts of EWS design and operation.
Step 5: Identify and discuss research, gaps, information, and technical advancements needed to
build future EWSs.
Step 6: Evaluate technologies for their capabilities and issues.
For the EWS commercial products identified in Steps 3 and 4, information collected includes a
general description of the product, the extent to which it is a comprehensive EWS for real-time
monitoring in a distribution system (e.g., detector only, not notification), method of detection,
contaminants detected, limits of detection, level of verification, potential for piloting, current use,
and other critiques. Complete information was not available for all the technologies and products
discussed. See Appendix C for a list of products and manufacturers mentioned in this report.
2.5 Sources of Information
2.5.1 Experts
Government, industry, academia, and water utility experts in the field were consulted at various
conferences, seminars, workshops, and by telephone and e-mail. Subject matter experts provided
their experience and knowledge on the subject of EWSs (See Acknowledgments). In many cases,
company representatives were contacted directly via e-mail and/or telephone for information
regarding products and the application of their products. In general, information about products or
August 2005 16
-------
U.S. EPA Office of Water Early Warning Systems
programs that was obtained through e-mail or telephone is cited as personal communication. All
correspondence took place between July 2004 and July 2005.
2.5.2 Conferences and Seminars
EWS design and its component technologies are part of the emerging field of water security. As
with all fields experiencing rapid growth, much valuable information is available from conferences
and seminars. Materials from several conferences were compiled and reviewed for building the
content of this state-of-the-art review. Conferences and seminars attended by the primary authors
for the purpose of gathering information for this review, are listed below:
• Water Security Congress, AWWA (April 2004, April 2005)12
• AWWA-sponsored seminar on Contamination Monitoring Technologies (Richmond,
Virginia; May 2004)13
• Water Infrastructure Security Enhancement Workshops Guidelines for Designing an Online
Contaminant Monitoring System for Water Utilities (May 19, 2004)14
• Rapid Detection Technologies for Water Supply Safety and Security (Washington, DC; June
2004)15
2.5.5 Published Literature
Information from peer-reviewed published literature, government agency publications, vendor
literature, and results of verification tests was gathered and critically reviewed for inclusion in this
report. The information sources include the following:
• articles, publications (e.g., Water Environment and Technology)
• reference texts such as Design of Early Warning and Predictive Source Water Monitoring
Systems (e.g., the American Water Works Research Foundation (AwwaRF))
• EPA Performance Verification Testing of commercially available rapid toxicity monitoring
systems
• Studies from other federal agencies working in this area (e.g., Department of Defense
(DOD), National Aeronautics and Space Administration [NASA])
2.5.4 Website Resources
Several hundred websites were evaluated for content applicable to this review. Electronic archives
and hard copies of websites cited were collected for future reference because web content evolves
so rapidly. The URLs are cited as endnotes throughout this report and were current as of April 2005.
2.5.5 Workgroup Efforts
Information from workgroups focused on EWSs was obtained, including EPA workgroups and the
Water Contaminant Detection Working Group.
August 2005 17
-------
U.S. EPA Office of Water Early Warning Systems
2.6 Criteria for Selecting Products/Technologies
The goal of this EWS document is to report on the state-of-the-art technologies and techniques for
detecting contaminants, particularly chemical, microbial, and radiological contaminants in drinking
water distribution systems. To focus on the most promising products and technologies in this
relatively new area, criteria were developed for including technologies and products in this
document. Most technologies and products that were either only conceptual or were not currently
envisioned to apply to water were omitted. A complete discussion of the criteria used to select
technologies, as well as a list of the products and technologies investigated, can be found in
Appendix C.
Three categories of technology development were designated: (1) available now (being used, off-
the-shelf technology, or could be used by water utilities) for use in EWSs; (2) potentially adaptable
technology for use in EWSs (in use, but needs additional steps to address specific challenges for use
with water distribution system); and (3) emerging technologies that may be applicable for use in
EWSs.
In this report, technologies are classified based on the three categories above (e.g., available,
potentially adaptable, emerging), and details are provided on the level of verification, if the
information was available. For most products, except where noted, manufacturer's claims have
not been evaluated by independent sources and products mentioned are not endorsed by EPA.
August 2005 18
-------
U.S. EPA Office of Water Early Warning Systems
3. Desired Characteristics and Features of Integrated Early Warning Systems
An EWS is much more than a collection of monitoring technologies. It is an integrated system for
deploying the monitoring technology; analyzing, interpreting, and communicating the results; and
using the results in making decisions that are protective of public health while minimizing
unnecessary concern and inconvenience within a community. EWSs should be viewed as a critical
part of the operation of a water system in general. EWSs can be used to identify intentional
contamination events and other non-intentional situations where water quality is impaired. To
become a widely used, effective, and reliable part of a water distribution security (and water quality
monitoring) system, there are characteristics that integrated EWSs should exhibit.
3.1 Desired Characteristics for Integrated Early Warning Systems
To become a widely used, effective, and reliable part of a water distribution security and quality
monitoring system, an ideal integrated EWS should demonstrate a number of characteristics, such
as the following (adapted from ILSI, 1999; Grayman, 2004a; Hasan, 2004):
• provide a rapid response
• include a sufficiently wide range of potential contaminants that can be detected
• exhibit a significant degree of automation, including automatic sample archiving
• allow acquisition, maintenance, and upgrades at an affordable cost
• require low skill and training
• identify the source of the contaminant and allows accurate prediction of the location and
concentration downstream of the detection point
• demonstrate sufficient sensitivity to detect contaminants
• permit minimal false-positives/false-negatives
• exhibit robustness and ruggedness to continually operate in a water environment
• allow remote operation and adjustment
• function continuously
• allow for third party testing, evaluation, and verification
Currently, an EWS with all of the above characteristics does not exist. However, there are parts of
an EWS that can meet certain core characteristics: (1) provide rapid response, (2) screen for a
number of contaminants while maintaining sufficient sensitivity, and (3) operate as an automated
system that allows for remote monitoring. Any EWS system that does not demonstrate these three
core characteristics could not be considered an effective EWS. Although emphasis is placed on
these three core characteristics, the other characteristics presented below cannot be ignored in the
design of an EWS. For example, consideration should be given to the rate of false-positive/false-
negative results and method sensitivity when interpreting the results. System operation and
maintenance costs, sampling rate, and reliability should be considered in the design of an EWS.
Furthermore, utilities are reluctant to invest in technologies that have not been third party verified.
3.1.1 Rapid Response Time
Response time for an EWS is typically from the time when the contaminant contacts the sensor to
when a result is reported and a response is initiated (Mays, 2004). An ideal EWS would detect,
interpret, and communicate the warning in sufficient time to take mitigation response actions before
August 2005 19
-------
U.S. EPA Office of Water Early Warning Systems
human health is jeopardized (ILSI, 1999). It is especially desirable that the EWS is rapid between
contamination occurrence and detection and identification of the contaminant. This time may be
influenced both by the technologies used and the overall approach to identification of the
contaminant. For example, an approach could involve an initial warning by one technology
followed by a confirmation by another technology. In most of the current literature, the speed of a
rapid detection technology refers mainly to the time between sample collection and final
interpretation of the results. The ILSI report Early Warning Monitoring to Detect Hazardous Events
in Water Supplies considers results in two hours or less to be rapid (ILSI, 1999). Some
manufacturers call their field portable grab sample kits (or devices) rapid detection technologies.
In these cases it is important to note that even if a technology boasts a 2-minute assay time, if the
equipment is in storage and requires 30 minutes of set-up time before the sample is tested, then the
effective time is really 32 minutes. In addition, the time it takes for the sample to be collected from
the field will affect the overall response time.
Response time can also include the time to inform the decision makers of the results of the
contaminant analysis, initiate the response decision making process, and initiate the implementation
of the response plan. Since the definition of sufficiently rapid EWSs includes "sufficient time for
action," the desired outcome of the action should be defined. For example, the desired outcome
could require actions as demanding as prevention of the contaminant reaching taps by mitigation
measures or pump shut down or a more simple response of issuing a boil-water notice.
Although detection, data analysis, decision making, and response implementation all factor into the
EWS's overall response time, it seems reasonable that the detection technology should not be the
bottleneck in the response process. Rapid detection technology should ideally have high throughput
(collecting data or samples every few minutes or less), rapid assay, and brief analysis time. In a
best-case scenario, a contaminant is detected, a decision is made, and a response is implemented
before the contaminant has time to reach consumer taps. In this sense, an EWS should strive to
"detect to protect" (incident detected and exposure prevented; may be instant) and "detect to
warn" (incident detected before significant exposure or manifestation of public health indicators; may
take hours). Systems that can only "detect to treat" (incident detected after exposure occurs or
manifestation of public health indicators, may take hours to days) qualify as contamination warning
systems but may not meet the rapid criteria of an EWS (Roberson and Morley, 2005). However,
with improvements in technology and the ability to detect specific contaminants or specific
contaminant categories in near real-time, it is anticipated that a contamination warning system will
move toward "detect to warn."
3.1.2 Range of Contaminants
It is impractical and probably unproductive to focus on specific drinking water contaminants when
designing an EWS. In addition, long and exhaustive lists of agents can give a misleading impression
of the extent of possible threats (WHO, 200416). Because the potential lists of contaminants and
potential threats are very large, it is unreasonable to expect to have a separate detection technology
in place for each contaminant or threat. The unlimited range of existing and emerging contaminants
would tax the resources and technical capacity of any water distribution system. Instead, there needs
to be an ongoing process of grouping known contaminants by such properties as their
physicochemical traits, source, public health impact, or the probability that they may be used to
disrupt a water distribution system. Characteristics like these can be useful because they help
August 2005 20
-------
U.S. EPA Office of Water Early Warning Systems
determine the appropriate type, placement, and cost of monitoring technologies. It is necessary,
therefore, that a suite or panel of technologies be capable of detecting broad groups of contaminants
and other threats, instead of separate detection technologies for each contaminant. Toxic chemicals,
radiological contaminants, and microbial pathogens are broad categories that require particular
detection and identification strategies. Additional subcategories with examples are provided in
Exhibit 3-1 (EPA, 2003/2004).
3.1.3 Automation and Remote Operation
Automated systems have several advantages over manual sampling assays. Sampling intervals are
easier to dictate and track with automated systems. Although human error and variability introduced
due to different human operators remain factors for automated systems, they are much reduced
compared to manual assays. With automated systems, remote monitoring is more feasible because
personnel do not need to travel to the sampling site each time a sample is taken. Remote operation
of the technology is also valuable. There are devices that are automated and can be placed remotely,
but if parameters need to be adjusted, calibrated or validated, then personnel must travel to the
devices. Performance optimization is cumbersome if traveling to each device is required to adjust
parameters. Systems amenable to adjustment, calibration, or validation via a central command
location are desirable but not available for many sensors used for detecting contaminants.
"Online" implies a certain degree of automation, remote control, and real-time capability. Online
at a minimum refers to the capability of a device to be permanently installed. "Continuously online"
can be used to emphasize real-time capability. This term should not be confused with "in-line" or
"in-pipe" which refers to the placement of a device within a pipe, so distribution system water does
not have to be re-routed out of the distribution system to be sampled. In-line or in-pipe technologies
are also online.
Automatic sample archiving is necessary when confirmatory testing is desired. When an alarm is
triggered, an automated sample collection device should store a grab sample so that the water sample
that triggered the alarm can be analyzed either in the field with portable devices or in a laboratory
for more sophisticated analysis. If automated sample collection is not in place, then it is problematic
to determine if a monitor detected a transient spike in contamination or if there was a false alarm.
Although flow rates vary throughout the distribution system, it is not unreasonable to predict that
some contamination spikes of concern could pass by sensors before response teams could make it
to the monitor location to perform a manual grab sample. Even if data from multiple sensors are
integrated with flow models to predict downstream contamination location and concentration, it may
still be desirable to have archived samples, particularly if a criminal investigation results from the
event.
August 2005 21
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 3-1. Drinking Water Contaminant Classes and Examples
Category
Examples
MICROBIOLOGICAL CONTAMINANTS
Bacteria
Viruses
Parasites
Bacillus anthracis, Brucella spp., Burkholderia spp., Campylobacter spp.,
Clostridiumperfringens, E. coli O157:H7, Francisella tularensis, Salmonella typhi,
Shigella spp., Vibrio choleraeOl, Yersiniapestis, Y. enterocolitica
Caliciviruses, Enteroviruses, Hepatitis A/E, Variola, Venezuelan equine encephalitis
virus
Cryptosporidium parvum, Entamoeba histolytica, Toxoplasma gondii
CHEMICAL CONTAMINANTS
Corrosives and caustics
Cyanide salts or cyanogenics
Metals
Nonmetal oxyanions, organo-
nonmetals
Fluorinated organics
Hydrocarbons and their oxygenated
and/or halogenated derivatives
Insecticides
Malodorous, noxious, foul-tasting,
and/or lachrymatory chemicals
Organics, water-miscible
Pesticides other than insecticides
Pharmaceuticals
Schedule 1 Chemical Weapons
Biotoxins
Toilet bowl cleaners (e.g., hydrochloric acid), tree-root dissolver (e.g., sulfuric acid),
drain cleaner (e.g, sodium hydroxide)
Sodium cyanide, potassium cyanide, amygdalin, cyanogen chloride, ferricyanide
salts
Mercury, lead, osmium, their salts, organic compounds, and complexes (even those
of iron, cobalt, copper are toxic at high doses)
Arsenate, arsenite, selenite salts, organoarsenic, organoselenium compounds
Sodium trifluoroacetate (a rodenticide), fluoroalcohols, fluorinated surfactants
Paint thinners, gasoline, kerosene, ketones (e.g., methyl isobutyl ketone), alcohols
(e.g., methanol), ethers (e.g., methyl ferf-butyl ether or MTBE), halohydrocarbons
(e.g., dichloromethane, tetrachloroethene)
Organophosphates (e.g., Malathion), chlorinated organics (e.g., DDT), carbamates
(e.g., Aldicarb) some alkaloids (e.g., nicotine)
Thiols (e.g., mercaptoacetic acid, mercaptoethanol), amines (e.g., cadaverine,
putrescine), inorganic esters (e.g., trimethylphosphite, dimethylsulfate, acrolein)
Acetone, methanol, ethylene glycol (antifreeze), phenols, detergents
Herbicides (e.g., chlorophenoxy or atrazine derivatives), rodenticides (e.g.,
superwarfarins, zinc phosphide, a-naphthyl thiourea)
Cardiac glycosides, some alkaloids (e.g., vincristine), antineoplastic chemotherapies
(e.g., aminopterin), anticoagulants (e.g., warfarin). Includes illicit drugs such as
LSD, PCP, and heroin
Organophosphate nerve agents (e.g., sarin, tabun, VX), vesicants, [nitrogen and
sulfur mustards (chlorinated alkyl amines and thioethers, respectively)], Lewisite
Plant, animal, microbial, and fungal derived toxins
(e.g. ricin, botulinum toxin, aflotoxins)
RADIOLOGICAL CONTAMINANTS
Radionuclides
Does not refer to nuclear, thermonuclear, or neutron bombs. Radionuclides may be
used in medical devices and industrial irradiators (e.g, Cesium- 137, Iridium-192,
Cobalt-60, Strontium-90). Class includes both the metals and salts.
August 2005
22
-------
U.S. EPA Office of Water Early Warning Systems
3.1.4 Affordable Cost
Affordability is essential. Although what is considered affordable will vary for different water
utilities, the desired characteristics and features of EWSs outlined in this chapter can serve as a
checklist to compare systems to each other and to buyer requirements. In addition, other factors will
likely be considered by potential users, such as how soon the equipment will become outdated,
supply and maintenance costs, and budget demands from other high priority areas. A cost-effective
EWS should allow for future growth and improvements. This might be accomplished by a modular
design that could be upgraded stepwise on a planned schedule to incorporate new technologies as
they become available. This type of spiral development should keep EWSs evolving, both in terms
of performance and cost.17 Besides the capital costs, there are ongoing operation and maintenance
costs.
3.1.5 Low Skill and Training
Skill level and training requirements will factor into the cost of the detection technologies and the
EWS as a whole, and will impact the effective use of the system. Complicated technologies that
require initial training along with further practice and hands-on experience for proper
implementation will suffer from periods of shut down if personnel turnover is too rapid. In contrast,
technologies that require a low skill level and minimal training to run effectively are more likely to
yield results on a consistent basis. Skill and training required should be considered for both the
actual sampling or assay technology, as well as the necessary software needed for analysis and
interpretation.
3.1.6 Source of the Contaminant
It is desirable that the site of introduction of contamination be identified as quickly as possible.
Although it is unlikely that an EWS will be able to pinpoint the point of entry, it should provide
guidance for investigation of the event and narrow the possible locations. Where contamination is
ongoing, this is valuable for halting contaminant entry. In the case of intentional contamination, the
entry site would be subject to criminal investigation protocols. Sites of accidental transient entry
would also need to be evaluated. Site locations for monitoring equipment throughout the system and
special contaminant flow models will assist in tracing a contaminant to its point of entry. In
addition, it is important to accurately predict the location and concentration of contaminants once
they pass the detector site. This is valuable for knowing where an intervention, grab sample, or
other response might be effective, for determining the particular users at risk, and for predicting
exposures.
3.1.7 Sensitivity
Sensitivity of an assay or test impacts the utility and often the cost of the detection technology.
Assays that can only reliably detect relatively high levels of contaminants would not be useful if
lower levels need to be detected. Often assays that can detect very low levels of contaminants are
swamped by high levels and may be more expensive than other appropriate options. An assay that
can quantitate abroad range of concentrations, however, is likely to be prohibitively expensive. For
regulated contaminants, an assay should be sensitive enough to detect concentration levels that
August 2005 23
-------
U.S. EPA Office of Water Early Warning Systems
bound the regulatory levels. For potentially hazardous contaminants, such as CBW agents, the
chosen thresholds of detection should be scientifically based and protective of public health.
3.1.8 Minimal False-positives/False-negatives
False-positives and false-negatives can make a system effectively useless. False-positive and false-
negative rates can be defined for individual monitors within the integrated EWS and for the EWS
as a whole. False-negatives lead to decreased protection because contaminants present at relevant
levels escape detection. This could result in catastrophic public health consequences and/or loss of
confidence in drinking water supplies. False-positives from devices or assays can slow down the
effective response time, because each time a positive result is obtained, it must be confirmed by
additional tests. If the consequences of a true positive are grave enough, then the initial presumptive
positive result would trigger a response. Any response has an associated cost measured in labor time
and direct costs. If the initial result is subsequently determined to be a false-positive, then money
has been spent needlessly and in some cases public trust is eroded. Each contamination warning that
the public finds inaccurate will increase the rate at which future warning are ignored, which would
increase the health impacts in a real contamination event. In many cases, the time needed to confirm
a result is added before any response plan is initiated.
Appropriate data analysis and event correlation techniques can reduce the false alarm rate for the
EWS. Only systems with either very low false-positive/negative rates or very rapid confirmation
methods should be considered early warning. The Defense Advanced Research Projects Agency
(DARPA) Chemical and Biological Sensors Standards Study presents methods for evaluating
sensors by capturing the performance trade-offs between sensitivity, probability of correct detection,
false-positive rate, and response time (DARPA, 2004). Hrudey and Rizak (2004) have developed
a statistical framework for hazard detection and judgment of the evidence to provide insight and
mathematical justification to decisions regarding balancing false-positive and false-negative errors
in the drinking water security context. In addition, Bravata et al. (2004) describe how sensitivity,
specificity, and pre- and post-test probability are reported in receiver operating characteristic (ROC)
curves to graphically communicate predicted false-positive and false-negative information. The
authors also suggest that published guidance for evaluating clinical diagnostic tests can be adapted
for evaluating detection systems because diagnostic test guidelines are well established and promote
study designs that provide unbiased estimates of both sensitivity and specificity (or likelihood ratios)
relative to an acceptable standard. Manufacturers and parties involved in third party verification
should quantitate the false-positive and false-negative rates.
Sources of false-positives and false-negatives can be due to human error, such as inaccurate
pipetting or mixing techniques. There are also common chemicals or conditions in the water
distribution system that can interfere with some types of assays. For example, the chlorine residual
can interfere with detection because chlorine itself may be measured as a toxin by some of the
assays. Chlorine can also be toxic to biomonitors (e.g., fish, bacteria). Copper and other metals can
affect the chemistry of some assays. Biofilms (microbial communities that coat the inside of pipes)
commonly occur in the distribution system, and can naturally slough off, releasing microorganisms
into the finished water. Microbial detection technologies could confuse this background level of
microbial occurrence with an intentional contamination event. Such interferents are discussed in
more detail in Chapter 9.
August 2005 24
-------
U.S. EPA Office of Water Early Warning Systems
The rates of false-positives and false-negatives for individual devices and assays should be
distinguished from the overall false rates of the EWS system. Although individual technologies may
have false-positive and false-negative rates that can be compensated for by overall design of the
EWS through the use of confirmatory and backup tests, the overall EWS should have a very low rate
of false-positives and false-negatives. If a response plan includes calling in emergency responders
from other community sources, utilities should consider that some local governments may have fines
in place for organizations and individuals that exceed a certain number of false alarms in a given
time period.
3.1.9 Robustness and Continuous Functioning
EWSs and the associated technologies should be resistant to damage or inaccuracies caused by
human error or environmental conditions, such as continuous exposure to the water environment.
This robustness decreases maintenance costs and provides increased reliability. These criteria apply
to both the hardware and the software. Human error during maintenance should be easily detectable.
Equipment that is highly sensitive to jarring, shaking, or falling has reduced utility and probably
higher maintenance costs. Environmental conditions such as humidity and temperature will
fluctuate, even in locations that are sheltered from direct sunlight and precipitation. Software that
is prone to crashing or is overly complicated to master compromises the usefulness of the system.
Battery life should be considered for hand-held units. For online units, automatic restart after a
power interruption is desirable, particularly for remotely operated devices. Continuous, predictable,
year-round functioning is a top priority for EWSs.
3.1.10 Third Party Verification
Third-party verification evaluation is desired to determine if specific devices and methods perform
as they are intended and advertised. EPA's ETV Program and TTEP provide third-party verification
evaluation reports for several products related to EWSs. Another organization that provides third
party verification is the AOAC International,18 whose mission statement is to serve "the communities
of analytical sciences by providing the tools and processes necessary for community stakeholders
to collaborate and, through consensus building, develop fit-for-purpose methods and services for
assuring quality measurements."
3.2 Design Features of Integrated Early Warning Systems
The design of an integrated EWS requires a conceptual framework of all its component parts.
Exhibit 3-2 outlines the component features of an integrated EWS. An integrated EWS is more than
just selecting sensors. It includes determining sensor locations, acquiring data, conducting data
analysis, developing communication and notification links, establishing decision making procedures,
and developing response protocols. Although public health surveillance, monitoring, consumer
complaints, and consequence management are features of a more broadly defined contamination
warning system (see Appendix A), this study focuses more on the use of contaminant
sensors/detectors and the associated data and communication networks in the drinking water
distribution system.
When considering an integrated EWS, especially for a distribution system, a utility should go
through a structured decision making process. Exhibit 3-3 presents a conceptual process for
August 2005 25
-------
U.S. EPA Office of Water Early Warning Systems
designing an EWS. The utility should (1) determine the need for an EWS, (2) conduct the proper
and necessary planning and coordination, (3) prepare the overall EWS approach, and (4) develop
the details of the EWS design. These steps are described in more detail below. Much of this
information on the EWS design process has been reviewed by Hasan and colleagues in Water
Resources Update (Hasan et al., 2004). In a related effort called the WaterSentinel Initiative (see
Appendix A), EPA is designing a contamination warning system. Furthermore, EPA, in partnership
with drinking water stakeholders, is developing a concept of operations which will be further
designed and piloted for deployment by select water utilities.
August 2005 26
-------
1 <
»B
I Decision Maker
^3
I Notification, an
PH
I Response
Decision software, E
„
X
o
~o
o
H
CD
O3
1
O3
bft
1 confirmation testin;
CD*
emergency respons
1 communication
I
5B
>%
GC
j
.0
1
«
^
{/}
a
\ informatioi
O5
a
O5
>^
O5
S
TO
S
1 software, s
e
—
e
o
OB
X.
S
e
t
I
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 3-3. Conceptual Process for Designing an Early Warning System
Evaluate Need for EWS
Determine Vulnerability/Threat (Contaminant List)
Consider EWS and Possible Alternatives and Backups
Establish EWS Plan
Obtain Management Commitment and Establish Teams
Define Objectives and Performance of EWS Plan
Examine Distribution System and Vulnerabilities
Determine Overall Approach to EWS Design
Develop Detailed Design of EWS
Select EWS Detection Technology
Determine Alarm Levels
Conduct Studies on Fate and Transport of Pathogens and
Chemicals
Determine Sensor Location and Density
Select Systems for Data Management, Interpretation, and
Reduction
Establish Response Communication Links, Notification, and
Decision Making
3.2.1 Evaluate the Need for an Integrated EWS
In evaluating the need for an integrated EWS, the utility should perform or revisit its vulnerability
assessment, especially with regard to the distribution system's vulnerability to intentional
contamination events. The vulnerability assessment should consider the threat of such an event, the
consequences of an event, and the status of the facility to prevent, identify, and respond to such an
event. Additionally, the utility should characterize the distribution system to determine if there is
a reasonable way for the utility to provide early warning to protect users. The utility's size, usage
patterns, and vulnerability should be factors. The utility should also weigh the cost-benefits of
integrating customer complaint monitoring and public health surveillance with contaminant
sensors/detectors and the associated data and communication networks. The utility should consider
August 2005
28
-------
U.S. EPA Office of Water Early Warning Systems
the cost and reliability of current EWS technologies. Although the protection of public health is the
primary goal of an EWS, public perception of the effectiveness of the warning system should also
be considered. Not only must the public actually be protected, the majority of the public should be
made to feel protected to the extent possible. In this regard, utilities should consider the appropriate
level of information to be released to the public. Although the level of information should be enough
to allow confidence in the EWS and deter attacks, it should not compromise the effectiveness of the
EWS. Utilities may choose to include additional design features that address the concerns of their
customers.
The size of the water utility could greatly influence the appropriate design for an EWS. Large,
medium, and small water utilities have significant differences that should be considered. For
example, the number and strategic placement of sensors in a finished water system will depend on
the size of the utility including the miles of pipeline in the distribution system, the populations
served, and the flow dynamics. The types of contamination threats may also differ from large to
medium to small systems so the EWS design should also reflect these differences. In addition, water
utilities of differing size will have vastly different budgets to spend on integrated EWSs. For
example, small or medium systems may need to rely mostly on lower cost screening techniques in
conjunction with some water quality monitors and customer complaint data rather than investing in
expensive online systems or confirmatory kits. Small systems may not have the more sophisticated
data collection and analysis systems such as SCAD A. Thus, other EWS designs may be needed with
less automated or even manual data collection/analysis techniques.
3.2.2 Establish EWS Plan and Coordination
A decision to develop an EWS capability requires a commitment from management. The planning
for an EWS usually requires the utility to assemble a team, which may include personnel from the
water utility, the local and state health department, emergency response units, law enforcement
agencies, and local political leadership. An EWS plan should be developed that outlines the
objectives for the EWS which should be clearly defined. The plan should also include how the
monitoring results or data will be interpreted, used, and reported. The plan can set performance
criteria and some essential design elements. The team should consider legal and regulatory issues
and the plan should define a budget and time line for the project. Note that the planning and
coordination for an EWS for finished water may be a separate, but linked, effort with an EWS for
source water contamination.
In developing the plan, the team should characterize the distribution system to be monitored. For
example, a distribution system should be characterized in terms of flow, pressure, access points,
water demand and usage patterns, extent of pipeline, and the locations of pipeline and pump stations.
A hydraulic model may be useful in this characterization.
Additionally, the team should specifically examine the vulnerability of the distribution system to
intentional contamination. Previously conducted vulnerability assessments, required by the
Bio terrorism Act, can be extremely helpful in evaluating aspects of physical security. An expanded
vulnerability assessment can help identify the contamination scenarios that could occur (e.g., pump
to overcome pipe pressure), as well as the insertion location and method (e.g., dump over a short
period; bleed, pump, or dissolving matrix over longer times). The duration of a contamination event
is an important parameter for an EWS. Hydraulic modeling may be useful in identifying those
August 2005 29
-------
U.S. EPA Office of Water Early Warning Systems
portions of the distribution system that, if compromised, would have the most time-sensitive
consequences. The vulnerability assessment should also help to identify target contaminants.
System vulnerabilities and the ability of existing treatment barriers to remove or neutralize specific
contaminants should be considered in choosing the contaminants of concern. The list could include
groups of contaminants that cover a broad range of specific contaminants. Lists of contaminants are
available from various agencies, including EPA. This step is important because different
components of the EWS may be chosen to detect toxic, microbial, or radiological contaminants.
During planning for an EWS, there should be a discussion of system maintenance requirements,
housekeeping, system administration issues, human resource requirements including security and
training, exercising the system and its human participants, and upgrading the system when improved
technology is available.
3.2.3 Determine Overall Approach to EWS Design
There are many ways to approach development of an effective EWS design. While real-time
continuous monitoring may be the ultimate long-term goal, there may be intermediate EWS
architectures that would be more effective until real-time monitors are available. Under the
proposed Fiscal Year 2006 budget, EPA will launch WaterSentinel, a proposed demonstration
project by which EPA, in partnership with select utilities and laboratories, would design, deploy,
and evaluate a model contamination warning system for drinking water security. EPA and its
partners will thus gain operational and tactical experience that can assist in developing standardized
and cost-effective approaches to coordinated surveillance and monitoring of drinking water. Until
this program is implemented and lessons learned however, a few possible approaches to EWS design
and some brief discussion on their capabilities and limitations are provided below.
• Some experts in the field of EWS design for drinking water distribution systems advocate a
"tiered" approach with two stages. The first stage uses continuous real-time sensors that can
provide a generic warning or trigger an alarm that a contaminant is in the water. Initial detection
could occur within seconds, if first-stage monitoring is effective and located properly. This
would trigger a second stage using more specific and sensitive technology to confirm and
identify the contaminant (ILSI, 1999; Hasan et al., 2004). A second-stage technology might
either be located in the field or be brought to the site as a field portable unit. Analysis could take
minutes or hours depending on the method of detection and the location in the distribution
system. This tiered approach addresses the current situation where limited or no technology
exists that allows for affordable, online, real-time monitoring, and identification of hundreds of
specific contaminants. The first screening stage alone of a monitoring system does not constitute
an EWS. Thus, confirmatory analyses used to verify a positive result from a screening analysis
should be integrated into the overall design of the EWS. A more complete discussion of a first-
stage/second-stage approach to an EWS is provided in Exhibit 3-4. Other experts are concerned
about the possibility of obtaining many false-positives for certain first-stage technologies that
could be sensitive to normal fluctuations in water quality (Wayne Einfeld, Sandia National
Laboratory, personal communication).
August 2005 30
-------
U.S. EPA Office of Water Early Warning Systems
Exhibit 3-4. Sample EWS Approach using First-Stage and Second-Stage Monitors
First-Stage Monitoring
A system that continuously monitors for general water quality parameters (e.g., turbidity,
temperature, pH, ORP, conductivity) in the distribution system has the potential of providing first-
stage monitoring for contamination events (intentional or accidental). Some, but not all,
contamination events can cause detectable changes in water quality parameters that are sufficient
level to provide an alarm. An intentional contamination event could even create changes in water
quality parameters that present a more specific fingerprint/signature indicating a possible group
of contaminants. In the laboratory, using confined distribution systems, online water quality
monitoring, and controlled spiking of known contaminants and contaminant cocktails, data are
being produced on how water quality parameters respond during contaminant influx. However,
an important concern is that baseline fluctuations in water quality parameters would need to be
determined before an anomaly in readings could be identified. This concern is the subject of
current research by EPA and other agencies. Some patterns in fluctuations may be associated
with storm events or certain operational conditions. It is likely that considerable experience with
daily, seasonal, and event-related fluctuations will be required for a specific finished water
distribution system before the monitoring system can be maximally useful for detecting
contamination events. Thus, false-positives are a large concern of utilities. Chapter 5 reviews
the specific technologies and research projects that are involved in developing first-stage
monitoring capabilities. Note that first-stage monitoring is not limited to traditional water quality
parameters; it could in the future also include biologically-based monitors.
Second-Stage Detection Confirmation
After routine first-stage monitoring has "raised a red flag," second-stage confirmation can be used
to verify detection and/or characterize specific contaminants. Second-stage confirmation differs
from first-stage monitoring in that second-stage technologies are not as amenable to high
throughput continuous sampling. For first-stage monitoring, speed is measured in seconds, while
for second-stage confirmation speed is more appropriately measured in minutes (-10 to 120
minutes). Because second-stage confirmation is often not in continuous operation, set-up time
(solution mixing, equipment warm-up), sample capture time (time to obtain an appropriate sample
to be tested), and assay and analysis time should all be included in the estimate for how much
time it will take from the initial "red flag" until second-stage confirmation is complete. The
information gained from second-stage confirmation is more specific and should assist in choosing
between more targeted response options. Rapid identification of the specific contaminant may
also help mitigate adverse health outcomes for potentially exposed populations.
Note: "Stage" refers to levels of detection within the same EWS. It does not refer to near-term
EWS design and long-term EWS design goals.
August 2005 31
-------
U.S. EPA Office of Water Early Warning Systems
• Another approach, perhaps an intermediate EWS design, is to (1) use multiple water quality
monitors to provide a red flag for contaminants, (2) perform frequent grab sample analysis by
rapid contaminant-specific monitors (e.g., arsenic/cyanide), and (3) take automated weekly
samples of water that can be analyzed for a suite of contaminants. The water samples could be
screened by field methods or analyzed with laboratory instruments, which are more developed
than online contaminant detectors. This approach has the advantage of monitoring for generic
warnings while using some established technology that is now available (e.g., chemical monitors
for continuous monitoring, automatic samplers, and established microbial tests). The approach
provides some intermediate level of EWS monitoring.
• Another more advanced EWS design is to use multi-parameter water quality monitors and
routine composite/grab sampling in conjunction with data collected from customer complaints
and public health surveillance as triggers for confirmatory testing. This design provides a
mechanism to identify potentially large scale events that are in the process of occurring.
Regardless of the approach, a properly designed EWS should include all other elements of a
monitoring program necessary to inform decision officials responsible for public health protection.
An EWS can be designed to be part of an overall water safety and security program for the utility.
In addition to providing warning for intentional contamination, the EWS fills a dual role of
providing water safety through routine and compliance monitoring. Other information sources such
as customer complaints could be indicative of intentional as well as accidental contamination events.
Utilities may want to consult with other utilities with additional experience, as well as keep informed
on the latest research development and test results. Such discussions need to be part of the overall
approach to EWS design.
3.2.4 Develop Detailed Design of an Integrated Early Warning System
This step is really a series of interrelated substeps, because the information from one may affect the
other. For example, a decision about which EWS detection technology to choose may depend on
the location and number of units to purchase. Whereas the alarm levels chosen may influence the
selection of EWS technology (i.e., make and model). Thus, all of these substeps will be discussed
together.
• Select EWS Detector Technology
Once target contaminants for the EWS have been identified, and the range of concentration
necessary to detect them has been established, it is necessary to select a monitoring technology for
the particular contaminant or class of contaminants. This assumes that a monitoring technology that
meets the core requirements of an EWS currently exists. Furthermore, the monitoring technology
should be capable of dealing with complex water matrices. This may require an extraction step to
remove the material from the water matrix and/or a concentration step to enhance detection and
quantification. Although techniques for isolating, concentrating, and purifying microbial and
chemical substances have been developed for many laboratory methods, they may not necessarily
be transferable to field deployable monitoring devices. The technology considered for use in an
EWS should be evaluated to ensure that all steps of the methodology perform correctly and can
detect the target contaminant(s) without excessive interference. Identifying a field-deployable
technology with an acceptable methodology is only the first step. Performance of the monitoring
August 2005 32
-------
U.S. EPA Office of Water Early Warning Systems
technology must also be adequate to meet the data quality objectives of the monitoring program.
These data quality objectives should be defined during the design of the EWS and include
specificity, sensitivity, accuracy, precision, and recovery, as well as rates of false-positives and
negatives. If the monitoring technology cannot meet the data quality objectives, then another
technology should be selected. If no technology can be identified that meets the objectives, then
either the EWS should not be implemented, or the data quality objectives will need to be revised.
If the later approach is taken, it will be necessary to modify the manner in which the results are used
to be consistent with revised data quality objectives. There are a great variety of EWS detection
technologies available (see Chapters 6 and 7). Certain technologies can be used online and provide
more real-time capability.
• Determine Alarm Levels
The basis for setting alarm levels will depend on the previously determined levels at which
contaminants need to be detected (based on human health risks) and on the type of EWS employed.
The response that is initiated by an alarm must be established before the EWS is deployed.
Automated systems with online monitoring especially should have alert capabilities. Operators
should be able to set threshold values so the system automatically triggers an alarm if readings move
outside of the range that has been defined as safe. Performance optimization of the system should
allow for alarm trigger rates that minimize false alarms, but still detect contamination events that
could pose a health risk. If a false alarm leads to a decision to issue a notice to the public to stop
using the water, public health as well as public confidence could be impacted.
• Conduct Studies on Fate and Transport Modeling of Pathogens and Chemicals
Chemical and microbial contaminants can behave in a variety of ways as they migrate through a
water system. Environmental conditions, the presence of oxidants or other treatment chemicals, and
the hydraulic characteristics of the system will affect the concentration and characteristics of these
contaminants. If information is available on contaminant characteristics that affect the
contaminants' fate and transport, it should be factored into the design of an EWS. For example, if
a target contaminant is known to chemically degrade at a certain rate in the presence of free chlorine,
it may be possible to use a hydraulic/chemical model of the distribution system to predict the
concentration profile through the system. This information in turn can be used to select optimal
locations for sensors.
• Determine Sensor Location and Density
Choosing the location or site of sensor placement within a distribution system is a complex task.
The location and density of sensors in an EWS is dictated by the results of the system
characterization, vulnerability/threat assessment, usage considerations, risk minimization, and cost.
Thus, the easiest site to place a sensor may not be a site that yields the most useful information. The
size of the population downstream from the sensor may be a defensible criterion for choosing a site,
but might not be the best site for system modeling. Due to the complexity and dynamic nature of
distribution systems, it may be beneficial to develop a hydraulic model of the system to assist in the
placement of sensors. Real-time integrated pressure and flow data can be used to build flow models
that have well characterized predictive capabilities. Other factors to determine sensor location could
be location of isolation valves, location of critical nodes (hospitals, emergency response), and
August 2005 33
-------
U.S. EPA Office of Water Early Warning Systems
physical security of the location. Even if sensors can be optimally located within a distribution
system, there may not be sufficient time to prevent exposure of a portion of the public to the
contaminated water. At best, monitoring conducted within the distribution system will provide time
to limit exposure, isolate the contaminated water, and initiate mitigation and remediation actions.
• Select Systems for Data Management, Interpretation, and Reduction
One of the challenges of a continuous, real-time monitoring system is management of the large
amounts of data that are generated. Use of data acquisition software and a central data management
center is critical. This will require that individual sensors deployed in the system be equipped with
transmitters, modems, direct wire, or some other means to communicate the data to the acquisition
and management systems. Furthermore, the data management system should be capable of
performing some level of data analysis and trending in order to assess whether an alarm level has
been exceeded. The use of "smart" systems that evaluate trends and can distinguish between
genuine excursions and noise could minimize the rate of false alarms. A decision will also have to
be made regarding the action that is taken when the data management system detects an excursion
above the alarm level. At a minimum, the system should notify operators, public health agencies,
and/or emergency response officials. If possible, redundant communication should be used (e.g.,
notifying multiple individuals through multiple routes such as telephone and fax). In some cases,
it may be appropriate to program the data management system to initiate preliminary response
actions, such as closing valves or collecting additional samples. However, these initial responses
should be considered simple precautionary measures, and response decision makers should make
decisions regarding decisive response actions. Integrated analysis of the data gathered on many
different parameters and contaminant detectors is what makes an integrated EWS more than a
collection of detection technologies and assays. Data validation helps ensure the integrity of data
by requiring that appropriate quality assurance and quality control (QA/QC) procedures are
followed, and that adequate documentation is included for all data generated. Data validation
guidance may be different for different types of data and the stringency depends on the planned
utilization of the data. Data security is also an essential part of the integrity of the system. It is
important during the transmission of data and during the analysis and storage of information. For
obvious reasons, it is important that data being transmitted from a remote sensor to a command
center be free from tampering or accidental degradation. Because previously obtained data are
accessed by the system to look for patterns, it is also important that old data be free from tampering
or accidental degradation.
• Establish Response Communication Links, Notification, and Decision Making
An integrated EWS would also include network and communication links between command center
operators and any people designated as response decision makers. The communications network
should assist with the rapid relay of critical information to the utility decision makers who
implement the response. The response may include the utility informing outside parties, such as
public health officials, police force, or other emergency responders. Utilities can refer to local
ordinances and laws to determine when they are required to inform outside parties. In general,
threats should be deemed credible before other agencies are alerted. Many responses are possible
when an early warning monitoring system triggers an alarm. Monitoring devices may report data
to a command center that will relay information to utility decision makers. In turn, the response may
use the same secure communications links to take actions in the distribution system (e.g., shut
August 2005 34
-------
U.S. EPA Office of Water Early Warning Systems
isolation valves). Further actions may include monitoring and sampling for the contaminant at
appropriate locations in the distribution system and monitoring for surrogate parameters that may
indicate contamination (e.g., increased chlorine demand, changes in pH). The EPA Response
Protocol Toolbox (EPA, 2003/2004) greatly assists with planning this part of the response.
After developing this comprehensive strategy for an integrated EWS, the utility is ready to develop
an implementation strategy that includes buying and installing the equipment; providing
maintenance and operation procedures; and providing training, testing, and exercises for systems
and personnel.
August 2005 35
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
4. Features of an EWS Related to Data Acquisition and Analysis; Contaminant
Flow; Sensor Placement; Alerts; Data Security; and Communication,
Response, and Decision Making
The purpose of this chapter is to characterize the state-of-the-art features of EWSs that are not
related to sensors. These EWS features include real-time data acquisition and analysis; contaminant
flow predictive systems; sensor placement; alert management; security enforcement; and
communication, response, and decision making. These features of EWS design and operation should
be part of an overall plan for the interpretation, use, and reporting of monitoring results (Hasan et
al, 2004).
4.1 Real-Time Data Acquisition and Data Analysis
4.1.1 Basics
Continuous, real-time water quality monitoring systems for an EWS have the potential to generate
large amounts of raw data, which can be unmanageable without the use of a data acquisition and
management system. Water utilities are already familiar with collecting information on water
quality and operating their drinking water treatment plant and distribution systems. Over 75 percent
of drinking water utilities already operate online analyzers for water quality parameters such as
chlorine and turbidity. However, not all of these utilities have monitors within the distribution
system. The 75 percent figure includes utilities that have monitors only at the drinking water
treatment plant (AwwaRF, 2002). Utilities often use the supervisory control and data acquisition
system known as SCADA. A SCADA system links monitoring instruments, remote telemetry units,
programmable logic controllers, and a host computer in order to integrate data collection and
processing into a single system-wide control center that can be accessed from various locations
(Carlson et al., 2004). Large utilities usually use a SCADA system for controls in the distribution
system. Such a SCADA system can often incorporate data from online or remote sensors in a cost
effective manner (Mays, 2004). The system triggers the collection of data. Remote locations could
use microprocessor-based "smart" SCADA systems. Although slower and more expensive than
programmable logic controller-based systems, the "smart" SCADA systems save on communication
media costs, maintenance, and travel. The simplest SCADA systems only include three or four
input/output channels for monitoring (Mays, 2004). SCADA systems that are already in operation
for daily water quality monitoring purposes may not be configured so that they can serve all the
locations where EWS detectors are needed (see Section 4.3 for discussion of sensor placement).
Various approaches taken and systems used for data transmission, validation, and analysis are
described below.
• Data Transmission
Data transmission to the central database occurs through either hardwired or wireless systems.
Hardwired transmission requires the physical connection of cable or wire, and can utilize either
coaxial or fiber optic technology. In some cases it may be difficult to hardwire remote locations.
Wireless transmission can use a variety of methods, including microwave, UHF or VHP radio, basic
telephone modems, cellular telephone modems, or satellite. Wireless transmission may require a
direct line of sight between the transmitter and receiver, or the use of re-transmitters (also known
as repeaters and amplifiers) (AWWA Workshop, 2004). The least expensive transmission systems
August 2005 37
-------
U.S. EPA Office of Water Early Warning Systems
are typically phone lines or direct wire. A combination of data transmission approaches could be
used throughout the distribution system, as long as the command center can achieve integration of
the information. The data transmission method must be compatible with monitoring equipment and
data acquisition equipment. In most cases, utilities would be expanding monitoring capabilities by
adding to existing equipment and the existing SCADA system so newer upgrades need to be
compatible with existing equipment and SCADA (AWWA Workshop, 2004). For water security
applications, the existing monitoring system would need to be evaluated for its vulnerability to direct
physical attack (e.g., wire cutting), and to cyber attack (e.g., tapping). Transmission of unencrypted
data is a security risk, so hardware and software should have encryption capabilities. Traditionally,
utilities have not seen a need to encrypt data from water quality monitors, however monitoring for
security applications would require data encryption to reduce vulnerability.
The quantity of data that needs to be transmitted should be considered. Factors that influence data
quantity are the number of instruments in the system, the number of data points each instrument
generates in a given time frame (e.g., sampling rate of I/second or I/minute or I/hour), and bits per
sample. Some monitoring equipment may generate a relatively manageable quantity of data,
whereas some equipment (e.g., video) may generate data that require considerably more
transmission capability.
Currently, data transmission capacity, computer data storage capacity, and software support are
adequate for online monitoring of contaminants. The water industry is not generally considered a
large market force; thus, similarly developed systems in other markets may push development of
technology for the water industry. For example, SCADA systems based on analog signals require
special drivers to accept data from monitors (e.g., particle counters) with digital signals. Products
for conversion between analog and digital formats are commercially available. Expected future
developments will probably alleviate this concern (AwwaRF, 2002).
• Data Verification
Manual verification of sensor data will not usually be an option with a continuous, real-time
monitoring system due to the large amount of raw data collected. Therefore, automated data
validation processes are indispensable to ensure accurate results from data analysis. A simple yet
effective protocol is to compare data received from monitoring sites with data stored at the sensor
locations to ensure accuracy and completeness (Carlson et al, 2004). Traditional SCADA systems
perform data validation processes such as range checking and data filtering (e.g., moving window
averaging). Other approaches under development for data validation are simple outlier detection
by commercial off-the-shelf data mining software and finding formal correlations among spatial and
temporal data attributes (e.g., pH, temperature, DO, specific conductivity, turbidity, chlorophyll)
(Mays, 2004). Data verification is part of an overall quality assurance plan which is a systematic
procedure for determining that all aspects of the EWS are functioning as expected.
• Data Analysis
Once obtained, the data may go through quality assessment and validation, aggregation,
transformation, and analysis (AwwaRF, 2002). Data analysis is performed by specialized software
and can take the form of univariate and multivariate analysis, Rule-Based Systems (RBS), or
Case-Based Systems (CBSs). Univariate analysis targets one parameter at a time, in order to note
August 2005 38
-------
U.S. EPA Office of Water Early Warning Systems
changes in a specific parameter or instrument response in the event of a change in water quality.
This separate monitoring of instrument response is useful in determining instrument sensitivity to
different contaminants, and in confirming the validity of other similar instrument responses when
considering potential false alarms (Carlson et al., 2004). Multivariate analysis uses inputs from all
instruments simultaneously to detect data anomalies. The benefits of multivariate analysis are the
potential to detect contamination events sooner and the means to learn more about the type of
contaminant that generated the alarm. RBSs are characterized by IF-THEN rules, which provide
real-time reasoning by looping through rules and polling for new data on a programmed schedule.
CBSs operate by comparing a collection of current measurements to a database of historical
measurements. Any deviations of the current state from past data will notify the operator, who can
run a predictive WHAT-IF model to evaluate scenarios (Carlson et al., 2004). The development of
logic systems (e.g., artificial neural networks and fuzzy logic) to interpret the data that are produced
by an EWS may be an integral part of some EWS designs. As more and more information is
collected by an array of instrumentation, there must be a dependable system developed to interpret
all the data produced. One example of such a project was published by IHT Delft, entitled Use of
Artificial Neural Networks and Fuzzy Logic for Integrated Water Management: Review of
Applications.19
When online instruments provide a stream of data from which the utility will need to manage and
potentially use for response decisions, the data quality has to be handled and analyzed carefully
(AwwaRF, 2002). The utility should consider the performance characteristics of the instrument, so
that the proper degree of validity of the online measurements can be determined. Online data
aggregation and the handling of those data should be considered so that large amounts of data can
be processed effectively.
For online data, a predictable percentage of the data will be dropped or lost, particularly for remote
sites. For these cases some methods for data validation include gap filling, range checking (when
data out of range, check along full working range of sensor), rate of change check (peaks/outliers
usually a result of disturbance of the sensor), and variance check (check small variations in
repeatability). Cross validation methods explore the connection between online measurements. This
is particularly valuable for detectors with multi-parameters. Highly correlated parameter measures
can be programmed into a model and used for determining confidence in measures.
An EWS may generate real-time data for quick analysis and decisions and action. There are various
techniques for assisting in real-time reporting and decision support. They include data filtering,
operational indexes (commonly used by operators to calculate measures that represent trends for
routine operational performance), short-term prediction using software sensors, and classification
and state description to reduce information overload. Predictive modeling can also be used for
assisting with data validation (AwwaRF, 2002).
Short-term and long-term data storage and backup needs also must be assessed. For some
parameters, baseline characterization may require a full year's worth of data. Data storage needs
will depend on how the utility wishes to use the data. The monitors discussed in Chapter 5
potentially serve the dual purpose of general water quality monitoring for operations and of red flag
alert for potential security threats. Although data analysis for the purpose of determining immediate
security threats may not require long-term storage of data, there may be other secondary uses for the
data that would warrant longer term storage. For example, older data could be released to the
August 2005 39
-------
U.S. EPA Office of Water Early Warning Systems
utility's systems modeling contractors or to the research community to answer future research
questions. A standard minimum storage of data for at least several years is not unreasonable to
expect. Utilities should consider developing a policy that maximizes the usefulness of the gathered
data, while also addressing the concerns of their legal counsel, and the budgetary constraints of data
storage and access.
4.1.2 Demonstration Projects, Tests, and Products
EPA recently conducted a project to evaluate and demonstrate the usefulness of SCAD A systems
for real-time monitoring of water sensors within distribution systems (EPA, 2004). The test was
conducted at the WATERS Center in Cincinnati, OH. Within the WATERS Center, there are two
DSSs, which were designed and fabricated to evaluate and interpret the dynamics that influence
water quality within water distribution infrastructure systems typical in the U.S. and abroad.20 EPA
concluded that a SCADA system is a critical feature for handling the data from online sensors in the
distribution system. Analyzing the data from multiple sensors is too difficult and time-consuming
without a centralized SCADA. If the data were collected separately from different sensors, the data
would have to be downloaded, processed with a computer, and analyzed. This would not provide
an effective real-time alert.
A centralized SCADA system also provides a time-stamp on the measurement for comparative
analyses. The EPA report also addressed the frequency of sampling for multi-parameter water
quality monitors. The EPA report suggested a frequency of between 2 and 10 minutes depending
on the ability of the SCADA system to take the data (based on SCADA system setup and
bandwidth), the location of the sensor, and the water flow rate. This sampling frequency would also
help utilities to establish trends in water fluctuations so as to reduce false alarms. Also, during the
test, a number of problems occurred in processing 4 to 20 milliampere signals through the SCADA
system so the report recommends purchasing monitors with the option of RS232 outputs, as these
are not subject to ground loop problems and other electrical interferences. Further challenges that
were identified are the need for regular maintenance and calibration schedules for the sensors and
that operator experience and expertise with the sensors and SCADA system are essential. It is
important to note that there are technologies other than SCADA that are available for combining and
integrating data.
Remote monitoring has applications to an EWS for finished water. An operation can monitor the
water quality on line for intentional contamination. A demonstration project conducted by EPA a
few years ago examined remote monitoring of drinking water treatment. The remote monitoring
system consisted of the sampling units and an associated data acquisition unit. The SCADA units
were programmed to monitor, record, and control water treatment and distribution system operation
in three locations: Washington, DC; McDowell County, WV, and a distribution system simulator
(DSS-1) at EPA's T&E Facility in Cincinnati. In Washington DC, online instruments measured
temperature, turbidity, pH, and residual chlorine at locations within the distribution system. In West
Virginia, EPA collaborated with the McDowell County Public Services Division to implement a
smart SCADA in 1998. The smart SCADA unit observed trends in pressure, chlorine, turbidity, and
flow at 15 minute intervals. The system also was programmed to provide an alert if certain
conditions were met. The project showed that remote monitoring can be a practical option for small
rural utilities (AwwaRF, 2002). Details on EPA's research in Cincinnati on online instruments are
discussed in Sections 5.2 and 5.3.
August 2005 40
-------
U.S. EPA Office of Water Early Warning Systems
There are several commercial products currently available on the market to help acquire and manage
sensor data remotely to improve water security. One company, PDA Security Solutions (Greer, SC),
is marketing the Hydra Remote Monitoring System. It captures data from sensing devices located
remotely throughout the water distribution system. It includes trend analysis tools, and it
continuously compares real-time water quality data against benchmark water profiles. Alarms are
programmed to alert operations and security personnel. Biometric login and data encryption are part
of the system. EPA has a Cooperative Research and Development Agreement (CRADA) with PDA
to develop a nationwide water quality surveillance project. Under the Federal Technology Transfer
Act, EPA can form CRADAs with the private sector in order to speed the development of
technologies for various programs. Currently, EPA's NHSRC and NRMRL are developing
CRADAs for the development of technologies specifically to accomplish the nation's homeland
security initiatives.21
During the 2002 Olympic Games in Salt Lake City, UT, the Hach Company (Loveland, CO)
demonstrated the remote transmission of data from a dozen continuously monitoring sensor
platforms in the distribution system. The data were transmitted through unbroken cellular telephone
data transfer or directly to the SCADA system. The data were monitored around the clock with
provisions for triggering alarms.22
An emerging technology project in Copenhagen, Denmark is integrating information from existing
data sources for a water distribution network in a utility. In the pilot project, the data from real-time
sensors are provided to a SCADA system. The sensor information is stored in a database allowing
for analysis. Automated checks of the system compare against baseline measurement. Data are
validated with standard modules that will flag potentially suspect or corrupt data. The project began
in 2001 and is expected to be completed somtime in 2005.
One commercial vendor, PureSense Environmental Inc. (Moffet Field, CA), has a PureSense System
which covers the full spectrum from data acquisition to alert and notification. The system includes
four components. The PureSense iNode™ is a remote data communication device that uses cellular
and Wi-Fi services to collect monitoring data and send commands to remote sensors. The PureSense
iWatch™ is an internet data management system that enables the integration of disparate data sets,
including data from remote online sensors. The PureSense iServe™ allows automated analysis of
real-time data, while PureSense AlertNet™ provides automated alerts. EPA had entered into a
CRADA with PureSense Environmental to test the system.
4.2 Contaminant Flow Predictive Systems
From the beginning of the planning phase, an EWS for water distribution systems should develop
the capability to predict the movement of flow and contaminants in the system. This predictive
ability is important not only to prepare for a potential contamination event, but also to determine the
effectiveness of the monitoring system.
Contaminant flow predictive systems are used to predict how a contaminant would move through
the distribution system. Such systems are built upon hydraulic and water quality modeling
technology known as water distribution system models that are now in wide use within the water
industry. This capability could be applied to accidental contamination events (e.g., backfiow,
cross-connections) or intentional contamination.
August 2005 41
-------
U.S. EPA Office of Water
Early Warning Systems
Hydraulic modeling dates back to the 1930s when Professor Hardy Cross of the University of
Illinois developed an iterative procedure for predicting flow and head within a distribution system
(Cross, 1936). This manual procedure was used throughout the water industry for almost 40 years.
With the advent of computers, computer-based models using the Hardy Cross methodology and
improved solution methods were developed and were in widespread use by the 1980s. These models
have become ubiquitous within the water industry and are an integral part of most water system
design, master planning, and fire flow analyses. In the 1980s and 1990s, hydraulic analysis was
extended to include the ability to model water quality, water age, and source tracing in distribution
systems. The usability of these models was greatly improved in the 1990s with the introduction of
the public domain EPANET model (Rossman, 2000) and other Windows-based commercial water
distribution system models (ASCE, 2004). Exhibit 4-1 provides examples of current water
distribution system modeling software.
In order to apply a hydraulic/water quality model to reliably predict the movement of a contaminant
in the distribution system, a calibrated, extended-period simulation (EPS) model is needed. An EPS
model represents the normal temporal variation in demands and operation. Such a model can be
used for planning purposes as part of vulnerability studies and emergency response plans, and as a
real-time tool during an actual contamination event. Another challenge is the effort to incorporate
pipe networks into the modeling framework, although geographic information systems (GIS) are
assisting utilities with this effort
Exhibit 4-1. Examples of Water Distribution System Modeling Software
Network Modeling
Software
AQUIS
EPANET
InfoWater/H2ONET/H2OMAP
InfoWorks WS
MikeNet
Pipe2000
PipelineNet
SynerGEE Water
WaterCAD/WaterGEM S
Company
Seven Technologies
U. S. EPA
MWHSoft
Wallingford Software
DHI, Boss International
University of Kentucky
SAIC, TSWG
Advantica
Haestad Methods
URL
http://www.7t.dk/company/default.asp
http://www.epa.gov/ORD/NRMRL/
wswrd/epanet.html
http://www.mwhsoft.com
http://www.wallingfordsoftware.com/
http://www.dhisoftware.com/mikenet/
http://www.kypipe.com/
http://www.tswg.gov/tswg/ip/Pipeline
NetTB.htm
http://www.advantica.biz/
http://www.haestad.com
4.2.1 State-of-the-Art Systems and Current Research and Development
Distribution system modeling in general and contaminant flow predictive systems in particular are
developing rapidly. Work is being sponsored by governmental agencies, professional organizations
such as ASCE and AwwaRF, and private companies. Predictive modeling is more extensively
integrated into European utilities than U.S. utilities. The state-of-the-art in contaminant flow
predictive models and some current active areas of research include the following:
August 2005
42
-------
U.S. EPA Office of Water Early Warning Systems
• Integration of models with GIS and computer-aided drafting (CAD) software packages has been
and continues to be an active area of research and development. GIS and CAD were initially
used to help build water distribution system models. Present research and development is
directed towards complete integration of distribution system models into the GIS or CAD
platform in order to facilitate modeling, display, and assessment. PipelineNet, an integration of
EPANET and Arc View GIS, and commercial integration packages such as WaterGEMS and
Info Water provide direct capabilities to assess impacts of contamination events. Additional
efforts by EPA are to expand EPANET to account for multiple interacting species and to better
involve hydrologists and biologists (DSRC Meeting, 2004).
• Calibration of a model involves adjustments in model parameters so that the model reflects
observed field behavior. Optimization techniques for calibrating hydraulic models based on
genetic algorithms and other mathematical methods have recently become part of many
commercial models (Walski et al., 2003). Continued research in this area is underway to expand
the use of these tools to extended period simulation calibration and calibration of water quality
parameters. Another area of development in the calibration field is the use of tracer studies in
distribution systems. A conservative tracer substance such as sodium chloride is injected into
the distribution system and monitored using online conductivity meters. The resulting dataset
can be used to calibrate and validate hydraulic models. Although there are currently no
established calibration criteria in the U.S., static and dynamic calibration methods exist (EPA,
2005). Also, a committee of the AWWA did propose a set of possible calibration guidelines
(ECAC, 1999). These guidelines have not been officially accepted however, there is no active
process underway to adopt them. Using these possible calibration guidelines as a catalyst or
starting point, it would be recommended to move forward on developing accepted calibration
guidelines or standards.
• Existing modeling software is being applied and advanced modeling-based tools are currently
under development to assist in evaluating the vulnerability of a distribution system to
contaminant events. Carlson et al. (2004) demonstrated the use of existing hydraulic models by
creating three case studies of the movement of a contaminant through a distribution system. The
capabilities of PipelineNet have been demonstrated through a series of applications in actual
distribution systems (Bahadur et al., 2003a). It has been applied as a modeling tool for use in
emergency response to contamination of a distribution system. TEVA, a probabilistic
distribution system simulation model, is being developed by EPA to evaluate both vulnerability
and sensor placement in distribution systems (Murray et al., 2004). Van Bloemen Waanders et
al. (2003) have developed a nonlinear programming method for tracing an observed
contamination event in a distribution system back to where it was introduced.
• Recent studies have recognized that there are both uncertainty and variability associated with
distribution system modeling. Application of such models in a purely deterministic framework
where all parameters are assumed to be known with certainty does not provide the information
needed for decision making. Baxter and Lence (2003) present a general framework for the
analysis of performance risk in a water supply. Kretzmann and van Zyl (2004) incorporated
uncertainty through a stochastic analysis of water distribution systems. Grayman et al. (2004b)
used Monte Carlo simulation in estimating contaminant exposures while reconstructing an
unintentional contamination event. The aforementioned TEVA system incorporates a
probabilistic framework for a large range of contamination senarios.
August 2005 43
-------
U.S. EPA Office of Water Early Warning Systems
• Consumption (demand) is an important factor in affecting movement of water and contaminants
in a distribution system. Typically, monthly or quarterly meter readings and approximate diurnal
water use patterns are used in estimating demands in models. This is recognized as inadequate
for detailed contaminant flow models and is being addressed through ongoing research and
development in demand models, demand metering, and customer information systems (CIS).
Li and Buchberger (2004) have developed and applied models using a Poisson Rectangular Pulse
method to simulate fine time scale demand patterns. Metering systems that will measure
customer water use at fine time scales and transmit the information to central locations are on
the market. CISs provide a mechanism for managing consumption data so that they can be used
as a basis for providing better demand data for distribution system models.
• Real-time use of water distribution system models is an active area of development with
applications in the improved operations of water systems leading to savings in energy usage and
water quality, and as a response tool to contamination events (Jentgen et al., 2003). This is
being accomplished through integration of models with SCADA systems that provide
information on operation of a water system on a continuous and real-time basis. Commercial
water distribution system software companies and companies offering SCADA systems are the
primary developers of such systems (Fontenot et al., 2003).
• Tanks and reservoirs have been identified as key elements in water systems that are especially
vulnerable to purposeful contamination. If a contaminant enters a tank in the inflow or through
direct contamination of the tank contents, the manner in which it mixes with the ambient water
in the tank and how it subsequently exits on the outflow affect how and when customers will be
exposed to the contaminant. Various mathematical modeling techniques have been developed
to assist in predicting the mixing within tanks. These include both detailed computational fluid
dynamics (CFD) models and conceptual systems models (Grayman et al., 2004c).
• Water quality models have most commonly been used to represent conservative substances,
chlorine residual, and trihalomethanes. Current research is directed towards improving the
capability to model disinfectants and disinfectant by-products and expanding models to include
bacteria and non-conservative substances that may be introduced in purposeful attacks (Powell
et al., 2004). An upgrade to EPANET will allow the simultaneous simulation of multiple
interacting chemicals in a distribution system (Uber et al., 2004b). For example, this will allow
a simultaneous simulation of both chlorine and a contaminant whose concentration is affected
by the chlorine residual.
• Incident Commanders Water Modeling Tool (ICWater)23 extends the capability of a previously
developed RiverSpill modeling tool to allow an incident commander to analyze and react quickly
to chemical and/or biological contaminants introduced into surface water sources. ICWater will
allow "plug-and-play" with existing incident commander and emergency response tools such as
the Chemical Biological Response Aide, Consequence Assessment Tool Set, Natural Hazard
Loss Estimation Methodology, and the EPA Emergency Response Analyzer.
• The hydraulic portion of flow contaminant models are still based largely on the methods and
assumptions of Hardy Cross developed over 75 years ago. More sophisticated representations
of the hydraulics may be required to adequately predict the behavior of water quality
August 2005 44
-------
U.S. EPA Office of Water Early Warning Systems
contaminants in the distribution system. The research community is starting to re-examine some
of the basic assumptions concerning hydraulic representation of phenomena such as dispersion,
pipe mixing, and flow dynamics (Li et al., 2005).
4.3 Sensor Placement
Historically, monitors and sensors have been placed in distribution systems to meet regulatory
requirements. Their locations have been determined based on ease of access and a general intuitive
assessment of representative locations. Lee et al. (1991) proposed a method for locating monitors
based on the concept of coverage, which is defined as the percentage of total demand that is sampled
by a set of monitors. Various other researchers further addressed this issue using alternative
mathematical methods (Kessler et al., 1998). Though widely cited, these methodologies have rarely
been applied in actual practice. However, following the events of September 11, 2001, there has
been a renewed interest in sensor placement primarily as a mechanism for detecting purposeful
contamination of distribution systems.
4.3.1 Current Research and Development
Many current studies are applying optimization techniques to determine the optimal placement for
monitors in distribution systems based on a defined objective function. Ostfeld (2004) and Ostfeld
and Salomons (2004) provide reviews of past work in this area and present example mathematical
formulations using genetic algorithm solution techniques. Their methodology finds an optimal
layout of an early warning detection system comprised of a set of monitoring stations aimed at
capturing deliberate external intrusions through sources, nodes, or tanks under an extended period
of unsteady conditions. The optimization considers the maximum volume of contaminated water
exposure at a concentration higher than a defined safe level. Berry et al. (2004) used an integer
programming optimization technique to place a limited number of perfect sensors in the pipes or
junctions of a water network so as to minimize the expected amount of damage to the public before
detection, assuming the attack occurs on a typical day. Watson et al. (2004) used a mixed-integer
linear programming model for the sensor placement problem over a range of design objectives.
Using two case studies, they showed that optimal solutions with respect to one design objective
(e.g., population exposed) are typically highly sub-optimal with respect to other design objectives
(e.g., time to detection). The implication is that robust algorithms for the sensor placement problem
must carefully and simultaneously consider multiple, disparate design objectives. Uber et al.
(2004a) describe an iterative numerical solution methodology using the "greedy heuristic" algorithm
for the sensor location problem.
In general, the optimization methods described above are experimental approaches that have been
applied only to hypothetical or small water systems and are based on assumptions about the
availability of monitoring technology, ability to define explicit objective functions, and limited
incorporation of the variability of water system operation. Further research, development, and
practical applications are needed before this technology is ready for routine use.
Research and development of formal methods for locating monitors in water distribution systems
are likely to continue in parallel with the development of monitors that can be effectively placed in
distribution systems. Though most of the cited research is related to monitors to be used to detect
intentional contamination events, related research efforts on sensor location are likely in the future
August 2005 45
-------
U.S. EPA Office of Water Early Warning Systems
to support future water quality regulations. For example, the Initial Distribution System Evaluation
(IDSE) component of the Stage 2 Disinfectants and Disinfection Byproducts Rule that is
approaching promulgation by EPA requires additional routine monitoring in distribution systems.
Similarly, re-evaluation of the existing Total Coliform Rule (TCR) will also focus on sampling
requirements. An AwwaRF sponsored study that is just commencing as this report neared
completion, Methodologies for Assessing and Improving Water Quality Sampling Programs in
Drinking Water Distribution Systems, will focus in part on sensor location in a probabilistic
framework (see Appendix D, Project #3017).
Sandia National Laboratories (SNL) is conducting a project to develop algorithms to identify and
quantify the threat to water systems, and thus determine optimum sensor placement locations. An
additional purpose of the project is to determine the contaminant source location in real-time. The
research team has identified the challenge of determining population density at any node and how
it changes throughout the day. Under an interagency agreement with EPA, SNL is planning to
develop a new set of mathematically based tools for designing an EWS. Scalability to very large
networks and the general uncertainty in the input parameters may be other issues (DSRC Meeting,
2004).
EPA's TEVA program is developing software and tools to examine system vulnerability and help
design response strategies. TEVA has developed a model to indicate sensor location, based on
protecting the greatest number of people. The TEVA program is comparing its model to other
models that have different sensor placement objectives, such as fastest warning or minimizing the
number of attacks that would be missed (system coverage, not based on population served). The
TEVA model uses a statistical approach, by simulating thousands of scenarios (different
contaminant injection locations) and calculating average impact for the entire set of simulations.
The comparative analysis should be completed in 2005.
There is a National Science Foundation project effort underway entitled Placement and Operation
of an Environmental Sensor Network to Facilitate Decision Making Regarding Drinking Water
Quality and Security. The objective of the project is to develop drinking water quality models for
multiple potential chemical and biological threats. It should also improve spatial and temporal
resolutions for sensor collection networks.24
Online sensors are typically installed using special sample taps that require interruption of water
flow through pipes. New installation techniques are being developed to enable installation without
interrupting the water flow or requiring major excavation. Sensors also need to be able to withstand
the harsh environment of different locations (AwwaRF, 2002).
4.3.2 State-of-the-Art Systems
With sophisticated models at their disposal, Bahadur et al. (2003b) pioneered a technique using
PipelineNet in which GIS data and hydraulic model results guide the manual placement of monitors
in order to fulfill certain general criteria. In a case study conducted with personnel at a water utility,
25 potential monitor sites were identified and subsequently reduced to the two best sites using the
GIS/PipelineNet framework. This approach is more closely related to the traditional methods for
locating monitors than are the optimization techniques currently in the research and development
stage.
August 2005 46
-------
U.S. EPA Office of Water Early Warning Systems
Due to budgetary and technical constraints, many water utilities will face a common situation— the
utility wishes to make a modest initial investment in sensors within its existing distribution system
and wants to pick the most appropriate locations. Without resorting to the sophisticated
experimental optimization techniques that have been previously described, a two-stage procedure
is typically followed. In the first stage, likely sites for sensors are determined based on
technological constraints for available sensors. Most current sensors require an external power
source, a secure location with protection from the elements, access to communications, and easy
accessibility for maintenance. These constraints typically result in a limited number of potential
sensor locations. In the second stage, the potential sites are evaluated in terms of the "information
content" that will be provided if a sensor is placed at that location. This typically translates into
placing sensors throughout the system on larger pipes that serve the most customers. This process
can be done informally by operations personnel who know and understand the distribution system
or more formally, using hydraulic models to identify high flow pipes. This procedure, in effect,
emulates the methodology developed by Professor Deininger at the University of Michigan (Lee et
al., 1991) based on the concept of coverage that is defined as the percentage of total demand that is
sampled by a set of monitors.
4.4 Alert Management Systems
Alert management systems consist of two general areas: (1) establishing parameters for alert triggers
and (2) reducing false alarms. During the data analysis phase, new data points are compared to
baseline data values. The baseline should incorporate all potential variations due to seasonal water
quality fluctuation and cover typical operational changes (Carlson et al., 2004). It may be necessary
to incorporate up to a year's worth of data into the baseline for it to adequately capture these
variations. The baseline must also differentiate between single data versus multiple data streams
(i.e., one physical/chemical parameter versus multiple parameters), but the amount of baseline data
needed will vary according to the technology used and the statistical variance of the data (AWWA
Workshop, 2004). Any anomalies in the comparison of new data to the baseline trigger alerts to the
operator. For example, Hach Company (Loveland, CO) has developed a trigger algorithm to create
an alarm when conditions in water depart from expected baseline parameter values. Because utilities
regularly experience changes in water quality, there are always concerns regarding false-positives.
Alert management systems usually rely on strict data validation protocols or specialized software
to reduce false alarms. PureSense Environmental, Inc. (Moffet Field, CA), for example, created a
software product to reduce false-positive and negative readings in standard water quality sensors.
EPA had entered into a CRADA with PureSense Environmental25 to test the system to determine if
the software can reduce false signals in EWS in water distribution networks. PureSense systems are
currently in use in public water systems and in the U.S. military.
4.5 Integrating Water Distribution Modeling and Data Acquisition Systems
The following systems comprise the state-of-the-art in integrating water distribution modeling and
data acquisition to support the goals of an EWS.
MIKE NET-SCADA unites EPA modeling software and SCADA systems in an effort to optimize
system performance to recognize and respond to alarm conditions. The system's online module
performs real-time comparisons of the measured and calculated data, automatic data pre-processing
August 2005 47
-------
U.S. EPA Office of Water Early Warning Systems
for the off-line module, and pressure/flow calculations at any point of the system. The model results
are stored back into the SCADA database and the online viewer is used to display detailed model
results (Fontenot et al., 2003). In addition, the online module features automatic data validation
procedures, in which all measurements are automatically checked and validated with standard
modules. These modules will tag questionable data and, if possible, fill in gaps in the time series.
This ensures that only validated data will be transferred and used as boundary conditions in the
strategic model, decreasing the potential for false alarms (Fontenot, 2003).
MIKE NET-SCADA' s off-line module models IF-THEN scenarios, models system breakdowns, and
predicts system behavior using demand and control rules prediction. It uses Microsoft Access to
store and maintain model alternatives. Coupling the online and off-line results of MIKE
NET-SCADA allows the operator to quickly detect abnormalities and help analyze ways in which
the abnormality can be remedied or its impact minimized (Fontenot, 2003).
Clarion Sensing Systems' (Indianapolis, IN) Sentinal™is a remote computing platform that features
logical data processing at the monitoring sites and compatibility with various forms of wireless and
wired data transmission. The system integrates sensor data into a single display that presents
information through the Internet, a local area network, or a local terminal. The data are presented
in a web page format with analytical and historical data storage capability. Each monitoring site has
its own Internet Protocol address and serves its own web page to allow for specific site monitoring
and remote configuration of the water quality profile of the site. The Sentinal™ system can be
integrated into an existing system such as SCADA, and its software is compatible with spiral
development approaches, since new sensor technologies can be integrated into the system (Martin
Harmless, Clarion Sensing Systems, personal communication).
AQUIS is a water network management system designed for both on- and off-line, real-time
monitoring. The software, produced by Seven Technologies (Denmark),26 is used to create models
to efficiently manage water resources. The models allow utility managers to minimize the impact
of operational disruptions in order to maintain continuity and quality of service. The software also
allows managers to explore strategies for responding to emergencies, including the introduction of
contaminants and increased demands placed on the system by extensive fire-fighting or other surge
demands. AQUIS is currently in use in 1,500 cities worldwide.
AQUIS offers a Contingency Management Software Package that has five modules designed to
establish a point of entry for contaminants, determine a method for limiting the spread of the
contaminant, and determine methods to mitigate any harmful effects. The modules include a model
manager for GIS data management and a hydraulic module for modeling throughout the distribution
system. A water quality module tracks the chemical composition of the water throughout the
system, and a diagnostic module identifies the source of contaminants. Finally, a flushing module
facilitates cleaning the distribution network.
4.6 Data Security
Utility management is provided routine characterization of water through daily, weekly, and
monthly reports. Monitoring data are often collected by SCADA information and then analyzed.
A major concern is that most data are not secure or encrypted by SCADA and could be subject to
security breaches. Thus, security enforcement is a major issue for all EWSs to ensure the integrity
August 2005 48
-------
U.S. EPA Office of Water Early Warning Systems
of the system. There are a number of security considerations related to the design of an EWS.
Hardwired data transmission systems have the benefit of being more secure than wireless, because
it is harder to obtain the physical connection to the cable required for data interception (AWWA
Workshop, 2004). Hardwired systems also may be more robust during a crisis situation because
there could potentially be issues with wireless network availability. A number of wireless
transmission systems rely on an outside network, which makes it difficult to ensure their security.
Similarly, Internet-based software applications like those in the Sentinal™ system are vulnerable
to viruses and hackers. Ideally, the SCADA system should be isolated from other systems to avoid
competition for bandwidth and prevent potential system crashes (Carlson et al., 2004).
In addition to a secure design, the water distribution facility should establish security policies to be
followed by all personnel and develop a security module for the EWS. Policies should restrict data
sharing and document access to the public, especially concerning sensitive information such as
individual sensor locations. Need-to-know requirements and data quality and integrity objectives
form the basis for these security policies (Mays, 2004).
The security module includes three areas of protection:
• Authentication. A password-based authentication mechanism should be employed for users
to access the sensor data and document databases.
• Access Control. A fine-grained access control system is capable of specifying access control
permissions based on the users' credentials, which are verified electronically.
• Secure Data Transfer. Encrypted communication using protocols such as SSL (Secure
Socket Layer) helps ensure confidentiality and integrity of the sensor data and documents
during their transit from the sensors to the system, as well as from the system to the users.
4.7 Communications, Response, and Decision Making
EPA's Response Protocol Toolbox (EPA, 2003/2004) provides guidance to utilities and response
agencies for evaluating, communicating, and responding to threats to the water system. Managing
the threat is based on the Incident Command System (ICS) in which the utility names a Water Utility
Emergency Response Manager (WUERM) to be the initial Incident Commander. EPA's guidance
specifically outlines a threat management system in which three main threat levels occur, "possible,"
"credible," and "confirmatory." At each level, there are evaluation, notification, and response
suggestions. Although information from many sources (e.g., outside law enforcement) can elevate
the threat from one level to another, the focus in this study is on how information on water
contamination provided by EWSs or subsequent laboratory analysis elevates the threat level.
A "possible" threat exists with the first sign of unusual water quality that significantly differs from
established baseline data. This information could be provided by multi-parameter water quality
monitors. The data should be compared with other monitoring locations to determine if changes in
source water could be the cause of the unusual data. Such possible classifications result in
notifications within the utility. Response may include identifying sites and initiating a site
characterization in order to rapidly test the water and collect samples for possible delivery to a
laboratory. Other responses include investigating unusual consumer complaints and consulting with
external information sources.
August 2005 49
-------
U.S. EPA Office of Water Early Warning Systems
A "credible" threat exists when additional information collected during the site characterization, as
well as other factors, corroborate the threat warning. Evaluation at this stage should determine if
the unusual data are substantially different from other water quality episodes, if the unusual water
quality is indicative of a specific contaminant, and if the unusual water quality is clustered in a
specific area. "Credible" threats result in notification of the drinking water primacy agency, state
and local public health agencies, local law enforcement, and the Federal Bureau of Investigation
(FBI). At this stage, appropriate response should estimate the affected area and isolate it if possible,
implement appropriate public health protection measures, and further analyze samples from the site
characterization in a laboratory.
"Confirmation" represents the transition from a credible contamination threat to a confirmed
contamination, based on definitive information that the water has been contaminated. Often at this
stage, the contaminant information is available. Emergency response agencies, as well as the
National Response Center, are notified. The WUERM is no longer responsible for incident
command at this stage, but still plays a vital role in helping other agencies. An external source that
may help determine the "confirmatory" stage is the WCIT, which is currently in development by
EPA. At this point, the local Emergency Operations Center (EOC) may be fully activated in order
to support an effective and coordinated response. All of the participating organizations will likely
be coordinated under existing incident command structures designed to manage emergencies at the
state or local level. One agency will be designated as a lead agency and will be responsible for
incident command. Public health protective measures are revised as necessary and may include a
"boil-water" notice, a "do not drink" notice, or a "do not use" notice, which involves consideration
of an alternate water supply for consumption, sanitation, and other uses.
Water utilities and public health officials should develop specific criteria for making important
notifications and to identify key contacts within each agency. These criteria will ensure that
effective communications occur and appropriate public health response actions are taken in the event
of a water-related public health incident. EPA recommends the use of the Water Information
Sharing and Analysis Center (WaterlSAC) for rapid information exchange.27 The WaterlSAC has
secure portals and established protocols for handling information. A WaterlSAC security analyst
follows up on incident reports. When responding to the threat, three factors should be considered:
(1) the credibility of the threat, (2) the potential consequences of the contamination incident, and
(3) the impact of the response action on consumers.
AwwaRF, with its partner the Water Environment Research Foundation (WERF), is conducting
research to assist utilities in efforts to communicate, respond, and make decisions during a disaster
such as a terrorist attack. In one research effort, AwwaRF/WERF is developing written and oral
message statements to be used by public agencies and elected officials to communicate with the
public in the event of a water contamination threat (see Appendix D, Project #3046). They will also
include an action plan for working with the public to increase public awareness of potential public
health risks, and appropriate responses. Another research effort is to provide a decision support
system for water distribution system security. The effort, involving Colorado State University and
Advanced Data Mining, will provide a broad and substantial knowledge base for utilities about the
impact of a toxicological attack on a distribution network and cost-effective approaches for detecting
and mitigating such attacks (see Appendix D, Project #3086).
August 2005 50
-------
U.S. EPA Office of Water Early Warning Systems
Summary reports containing monitoring data could assist in the response. Such reports are
increasingly being developed in electronic format. To support the communications of contamination
information, there are the beginnings of web-based communication of results to critical response
agencies (AwwaRF, 2002). Data communication between online monitors is becoming part of a
"knowledge-based" management environment (Rosen et al., 2003). Furthermore, water utilities
serving more than 3,300 people have developed emergency response plans under the Public Health
Security and Bioterrorism Preparedness and Response Act of 2002. These plans help in facilitating
response and decision making.
August 2005 51
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
5. Multi-Parameter Water Quality Monitors as Candidates for Early Warning
Systems
As noted elsewhere in this report, a first-stage approach to online screening for contaminants has
typically been the use of readily available online sensors that measure simple physicochemical
parameters typical for monitoring water quality (e.g., temperature, pressure, pH, conductivity,
chlorine residual). A second-stage is analysis that confirms and identifies the contaminant. Exhibit
3-3 provides a more complete discussion of the Stage One/Stage Two approach to an EWS.
Technologies to continually measure basic water quality parameters have been commercially
available and widely used by utilities for some time. They are used for process control and to ensure
regulatory compliance. These technologies are rapid and relatively easy to use on a continuous basis
via remote access to data and are available from a variety of manufacturers. Vendors have
developed panels of sensors that monitor multiple water quality parameters. The most basic
application for such sensors is to detect a physical and/or chemical change in water quality (e.g.,
change in state) that might suggest that a contaminant had been accidentally or intentionally added
to the water. The goal is to use multi-parameter water quality monitors to provide an online early
warning or red flag of an unspecified contaminant. This has been referenced by some in the field
as a Stage One EWS (Hasan et al, 2004).
A more advanced application for the use of routine physicochemical sensors is the attempt to
establish a characteristic pattern of changes in multiple parameters that might be used to actually
presumptively identify the contaminant. Such a characteristic pattern is often termed a contaminant
signature. This application is currently in the development and testing phase and is being actively
pursued by several manufacturers and government research groups. Applications utilizing routine
multi-parameter water quality monitors (e.g., simple detection of change of state as well as
establishment of a contaminant signature) are described in this section.
5.1 Descriptions of Various Multi-Parameter Water Quality Monitors
Typical multi-parameter water quality sensors for finished water have the following types of water
monitoring methods:
• colorimetric and membrane electrode for chlorine
• thermistor for temperature
• membrane electrode or optical sensors for DO
• potentiometric method for ORP
• glass bulb electrode for pH
• nephelometric method or optical sensor for turbidity
• conductivity cell method for specific conductance
• ion-selective electrodes for Cl", NO3 and NH4+
In addition to basic screening of single parameters by individual sensors, some vendors are now
offering preassembled packages consisting of several conventional water quality sensors. The
following are a few example multi-parameter platforms of water quality sensors. All of these are
considered available technologies except for STIP-Scan, which is designed for wastewater
applications but is potentially adaptable for drinking water. EPA does not endorse or recommend
August 2005 53
-------
U.S. EPA Office of Water
Early Warning Systems
any of the following technologies. The summary information below was obtained from company
websites, promotional literature, and personal communication with company representatives.
The Hach Corporation (Loveland, CO) is marketing the Water Distribution Monitoring Panel. This
panel combines established instrumentation into a preconfigured system for more comprehensive
monitoring. The basic model includes the following:
• Hach CL17 Chlorine Analyzer
• Hach 1720D Low Range Turbidimeter
• Hach/GLI pH Controller
• Hach/GLI Oxidation Reduction Potential Controller
• Hach/GLI Conductivity Controller
• GEMS Pressure Sensor
The expanded model also includes a Hach Astro UV TOC analyzer and an American Sigma 900
MAX auto sampler that can be activated to collect and archive samples when pre-specified setpoint
values are exceeded for any of the parameters being measured. The Hach Distribution Monitoring
Panel is designed to continually measure these six to seven physicochemical parameters from a side
stream of water in a municipal distribution system, and the results can be reported directly to the
utility SCADA system.29
In addition to the system H
designed to continually sample a
side stream of water, the Hach
Corporation is also marketing a
multi-parameter probe that is
installed directly into a water
distribution pipe. The Hach
Water Distribution Monitoring
PipeSonde In-Pipe Probe can be
installed into any water pipe (at
least eight inches in diameter),
via a two-inch corporation stop
(ball valve), and is designed to withstand water pressures of up to 300 psi. The Pipe Sonde can
measure the following parameters: pressure, temperature, conductivity, turbidity, ORP, DO, and
chlorine concentration. A sample port is available for attachment of an autosampler and a TOC
analyzer. As was the case with the Hach Water Distribution Monitoring Panel, the PipeSonde In-
Pipe Probe can be configured to communicate directly with a utility's SCADA system. The Hach
Event Monitor Trigger System allows for real-time analysis of data from the Water Distribution
Monitoring Panel, PipeSonde In-Pipe Probe, and the online TOC analyzer. It triggers an alarm when
water quality deviates from a baseline. It can thus profile and catalog events. The trigger signal and
all parameter measurements can be viewed from the main touch screen interface30. (Image
reproduced with permission from Hach Company.)
Dascore, Inc. (Jacksonville, FL) markets a multiarray sensor called Six-Cense™. The Six-Cense™
system, like Hach's Pipe Sonde, is designed for permanent insertion into a pressurized water main.
However, unlike the Hach products, Six-Cense™ consists of electrochemical sensors mounted on
Water Di'>fributiou
Mouitoriug Pnuel (Hacli)
Pipt&onde (Hack)
Eveut Monitor Trigger
(Huh)
August 2005
54
-------
U.S. EPA Office of Water
Early Warning Systems
a one square inch ceramic chip layered with gold. Measurement is accomplished via
electrochemical methods, rather than through the use of reagents. The system can continuously
monitor six parameters including chlorine or chloramine, DO, pH, ORP, conductivity, and
temperature. The system can be operated remotely with data reported to the utility SCADA
system.
31
Emerson Process Management - Rosemont
Analytical (Columbus, OH) is marketing a
continuous-monitoring system for freshwater and
water distribution networks. The Model WQS
Multi-Parameter Electrochemical/Optical Water
Quality System (Model 1055 Solu Comp II)
continually analyzes a low flow (< 3 gal/hr) side
stream of water. No reagents are utilized. Six
parameters are analyzed by electrochemical
methods (pH, conductivity, ORP, DO, free
chlorine, and monochloramine). Two parameters
are assayed via optical methods (turbidity and
particle index). The particle monitor counts
particles utilizing optical laser technology, and
reports particle concentration as a particle index.32
(Image reproduced with permission from Emerson Process.)
Model 1055 Solu Comp Analyzer
i Z LnfT'oiL Process)
(Y&I Environmental)
YSI Environmental, Inc. (Yellow Springs, OH) produces standard
equipment to monitor drinking water for ORP, DO, pH, specific
conductance, and temperature. In addition, YSI systems measure turbidity,
and levels of chloride, ammonia nitrogen, and nitrate nitrogen. YSI
currently uses its technology in surface water applications. (Image
reproduced with permission from YSI Environmental.)
Analytical Technology, Inc.'s, (Collegeville, PA) Series C15 Water
Quality Monitoring system allows the user to choose those parameters for
which monitoring is desired and to integrate those components into a
monitoring package suitable for continuous monitoring, alarming, and data
collection. System components are currently available for free chlorine, combined chlorine (for
chloramine treated systems), dissolved ozone, pH, ORP, conductivity, and temperature. In addition,
DO and turbidity modules will be added to the
system in the future.33
Clarion Sensing Systems' (Indianapolis, IN)
Sentinal™ integrates sensor data into a single
display which can be viewed remotely (e.g., over
the Internet). Clarion sells the complete system
including sensors, or can integrate a utility's
existing sensors from a variety of manufacturers.
The system is modular, so utilities can select to
monitor various parameters including: chlorine,
SentinaP" (Clarion Sensing Systems)
August 2005
55
-------
U.S. EPA Office of Water Early Warning Systems
pH, temperature, flow, pressure, conductance, turbidity, ORP, DO, radiation, TOC, VOCs, and
certain chemical weapons. The Sentinal™ software is compatible with spiral development
approaches, since new sensor technologies can be integrated into the system. The system can run
off of AC current or solar power and automatically re-boots if power is interrupted. Data can be
transmitted via LAN or satellite link (Martin Harmless, Clarion Sensing Systems, personal
communication) (Image reproduced with permission from Clarion Sensing Systems.)34
STIP-scan from STIP Isco GmbH (Germany) analyzes multiple wastewater parameters with a single
device. Although designed for wastewater applications, the equipment is potentially adaptable for
drinking water distribution systems. Designed to operate in municipal and industrial wastewater
treatment plants at the inlet, in the aeration basin, and in treated effluent, the STIP-scan UV/Vis
spectroscopic sensor is capable of simultaneous measurement of nitrate, chemical oxygen demand
(COD), TOC, spectral absorption coefficient (SAC254), total solids, sludge volume, sludge volume
index, and turbidity. It can also be used for river monitoring. In addition to these parameters, the
STIP-scan measures absorption in any specified range within the wavelength spectrum from 190 to
720 nm for detection of other compounds. No sample filtration or preparation is required, and the
wiping action of the piston seals cleans the measuring cell on each cycle. The controller is equipped
with analog outputs and a bidirectional serial interface to transmit data. A color display presents the
data as continuous daily graphs of nitrate, COD, TOC, or SAC254. Data intervals are two minutes
and can be stored up to 14 days.35
5.2 Efforts to Determine Performance of Multi-Parameter Water Quality Monitors and
Establish Water Quality Baselines
During the process of developing a workable E WS based on multi-parameter water quality monitors,
there are various validation steps that are occurring simlutaneously. These include efforts to
determine the performance of multi-parameter water quality monitors as well as establishing
operating water quality baselines so that anomalies can be spotted. Several tests have and are being
performed at EPA. Some efforts have occurred under CRADAs. Under the Federal Technology
Transfer Act, EPA can form CRADAs with the private sector in order to speed the development of
technologies for various programs. In forming these CRADAs, EPA is aiming to continue research
into the detection and identification of contaminants, response and mitigation, and prevention and
protection. Private industry, as well as state and local governments, can utilize CRADAs to access
federal laboratory equipment, personnel, and services.36
EPA has used its T&E Facility in Cincinnati, OH, for much of the past research. The specific center
is called the WATERS. Within the WATERS Center, there are multiple DSSs, which were designed
and fabricated by the NRMRL to evaluate and understand the dynamics that influence water quality
within water distribution infrastructure systems typical in the U.S. and abroad.37 EPA research
investigators selected a platform array of real-time online sensors and instrumentation that represent
the various types of technologies currently used by utilities to monitoring water quality.
Experiments are underway to evaluate the selected online water monitoring sensors in the various
DSS units for their ability to detect changes in water quality due to chemical, physical, and microbial
contaminants at concentrations that would present a public risk in a distribution system. Water
quality sensors utilized for the EPA WATERS Center studies can be classified in two
groups—traditional sensors and continuous monitors. EPA is conducting studies to evaluate the
sensitivity, response, limit of detection, reproducibility, potential for false-positives/false-negatives,
August 2005 56
-------
U.S. EPA Office of Water Early Warning Systems
and other limitations of the selected sensors. The individual conclusions of these evaluations are
discussed in Chapter 9.
5.2.1 Evaluate Sensor Performance
An important part of developing a first stage EWS is to evaluate whether the normal operation of
water distribution systems can be documented in terms of sensor response. At the WATERS Center,
EPA has conducted tests to understand which sensors can determine a baseline water quality and
if sensor drifting takes place. A basic conclusion was that specific-conductance, TOC, and free
chlorine monitors drift very little when properly calibrated and serviced, and therefore these sensors
are ideal for characterizing normal or safe conditions (EPA, 2004).
5.2.2 Investigate Baseline for Water Quality
The USGS New Jersey District, EPA, and a local water utility, under an Interagency Agreement,
will plan a study to implement and test an EWS in an actual water distribution system. The study
will test sensors, optimize the sensor location, and develop a baseline water quality profile of the
distribution system (DSRC Meeting, 2004). AwwaRF has two projects that analyze online water
quality data to address normal fluctuations in water quality parameters and develop methods to
differentiate them from contamination events (see Appendix D, Projects #3035 and #3086).
5.2.5 Verify Sensor Performance
Another series of tests is currently being conducted at the EPA WATERS DSS facility to verify the
capability of multi-parameter water monitors for normal daily operation of distribution systems. The
testing is being performed under the auspices of EPA through the ETV Program. The work is being
performed by Battelle Laboratories (Columbus, OH), which is managing the ETV AMS Center
through a cooperative agreement with EPA. Vendor representatives are installing, maintaining, and
operating their respective technologies throughout the test (Ryan James, Battelle, personal
communication38) A CRADA with EPA will test YSI technology in water distribution systems to
determine how applicable the technology is to online monitoring of water distribution. This project
is separate from the research project described below, because the sensors are being verified for their
ability to perform as basic water quality monitors, and their response to injected contaminants is not
being tested.
5.3 Efforts to Provide Red Flag EWS and to Identify Specific Contaminants via Signature
Using Multi-Parameter Water Quality Monitors
There is an increasing level of activity to examine the use of multi-parameter water quality monitors
to provide a red flag EWS and to identify the actual contaminant using signatures. Historically,
utilities have invested in multi-parameter water quality monitors to enhance the management
capabilities of drinking water treatment plants' and distribution systems' daily operations. If these
same monitors prove to be useful for detecting even a subset of intentional contamination events,
then the monitors would serve dual purposes. If the same equipment can be used for both daily
operations and for detecting system perturbations due to intentional contamination, then their value
to utility companies would be enhanced. Further details of the evaluation are provided in Chapter 9.
August 2005 57
-------
U.S. EPA Office of Water Early Warning Systems
5.3.1 Sensor Response to Contaminants
At the WATERS Center, EPA investigated whether various sensors could identify contaminants.
The conclusion was that certain sensors can provide only a general indication of the contaminant
class (such as inorganic, organic, or a reactive species producing chlorine demand) (EPA, 2004).
5.5.2 Multi-Parameter Response to Chemical or Biological Agent Simulants
Another study at EPA's WATERS Center investigated the response of a combination of off-the-shelf
sensors to detect changes caused by the injection of wastewater, groundwater, a chemical mixture,
and individual chemicals. The sensor system showed promise for providing quick detection of water
quality changes caused by these contaminants. With additional optimization, a system of sensors
may be used as an EWS (EPA, 2004). However, because the range of contaminants was narrow,
additional types of contamination and scenarios should be examined to further test this conclusion.
5.5.5 Signature Development
An advanced application of conventional physicochemical sensors for security monitoring is the
presumptive identification of a specific contaminant through interpretation of a characteristic pattern
of changes (signature) in multiple parameters. Typically, the interpretation of such a pattern of
changes is accomplished with the help of a computerized data system. Just as in the previously
described effort to infer the occurrence of an unidentified contaminant by observing a
physicochemical change, a reliable parameter baseline must first be established for successful
interpretation of sample results. Furthermore, when attempting to actually identify a contaminant
through observation of a characteristic pattern of changes in parameter values, it is also necessary
to characterize the expected changes in multiple parameters ahead of time. The signature data for
the characterization would need to be obtained empirically. Several manufacturers are exploring
this signature approach for water contaminant monitoring. One company, Hach (Loveland, CO),
has tested this signature approach.39 The signatures being developed by Hach and others to identify
contaminants or classes however, are difficult to independently validate because their methods and
algorithms are not available to the public. Also, the examination of water quality parameters for use
in detecting and identifying contaminants is still being evaluated by EPA, USGS, the Army, and
other organizations, and there has yet to be implemented a field-scale test of a full EWS with these
water parameter components. These add to the caution in currently recommending use of these
water quality parameter-based EWSs.
5.3.4 Further Testing of Signature Concept
To quickly facilitate the development and use of multi-parameter water quality monitors as part of
an integrated EWS for water distribution systems, EPA has entered into a CRADA with Hach
Company. Hach Company currently produces a real-time water distribution monitoring panel that
is composed of several different types of sensors. The CRADA will determine if this technology
can be adapted to real-time monitoring of water distribution systems to detect contaminants such as
pesticides, herbicides, industrial chemicals, and wastewater.
August 2005 58
-------
U.S. EPA Office of Water Early Warning Systems
5.3.5 Multi-Parameter Response to Actual Chemical and Biological Agents
Working with EPA, the Edgewood Chemical Biological Center (ECBC) is planning to test multi-
parameter water quality monitors using actual CBR Agents.
August 2005 59
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
6. Technologies that Detect Chemical Contaminants for Early Warning Systems
6.1 General Introduction to Assays and Sensors
The technologies presented in this chapter are for the detection of chemical contaminants. However,
several of the technologies are also capable of detecting microbial pathogens. For EWSs, second-
stage confirmation technologies can come mostly in the form of portable field kits or hand-held
sensor devices. Online and kit assay biomonitors that measure effects on biological organisms are
also included even though they do not identify specific contaminants. Online gas chromatography
and mass spectroscopy (GC-MS) is also included, and although it may be too expensive for
continuous operation, it would be valuable for remote identification when triggered by a stage one
alarm (e.g., red flag). Kits are generally portable versions of bench top assays that require pipetting,
mixing, and reaction containers. They may also require a reader device for monitoring the assay
reaction. Sensors and detection devices can be based on a variety of technology platforms, many
of which are described in this chapter and Chapter 7, and can be of a suitcase, backpack, or hand-
held size.
EPA does not endorse or recommend any of the following technologies. Much of the summary
information below was obtained from company websites, promotional literature, personal
communication with company representatives, and some government sources.
6.2 Available Technology
6.2.1 Detection of Arsenic
To evaluate the presence and concentration of arsenic in water there are two basic types of
technologies that are used in commercially available tests, which have been third party verified by
EPA's ETV Program (see Chapter 9 for verification results). Both types of technologies are used
in portable devices designed for onsite rapid analysis of arsenic in water. The first type involves a
color reaction kit in which a water sample is mixed with a series of reagents producing a color
change in an indicator, which is then compared to a standard color gradient that corresponds to the
concentration of arsenic in the water.
Industrial Test Systems, Inc. (Rock Hill, SC) offers five Quick™ test reaction kits that identify
arsenic levels at varying concentrations, depending on which kit is used. For these products the
indicator strip, in addition to being read visually, can be read with the hand-held instrument or with
a portable scanner and laptop system. Both operate on the same principle as a colorimeter, provide
a quantitative result, and are available separate from the test kits. All five Quick™ tests are easy
to use and easily transportable to the field. The time to analyze one sample is approximately 15
minutes. Peters Engineering (Austria) offers the AS 75 arsenic test kit, another color reaction kit,
which is also field portable. It measures the color change in a filter after reagent tablets are dropped
into the sample by visual comparison to a color chart, or by the battery operated AS 75 tester.
Another color reaction test kit is the As-Top Water test kit offered by Envitop, Ltd. (Oulu, Finland).
The kit is easily transportable to the field and analysis takes 35 minutes (EPA-ETV, 2004).
A second type of test to evaluate arsenic concentrations employs anodic stripping voltammetry
(ASV), which can be used to analyze various analyte metal ions. The measurement is performed
August 2005 61
-------
U.S. EPA Office of Water Early Warning Systems
in an electrochemical cell. A reducing potential is applied to the working electrode. When the
electrode potential exceeds the ionization potential of the analyte metal ion in solution (in this case,
arsenic), it is reduced to the metal which plates on the working electrode surface. The analyte is
then stripped (e.g., oxidized) off the electrode by an applied potential. The electrons released by this
process form a current, which is measured and can be plotted as a function of applied potential to
give a "voltammogram." The current at the oxidation/stripping potential is read as a peak, and to
determine analyte concentration the height or area of the peak can be measured and compared to that
of a known standard solution (EPA-ETV, 2004).
Monitoring Technologies International Pty., Ltd. (Perth, Australia) offers the PDV 6000 portable
analyzer which measures arsenic in water using ASV. For this device, the sample concentration
result is provided as a digital readout on a hand-held controller, or can be measured using VAS
Version 2.1 software. The PDV 6000 portable analyzer is easily transported from the field to a
storage shed where samples are analyzed. Instrument setup and calibration takes about 30 minutes
and analysis time per sample is about five minutes. Another device that employs ASV technology
is the Nano-Band™ Explorer made by TraceDetect (Seattle, WA). This device has a three-electrode
cell which combines a Nano-Band™ Explorer electrode with a reference and auxiliary electrode.
The samples take one hour to prepare prior to analysis, then the concentration is detected within
seconds and displayed in real-time using software run on a laptop computer. The Nano-Band™
Explorer is optimized for trace metal analysis and allows for detection of some metals as low as
0.1 parts per billion (ppb). The Nano-Band™ Explorer measurement system includes Explorer
software, but not a laptop computer (EPA-ETV, 2004).
6.2.2 Detection of Cyanide
To evaluate the presence and concentration of cyanide in water there are two basic types of
technologies in use by commercially available tests, both of which have been third party verified by
EPA's ETV program (see Chapter 9 for verification results). Both types of technologies are used
in portable devices designed for onsite rapid analysis of cyanide in water. The first is a portable
colorimeter in which a sample and reagent(s) are mixed producing a color whose intensity is
proportional to the cyanide concentration. The color is measured photometrically to provide a
quantitative determination of cyanide in the sample (EPA-ETV, 2004).
The VVR V-1000 multi-analyte photometer was used with the V-3803 cyanide module and self-
filling reagent Vacu-vial® ampoules, all by CHEMetrics, Inc. (Calverton, VA) to test for cyanide
concentrations. The CHEMetrics VVR operates on four AA batteries, was easy to operate, and was
easily transported to the field. The 1919 SMART 2 Colorimeter with the 3660-SC Reagent System
by LaMotte Company (Chesterton, MD) was also tested by ETV. The LaMotte SMART 2 operates
at 120V/60Hz and was easy to use and transport to the field. Analysis for one sample takes
approximately 22 minutes. The Mini-Analyst Model 942-032 by Orbeco-Hellige (Farmingdale, NY)
also operates on four AA batteries. Analyzing one sample takes approximately 18 minutes. The
AQUAfast® IV AQ4000 colorimeter by Thermo Orion (Beverly, MA) automatically identifies the
species to be measured and selects the method, wavelength, and reaction time. It also operates on
four AA batteries and can measure cyanide concentration when used with AQ4006 cyanide reagents.
The AQ4000 was easily transported to the field and clear instructions assured easy operation.
However, sample throughput for only one sample would take approximately 17 minutes (EPA-ETV,
2004).
August 2005 62
-------
U.S. EPA Office of Water Early Warning Systems
The second basic technology type that is used to measure cyanide concentrations in water is found
in a device that consists of a solid sensing element containing a mixture of inorganic silver
compounds bonded to the tip of an epoxy electrode body. When the sensing element is in contact
with a cyanide solution, silver ions dissolve from the membrane surface. Silver ions within the
sensing element move to the surface to replace the dissolved ions, establishing a potential difference
that is dependent on the cyanide concentration in the solution. Upon calibration with solutions of
known cyanide concentrations, these potential differences are converted to concentrations and
displayed on the digital readout at milligrams per liter (mg/L) (EPA-ETV, 2004).
One device that employs this technology is the Thermo Orion Model 9606 Cyanide Electrode with
the Model 290A+ Ion Selective Electrode Meter, which operates on a 9-volt battery. The Thermo
Orion ISE was easily transported to the field and the instruction manual for its use was clear and
concise. Sample preparation takes 1-2 minutes per sample, calibration takes between 15 to 20
minutes and each sample took approximately 5 minutes to attain a reading. The Cyanide Electrode
CN 501 with the Reference Electrode R503D and Ion Pocket Meter 340i (WTW ISE) by WTW
Measurement Systems (Ft. Myers, FL) operates on four AA batteries and was easily transported and
used in the field setting (EPA-ETV, 2004).
6.2.3 Gas Chromatography
GC can be used to analyze for a wide variety of organic chemicals such as industrial chemicals and
components of fuel oils. Individual organic compounds are separated as the compounds are carried
by a carrier gas (e.g., nitrogen, helium, argon, hydrogen) through a packed or coated column
containing a stationary phase. The columns are coiled in an oven. At the beginning of the coil, the
compound is vaporized into a gas. The compounds separate from each other within the column due
to differing affinity among the sample compounds for the gas and liquid phases. As a result, the
individual compounds travel through the column at different speeds. The time required to pass
completely through the column and reach the detector varies from compound to compound. The
components of a complex sample matrix can be separated, compared to known standards, and the
concentrations then quantified. For smaller devices, photolithographic machining techniques are
used to produce injection and detection systems on silicon microchips (see Section6.3.5 for general
description of microchips). GC alone can only provide tentative identification. For more definitive
identification, traditional and microfabricated GC columns can be coupled to detectors such as
thermal conductivity detector (TCD), surface acoustic wave (SAW) detector, electrolytic
conductivity detector (BCD), electron capture detector, flame ionization detector (FID), photo
ionization detector (PID), and mass spectrometer. Volatile organic compounds (VOCs) can be
extracted and concentrated from water samples using purge and trap technology. The volatile
compounds are initially purged from the water sample using a purge gas such as helium and are
sorbed onto an organic resin trap. The compounds are subsequently desorbed from the trap by flash
heating and enter the GC column. Purge and trap gas chromatography has been used for years to
monitor raw waters such as the Ohio River and Rhine River for industrial spills (ILSI, 1999). In
some cases analyses are conducted on grab samples one or more times each day. In other cases
sample collection has been automated and occurs at regular intervals throughout a 24-hour period.
The units require skilled operators and regular maintenance.
One commercially available automated gas chromatograph that is being used as a continuous online
monitoring system for the analysis of VOCs is the Scentograph CMS500 manufactured by INFICON
August 2005 63
-------
U.S. EPA Office of Water
Early Warning Systems
(East Syracuse, NY). It can operate automatically and unattended providing real-time
measurements. The system can measure concentrations ranging from parts per million (ppm) to
parts per trillion (ppt) with sampling and reporting of results at programmable intervals. The
Scentograph CMS500 analyzes VOCs in water using a modified EPA purge and trap protocol (the
SituProbe). Since no pumps, valves, or cells are exposed to the water matrix, there is no need for
sample pretreatment or filtration, allowing the analysis of even complex water samples. Currently,
a medium-sized U.S. city is using an automated gas chromatograph (INFICON Scentograph
CMS500) in its water distribution system to detect trihalomethanes and other chemicals.
HAP SITES SituProbe Purge and
Trap GC/MS System (Infkoii)
CT-1128 Portable GC/MS
(Constellation Technology Corporation)
Portable versions of GC instruments are also available. INFICON's, Scentograph CMS200 is a
portable version of the online instrument described above. INFICON's HAPSITE® GC/MS is used
in 15 different countries for military and homeland security applications. It is backpack portable,
can be operated by individual soldiers, and yields results in minutes.40 Constellation Technology
Corporation's (Largo, FL) CT-1128, is also a portable GC-MS, but weighing 70 pounds, is set up
from the back of a truck41. (Images reproduced with permission from INFICON and Constellation
Technology.)
6.2.4 Enzyme-Based Detection
• Inhibition of Cholinesterase
The Severn Trent Field Enzyme Test (distributed by Capital Controls a division of Severn Trent
Services, Fort Washington, PA) was developed for qualitative detection of nerve agents in a field
setting. The test is based on the inhibition of the enzyme cholinesterase. A membrane disk is
saturated with cholinesterase and dipped into a water sample for one minute. If no pesticide/nerve
agent is present in the water sample, the cholinesterase on the membrane disk hydrolyzes esters
resulting in the formation of a blue color. If sufficient concentrations of pesticide/nerve agent are
present, the cholinesterase on the membrane is inhibited, the ester is not hydrolyzed, and there is no
change in color (i.e., remains white). The detection limits cited by the manufacturer for pesticides
are: carbamates (0.1 to 5 mg/L); thiophosphates (0.5 to 5 mg/L); and organophosphates (1 to
5 mg/L). No data are available for detection limits for nerve agents.42
August 2005
64
-------
U.S. EPA Office of Water
Early Warning Systems
• Inhibition of Horseradish Peroxidase - Indicated by Decreased Chemiluminescence
A chemiluminescence detection technique based on the reaction of luminol and an oxidant in the
presence of the enzyme horseradish peroxidase (HRP) can be used to indicate the presence of toxins
in a sample. The HRP mediated reaction produces light that is measured by a luminometer.
Phenols, amines, heavy metals, or compounds that interact with the enzyme reduce light output and
indicate contamination.
Eclox™, developed by Severn Trent Services (Fort Washington,
PA), is a broadband chemiluminescence test that qualitatively
assesses a water sample to determine contamination by a variety
of chemical and biological agents. The Eclox™ portable
kit/luminometer is applicable for both laboratory and field use
(EPA-ETV, 2004, photo from ETV Report 2003/2004).43
Aquanox™, developed by Randox Laboratories (Co. Antrim,
United Kingdom), is a hand-held water quality-monitoring
instrument based on enhanced chemiluminescent technology.
Aquanox™ offers onsite analysis of water and effluent waste
testing in a range of industrial applications including chemical
wastewater treatment plants.44
Edoi™ \\ ater Test Kit
(Severn Trent Services)
6.2.5 Biosensors
Biosensors work to identify toxic substances in the water using whole-organism or cellular response
approaches. Biosensors measure changes in physiology or behavior of living organisms resulting
from stresses induced by toxins. This type of biosensor does not identify the specific toxin, but
indicates that there is an unusual condition in the water. The overall rationale is that the organism
can respond with sensitivity to all the factors that contribute to stress. Fast acting toxins associated
with acute effects are most quickly detected. However, slower acting toxins or toxins with chronic
effects would not be rapidly detected if they do not also have acute effects. It is important to note
that the biosensors presented are not effective for detecting human pathogens because pathogens are
often species- or tissue-specific and require incubation times (typically days or weeks) before
disease symptoms are noticeable.
Biosensors have used bacteria (prokaryotic cells) and eukaryotic cells, as well as organisms such
as daphnia, mussels, algae, and fish. Organisms or cells that have been genetically modified to
contain specific response elements and reporter constructs such as bioluminescence can be designed
to respond to specific contaminants. Because the living organisms are sensitive to chlorine and other
water treatment chemicals, many biosensors are currently limited to raw water application. If such
treatment chemicals are sufficiently removed from the sample, the biosensors have the potential to
be used at critical points in the water distribution system. Portable sensors can be used to monitor
grab samples.
August 2005
65
-------
U.S. EPA Office of Water
Early Warning Systems
• Bacteria-Based Biosensors
Biosensors using bacterial responses show promise for use in early warning screening because
bacteria can rapidly react to toxins (biological or chemical). The metabolism of bacteria is disrupted
as a result of exposure to toxins. Some detection techniques work by monitoring for reduction in
bioluminescence. Some bacteria have natural or engineered bioluminescence, which emits light
when the cells are healthy. This bacterial bioluminescence is closely tied to respiration, so that
changes in the cellular metabolism or disruption of the cell structure decreases luminescence, which
is measureable. Kits come with freeze-dried or customer-activated bioluminescent bacteria. The
reconstituted or cultured bacteria are exposed to the water sample and the luminescence is compared
to bacteria exposed to control water. A decrease in bioluminescence (light reduction) suggests the
presence of toxins. Other detectors monitor bacterial metabolism based on color changes or on
bacterial oxygen demand. Substances that interfere with such bacterial monitoring include chlorine,
chloramine, and copper.
There are currently several commercially available bioluminescent bacterial monitoring systems that
have been verified under EPA's ETV Program,45 including Tox Screen II (Check Light, Ltd),
BioTox™ (Hidex Oy), MicroTox®/DeltaTox® (Strategic Diagnostics Inc.), ToxTrak™ (Hach
Company), and POLYTOX™ (InterLab Supply, Ltd.). Brief descriptions of these technologies are
provided below.
ToxScreen-II
ToxScreen-II Rapid Toxicity Testing System was
developed by CheckLight, LTD (Qiryat Tivon, Israel).
The basis for this technology rests on bacterial metabolism
as indicated by bioluminescence. This product uses the
luminous bacteria Photobacterium leiognathi and special
assay conditions to detect toxins in water samples. There
are two assay buffers designed to discriminate between
organic pollutants and metal toxins at sub-mg/L
concentrations. Changes in bioluminescence indicate
water toxicity. The system involves a series of steps
including sample preparation and a 90-minute incubation
period. Results are captured using a portable luminometer, which can
be integrated with a personal computer for data acquisition,
evaluation, and storage.46 The ToxScreen-II test kit costs $300 and
luminometer costs $2,895 (EPA-ETV, 2004, photo from EPA-ETV,
2004).47
BioTox™
BioTox™ Rapid Toxicity Testing System was developed by Hidex Oy
(Turku, Finland) and bases its technology on bacterial metabolism as
indicated by bioluminescence. The test uses the photobacteria Vibrio
fischeri, which reduce their bioluminescent output when exposed to
toxic chemicals. The BioTox™ Flash Test is an improved Vibrio
TosScreen-n (theckLiglit, Ltd.)
TiintKlei1™ Luminometei
ivith Injector (Hides Oy)
August 2005
66
-------
U.S. EPA Office of Water
Early Warning Systems
fischeri test for rapid screening of water and sediment samples. The detection process is the same
as BioTox™, but BioTox Flash™ automatically corrects for color/turbidity interference and can
screen the majority of the samples within a few seconds. The system uses freeze-dried BioTox™
reagent ($128 for entire kit) together with the Hidex Oy Triathler™ (portable combination liquid
scintillation counter, luminometer, gamma counter, totaling $8,900 with injector)48 Results take
from 5 to 30 minutes. The product is small enough to be portable, but can only be operated on
110-volt AC electricity (EPA-ETV, 2004, photo from EPA-ETV, 2004).49
DeltaTox®
DeltaTox®, developed by
Strategic Diagnostics Inc
(Newark, DE), is the portable,
field-applicable version of
MicroTox®, a laboratory testing
system based on bacterial
metabolism as indicated by
bioluminescence.50 Both
products can detect multiple
kinds of toxins by measuring the
light output of the photobacteria
Vibrio fischeri. The organisms'
metabolism and thus light output are reduced in the presence of toxins, indicating sample
contamination. Results are obtained in 5 to 15 minutes. DeltaTox® is a self-calibrating photometer
that incorporates a photomultiplier tube, a data collection and reduction system, and software. The
system costs $5,900 and the test kit is $370. MicroTox® Model 500 costs $17,895 and the reagents
are $360. The DeltaTox® lacks the temperature control chambers of MicroTox® (EPA-ETV,
2004).51 Both MicroTox® and DeltaTox® are sensitive to chlorine, which makes them difficult for
use in water distribution systems. The maker of MicroTox® is developing a commercial online
system that can remove the chlorine residual. (Image reproduced with permission from Strategic
Diagnostics Inc.)
ToxTrak™
MicrotoxS Model 500
(Strategic Diagnostics Inc.)
Delta toil
(Strategic Diagnostics Inc.)
ToxTrak™ Rapid Toxicity Testing System was developed by the
Hach Company (Loveland, CO) as a colorimetric test based on
resazurin dye chemistry. The basis for this technology is bacterial
metabolism as indicated by a color change. The process uses
resazurin reduction to measure respiration, a critical pathway for
cell viability. Resazurin is a redox-active dye which, when reduced,
changes color from blue to pink. Substances that are toxic to
bacteria can inhibit their metabolism and thus, inhibit the rate of
resazurin reduction. The test indicates the presence of a toxic
substance if the color of the dye does not change; a colorimeter or
spectrophotometer must be used to determine color changes (not
included in the kit). ToxTrak™ costs $280 for the kit, $100 for the reagent set, and
spectrophotometer cost approximately $3,950 (EPA-ETV, 2004, photo from EPA-ETV, 200452).
ToiTrak™ (Hach)
August 2005
67
-------
U.S. EPA Office of Water
Early Warning Systems
The Hach Company supports the use of this system as a cost effective and wide scope monitoring
method for use as an EWS by drinking water utilities.
POLYTOX™
POLYTOX™, developed by InterLab Supply™, Ltd (The
Woodlands, TX) uses the respiration of microorganisms to
indicate the toxicity of a water or wastewater stream,
including the presence of chemical and biological
contaminants.53 When activated in water, the mixture of
bacterial cultures in POLYTOX™ breathe in oxygen and
respire carbon dioxide at a rate that can be monitored and
measured as milligrams (mg) of oxygen consumed per liter
per minute. A change in respiration rate is an indication of
toxins in the sample. The portable dissolved oxygen probe
and meter cost $1,600 and the test kit costs $ 147 (EPA-ETV,
2004, photo from EPA-ETV, 2004).54
microMAX-TOX
POLYTOX™ (Inrei Lab Supply, Ltd.)
None of the preceding technologies is designed for use in water distribution systems. However, a
system called microMAX-TOX Screen is being adapted to distribution systems. Tox Screen is
manufactured by SYSTEM Sri. (Italy) and is based on the measurement technology developed by
an Israeli company (Check Light, Inc.). It is similar to MicroTox® but applied in continuous online
mode. It has two online analyzers, one colorimetric and the other ion selective. The detection
mechanism uses freeze-dried bioluminescent bacteria. It is fully automated to activate alarms and
conducts the analysis every 30 to 60 minutes (grab samples).
Every two weeks, the instrument is re-supplied with a new set
of liquid buffers and a freshly hydrated suspension of bacteria.
The interfering chlorine residual is continuously removed by
sodium thiosulfate. It is expected to be commercially available
in 2005 (Grayman et al, 2003).
• Organism-Based Biosensors
MosselMonitor®
Mussels will change their behavior in response to toxins, such
as closing their shells to reduce exposure to the toxin. Thus,
the frequency of valve opening and closing can be monitored
to indicate toxin avoidance behaviors. Delta Consult
(Netherlands) has a commercially available MosselMonitor®
that can be used to monitor chlorinated drinking water by
applying a thiosulphate pre-treatment to remove chlorine.
Thiosulfate does not remove all disinfectants used in finished
water, therefore systems that use chloramines would not be
able to use this technology. For drinking water applications,
(Delta Consult)
August 2005
68
-------
U.S. EPA Office of Water
Early Warning Systems
an Automated Food Device (AFD) is used to provide a continuous supply of nutrients. The
MosselMonitor® can be run online continuously for up to two to three months before replacement
of mussels is necessary. The data presentation software allows for near real-time graphical
presentation at a remote location or on the Internet. Only eight mussels are required because each
mussel's behavior is analyzed against its own previous behaviors, then the combined results from
all eight mussels are analyzed. The MosselMonitor® has been used with five different bivalve
species (three for freshwater and two for marine water55). The MosselMonitor® has been used by
Waterworks of Budapest, Hungary, to monitor chlorinated drinking water. Although the company
recommends replacement of mussels every three months, in the Budapest installation the mussels
survived 10 months before needing to be replaced (Jan de Maat, Delta Consult, personal
communication). (Images reproduced with permission from Delta Consult.)
Fish Bio-sensor®
Biological Monitoring, Inc. (Blacksburg, VA) makes
Bio-Sensor®, which monitors the bioelectric field of 8-12 fish to
assess if abnormal behavior indicates the presence of a toxin.56
Concurrent automated and continuous physicochemical
monitoring enhances the reliability of the results. In an event
when a toxin is indicated, an alarm is generated both locally and
remotely, so as to inform the users. A water sample is also
automatically collected should advanced (chemical) analytical
verification be desired. A dechlorination module can be added if
Bio-Sensor® is installed in the distribution system. An automated
feeder reduces the maintenance interval to a recommended once
monthly. Bio-Sensor® is currently installed in distribution
systems in Singapore, Australia, and South Africa (Joe Rasnake,
Biological Monitoring Inc., personal communication). (Image
reproduced with permission from Biological Monitoring Inc.)
6.3 Potentially Adaptable Technology
Bio-SensorS
(Biological Monitoring Inc.)
Several technologies that have been developed for other applications, such as monitoring in source
water or air, may be adaptable for use in distribution systems.
6.3.1 Enzyme-Based Detection
• Inhibition of Photosynthetic Enzyme Complexes
(PEC)
LuminoTox developed by Lab_Bell Inc. (Shawinigan,
Canada) utilizes either photosynthetic enzymes isolated
from plants, or detection of photosynthetic activity of
algae, to detect toxic analytes. Photosynthetic enzyme
complexes (PECs) isolated from plants are stabilized by
vacuum evaporation for storage. The LuminoTox system
using PECs can detect toxic molecules such as
LuminoTos?
(Lab Bell Inc.)
August 2005
69
-------
U.S. EPA Office of Water Early Warning Systems
herbicides, hydrocarbons, phenols, divalent cations, polycyclic aromatic hydrocarbons (PAHs), and
aromatic hydrocarbons. LuminoTox used with photosynthetic algae allows for the specific detection
for herbicides, organic solvents (e.g., gasoline, hydrocarbons), ammonia nitrogen and organic
amines. In the urban and industrial effluents the company has tested, toxicity can be measured in
10 to 15 minutes, while sensitivity of detection can be increased by lengthening the incubation time.
The hand-held portable luminometer allows the system to be field portable. In 2005, the company
introduced an online version called Robot LuminoTox, which is self-cleaning, monitors pH and
temperature, and records a toxicity measurement every 30 minutes. The data are stored in an Excel
file. Robot LuminoTox is operated via Windows and can be processed via a SCADA system57.
(Images reproduced with permission from Lab_Bell Inc.)
Researchers from the Institute of Microbiology (Czech Republic) have demonstrated that
Photosystem II complexes coupled to a screen-printed electrode can detect triazine and phenylurea
herbicides. The biosensor is reusable with a half-life of 24 hours and limit of detection of
approximately 10"9 M for diuron, atrazine, and simazine (Koblizek et al., 200258). However, this
system is at the research stage.
• Inhibition of Submitochondrial Particles
Harvard BioScience, Inc. (Holliston, MA) produces a toxicity detection kit called MitoScan that uses
fragmented inner mitochondrial membrane vesicles isolated from beef heart. The Submitochondrial
particles (SMPs, called "micelles") contain complexes of enzymes responsible for electron transport
and oxidative phosphorylation that are inverted from their in vivo orientation. The enzymes in the
SMPs produce a hydrogen ion gradient which is used to produce ATP from adenosine diphosphate,
and this process (oxidative phosphorylation) is coupled to electron transport as it occurs in vivo.
Progress of this reaction is directly proportional to the redox state and can be monitored
spectrophotometrically at 340 nm. When specific inhibitors or toxicants are added to the SMPs the
reactions are slowed or inhibited. The SMP vials from the MitoScan kit need to be stored at -20°C
and are viable for up to four weeks. For longer storage, SMPs should be stored at -80°C. The test
kits include all necessary reagents and SMPs in lOOul or 500ul vials. MitoScan bioassays can be
configured for microwell plate readers or for manual cuvette formats using single-beam
spectrophotometers. Cuvettes with a portable spectrophotometer capable of running kinetic assays
at 340 nm make the kit field portable. The assay also requires cuvettes, sample dilution tubes,
pipettors, and pipettor tips. MitoScan tests can be formatted to provide results in less than 30
minutes.59 The kit has not yet been third party verified.
6.3.2 Organism-Based Biosensors
Several biomonitors are available and used for monitoring surface water. They are not currently
used in treated water distribution systems because the chlorine residual is toxic to the organisms.
Thus these may be adaptable for use in distribution systems if chlorine can be removed, similar to
what been done with MosselMonitor® and Bio-Sensor®.
• Daphnia-Based Biosensors
Daphnia (water fleas) are small free-swimming organisms that are very sensitive to toxins. A
daphnia toximeter consists of daphnids contained in a glass chamber through which sample water
August 2005 70
-------
U.S. EPA Office of Water Early Warning Systems
continuously flows. The swimming behavior is monitored by closed circuit TV and analyzed via
integrated computer. Variations in speed, altitude, and frequency of turning can indicate a potential
contaminant. Sometimes the method for measuring toxins is based on feeding the daphnids
fiuorogenic food, which when metabolized by healthy daphnids, generates a fluorescent glow. In
either identification method, the setup requires high maintenance (e.g., changing daphnia
periodically) and the daphnia are sensitive to changes in temperature. The method was extensively
used in Europe and during the 2002 Salt Lake City Olympic Games. The method is primarily used
to monitor raw water because the daphnids are sensitive to the chlorine in finished water. This
makes the method more difficult for use in treated water distribution systems.
The IQ Toxicity Test™ kit for grab samples was developed by Aqua Survey, Inc. (Flemington, NJ)
and detects various chemical and biological contaminants, including nerve agents and biotoxins, in
potable water.60 This method is based on fluorescent tagging and metabolism of live multicellular
organisms. In the presence of toxins, the metabolism of Daphnia magna is reduced, blocking their
normally visible light emittance. Test preparation includes growing and feeding Daphnia magna
a fluorescent sugar reagent, while the test itself takes 75 minutes. Aqua Survey, Inc., has packaged
the IQ Toxicity Test™ into a product called Threat Detection Kit™, which the company claims can
detect 9 different toxins at levels 2 to 20 times below the human lethal threshold. Based on EPA's
ETV study, IQ-Tox™ can detect both nerve agents and biotoxins (EPA-ETV, 2004). However, the
results also may suggest that the test is too chlorine sensitive for use in finished water and
distribution systems.61
The Daphnia Toximeter, developed by bbe moldaenke (Kiel-Kronshagen, Germany), is used on line
with a measuring cycle of 30 minutes. Toxicity is assessed based on monitoring the behavioral
parameters, swimming speed, swimming altitude, turns and circling movements, growth rate, and
number of live daphnia. Allowable temperature range is between 0 to 30°C. The maintenance
interval is greater than seven days.62
• Algae-Based Biosensors
Algae can be used to detect the presence of toxic compounds by
monitoring chlorophyll fluorescence. An Algae Toximeter (bbe
moldaenke) cultures algae in a fermenter that regulates the concentration
and activity of the algae. A water sample is automatically injected with
standardized algae and monitored for changes in fluorescence. Activity
of the algae is constant if no toxic substances are present. Monitor
maintenance is required, at most, once every seven days.63 The Algae
Toximeter has not been adapted for use with finished drinking water.
(Image reproduced with permission from bbe moldaenke.)
Algae Tosimeter
• Fish-Based Biosensors (bbe m°Ida*nk*>
The German company that manufactures daphnia and algal taximeters (bbe moldaenke) also makes
a Fish Toximeter and a combined Fish and Daphnia Toximeter. The system uses zebrafish and
daphnia which are both fed by an integrated algae fermenter. Both toximeters use visual monitoring
for analysis of behaviors to assess organism health. Parameters measured are speed, altitude, turns,
circling movements, growth, and number of living fish. The presence of toxins is indicated by
August 2005 71
-------
U.S. EPA Office of Water
Early Warning Systems
changes in behavioral parameters. The measuring cycle is 1 to 30 minutes and the maintenance
interval is greater than seven days64. (Images reproduced with permission from bbe moldaenke.)
D aphniii and fish Toiimeter
(bbe raoldaenke)
Dinoflaggelate-Based Biosensor
Fish Tosimeter Real Time Bio monitoring
(bbe moldnenke)
Because human cells are eukaryotic (contain a nucleus and organelles), eukaryotic cell responses
could be better models for toxicity to humans than bacterial cell responses. The only eukaryotic
cell-based biomonitor currently commercialized for use with water is the dinoflaggelate-based
Lumitox®. Lumitox® (Lumitox Gulf L.C., River Ridge, LA) uses bioluminescent dinoflaggelate
mutants to detect the presence of toxins in the ppb range. It is field portable and the company
indicates it is unaffected by a broad range of pH, turbidity, and salinity. The American Society for
Testing and Materials (ASTM) has published a guide (ASTM El924-97) for using Lumitox®. It
can test both marine and non-marine fluids, soils, and chemicals (either water-soluable or
lipophylic). The patented TOX BOX® testing instrument is considered easy to operate, and no
computer is required. Grab sample screening can be completed in two to four hours.65
6.3.3 Infrared Spectroscopy
HazMatID™ is the newest product from SensIR Technologies (now
merged with Smiths Detection, Edgewood, MD)66 for identification of
a variety of weapons of mass destruction, toxic industrial chemicals,
narcotics, and explosives. This is a portable tool using Fourier
transform infrared (FT-IR) attenuated total reflection (ATR)
spectroscopy for field identification of analytes in the solid or liquid
phase. HazMatID™ features an integrated computer system with
wireless remote control capabilities to instantly compare the
spectroscopy of the unknown contaminant against an onboard
spectroscopy database of known substances. Its Bio-CheckO? software
can alert the user when at least 10 percent of the sample is composed of
proteins, indicating that a possible biological material is present (Mark
Norman, Smiths Detection, personal communication). The sample
interface is a diamond sensor with integrated video monitoring and is operational in extreme weather
and temperatures. Since infrared analysis is limited in aqueous samples containing less than 10
percent product, SensIR has developed an accessory product to HazMatID™, called ExtractIR™.
This portable tool physically removes nonvolatile organic chemicals from the interfering water
HazMat ID™
(Smiths Detection)
August 2005
72
-------
U.S. EPA Office of Water Early Warning Systems
matrix. By doing this, chemical levels as low as 100 ppm in water can be identified. ExtractIR™
can be used in a hot zone by a responder in full Level A protective gear, and the entire process takes
about 10 minutes.67 The suitcase unit can be operated in extreme temperatures and can be totally
immersed in water for decontamination.68 (Image reproduced with permission from Smiths
Detection.)
6.3.4 X-ray Fluorescence
ITN Energy Systems, Inc. (Littleton, CO)69 has been given an EPA Small Business Innovation
Research (SBIR) Phase I award for the period of March 1,2005 to August 31, 2005.70 The ultimate
project goal is to adapt ITN's x-ray fluorescence technology to provide a smart, automatic early
warning sensor for trace levels of toxic metals in water on a ppb scale. This technology has been
proven in solar cell manufacturing where the sensor continuously monitors very small amounts of
metal in products and automatically provides feedback to process control. The sensor is capable of
simultaneously detecting multiple metals, including mercury, arsenic, and lead. In Phase I, the
feasibility of this approach will be demonstrated by showing the sensor can detect 20 ppb of mercury
in water without interferences from other metals, chemical state of the metals, or organic material.
6.3.5 Ion Mobility Spectroscopy
Ion Mobility Spectroscopy (IMS) is a technique for identifying and measuring volatile compounds.
An ambient air or vapor sample is drawn over a semi-permeable membrane. Smaller volatile
compounds pass through the membrane into the detection cell, where the sample is ionized by a
weak plasma formed by a nickel-63 radioactive source. The ionized sample molecules drift through
the cell under the influence of an electric field. An electronic shutter grid allows periodic
introduction of the ions into a drift tube where they separate, based on charge, mass, and shape.
Smaller ions move faster than larger ions through the drift tube and arrive at the detector sooner.
The amplified current from the detector is measured as a function of time and a spectrum is
generated. A microprocessor evaluates the spectrum for the target compound, and determines the
concentration based on the peak height.71 IMS is used in explosives detection equipment at airport
security check points.72 There are several portable IMS sensors for chemical detection but all have
been designed for use with air/vapor samples.73 One example is described below.
Smiths Detection produces a hand-held chemical vapor detector
called the SABRE 4000, that utilizes IMS and can detect and
identify over 40 threat substances (explosive, chemical warfare
agent, toxic industrial chemical, or narcotic substances) in
approximately 15 seconds. It takes 10 minutes to warm-up and
weighs seven pounds including the four-hour battery.74 The Sabre
4000 can be used in two modes, vapor mode or direct thermal
desorption mode. In the latter mode, the Sabre 4000 can test liquid
,., , ^v, . ' A 11 j • . SABRE 4000 (Smiths Detection)
or solid samples. Either temperature controlled ramping to
evaporate the sample or a fiber solid phase microextraction
(SPME) probe are needed to measure aqueous samples. IMS technology can detect volatile organic
and inorganic compounds (MW < 1,000) above about 5 to 10 ppb concentrations. The technology
could be used to detect many different substances, but only about 40 to 50 profiles can be saved in
August 2005 73
-------
U.S. EPA Office of Water Early Warning Systems
the current versions of the sensor devices (Rachel Kohn, Smiths Detection, personal
communication). (Image reproduced with permission from Smiths Detection).
6.3.6 Microchips for Portable Chemical Sensors
Not all portable sensors are based on microchip technology, but many are. The name, microchip,
is derived from the micrometer dimensions of the individual components that comprise microchips.
Any technology that can be minaturized to the micrometer scale could be mounted on a solid
platform and form the basis of a microchip technology. Minaturization is a key development for
making small portable sensors. When semiconductor devices (e.g., transistors and resistors) were
minaturized into the form of microprocessors, the computer industry was revolutionized. Since the
early 1990s, other technologies have been adapted to microchip dimensions and technologies that
would not be feasible on a macro level have also been invented for microchip platforms. This newer
class of microchips is often referred to as "lab-on-a-chip," because the kinds of data generated by
the microchips are comparable to data that would have previously required bench-top sized
equipment to gather. "Lab-on-a-chip" generally refers to microetched chambers that can contain
nanoliter volumes of liquid to carry out multiple, small-scale, side-by-side chemical and/or
biological reactions. Microfluidics technology which allows minute quantities of liquids (reagents
and sample) to be manipulated and delivered to microchip components is an essential part of any
chip design that utilizes aqueous solutions.75
"Microarray" refers to microchips that have zones of unique characteristics. Current technology can
allow for as many as 100,000 distinctly unique zones or elements in a microarray. Each element
within an array can be designed to react with or respond to a different sample component. For
example, if the microarray has elements that recognize different DNA sequences, when a sample
of mixed DNA sequences is delivered to the microarray elements, only a subset of the elements will
respond to the sample. The response of the elements is detected or "read" by a chip reader device.
For example, the chip transducer could be based on optical, piezoelectric, magnetic, electrochemical,
or thermometric mechanisms. Although there are many different technologies for reading
microchips, the microchip design and the microchip reader device are designed as a single system
and in some cases are fully integrated. Microchips may or may not be reuseable. Several companies
offer custom designed microchips, which means they will incorporate subsets of elements specific
for the particular targets the customer wants to identify.
Micro-Electro-Mechanical System (MEMS) microchips have miniaturized mechanical and electrical
components.76 Microcantilevers andmagnetoelastic sensors are examples of MEMS. Research and
development of MEMS-based technologies is a large field covering many potential applications.
As this field grew out of the integrated circuit field, much of the technology is built on silicon
wafers.77
Biochips, which often come in the form of microarrays, are based on biological reactions using
biological components such as nucleic acid hybridization, antibody reactions, and enzymatic
reactions. Biomolecules are also used for bioelectronic applications in a microchip format.
Biochips could be designed to detect biological molecules (DNA, RNA, proteins, biotoxins), or non-
biological chemicals.78 It's important to note that the term "biochip" refers to the biological basis
of the chip and does not limit the class of targets to those that are biological. Because enzymes,
August 2005 74
-------
U.S. EPA Office of Water
Early Warning Systems
DNA, and antibodies all work in aqueous environments, microfluidics is an essential part of biochip
technology.
There are many kinds of prototype microchips demonstrating proof of concept that have not been
commercialized. Commercialized microchips are common in research and diagnostic laboratories,
however the support equipment to use the microchips, such as microfluidics stations and chip
readers, requires a laboratory setting, including trained personnel. Although microchips are a highly
developed technology for genomics, they are an emerging technology for other applications, such
as medical devices and environmental sensing.
Because of programs like the National Nanotechnology Initiative79 (NNI, a federal R&D program
established to coordinate multiagency efforts in nanoscale science, engineering, and technology) and
the intense interest and promise in the field in general, it is likely that micro- and nano-based
technologies will significantly contribute to product development and commercialization in the near
future.
6.5.7 Microchip Surface Acoustic Wave (SAW) Technology
SAW technology has been used for decades in transceiver technology and cell phone technology.
For chemical detection, SAW sensors can be configured in amicroarray, with each element uniquely
coated. Mass changes in a subset of elements due to interaction with a particular volatile chemical
causes surface acoustic waves (~10 A in amplitude, 1 to 100 micrometers in wavelength) which are
detected by piezoelectric materials. The subset of elements that respond to a specific VOC can be
recognized by software included in the sensor, allowing for a diverse list of detectable analytes.80
HAZMATCAD™, from Microsensor Systems, Inc. (Bowling
Green, KY), features three 250 MHz SAW sensors in a hand-held
portable Chemical Agent Detector instrument. Each sensor is
coated with different polymers that provide a multi-pattern sensor
response (fingerprint) to indicate the presence of contaminants in
vapor samples.81 HAZMATCAD™ detects and identifies trace
amounts of chemical warfare agents, including nerve and blister
agents, and can be configured to detect phosgene and/or hydrogen
cyanide. HAZMATCAD™ Plus supplements the SAW technology
with electrochemical sensors for additional detection of up to four
classes of toxic industrial chemicals, specifically hydride, halogen,
choke, and blood agent vapors. The system operates on 20 to 120
second analysis cycle, depending on the mode selected, and the
typical time to alarm is less than 60 seconds in "Fast Mode." This
device has price range of $4,850 to $7,950 and has been evaluated
by the Army.82 (Image reproduced with permission from
Microsensor Systems Inc.)
HAZMATCAD™
ficrosensor Systems, Inc.)
August 2005
75
-------
U.S. EPA Office of Water
Early Warning Systems
Cyi an os e™ 320 and Nose Chip™
(Smiths Detection)
6.3.8 Microchip Chemiresistors
Cyrano™ Sciences, now a part of Smiths Detection
(Edgewood, MD),83 has developed miniature "electronic
nose" sensors using conductive polymer films deposited
in an array on a ceramic substrate. Each individual
detector of the sensor array is a composite material
consisting of conductive carbon black homogeneously
blended throughout a non-conducting polymer. The
detector materials are deposited as thin films on an
alumina substrate that lie across two electrical leads,
which create conducting chemiresistors. The output from
the device is an array of resistance values as measured
between each of the two electrical leads for each of the detectors in the array. The polymer
composite chemiresistors, are designed to absorb a diversity of analytes. The manufacturers claim
that the polymer composite sensors respond to a wide range of organic compounds, bacteria, and
natural products in a vapor form. The signature of 100 different analytes can be stored in the
memory of the Cyranose® 320 hand-held device, which is commercially available. A penny sized
version called the NoseChip™ is being used to develop future products called ChemAlert™ and
ChemBioAlert™, which will be integrated into online sensor networks. Although these "electronic
noses" are not able to directly measure water samples, if used with vaporizing technology, they
could be adapted for water sampling (Rachel Kohn, Smiths Detection, personal communication).
(Images reproduced with permission from Smiths Detection.)
6.4 Emerging Technology
Although the biosensor, IMS, and SAW technologies discussed above (Section 6.3) have
commercially available products on the market for surface water or vapor media, these technologies
are also being used to develop the next generation of sensors so they can also be considered an
emerging technology. In addition, fiber optic cables, which are being designed to serve as
continuous sensors, are discussed below. These emerging technologies are being developed by
companies, national labs, and other research institutes.
6.4.1 Organism-Based Biosensors
• Clam Biomonitor
Similar to MosselMonitor®, discussed previously, other groups are investigating using clam
responses to detect water quality issues that affect clam behavior. However, these approaches have
been limited to source water monitoring. The University of North Texas, Little Miami, Inc.
(Milford, OH), and EPA have a joint project to develop a Clam Biomonitoring System. The gape
of 15 clams is measured at one-minute intervals and is plotted along with temperature, pH,
conductivity, and dissolved oxygen. A cellular modum is used to connect to the Internet for data
relay. The system is installed on the Little Miami River in Miamiville, OH.84 This system has not
been tested in chlorinated drinking water.
August 2005
76
-------
U.S. EPA Office of Water Early Warning Systems
• Fish Biomonitors
The Great Lakes Water Institute, a University of Wisconsin research group, is developing a
biosensor system using transgenic zebrafish to monitor water distribution systems. The embryos
of zebrafish are injected with pollution responsive reporter genes shortly after fertilization. The
transgenic fish are designed to detect 18 chemical contaminants, including the biological warfare
agents paraoxon and parathion, which are relatives of sarin. The contaminants trigger the pollution
response elements, which then activate the production of luciferase in the fish. The enzyme causes
the fish to emit light, thereby signaling the presence of a toxin in the water. The fish can be used
to repeatedly analyze the same site as the monitoring process does not kill the fish.85
The U.S. Army Center for Environmental Health Research (USACEHR) has developed an
automated biomonitor-based on bluegill (Lepomis macrochirus) ventilatory and body movement
patterns.86 It was developed to monitor toxins in treated wastewater but has not been adapted for
chlorinated drinking water.
6.4.2 Eukaryotic Cell-Based Biosensors
There are emerging technologies that incorporate eukaryotic cells or tissues in assays on microchip
platforms for biosensors. The B-cell, cardiac-cell, and fish-cell systems described in this section
have not yet been commercialized and are not being developed specifically for drinking water.
These emerging technologies could be utilized for drinking water monitoring provided that sample
volume concentration and disinfectant residual issues are addressed.
Fish cells could be used to detect toxins through changes in color and movement of chromatophores
(primary cells). The cells are stored in disposable cartridges which can be monitored by a video
camera through a microscope. The data would be analyzed by a computer for the changes described
above. Such a system could show changes within 1 to 100 minutes and could be used in continuous
or grab sample mode (Grayman et al., 2003). A prototype system, SOS Cytosensor system, is being
developed by Adlyfe, Inc. (Rockville, MD).87
A biosensor-based on mammalian cells (cardiac cells) is being developed by Dr. Gregory Kovacs
at Stanford University. The cells are cultured on disposable microelectrode arrays. The toxins are
detected by changes in electrical discharge, beat rate, and signal propagation velocity. The system
is being designed for grab sample analyses using portable hand-held equipment. The Portable Cell-
Based Biosensor is in the prototype stage. Mammalian neuronal cells have also been explored as
a biosensor by the U.S. Naval Research Laboratory. The neuronal cells are carried in a transportable
cell cartridge which is placed in a monitoring device for continuous monitoring for up to two days.
A computer program evaluates changes in electrical patterns such as mean spike rates. The system
called Portable Neuronal Microelectrode Array is commercially available (Shaffer et al., 2003).88
6.4.3 Fiber Optic Cable-Based Sensors
The Great Lakes Water Institute, aUniversity of Wisconsin research group, is developing a real-time
water distribution system monitor utilizing fiber optic cables. The Water Security Division of the
Great Lakes Water Institute is supported in part by the Water Harvesting and Water Purification
Program at DARPA. The monitor consists of a fiber optic cable run through the water conduit, and
August 2005 77
-------
U.S. EPA Office of Water Early Warning Systems
various chemical receptors and a fluorescent group are attached to a gel coating the cable. When
a toxin binds to a receptor, the characteristics of the fluorescent group, or fluorophore, are altered.
Laser pulses passing through the cable detect the change in the fluorophore, providing for detection
from a central monitoring station. Spatial maps are generated, based on the data from the laser
pulses. The Institute is developing technology to add both biological and chemical receptors to the
cable in order to provide a more complete monitoring system, since the only toxins that are detected
are the specific ones that the system is designed to detect.89
Intelligent Optical Systems, Inc. 's (Torrence, CA) DICAST® technology is composed of fiber optic
cables that have a glass core coated with a permeable indicator-doped cladding to achieve chemical
sensitivity over their entire length.90 Instead of having several sensors at various locations along its
length, the entire length of the fiber is a sensor. Thus, it has more sensing area and less probability
of missing a target molecule. Although DICAST® has been developed for sensing in air, it could
be modified to work in water. The applications of this and other fiber optic sensors include
monitoring water for dissolved gases, pH, bio-content, and toxic chemicals and their byproducts
such as cyanide (Steven Cordero, Intelligent Optical Systems, Inc., personal communication).
6.4.4 Ion Mobility Spectroscopy (IMS)
Sionex Corporation's (Waltham, MA), MicroDMx™, technology is based
on IMS in a MEMS format, but is referred to as differential mobility
spectroscopy because a varying radio frequency field pulls the ions in a
zigzag pattern increasing the distance they travel thereby increasing their
separation.91 The company has prototype hand-held and online products
in development for detection of VOCs in air. The company also has
research projects ongoing for adapting the prototypes for water matrices
(Joe Santos, Sionex, personal communication). (Image reproduced with
permission from Sionex Corporation.) . , n\r TM r-i •
(Sioues)
6.4.5 Surface Acoustic Wave (SAW) Technology
S-CAD, developed by Science Applications International Corporation (SAIC; San Diego, CA), is
a portable, hand-held chemical agent detection system for air. This product features the dual
detection power of an IMS cell and a SAW sensor, combined with a data fusion algorithm to reduce
false alarms without affecting detection performance. The system can both identify and specify the
concentration of different chemical agents, including nerve, blister, and blood agents. S-CAD has
the capability to gather and store data for future analysis, and its modular design allows easy
integration with nuclear and biological agent detectors and/or other application specific sensors.92
The company is currently examining issues regarding adaptation for water sampling (Steven Haupt,
SAIC, personal communication).
Sandia National Laboratories (SNL) has developed the Micromachined Acoustic Chemical Sensor
to detect VOCs, explosives, illicit drugs, and chemical warfare agents.93 These miniature sensors
(as small as 0.5 mm) incorporate a micromachined flexural plate wave (FPW) device as a general
purpose chemical sensing platform.94 FPW technology uses polymer films to selectively absorb the
analyte of interest and is compatible with both gases and liquids. These devices are analogous to
the SAW sensor, which is an extremely sensitive gravimetric detector that can be coated with a film
August 2005 78
-------
U.S. EPA Office of Water Early Warning Systems
to collect chemical species of interest down to ppm and ppb levels of airborne contaminants.95 The
Micromachined Acoustic Chemical Sensor can be fabricated on silicon with integrated
microelectronics to complete the sensor capability. Although none of these technologies has been
integrated into a commercially available product nor has been designed for use in water distribution
systems, SNL believes the technology has high potential for in situ chemical detection applications.
SNL has also developed two field prototypes for portable chemical
analysis systems, called uChemLab, which researchers hope to
combine into one complete system by 2008 (uChemLab/CB™96).
The Gas-Phase uChemLab uses a combination of GC channels
coupled with an array of SAW sensors to monitor volatile and semi-
volatile organic species in air. This product can detect chemicals at
levels as low as 10 to 100 ppb and analysis takes just seconds-to-
minutes. The Liquid-Phase uChemLab hand-held analyzer utilizes
various chip-based technological innovations such as microfluidics, nChemLab™ Liquid Phase
capillary gel, and zone electrophoresis columns that are combined (Saudia National Lab)
with small laser-induced fluorescence detectors. The detectors can
analyze biotoxins and other inorganic and high MW chemical contaminants. SNL scientists plan
to further develop the system so it can detect viruses and bacteria. Nanodetex97 (formerly MCL
Technologies, Albuquerque, NM) is an SNL spin-off company licenced to develop uChemLab for
chemical warfare agents, drug detection, and health monitoring. There is also ongoing research to
adapt the Gas-Phase uChemLab to water applications, including automated onsite measurement of
trihalomethanes, petroleum hydrocarbon contaminants, and chemical warfare agents and their
hydrolysis products. The ultimate goal is to develop a low-cost, quickly deployable, and real-time
discriminatory sensor for water quality determination as well as online versions of the equipment.
This technology is funded under the DOE's Chemical and Biological Non-Proliferation program and
is also a candidate system for the DOD's Joint Chemical Biological Agent Water Monitor project
(Wayne Einfeld, SNL, personal communication). SNL, CH2M Hill (Colorado), and Tenix
Investments (Australia) have signed an agreement that calls for an online water monitoring
prototype, based on uChemLab, to be developed and begin testing by June 2005.98 The first phase
of testing will focus on detecting ricin and botulinum toxin. The development team eventually also
hopes to address viruses, bacteria, and parasites. (See Appendix B for summaries of SNL research
projects related to detection technologies; photo from SNL website.)
Pacific Northwest National Laboratories (PNNL; Richland, WA) has designed a S AW-based sensor
system to detect chemical warfare agents in the field. The SAW sensors have chemically selective
polymer coatings to provide rapid, reversible analysis of multi-component chemical vapors.
Chemometric analysis is used to distinguish between multiple vapor signals, which is the application
of mathematical, statistical, graphical, or symbolic methods to maximize chemical information
which can subsequently be extracted from data. The system is integrated with computer control and
data acquisition and has been deployed and demonstrated in the field. PNNL researchers have also
improved the sensitivity of traditional SAW sensors by coating them with synthesized novel
hydrogen bonding organic/inorganic copolymers. The sensors reach 90 percent of full response
within six seconds of the first indication of response when exposed to nerve agent simulants and
organic solvents. In tests using actual nerve agents, response sensitivity was increased at least four
times over traditional sensing polymers. PNNL's new copolymers are currently being used in
August 2005 79
-------
U.S. EPA Office of Water Early Warning Systems
commercially available chemical sensors for the detection of nerve agents. Copolymer technology
is available for licensing."
6.4.6 Raman Spectroscopy
Real-Time Analyzers, Inc. (East Hartford, CT)100 has recently been received an EPA Small Business
Innovative Research (SBIR) Program Phase I award for the period of March 1, 2005, to August 31,
2005, to provide the Agency with a chemical sensor that can be multiplexed into water distribution
systems to provide an EWS capability. Surface-enhanced Raman scattering (SERS) sensors will be
coupled to a central Raman analyzer via fiber optics. The objective of the project is to develop
sensors that will selectively detect several chemical agent hydrolysis products, toxic industrial
chemicals, and pesticides at concentrations below 1 mg/L in flowing streams in 10 minutes or less.
August 2005 80
-------
U.S. EPA Office of Water Early Warning Systems
7. Technologies that Detect Microbial Contaminants for Early Warning Systems
7.1 General Introduction to Assays and Sensors
Currently, microbial culture methods are relatively slow, requiring at least 24 to 48 hours before
results are available. With traditional techniques, growth of the target organisms in culture is
required for identification. Ideally, microbial monitoring methods should be rapid, providing results
in two hours or less. The methods presented in this chapter begin to meet these criteria. Sensors
for living organisms can target detection of genetic material (nucleic acids), proteins, or other
components or activities of living cells such as adenosine triphosphate (ATP). Most sensors for
detecting microbes are based on biological interactions, so they have biological components
incorporated into the sensor technology. The sensor components that interact directly with sample
components are called capture or recognition components. Capture molecules can be DNA,
antibodies, or other molecules that bind or react with sample components. The sample component
that will be detected is called the target. Generally, target refers to the actual molecule that interacts
with the capture molecule, but in some cases it is used to refer to the whole organism that is
indicated by the presence of the target molecule. Examples of target molecules are specific DNA
sequences and antigens. Target molecules can also be chemicals. There are some sensors presented
in this chapter that can detect both pathogens and chemicals, but detection of chemicals is not the
primary application of the technology.
The specificity of a sensor is determined by how reliably and firmly capture molecules bind or
interact with the specified target. The capture molecule-target molecule interaction is detected
through various mechanisms such as light production or mass change. The sensitivity of a sensor
is based both on how well the capture molecule and target interact and on how many molecules need
to interact before the reaction can be detected. The reaction kinetics (binding and release of target
molecules) requires that targets be relatively concentrated, so methods for concentrating microbes
from larger volumes of drinking water are needed. In the case of intentional contamination, a
sample concentration step may not be needed. The above discussion applies to multiple sensor
platforms, such as immunoassay strips, microchips, and solution phase systems.
EPA does not endorse or recommend any of the following technologies. The summary
information below was obtained from company websites, promotional literature, and personal
communication with company representatives.
7.2 Available Technologies
7.2.1 Immunoassays
The principle behind rapid immunoassay technologies is to detect the antigen-antibody reaction.
Specific antigens in the water are bound to corresponding antibodies through targeting of specific
proteins. The presence of a microbial contaminant is "seen" when specific antigen proteins in the
sample bind with the corresponding antibodies. Immunoassays have been used since the early 1980s
for many research, clinical, and safety/quality control applications. A familiar example of a strip
type immunoassay is a home pregnancy test.101 Classic examples of immunoassays are the enzyme-
linked immunosorbent assay (ELISA) and the enzyme-linked fluorescent immunoassays (ELFA).
Immunosorbent refers to the immobilization of the capture molecules on a surface, such as a
August 2005 81
-------
U.S. EPA Office of Water Early Warning Systems
membrane (Alberts, 1994). The capture molecules can be either antigens or antibodies. The target
molecules in the sample are antibodies if antigens are immobilized and antigens if antibodies are
immobilized. Another antibody that recognizes either the sample antigens or sample antibodies,
called the secondary antibody, is added. The secondary antibody is conjugated (linked) to an
enzyme that forms a colored precipitate (for ELISA) or gives off light (for ELFA) in the presence
of substrate. Each enzyme molecule acts catalytically, thereby amplifying the signal of a successful
binding interaction. When the assay consists of capture antibodies recognizing antigens, which are
in turn recognized by secondary antibodies, it is called an antibody sandwich assay. ELISAs for
quantitative assays are commonly done in microtiter plates in laboratory settings and may require
time consuming pipetting. However, the technology has been advanced for strip assays for onsite
uses. One of the issues for many immunoassays is cross reactivity with other microbes which leads
to high false-positives rates. This can be addressed with the use of specific epitope targeted
monoclonals that have been verified for low cross-reactivity with other potential sample
components. Immunoassay methods have primarily been developed for grab samples, but have not
been applied for online water distribution monitoring systems. Variations on the basic ELISA and
ELFA concepts have also been incorporated into microchip designs.
Immunoassays designed for onsite grab sample screening for intentional chemical or microbial
contamination of air, food, and water are desirable because the method can determine the presence
of a specified microbial contaminant or contaminants and can be completed in less than 15 minutes.
Test strips are generally not quantitative, so the results usually need to be confirmed by other
methods.
• Strip Tests - Lateral Flow Assay
The lateral flow assay is a general technique for detecting antigens.102 It is a simplified version of
ELISA. The test strip is an absorbent membrane mounted on a stiff (plastic) backing, often
contained in a plastic cassette. A liquid sample is applied to one end of the strip, the sample diffuses
along the length of the strip, passing several stripes (narrow regions) impregnated with high
concentrations of specific antibodies, which are labeled with colored dyes or fluorescent agents.
When the antigen-labeled antibody complex migrates to the test stripes, a color change or
fluorescent signal indicates the presence of the target antigen. A control stripe beyond the test stripe
serves as a positive control, indicating proper diffusion and appropriate functioning of strip reagents
(Exhibit 7-1). Home pregnancy tests are lateral flow assays.103 Traditional lateral flow assays
produce color changes that can be detected with the naked eye. Newer fluorescent and
phosphorescent reporters require excitation light and/or light detection devices.
A commercially available lateral flow assay test strip is Bio Threat Alert® (BTA) made by Tetracore
(Gaithersburg, MD104). Samples are put on the test strip and move along the strip membrane. A
reddish band that can be seen with the naked eye on the positive line indicates the presence of
specific contaminants. The following are the currently available tests and detection limits, expressed
as colony forming units per milliliter (cfu/ml) (images reproduced with permission from Tetracore):
• Bacillus anthracis (1 x 105 cfu/mL)
• Yersinia pestis (2 x 105 cfu/mL)
• Francisella tularensis (1.4 x 105 cfu/mL)
August 2005 82
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 7-1. Lateral Flow Assay
(image from: http ://spaceresearch.nasa. gov/general info/homeplanet.html)
Analyte
Lateral Flow Assay Architecture
Antibodies conjugated Tag Test Line Control Line
(Gold, Latex, Fluorophore, etc.) (Antibodies) (algG Antibodies)
Sample Conjugate Nitrocellulose
Pad Pad Membrane
Wlcking
Pad
Backing
•p • I
A A
est Line Control Line
(Positive) (Valid Test)
BioTbreat Alertf (Tetracore)
• Botulinum toxin (10 ppb)
• Staphylococcal enterotoxin B (2.5 ppb)
• Ricin (50 ppm)
New Horizons Diagnostics' (Columbia, MD) SMART™ (Sensitive
Membrane Antigen Rapid Test) Tickets use a similar method. The
detection is provided by immuofocusing of colloidal gold-labeled
antibodies and their corresponding target antigens onto small
membranes. Two red lines on the strip indicate positive control and
positive sample and the test is complete in 15 minutes. SMART™ tests
currently available are the same as BTA with the inclusion of Vibrio
cholerae Ol. The detection limits of bacteria are 105 cfu/mL and for biotoxin is 50 ppb.
SMART™ Tickets have been incorporated into Bio-HAZ™ (EAI
Corporation, Abingdon, MD), a portable field kit for sample
collection and analysis to detect the presence of biological
contaminants (ECBC, 2002). Designed primarily for use by
emergency response and forensics evidence personnel, the kit
contains materials needed for liquid, solid, and air sampling and
onsite biological screening. Fluorometry, luminescence,
colorimetry, and sample-specific analyses, performed with hand-
held equipment, identify the presence of biological contaminants
105
Bio-HAZ™ (EAI Corporation)
August 2005
83
-------
U.S. EPA Office of Water Early Warning Systems
in the field. Ruggedized for field use, the kit also includes instructions to assure evidentiary
integrity.106 (Image reproduced with permission from EAI Corporation.)
ADVNT (Phoenix, AZ),107 immunoassay detection strips called BioWarfare Agent Detection
Devices (BADD), have been used by U.N. weapons inspectors in Iraq.108 BADD strips are
self-contained, qualitative assays for screening environmental samples for the presence of anthrax,
botulinum toxin, and ricin. After a sample is transferred to the BADD test strip, dye-labled
antibodies detect trace amount of the contaminant and show its presence with two bands. After 15
minutes, the results are read visually.
RAMP Anthrax Assay by Response Biomedical Corporation (Vancouver, Canada) is a rapid
immunochromatographic system. The system has a portable fluorescence reader and test cartridges
for detecting anthrax, botulinum toxin, ricin, and smallpox. The cartridge is a lateral flow
immunoassay device with a test and control line. The detector is an antigen-specific antibody
attached to a fluorescent bead. The RAMP instrument detects the presence of fluorescent beads that
attach to the capture assay line.109 The RAMP Anthrax Assay has not been tested for use in drinking
water systems.
7.2.2 Detection of Bacterial-ATP
A common indicator for microbial presence used in the food and beverage industry is ATP. Tests
for ATP are now being used for rapid detection of microbes in water, such as cooling towers. A
small volume of water (typically < 20 milliliters) is required and the test takes between 30 seconds
and a few minutes depending on the kit. Either free ATP or microbial ATP can be measured. To
measure microbial ATP, cells in the water sample are lysed to release ATP into the solution. To
detect only ATP contained within living bacterial cells, the water sample must first be filtered to
collect the cells and rinse away non-bacterial ATP, then the cells are lysed to release ATP.
An enzyme, luciferase, and a substrate, luciferin, are present in the reaction solution. The reaction
catalyzed by luciferase, that breaks down ATP and releases a photon of light from luciferin. A small
hand-held/portable luminometer reads the quantity of light emitted from the reaction. The light
intensity is directly related to the concentration of ATP in the sample, which is in turn an
approximate indicator of biomass in the sample.110 The ATP test requires users to purchase a
luminometer and consumable supplies (enzyme, substrate, and sample container). The ATP reaction
can be amplified with adenylate kinase (AK) to measure very low numbers of cells.111
The concentration of ATP in microbes depends on species, strain, environmental, and metabolic
factors. Therefore, ATP is only an approximate indicator of biomass. Using bacteria-spiked water
samples, the kits available have lower detection limits, around 1,000 cfu/mL. Since all living cells
have ATP, microbial ATP needs to be separated from non-microbial ATP. However it cannot
differentiate between bacterial species. In general, a single relative light unit (RLU) reading is not
adequate for assessing the degree of microbial presence in a sample. It is important that routine
testing establish a baseline trend for ATP results, then subsequent fluctuations in ATP can indicate
a change in microbial status for the system. ATP from the human user can also be a source of false-
positive readings. Numerous companies have products on the market, a few that design products
for water are presented below. No third party verification was indicated for any of the products.
August 2005 84
-------
U.S. EPA Office of Water
Early Warning Systems
AMSALite™ AMSALitelH
(Antimicrobial Specialist's and Associates Iiic.)
A few products that can detect total
ATP are included because they may be
useful for monitoring highly pure water
samples that do not have background
sources of ATP. AMSALite™
(Antimicrobial Specialists and
Associates, Inc., Midland, MI)112 sells
an ATP detection kit that is specifically
for industries using high purity water,
such as printing.113 The luminometer
costs about $2,000 and several versions
of the kit available. The WaterGiene™
(Charm Sciences, Inc., Lawrence,
MA)114 test swab has a chamber with a cell lysing agent to expose cellular ATP, but it does not first
rinse extracellular ATP away. (Images reproduced with permission from AMSA Inc.)
Bio Trace International (Bridgend, UK) sells an online Continuous
Flow ATP Detector115 that can be used to discriminate between
background ATP and bacterial ATP. The company claims "near real-
time" results and a testing capacity that provides readings every
second.
New Horizons Diagnostic Corporation's (Columbia, MD) Profile®-1
uses its Filtravette™ disposable cartridge system to remove non-
bacterial ATP arising from somatic cells (other sources of non-
bacterial ATP) and other interfering compounds.116 This system has
been demonstrated to be useful for measuring only bacterial ATP in
distribution system water by Deininger and Lee (2001). The
Filtravette™ allows for the free ATP to be rinsed away before the
bacterial ATP is released into the assay solution. (Image reproduced
with permission from New Horizons Diagnostics Inc.)
7.2.3 Flow Cytometry- and Micro-Flow-Based Technology
Flow Cytometry is a general technique that has been used in laboratories since the 1960s for analysis
of cells. More recently, flow cytometry has been used in medical applications and analysis of
environmental microbial populations. A monodisperse suspension of cells flows past a laser beam
(or in some more complicated instruments, multiple laser beams) and the device measures properties
of each cell, such as size, granularity, green fluorescence, red fluorescence, and far red fluorescence
intensities.117 Fluorescent tags can be used for a variety of general or specific cell components, such
as DNA, RNA, proteins (antigens), or other target molecules. Some microorganisms can be
distinguished on the basis of differential light scatter properties with the addition of fluorescent tags.
Dyes that stain only live cells can be added to samples so that the flow cytometer can quantify the
level of live versus dead cells. Nucleic acid intercalating dyes can be used to determine DNA/RNA
ratios and adenine-thymine/cytosine-guanine content which helps further characterize the cells in
a sample and could be used to identify microorganisms in some cases. Fluorescently labeled
antibodies can be used to identify specific organisms in aerosols, water, soil, and food.
Profile!:-!
CN'eiv Horizons Diagnostics Inc.)
August 2005
85
-------
U.S. EPA Office of Water
Early Warning Systems
Alternatively, attachment of specific antibody to fluorescent microspheres can detect both toxins and
viruses in multicolor assays. The same technology that can "see" and sort individual cells can also
be used to analyze micro particles.
Microcyte Aqua® and Microcyte Field® by BioDetect (Houston, TX)
118 are suitcase sized flow cytometers that can be used in the field to
characterize particles and identify microbes if used in conjunction with
fluorescent tags. Microcyte Aqua® is targeted at routine analysis of
algae and other microorganisms in water. BioDetect claims that the
minimal sample preparation makes the system suited for integration into
an online, continuous water surveillance system. The instrument can
differentiate between biological and non-biological particles, which is
important for applications where correlation between the total number
of particles and microorganisms does not always exist. (Image used
with permission from BioDetect.)
"
Microcyte Aqua:I
(BioDetect)
Micro-Flow Imaging (Brighnvell Technologies)
Brightwell Technologies (Ottawa, Canada)119
produces an automated Micro-Flow Imaging
instrument that acts as a particle counter but
actually captures digital images of the particles in
the water sample. The sample is drawn through
a micro-fluidic flow cell. One digital image per
second is taken of the sample and images are
stored that meet user-defined parameters. It
takes about 5 minutes to analyze 1 mL of sample.
Particles as small as 1 um can be seen by the
camera, which has a 0.2 um resolution. Particle
sizes and concentrations are analyzed and the
data are presented graphically. The system requires no sample preparation and can be run
continuously or intermittently for hours. The company has tested the system in drinking water and
wastewater treatment applications. (Image reproduced with permission from Brightwell
Technologies.)
7.2.4 Bioparticulate Monitors - Light Scattering Technology
Light scattering technology is a simple scanning procedure that provides information about the
presence of particles of a certain size. When a laser beam is sent through flowing water, the laser
light is scattered at right angles by the presence of particles in the water. Optical devices such as
photodiodes collect the scattered light, which can be analyzed to determine the size and number of
particles present in the water sample.
An alert to the presence of particles in the water is useful because it can quickly alert water operators
of a possible contaminant. However, the technology does not allow for the determination of specific
information about the identity of the particle; furthermore, it can only detect particles within a
certain size range. It often cannot differentiate between a grain of sand and a harmful
microorganism. As a result, there can be a high false-positive rate associated with the technology. 12°
August 2005
86
-------
U.S. EPA Office of Water Early Warning Systems
Online turbidimeters, which measure the cloudiness of water using continuous light scattering
technology, are well known and used within the water utility industry. Turbidimeters use heated
tungsten filaments, light-emitting diodes, or lasers. Turbidity monitoring is required on any water
produced for public consumption, and tests alert water operators that filters are failing or that the
water is more conducive to bacterial growth.121 The same technology that supports turbidimeter
technology can also be applied to continuous online monitoring of drinking water distribution
systems to screen for the intentional introduction of pathogens. Optical techniques can overcome
many of the shortfalls of traditional detection methods, which are slow, labor-intensive, and can
have high variability and low recoveries.
An AwwaRF research proj ect determined that Multi-Angle Light Scattering (MALS) can distinguish
between Cryptosporidium and finished water matrix particles at a level great enough to serve as an
early warning tool for water systems. The identification rate of Cryptosporidium parvum oocysts
varied from 11 percent to 45 percent, and false-positive rates varied from 0.3 percent to 3 percent.
The MALS system may be tuned by the user, who must understand that a higher identification rate
will be accompanied by a higher false-positive rate. MALS was also able to differentiate between
different physical states of Cryptosporidium oocysts, including oocysts treated with ozone, heat
treated, or excysted from live untreated oocysts. The technology uses optical fingerprints in order
to identify the different types of oocysts. This finding is important to water system operators
because it allows users to determine if the potential harm from Cryptosporidium oocysts has been
reduced by mitigation actions. The limit of detection was such that MALS could be used as an early
warning tool for water contamination outbreaks. For purified water, the estimated limit of detection
(ELOD) was found to be 7, 0.7, and 0.1 oocysts/mL in 1, 10, and 60 minutes, respectively. For
finished drinking water samples, theELOD was 75,7.5, and 1 oocysts/mLin 1,10, and 60 minutes,
respectively. The researchers concluded that MALS technology has suitable applications to water
distribution system monitoring (Quist et al., 2004122). It was tested at the San Diego State University
"Shadow Bowl," which was organized to test new rapid-response measures to deal with a potential
national security emergency during the 2003 Super Bowl.123
The AwwaRF research project on MALS was developed into a
commercial product called BioSentry™, by JMAR Technologies, Inc.
(Carlsbad, CA; formed from LXT Group and PointSource
Technologies). BioSentry™ beta units have been field tested and the
product is scheduled for commercial production in late 2005. The
system is composed of multiple, laser-illuminated sensor units that are
networked together to provide continuous, real-time monitoring of
water distribution systems. It uses 660 nm wavelength laser light and
a charge-coupled device (CCD) detector to collect the scattered light.
The technology uses Mie scattering, which is measured from the
direction of the incoming light and is not dependent on the specific
wavelength of the light. Although mie scattering has been used in
"D * C.' • t' T"l ¥ ii "D \
other particle counting technology, this is the first application to water '!'
monitoring. The light is then collected from multiple angles, allowing
for the compilation of more information about the particles than is possible with a single collection
point. Desktop computers at a central location utilize LXT's proprietary algorithms to analyze the
shape, size, index of refraction, and internal structure of particles in order to identify the
contaminant. The analysis of all of the factors helps reduce false-positives generated by systems that
August 2005 87
-------
U.S. EPA Office of Water Early Warning Systems
examine particle size alone. BioSentry™ also has the ability to alert system operators to the
presence of unknown or unrecognized contaminants. (Image reproduced with permission from
JMAR.)
The two main challenges facing this technology are false-positives and the level of sensitivity. The
application of light scattering technology using multiple collection angles and complex algorithms
will yield a lower false-positive rate than its predecessor technology that simply counts particles of
a certain size. However, in order for the system to be effective in protecting humans from
contaminated water, the technology must be sensitive enough to detect microorganisms well below
dangerous levels, perhaps one spore in 100 liters. JMAR hopes to be able to more accurately assess
the effectiveness of BioSentry™ upon completion of field tests at water utilities in 2005.124
Rustek, Ltd. (UK) has developed a Multi-Angle Laser Light Scattering (MALLS) technology that
combined with pattern recognition techniques increases the technology's ability to monitor for
microbial contaminants.125 The Computing Research Centre at Sheffield Hallam University,
(Sheffield, UK), is using the MALLS device patented by Rustek, Ltd. The technology is currently
being used to detect bacteria in different elements of the water industry, including water companies,
the bottled water industry, and breweries; in addition, the technology has applications in the medical
field. The device analyzes how light is scattered in order to determine the particle content in the
water sample. When a laser beam is projected through a sample, it is disrupted by particles. The
resulting light picture, or how the laser beam appears after it is disrupted by particles, will vary in
both the direction and speed of the light beams. Thus, the fluctuation in amplitude and frequency
of intensity are measured to determine the amount of microbes present in the sample.126
7.3 Potentially Adaptable Technology
The technologies presented in this section are potentially adaptable for finished water and
distribution systems, but have not reached the fully commercial stage or undergone third party
verification. The technologies either have been developed originally for source water or for non-
water media (e.g., air).
7.3.1 Fiber Optic-Based Biosensor
RAPTOR™, developed by the U.S. Naval Research Laboratory and
Research International (London, UK), is a portable, rapid, automatic
fiuorometric assay system for monitoring biological contaminants,
toxins, explosives, and chemical contaminants.127 The self-contained
instrument integrates optics, fluidics, electronics, and software and is
suitable for laboratory and field assays. It performs user-defined,
multi-step assay protocols for monitoring fiuorescently labeled
chemical reactions occurring on the surface of each of the system's
four disposable optical waveguide sensors. Research International's
RAPTOR™
biosensor systems are based on monolayer receptor-ligand reactions ' ;
J / , , j , , (Research International)
taking place on the surface of injection molded polystyrene
waveguides. The baseline protocol used to identify specific pathogens is called the "sandwich
format" fiuoroimmunoassay. Toxins and bacteria such as ricin and B. anthracis have been detected
at levels below <1.0 ng/mL and 100 cfu/mL, respectively. According to Research International,
August 2005 88
-------
U.S. EPA Office of Water Early Warning Systems
RAPTOR™ is capable of real-time detection of microbial pathogens with or without conventional
culture. The portable unit can simultaneously process four analytes in 7 to 12 minutes. The hand-
held model in development will be capable of 12 to 16 simultaneous assays (Bunk, 2002). (Image
reproduced with permission from Research International.)
Daniel Lim's laboratory (University of South Florida) was involved in the development of
RAPTOR™ and is currently in the process of placing a prototype array biosensor linked to a
filtration/concentration system for automated, continuous online monitoring of potable water at a
local water utility. The filtration/concentration system uses a hollow fiber filter that concentrates
microorganisms from large volumes of water. The system will be back flushed and the back flush
will be sent directly to the biosensor. It is anticipated that the biosensor will be able to identify
specific microorganisms in the distribution system, including biothreat agents, should they occur
(Daniel Lim, University of South Florida, personal communication).
7.5.2 Dye-LoadedMicrospheres
The Luminex® Corporation's (Riverside, CA) xMAP®
system consists of 5.6 micrometer polystyrene
microsphere particles that are loaded with red and
infrared fluorophores in ratios that allow for 100
distinct color-codes. Each microsphere can be coated
with a separate capture molecule which can be
involved in either nucleic acid hybridization, antibody
recognition, areceptor-ligand reaction, or an enzymatic
reaction. The reaction times are about three times Lumiliesg SM.\PS (Lumin«Coi-p.)
faster than standard microarrays because the
microspheres are in solution have 3-D exposure allowing for nearly solution-phase kinetics, whereas
flat microarrays are limited by solid-phase kinetics. The microspheres pass through the detection
chamber single file and are optically measured.128 Luminex® markets bench top readers for
genomics and proteomics applications. Researchers from Lawrence Livermore National Laboratory
(LLNL) utilized Luminex® technology to develop an Autonomous Pathogen Detection System
(APDS). The system has an automated sample preparation module, based on sequential injection
analysis (SIA). The APDS interfaces aerosol sampling with multiplexed microsphere immunoassay-
flow cytometric detection. The system performed well over five days of unattended continuous
operation (Hindson et al., 2004).129 (Image reproduced with permission from Luminex Corp.)
7.5.5 Detection of ATP
Several ATP detection systems are available for use with water samples. Another commercially
available system deserves note because it has a long-standing history of use in the food and beverage
industry and may be applicable for finished drinking water. Celsis-Lumac (Landgraaf, The
Netherlands) sells a cellular ATP detection system, RapiScreen™, that can be used to measure
bacteria on the surfaces of meat products and in beverages.130 Hygiena (Camarillo, CA), which also
markets some of the Celsis-Lumac technology, has developed liquid stable reagents and freeze-dried
reagents for ATP bioluminescence applications.131
August 2005 89
-------
U.S. EPA Office of Water
Early Warning Systems
BioFl.ish™ (Innovative Biosensors Inc.)
7.3.4 Cell-Based Biosensor
The Massachusetts Institute of Technology
(MIT) developed Cellular Analysis and
Notification of Antigen Risk and Yields
(CANARY™), which was licensed to
Innovative Biosensors, Inc. (College Park,
MD), and is now marketed as the BioFlash™
System. With this system, transgenic B-cells
express antibodies against target antigens.
When the target antigen is present in the test
sample, the B-cells emit light (via green
fluorescent protein), which is detected by a
portable luminometer. Either liquid or solid samples can be tested. The assay protocol for liquid
samples has five steps, which collectively take about 5 minutes to complete, including reading time.
The sample preparation removes inhibitors such as chlorine (Hollie Kephart, Innovative Biosensors,
Inc., personal communication). The original development paper cites the sensitivity as 200 cfu for
Y. pestis (in 20 uL reaction volume), 1,000 cfu B. anthracis (from a swab rinsed in 1 mL of
extraction buffer), and 500 pfu vaccinia virus (in 20 uL reaction volume) (Rider et al., 2003). A
separate cell line must be developed for each target antigen. This system has not been verified by
a third party. The company markets CANARY™ as having broad applications in food testing,
animal health, biodefense, and human health care, including drug discovery and development, and
disease diagnosis.132 (Image reproduced with permission from Innovative Biosensors, Inc.)
7.5.5 Polymerase Chain Reaction
Polymerase chain reaction (PCR) is an analytical technique that detects/identifies organisms by
targeting their nucleic acids (DNA/RNA). PCR is a highly developed technology for molecular
biology applications. It has broad applicability for almost any situation where DNA needs to be
detected and identified.
Nucleic acids are synthesized in great numbers and subsequently identified by various techniques.
In terms of security applications, it can be used to detect and identify biological contaminants. The
following illustrates the PCR process:
• Microbial cells are disrupted to expose the DNA/RNA. Disruptive techniques include use
of lytic enzymes, freeze-thaw cycles, or bead-beating.
• DNA/RNA is extracted and purified to remove interferences such as folic acids.
• Various reagents are added (e.g., DNA primers, excess nucleotide bases, and enzymes to
produce DNA).
• DNA is synthesized in great numbers via thermal cycling. The amplification of DNA
occurs through a series of 30 to 45 temperature cycles producing billions of copies.
• The amplified target DNA is detected using various methods including electrophoresis,
fluorescent gene probes, and fluorescence melting curves.
PCR is a sensitive and potentially rapid detection method. It can detect any organism that contains
nucleic acids, including viruses, bacteria, and protozoa. The assay is selective, can be used to screen
August 2005
90
-------
U.S. EPA Office of Water Early Warning Systems
for selected contaminants, and is getting easier to conduct using prepackaged reagents. However,
the method cannot distinguish between live and dead microorganisms and can be negatively affected
by natural interferences such as soil-derived humic and fulvic acids. Because PCRreactions are also
carried out in small volumes, samples need to be reduced to microliter volumes. Several portable
PCR-based identification devices are available.
The Ruggedized Advanced Pathogen Identification Device (RAPID), developed by Idaho
Technologies (Salt Lake City, UT), has been used extensively by the military.133 It can screen up
to eight contaminants simultaneously, uses a closed system to reduce contamination, and all reagents
are freeze-dried for convenience. It is an automated system which includes cell disruption and DNA
extraction-purification necessary for PCR. Reaction volume is 10 to 20 uL. The thermal cycler has
preprogrammed tests and automatic data interpretation. It is field deployable weighing 35 pounds
and can analyze samples in 30 minutes. The microbes that can be detected by the RAPID System
are listed below. The nucleic acid limit of detection (NALOD) is indicated where available in
parenthesis (units are genomic equivalent [GE] or plaque forming unit equivalents [pfu-e]; Tuck et
al., 2005):
• Bacillus anthracis (5 GE)
• Brucellaspp. (10 to 20 GE)
• Salmonella spp.
• Yersinia pestis (5 to 40 GE)
• Francisella tularensis (2.3 to 7 GE)
• E. co/zO157:H7
• Listeria monocytogenes
• Campylobacter
• Clostridium botulinum
• Orthopox (200 to 350 GE) RAPID (JBAIDS)
• Smallpox (40 to 125 GE) (Idaho Technologies)
• Qfever(5to31GE)
• Typhus (10 GE)
• Glanders (5 GE)
• Ebola virus (260 to 706 pfu-e)
• Marburg virus (1.9 to 4 pfu-e)
• Easter Equine Encephalitis virus (20 to 5,000 pfu-e)
RAPID can be used to detect pathogens in water samples. The unit costs around $55,000 with each
test costing $50. EPA and the Army are developing a prototype for water. Currently, EPA's ETV
program is investigating the sensitivity, interferences, and cross-reactivity of the unit.134 In
September 2003, Idaho Technology was awarded the Joint Biological Agent Identification and
Diagnostic System (JBAIDS) contract.135 In March of 2005, RAPID underwent a two-week
operational test at Brooks City-Base, TX. The Air Force Operational Test and Evaluation Center,
based at Kirtland Air Force Base, NM, took the lead on the exercise, while the Army Medical
Department provided training and technical assistance. After validation by a joint-service Data
Authentication Group, the operational test results will be forwarded to the Joint Program Executive
Office for Chemical and Biological Defense for final approval. If approved, JBAIDS will enter
full-rate production in September 2005 and the DOD will distribute 450 systems throughout the
services over the next three years.136 (Image reproduced with permission from Idaho Technologies.)
August 2005 91
-------
U.S. EPA Office of Water
Early Warning Systems
RAZOR
PathFinclei™ Pouches
(Idaho Technologies)
Idaho Technology's latest
portable PCR device, called
RAZOR, comes with
prepackaged freeze-dried
reagents for 12 different targets
already loaded into the clear
flexible plastic PathFinder™
reaction pouches. Reagent
grade water is required but the
sample ports attached to the
reaction pouches are designed
so that volume measuring is not required. The reaction pouches are loaded into the cycling device
and results are obtained after about 30 minutes. The device weighs nine pounds, including
battery.137 (Images reproduced with permission from Idaho Technologies.)
Bio-Seeq™, developed and commercially available through Smiths
Detection (Edgewood, MD), is a hand-held PCR-based biological
detector. A sample preparation cartridge allows samples to be
taken in the field and tests run on the spot. All the necessary
reagents, filters, and mixing chemicals required to process a
biological or viral sample are indued in the sample prep cartridge,
eliminating the need for pipettors, tips, and sample vials. The unit
has six detection modules (thermocycler/optics modules) that
perform the thermal cycling, optical reading, and alarm detection
for each test. Each module has two independent optical channels
that can be used during a single test. With suitable reagents, these
channels can allow the user to run a target sample with a positive control in the same tube,
eliminating the need to prepare a separate positive control. The device is capable of detecting 1 cfu
(in -28 \\L sample volume) in 30 minutes and costs $25,000.138 (Image reproduced with permission
from Smiths Detection.)
LLNL has developed a "hand-held nucleic acid analyzer" (HANAA) based
on real-time PCR (TaqMan). This specific technology has been licensed to
Cepheid, a Sunnyvale, CA-based company that is developing biosensors for
the U.S. Postal Service. HANAA can identify an organism in less than 10
minutes (Perkel, 2003).139 It requires the operator to prepare the sample by
adding reagents to a reaction tube and selecting which pathogen will be
targeted. Water Environment Research Foundation (WERF) has tested
HANAA with the waterborne pathogens, Cryptosporidium parvum and
E. coli O157:H7 (photo from LLNL website).
1 HANAA fLaw
Livermore National Lab)
Cepheid, Inc. (Sunnyvale, VA) Smart Cycler® XC is a portable PCR
machine that uses the company's patented I-CORE (integrated cooling/heating optics reaction)
modules to amplify up to four targets simultaneously from one sample. The amplification is
monitored in real-time and can be completed in as little as 30 minutes. The company's newer
GeneXpert System includes a cartridge-based sample preparation system that takes about five
minutes to prepare.140 In 2001, the U.S. Centers for Disease Control and Prevention (CDC)
Bio-Seeq™ (Smiths Detection)
August 2005
92
-------
U.S. EPA Office of Water
Early Warning Systems
Smart Cvclert
(Cepheid Inc.)
GeneXpei tl
developed and validated test
kits for several bio-threat
agents optimized for use with
the Smart Cycler®. The
Laboratory Response
Network (LRN), uses the
CDC validated kits and Smart
Cycler® to provide
nationwide screening and
reference testing to respond
to a bioterrorism event. In
2002, Cepheid, Inc.,
delivered to the U.S. Army
Medical Research Institute of Infectious Diseases (US AMRIID) field-ready DNA test kits for rapid
detection of four bio-threat agents, Bacillus anthracis (anthrax), Yersiniapestis (plague), Francisella
tularensis (tularemia), and Clostridium botulinum (botulism). Developed in collaboration with
USAMRIID under a DOD contract, the tests combine DNA sequences identified and validated by
US AMRIID for bio-threat agent detection with Cepheid's proprietary reagent formulations and
"freeze-dry" processing for prolonged stability and ease-of-use.141 This technology is being used
in many post office sorting facilities throughout the United States. The system has a sensitivity of
< 30 spores/reaction in water or buffer, a false-positive rate target of < 1:500,000 samples (99.9998
percent), no cross-reactivity with nearest neighbor organisms, and a non-determinate rate of
< 1 percent. The anthrax assay has been validated by third party government agency evaluations.
In 2005, the company expects to introduce a triplex cartridge for anthrax, tularemia, and plague, and
a separate Orthopox cartridge (Jaymee Rosenberger, Cepheid, personal communication). For use
with drinking water samples, concentration techniques would be needed to reduce large sample
volumes to appropriately smaller volumes. (Images reproduced with permission from Cepheid, Inc.)
The PathAlert™ Detection System from Invitrogen Federal
Systems (Frederick, MD), in conjunction with the 2100
Bioanalyzer micro-fiuidics based electrophoresis system from
Agilent Technologies (Palo Alto, CA),142 is a PCR based system
that is able to detect biothreat agents in a single agent assay or
multi-agent assay formats. Existing products include single agent
assays for B. anthracis, Y. pestis, vaccinia, and F. tularensis with
multiple target loci for each reaction. In addition, a multi-agent
single reaction assay is available for the detection of these four
agents. Under development are assays for additional water borne
pathogens such as E. coli O157:H7, Cryptosporidium parvum,
Giardia lambia, Salmonella species, Shigella sp., and others.
prepackaged or custom formats, where the user is able to pick four to six targets of choice for
inclusion in a single multi-agent assay. PathAlert™ was tested in June 2004 as part of EPA's ETV
program. During this testing, the system was able to overcome environmental inhibition from fulvic
and humic acids. The system is able to operate in a standard stationary or mobile laboratory
environment (Willem Folkerts, Invitrogen Federal Systems, personal communication). Although
PathAlert™ is not marketed as portable, it is included in this report because the developers have
specifically addressed biothreat agents and the technology has been tested in EPA's ETV program,
PathAlert™
(Iiivitrogen Federal Systems)
The product is available in
August 2005
93
-------
U.S. EPA Office of Water
Early Warning Systems
Deck A
TIGER , {Jbis Robo
DeckB
illternatL(mal)
as well as in a technology readiness assessment (TRA) conducted at the U.S. Army's Dugway
Proving Ground. (Image reproduced with permission from Invitrogen.)
One method developed around DNA base
composition and PCR, that has successfully
demonstrated proof-of-principle, is called
Triangulation Identification Genetic Evaluation of
Risks (TIGER). TIGER has been developed by
Isis Pharmaceuticals, Inc. (within their Ibis
Therapeutics program, Carlsbad, CA), in
collaboration with SAIC, and funded by
DARPA.143 The TIGER biosensor system can
identify a broad range of infectious organisms,
including known, unknown, non-culturable, or
bioengineered elements, within a few hours. PCR primers are designed to place unknown organisms
with their related neighbors, and multiple primer pairs target multiple locations within pathogen
genomes. Mass spectrometry is used to obtain mass signature data. RoboDesign International, Inc.
(Carlsbad, CA) is developing TIGER 2.0, which will be an automated system and will require
minimal intervention from a technician (Bunk, 2002). The footprint will be ~ 8 x 8 feet and is being
transitioned to USAMRIID and CDC. The culture-independent methodology allows for flexibility
with respect to the types of starting sample that can be used (e.g., blood, urine, soil, other
environmental samples) (Kumar Hari, Ibis Therapeutics, personal communication). Although this
technology is not field portable or online, it presents an advancement beyond standard PCR
approaches where primers are designed for detecting only one known pathogen at a time. (Image
reproduced with permission from Ibis Theraputics.)
7.5.6 Bio-Optoelectronic Sensor Systems (BOSS)
DARPA funds the bio-optoelectronic sensor systems (BOSS) center, which includes team members
from the University of California at Berkeley, Colorado State University, Columbia University,
Georgia Institute of Technology (Georgia Tech), the University of Illinois at Urbana-Champaign,
the University of Michigan at Ann-Arbor, and the University of Texas at Austin, who collaborate
to develop technologies to detect chemical and biological warfare threats.144 Georgia Tech's,
Applied Sensors Laboratory is using fiberoptic evanescent wave spectroscopy in the mid-infrared
region to sense capture-target molecule interactions. The sensing area is coated with a thin polymer
layer (can be a molecularly imprinted polymer; see Section 7.4.13) that forms a hydrophobic
membrane, serving both as an extractor phase to enrich hydrophobic analytes in close vicinity of the
sensor surface and as a suppressor of matrix interference by water absorptions.145 Georgia Tech has
developed sensors for marine and groundwater applications (EPA, 2003). The developers claim that
"the sensor is fast, highly sensitive, and provides real-time direct measurement with no additional
steps or consumable reagents...[and is] capable of detecting a wide variety of chemical and
biological species in air, water, and biological samples" (Bodurow, 2005).
7.5.7 Surface Plasmon Resonance (SPR)
Surface plasmon resonance (SPR), detects changes in mass by measuring changes in refractive
index.146 The sensor chip consists of a glass support surface coated with a thin layer of gold. The
August 2005
94
-------
U.S. EPA Office of Water Early Warning Systems
gold surface can be modified in a variety of ways to immobilize different compounds. For example,
if the gold surface is modified with a carboxmethylated dextran layer, various biomolecules can be
attached to this hydrophilic layer without causing denaturation. When a sample is passed over the
chip surface (via microfluidics), molecules that bind/interact with the immobilized target are
captured. Interactions between proteins, nucleic acids, lipids, carbohydrates, and even whole cells
can then be studied. When binding occurs, the mass increases, and when disassociation occurs, the
mass deceases. These changes in mass can be detected as they occur and yield quantitative
information such as kinetics, affinity, and concentration of the molecule in the sample. Binding of
molecules as small as 100 Daltons can be detected. Portable instruments are being developed (Karl
Booksh, DMS Lab, University of Washington147). A related technique, called surface plasmon-
coupled emission (SPCE), is being developed at the University of Maryland, and has the potential
to improve sensitivity 1000-fold over other fluorescence technologies.148
Nomadics® Advanced Instrumentation Group
(Stillwater, OK) offers a Surface Plasmon
Resonance (SPR) Evaluation Module for
researchers interested in studying specific
chemical and biological contaminants on a
portable SPR platform.149 The Nomadics
Evaluation Module is based on Texas
Instruments' (Dallas, Texas) Spreeta™
biosensor that enables real-time quantitative
measurement of bio-molecular interactions.
The Module contains 50 sensors (chips), a flow
cell with an electronic PC interface control, and a Windows based operating system. Spreeta's design
encapsulates the entire SPR optical system making the device compact and able to be integrated into
various instrument designs. The sensor can be used to detect and quantify the presence of specific
contaminants for applications such as agriculture, water quality, medical, and food safety
applications.150 Researchers at Stanford University tested Spreeta and conclude that the sensor
"shows promise as an inexpensive, portable, and accurate tool for bio-analytical applications in
laboratory and clinical settings" (Whelan et al., 2002).151 The same research team reported that the
analyte concentration in their test method corresponds to the detection of 90 fmol IgG (Whelan and
Zare, 2003). In 2006, Nomadics anticipates launching a Spreeta-based Life Science platform.
(Images reproduced with permission from Nomadics.)
7.3.8 Electrochemiluminescence (ECL)
ECL is a detection technology that utilizes light generated by the oxidation and reduction of a
ruthenium metal ion as a labeling system. Capture molecules (e.g., antibodies) are absorbed onto
a support surface, such as magnetic beads or a microarray plate. The addition of target molecules
from the sample and ruthenylated antibodies creates an antibody sandwich which is detected when
the ruthenylated antibodies are stimulated to glow with an electrode. Background signal is limited
because the electrodes only stimulate nearby ruthenium, quenching is not a problem with the 620 nm
emitted light.152 As with other antibody-based technologies, volume reduction would most likely
be required for ECL technologies to be adapted for distribution systems.
August 2005 95
-------
U.S. EPA Office of Water
Early Warning Systems
ECL was developed by BioVeris (formerly IGEN; Gaithersburg, MD).
BioVeris' Bio Verify Tests utilize two antibodies which recognize the
pathogen or toxin, one immobilized on paramagnetic microparticles and
the other labeled with BioVeris' BV-TAG™ label. The sample mixed
with the antibody reagents is loaded into the flow based MlM Analyzer,
which transports this assay mixture into a measurement cell and collects
the microparticles on an electrode. The electrode stimulates the
BV-TAG™ labels bound (via the antibodies and the spore) to the
microparticles and the emitted light is measured. Assays that are
available for Ml M include Botulinum neurotoxins (A, B, E, F), anthrax,
ricin, Staphylococcal enterotoxins A and B, E. coli 0157:H7, Salmonella,
Listeria, and Campylobacter. The M1M comes packed in a transportable
suitcase with a separate reagent suitcase.153 The company markets the
system for biodefense applications and claims it can be used for research
and environmental samples.
M1M Analvztr
(BioVeih)
Meso Scale Defense154 (Gaithersburg, MD), a division of Meso Scale Discovery, also sells systems
utilizing ECL technology. The company's MULTI-ARRAY™ and MULTI-SPOT™ microplates
have electrodes integrated into the bottom of the plate. Capture molecules are immobilized onto the
plates, then sample and MSD-TAG™ are washed over the array. Antibody sandwiches, which are
formed when target molecules are present, are detected by a reader. Meso Scale Defense has
benchtop readers on the market and is designing a portable cartridge reader for first responders.155
7.4 Emerging Technology
The emerging technologies are prototype field or laboratory devices as well as technological
advancements that could be integrated into a system. Systems discussed include advancements in
immunoassays, proof-of-concept and prototype microchip technologies, micro beads, and light
scattering technology. Immunoassay technology in general is described in Section 7.2. Some of the
microchip technologies presented are used in commercial products for other applications such as
genomics research or clinical analysis. The potential range of applications of these technologies for
other uses is diverse and includes sensors for drinking water monitoring. Whether or not any of
these technologies are ever used to develop products specifically for use with drinking water
depends mostly on cost. Because it is beyond the scope of this document to analyze economic
factors, technologies that could be developed into products for use in drinking water distribution
systems will be presented regardless of whether costs might eventually prevent their development
for water application. In addition, it is difficult to determine what the potential detection limits
could be for these technologies because detection limits are based on specific contaminants in a
specified matrix and these technologies have not been adapted for distribution system water or the
range of threat agents that are of concern.
7.4.1 Lateral Flow Assay
NASA's Jet Propulsion Laboratory has developed a quantitative lateral flow assay (QLFA) for
testing drinking water samples in space.156 Depending on the design and the specific antibodies
used, the test strip yields a rough count of the total cfus in the water sample, and a preliminary
classification as to the types of organisms present—for example, viruses versus different major
August 2005
96
-------
U.S. EPA Office of Water Early Warning Systems
classes of bacteria. The test takes only minutes to conduct and does not require growth of bacteria.
Relatively low levels of antigens can be detected using newer fluorescent dyes, such as Qdots®,
which are much brighter than traditional fluorescent labels (see more below157).
7.4.2 Labels
Quantum Dot Company's (Hayward, CA) Qdots® are nanocrystal
spheres that fluoresce at a variety of different wavelengths. The exterior
surface can be coated with biological molecules such as DNA,
antibodies, or receptors. Although colloidally dispersed pigments occur
naturally and have been used in paints for centuries, the breakthrough has
been creating fluorescent versions that are water soluble and non-toxic
to cells.158 Qdots® are currently being marketed for labeling subcellular
components of living cells for imaging applications.159 Other researchers
are also developing quantum dot technologies (Gorman, 2003).16° QDot^t
Quantum dots have the ability to quantitatively detect biological (Quantum Dot Co.)
molecules in samples. The EPA Office of Ground Water and Drinking
Water (OGWDW) Technical Support Center located in Cincinnati, OH, is using quantum dots to
as part of a sensor technology for determining the occurrence and prevalence of cyanobacteria and
their toxins in surface and finished water (Gerald Stelma, EPA, personal communication). The
biosensor EPA is developing will be portable for use in the field and eventually adapted for
continuous monitoring. The goal of EPA's research is to develop molecular methods to detect
cyanobacteria and to simultaneously extract and detect cyanotoxins of interest to the Agency.161
(Image reproduced with permission from Quantum Dot Co.)
Another newer labeling technology is Upconverting Phosphor Technology™ (UPT). SRI
International (Menlo Park, CA), in collaboration with OraSure Technologies (Bethlehem, PA) and
under DARPA support, has developed a lightweight, battery operated, hand-held sensor to detect
multiple pathogens (bacteria, viruses) and their toxins simultaneously.162 The system uses UPT™
to color-code multiple pathogens simultaneously in a lateral flow immunoassay test strip.163 The
upconverting phosphor reporters emit visible light upon excitation by near infrared light and have
the advantages of (1) single particle detection sensitivity, (2) multiplexing, (3) no autofluorescence,
and (4) no photobleaching. SRI has thus far developed ten UPT™ phosphors, each producing a
different color, which allows for the simultaneous detection of several contaminants in the same
sample.164 The sensor can detect 10 to 1,000 picograms (pg)/mL of small (e.g., virus, toxin) target
antigen in < 15 min with a sample volume of less than 300 uL. For spores and bacteria, the
sensitivity is as low as 1000 cfu/mL. So far, this technology has been developed for sampling of
biological fluids (oral fluids, blood, etc.) but future research aims to apply UPT™ to environmental
testing, including drinking water.165 OraSure Technologies owns commercial rights and may
develop the technology for other emerging applications, such as biological warfare defense,
combinatorial chemistry, biomolecular screening, medical diagnostics, and drug testing.
7.4.3 Magnetic Beads
Dynabeads®, developed by Dynal Biotech LLC (Oslo, Norway), is a product used for rapid
separation and detection of microbes, nucleic acids, proteins, and other biomolecules in liquid or
viscous samples. Dynabeads® technology is based on immunomagnetic separation. The polymer
August 2005 97
-------
U.S. EPA Office of Water Early Warning Systems
shells of these microscopic (1 to 4.5 um) beads can be coated with a variety of ligands (antibodies,
oligos, proteins, DNA/RNA probes, etc.) that will bind to specific targets. The resulting target-bead
complexes are then isolated using magnetic attraction and detected using an ultraviolet laser
system.166 The variable combinations of bead-size and ligand type facilitate detection and
identification of a wide variety of targets. This method has successfully detected E. coli in water
samples in less than five hours (Pyle et al., 1999). Although Dynabeads® are not being marketed
by Dynal as tools for CBW detection, they could be used in flow cytometry devices and other
technology that can detect small labled particles.
7.4.4 Flow-Through Columns
The Biodetection Enabling Analyte Delivery System (BEADS), developed by the PNNL (Richland,
WA), is a portable, automated front-end sample preparation device for pathogen detection. The
system features micro-sized glass, polymer, or magnetic beads coated with antibodies specific to a
chemical or biological species of interest. The beads are color-coded to differentiate their specific
chemistries for extracting and detecting multiple pathogen signatures. Liquid samples flow over
renewable bead-based immunoassay columns, which serve to isolate and concentrate whole cells,
proteins, nucleic acids, and/or chemicals as they bind with the beads. In addition to sample
purification and concentration, BEADS has its own PCR detector, or can be linked to other detector
technologies. No human interaction with the system is required for sample preparation or analysis,
and field test results can be sent electronically to a remote location. The BEADS system has
successfully detected trinitrotoluene (TNT), pesticide/herbicide, botulinum toxin, E. coli, and
anthrax in a process that takes about four hours to complete (ACS, 2002).167 This technology has
not been third party verified.
7.4.5 Raman Spectroscopy
Biopraxis (San Diego, CA) is developing a reagentless, portable, biosensor, whose first version is
known as "Doodlebug." The biochip has biomolecules immobilized on a surface-enhanced
Raman168 scattering (SERS)-active metal surface. When a sample is added to the surface of the chip,
the immobilized capture biomolecules selectively bind their ligands. The chip reader, a Raman
microscope, illuminates the surface of the chip with a laser and the scattered light is collected. The
wavelengths and intensities of the scattered light are used to analyze the unique molecular structure
of the cross-reactions. The technology can detect chemicals (including explosives) and biologicals.
Biopraxis is developing a biochip to detect 8 to 10 different targets at the same time (Bunk, 2002).
A WERF study showed that Doodlebug could distinguish between six Legionella species, recently
passaged (fresh) oocysts from six C. parvum Genotype 2 strains, three Genotype 1 strains, a
C. meleagridis strain, and a Giardia sample. Results take about 60 seconds to obtain. Experiments
involving the impact of environmental and water treatment conditions suggested that this technology
will be able to differentiate between viable and nonviable, and possibly injured, organisms. The
SERS fingerprint may even be useful in determining the "age" of an oocyst (e.g., whether it is too
old to be infectious). The sensitivity of the technique eliminates the need for amplification
techniques, such as PCR, fluorescent labels, and enzymatic reactions, thereby greatly reducing the
potential for false responses from sample constituents that either mimic or inhibit the signals from
labels or interfere with enzymatic reactions (Grow et al., 2003).169 Several other companies have
portable Raman spectroscopy-based instruments.170
August 2005 98
-------
U.S. EPA Office of Water
Early Warning Systems
Sen-ZTJI (CombiMatrix Corp.)
7.4.6 Microelectrode Arrays
The CombiMatrix Corporation (Mukilteo, WA) is in the
process of testing its biothreat detection system, Sen-Z™,171
a hand-held, portable, self-contained unit that captures and
electronically detects a range of threat agents (e.g., anthrax
spores, smallpox virus particles, ricin, and saxitoxin).
CombiMatrix's core technology consists of microarrays of
1,000 to 12,000 microelectrodes in one cm2. Overlaying each
microelectrode is a porous reaction layer which serves as a
reaction "flask." The microelectrode creates a local change
in pH, which dictates where on the microarray capture
molecules are synthesized or deposited. CombiMatrix
microarrays have been made to detect DNA hybridization and antigen-antibody reactions. Some
of the key features of the Sen-Z™ are multiplex immunochemical assays; a platform that can be
quickly configured to detect a broad range of threat agents; real-time fluorescent-free electronic
detection of signals by electrochemical methods; automated hands-free sample collection,
preparation, detection, and analysis; and high sensitivity (ricin at 60 pg/mL). At the present time,
this technology focuses on air sampling, but the company believes that the integration of agent
isolation and processing technology could adapt this product for a water monitoring system (David
Danley, CombiMatrix, personal communication). (Image reproduced with permission from
CombiMatrix.)
7.4.7 Microarray of Gel-Immobilized Compounds
The Biochip Technology Center at Argonne National Laboratory developed a reuseable "Micro
Array of Gel-Immobilized Compounds" (MAGIChip™) that can perform thousands of biological
reactions in a few seconds. MAGI Chip™ is a small glass slide with up to 10,000 3-D gel pads that
serve as micro-test tubes. Robots load the pads with DNA or protein fragments from bacteria,
viruses, or chemicals. Bench top equipment is required to analyze the chips. Researchers at
Argonne's Biochip Technology Center are developing new applications for the biochip, writing
faster sample analysis programs, and working to shrink portable biochip readers. This technology
has been used for gene expression, diagnosis and
monitoring of genetic diseases, microbial analysis in
environmental cleanups and agriculture, routine protein
analysis of blood and urine, exploration for life in outer
space, and forensic DNA testing.172 Although they have
the potential for detecting biological contaminants in an
EWS context, they have not yet been adapted for this
use. In addition, volume reduction would most likely be
required for this technology to be adapted for
distribution systems.
Control
Chip Cin»r Board
7.4.8 MagneticMicrobeads
The Bead ARray Counter (BARC) chip, developed at
the Naval Research Laboratory, consists of an array of
QOH >'
I1IIL I
BARC Chip-^
r ••^•••^^^•^^^^^H
BARC Chip (Naval Research Lab)
August 2005
99
-------
U.S. EPA Office of Water Early Warning Systems
DNA spots immobilized on a surface.173 Hybridization of sample DNA with chip immobilized DNA
is detected by magnetic beads 1 to 3 um in diameter. The chip's magnetic field microsensors are
um-scale, wire-like structures made with giant magnetoresistive (GMR) material, which can detect
individual magnetic beads and is more sensitive and more compact than the optical detectors
required for fluorescence-based techniques (Whitman et al., 2001; Tamanaha et al., 2002; Rife et
al., 2003174). The developers envision that millions of sub-micrometer GMR elements enabling
simultaneous detection of thousands of DNA sequences with high sensitivity and dynamic range will
be possible as the technology advances.175 This technology is not yet field ready. (Photo from NRL
website.)
7.4.9 DNA Microarrays
DNA microarrays (DNA chips) can contain 100,000 different spots of DNA printed on a glass
microscope slide (Fitzgerald, 2002). Alternatively, photolithography and solid-phase chemistry can
be used to produce microarrays containing 500,000 oligonucleotide (single stranded DNA) probes
within 1.28 cm2 (Affymetrix Genechip®).176 When a sample with unknown DNA is exposed to the
spots, it will hybridize (match-up) with the spots of DNA that are complementary. The sample DNA
is labeled (through a PCR reaction) so a chip reader device is able to detect where on the microarray
hybridization occurs. The microarrays could be designed to detect a multitude of sequences that
would be unique to specific pathogens of concern.177 There are several companies producing and
custom designing microarrays. Microarrays are widely used for genomics research,178 but face
sample volume reduction problems for use with environmental samples and drinking water.
7.4.10 Micro-Cantilever System
The VeriScan™ 3000 System, produced by Protiveris
(Rockville, MD),179 utilizes technology licensed from
Oak Ridge National Laboratory.180 The bench-top
system can conduct 64 simultaneous assays using a
proprietary Micro-Electro-Mechanical System181
(MEMS)-based biochip, a patented laser reader
technology, microfluidics, and advanced custom analysis
software. The biochip has an array of microcantilevers
that can detect interactions between proteins, antibodies,
antigens, or DNA. The system, which does not require VeriScan™ 3I)00 Svstem (P| otive,is)
labels or amplification, delivers data in real-time as the
binding interactions take place. The lower limit of detection is 0.2 ng/mL, making it competitive
with the more traditional ELISA assay (Daviss, 2004). As with other technologies for detecting
microbes on microchip platforms, sample volume reduction for the concentration of cells would be
necessary before the technology is viable for drinking water. (Images reproduced with permission
from Protiveris.)
7.4.11 Photoluminescent Biochips
latroQuest's (Verdun, Canada) Bio-Alloy™ biochips are made of silicon-based semiconductor
materials that are nanostructured and chemically modified to bind to a variety of molecules,
including antibodies, enzymes, nucleotides, and chemicals as recognition elements. The underlying
August 2005 100
-------
U.S. EPA Office of Water
Early Warning Systems
detection principle, based on a photoluminescence response, relies on quantum confinement and
changes in the surface energy state when the material is excited with low-power blue LED light.
Upon affinity binding of target contaminants to the recognition elements linked to the surface,
surface energy perturbations result in an immediate change in the photoluminescence response,
which is detected as an increase in green light intensity. Different formats of the material can be
produced, including chips, particles, or microspheres for more end-product versatility. latroQuest
has a portable demonstration system, but no products are yet developed.182 The company has been
awarded a $3 million (Canadian dollars) development contract from the CRTI Counter-Terrorism
Initiative.
7.4.12 Polymer Microbeads - Taste Chip
Scientists at the University of Texas at Austin have
developed the "electronic taste chip," based on a system of
polymer microbeads that mimic human taste buds.183 The
system has been licensed to LabNow, Inc., for commercial
product development. This multi-sensor array technology
can generate digital fingerprints of complex fluids in near-
real-time. The device functions by using a combination of
micromachining, nano-chemistry sensing schemes, molecular
engineering of receptor sites, and pattern recognition
protocols to detect a variety of biological and chemical
contaminants (e.g., electrolytes, toxins, drugs, metabolites,
bacteria, blood products). Taste chips can be adapted within months to respond to new analytes,
which provide for customized applications such as drinking water analysis. Their analytical
characteristics (sensitivity, selectivity, detection thresholds, assay variance) have been shown to be
comparable to or better than many well-established macroscopic analytical methods (Goodey and
McDevitt, 2003; McCleskey et al., 2003; Kirby et al, 2004). Fully developed pro to type instruments
and customized microchip sets have been designed, constructed, and tested in numerous application
areas including homeland defense, where hand-held units have been sent to the Defense Threat
Reduction Agency for further testing (John T. McDevitt, U. Texas, personal communication; image
reproduced with permission).
7.4.13 Molecularly Imprinted Polymers
Molecularly imprinted polymers (MIP) are synthetic receptors which can be designed for a range
of toxins and some microorganisms.184 MIPs have greater stability, being able to withstand climate
extremes and larger sensitivity ranges, than antibodies (Haupt, 2002).185 The technology is being
incorporated into the UK Ministry of Defense's development of an integrated biological detection
system for battlefield use against biowarfare agents (Bunk, 2002). Some analytes for which MIPs
have been developed include the algal toxins, domoic acid, and microcystin, and the fungal toxins,
afiatoxin Bl, and ochatoxin A.186 MIPs are used very successfully in "at home" glucose detection
devices for diabetics. Further refinement of MIPs for CBW detection is being pursued by multiple
laboratories (Mays, 2004; Pesavento et al., 2004).
Taste Chip (University of Texas)
August 2005
101
-------
U.S. EPA Office of Water Early Warning Systems
7.4.14 Magnetoelastic Sensors
Mass-sensitive magnetoelastic sensors can be used to detect antibody-
antigen interactions. However the mass change must be amplified by
biocatalytic precipitation. The sensor platform has immobilized
capture antibodies, which recognize target antigens, then alkaline
phosphatase-labeled antibodies are added to form an antigen-antibody
sandwich complex. The mass of the sandwich complex is amplified by
the precipitation of 5-bromo-4-chloro-3-indolyl phosphate disodium "'j^™^,'"oi'^t i"~v~
salt hydrate (BCIP). In response to an externally applied time-varying ma»netoelastic senior
magnetic field, steady-state or pulse, the ribbonlike magnetoelastic (Craig Grime?. PennState)
sensors mechanically vibrate at a characteristic resonance frequency.
These mechanical vibrations can be detected in several ways: optically from the amplitude
modulation of a reflected laser beam, acoustically using a microphone or hydrophone, and by using
a pickup coil to detect the magnetic flux emitted from the sensor.187 E. coli, enterotoxin, and ricin
have been detected with laboratory prototypes (Ruan et al., 2003, 2004a, 2004b). Although, this
laboratory prototype sensor will not likely be adapted for water in the near future, its very low cost
makes it an attractive technology. (Image reproduced with permission from Craig Grimes.)
7.5 Concentration Methods
Two AwwaRF research papers scheduled for publication in 2005 (Extraction Methods for
Early/Real-Time Warning Systems for Biological Agents - Project A and B; see Appendix D)
address methods for extracting biological contaminants from large sample volumes.
AwwaRF Project #2985 seeks to develop a large-volume extraction method for biological
contaminants that takes less than three hours and has at least a 60 percent to 70 percent recovery
efficiency. CDC is a project partner. The project builds on research being done by the DOD on the
Joint Service Agent Water Monitor, which serves all of the DOD services (e.g., Army, Navy, Air
Force, Marines). The objective of the program is to develop a water monitor that is portable,
preferably hand-held, near real-time, and capable of detecting all agents harmful to service members
in the field, while providing no false-positives. Because intentional contamination of water has
historically been a concern for the military, DOD has the most focused and developed information
concerning purposeful contamination of drinking water.
AwwaRF Project #2908 (see Appendix D) seeks to screen three to five different water extraction
methods for biological contaminant surrogates. DOD is a research partner. AwwaRF indicates that
release of the final report may involve a special protocol that requires signing a non-disclosure
agreement.
7.5.1 Hollow Fiber Ultrafiltration
A persistent challenge for many detection methods is the need to concentrate the contaminants
before identification and quantification. Hollow fiber ultrafiltration is a technology to
simultaneously concentrate viruses, bacteria, and protozoa from large volumes of water. The
ultrafiltration method can concentrate 100 liters of drinking water into 250 mL within 1 to 2 hours.
The water is circulated through a filter system to catch certain molecule sizes. The retentate can be
August 2005 102
-------
U.S. EPA Office of Water Early Warning Systems
further subdivided to permit detection of various microorganisms. This method may still have to
remove concentrated inhibitors that can interfere with certain assay tests like PCR. The method is
currently being developed by EPA, CDC, Army, and the Metropolitan Water District of Southern
California. A study of the hollow fiber techniques on source water from four water districts showed
a mean recovery of approximately 48 percent of Cryptosporidium oocysts. The results were
comparable to the Envirochek filter. The conclusion was that hollow-fiber ultrafiltration can
efficiently recover oocysts from a variety of surface waters (Kuhn and Oshima, 2002).
EPA-NHSRC and the Idaho National Laboratory (INL) have an interagency agreement to develop
and produce a next generation prototype of the Ultrafiltration Concentration (UC) device previously
developed by NHSRC and other stakeholders. The UC benchtop device concentrates microbial
pathogens within a 100 L municipal drinking water sample into a 250 mL volume in approximately
a 2-hour time frame (400-fold concentration). INL hopes to uses the benchtop UC system that has
been tested at the NHSRC in Cincinnati, OH, to redesign/repackage and automate the components
such that the new device can be operated in the field as a near-commercial, or field prototype system
(Vincente Gallardo, EPA, personal communication).
7.5.2 Hydroxyapatite Whole Cell Capture
Hydroxyapatite (HA) whole cell capture is a technology that could be applied to concentrate
microbiological contaminants in water supplies in order to detect them. For both pathogenic and
non-pathogenic organisms, the presence of anionic polymers on cell surfaces can be used to capture
both Gram-positive and Gram-negative eubacteria. Hydroxyapatite is a form of calcium phosphate
that can bind to bacterial cells with high affinity. It has been demonstrated by Berry and Siragusa
(1999) that positively charged HA particles can concentrate and purify bacteria from complex
matrices such as suspensions of ground beef and bovine feces. The bacteria would then be ready
for identification by PCR analysis. Because the cell capture is based on van der Waals and
electrostatic interactions between the bacteria cells and the HA particles, the affinity to the HA
particles depends on the specific cell type. When tested, the efficiency of capture varied from
46 percent for E. coli to 99 percent for Yersinia enterocolitica (Mays, 2004).
7.5.5 Lectin and Carbohydrate Affinity
Another approach that could be used to capture microbial cells is to use lectins that target the
carbohydrate rich cell envelope polymers of microbes. The carbohydrate pieces are usually
fundamental structural elements of cell walls or proteins so are less likely to vary than protein
sequences that can mutate or are dependent on environmental conditions. Therefore, the use of
lectins is an attractive candidate for concentration methods. Lectin-based capture of several
eubacteria including E. coli and Salmonella spp. has been demonstrated. A microbial capture
approach analagous to using lectin affinity is to use the carbohydrate binding properties of the
microorganisms themselves. It is normally critical for pathogenic bacteria to adhere to the lining
of the gut in order to colonize. Well studied examples of carbohydrate binding in bacteria are
adhesions used by E. coli and S. flexneri. Also, necessary host cell recognition through
carbohydrates has been studied for certain viruses, such as rotavirus. It can be speculated that a
collection of lectins and/or carbohydrates could be chosen to bind organisms in a semiselective
manner to concentrate and purify them for detection.
August 2005 103
-------
U.S. EPA Office of Water Early Warning Systems
Anticipated challenges to the use of hydroxyapatite and lectin/carbohydrate affinity include the
difficulty of immobilization of HA and lectins onto magnetic or polystyrene beads. Also, tests
would need to be performed to determine capture efficiency of the microbial contaminants of
interest, and tests would be needed to determine if these methods would co-concentrate inhibitors,
causing them to be of little use for detection (Mays, 2004).
August 2005 104
-------
U.S. EPA Office of Water Early Warning Systems
8. Technologies that Detect Radiological Contaminants for Early Warning
Systems
Radiation is a possible contaminant in water and is associated with carcinogenic and non-
carcinogenic adverse health effects. The Federal Water Pollution Control Act (Clean Water Act),
the Safe Drinking Water Act (SDWA), and the Maximum Contaminant Levels (MCLs) address
protection of water systems from radiation and other contaminants. Radiation MCLs are measured
at entry points, do not require routine monitoring, and currently exist for beta/photon emitters
(includes gamma radiation), alpha particles, combined radium 226/228, and uranium. These
regulations had been thought to be adequate in ensuring a long-term distribution of clean and safe
water. However, now that terrorism is a major security concern in the U.S., the importance of the
water sector's preparedness for potential attacks and accidents has become increasingly important.
The Homeland Security Presidential Directives (HSPDs) and the Public Health Security and
Bioterrorism Preparedness and Response Act of 2002 (Bioterrorism Act) have obligated EPA to
focus on the water sectors' strategies for emergency prevention and response.
In the case of intentional contamination, realtime monitoring of radiation is important for immediate
detection and response. Currently there is radiation measurement equipment that detects the total
amount of radiation, as well as equipment that detects specific types of radiation by the energy levels
emitted from a given source. SSS-33-5FT Drinking Water Rad-safety Monitor by Technical
Associates and 3710 RLS Sampler by Teledyne Isco, Inc. (Los Angeles, CA)188 (both mentioned
below) are examples of devices that analyze the aggregate radiation of alpha, beta, and gamma rays.
These instruments alert operators when the water has been radiated, but do not identify the specific
contaminant. Other instruments identify alpha, beta, and gamma radiations separately, and will be
discussed further in this section. The general information as well as many of the technology costs
presented in this section are available on EPA's website,189 Water and Wastewater Security Product
Guide, which addresses radiation detection equipment for monitoring water assets. This website is
based on information culled from the Multi Agency Radiation Survey and Site Investigation Manual
(EPA, 2000), developed by EPA, DOE, DOD, and the U.S. Nuclear Regulatory Commission.
EPA does not endorse or recommend any of the following technologies. The summary
information below was obtained from company websites and promotional literature.
8.1 General Introduction to Methods of Detection
Gamma radiation emissions are long-range electromagnetic waves that penetrate many objects and
can be measured in the field with a sodium iodide (Nal) scintillation survey meter. On the other
hand, it is difficult to have in-field rapid detection technology for alpha and beta radiation in water
due to their physical properties. Alpha radiation emissions are positively charged particles that
cannot penetrate through objects while beta radiation emission are negatively charged particles that
have a moderate capacity to penetrate objects. Difficulties arise in measuring alpha and beta
emissions in water because these short range types of radiation are easily blocked (attenuated) by
water before they reach the detector.
Therefore, the instruments need to be placed close to the source without blocking the path of the
radiation to the detector. Furthermore, gas-flow proportional counters typically evaluate alpha and
beta radiation from smooth, solid surfaces. However, since water surfaces are not smooth, a large,
August 2005 105
-------
U.S. EPA Office of Water
Early Warning Systems
sensitive liquid scintillation counter in a laboratory is frequently required; thus, in-field
quantification of alpha and beta radiation in water is a rare practice. This chapter, however,
introduces certain devices that can detect and quantify radiation onsite.
The instruments and methods can be evaluated in terms of specificity and sensitivity. EPA defines
specificity as the ability of an instrument to quantify or evaluate the specific type of radiation or
radionuclide for which it is designed without false-positives, (e.g., without interference from other
radiation or radionuclides). Sensitivity is defined as the radiation level or quantity of radioactive
material that can be measured or detected with some estimated level of confidence, and is a function
of the instrumentation and the technique used. With regard to the specificity in measuring alpha and
beta radiations, liquid scintillation counters are generally extremely flexible and accurate when
properly calibrated and quenching effects are compensated (the full energy pulse may not reach the
photo-multiplier detector). The complex multi-energy spectra of beta radiation can be quantified
because its energy spectra are 10 to 100 times broader than the gamma spectra. With regard to
sensitivity, this scintillation survey meter is ideal for moderate to high energy beta and alpha
emitters because different radiation types can be easily distinguished by their pulse shape.
With regard to the specificity in measuring gamma radiation, some of these scintillation survey
meters make the preliminary identification of specific isotopes possible with their ability to analyze
selected ranges of gamma energies. The minimum sensitivity of these scintillation meters is 200 to
1000 counts per minute (cpm), and can be lower when switched to digital integrate mode. The cost
of the sodium iodide scintillation survey meter is approximately $2,000. Typically these in-line
gamma radiation detectors are used only at special facilities that handle radioactive materials.
Continuous online monitoring systems can monitor the water in real-time however, there are few
such units commercially available. These systems can be installed along with an alarm system that
would alert operators of unusual radiation measurements. There are units for measuring radiation
in wastewater but these would have to be adapted for drinking water. Grab sample units for drinking
water are much more common. Very few water districts have real-time radiation monitors in place
to protect water and the public.
8.2 Available Technology
The SSS-33-5FT190 by Technical Associates (Los Angeles, CA) is
a real-time, in-line, continuous flow-through scintillation detector of
alpha, beta, and gamma radiation in ground, surface, or waste water.
The detector can be applied to measure one type of radiation or all
radiations combined. This easily calibrated apparatus uses ion
exchange resin beads and charcoal filters, and does not require liquid
scintillant. The ion exchange resin collects ions from dissolved
metals, which are then measured for activity by gamma spectrum
detectors. The charcoal filter collects non-ionized stray radioactives.
Crushed anthracene scintillation crystals are the final detector of the
radiation. The instrument measures tritium up to 100 picoCurie/mL
and is equipped with a system that sends an alert if unusual readings
are made. All the data can be retrieved in a spreadsheet format.
^ SSS-J3-5FT
(Technical Associates)
August 2005
106
-------
U.S. EPA Office of Water
Early Warning Systems
This instrument is on the market for approximately $58,000. (Image reproduced with permission
from Technical Associates.)
MEDA-5T191 by Technical Associates is a continuous monitor of
intentional contamination or accidental spills of gamma radiation
into the water source. The instrument is equipped with pumps and
a scintillation detector. An automatic quick alarm will sound in the
event of a radioactive water contamination. The monitor is
available for approximately $25,000. (Image reproduced with
permission from Technical Associates.)
MEDA-5T
(Technical Associates)
The 3710 RLS Sampler192 by
Teledyne Isco, Inc. (Los Angeles,
CA), detects radionuclides using
3M Empore™ Rad Disks and a
known amount of flow-through. The sampler continuously
monitors in water for all types of radiation. (Image reproduced
with permission from Teledyne Isco Inc.)
The down-hole tritium in water detectors, SSS-33DHC and SSS-
33DHC-4193 by Technical Associates, are used to continuously
monitor and detect underground plume or tritium leakage. These
detectors fit in bore holes, are not influenced by other nuclides, and
require no liquid scintillant. The monitors are sensitive below
EPA clean drinking water levels, have a sensitivity of 1
nanoCurie/ml in 100 seconds, and their lower limit of detection is
better than the FDA drinking water standard of 20,000 pCi/L
average over 24 hours. These monitors cost $72,000.
3 711) RLS Sampler
(Teledyne. I^eo, Inc.)
The SSS-33M8 monitor194 by Technical Associates is a real-time continuous monitor for tritium in
water. No liquid scintillant is required and it is sensitive to 0.1 nanoCurie/mL without being
influenced by other nuclides. It is useful in monitoring leaks in type reactors, tritium in
groundwater, and laboratory or plant liquid waste streams. This monitor costs $16,500.
8.3 Potentially Adaptable Technology
Canberra (Meriden, CT) has developed several devices that detect radiation in liquid pipes such as
those carrying waste streams, but these monitors are not intended for drinking water distribution
systems. All of Canberra's devices can monitor liquid streams in realtime using the Radiological
Assessment Display and Control Software (RADACS), which allows online access to the monitors
from remote locations.
The LEMS600 Series Liquid Effluent Monitoring System (LEMS)195 by Canberra has the capacity
to continuously evaluate the gross gamma/beta radiation. The series consists of LEMS614,
LEMS615, and LEMS616. The detectors are equipped with alarms in case of high radiation or
failure conditions. The LEMS614 detects the combination of beta and gamma radiation, while
LEMS615 detects gamma radiation in liquid samples between 0 to 50° C. A similar gamma
August 2005
107
-------
U.S. EPA Office of Water Early Warning Systems
detector, LEMS616, incorporates a cooling system for liquids with higher temperatures. The cost
of LEMS technologies is between $100,000 and $150,000.
The OLM100 Online Liquid Monitoring System196 by Canberra continuously monitors liquid and
gas gamma radioactivity in liquid and gaseous streams. It is available both as a clamp on model as
well as a clam shell model so as to fit various pipe sizes. The device uses a gain-stabilized
scintillation detector and has earned Class IE Safety Qualifications. The instrument in an online
monitor that is attached to the exterior of pipes so as to not disrupt the flow rate within. Its
sensitivity/detection limit depends on a preprogrammed lower limit of detection and the normal
background. The cost of OLM-100 is between $35,000 and $70,000.
The ILM-1OO197 by Canberra is similar to the OLM-100 but is installed within the pipe system. The
ILM-100 and OLM-100 cost between $35,000 and $70,000. The OLM system is typically cheaper
than the ILM system because it can be clamped onto an existing pipe, whereas the ILM system must
be fitted into the pipe. Both systems can be fitted into pipes of Vz inch to 16 inches, but the cost
increases with the pipe size because of the additional expense of ensuring that the detector is
properly installed in the pipes.
8.4 Emerging Technology
Clarion Sensing Systems, Inc. (Indianapolis, IN)198 has developed an in-pipe radiation detector that
the company plans to launch on the market in late 2005 (Martin Harmless, Clarion, personal
communication). The Gamma Shark™ sensor detects gamma radiation in water above background
levels. The device is able to expose more surface area for the radiation to contact by inserting the
scintillator tube in the water stream. The monitor logs fissions detected and converts the counts per
minute to normal units. The Gamma Shark™ compares the radiation levels in the water to
background and detects increases in radiation in the water. The company is in the process of third
party verification and the instruments are expected to be cost-effective compared to currently
available technology. Clarion's radiation detectors will stand alone or interface with the company's
Sentinal™ unit (see Chapter 4) to provide website display results as they are obtained.
According to a DOE publication in 2000,199 Thermo Power Corporation (Waltham, MA) is
developing the Thermo Alpha Monitor (TAM)200 under the sponsorship of DOE. This instrument
is a near real-time monitor of alpha radiation, which was estimated to have a cycle time of
approximately 30 minutes for 1 ppb Uranium, and approximately 5 minutes for 10 ppm Uranium.
The concept was developed around the simultaneous in situ collection and quantification of
radioisotopes on a silicon detector using a solid-state semiconductor counter. The detector is similar
to those that use ionization chambers, but measures the lost energy caused by ionized radiation as
an electrical current.
The PNNL (Richland, WA) is in the process of developing a in-field radionuclide sensor to detect
technitium-99 (Tc-99) in groundwater.201 The technology will incorporate the use of chemically
selective beads to preconcentrate Tc-99 in the sensor in order to increase the sensitivity and
selectivity of the direct measurements. The laboratory strives to create a device that can demonstrate
reversible operation and the required sensitivity as well as the capability to operate the embedded
controls from remote locations.
August 2005 108
-------
U.S. EPA Office of Water Early Warning Systems
In addition, an online real-time radiation detection instrument will be available for alpha radiation
in the near future. DOE tested this prototype, and further technological development has been
conducted under the Los Alamos National Laboratory since 2001,202 The detector is expected to be
used in monitoring radioactive liquid waste and groundwater using Long Range Alpha Detection
(LRAD) technology. The monitor is in real-time and is non-intrusive.203
August 2005 109
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
9. Technical Evaluation of Early Warning Systems
This chapter provides a technical evaluation of the various components of an Early Warning System
(EWS) for drinking water, as characterized in Chapters 3 and 4. Such an evaluation is important to
allow utilities and drinking water quality officials to identify technologies appropriate to particular
situations and systems. Researchers and utilities need a better sense of what systems are becoming
fully developed and where there are research gaps. False-positives and false-negatives have
occurred with various devices and have caused great concern among first responders, emergency
agencies, and health and law enforcement officials. There are claims of performance that would
benefit from additional testing and evaluation. In addition, technology is rapidly evolving and at
various stages of development to detect chemical and biological contaminants. This report is
divided by the level of development of early warning. The level of development is set by three
categories, (1) available, (2) potentially adaptable, and (3) emerging. In Chapter 3, the desired
characteristics of an EWS (e.g., detects range of contaminants, sensitive) were presented. In this
chapter, an analysis is conducted on how close the existing EWS technologies meet these desired
characteristics.
9.1 Approach for Technical Evaluation
A scientific and technical evaluation of EWS technologies is based on expert review of qualitative,
semi-qualitative, and quantitative information sources as described below. It is important to note
that the evaluation did not involve actually testing equipment or assays. The information sources
include the following:
• verification studies
• degree of government involvement, support, and development for technology
• field experience and case studies
• other studies
• expert opinions
The sources are described in detail below.
9.1.1 Verification Studies
During the anthrax attacks of 2001, when hand-held assays proved unreliable, it became particularly
clear that the government has a responsibility for validating the performance of CBR detection
equipment (Emanuel et al., 2003). Verification, feasibility, and proof-of-concept studies of CBR
detectors, with a special focus for use in water, are underway at several government and private
facilities. Examples include the U.S. Army Edgewood Chemical Biological Center facility, the
DOD Chemical and Biological Defense Program Test and Evaluation Executive Agent Facility for
Water Monitor Test Methodology and Instrumentation Development, the EPA WATERS Facility,
and various contractor facilities. Specific efforts at evaluation include the following:
• EPA's Environmental Technology Verification (ETV) program evaluates various
technologies, including sensors for chemical, microbial, and radiological contaminants.
• EPA's Technology Testing and Evaluation Program (TTEP) tests the performance of
technologies related to homeland security applications.
August 2005 111
-------
U.S. EPA Office of Water Early Warning Systems
• The National Technology Alliance, through the Chemical, Biological, and Radiological
Technology Alliance, examined and reported on state-of-the-art and emerging technologies
in water monitoring (Black & Veatch, 2004).
• Some local water utilities, including the Pittsburgh Water and Sewer Authority, have
conducted verification tests (States, 2004).
• AwwaRF has an extensive number of evaluation projects that deal with EWSs (see
Appendix D). However, most are on going, and were not available for this evaluation.
The challenge remains that only a limited number of facilities can test this equipment with real
chemical or biological agents.
9.1.2 Degree of Government Involvement, Support, and Development for Technology
Government and industry have sponsored research into the development of various water monitoring
technologies, as well as the verification of their performance. Such research could be an indication
of the potential development of the technologies. Sponsoring agencies include DHS, EPA, U.S.
Army (ECBC), FDA, and CDC. For example, ECBC has an active program, including projects such
as "Development of Novel DNA Probes for Emerging BW Agents," "PCR Assay Optimization for
BW Detection," Validation of BW Detectors," and "Development of Enzyme-based CW Sensors."
FDA is investigating several technologies for identification of microbial contaminants, most of
which are food-related, but some have applicability to the water sector. For example, in September
2003, FDA awarded five research grants on;204 (1) development of the Waveguide Immunoassay for
Yersinia pestis; (2) a rapid immunoassay silver amplification test system; (3) uses of a novel,
compact microchip sensing system for rapid food screening; (4) development of PCR-based assays
on a microchip, and (5) use of thin layer chromatography and bioluminescence.
9.1.3 Field Experience and Case Studies
Some of the technologies have been widely used in source water or in the food industry that may
enable the technology to be applicable to finished drinking water. Some utilities are using these
technologies for grab samples from treated water, and a few utilities have online systems. The
limited field experiences and case studies that were examined during the development of this report
helped to provide insight into the technologies' current uses. It is likely that further examination of
more detailed (and confidential) case studies would yield a wealth of valuable information for the
advancement of the EWS field.
9.1.4 Other Studies
There are several studies that provide evaluation information. These include the ASCE White
Papers andlnterim Guidelines for Designing Online Contaminant Monitoring System (ASCE, 2004),
the CBRTA report Water Monitoring Equipment for Toxic Contaminants Technology Assessment
(Black and Veatch, 2004) and various AwwaRF research studies (see Appendix D; Roberson and
Morley, 2005).
August 2005 112
-------
U.S. EPA Office of Water Early Warning Systems
9.1.5 Expert Opinions
Experts were contacted in various agencies and organizations, including DHS, USGS, EPA, DOD,
various national labs, water associations, utilities, and consultants contracted for this project.
Using the above sources, a technical evaluation was conducted on various operational features of
anEWS (e.g., data management, acquisition, security), on multi-parameter water quality monitors,
on chemical sensors, on microbial sensors, and on radiological sensors.
9.2 Evaluation of Various Operational Features of EWS
This section addresses the features of the EWS that are not related to sensors. These EWS features
include real-time data acquisition and analysis, contaminant flow predictive systems, sensor
placement, alert management, security enforcement, and communications, response, and decision
making.
9.2.1 Issues and Gaps
• Real-time Data Acquisition and Analysis
The SCADA interface is key to managing the data and identifying a contamination event. Most
utilities are familiar with SCADA systems. Remote data acquisition systems are commercially
available and used in many utilities for basic water quality control and monitoring. Adapting the
SCADA to track sensor data is not a large challenge. However, challenges include the ability to
handle the data load as well as to interpret the data that are collected. Analyzing large streams of
data requires special training. Various software exists, but many algorithms for data analysis are
not verified or demonstrated. Standardized methods for data analysis do not exist and may need to
be developed for such EWSs. Documentation of case studies could also further guide the proper use
of such data analysis techniques.
• Contaminant Flow Predictive Modeling Systems
There are some basic contaminant flow predictive systems, but their use by utilities for the purpose
of modeling the movement of contaminants is not widespread. Many utilities do use models for
tracking chlorine residuals and disinfection byproducts. However, as models undergo further
development by researchers, it is necessary to calibrate and verify the models and make the models
useful tools for sensor placement, for real-time contaminant flow prediction, and for identifying the
location of the contaminant source. Utilities would need to expand their current use of models to
include modeling for intentional contamination events. Also, utilities would either need to train
personnel to operate the expanded models, or hire contractors to run the predictive flow models.
• Sensor Placement
Sensor placement has implications for cost. Placement is often determined by logistic factors such
as location security and convenience, and access to a power source or data transmission rather than
risk minimization. Although current research on combining flow models and sensor technology is
a move in the right direction, such models must be verified before difficult and costly decisions are
August 2005 113
-------
U.S. EPA Office of Water Early Warning Systems
made by the utilities. Simple guidance is needed for situations where the number of sensors is
limited.
• Alert Management
Managing the alert process will help determine when certain proper response actions are needed.
Proper alert levels will also minimize false-positives/false-negatives. It is not possible to eliminate
all false-positives and false-negatives at the same time. Therefore, it is preferable to optimize the
system to eliminate false-negatives and manage the inevitable false-positives in a way that
minimizes undesirable impacts on the utility and community. Currently, only some utilities have
experience with this as a part of their normal operations. The use of water quality parameters as a
first stage alert is a particular challenge for EWSs for terrorist attacks. Detailed baseline water
quality data are needed to set alerts with reasonable confidence that they do not result in too many
false-positives or false-negatives. More demonstration projects are needed to assure the
reasonability of certain approaches to alert management and further guide utilities to the proper use
of such alert management.
• Data Security
Current remote monitoring products are beginning to incorporate security precautions, including
encryption. However, demonstration projects are not frequent. Standardization from other data
security efforts could be applied to the water sector. Programs should be developed to link utility
security efforts surrounding SCADA generally with data security geared to EWSs.
• Communications, Response, and Decision Making
The process linking the analysis of contamination data with the decision making and response is
outlined in EPA's Response Protocol Toolbox. While this guidance provides a process, the
notification and communication equipment to effectively implement the process has not been
extensively developed for water utilities. Tools(e.g., Water Contaminant Information Tool) to assist
decision making and response are being developed.
9.2.2 Conclusions and Recommendations
Much of the data acquisition software and hardware already exist. The data acquisition systems do
not represent a major issue given currently recommended sampling times by EPA. Security of
SCADA systems for EWSs is a challenge but can probably be handled as the general security of
SCADA systems is addressed by utilities. Recommendations on the topics in the section include
the following:
• Standardized methods and guidance for data analysis and interpretation are needed. Some
of the efforts by ASCE will help to guide utilities in the use of such systems.
• Projects are needed to verify contaminant flow models and then adapt the models relatively
easily for use by various sized utilities.
• Simple guidance is needed, such as what to do if only a small number of sensors are
available.
August 2005 114
-------
U.S. EPA Office of Water Early Warning Systems
• A demonstration project is needed to assure the reasonableness of certain approaches to alert
management. The USGS project is an example where multi-parameters sensors for water
quality are being used in conjunction with alert management. Additional projects should
examine alerts for other promising sensors such as mussel or bacteria monitors.
• Effective technologies should be developed to facilitate the rapid and efficient alert of
decision makers and response personnel with response information.
9.3 Evaluation of Multi-Parameter Water Quality Monitors
Traditional drinking water quality monitors have been bundled together and sold commercially.
They can now be monitored remotely, continuously, and in real-time. Several vendors have modular
systems, so that utility companies can choose which parameters they want to measure. These multi-
parameter monitors have proven valuable for maintaining daily water quality, and more recently
have been evaluated as a first-tier warning of the presence of an intentional contaminant. However,
combining monitors and sensors from different manufacturers is still problematic because these is
limited uniformity and interchangeability in hardware, signal generation and processing, and
connectivity. Several cities have used multi-parameter probes in their distribution systems both for
ensuring general water quality and for water security. Usually the probes were placed at convenient
sites with ready access.
An important part of developing a first stage EWS is to evaluate whether the normal operation of
water distribution systems can be documented in terms of sensor responses. One purpose of the test
at the WATERS Center was to determine if sensors can identify a baseline water quality or if sensor
drifting takes place. Thus, the basic performance characteristics of individual commercially
available sensors (in terms of use for early warning in water) were evaluated. The parameters
measured include pH, DO, turbidity, free chlorine, conductivity, ORP, TOC, and ion-selective
electrodes (Cl~, NO3 and NH4+). A basic conclusion was that specific-conductance, TOC, and free
chlorine monitors drift very little when properly calibrated and serviced; therefore, these sensors are
ideal for characterizing normal or safe conditions. However, it was also found that free chlorine
interferes with some of the above water parameters. Additionally, the test examined if the sensors
can qualitatively detect the contaminants. The contaminants injected into the system included
wastewater, potassium ferricyanide, malathion, and glyphosate. The test concluded that the sensors
can only provide a general indication of the contaminant class such as inorganic, organic, or a
reactive species producing chlorine demand (EPA, 2004).
The USGS and EPA have an Interagency study to implement and test an EWS at actual field sites.
The investigation team, composed of scientists and researchers from EPA, the USGS New Jersey
District, and a water utility are involved in the project whose purpose is to identify potential EWS
field sites, select sensors based on the previous EPA and USGS efforts, evaluate the distribution
system hydraulics and water quality, determine sensor locations, and collect sensor data. However,
no contaminants will be inj ected into the distribution system. The data will help determine how well
the sensors work, optimize the sensor location, and develop a baseline water quality profile of the
distribution system.
Another study from EPA's WATERS test facility investigated the response of a combination of off-
the-shelf sensors to detect changes caused by the injection of wastewater, groundwater, a chemical
cocktail, and individual chemicals, such as potassium ferricyanide, malathion, and glyphosate. The
August 2005 115
-------
U.S. EPA Office of Water Early Warning Systems
feasibility of using the sensor information as an early warning of an attack was investigated. Initial
results show that water quality sensors responded to the injection of test materials. The parameters
monitored by the sensors include chloride, free chlorine, ORP, specific conductance, TOC, and
turbidity. These parameters exhibited a unique and consistent pattern of signal changes upon
injection of wastewater, potassium ferricyanide, malathion, glyphosate, and groundwater into a
drinking water distribution system simulator. The sensor system showed promise for providing
quick detection of water quality changes due to introduction of these test materials in the pipe loops.
Because of the unique physicochemical properties of these test materials, a specific sensor response
pattern was observed for each substance. This suggests that the sensor system may provide
information on the characteristics of unknown contaminants and facilitate the subsequent
identification with more sophisticated instruments. With further optimization, a system of sensors
may be used as an EWS (EPA, 2004). However, because the range of test materials was narrow,
additional types of contamination and scenarios should be examined to test this conclusion further.
Currently, EPA has several planned efforts to examine the use, development, and testing of multi-
parameter probes. EPA has an interagency agreement with the USGS and initiated CRADAs with
Hach Company, PureSense Environmental, Inc., and YSI, Inc. The goal of the partnerships is to
develop technologies for both detection devices (e.g., multi-parameter probes) and data analysis
software to facilitate the installation of early warning systems in the distribution networks of local
water utilities. The types of multi-parameter probes include pH, ORP, specific conductance, residual
chlorine, and temperature.
Another series of tests currently being conducted at the EPA WATERS facility is verifying the
capability of multi-parameter water monitors for distribution systems. The testing is being
performed under the auspices of the EPA-ETV Program. The multi-parameter monitors for
distribution systems being used during these verification tests consist of instrument packages that
can be connected to or inserted into distribution system pipes for continuous monitoring. Also
included in this category are technologies that can be programmed to automatically sample and
analyze distribution system water at regular intervals, as well as hand-held technologies requiring
technicians to manually collect samples and perform the analysis. The monitors must be able to
measure free chlorine as well as at least one other water quality parameter (e.g., alkalinity, pH, DO,
ORP, temperature, turbidity, conductivity, ammonia, calcium, total carbon, chloramines).
The multi-parameter water monitors are being evaluated for the following parameters:
• Accuracy: Comparison to results from standard laboratory reference analyses.
• Response to individually injected contaminants: detection of changes in pipe loop water
chemistry. (Contaminants tested will include aldicarb, arsenic trioxide, and either E. coli
bacteria or nicotine.)
• Inter-unit reproducibility: comparison of results from two simultaneously operating
monitors.
• Ease of use: general operation, data acquisition, set up, demobilization, required
maintenance.
• Presence and identification of injected contaminants (if applicable): A total of 17 different
contaminants will be tested for identification by the appropriate monitoring system.
August 2005 116
-------
U.S. EPA Office of Water Early Warning Systems
The monitors being verified during this series of tests include the following:
• Clarion Systems Sentinal™
• Emerson Model 1055 Solu Comp II Analyzer
• Man-Tech TitraSip SA (multi-parameter but not online)205
• Hach Event Monitor206
• Analytical Technology, Inc., Series C15 Water Quality Monitoring Module
Several manufacturers are exploring the signature monitoring approach. Using data from multi-
parameter continuous monitoring devices, one company tested a number of potential water
contaminants in an effort to establish signatures that could provide detection and a tentative
identification. A list of 60 contaminants (chemicals, toxins, biologicals) were analyzed for possible
detection. The sensors included pH, chlorine, conductivity, turbidity, and TOC. The sensors
showed response to the various contaminants, however, a few contaminants did not trigger a sensor
response at concentrations that are considered harmful. The manufacturer has developed a trigger
algorithm to set off an alarm when conditions in water depart from expected baseline parameter
values. Because utilities regularly experience changes in water quality, there are always concerns
in this type of system for false-positives (King, 2004).
It was announced in October 2004 that the Army and Hach Company have signed a CRADA to
complete testing of the Army's new real-time water security detection and response technology by
the end of 2004. Under the agreement, ECBC, the Army Corps of Engineers, and Hach will conduct
live-agent testing using Hach monitoring equipment (GLI International panel, Cl-17, turbidmeter,
pH, and specific conductance, as well as Hach's TOC monitor) to detect terrorist attacks on drinking
water distribution systems. The ECBC is one of the few sites in the U.S. where testing on actual
chemical and biological contaminants can take place. Pending verification tests, the technology is
scheduled for commercial production in 2005.
Exhibit 9-1 provides an evaluation of specific water quality parameter probes in the distribution
system. Exhibit 9-2 summarizes how the probes (both currently available or potentially adaptable)
compare with the desired characteristics of EWSs (as described in Chapter 3).
9.3.1 Issue and Gaps
The following discussion highlights various issues and gaps in using multi-parameter water quality
monitors in an EWS.
• Baseline Data are Needed
Although research has demonstrated proof of concept for using water quality parameter fluctuations
as a signal that a contamination event has taken place, baseline data to calibrate the alarm triggers
may need to be gathered over months or years for each independent system. This will likely prove
to be sufficiently expensive that budgetary impacts would need to be addressed. Daily, seasonal,
and event- (i.e., storm-) related fluctuations will need to be identified and characterized, so they are
not confused with contamination events. Thus, systems with highly fluctuating source waters may
have considerable noise in the baseline.
August 2005 111
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 9-1. Evaluation of Water Quality Parameter Monitors
Technology
Manufacturer
Evaluation
Single Parameter Sensors
Model A 15/B-2-1
- Free Chlorine
Hach TOC process
analyzer
International
Model 5500
Dissolved Oxygen
Ion Selective
Electrodes
Specific-
Conductance
Dissolved Oxygen
Oxidation-
Reduction
pH
Temperature
sensor
Turbidity
Monitors
Free Chlorine
Total Organic
Carbon
Analytical
Technology, Inc.
Hach
GLI
Various vendors
Various vendors
Various vendors
Various vendors
Various vendors
Various vendors
Various vendors
Various vendors
One vendor
This is an online sensor that uses a polargraphic method to
monitor free chlorine. It showed less baseline stability and
sensitivity than Hach's AquaTrend Panel, which also
monitors chlorine.
This is an online sensor that uses UV persulfate oxidation
to monitor total organic carbon. A good baseline and high
sensitivity was maintained when tested with all
contaminants. Cost is $30,000, higher than most of the
other devices tested in the study.
This is an online sensor that uses a membrane electrode to
monitor dissolved oxygen concentration.
No difference in chloride, nitrate, and ammonium analyzers.
Nitrate electrode not properly calibrated after exposure to
Cl. Chloride and ammonium sensor fail at a rate of 3-6
months. Recommend using 3 point calibration.
No difference with vendors. Easy calibration, sensors last
years, care during cleaning of annular-ring carbon
electrodes
Low flow technology no advantage over larger cell. DO
sensor based on plan sensor technology gave positive
deviation with sudden changes in Cl. Had failure rates for
chip higher than manufacturer anticipated.
Similar performance of vendors, ORP robust, other sensors
(pH/electrodes) combined with ORP will fail before ORP
Failure rates of 6 months in chlorinated water (planar sensor
failed at 1-2 months, and required at least weekly
calibration).
Rarely fail.
Very stable when properly cleaned (cleaning procedure is
more difficult but only required quarterly. Calibration
requires practices. Some false-positive readings when
vibrated or change pressure.
Colorimetric reliable when serviced properly. However,
sampling interval is 3 minutes. Polarographic technique
very reliable but must change membranes every 2 months
and need rigorous cleaning. Planar measurement failed
most frequently.
Very stable but need experience to operate.
August 2005
118
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 9-1. Evaluation of Water Quality Parameter Monitors
(continued)
Technology
Manufacturer
Evaluation
Multi-Parameter Sensors
Six-Cense™
continuous
monitor
AquaTrend panel
DataSonde 4a
In-Situ Model
Troll 9000
Signet Model 8710
YSI Model 6000
continuous
monitor
Zero Angle
Photon
Spectrometer
MP-1
STIP-scan
Dascore
Hach
Hydrolab
In-Situ
Signet
YSI
Oregon State
University
STIP Isco GmbH
(Germany)
This online sensor monitors dissolved oxygen, free
chlorine, ORP, pH, specific conductance, and temperature.
It showed an unstable baseline when monitoring ORP,
specific conductance, and free chlorine.
This online sensor monitors free chlorine, pH, specific
conductance, temperature, and turbidity. For free chlorine
detection, it showed the highest sensitivity and most stable
baseline compared to ATI's and Dascore's free chorine
sensors. It also had the most stable baseline for turbidity
monitoring. Overall, the AquaTrend showed consistent
responses, stable baselines, and high sensitivity in almost all
parameters measured in the study.
This online sensor monitors ammonia nitrogen, chloride,
dissolved oxygen, nitrate nitrogen, ORP, pH, specific
conductance, temperature, and turbidity. It performed well
compared to other sensors in measuring ORP.
This online sensor monitors dissolved oxygen, ORP, pH,
specific conductance, temperature, and turbidity. In
measuring ORP, it was noted that the In-Situ model had a
higher rate of failure than other sensors.
This online sensor monitors ORP and pH.
This online sensor monitors ammonia nitrogen, chloride,
dissolved oxygen, nitrate nitrogen, ORP, pH, specific
conductance, temperature, and turbidity. The YSI model
showed a more stable baseline for ammonia nitrogen and
nitrate nitrogen monitoring and a higher sensitivity for
chloride monitoring. Overall, the YSI model showed
consistent responses, stable baselines, and high sensitivity
in almost all parameters measured in the study.
This online sensor uses optical measurement to monitor
bacterial fluorescence, humic fluorescence, nitrate nitrogen,
total fluorescence, transmission, and 245 nm UV
absorbance.
UV/V is a spectroscopic sensor that is capable of
simultaneous measurement of nitrate, COD, TOC, spectral
absorption coefficient (SAC254), total solids, sludge
volume, sludge volume index, and turbidity
Sources: EPA, 2004a, 2004b.
August 2005
119
-------
bo
c
IK
§
I
x.
^
o
GC
I
•d
x.
e
o
5B
a
S
o
O
5B
o
s
O*
o\
+j
z
fl
.S
1
"3
H
a 3
- s
s a
U a
'[« aj
^
IS
C/3
L.
O
ta
o>
O.
O
ta
O
U
~OJ .—
.S ^
1 ^
,P
(._ a
= a
o> .5
CM g
g i
OS g
u
Description
Product
c/3
W
H
^
^
OH
W
tu
cd
•a
§
.a
C/J
W
H
^
^
OH
W
tu
IS
"a
§
.a
.a
^
rA
g
•3
S
0
o
iy^°
(N
6«
.a
'g
0
8
H
,_
le-parameter wa
ity monitoring
.al
™ *
Various Total Organic
Carbon monitors
c/3
tt!
W
H
^
^
OH
tu
cd
'a
&
1
^
rA
•s
cd
s
•g
^
.-
^
0
o
0_
^'
(ft
s
'g
0
'x
o ^
-^ ^
'-j Q
60 0
S 2 'S1
le-parameter wa
ity monitoring (1
gen concentratio
s S &>
r! r
Hach TOC Process
Analyzer
C/3
W
H
^
^
OH
W
1>
IS
"a
§
.a
IT)
V
•s
s
•s
:$
.-
.2
0
o
r^'
, |
6ft
^g
"3
0
'S
O <£
-^ ^
'—I (J
^H rH
cu H
a 53
60 0
53
ti-parameter wat
ity monitoring
1 g
g cr
o
Hach Water Distributi
Monitoring Panel
c/3
W
H
^
^
OH
W
0 0 g
a * f
tu
^
0
"><
O cc
"*""* 13
Id u
53 'S
a 53
tu ^
60 0
ti-parameter wat
ity monitoring
r- r^
n-i
Emerson Model 1055
Solu Comp II Analyzt
-------
bo
S
OB
• *4
•^
OB
O
GC
I
•d
._
•,—
•»^
._ v
5« S
& a
OB
O
•I—
'ia
o
S
-^
"es
S
o*
o\
+j
z
a
.s
1
"3
,*
a 1
S a
- s
S 'a
U a
.2
p^ S
a 3
e H
3
C/3
L.
O
ta
o.
O
o
U
.S *
"3 •£
O o
+•.*
Is
M §
§ «
OS |
U
=
o
'•S
&
*c
u
o
Product
rH
O
a
a>
ta
•a
.a
g
in
V
•s
13
a
t^
a
|
a>
IS
ta
> •'S
^ g _o
!Ǥ i*
a
^
0
^•i
"s •"
a> ^
^
| 60
$_, -9
^ O
i!
•73 ~a
a cr
60 U
•§£ g>
Analytical Techn
Inc., Series CIS A
Quality Monitoril
c/3
W
H
j>
^~
<£
OH
W
1>
IS
13
.a
a
^
60 U
VH 8
13 - — -
B '§
u S
!!i
60"3 J3
•to Er-3
00 i
O c~)
Analytical Techn
Inc Model A 15/1
aj
o
a
a>
ta
•a
.a
in
V
•s
13
a
t^
a
g
-2
a>
IS
_c«
^ «
B al
1^ %
.a
•a
0
^•i
Id u
B 'a
S s>
60 0
^>
^ -P
H
3 ~a
a &
Clarion Systems'
Sentinel
C/!
W
H
j>
^~
^
OH
W
a>
td
•g
&
.a
a
^
60 0
b
1 60-J
^3 "§ M
"S 2 ^
i "^ o
2 ^ T3
"60 ^ _S
•to g
^~
^
OH
W
a>
td
•a
.a
a
60
H
£2 O
;i-parame
ity monit
3 g
a>
"H
Hydrolab DataSo
C/!
W
H
j>
^~
^
OH
W
a>
a
•a
.a
in
V
•s
cd
a
^
ca
g
-2
0
o
oo
Ovf
g
•g
0
o ^
13 u
B 'a
a B
a> ^
60 0
^
| 60
^ •§
;i-parame
ity monit
3 §
g
&
Hach AquaTrend
C/!
<*
W
H
5
^~
^
OH
W
a>
C3
•a
.a
in
V
•s
13
a
^
ca
^
•2
0
o
(ft
g
*a
0
^•i
'rt u
a I
a> ^
60 0
^
| 60
^ •§
;i-parame
ity monit
r- ^
0
O
O
•Q
In-Situ Model Tn
C/J
Pi
W
H
j>
^~
^
OH
W
a>
IS
13
.a
a
^
60 0
^
1 ^
}_, -9
||
a §-
o
r-
oo
1
I
C/!
C/!
W
H
i>
^~
^
OH
W
a>
IS
^
&
.a
a
g
60 0
*
13
t> 60
H
;i-parame
ity monit
3 g
a cr
o
YSI Model 6000
continuous monit
aj
o
a
^.^
a *^
^> &
0 T3
a
'g
in
V
a>
13
a
0
tj
ca
£
•2
a>
IS
_ca
13 «
B a!i
1^ 1
.a
•a
0
o ^
13 u
B 'a
a B
a> ^
60 0
^
t5
I> 00
PH
OJ .tt
II
^h
=i S
r- r->
fl
pi
H
C/!
a>
o
a>
IS
13
.a
a
IS
ta
> •'S
^ g _o
!Ǥ *
.a
•a
0
O _^j
^ ta
13 u
1 "
60 0
^
1 ^
}_, -9
^ o
i!
3 "3
a &
Clarion Systems'
Sentinel
g,
s
-------
U.S. EPA Office of Water Early Warning Systems
• Contaminant Specific Signatures are Needed
Research has begun to develop contaminant specific signatures; however, the limited number of
contaminants examined lowers the confidence of the claim that the signature really is unique.
Classes of contaminants may be identifiable, but whether a wide range of specific contaminants can
be identified has not yet been determined. Regular changes in water quality can be a source of false
alarms (i.e., false-positives). Additionally, it is not known yet whether there are specific signatures
for biological contaminants.
• Data Storage and Manipulation are Needed
For continuous real-time monitors, raw data can be generated on a scale that may be too large for
manual manipulation in a spreadsheet. Monitors that gather this volume of data would require
customized software (which vendors would usually supply) for data analysis. Utilities may choose
to archive summary data in a compressed form, to meet possible later needs. For many of the
techniques and devices reviewed in this report (see Chapters 5 to 8) the quantity of data generated
will not present a challenge. However, utilities should be mindful of what they may want their
system to handle if upgrades are planned. Approaches for storing and analyzing data are presented
in Chapter 4.
• System Can be Expensive
Not all vulnerable utility companies are able to afford the monitoring systems currently being
evaluated as EWSs. Reductions in price due to competition and technological advancements may
remedy this situation in the future, but in the interim, utilities with limited financial resources will
face a challenge in implementing online water quality monitoring.
• Cost Decisions
Current multi-parameter units cost approximately $10,000 without TOC. Although TOC appears
to be a valuable parameter to measure, it adds $18,000 to 29,000 per unit. Research needs to
determine the cost/benefit relationship of including this technology in an E WS. A basic system with
10 microprobe monitors (without TOC) linked to an existing SCADA system is estimated to cost
approximately $150,000, plus $60,000 per year in operational costs (DSRC Meeting, 2004).
Exhibit 9-3 provides a snapshot of capabilities, issues, and gaps for multi-parameter water quality
monitors.
9.3.2 Conclusion and Recommendations
Given the current developmental stage of multi-parameter technologies and based on preliminary
EPA tests, parameters that appear stable in monitoring a distribution system include chloride (ISE),
specific conductance (electrode), turbidity (nephelometric), free chlorine, and ORP. TOC also
appears to be extremely helpful, but it is expensive to monitor. Sensor probes and monitoring
systems are being developed by manufacturers. Such probes include free and total chlorine, pH,
temperature, specific conductance, chloride, nitrates, turbidity, and ORP. The cost per probe ranges
from a few hundred to several thousand dollars.
August 2005 122
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 9-3. Water Quality Monitors as EWS
Capabilities
Issues and Gaps
Generally reliable and accurate for WQ
parameters
Training and maintenance needs
considered reasonable by water utilities
Have demonstrated the ability to identify
some chemical contaminants
Wealth of multi-parameter monitors
available from a variety of companies
A few utilities have experience in using in
distribution systems
Significant testing of systems is ongoing
and will further evaluate use as an EWS
component
Generally not prohibitively expensive
Selection of parameters will determine
effectiveness
Free chlorine sensor had the best response
for contaminants tested at EPA WATERS
laboratory
Baseline data needed on regular water
quality fluctuations
Contaminant specific signatures needed
Large scale data storage and manipulation
needed
Systems can be expensive to install and
operate depending on the parameters
chosen
System not yet demonstrated for chemical
warfare agents, but future efforts planned
Easy to place, but may present challenges
for placement based on risk evaluation
(space constraints, protection of the
equipment). Also, location may be
influenced by interferants created by
water source mixing and
routine/unexpected water quality
fluctuations.
Using the multi-parameter techniques, there is some preliminary evidence to suggest that such a
system appears to be able to detect an anomaly in the system. However, it is also reasonable to be
concerned about false-positives and whether the system can provide definitive indications of
contamination by a terrorist. Just gathering the baseline data may be prohibitively expensive.
Currently, these technologies need to demonstrate the ability to detect biological contaminants or
dangerous chemicals or develop a track record of field performance (e.g., how will it behave with
biofilms). Additional testing is needed, most likely in the field, before widespread use could be
recommended. For example, no tests have been performed on chloraminated systems. Full-scale
testing by the USGS New Jersey District, EPA, and a water utility during 2006 and 2007 may help
shed light on concern over false-positives and whether a system can function within the fluctuations
of normal water quality.
The signatures being developed by Hach and others to identify contaminants or classes are difficult
to evaluate independently or validate because their methods and algorithms are considered
proprietary and therefore are not available to the research community. As additional testing is
conducted using these methods and their performance is confirmed, there is less of a reason to be
cautious in proposing the use of these signature methods. Also, the examination of water quality
parameters for use in detecting and identifying contaminants is still being evaluated by EPA, USGS,
Army, and other organizations. Yet, no field scale tests of a full EWS with these multi-parameter
components have been implemented to date. This adds to the caution in currently recommending
use of these water quality parameter-based EWSs.
August 2005
123
-------
U.S. EPA Office of Water Early Warning Systems
9.4 Evaluation of Chemical Sensors
Online continuous sensors and hand-held sensors for the detection of chemical/toxins in vapors or
air have been on the market and in use since before 9/11. The number of potential users of CBW
sensors greatly expanded after vulnerabilities to terrorist attacks were recognized. Researchers and
companies are rapidly developing technologies and products that will meet the needs of this
increased pool of end users.
• Microchip and microfiuidics technology is advancing the sensor field by enabling
miniaturization of traditional methods (e.g., GC), as well as design of new methods.
• Portable and online gas chromatographs are available and in use by first responders. GC
can reliably identify a wide range of VOCs. Several portable GCs have been tested under
EPA's ETV program. In one case, an online GC has been used in water distribution
systems.
• Kits that utilize bacteria to detect toxins have been developed for use with drinking water.
Several kits have been verified under the ETV program.
• Daphnia, mussels, algae, and fish have been incorporated into sensors for toxins in treated
waste effluent and source waters. However, only a mussel-based and a fish-based system
have been used to date with chlorinated drinking water.
• Portable infrared spectroscopy, ion mobility spectroscopy, surface acoustic waves, and
polymer composite chemoresistors technologies have been incorporated into portable
devices that can be used by first responders for the identification of numerous toxic
chemicals.
• Fiber optic cables are being coated with sensor materials to engineer continuous flexible
sensors for use in water and in air.
EPA's ETV Program has investigated a number of the sensors that are sensitive to chemical
contaminants. A few utilities have experience using the biosensors for grab samples in distribution
systems. Another utility has experience using portable GC in the distribution system.
The following are evaluations of specific chemical detection devices starting with arsenic and
cyanide and then continuing with other specific detector systems. Exhibit 9-4 summarizes how the
chemical detectors (only those available now or potentially adaptable) compare favorably against
the desired characteristics of an EWS (as described in Chapter 3).
• Arsenic Sensor
There are two basic types of technologies that are used in commercially available tests, both of
which have been third party verified by EPA's ETV Program. The first type involves a color
reaction kit (three manufactures evaluated) and the second employs anodic stripping voltammetry
(AS V; two manufacturers evaluated). The arsenic monitor manufactured by Industrial Test Systems,
Inc. (Rock Hill, SC) showed that low and high levels of interferents did not appear to affect the
detection of arsenic. A very low rate of false-positives was reported, but a variable rate of false-
negatives was reported. One source of error that ETV testers noted was that when samples contain
concentrations exceeding optimal detection, range accuracy and precision of the associated results
were reduced because of the difficulty in performing accurate dilutions in a field setting. The AS
August 2005 124
-------
bo
-&"
a
O
GC
I
•d
x.
a
o
OB
&
S
o
O
O
OB
a
v
GC
|
O
d\
•^
z
IS
a
o
u
s
0>
^ .—
fl ,c
t- p2
s 1
U a
^
^
a
C/3
"S a;
•3 .5
a H
o
cs 3
!«
o
an
o
U
*« S
fl 42
"3 *
C ^H
O o
•S
1= S
Range c
Contain in;
Description
u
Produ
>
cq
OH
W
0>
1
'3
C3
JO
0
V
•3
IT)
1 '
-M
0 ^
"o a
^.2 C1
V I ^
,2 £ ^
^.
g ^^
| ° 1 |
6ft ^ IH 6ft
1>
JO
^
O
&
arsenic
b ^n
color reaction c
anodic strippin;
voltammetry
(ASV)
CJ
_0
^
Arsenic dete
devices
>
cq
W
a>
jj
^H
cd
OJ
1 1
g
o
V
.a
g
0
1 '
i?w-
t> ti
•1 1
S 13 o>
,2 -S g
0
o
0
6ft
1>
JO
•§
o
&
1
colorimeter or
ion selective
electrodes
a
•2
fl^
Cyanide dett
devices
a
o
a
a>
JO
r^
'a
%
JO
&
&
.a
G
o
o
£
•2
cS j§
-t-j ^
a <~
a>
^
"3
0
0
•a
volatile orga
compounds
automated gas
chromatograph
INFICON
Scentograph
CMS500
a
o
a
a>
JO
*^
'a
cS
-a
&
.a
s
o
o
.60
jo '55
TO 0
1 §
a q-H
1>
0
-£
o
&
0
•a
volatile orga
compounds
gas
chromatograph
INFICON
Scentograph
CMS200
^" fe
^ >
w <;
o>
JO
*^
'a
§
JO
&
&
.a
3
0
0
.60
O
(ft
JO
•§
o
&
0
•a
volatile orga
compounds
gas
chromatograph'
mass
spectrometer
INFICON
HAPSITE®
a
o
a
a>
JO
*^
'cS
§
JO
&
&
_a> ,J>
53 -55
13 w
§ g
a *
"60
^
a> Q
CS j?
rS a>
-t-j ^
a ^
a>
0
-^
o
&
0
•a
volatile orga
compounds
gas
chromatograph'
mass
spectrometer
0
n O
Constellatiol
Technology
CT-1128
u.
g
1
0>
3
^H
cd
cd
"I
a
'g
IT)
£
-^
1> ^
*£ "S
ra ^ o
r^n a>
il
a ^
a>
JO
-^
o
&
insecticides,
nerve agents
rapid enzyme
inhibition
2
fc
Severn Trem
Enzyme Tes
£&
w *3
w •<
o>
JO
'^
'cS
&
i_j -g
» o 3
u J s
•K ~£o a
U r-< O
T3 3 0
fl
'S
in
£
-2
o
0
r-"
6ft
1>
0
-^
O
&
T3
chemicals an
biotoxins
enzyme
inhibition
ii
tu 0
60 W
a
o
a
a>
JO
^^
'cS
§
JH -^
jo '55
TO 0
'> ^
SI
_a> ^>
"I jl
rS a>
^ P
a ^
£
-2
a> ^>
Jo '55
cd o
rS a>
-t-j ^
a ^
a>
0
-^
o
&
T3
chemicals an
biotoxins
enzyme
inhibition
Randox
Laboratories
Aquanox™
§5
-------
bo
c
•&"
a
O
GC
I
•d
x.
•=.=
O
5B
X.
O
^
a
|
O
d\
•^
z
IS
=
o
u
s
"C
>
^ .—
fl ,a
t- •—
s *
U a
^
^
•^
=
cc
— ' S
Q H
o
« 3
a> •£
ft ^
O
+j
[«
o
U
1 «
O o
09
<« H
^ «
« .s
M B
S 1
* g
u
=
o
Descripti
Product
S
o
0>
JO
'3
&
JO
&
.a
g
IT)
1—1
V
1
^
S
n>
r^ Cfl
C3 0
^H U
§5
§£
1
o
&
g
8
chemicals
biotoxins
x
U ^H
Photosynthe!
enzyme com
inhibition
Lab Bell inc.
LuminoTox
g
o
0>
JO
"3
§
'o "55
cd o
'3 |
S |
a *
.a
0
1
T3
s
ss
§£
1
o
&
g
8
chemicals
biotoxins
,
submictocho
drial particle
inhibition
Harvard
BioScience, Inc.
MitoScan
^
cq
<;
OH
W
0>
JO
'3
a
g
&
&
V
.a
s
0
£
.2
, 8
-t-j ^
'CJ v
™ x
§ rn
O H tt! H
^
cq
<;
OH
W
0>
JO
'3
&
_o> ^>
Jo '55
cd o
rS o>
-t-j 8
a *
a
'g
o
m
iri
"Si
2
, 8
^ ^>
'CJ v
™ x
§5S
a>
1
o
&
a>
^
a
JO
,0
'x
0
biomonitor
(bacteria)
Hidex Oy BioTox
Flash™
^
cq
<;
OH
W
0>
JO
'3
§
h-J
g
^
O
0
o
a
'S
IT)
^
IT)
£
-S
, s
+_, a>
'^ ^
|«b-|
gss"
a>
1
o
&
a>
^
rt
M
JO
3
M
,0
'x
0
biomonitor
(bacteria)
Strategic
Diagnostics Inc
DeltaTox®
^
w
•< g
OH §
W §
0>
JO
'3
§
'o "55
cd ^o
'3 |
- i
a *
.a
in
' '
^
^
, s
-t-J ^
'^ ^ l/^
o> ^-r oo
£P S r-"
g m ^H
e
a
3
u
^
B
,0
'x
0
biomonitor
(bacteria)
Strategic
Diagnostics Inc
MicroTox®
^
cq
<;
OH
W
0>
JO
'3
&
r^> "55
cd ^o
1 »
^ s
a *
.a
S
in
^
^
.2
, 6
-t-j ^
'CJ M
•^ x
S " 0
£ S S
1
o
&
u
^
*3
JO
.0
0
biomonitor
(bacteria)
Hach ToxTrak™
Rapid Toxicity
Testing System
^
cq
<;
OH
W
0>
JO
'3
§
jo 'S
Ctf 0
"| 1
3 |
S *
.a
0
(N
^
^
, S
-t-J ^
'CJ M
™ x
ess
1
o
&
u
^
rt
[/}
JO
3
M
,0
'x
0
biomonitor
(bacteria)
InterLab Supply
POLYTOX™
-------
bo
•S
c
E^.
a
•-a
5B
O
GC
I
•d
x.
•=.=
O
5B
X.
O
tt>
a
i
o
d\
•^
z
IS
=
o
u
|S
"^
0>
^ e,.
^ p3
fl ^
iL ^
>" '3
s 5
U a
_^
.f
=
C/3
.2
•-1 S
a H
o
ta 3
ft ^
O
V3
O
U
= z
if
OH
*2
*M S
= 1
Ef S
53 2
Di a
o
U
=
o
+J
u
o
Product
tu
^
o
o
0
(N
*^3
' — ' -H
1 »
ca - — -
'o
a
_S
informal
availabl(
g|
'S ^H
g >
t2 -^
.a §
§
o
§1
.a ca
o
a
8
v£ a>
II
t§ ^
• 3 CJ
I
"3
0
to
U
jl
55
^0
0
2 ^>
biomoni
(bacteria
SYSTEM Sri.
microMAX-TOX
1>
£^
O
. — 1
jl
a
M
^H ~TH
2 g
•1 'i
i> S
% 0
u
M (N
.a
a
0
(N
^
f-
0
m
rf
1
"3
0
to
g
jl
M
^0
0
0 ^~,
biomoni
(mussels
Delta Consult
Mus selMonitor®
tu
a
o
a
i — i
'3
TO
a> ^J
"§ jl
r^n a>
^_, S
»3
V
-^
i> ^
r^> "55
cd ^o
r^n a>
+-• ^
a *
g
"rt
0
tu
§
JO
M
,0
'x
0
3"
s
0
1
_g
Biological
Monitoring Inc.
Bio-Sensor®
^
W
L
*^<
OH
U
JH tU
• S JO
1 &
t5 T3
& ca
tu 1>
Jo '55
_ca jo
-t-j ^
.a
a
[C
1
a
^ tu
TH ^
S §
1 ' o i>
"*• ^ "c3 s-^
C^\ -"tn S -^
6ft ^ -^^
1
o
to
§
JO
55
,0
'X
0
° 0-
•Tn ra
1!
^H H
>^ en
1> 1>
0>
O
r— i^ (1>
cd «3
If
0 T3
_u B
jo 'S
cd jo
not avail
from we
.a
a
0
V
•2
tu «
jo • S
TO JO
rS tu
£ 3
5 a
0 g
ri ^M
g
"3
0
to
|
jl
M
.0
0
0 ^-,
biomoni
(daphnia
bbe moldaenke
Daphnia Toximeter
tu
a
o
a
JH tU
^ jo
1 ft
t5 T3
_a> ^>
jo '£
JS JO
not avail
from we
•2
tu 1>
jo • S
TO JO
rS tu
g !*
2 a
0 g
g
"rt
0
tu
§
jl
55
,0
0
0
a ^
M £J}
O . — i
bbe moldaenke
Algae Toximeter
tu
a
o
a
JH tU
!s j3
1 ft
'S "73
& cS
a> .JJ
1 il
not avail
from we
.a
s
0
V
•2
tu
1!
US Army Center
for Environmental
Health Research
tu
s
o
JH OJ
"S JO
S 3
S ft
-t-j :3
0 T3
C3n :3
tu 1>
_ca j?
* a
g
fr
(N
•2
tu i>
4= ' S
TO JO
rS tu
£ *
~ i
a *
tu
•K
o
tu
rt
jl
t/}
Q
"><
0
f ^
s
O 1>
biomoni
(dinoflaj
Lumintox Gulf
L.C. Lumintox
-------
bo
c
-&"
a
O
GC
I
•d
'—
•=.=
O
5B
X.
O
^
a
|
O
d\
•^
z
1=
o
u
jS
"£
fli
^ .—
fl ,a
!H '2
s «
U a
£,
1?
+J
[«
=
OJ
cc
.2
•£• 9
a -S
a H
o
« 3
!«
O
[«
o
U
S *
O o
Range of
Contaminants
=
o
o.
Descri
Product
tu
a
o
a
c3 ^ o
I ^
'S "73
5
"K s
.y &
l§
0 i— i
.a
0
1
•S 's
TO 0
rS tu
£ *
-t-j a
1
o
&
Ipl
^
, Tj O
1 1 1 1 -i 1
iiiin
M ^
£H ° ^
tu
a
o
a
!s 2
S ft
S T3
tu ^>
53 -53
TO 0
rS tu
g !*
2 a
§£
3.^
^ 'ii
^ g
o ^
a *
^
^
o
•-S tu
a "§
^2 '^
_o> ^>
CS ^
rS tu
> ^
-t-J ^
1
"3
0
13
1
ITN
X-Ray
fluorescence
tu
^
o
"3 ^
If
0 T3
&
&
O
A
.a
a
1 '
V
1
53 ~53
Cd 0
> f
SI
1
o
&
explosives,
chemicals
warfare agents,
toxic chemicals
5 °
1 s —
a -5 a
III
o
Smiths Detecti
SABRE 4000
tu
a
o
a
3 53
S &
t5 T3
tu ^>
53 -53
TO 0
rS tu
^ *
« a
§£
r^ "c/}
^ --?
I g
Q O
a ^
_a> £>
42 'S
^ ^
> S
SI
?!
> ^
^_, S
1
o
&
toxic chemicals
in vapor form
tu ™
polymer
composit
chemires:
o
Cyranose® 32'
tu
a
o
a
c3 ^ o
I ^
B T3
& Cd
1> ^
3 '55
ra ^o
r^n a>
^ s
2 a
§^
'frt "^
^ 'ii
B s
Q O
a ^
i-^H -H
'S "^
r^ i>
^ s
2 a
r^> "55
cd o
> ^
^_, S
1
"3
0
toxic chemicals
in vapor form
tu .2
polymer
composit
chemires:
tu
Cyrano Scienc
Nosechip™
oo
-------
U.S. EPA Office of Water Early Warning Systems
75 arsenic test kit manufactured by Peters Engineering (Austria) showed that low and high levels
of interferents did not appear to affect the detection of arsenic. A very low rate of false-positives
and negatives was reported. One problem ETV testers noted was that the reagent tablet took up to
1.5 hours to dissolve.
The As-Top Water test kit offered by Envitop, Ltd. (Oulu, Finland) indicated the presence of
interferents did not affect the arsenic test. The effect of operator skill level appears to be a
significant factor with the As-Top Water test kit. The non-technical operator rarely detected arsenic
in any samples, even those containing arsenic at over 90 ppb. The technical operator detected
arsenic more frequently, though rarely at the same concentration determined by the reference
method. One problem ETV testers noted was the color on the indicator did not correspond exactly
to colors on the comparison card.
Monitoring Technologies International Pty., Ltd. (Perth, Australia) offers the PDV 6000 portable
analyzer, which measures arsenic in water using anodic stripping voltammetry (AS V). ETV testers
noted that instructions in the operation manual for water analysis were difficult to follow, indicating
that the level of experience in the operation of the PDV 6000 analyzer and associated software is
likely to influence the reliability of results. Low and high levels of interferents (iron and/or sulfide)
adversely affected the detection of arsenic.
Another device that employs ASV technology is the Nano-Band™ Explorer made by TraceDetect
(Seattle, Washington). ETV testers noted the Nano-Band™ Explorer did not appear to be affected
by matrix interferences added to the samples. However, the data from the two operators were quite
different, with the non-technical operator reporting no detectable arsenic in any of the 16 samples.
Cyanide Sensor
There are two basic types of technologies (colorimetric and solid sensing element) that are used in
commercially available tests, both of which have been third party verified by EPA's ETV Program.
Both types of technologies are used in portable devices designed for onsite rapid analysis of cyanide
in water. One problem common to all four portable colorimeters tested by the ETV was that
performing analyses under extremely cold conditions (sample water 4 to 6° C) negatively affected
the performance of the reagents. Also, for all four colormetrics, when a lethal amount of cyanide
was present, a drastic, rapid change in color was visible and no reading by the colorimeter was
necessary, making lethal amounts of cyanide in water quickly detectable. Using the WR V-1000
multi-analyte photometer, there was a slight bias for a technical operator compared to a non-
technical operator. It takes approximately 17 minutes to obtain and analyze one sample. Using the
1919 SMART 2 Colorimeter, technical verses non-technical operator did not impact results. For the
Mini-Analyst Model 942-032 by Orbeco-Hellige (Farmingdale, NY), the manufacturer recommends
adjusting the pH of water samples to between 6.0 and 7.0. Since gaseous hydrogen cyanide can be
released at a pH less than 9.0, this adjustment is not desirable from a safety standpoint, especially
if lethal/near-lethal concentrations of cyanide are present. Although the model was easily
transported to the field and the sample preparation instructions were clear, the liquid pyridine
reagent had an offensive odor, and the granular reagent tablets were difficult to open. Also, the
operators stated that it was inconvenient to keep track of the mixing and waiting times during the
analysis. There was a slight bias for a technical operator. Using the AQUAfast® IV AQ4000
August 2005 129
-------
U.S. EPA Office of Water Early Warning Systems
colorimeter by Thermo Orion (Beverly, MA), there was very little difference in results generated
by the non-technical operator compared with those of the technical operator.
Using the solid sensing element, the Thermo Orion Model 9606 Cyanide Electrode required
calibration and electrode polishing before every sample set. Operator bias was not tested. Also,
using the solid sensing element, the Cyanide Electrode CN 501 with the Reference Electrode R503D
and Ion Pocket Meter 340i (WTW ISE) by WTW Measurement Systems (Ft. Myers, FL) had a
difficult to understand instruction manual. A one-hour telephone consultation with WTW was
required before the WTW ISE was easy to operate. Operator bias was not evaluated.
MosselMonitor® by Delta Consult
MosselMonitor® detects toxins. One of the major problems for using the MosselMonitor® was the
low food content of the water to be monitored (fast running surface waters, groundwater). An
"Automated Food Device" (AFD) was developed to automatically and continuously feed algae to
the mussels using a flow-though system. Mussels are very sensitive to chlorine, and addition of
thiosulphate to the water minimizes the effects of free chlorine (Jan de Maat, Delta Consult, personal
communication). As a result of the adaptations, the Budapest Waterworks have now successfully
applied the MosselMonitor® in monitoring of chlorinated drinking water for 10 months.
Eclox™ by Severn Trent Services
Eclox™ detects chemicals and biotoxins. Sample analysis is simple and takes only five minutes.
Contaminant concentrations could be detected to a limit of ug/L to mg/L, but results were not
consistently reproducible. When compared to the similar detection device Microtox®, the actual
contaminant values generated by these devices may vary in different types of water, especially in
distilled water. It is necessary to establish site-specific baseline values (States, 2004). In another
study, clean chlorinated and chloraminated water samples produced very low inhibition of light,
indicating that the disinfection byproducts that may be present in drinking water do not interfere
with Eclox™ results. However, false-negative results were produced by lethal doses of soman and
Botulinum toxin. Eclox™ is easily transported and operated in the field, where similar results were
produced as in the lab (EPA-ETV, 2004).
MicroTox® and DeltaTox® by Strategic Diagnostics, Inc.
MicroTox® and DeltaTox® detect chemicals and biotoxins. Sample analysis is moderately difficult
and takes 45 minutes. Copper was shown to be a potential interferent. Contaminant concentrations
could be detected to a limit of ug/L to mg/L, but results were not consistently reproducible. When
compared to the similar detection device Eclox™, the actual contaminant values generated by these
devices may vary in different types of water, especially in distilled water. It is necessary to establish
baseline values for each site (States, 2004). In another study, MicroTox® had false-positive results
when testing clean chloraminated water, but not when testing clean chlorinated water. Half the
contaminants tested produced false-negatives when lethal doses were present. The operation of
MicroTox® was easy to understand for in-lab operation. This product is not field portable. For
DeltaTox®, false-positive results were produced when testing clean chloraminated water, but not
when testing clean chlorinated water. Over half the contaminants tested produced false-negatives
August 2005 130
-------
U.S. EPA Office of Water Early Warning Systems
when lethal doses were present. DeltaTox® is straightforward to operate and easily transported to
the field (EPA-ETV, 2004).
Tox Screen II by Checklight, Ltd.
When using the pro-organic buffer, the low level of light production in water disinfected by
chlorination or chloramination may interfere with the ToxScreen II results, resulting in false-
positives. However, residual sodium thiosulfate from dechlorination may have caused these results.
When using the pro-metal buffer, as long as a similar reference sample is used, water disinfected
using either process is not likely to interfere with the ToxScreen II results. Operation of the
ToxScreen II was relatively straight forward and the instrument was easily transported to the field
when similar results as in the lab were produced (EPA-ETV, 2004).
ToxTrak™ by Hach Company
Half of the samples from a water system using chlorination were analyzed in July and the other half
in September. In July, a significant amount of inhibition was noted, while in September, the same
sample was largely non-inhibitory. There seems to be a risk for false-positive results due to
interference of previously chlorinated water, though the reason for the difference in these results is
not clear. False-positives were also experienced due to iron in the water samples. False-negative
results were experienced when samples with lethal doses of contaminants were tested. The
ToxTrak™ was easy for testers to operate and is field portable, but the ToxTrak™ reagent must be
incubated over night at 35°C, which could be problematic for field deployment (EPA-ETV, 2004).
IQ Toxicity Test™ by Aqua Survey
Aluminum, copper, and iron are potential interferents because they adversely affected 90 percent
to 100 percent of the Daphnia organisms. In addition, all Daphnia organisms exposed to drinking
water from a system disinfected by chloramination were adversely affected, and therefore produced
false-positives. However, the water sample from a drinking water system that uses chlorination did
not adversely affect the daphnia. There were no false-negative results produced when using this test.
The IQ Toxicity test™ instruction manual is easy to understand and the test is field portable, though
a stock of Daphnia must be maintained to facilitate short-notice field testing (EPA-ETV, 2004).
BioTox Flash™ by Hidex Oy
Bacterial metabolic inhibition caused by copper and zinc may result in interference with BioTox™
results. Slightly exaggerated inhibition may result when using chloraminated water, resulting in a
possible false-positive, while there is a risk for false-negatives when using chlorinated water
samples. The BioTox™ is field portable, but a flat, steady surface is needed to operate the
BioTox™. ETV testers found it difficult to operate the BioTox™ without an instruction manual,
but once the correct procedure was determined, operation was easy (EPA-ETV, 2004).
POLYTOX™ by Interlab Supply, Ltd.
Without a baseline water sample of a matrix similar to the test water sample, there is a considerable
risk that an analysis of POLYTOX™ in clean chloraminated water would produce organism
August 2005 131
-------
U.S. EPA Office of Water Early Warning Systems
respiratory inhibition great enough to yield a false-positive result. In clean chlorinated water
inhibition was low enough that false-positives were not produced. Over half of the contaminants
tested produced a false-negative when present at lethal doses. The POLYTOX™ is field portable
and ETV testers had no difficulty operating it (EPA-ETV, 2004).
Pesticide/Nerve Agent by Severn Trent Services
Pesticide/Nerve Agent is a rapid enzyme test that detects pesticides and nerve agents. Sample
analysis is performed in five minutes and is generally simple. This test is easily conducted using
concentrated or non-concentrated sample water (States, 2004).
HAPSITE® by INFICON Inc.
HAPSITE® is a field deployable gas chromatograph-mass spectrometer (GC-MS) that detects VOCs
found in toxic substances and chemical warfare agents. The recent addition of a "Situ Probe" purge-
and-trap sampling device enables water samples to be analyzed. The GC portion of the device
detects volatile substances with a molecular weight between 45 and 300, and the MS portion of the
device identifies the compound from a library of 170,000 organic compounds (States, 2004).
Sample analysis takes 60 minutes but is very difficult to perform. This system has been deployed
in a distribution system and could serve as a case study.
9.4.1 Issues and Gaps
This section highlights various issues and gaps in using chemical detectors for early warning for
finished drinking water.
• Cost for Some Available Technology is High
Online and portable devices such as GC-MS are expensive, ranging from $75,000 to $95,000.
• Field Kits are Not Optimal
The bacterial monitoring kits tested by EPA' s ETV have high false-positive and false-negative rates.
A common drawback for kits is the stability of reagents. Often reagents require reconstitution (if
they are lyophilized) or careful measurement of reaction components to constitute a fresh reaction
mixture. Kits can be subject to variability due to different users because it is difficult for users to
mix and pipette in a consistent manner. They require trained personnel and have set up
requirements, such as culturing log phase growth bacteria. The results do not provide specific
identification of a toxin. Although these kits may be suitable for confirming the presence of a toxin,
further methods would need to be utilized for specific identification.
• Some Detection Methods Face Challenges from the Chlorine Residual
Organism-based biomonitors are sensitive to the chlorine residuals in drinking water. Although the
fish-based Bio-Sensor® and MosselMonitor® remove chlorine, broadly applicable methods for
chlorine removal have not yet been developed. Although one company, Checklight, is developing
August 2005 132
-------
U.S. EPA Office of Water Early Warning Systems
a system for removal of chlorine, the effect of chlorine removal on other biosensors has not been
demonstrated.
• Many Technologies Have Not Been Adapted for Water
Portable infrared spectroscopy, ion mobility spectroscopy, surface acoustic waves, and polymer
composite chemoresistors technologies are being aggressively pursued for air and vapor
applications, but have not been developed specifically for drinking water monitoring. These
techniques could be adapted for water if companies' market research indicates there is a potential
market.
Exhibit 9-5 provides a snapshot of capabilities, issues, and gaps for chemical sensor technologies
and techniques.
9.4.2 Conclusion and Recommendations
Portable technology (e.g., GC) is available for conducting analysis of many possible chemical
contaminants. This area will continue to improve as high technology equipment is based on
microchip technology (e.g., nose chip). Certain biomonitors are portable and can be used for site
assessments. Chlorine would have to be removed for such analyses. The technology that is readily
available and reliable are specific probes like arsenic and cyanide probes that could be effective
against a narrow selection of contaminants. Online technologies are not cost effective and are not
reasonably available. GC and ion mobility have cost and technical challenges. The current
experience of using high-tech GC in a utility's distribution system has not been determined to be
cost-effective to cover the distribution system.
Certain biomonitors may be promising if the issue of chlorine and chloramine residual interference
can be resolved. For example, the mussel monitor has been recently demonstrated in finished water
in Europe and there are various efforts to make other monitors (MicroTox® and ToxScreen)
adaptable to finished water. The mussel monitor could be a good candidate for an EPA-ETV
Program verification and perhaps used for a laboratory and/or field study on finished water or with
CBR surrogates or agents. In the next three years, the field should show further development in
terms of cost effective and reliable devices. A few new technologies (such as microchips) could
revolutionize the chemical detection field for drinking water.
August 2005 133
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 9-5. Chemical Sensor Technologies and Techniques
Method
Arsenic and
cyanide probes
GC
Enzyme
Inhibition
Bacterial
biomonitors
Daphnia, fish,
algae biomonitors
Mussel and fish
biomonitors
Eukaryotic cell
and tissue
biomonitors
Fiber optic cable-
based sensors
Infrared
Ion mobility
SAW -based
sensors
Microchip
chemoresistors
Capabilities
• established monitors for arsenic/cyanide
• reliable
• ETV reports available
• range of VOCs detected
• online for drinking water
• detects chemicals (phenols, amines, heavy
metals)
• detects substances toxic to cholinesterase
(nerve agents and pesticides
• portable (kit)
• detects substances toxic to bacteria
• portable (kit)
• ETV reports available
• online for surface water
• detects substances toxic to daphnia, fish
or algae
• online for source and treated waste water
monitoring
• detects substances toxic to mussels or fish
• online systems exist
• in use in a few distribution systems
• can potentially detect substances toxic to
human cells
• detects toxic chemicals
• being designed to be on line for water
distribution systems
• can cover long continuous areas
• identifies wide range of substances
• portable (suitcase)
• detects wide range of compounds
(explosives, chemical warfare agents,
toxic industrial chemicals or narcotic
substances)
• portable (hand-held)
• detects VOCs, explosives, illicit drugs,
and chemical warfare agents
• online for groundwater monitoring
• detects wide range of volatile compounds
• portable (hand-held)
Issues and Gaps
• limited to one parameter identified at a
time
• various issues (reliability) with certain
methods
• expensive
• not widely used for drinking water
• requires mixing and pipetting
• chlorine residual interferes (false-positive
source)
• requires mixing and pipetting
• false-positive/false-negative rate high
(ETV reports)
• chlorine residual interferes
• chlorine residual adversely affects the
living organisms
• not proven for drinking water
• no third party verification
• emerging technology, commercial
products not available
• emerging technology
• most advanced products are for air
• water matrix interference requires use of
extraction method
• unidentified substance must be
concentrated
• expensive
• developed for vapor sensing, would
require accessory equipment for water
• MEMS-SAW is emerging technology
• water applications will lag behind air
applications
• developed for vapor sensing, would
require accessory equipment for water
August 2005
134
-------
U.S. EPA Office of Water Early Warning Systems
9.5 Evaluation of Microbial Sensors
Rapid microbial detection technologies are not as advanced as chemical detection technologies.
However, advances in the last few decades in molecular biology, genomics, and microfluidics, have
stimulated the imagination of researchers and resulted in the appearance of the first generation of
detection devices for microbes. There are many technological approaches with laboratory
prototypes and the beginning of commercial product development.
• Immunoassays are available in several formats including: test strips, columns, linked to fiber
optics, attached to microspheres and quantum dots, and incorporated into a wide array of
microchip technologies. Specific contaminants can be identified with antibody-based
technologies.
• ATP is a general indicator for the presence of live cells. There are numerous small portable
kits for detecting ATP that are used in the food industry. There are also portable kits for
testing grab samples in water.
• Portable and rapid (30 minutes or less) PCR is a reality, with four systems currently on the
market.
• Light scattering technologies have a long history of use for measuring turbidity and are now
being adapted for detecting microbial cells in drinking water.
• Technologies that utilize microchips/microarrays are being aggressively pursued. The
proliferation of approaches should yield some viable options for drinking water applications
in the long-term.
ECBC conducts research applicable to biological and chemical defense sensors and integrated
detection systems. It has conducted many studies, including some that can detect contaminants in
water. In March 2002, ECBC issued the Bio-Detector Assessment (ECBC, 2002). The report
evaluated several biological detection devices including immunoassay-based or nucleic acid-based
technologies. The devices include Bio-HAZ™, FACSCount, Luminex 100, ANALYTE 2000,
BioDetector, Hand-Held Assays, ORIGEN Analyzer, Tetracore Tickets, Cepheid Smart Cycler®,
and Rapid System. The ANALYTE 2000 technology has been replaced by the company's newer
technology, RAPTOR™. ORIGEN Analyzer technology is now being further developed by
BioVeris. The evaluation used criteria including portability, reliability, time to analyze, classes
detected, viability, sensitivity to bacteria, toxins, viruses, ease of use, rate of processing samples,
portability, and price for consumables. The methods were evaluated using a quantitative scale and
input from experts. Hand-Held Assays, Tetracore, andNew Horizon (referred to as Smart™ Tickets
elsewhere in this report) received the highest score of all immunoassay-based devices. PCR was
credited with being specific and sensitive and not prone to false-positives. Specifically, RAPID
from Idaho Technologies received high scores. Freeze-dried PCR assays are already available, and
the system can operated from a battery and connected via Internet to transmit data to remote sites
to assist in response. ECBC also has a program (Early Sentinal Biomonitoring System Program)
to screen and examine, and verify the performance of up to 28 technologies for rapid detection for
drinking water (Stanley States, Pittsburgh Water and Sewer Authority, personal communication).
Evaluations of specific microbial detection devices are provided below. Exhibit 9-6 summarizes
how the microbial detectors (those available now or potentially adaptable) compare with the desired
characteristics of an EWS (as described in Chapter 3).
August 2005 135
-------
bo
c
-&"
a
O
GC
I
•d
h
a
o
OB
&
S
o
O
o
OB
a
V
GC
o
o\
+j
z
=
o
+j
u
s
^
OJ
a 2
*- •=
3 «3
an
C/l
an
j? I
cs •"
a H
o
« s
o
V3
O
u
= 3
•— C3
fl "t^
O o
Range of
Contaminant!
=
o
+j
o.
u
OJ
C
Product
O
- ffl
ca ^
s *•>
< W
IS
'3
ca
iiVH1
C3 w ^". • *>
.a
in
^
•9
fo?
ill
in ^* ft
C~l /"^ 'Pn
S^^
i>
0
C
&
biological agents
including anthrax.
botulinum toxin,
ricin
1
•c3
S
^>^
ll
a a
Tetracore Bio
Threat Alert (BTA)
U-
^
s
1
IS
Is
ca
bacteria 10s
cfu/mL;
biotoxin 50
ppb
.a
IT)
£
-^
&
0
^0
e
0
&
biological agents
including anthrax.
botulinum toxin,
ricin
1
^
}-H
r^
!>x
o ^
s »
H M
a a
New Horizons
Diagnostics
SMART™ Tickets
O
ffl
O
w
1>
1
'ca
ca
see above
.a
s
0
V
^
•9
0
o
0
o"
(N
6«
1>
IS
£
0
&
biological agents
including anthrax.
botulinum toxin,
ricin
1
•c3
S
^
o ^
f~\ r*~t
EAI Corporation
Bio-HAZ™ (with
SMART™ tickets)
U
I
x
"3 2
S -
S &
-t^ TO
0 T3
anthrax
detected at
< l.Ong/mL
.a
g
(N
r--
|
•3
2
0
o
0
o"
in
6ft
1>
0
e
0
&
biological agents,
toxins, chem
contaminants
^
t/}
Cfl
U
"S
K^ lij
"6" S
o o
1 §
cd ^H
Research
International
RAPTOR™
L
W
^
OH
W
"I
It
a
s i ^ 7
pii
ra en ,£! G
.a
in
^
•9
in c
2
£
o
&
g
biological agents
(anthrax, botulinu
toxin, ricin)
1
•c3
S
^
ll
Response
Biomedical Corp
Test Cartridges
^
W
^4
OH
W
IS
"a
>
.B
in
^
•9
0
in .B*
(N h
6ft to
1>
IS
e
o
&
s
biological agents
(anthrax, botulinu
toxin, ricin)
1
^
-S
r^
!>x
o .
O P*~l
1 1
ADVNT
BADD Test Strips
U
i
x
3 is
s ^
° S1
0 T3
'I '^
ll
||
o
ll
60
a
a -^
1 ^
1 §
£ ~
1 §
"a
o
_s
biological
contaminants
•S g
^ •§
£ o
r^H ^ 1>
§"'60 g
Sou
a "3 "3
s is a
LLNL Autonomous
Pathogen Detection
System using
Luminex
Corporation's
xMAP®
CO
-------
bo
c
•&"
a
£
O
GC
I
•d
'—
§1
•el
5« -g
^ S
So
U
O ^
O
5B
X.
O
^
a
v
GC
o
o\
+j
z
a
o
U
Oj
^i ^""
— " U
a 3
U CB
!H ™"
>• 'S
a ^
^
•—
.^
an
C/l
an
'a? OJ
S.§
a H
^
o
"eg ^
"" 3
O
•M
O
U
a 3
•— C5
a "t^
0 (2
+^
tM fl
o 2
o> — «
g 1
OS a
o
U
a
O
B.
(j
an
Q
Product
a
g
^
Cd
Cd
a>
'o 1>
cd .-y
cd ,C
o a
a o
tu
o a
a o
0
^ r5
•2 ^d
2 "cd
a i>
C cd
11
« o
cd "55
•"^ . ^
"o n
d
1>
^o
B
&
en
1
i>
-*-j
a>
o
^
u
g
c
BioDetect
Microcyte Aqua®
a
g
JH
ca
ca
>
S
N
[/}
rt
r-H S
•rt 3.
1 ^
& A
fl
'g
0
^ 2
• 2 ^ca
1 '3
P >
C ca
11
« o
•i ' ™
II
"o n
d
0>
.g
o
[/}
^
0
"S
[/}
JH
e
ca
&
60 tu
'II
il
"bO ^
^ &
Brightwell
Technologies
Micro-Flow
Imaging
g
g
r^H
cd
>
c5
»
r^ tU
ca .-y
'8 ,0
> tu
'o p
S 0
0
V
|
o
0
o_
fN
6ft
tu
*o
ca
B
&
OH
H
1
t/}
.1
£H
^
AMSALite™
Antimicrobial
Specialists and
Associates
a
g
r^H
^
^
>
cd
h-1
s
1
0
o
'g
0
V
|
a>
2 CL>
cd .-H
"c^ £
> d>
t^
o a
a o
a>
'o
B
&
OH
H
1
1
|
WaterGiene™
Charm Sciences, Inc
g
g
JH
ca
ca
tu
*o tu
ca ,-y
ca ^
0 S
S 0
S
'g
tu
*7^ j-j
^ca .-y
ca ^
'o S
S 0
0
o
0
o"
6ft
tu
.g
o
OH
H
tu
S
tu
tu
R
£
Continuous Flow
ATP Detector
BioTrace
International
a
g
Jj. tu
"c3 2
S -
S &
-t-j ca
O T3
& ca
tu
'o tu
ca ,-y
'S ^3
> tu
o a
a o
.g
0
V
tu
*o tu
r3 '^
'S f>
o a
a o
tu
1 tt
II
t^
0 fl
fl o
a>
'o
B
&
OH
H
1
1
S
•3
Celsis-Lumac
Landgraaf, the
Netherlands
a
g
JH
ra
ca
ca
tu
*o tu
ca ,-y
ca ^^
> tu
o a
a o
.g
V
»
^Q 1>
r3 "^
cd ,c
!> a>
1 §
a>
^ a>
!> a>
^j t^
o fl
fl o
a>
'o
B
&
OH
H
tu
S
1
£H
•3
Profile™ -1 (using
Filtravette™)
New Horizons
Diagnostic Corp
s
g
.g
O
+£
"B ^
'M'S
-2 |
•2 §
r^ CJ
d5
"B
t!
cd
u
en
r^t
^
LXT/JMAR
BioSentry
a
g
^ tu
3 2
'•g ^ca
8 &
-t-j ca
O T3
& ca
tu
2 tu
ca .-y
ll
o a
a o
tu
2 tu
not avails
on websi!
»
2 tu
ca .-y
'> -^
o a
a o
»
1 s
||
t^
o a
a o
i>
^
o
•^
"B ^
'Bo S
1!
r^ U
60
.g
^a>
Id
|2
^n "^
a S
3
CO
-------
60
•S
c
E^.
a
O
GC
I
•d
x.
§1
•el
5« -g
^ S
So
U
O ^^
O
5B
X.
O
tt>
a
o
o\
+j
15
a
O
*z
U
S
tu
>> ^-
^ —
a 3
^ ^
g 'S
^* ^
Sensitivit
[«
II
a H
o
a 5
Q. ^
O
^
O
U
a 3
"^ +j
O o
&H
+J
*i-( S
° fl
aj — •
a |
OS a
o
U
a
O
*^
=•
^
U
s
C
Product
u.
&
g O
•< w
^,>
ca 2
S ^
'S T3
J
g
0
O
„
fl
'g
o
g
^
•3
g
0
o
I/-1
v$
^
e
Q
&
-1
11
l§
^
O
OH
Idaho Technology
RAPID
§
^ tu
a f,
'% ^
0 "3
"On
S S S"
'g
o
g
^
•3
g
0
o
(N
6«
^
e
Q
&
_1
11
3 §
J2 0
^
O
OH
Smiths Detection
Bio-Seeq™
H °
W "
^ o>
ca 2
If
0 T3
J
1
.a
0
.£P
o>
42 0>
> tu
o a
a o
a
tu
2
2
e
t/}
r-9 -^
•^H O
r^> O
0<
o
OH
Invitrogen
PathAlert™
i
^ tu
.2 3
S 'a
-^^ TO
0 T3
CU
information
not availabl
'g
0
V
a -^
1 ^
g >
11
tu
"§ B
'> ^
o a
a o
tu
^
e
0
&
•al
O ^
•^H O
T^> Ci
&,
o
OH
Cepheid Hand-Held
Nucleic Acid
Analyzer (HANAA)
O
ffl
O
W
^ tu
.2 3
tU ^H
0 T3
PH C3
k>
< 30 anthra)
spores
„
fl
'g
o
g
^
•3
g
0
o
0_
*?
v$
^
e
Q
&
•al
'5b g
0 ^
' — ^ G
•^ O
42 U
0<
O
OH
Cepheid Smart
Cycler® XC
i
.^ ,u
ca 2
If
0 T3
CU
not availabl
on website
ji
1
^
3 „
^ 42
> tu
0 S
a o
tu
42 ^tU
|1
'o fl
a o
a>
^
e
Q
&
•al
'5b g
_o ^
• ^ o
42 U
0<
O
OH
Ibis Pharmaceuticals
and SAIC
T.I.G.E.R.
1
4^ tu
.2 42
tu ^H
0 T3
not availabl
on website
A
ca 42
0 S
a o
JJ
^ 42
> tu
0 S
S 0
^
OS
ON"
6«
S
V$
^
e
0
1
_o g
"2 4d
B o?
C3 OH
S S
r^ tU
& 0
« o
B tu
Nomadics®
Spreeta™
Evaluation Module
i
4^ u
a f,
S -
U PH
0 T3
information
not availabl
en
^
H
a -^
1 ^
G >
11
r-^H
§ ^
1 ^
^ o
•9 d
aJ
2 i>
^ -S
Q pH
& 0
1-
c?'g
.2^
r^> U
a>
> .^
ft
ri o
•-H d> ^
PH S 0
o a> i2
*H d Q
3: ra 1>
^— i 1> t/}
Georgia Tech
BOSS
i
4^> tu
a f,
If
0 T3
not availabl
on website
d
'g
IT)
g
^
S
tu
"ca B
> tU
a ij
'o fl
a o
tu
^
e
0
^|
o fS
° §
42 U
8
i
o
J3
r^
tu
•?
a>
0
Innovative
BioSensors Inc.
BioFlash™
i
!=l u
.2 3
<3 'a
^-J TO
0 T3
not availabl
on website
d
'g
1 '
.£P
tu
"§ B
ca tj
o a
a o
^
tu
2
2
e
, i
0
'«>
"o
j
O
w
BioVeris
MIM
oo
CO
-------
U.S. EPA Office of Water Early Warning Systems
BTA Test Strips by Tetracore (Gaithersburg, MD)
BTA Strips detect pathogens by rapid immunoassay. In one evaluation study, the detection limit
was found to be 105 cfu/mL for pathogenic bacteria. They are useful for screening acute hazards,
but poor sensitivity limits their usefulness in detecting low levels of contaminants in water. Analysis
is simple and takes 15 minutes (States, 2004). EPA conducted a second evaluation test for anthrax,
botulinum, and ricin. The anthrax test strips generated one false-positive when testing each Florida
and New York drinking water sample. False-negatives were generated in water due to Ca and Mg
ion presence, and in concentrated drinking water from California and New York. The strips were
not able to detect spores at the vendor indicated concentration limit of 105, only at levels 100 to
1,000 times greater. For Botulinum toxin test strips, false-positives resulted due to the presence of
humic acid, fulvic acid, and lipopolysaccharides in water. There were no false-negatives. Both
toxin types were detected near the vendor-indicated limit of 0.1 mg/L. For ricin, false-positives
occurred in Florida drinking water while false-negatives occurred in New York drinking water. Test
strips can detect to concentrations of 0.035 mg/L, as stated by vendor. All types of BTA test strips
had consistency near 100 percent. A 25 test strip kit costs $625 and the Alexeter strip reader costs
$4,000 (EPA-ETV, 2004).
In a third evaluation study, BTA Tetracore Tickets identified four of eight contaminants. Many
BTA reagents are available commercially or through the government and are not likely to produce
false-positives or negatives. Furthermore, there was low opportunity for error due to ease of use and
it takes less than 30 minutes to set up the device and analyze one sample. It is a hand-held device
and costs $4,500. Tetracore scored one of the highest of all immunoassay-based devices in this
evaluation (ECBC, 2002).
SMART™ Tickets by New Horizons Diagnostics (Columbia, MD)
SMART™ Tickets detect biotoxins by rapid immunoassay with a detection limit of 2 to 50 ug/L.
Analysis is simple and takes 15 minutes; however, poor sensitivity limits their usefulness in
detecting low levels of contaminants in water (States, 2004). SMART™ Tickets have been
incorporated into Bio-HAZ™. In a study of Bio-HAZ™ with New Horizon SMART™ Tickets, it
was able to identify all four of the bioagent classes (sporulated bacteria, vegetative bacteria, toxins,
and viruses), and identified four of eight traditional bioagents identified by DOD. Some SMART™
Tickets reagents are available commercially or through the government and the method is not likely
to produce false-positives or negatives. There is also low opportunity for error due to ease of use
and it takes less than 30 minutes to set up and analyze one sample. It is field portable and costs
$20,000. Bio-HAZ™ scored one of the highest of all immunoassay-based devices (ECBC, 2002).
R.A.P.I.D. (Ruggedized Advanced Pathogen Identification Device) by Idaho Technology
R.A.P.I.D. uses PCR to detect pathogens and biotoxins to a limit of 103 cfu/mL. This device is used
extensively by the military, but the detection capabilities suggest potential application for water
utility personnel or contaminated drinking water investigators. The sample preparation is
standardized, the kit includes positive and negative DNA controls, and initial data interpretation is
automated. The effectiveness of this test is limited by its sensitivity. Analysis is moderately
difficult and takes 90 minutes (States, 2004). In another evaluation study, the Rapid System
identified four of eight contaminants with half of reagents being available commercially or through
August 2005 139
-------
U.S. EPA Office of Water Early Warning Systems
the government. It was not likely to produce false-positives or negatives, but had a high opportunity
for error due to difficulty of use. The Rapid System takes approximately three hours to set up the
device and analyze one sample. It is field portable and costs $55,000 (ECBC, 2002).
Eclox™ by Severn Trent Services
Eclox™ detects chemicals and biotoxins. Sample analysis is simple and takes only five minutes.
Contaminant concentrations could be detected to a limit of ng/L to mg/L, but results were not
consistently reproducible. When compared to the similar detection device MicroTox®, the actual
contaminant values generated by these devices may vary in different types of water, especially in
distilled water. It is necessary to establish baseline values for each site (States, 2004). In another
study, clean, chlorinated, and chloraminated water samples produced very low inhibition of light,
indicating that byproducts of either disinfection process that may be present in drinking water do
not interfere with Eclox™ results. However, false-negative results were produced by lethal doses
of soman and butulinum toxin. Eclox™ is easily transported and operated in the field, where similar
results were produced as in the lab (EPA-ETV, 2004).
Hand Held Assays from the Department of Defense
HandHeld Assays can identify seven of the eight traditional bioagents identified by DOD. Reagents
are mostly available commercially or through the government with the assays being found to be less
unlikely to produce false-positives or negatives. They have a low opportunity for error due to ease
of use and take less than 30 minutes to set up and analyze one sample. HandHeld Assays scored
high in the ECBC study (ECBC, 2002). In another study, HandHeld Assays were determined to
have limits of detection that are many times the infectious dose. There can be false-positives
because of environmental contamination and they are sometimes not used properly. The DOD
determined that the HandHeld Assays are effective if used in concert with other confirmatory
detectors (Emanuel et al., 2003). In another report generated by scientific experts from 14 federal
agencies,207 HandHheld Assays had either high false-positive rates (ranging from 3 percent to
83 percent) or sensitivity problems. Thus, none of the assays could be considered reliable for field
detection.
Smart Cycler® from Cepheid
Smart Cycler® identified four of eight traditional bio agents identified by DOD, but the reagents are
not currently available commercially or through the government. It is less likely to produce false-
positives or negatives, has high opportunity for error due to difficulty of use, and takes
approximately three hours to set up the device and analyze one sample. It is field portable and costs
$35,000 (ECBC, 2002).
Light-Scatter
The identification rate of Cryptosporidium parvum oocysts varied from 11 percent to 45 percent,
and false-positive rates varied from 0.3 percent to 3 percent. The MALS system may be tuned by
the user, who must understand that a higher identification rate will be accompanied by a higher
false-positive rate. MALS was also able to differentiate between different physical states of
Cryptosporidium oocysts, including oocysts treated with ozone, heat treated, or excysted from live
August 2005 140
-------
U.S. EPA Office of Water Early Warning Systems
untreated oocysts. The limit of detection was such that MALS could be used as an early warning
tool for water contamination outbreaks. For purified water, the estimated limit of detection (ELOD)
was found to be 7, 0.7, and 0.1 oocysts/mL in 1, 10, and 60 minutes respectively. For finished
drinking water samples, the ELOD was 75, 7.5, and 1 oocysts/mL in 1, 10, and 60 minutes
respectively. The researchers concluded that MALS technology has suitable applications to water
distribution system monitoring (Quist et al., 2004, AwwaRF Project #2720, see Appendix D).
RAMP Anthrax Assay by Response Biomedical Corporation (Vancouver, Canada)
RAMP uses a cartridge that is a lateral flow immunoassay device with a test and control line. The
detector is an antigen-specific antibody attached to a fluorescent bead. The RAMP instrument
detects the presence of fluorescent beads that attach to the capture assay line. In the test study, three
non-pathogenic strains of Bacillus anthracis and three non-anthracis Bacillus strains were tested.
The detection limit of the three Bacillus anthracis strains ranged from 1,000 to 2,000 spores. The
non-anthracis Bacilli did not cross-react and no false-positives were produced in the presence of
interferents. The RAMP Anthrax Assay has not been tested for use in water systems (Heroux and
Anderson). In another verification test conducted by EPA, three types of test cartridges were
available for anthrax, botulinum toxin, and ricin. The anthrax cartridges produced no false-positives
or negatives due to interferants, though were not able to detect anthrax spores at the vendor indicated
concentration of 4xl05 spores/mL (only at levels 100 to 1,000 times greater). The botulinum toxin
cartridges had no false-positives, but were not able to detect Type B. They detected Type A to a
concentration of 2 mg/L, though the vendor indicated detection limit was 0.5 mg/L. No false-
positives or negatives resulted when testing the ricin cartridges, which could detect to a
concentration of 5 mg/L, thought the vendor indicated limit was 1 mg/L. Results from all types of
cartridges were very consistent. Sample throughput was four samples per hour. The cartridges were
portable and easy to use with minor direction from a trained operator. Cost is $10,000 for 25
cartridges, a reader, carrier, and printer (EPA-ETV, 2004).
BADD Test Strips by ADVNT
Three types of test strips are available that detect either anthrax, botulinum toxin, or ricin. The
anthrax strip had no false-positives, but there was one false-negative in concentrated New York
drinking water. Consistency was 90 percent. Sensitivity was 4 x 107 to 8 x 107 spores/mL. The
botulinum toxin strips had no false-positives or negatives due to interferents. However, the strips
were not able to reproducibly detect Type B toxin, and only detected Type A at a concentration of
5 mg/L. However, the vendor stated that the limit was 0.4 mg/L for either type. Consistency was
84 percent. The ricin strips produced no false-positives, but produced false-negatives due to
interferents in drinking water. Consistency was 100 percent. Sensitivity was 20 mg/L, higher than
the 0.4 mg/L detection level indicated by the vendor. Cost is $250 for a box of 10 strips. Strips are
portable and easy to use; although the indicator line color was very faint sometimes, increasing the
risk for false-negatives. It takes 15 minutes for an indicator line to appear. Sample throughput was
20 to 30 samples per hour (EPA-ETV, 2004).
ELISA by Tetracore
Anthrax ELISA produced false-positives due to water with humic and fulvic acid in it and no false-
negatives were produced. However, the ELISA was not able to detect anthrax at the vendor-
August2005 141
-------
U.S. EPA Office of Water Early Warning Systems
indicated level of 2x104, only at concentrations 100 times greater. Botulinum toxin ELISA had no
false-positives, but false-negatives were produced when humic and fulvic acid were present in the
water sample. The lowest detectable concentration for Type A was 0.02 mg/L, but it was not clear
for Type B. The ricin ELISA showed no false-positives or negatives and could detect ricin to a
concentration of 0.0075 mg/L, slightly higher than vendor indicated level. ELISA was easily
portable, but not easily operated by an untrained user. The cost for one Tetracore ELISA (96 well
plate) is $400 (EPA-ETV, 2004).
9.5.1 Issues and Gaps
The following section highlights various issues and gaps in using microbial detectors for early
warning for finished drinking water.
• Contaminants Need to be Concentrated to be Detected
Many microbial pathogens are a risk to human health at low concentrations. Detecting low
concentrations is difficult for the binding assays that most capture-target technologies utilize. Thus,
it may be necessary to concentrate large volumes of water samples to gather enough contaminant
to detect. Two AwwaRF research papers expected in 2005 (Extraction Methods for Early/Real-
Time Warning Systems for Biological Agents - Project A and B) may help with this issue (see
Appendix D).
• Concentration Techniques can also Concentrate Interference Compounds
Techniques that concentrate the target contaminant will very often also concentrate other non-target
contaminants and interferants that will interfere with the biosensor assay. For example, to
successfully perform PCR analysis of environmental waters, it is necessary to remove analytical
interferences such as humic and fulvic acids.
• Antibodies Cross-React and are Subject to Binding Kinetics
Antibodies may bind with known or unknown affinity to non-target antigens. The sensitivity of
antibodies needs to be calibrated for each batch, even for monoclonals. Cross-reactivity is a problem
if different target molecules have overlapping epitopes. Antibodies designed against specific unique
epitopes would have less cross reactivity. Even the best antibody will only have detectable binding
if the target antigen is sufficiently concentrated; therefore, it is necessary to concentrate drinking
water samples before testing.
• Emerging and Bioengineered Microbes Could Escape Detection
Even with a broad selection of capture molecules, microbes will evolve such that they may lose
target epitopes or DNA, and therefore escape detection. Bioengineered pathogens could be designed
to evade detectors, if detailed information on capture molecules was obtained. The only technology
that minimizes this problem is Triangulation Identification Genetic Evaluation of Risks (TIGER),
developed by Isis Pharmaceuticals, Inc., because it integrates a variety of approaches (DNA base
composition and PCR) into one analysis.
August 2005 142
-------
U.S. EPA Office of Water Early Warning Systems
• Reagents May Not be Stable in Field Environmental Conditions
Reagents containing biological molecules usually degrade or become inactive within hours of being
subjected to room temperature conditions. Even if partial activity is maintained, quality control is
jeopardized. This problem can partially be addressed with freeze-dried reagents, but reagent-grade
water at the field site is required to reconstitute the reaction solutions. Molecular stability of
biomolecules on biochips is even more difficult.
• Most Microbial Detector Technologies are for Grab Samples, Not Online
With the exception of Micro-Flow Imaging, there are no commercial online technologies for
detecting microbes. Two light scattering technologies (BioSentry and MALLS) are online and are
undergoing beta testing, but are not yet on the market.
Exhibit 9-7 provides a snapshot of capabilities, issues, and gaps for microbial sensor technologies
and techniques.
9.5.2 Conclusion and Recommendations
Development of an online microbial technique appears years away. Light scattering methods show
some promise, but most methods are not suited for continuous online monitoring or differentiating
among microbes. However, there are several potentially adaptable methods, including
immunoassay, PCR, and ATP, that could be used for confirmation testing of grab samples. For
drinking water, the challenge that remains with most methods is the issue of concentrating the
sample. A few methods of concentration show promise including the hollow fiber, the micropump,
and the PNNL BEADS technology. Generally, this does not seem like an insurmountable obstacle
for some methods. However, for PCR, current technology and concentration methods still do not
have the adequate level of detection. Overall, none of the methods met all requirements for rapid
detection technology. Therefore, the recommended approach was to screen a sample with a generic
detector (e.g., multi-parameter probe or perhaps light scatter), then use an immunoassay device in
tandem with another method for identification. Detectors based on ATP are promising, but have not
been verified for water (one is online for water). An ETV Program testing ATP products is
recommended. In the future, microchips have great potential in online measurement, but presently
the field is not sufficiently mature.
August 2005 143
-------
U.S. EPA Office of Water
Early Warning Systems
Exhibit 9-7. Microbial Sensor Technologies and Techniques
Method
Capabilities
Issues and Gaps
Strip tests
(immunoassay)
• detects 1-4 antigens in one test
• portable
• seconds to complete test
• EPA conducting research on use
for finished water
sample concentration needed
lack of necessary sensitivity
Fiber optic-based
biosensor
can detect specific antigens
portable
high potential to be online
emerging for water
Micro spheres
can detect specific antigens
• emerging for water
• not in portable format
Flow cytometry and
micro-flow imaging
potentially identify specific
microbes
can quantify
portable
sample concentration may be
needed
ATP
detects cellular components
portable (kits)
chlorine residual is not an issue
not independently validated for
water
cannot provide microbe specific
information
PCR
detects specific DNA sequences
portable (suitcase and hand-held)
not designed for water
sample concentration required
current level of technology has
limits of detection that are still not
adequate
Light scattering
• detects cells, oocysts, spores,
possibly species specific
• potential for online
• does not require concentration
techniques
• not widely used for pathogen
detection
• not independently validated for
water
• may be prone to false-positives
• cannot detect viruses
Microchips and
microarrays
potential to identify many
specific pathogens
small size is good for minaturized
devices
portable versions are emerging
technology
sample concentration probably
required
9.6 Evaluation of Radiological Sensors
With the possibility of accidental radioactive spills and emerging terrorist threats, it is important that
water systems have the capability to detect a surge in radioactivity as soon as it occurs.
Furthermore, identifying the type of radiation and its source will greatly contribute to a rapid
response and recovery effort. Continuous, online monitoring in real-time will allow for rapid
detection of intentional or accidental contamination of the water system. The use of an alarm system
will help alert the appropriate operators. For intentional contamination of distribution systems,
radionuclides that have a high specific activity (Curies/gram), deliver a relatively high gamma and
August 2005
144
-------
U.S. EPA Office of Water Early Warning Systems
maybe alpha/beta dose, and can be obtained and solubilized in high concentrations are the most
likely contaminants of concern.
The devices described below are the instruments used for measuring alpha, beta, and gamma
radiation in liquid or water. New technology that allows detection of radiation at lower levels has
been emerging. In-field, gamma radiation detectors are more common than detectors for alpha and
beta radiation, which have properties that make them difficult to detect. There were very few
verification studies of radiation detectors for use in drinking water. They are provided below.
Exhibit 9-8 summarizes how the radiological detectors (those available now or potentially adaptable)
match up against the desired characteristics of radiation sensors for use in an EWS.
Isco 3710 RLS Samplers
A study by Westinghouse Savannah River Company, using Isco 3710 RLS samplers, showed that
"Data collected over a four-month field test period compares very favorably with concurrent
laboratory-based analyses and historical data, at a reduction in cost and significant time savings."208
Thermo Alpha Monitor
Oak Ridge National Laboratory has tested this instrument and demonstrated its capabilities on water
under 1 picocurie/L, and has analyzed isotopic U levels of 10 ppt natural U (15 femtocuries/L), as
well as 20 ppb natural U (30 pCi/L) in under 30 minutes. This detector is still under development
and requires further testing of durability and accuracy as well as peer reviews and EPA approval.
9.6.1 Issues and Gaps
The following section highlights various issues and gaps in using radiation detectors for early
warning for finished drinking water.
• Devices and Results Show Variability
The appropriate devices and methods will vary according to local conditions such as temperature
and humidity, or the properties of the radionuclide at the source of the radiation.
• Require Special Expertise
All of the devices mentioned in this chapter usually require specialized expertise for installation,
setup, and routine calibration, even if they are labeled maintenance-free.
• Costly for Online Monitoring
Although online analyzers are efficient in monitoring water quality, they are expensive and limited
in number. Many facilities may find grab samplers to be more appropriate. Manufacturers are in
the process of developing and refining other applications of small-scale flow-through scintillation
technology. Utilities will need to collaborate with manufacturers to customize the monitors for
small-scale needs.
August 2005 145
-------
bo
c
•&"
a
u
:_
O
GC
I
•d
'—
a
o
5B
A
S
o
O
IK
O
a
&£
_©
]o
•3
oo
d\
+j
z
IS
G
O
U
0>
"*"
£? 'oi
a A
OJ CS
s!
u <<
^
•™
s
'^
a
0>
C/3
£? S
ca —
a H
•-
o
^ !^
2 S
O
o
u
a S
— «
a t;
O o
PH
an
<+4 fl
= a
WD a
g 5
CS ^^
o
U
Description
Product
§
§
cd
"S3
cd
3
_g
'S ° *g
o ^
8 2 'S
.a
V
0
|
tu 60
0
o
^-~l.
oo
6ft
1>
.3
"3
o
„ »
+3 ^
U ^H
•" ca
S 6
tl
continuous flow
through
scintillation
detector
Technical
Associates SSS
33-5FT
§
§
1
'8
ca
a
0
JH
JS
'8 ^
H
.a
V
0
|
sa
0
o
<~-^
(N
6«
1>
.a
"3
o
i
1
60
scintillation
detector
Technical
Associates
MEDA 5T
§
§
1
ca
j
g
11
a
1
V
0
|
S
^
o
|
•fi 0
e 3
o a
If
no scintillation
detector
i t/j
Technical
Associates SSS
33DHC and SS
33DHC -4
§
§
1
'8
ca
j
2 g
- 1
d O
.a
V
0
g
1> ^60
0
O
^\
&
1>
.a
"3
o
1
"S
^
0
'3
o
no scintillation
detector
Technical
Associates SSS
33M8
§
§
'8
ca
a
0
JH
JS
'8 ^
ll
a
1
V
0
|
.a
^
o
=8
1
p" ° n
>^ rC ' — i
"o m E
•o a i
H iS 60
T3
|
"ft
O
O
(N
1
ca
a
0
^
JS
'8 ^
11
1
V
_1>
o 33
1 ^
P >
C ca
ll
< — I
"S ^
^*~i ' — '
O >
1>
M
fH
o
1
1
60
details not yet
available
GammaShark™
§
§
>>~>
O 'O
& ca
p
0
4n
!§
'8 ^
ll
.a
V
0
g
tU 60
3 a
o o
o o
0 0
o" o"
O "n
6ft 6ft
1>
.a
"3
o
3
1
60
r-rt
i
ca ^
a
o
monitors radiati
in liquid from
outside of pipe
to
Canberra
LEMS600 Serii
Liquid Effluent
Monitoring
§
§
>>~>
^H 0>
11
ll
0
4n
!§
'8 ^
11
.a
V
0
|
sa
o o
<^, ^^,
& &
1>
.a
"3
o
£
ca
^H
ca
ca
60
P
0
monitors radiati
in liquid from
outside of pipe
»
Canberra
OLMIOO Onlin
Liquid
Monitoring
§
§
>>~>
^ Io
11
O 'O
& ca
p
0
JH
ca
'8 ^
Bl
§ %
1
V
0
|
.a
"3
o
>,
ca
ca
60
a
monitors radiati
in liquid from
within pipe
Canberra ILM
100
-------
U.S. EPA Office of Water
Early Warning Systems
• Few Detectors are Available or Verified
There are only a few detectors that are designed to monitor for alpha and beta radioactivity. There
are also only a few detectors that are designed for online monitoring of gamma radioactivity.
Verification studies have not been extensively performed on radiological detectors for finished
drinking water.
• Not Geared to Distribution Systems
Some of these monitors are intended for waste streams rather than distribution systems due to the
more stringent requirements for drinking water monitors. Waste stream monitors would be geared
more towards the detection of accidental spills rather than intentional contamination.
Exhibit 9-9 provides a snapshot of capabilities, issues, and gaps for radiation sensor technologies.
Exhibit 9-9. Radiation Sensor Technologies
Radiation Detected
Alpha
Beta
Gamma (liquid scintillation)
Capabilities
• none
• continuous and real-time
measurements
• continuous and real-time
measurements
• has required sensitivity in
water
Issues and Gaps
• difficult to detect in water
• difficult to detect in water
• typically designed for wastewater
and groundwater
• not verified for drinking water
• typically developed for wastewater,
not drinking water
• not verified for drinking water
• requires special expertise for setup,
operation and maintenance
• costly for online monitoring
9.6.2 Conclusions and Recommendations
There are demonstrated technologies for examining radiation in wastewater, but the transfer or
adaptation to drinking water has not taken place. Only a few products claim applicability to water,
some on a grab-sample basis. More products are being developed by a few vendors, but it is still
unclear whether the threat merits use of these expensive products on a real-time basis. The few
items that are commercially available should be verified either by EPA or by a national laboratory
that specializes in radiation. Grab sampling can be performed by a number of products, but it is
unclear if there is any generic monitor that can trigger this more detailed analysis. Thus, early
warning for radiation detection is not currently available, and the market forces for such
development may not be strong.
August 2005
147
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
10. Conclusions and Recommendations
The following are conclusions and recommendations from this state-of-the-art review of
technologies and techniques for EWSs. General conclusions and recommendations are provided
first. These are followed by specific conclusions and recommendations organized by features of the
EWS: data acquisition and analysis; flow modeling; sensor placement; alarm management; decision
making and response; multi-parameter water quality technologies; and detection of chemical,
microbial, and radiological contaminants. The recommendations include a list of near- and long-
term knowledge and research gaps.
10.1 General Conclusions and Recommendations
Viable integrated EWSs that meet the desired characteristics and can be routinely used are several
years away. Some individual components are available currently, however, others need further
development. Designs of EWSs for water distribution systems are largely theoretical or in
preliminary stages. Much of the required data acquisition software and hardware already exists, but
software for security of EWS SCADA systems is still under development and needs verification.
Distribution system modeling in general, and contaminant flow predictive systems in particular, are
developing rapidly with the incorporation of graphic software; however, most utilities have not
implemented software for modeling intentional contamination events. Most sensor and EWS
components have not been tested or verified, and the types of contaminants and levels of exposure
have not been well defined to support selection of sensor technologies. Several companies are
working on alert management methods but are in preliminary stages of research, with often
proprietary trigger algorithms. Linking contamination data analysis with decision making and alert
response has been outlined; however, the equipment to effectively implement the process has not
been extensively developed for water utilities. Research on approaches and technologies to detect
engineered microbes is needed. Also, all of these technologies should be verified and affordable,
and should operate consistently in the field.
Short-Term Research Needs
• An in-depth review of EWS architecture and implementation should be conducted.
Because a detailed examination of the basic design and features of an EWS architecture is
beyond the scope of this study, follow-up studies to prioritize certain features and provide
comprehensive guidance to implement the selection, linking, and testing of various EWS
components are recommended. Any guidance on design should include considerations for small
and large systems, public health surveillance, and consumer complaint monitoring. Also, the
research should provide guidance on EWS performance, alarm, and response criteria.
• Methods for vulnerability assessments should be adapted to focus on contamination scenarios.
There are various vulnerability assessment methods that utilities have used. To develop an
EWS, vulnerability to contamination has to be examined. It is unclear whether utilities have
specifically examined their vulnerabilities for contamination or whether the existing
methodologies can adequately assess contamination vulnerabilities. This is an area of further
research to see how adequately utilities have examined contamination vulnerability and how
such information can be incorporated into EWS architecture.
August 2005 149
-------
U.S. EPA Office of Water Early Warning Systems
• Research on international efforts on EWS is needed.
EPA may find it useful to continue to engage the international research and development
community to gain access to innovations and advancements being made in other countries.
• Grab-sampling protocols and analysis technologies should be developed quickly.
Online detectors, as part of an integrated EWS, may be years away. However, near real-time
monitoring can be approached by establishing periodic grab samples that can be analyzed as
appropriate by field or laboratory instruments.
Long-Term Research Needs
• Survey case study and analysis should be performed on monitors/sensors/detectors used by
utilities.
Some water utilities have installed or tested EWS component technology, such as monitors,
sensors, or detectors, in the field but the experiences are not extensive or shared. Experience
gained from using these technologies in a distribution system, even if they are not part of a fully
integrated EWS, as defined by this report, should be captured. A survey of such case studies can
provide great insights into the validated use and capabilities/disadvantages of certain
technologies. Further research efforts should be undertaken to more fully document and
evaluate the case study experience of utilities using EWS components in the field.
• EWS technologies and techniques need to undergo verification testing.
The ETV or TTEP program should test various sensor technologies, including for example, ATP
products, mussel biodetectors, and radiological online detectors.
• Potential contaminants list should be reviewed on an ongoing basis.
The types of contaminants of concern for public health are important to evaluating EWS
adequacy. Determining the specific contaminants is an area of ongoing work that is not publicly
available and is generally under classified review. There will be further such classified efforts.
There is a research need to continue to determine the types of contaminants that are harmful and
that need to be detected by EWSs.
• Concentrations that must be detected by a sensor should be reviewed on an ongoing basis.
Another recognized area of research is to determine the concentration of contaminants that need
to be detected by EWSs. The concentration depends on the properties of the contaminant, the
routes and magnitude of exposure, and the susceptibility of the exposed population to particular
contaminants. Without a determination of these concentrations, the adequacy of any specific
instrument or set of instruments in reducing public health risk is difficult to determine.
Continued research is recommended to fully determine concentrations that must be detected.
Related to concentrations, research should also be conducted to determine the anticipated
exposures and doses based on specific concentrations that impact human health, as well as to
determine if decontamination efforts were sufficient to protect human health.
• The fate and transport (including exposure levels, doses, and detectible concentrations) of
contaminants, especially toxic byproducts, should be examined.
As research occurs on the contaminants of concern, additional research should be conducted to
determine the toxic byproducts of certain contaminants and to model the fate and transport of
these contaminants in real world water distribution systems. The persistence, stability,
August 2005 150
-------
U.S. EPA Office of Water Early Warning Systems
resistance to chlorine, and ease of dispersion of a contaminant in an aqueous environment are
important fate and transport factors. This research area will assist with the evaluation and
selection of EWS technologies.
• Results from laboratory research by various agencies on EWSs should be replicated and the
conclusions of EWS studies should be shared among government agencies and water utility
stakeholders.
There are various laboratory research efforts on EWSs in distribution systems that are sponsored
or conducted by the EPA WATERS Center, U.S. Army ECBC facility, USGS facilities, and
specific utilities. These efforts should be replicated when possible to verify results. The
information from these efforts and other EWS studies should be shared among government
agencies and water utility stakeholders.
10.2 Specific Conclusions and Recommendations
10.2.1 Data Acquisition and Analysis
Data collection by Supervisory Control and Data Acquisition (SCADA) or other automated systems
is essential to handle the large volume of data from online sensors in an EWS. Existing data
acquisition systems do not present a major issue given currently recommended sampling times by
EPA (between 2 and 10 minutes, depending on the SCADA system setup and bandwidth, the sensor
location, and the water flow rate). Because of the large amount of data generated, automated data
validation processes are indispensable to ensure accurate results from data analysis. Data
transmission to a central database, through hardwired or wireless systems, requires a simple yet
effective protocol to ensure accuracy and completeness, and can be done by comparing data received
from monitoring sites with data stored at the sensor locations. Much of the data acquisition software
and hardware already exists. Software for security of SCADA systems for EWSs is still under
development and needs verification, but can probably be addressed by utilities simultaneously with
general security issues (e.g., encryption).
Short-Term Research Need
• Standardized methods and guidance for data analysis and interpretation are needed.
Specifically, further research and development in mining software to identify outlier detection
is needed. Verification programs are needed to substantiate the data analysis algorithms. Some
of the efforts by ASCE will help to guide utilities in the use of such systems (ASCE, 2004).
Long-Term Research Needs
• Large-scale data storage and manipulation techniques and technologies are needed.
For continuous real-time monitors, data are generated on a large scale. SCADA is the likely
vehicle to collect the data. New ways of storing and manipulating the information for immediate
and long-term uses are needed.
• SCADA data security programs should be developed to link existing utility efforts with security
characteristics inherent in EWS.
Current remote monitoring products are beginning to incorporate security precautions including
encryption. However, demonstration projects are not frequently undertaken. Standardization
August 2005 151
-------
U.S. EPA Office of Water Early Warning Systems
from other data security efforts could be applied to the water sector. Programs should be
developed to link utility security efforts surrounding SCADA with data security geared to EWSs.
10.2.2 Flow Modeling
Predicting the movement and flow of contaminants in a water distribution system EWS is important
not only to prepare for a potential contamination event, but also to improve the effectiveness of the
monitoring system. Distribution system modeling in general, and contaminant flow predictive
systems in particular, are developing rapidly. Current contaminant flow models can also integrate
data from geographic information systems and display results using computer-aided drafting (CAD)
software. PipelineNet, an integration of EPANET and Arc View GIS, and commercial integration
packages such as WaterGEMS and Info Water provide direct capabilities to assess impacts of
contamination events. Calibration using optimization methods or tracer studies has been
increasingly used in distribution systems. Utilities are not generally using models specifically for
the purpose of modeling intentional contamination events. Consumption-of-water models are also
being occasionally incorporated. Utility efforts to validate and develop predictive flow models
would meet the dual purposes of general planning (expansion, upgrades, repairs, maintenance) and
testing intentional contamination scenarios. There are currently few established calibration criteria
in the U.S., although a committee of the AWWA did propose a set of possible calibration guidelines
in 1999 and the EPA is preparing a distribution system handbook which will include a section on
hydraulic model calibration and validation. These possible calibration guidelines should serve as
a catalyst or starting point to move forward on developing accepted calibration guidelines or
standards. Incident Commanders Water Modeling Tool (ICWater) extends the capability of a
previously developed RiverSpill modeling tool to allow an incident commander to analyze and react
quickly to chemical and/or biological contaminants introduced into surface water sources. EPA's
TEVA program incorporates a probabilistic framework for a large range of contamination attacks
in assessing vulnerabilities and estimating the most appropriate sensor placements.
Short-Term Research Need
• Improved contaminant flow models are needed.
Research is recommended to develop improved models and methods for validating the models
in specific applications. The models should be improved to better include the effect of chemical
fate and byproducts.
Long-Term Research Need
• Flow models need to be verified and then used to improve EWS design.
Projects are needed to verify contaminant flow models and to make the tools useful for sensor
placement, real-time contaminant flow prediction, and identifying the likely location of the
contaminant source. These models need adaptation for ease of use by various sized utilities.
There are no established calibration criteria for contaminant flow models in the U.S., although
a committee of the AWWA did propose a set of possible calibration guidelines (ECAC, 1999).
However these guidelines have not been officially accepted and there is no active process
underway to adopt them. Using these possible calibration guidelines as a catalyst or starting
point, it would be recommended to move forward on developing accepted calibration guidelines
or standards.
August 2005 152
-------
U.S. EPA Office of Water Early Warning Systems
10.2.3 Sensor Placement
Because of both budgetary and technical constraints, utilities can make only a modest initial
investment in sensors within their distribution system, and therefore the utilities want to determine
the most appropriate locations. Without resorting to the sophisticated experimental optimization
techniques, a two-stage procedure is typically followed. In the first stage, likely sites for sensors
are determined based on technological (e.g., available power source, access to communications) and
physical (e.g., access) constraints. In the second stage, sensors are placed throughout the system on
larger pipes that serve the most customers. Current research combining flow models and sensor
technology is beginning to be developed, but such models must be verified before difficult and
costly decisions are made by the utilities.
Short-Term Research Need
• Hardware and materials to protect remote sensors are needed.
Online sensors are typically installed using special sample taps that require interruption of water
flow through pipes. New installation techniques are being developed to enable installation
without interrupting the water flow or major excavation. Sensors also need to be able to
withstand the harsh environment of different locations. New development in construction of
materials and protective hardware is necessary to enable sensor systems to be installed in areas
open to the environment (AwwaRF, 2002).
Long-Term Research Need
• Research into sensor placement parameters is recommended.
Simple guidance, such as what to do with a limited number of specific sensors, is needed. Also,
validation of sensor placement strategies should be investigated by perhaps comparing results
of several models that determine sensor placement and optimization.
10.2.4 Alert Management
Alert management systems typically consist of two general areas: (1) establishing parameters for
alert triggers and (2) reducing false alarms. Any anomalies in the comparison of sensor data to the
baseline trigger alerts to the operator. Establishing reliable baseline data is key, especially when
water quality fluctuates. Alert management systems usually rely on strict data validation protocols
or specialized software to reduce false alarms. Several companies are working on alert management,
but are in preliminary stages of research, with often proprietary trigger algorithms.
Long-Term Research Need
• Alert management approaches/technologies should be examined and sensitivity to false-positives
and false-negatives should be quantified.
A demonstration project is needed to ensure the reasonableness of certain approaches to alert
management. The relationship between alert sensitivity and the potential for adverse
consequences (false-positives and false-negatives) should be quantified. Additional projects
should examine alerts for other promising sensors such as mussel or bacteria monitors.
August 2005 153
-------
U.S. EPA Office of Water Early Warning Systems
10.2.5 Decision Making an d Response
The process linking the analysis of contamination data with the decision making and response is
outlined in EPA's Response Protocol Toolbox; however, additional tools to effectively implement
the process are needed for water utilities.
Long-Term Research Need
• Technology to support and implement decision making and response are needed.
Tools to assist decision making and response, such as the Water Contaminant Information Tool,
are being developed and will help fill a current void.
10.2.6 Multi-Parameter Water Quality Technologies
There is an active research program in the use of the multi-parameter water quality monitors as part
of an EWS for distribution systems. Preliminary evidence suggests that such monitors can detect
an anomaly in the distribution system and provide an initial red flag warning. However, it is also
reasonable to be concerned about false-positives and whether the system can provide definitive
indications of intentional contamination. Gathering baseline data may also be prohibitively
expensive. Currently, these technologies need to be demonstrated to detect biological contaminants
or dangerous chemicals or develop a track record of field performance. The technology has not been
sufficiently evaluated to recommend widespread use; for example, no tests have been performed on
chloraminated systems. However, full-scale testing by USGS, EPA, and a water utility during 2006
and 2007 may help shed light on concern over false-positives and whether a system can work with
the fluctuations of normal water quality.
Given the current developmental stage of multi-parameter technologies, parameters that appear from
EPA's preliminary tests to be useful to monitor in distribution systems include chloride (ISE),
specific conductance (electrode), turbidity (nephelometric), free chlorine, and ORP. TOC is
extremely helpful, but may be too expensive to be widely used. Probes are being developed by
manufacturers to include free and total chlorine, pH, temperature, specific conductance, chloride,
nitrates, turbidity, and ORP. Not all utility companies are able to afford the monitoring systems
currently being evaluated as EWSs. Reductions in price due to competition and technological
advancements may remedy this situation in the future, but in the interim, utilities with limited
financial resources will face a challenge in implementing online (e.g., continuous) water quality
monitoring.
The signatures being developed by Hach Company and others to identify contaminants or classes
are difficult to independently validate or be understood because their methods and algorithms have
not been made public. Also, the examination of water quality parameters in detecting and
identifying contaminants is still being evaluated by EPA, USGS, the Army, and other organizations.
There has yet to be a field-scale test of an EWS with these water quality parameter components.
These are reasons for caution in recommending the use of water quality parameter-based EWSs.
Short-Term Research Needs
• Verified baseline data to calibrate EWS alarm triggers are needed.
Although research has demonstrated proof-of-concept for using water quality parameter
fluctuations as a signal that a contamination event has taken place, baseline data to calibrate the
August 2005 154
-------
U.S. EPA Office of Water Early Warning Systems
alarm triggers may need to be gathered over months or years for each distribution system.
Because this is very expensive, research demonstration projects such as the USGS project are
needed to advance the understanding of how baseline water quality data might influence the
performance of EWSs in the field.
• Contaminant specific signatures are needed.
Research has begun to develop contaminant specific signatures; however, the limited number
of contaminants examined to date lowers the confidence of the claim that the signature really is
unique. Classes of contaminants may be identifiable, but whether a comfortable range of
specific contaminants can be identified has not yet been determined. Additionally, it is not
known yet whether monitors can detect biological contaminants and dangerous chemicals.
Further research is needed to develop specific signatures for a wider range of contaminants and
concentrations, including actual agents.
• Validation of event detection algorithms is needed.
It is important that the algorithms used to determine when alarm conditions are encountered are
validated for a variety of real world operational conditions.
Long-Term Research Need
• Costs and benefits (e.g., ability to detect contaminants with multi-parameter water quality
monitors) of using TOC sensors should be determined and more affordable and reliable TOC
sensors should be developed.
Current multi-parameter units cost approximately $10,000 without TOC. Although TOC
appears to be a valuable parameter to measure, it adds $18,000 to 29,000 per unit. Research
needs to determine the cost/benefit relationship of including this technology in an EWS. A basic
system with 10 microprobe monitors (without TOC) linked to an existing SCADA system is
estimated to cost approximately $150,000, plus $60,000 per year in operational costs. Also, a
research/development need is to have cheaper and reliable online TOC monitors. A TOC
monitor for an EWS would not need to be capable of compliance monitoring but would only
need to detect gross changes in TOC levels that warrant further investigation.
10.2.7 Detection of Chemical Contaminants
Portable field technology is available for conducting analyses of grab samples on site for many
possible chemical contaminants. This area will continue to improve as high technology equipment
evolves based on microchip technology (e.g., taste-chip). The technology that is readily available
and reliable uses specific probes, such as arsenic and cyanide, that could be effective against a
narrow selection of contaminants. Certain biomonitors are portable and can be used for site
assessments. Chlorine in the water would have to be removed for many such analyses. In contrast
to portable field technology, online chemical detection technologies are not reasonably available or
are not cost-effective. GC and ion mobility have cost and technical challenges. Certain biomonitors
may be promising, if the interference issue of chlorine and chloramine residual can be resolved. In
one case, the mussel monitor has been recently demonstrated in finished water in Europe. There are
various efforts in the U.S. to make other monitors (MicroTox® and ToxScreen) adaptable to finished
water. The MosselMonitor® or Bio-Sensor® could be a good candidates for an EPA-ETV Program
study and perhaps used for a laboratory and/or field study on finished water or with CBR surrogates
or contaminants. In the next three years, the field should show further development in terms of cost-
August2005 155
-------
U.S. EPA Office of Water Early Warning Systems
effective and reliable devices. A few new technologies (such as microchips) could revolutionize the
chemical detection field for drinking water.
Short-Term Research Need
• The impact and removal of chlorine and other residues on accuracy of detection should be
examined.
Organism-based biomonitors are sensitive to the chlorine residuals in drinking water. Although
the fish-based Bio-Sensor® and MosselMonitor® remove chlorine, broadly applicable methods
for chlorine removal have not yet been developed. Although one company, Checklight, is
developing a system for removal of chlorine, the effect of chlorine removal on other biosensors
has not been demonstrated.
Long-Term Research Needs
• Reliable field kits should be developed.
The bacterial monitoring kits tested by EPA's ETV Program have high false-positive and false-
negative rates. A common drawback for such kits is the stability of reagents. Often reagents
require reconstitution (if they are lyophilized) or careful measurement of reaction components
to constitute a fresh reaction mixture. Kits can be subject to variability due to different users
because it is difficult for users to mix and pipette in a consistent manner. They require trained
personnel and have set-up requirements, such as culturing log phase growth bacteria. The results
do not provide specific identification of a toxin. Although these kits may be suitable for
confirming the presence of a toxin, further methods would need to be utilized for specific
identification.
• Existing state-of-the-art detection technologies should be adapted for use in EWSs.
Portable infrared spectroscopy, ion mobility spectroscopy, surface acoustic waves, and polymer
composite chemoresistors technologies are being aggressively pursued for air and vapor
applications, but have not been developed specifically for drinking water monitoring. Research
could examine whether these techniques could be adapted for water with further research if there
is a potential market.
10.2.8 Detection of Microbial Contaminants
Development of an online microbial technology appears years away. An exception to this may be
the light-scattering methodology, which shows some promise. Most methods are not suited for
continuous online monitoring or for differentiating among microbes. However, to confirm a
contaminant using a portable field unit with grab samples, there are several potentially adaptable
methods from which to choose, including immunoassay, PCR, and ATP. These methods have not
yet been exploited to their full potential, so will likely continue to be incorporated into new
monitoring devices and systems. Grab sampling could include sampling at scheduled intervals or
taking composite samples (e.g., collecting small volumes of sample continuously over time). In any
sampling, the microbial integrity must be ensured. For drinking water, the challenge for most
methods is the need to concentrate the sample. A few methods of concentration show promise
including the hollow fiber, the micropump, and related efforts by Pacific Northwest National
Laboratory. Generally, this does not seem like an insurmountable obstacle for some methods.
However, for PCR, current online technologies and concentration methods are not sufficiently
advanced to detect microbes at levels that threaten public health. Overall, none of the methods by
August 2005 156
-------
U.S. EPA Office of Water Early Warning Systems
themselves provide sufficiently rapid detection. Therefore, one recommended approach is to screen
a sample with a generic detector (e.g., multi-parameter probe or perhaps light scatter), then use an
immunoassay device in tandem with another method for identification. Detectors based on ATP are
promising, but have not been verified for water (one is online for water). ATP detection products
should be third party evaluated for EWS application. In the future, microchips have great potential
in online measurement, but presently the field is not sufficiently mature.
Short-Term Research Needs
• Extraction and concentration technologies need improvement.
Many microbial pathogens are a risk to human health at low concentrations. The concentration
of pathogens that needs to be detected to adequately protect public health is a policy issue.
However, it is widely acknowledged that current detection methods for online or near continuous
use are not sensitive enough to detect the lowest microbial concentrations that may be a concern
for human health. Therefore, microbial contaminants need to be sufficiently concentrated to be
detected. Detecting low concentrations is difficult for the binding assays that most capture-
target technologies utilize. Thus, it may be necessary to concentrate large volumes of water
samples to gather enough contaminant to detect. Two AwwaRF research papers expected in
2005 (Extraction Methods for Early/Real-Time Warning Systems for Biological Agents - Project
A and B) and research at Idaho National Laboratory in Idaho Falls, ID may help with this issue.
• Methods to distinguish concentrated interferants from target compounds are needed.
Concentration techniques could also concentrate interference compounds. Techniques that
concentrate the target contaminant will very often also concentrate other non-target
contaminants that will interfere with the biosensor assay. For example, to successfully perform
PCR analysis of environmental waters it is necessary to remove analytical interferences such as
humic and fulvic acids.
• The development of field-stable reagents is needed.
Reagents may not be stable in field environmental conditions. Reagents containing biological
molecules usually degrade or become inactive within hours of being subjected to room
temperature conditions. Even if partial activity is maintained, quality control is jeopardized.
This problem can be partially addressed with freeze-dried reagents, but reagent-grade water at
the field site is required to reconstitute the reaction solutions. Molecular stability of
biomolecules on biochips is even more difficult.
Long-Term Research Needs
• Antibodies for unique epitopes that would show less cross-reactivity should be developed.
Antibodies cross-react and are subject to binding kinetics. Antibodies may bind with known or
unknown affinity to non-target antigens. The sensitivity of antibodies needs to be calibrated for
each batch, even for monoclonals. Cross-reactivity is a problem if different target molecules
have overlapping epitopes. Antibodies designed against specific unique epitopes would have
less cross-reactivity.
• Research on additional approaches and technologies that can detect emerging, evolving, and
engineered microbes is needed.
August 2005 157
-------
U.S. EPA Office of Water Early Warning Systems
Emerging and bioengineered microbes could escape detection. Even with a broad selection of
capture molecules, microbes will evolve such that they may lose target epitopes or DNA, and
therefore escape detection. Bioengineered pathogens could be designed to evade detectors, if
detailed information on capture molecules was obtained. The only existing technology that
minimizes this problem is Triangulation Identification Genetic Evaluation of Risks (TIGER),
developed by Isis Pharmaceuticals, Inc., because it integrates a variety of approaches (DNA base
composition and PCR) into one analysis.
10.2.9 Detection of Radiological Contaminants
There are demonstrated technologies for examining radiation in wastewater, but the transfer or
adaptation to drinking water has not yet taken place. Only a few products claim applicability to
water, some on a grab-sample basis. Also, the appropriate devices and methods will vary according
to local conditions, such as temperature and humidity, or the properties of the radionuclide at the
source of the radiation. More products are being developed by a few vendors, but it is still unclear
whether the threat merits use of these expensive products on a real-time basis. The few items that
are commercially available should be verified either by EPA or by a national laboratory that
specializes in radiation. Grab sampling can be performed by a number of products, but it is unclear
if there is any generic monitor that can trigger this more detailed analysis. In addition, all of the
radiological monitoring devices mentioned in this study usually require specialized expertise for
installation, set-up, and routine calibration, even if they are labeled maintenance-free. Thus, early
warning for radiation detection is not currently available, and the market forces for such
development may not be strong.
Short-Term Research Need
• Beta and gamma radioactivity detectors should be developed and verified for drinking water
monitoring applications.
There are only a few detectors that are designed to monitor for alpha and beta radioactivity.
There are also only a few detectors that are designed for online monitoring of gamma.
Verification studies do not seem to have been performed on detectors that are geared to radiation
terrorist attacks on drinking water.
Long-Term Research Needs
• Low-cost, online, radioactivity monitors are needed.
Although online analyzers are efficient in monitoring the water quality, they are expensive and
limited in number. Many facilities may therefore find grab samplers to be more appropriate.
Manufacturers are in the process of developing and refining other applications of small scale
flow-through scintillation technology. Utilities will need to collaborate with these manufacturers
to research and customize the monitors for small scale needs.
• Monitors should be developed that are specifically intended for water distribution systems.
Some of these monitors are intended for waste streams rather than distribution systems due to
the more stringent requirements for drinking water monitors. Waste stream monitors would be
geared more towards the detection of accidental spills rather than intentional contamination.
August 2005 158
-------
U.S. EPA Office of Water Early Warning Systems
REFERENCES
ACS (2002) American Chemical Society. "PNNL Research: Biodetection Enabling Analyte
Delivery System." Chemical and Engineering News. 80(20):32-36.
ASCE (2004) Interim Guidelines for Designing an Online Contamination Monitoring System,
American Society of Civil Engineers.
Alberts, B; Bray, D; Lewis, J; Raff, M; Roberts, K; Watson, JD. (1994) Molecular Biology of the
Cell. Third Edition Garland Publishing Inc. NewYork.
AWWA Workshop (2004) Contamination Monitoring Technologies sponsored by the American
Water Works Association, Richmond, VA; May 2004.
AwwaRF (2002) Online Monitoring for Drinking Water Utilities (Project #2545) Order number
90829. Report prepared by: E. Hargesheimer, O. Conio, and J. Popovicova.
Bahadur, R; Samuels, W; Pickus, J. (2003a) Case Study for a Distribution System Emergency
Response Tool. AwwaRF, Denver, CO.
Bahadur, R; Samuels, W; Grayman, W; Amstutz, D; Pickus; J. (2003b) PipelineNet: A Model for
Monitoring Introduced Contaminants in a Distribution System. World Water & Environmental
Resources Congress, Environmental & Water Resources Institute - ASCE.
Baxter, CW; Lence, BJ. (2003) A Framework for Risk Analysis in Potable Water Supply. World
Water & Environmental Resources Congress, Environmental & Water Resources Institute - ASCE.
Berry, ED; Siragusa, GR. (1999) Integration of hydroxy apatite concentration of bacteria and semi-
nested PCR to enhance detection of Salmonella typhimurium from ground beef and bovine carcass
sponge samples. J. Rapid Methods Automation Microbiol. 7:7-23.
Berry, J; Hart, W; Phillips, C; Uber, J. (2004) A General Integer-Programming-Based Framework
for Sensor Placement in Municipal Water Networks. World Water & Environmental Resources
Congress, Environmental & Water Resources Institute - ASCE.
Black & Veatch (2004) "Water Monitoring Equipment for Toxic Contaminants Technology
Assessment" General Dynamics, and Calspan/University of Buffalo Research Center, October 2004.
Bodurow, CC; Campbell, DP; Gottfried, DS; Xu, J; Cobb-Sullivan, JM; Caravati, KC. (2005) A
low-cost, real-time optical sensor for environmental monitoring and homeland security. EPA
Science Forum 2005. http://www.epa.gov/sciforum/2005/pdfs/oeiposter/bodurow catherine.pdf
Bravata, DM; Sundaram, V; McDonald, KM; Smith, WM; Szeto, H; Schleinitz, MD; Owens, DK.
(2004) Evaluating detection and diagnostic decision support systems for bioterrorism response.
Emerging Infectious Diseases.10(1):100-108.
Bunk, S. (2002) Sensing Evil. The Scientist. 16(15): 13.
August 2005 159
-------
U.S. EPA Office of Water Early Warning Systems
Carlson, K; Byer, D; Frazey, J. (2004): Section 5.2 - Candidate Instruments and Observables.
Section 5.3. - Models. May, 2004 Draft of Methodology and Characteristics of Water System
Infrastructure Security.
Cross, H. (1936) Analysis of Flow in Networks of Conduits or Conductors. Univ. of Illinois Eng.
Experiment Station Bulletin. Page 286.
DARPA (2004) Chemical and Biological Sensor Standards Study. LTC John Carrano, Study Chair.
Daviss, B. (2004) A Springboard to Easier Bioassays. The Scientist. 18(4):40. March 1, 2004.
Deininger, R; Lee, JY. (2001) Rapid determination of bacteria in drinking water using an ATP
assay. Analytical Chemistry and Technology. 5(4): 185-189.
DSRC Meeting (2004) Distribution System Research Consortium Draft Summary Report.
Distribution System Research Consortium Meeting, presentations by Vowinkle, Uber, and
McKenna: August 25-26, 2004, Cincinnati, Ohio USEPA.
ECBC (2002) Bio-detector Assessment. ECBC-TR-171, March 2002, Walther, Emanuel and Goode.
Edgewood Chemical Biological Center.
Emanuel, P; Chue, C; Kerr, L; Cullin, D. (2003) Validating the Performance of Biological Detection
Equipment: the Role of the Federal Government, BioSecurity and Bioterrorism: Biodefense
Strategy, Practice and Science.
ECAC (1999) Engineering Computer Applications Committee Calibration Guidelines for Water
Distribution System Modeling. Proc. AWWA ImTech Conference, 1999.
EPA (2000) Multi Agency Radiation Survey and Site Investigation Manual. (EPA 402-R-97-016),
August, 2000.
EPA (2003) A Review of Emerging Sensor Technologies for Facilitating Long-Term Ground Water
Monitoring of Volatile Organic Compounds. (EPA 542-R-03-007.)
http://www.epa.gov/tio/download/char/542r03007.pdf
EPA-ETV (2004) U.S. Environmental Protection Agency, Environmental Verification Program
http://www.epa. gov/etv/
EPA (2003/2004) Response Protocol Toolbox: Planning for and Responding to Drinking Water
Contamination Threats and Incidents. Modules 1-6. 2003/2004.
EPA (2004) Office of Research and Development, Shaw Environmental, Draft Report Evaluation
of Water Quality Sensors in Distribution Systems; May 2004 and EPA, Office of Research and
Development, Shaw Environmental, Draft Report: Water Quality Sensor Responses to Chemical and
Biological Warfare Agent Simulants in Water Distribution Systems July 2004.
EPA (2005) Distribution System Water Quality Report: A Guide to the Assessment and
August 2005 160
-------
U.S. EPA Office of Water Early Warning Systems
Management of Drinking Water Quality in Distribution Systems. ORD NRMRL Water Supply and
Water Resources Division. June 2005 Draft.
Fitzgerald, DA. (2002) Microarrayers on the spot. The Scientist. 16(4):42.
Fontenot, E; Ingeduld, P; Wood, D. (2003) Real-time Analysis of Water Supply and Water
Distribution Systems, World Water & Environmental Resources Congress, Environmental & Water
Resources Institute - ASCE.
Goodey, AP; McDevitt, JT. (2003) Multishell Microspheres with Integrated Chromatographic and
Detection Layers for Use in Array Sensors. J. Am. Chem. Soc. 125(10):2870-2871.
Gorman, J. (2003) NanoLights! Camera! Action! Tiny semiconductor crystals reveal cellular
activity like never before. Science News. 163(7): 107.
Grayman, W; Roberson, A; States, S. (2003) AWWA Contamination Monitoring Technologies
Seminar; Richmond, VA. May 2003.
Grayman, W; Deininger, R; Males; Gullick, R. (2004a) Source water early warning systems.
Chapter in Water Supply Systems Security. Edited by Larry Mays, McGraw-Hill and Companies,
New York, NY.
Grayman, WM; Clark, RM; Harding, BL; Maslia, M; Aramini, J. (2004b) Reconstructing Historical
Contamination Events. Chapter in Water Supply Systems Security. Edited by Larry Mays,
McGraw-Hill and Companies, New York, NY.
Grayman, W; Rossman, L; Deininger, R; Smith, AC; Smith, J. (2004c) Mixing and aging of water
in distribution system storage facilities. J. AWWA. 96(9).
Grow, AE; Deal, MS; Thompson, PA; Wood, LL. (2003) Evaluation of the Doodlebug: a biochip
for detecting waterborne pathogens. Water Intelligence Online IWA (WERF).
Hasan, J; States, S; Deininger, R. (2004) Safeguarding the security of public water supplies using
early warning systems: A Brief Review. J. of Contemporary Water Research and Education.
129:27-33.
Haupt, K. (2002) Creating a good impression. Nature Biotechnology. 20:884-885.
Heroux, K. and Anderson, P. (No date) "Evaluation of a Rapid Immunoassay System for the
Detection of Bacillus anthracis Spores." U.S. Army Edgewood Chemical Biological Center,
Aberdeen Proving Ground, MD. http://www.responsebio.com/pdf/EdgewoodEvaluation.pdf
Hindson, BJ; Brown, SB; Marshall GD; McBride, MT; Makarewicz, AJ; Gutierrez DM; Wolcott
DK; Metz, TR; Madabhushi, RS; Dzenitis, JM; Colston, BW Jr. (2004) Development of an
automated sample preparation module for environmental monitoring of biowarfare agents.
Analytical Chemistry. 76(13):3492-3497.
August 2005 161
-------
U.S. EPA Office of Water _ Early Warning Systems
Hrudey, SE; Rizak, S. (2004) Discussion of rapid analytical techniques for drinking water security
investigations. J. AWWA. 96(9): 110-1 13.
Kessler, A; Ostfeld A; Sinai, G. (1998) Detecting accidental contaminations in municipal water
networks. J. of Water Resources Planning and Management. 124(4):192-198.
King, KL. (2004) Presentation by K.L. King, Event Monitor for Water Plant or Distribution System
Monitoring; Hach Homeland Security Technologies; AWWA Water Security Congress; April 25-27,
2004.
ILSI ( 1 999) Early Warning Monitoring to Detect Hazardous Events in Water Supplies. International
Life Sciences Institute-Risk Science Institute. http://rsi.ilsi.org/file/EWM.pdf
Jentgen, L; Conrad, S; Riddle, R; Von Sacken, EW; Stone, K; Grayman, WM; Ranade, S. (2003)
Implementing a Prototype Energy and Water Quality Management System. AwwaRF/AWWA.
Kirby, R; Cho, EJ; Gehrke, B; Bayer, T; Park, YS; Neikirk, DP; McDevitt, JT; Ellington, AD.
(2004) Aptamer-based sensor arrays for the detection and quantitation of proteins. Analytical
Chemistry. 76(14):4066-4075.
Koblizek, M; Maly', J; Masoji'dek, J; Komenda, J; Kucera,T; Giardi, MT; Mattoo, AK; Pilloton,
R. (2002) A biosensor for the detection of triazine and phenylurea herbicides designed using
photosystem ii coupled to a screen-printed electrode. Biotechnology and Bioengineering.
Kretzmann, H; Van Zyl, J. (2004) Stochastic Analysis Of Water Distribution Systems. World Water
& Environmental Resources Congress, Environmental & Water Resources Institute - ASCE.
Kuhn, R; Oshima, K. (2002) Hollow-fiber Ultrafiltration ofCryptosporidiumparvum oocysts from
a wide variety of 10-L surface water samples. Canadian J. of Microbiology, 48(6):542-549.
Lee, B; Deininger, R; Clark, R. (1991) Locating monitoring stations in water distribution systems.
J. AWWA.83(7):6Q-66.
Li, Z; Buchberger, S; Tzatchkov. (2005) Importance of Dispersion in Network Water Quality
Modeling. Proc. World Water & Environmental Resources Congress, Environmental & Water
Resources Institute - ASCE. 2005.
Li, Z; Buchberger, S. (2004) Effect of Time Scale on PRP Random Flows in Pipe Network. World
Water & Environmental Resources Congress, Environmental & Water Resources Institute - ASCE.
Mays, LW. (2004) Water Supply Systems Security. New York: McGraw-Hill.
McCleskey, SC; Griffin, MJ; Schneider, SE; McDevitt, JT, Anslyn, EV. (2003) Differential
receptors create patterns diagnostic for ATP and GTP. J. Am. Chem. Soc. 125(5): 1 1 14-1 1 15.
August 2005 162
-------
U.S. EPA Office of Water Early Warning Systems
Murray, R; Janke, R; Uber, J. (2004) The Threat Ensemble Vulnerability Assessment Program for
Drinking Water Distribution System Security, Proceedings of the World Water and Environmental
Resources Congress, Salt Lake City, June 2004.
Ostfeld, A. (2004) Optimal Monitoring Stations Allocations for Water Distribution System Security.
Chapter in Water Supply Systems Security. Edited by L. Mays, McGraw-Hill.
Ostfeld, A; Salomons, E. (2004) Optimal layout of early warning detection stations for water
distribution systems security. J.WRPM, ASCE. 130(5):377-385.
Perkel, JM. (2003) Microbiology vigil: probing what's out there. The Scientist. 17(9):40.
Pesavento, M; D'Agostino, G; Alberti, G. (2004) Molecular Imprinted Polymers as Sensing
Membrane for Direct Electrochemical Detection of Pollutants, IAEAC: The 6th Workshop on
Biosensors and BioAnalytical u-Techniques in Environmental and Clinical Analysis
ENEA - University of Rome "La Sapienza": October 8-12, 2004 - Rome, Italy.
Powell, J; Clement, J; Brandt, M; Casey, R; Holt, D; Grayman, W; LeChevallier, M. (2004)
Predictive Models for Water Quality in Distribution Systems. AwwaRF-AWWA, Denver, CO.
Pyle, BH; Broadway, SC; McFeters, G. (1999) Sensitive detection ofEscherichia coli O157:H7 in
food and water by immunomagnetic separation and solid-phase laser cytometry. Applied and
Environmental Microbiology. 65:1966-1972.
Quist, GM; DeLeon, R; Felkner, 1C. (2004) Evaluation of a Real-time Online Continuous
Monitoring Method for Cryptosporidium. AwwaRF Project #2720.
Rider, TH; Petrovick, MS; Nargi, FE; Harper, JD; Schwoebel, ED; Mathews, RH; Blanchard, DJ;
Bortolin, LT; Young, AM; Chen, J; Hollis, MA. (2003) A B cell-based sensor for rapid
identification of pathogens. Science. 301:213-215.
Rife, JC; Miller, MM; Sheehan, PE, Tamanaha, CR; Tondra, M; Whitman, LJ. (2003) Design and
performance of GMR sensors for the detection of magentic microbeads in biosensors. Sensors and
Actuators A. 107:209-218.
Roberson, JA; Morley KM. (2005) Contamination Warning Systems for Water: An Approach for
Providing Actionable Information to Decision-makers. American Water Works Association.
http://www.awwa.org/Advocacy/pressroom/pr/index.cfm?ArticleID=424
Rosen, JS; Miller, DM; Stevens, KB; Ergul, A; Sobrinho, JAH; Frey, MM; Pinkstaff, L. (2003)
Application of Knowledge Management to Utilities. AWWA Research Foundation. ISBN
1-P-4.5C-90895-2/03-CM.
Rossman, L. (2000) EPANET Version 2 User's Manual. USEPA.
Ruan, C.; Zeng, K; Varghese, OK; Grimes, CA. (2004a) A staphylococcal enterotoxin B
magnetoelastic immunosensor. Biosensors and Bioelectronics. 20:585-591.
August 2005 163
-------
U.S. EPA Office of Water Early Warning Systems
Ruan, C; Varghese, OK; Grimes, CA; Zeng, K; Yang, X; Mukherjee, N; Ong, KG. (2004b) A
magnetoelastic ricin immunosensor. Sensor Letters. 2:1-1.
Ruan, C; Zeng, K; Varghese, OK; Grimes, CA. (2003) Magentoelastic immunosensors: amplified
mass immunosorbent assay for the detection of Escherichia coli O157:H7. Analytical Chemistry.
75:6494-6498.
Shaffer, KM; Gray, SA; Fertig, SJ; Selinger, JV. (2003) Neuronal Network Biosensor for
Environmental Threat Detection. Internet abstract found at:
http://www.nrl.naw.mil/content.php?P=04REVIEWl 18.
States, S. (2004) Rapid Screening, How Good Is it for Security Investigations? AWWA Jan 2004.
Tamanaha, CR; Whitman, LJ; Colton, RJ. (2002) Hybrind macro-micro fiuidics system for a chip-
based sensor. J. of Micromechanics and Microengineering. 12:N7-N17.
Tuck, K; Powers, M.; Millward, H; Chen, Y; Zharkikh, L; Wall, M; Li, W; Gundry, C; Pavlov, I;
Ditty, S; Hadfield, T; Karaszkiewicz, J; Nielsen, D; Teng, DH-F. (2005) Joint Biological Agent
Identification and Diagnostic System (JBAIDS): Stability and Sensitivity of Freeze-Dried Real-Time
PCR Assays. American Society of Microbiology 2005. Abstract 089/Y-018.
Uber, J; Janke, R; Murray, R; Meyer, P. (2004a) Greedy Heuristic Methods for Locating Water
Quality Sensors in Distribution Systems. Proc. World Water & Environmental Resources Congress,
Environmental & Water Resources Institute - ASCE, 2004.
Uber, J; Shang, F; Rossman, L. (2004b) Extensions to EPANET for Fate and Transport of Multiple
Interacting Chemical or Biological Components. World Water & Environmental Resources
Congress, Environmental & Water Resources Institute - ASCE.
van Bloemen Waanders, BG; Bartlett, R; Biegler, L; Laird, C (2003) Nonlinear Programming
Strategies for Source Detection of Municipal Water Networks. World Water & Environmental
Resources Congress, Environmental & Water Resources Institute - ASCE.
Walski, TM; Chase, DV; Savic, DA; Grayman, W; Beckwith, W; Koelle, E. (2003) Advanced
Water Distribution Modeling and Management, Haestad Methods. Waterbury CT: Haestad Press.
Watson, J; Greenberg, HJ; Hart, WE. (2004) A Multiple-Objective Analysis of Sensor Placement
Optimization in Water Networks. World Water & Environmental Resources Congress,
Environmental & Water Resources Institute - ASCE.
Whelan, RJ; Wohland, T; Neumann, L; Huang, B; Kobilka, BK; Zare, RN. (2002) Analysis of
biomolecular interactions using a miniaturized surface plasmon resonance sensor. Analytical
Chemistry. 74:4570-4576.
Whelan, RJ; Zare R.N. (2003) Surface plasmon resonance detection for capillary electrophoresis
separations. Analytical Chemistry. 75:1542-1547.
August 2005 164
-------
U.S. EPA Office of Water Early Warning Systems
Whitman, LJ; Sheehan, PE; Colton, RJ; Miller, RL; Edelstein, RL; Tamanaha, CR. (2001) Naval
Research Laboratory Review/Chemical/Biochemical Research.
WHO (2004) Public Health Response to Biological and Chemical Weapons: World Health
Organization Guidance, http://www.who.int/csr/delibepidemics/biochemguide/en/index.html
August 2005 165
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
APPENDIX A
WaterSentinel Overview
Promoting the security of the nation's water infrastructure is one of the most significant
undertakings and responsibilities of the U.S. Environmental Protection Agency (EPA) in a
post-September 11 world. An attack, or even a credible threat of an attack, on water infrastructure
could seriously jeopardize the public health, infrastructure, and economic vitality of a community.
Although historical evidence suggests that the probability of intentional contamination of the
drinking water supply is relatively low, experts agree that it is possible to contaminate a portion of
a drinking water system, resulting in adverse public health consequences. Furthermore, the
probability of a contamination threat (the mere indication that contamination of the drinking water
supply may have occurred) is relatively high. Given that it is possible to contaminate drinking water
at levels of public health concern, and the probable occurrence of contamination threats in the water
sector, there is a need to evaluate the credibility of any contamination threat and identify appropriate
response actions in a very short period of time.
In recognition of this threat and in response to Homeland Security Presidential Directive 9, EPA
developed the proposed WaterSentinel initiative. HSPD-9 directs EPA to:
• "develop robust, comprehensive, and fully coordinated surveillance and monitoring systems .
.. for ... water quality that provide early detection and awareness of disease, pest, or poisonous
agents"; and
• "develop nationwide laboratory networks for... water quality that integrate existing Federal and
State laboratory resources, are interconnected, and utilize standardized diagnostic protocols and
procedures."
The proposed WaterSentinel initiative will build on existing EPA efforts to design and deploy a
contamination warning system (CWS). A CWS is an evolution of the "early warning system"
concept and involves the active deployment and use of monitoring technologies/strategies and
enhanced surveillance activities to collect, integrate, analyze, and communicate information to
provide a timely warning of potential water contamination incidents and initiate response actions
to minimize public health and economic impacts.
The key to an effective and timely response to a water contamination threat is to minimize the time
between indication of a contamination incident or change in water quality and implementation of
effective response measures through early detection of threat warnings of potential contamination.
Identification of a contamination threat leads to response actions designed to determine whether or
not a threat is credible and to protect public health in the case of a credible threat. Early detection
can be achieved through the implementation of a CWS. A CWS is not merely a collection of
monitors and equipment placed throughout a water system to alert of intrusion. Fundamentally, it
is an exercise in information management. Different information streams must be managed,
analyzed, and interpreted in a timely manner to recognize potential contamination incidents in time
to respond effectively.
Figure 1 presents an overview of the components of the proposed concept of operations for the
WaterSentinel contamination warning system (WS-CWS). Although an effective CWS should be
designed to maximize the detection of contamination incidents, accidental or intentional, it is
August 2005 A-l
-------
U.S. EPA Office of Water
Early Warning Systems
Figure 1 Overview of the WS-CWS Concept of Operations
o
Di
Contamination Warning
System
Monitoring
0 i -III e •.',• .n tr i • | < ,-i lit. pa ram « te rs
Roitlie t am p III g aid .n i .n 1. : 1:
E i t -j i - T • 1 : T • i i It .
S u rveillance
P ih Me 1 ea Itl ; i rue II,: i • T
Coif im e r com | • 1,1 n t : i r v e II l.n i • r
0
Event Detection
Change in water quality
Deviation from baseline
N otification
Security breach
C redibility Determ ination
R ule out— return to
routine operations
C re d ib le
C o n firm e d
o
Q.
M
(D
tt
Remediation and Recovery
D econtam ination
Sampling and analysis
C le a r a n c e
Consequence Management
0 perational response
Pub lie health response
Laboratory confirm ation
important to demonstrate the effectiveness of the design and integration of CWS components
through a systematic process that can be expanded and adapted over time to ensure sustainability.
In the design of the WS-CWS, EPA will partner with drinking water utilities, key water sector
stakeholders, technical experts, representatives from public health, law enforcement, and other
federal agencies to focus on first-generation CWS components that initially address a representative
subset of priority contaminants to improve a utility's ability to respond to any contamination threat
or incident. In addition, the WS-CWS will yield operational benefits for non-security related water
quality issues and enhance collaboration/integration of water utilities and local health departments.
Through working with these partners, EPA will use the results of the WaterSentinel initiative
demonstration proj ect to develop a sustainable model for a CWS that can be implemented by utilities
throughout the nation.
Major Components of the WS Initiative
The major components of the WS initiative include the following:
• System architecture and program design
• Selection of baseline contaminants
• Laboratory support and the Water Laboratory Alliance
• Event detection and credibility determination
August 2005
A-2
-------
U.S. EPA Office of Water
Early Warning Systems
Figure 2
Contamination
Warning
System
• Consequence management
• Data management
System Architecture and Program Design
The WS system architecture will define the
conceptual approach for the WS-CWS and
document the most effective combination of
CWS components to yield a sustainable
program that can be adopted and implemented
by drinking water utilities. The key
components of a CWS are shown in Figure 2
and are described as follows:
• Water quality monitoring. Multiple
approaches are available for monitoring
water quality as part of a CWS. The
WS-CWS will focus primarily on two
options as described below.
° Online monitoring for changes in
water quality parameters. Online
monitors for water quality parameters, such as chlorine residual, pH, conductivity, turbidity,
etc., can potentially detect an identifiable change from an established water quality baseline
and serve as an indicator of potential contamination in the WS-CWS.
° Routine sampling for select contaminants. Water samples can be collected at a
predetermined frequency or in response to a trigger and analyzed for specific, targeted
contaminants. It may also be possible to detect some non-target analytes if the analytical
techniques used in the routine monitoring program are sufficiently robust and if the analysts
are trained and encouraged to investigate tentatively identified contaminants.
• Consumer complaint surveillance. Consumer complaints regarding unusual taste, odor, or
appearance of the water are often reported to and recorded by water utilities, which
conventionally use them to identify and address water quality problems. Using an appropriate
methodology, WS could track and analyze these complaints to look for unusual trends that may
be indicative of a contamination incident.
• Public health surveillance. Syndromic surveillance by the public health sector as well as
reports from 911 call centers and poison control hotlines might serve as a warning of a potential
drinking water contamination incident if there is a reliable link between the public health and
drinking water utilities.
• Enhanced security monitoring. Security breaches, witness accounts, and notifications by the
perpetrator, news media, or law enforcement can be monitored through enhanced security
practices.
August 2005
A-3
-------
U.S. EPA Office of Water Early Warning Systems
Factors for consideration in the development of the WS system architecture for a given utility
include the following:
• Sustainability and dual use: The ability of utilities to operate and maintain the WS-CWS to
benefit water security as well as routine operations and water quality concerns.
• Cost-benefit of implementation: The ability to demonstrate that the cost of implementing,
operating and maintaining the WS-CWS is justified by the benefits. This factor is directly
related to the previous consideration of Sustainability and dual use.
• Universal application: The ability to adapt and implement the WS-CWS design, in some
manner, at any drinking water utility regardless of size, treatment type, location, or complexity.
Under the proposed initiative, EPA would work with utilities to determine sensor placement and
sampling locations, develop and enhance communication and coordination between the utility and
public health community, identify approaches for enhancement and integration of consumer
complaint and public health surveillance, and define dual use benefits of implementation of the
WS-CWS.
Selection of Baseline Contaminants
Many potential monitoring and surveillance components that could be integrated into a CWS, and
evaluated during WS demonstration project, have not been thoroughly demonstrated in a CWS
application. Thus, it is critical that the WS demonstration project be limited to contaminants for
which technologies/strategies are well understood or characterized through other water monitoring
application. For example, there is substantial experience with technologies used to monitor basic
water quality parameters, such as chlorine residual, pH, and conductivity. Use of such established
technologies during the WS demonstration project will allow for the focus on the performance of
the contamination warning system, without introducing additional uncertainty associated with
research techniques. Once the concept of a CWS has been demonstrated, novel detection
technologies could be evaluated in the context of this proven system.
The WS initiative is intended to increase protection against those contaminants that could cause
serious harm to public health or economic well-being if introduced into a drinking water system.
Thus, selection of candidate contaminants for the WS initiative should consider the threat posed by
potential contaminants. Therefore, the first step in the development of the WS initiative is to
identify those baseline contaminants that will be included in the demonstration project.
The objectives of the WS contaminant selection process are to:
• Select a reasonable number of baseline contaminants that provide coverage for all categories of
priority contaminants using a combination of different monitoring and surveillance
technologies/strategies.
• Identify research priorities related to sampling and analysis methods that can be initiated in the
near term for evaluation in the WS demonstration project.
• Identify long-term research priorities related to sampling and analysis to ensure future coverage
of all priority contaminants.
August 2005 A-4
-------
U.S. EPA Office of Water Early Warning Systems
Laboratory Support and the Water Laboratory Alliance
To provide necessary analytical support for both routine monitoring and response actions, the Water
Laboratory Alliance (WLA) is intended to establish capability and capacity at drinking water utility
laboratories for routine monitoring of baseline contaminants. The WLA is a network of laboratories
with extensive capability for the analysis of water samples for a wide range of potential
contaminants. It is proposed that the WLA integrate existing water quality labs with the existing
Laboratory Response Network (LRN), established by the U.S. Center for Disease Control and
Prevention (CDC) to support analysis of potential biothreat agents.
Required confirmatory and response analysis capabilities and capacities should be established to
support the WS initiative. Since many CWS components may provide non-specific indicators of
potential contamination, laboratories that are part of the WLA should be proficient in the screening
and analysis of unknown samples for chemicals, pathogens, and radionuclides. In addition to the
development of laboratory capabilities to support the WS-CWS, the development of standardized
analytical methods that can be utilized for screening, presumptive, and confirmatory analysis will
be a key focus of research activities.
Event Detection and Credibility Determination
While the WS initiative is designed to gather and integrate information from a number of sources
that might be indicative of a contamination threat, the information is only useful in the context of
a CWS if it can be quickly and effectively used to make appropriate response decisions. Thus, there
is a need for decision support tools.
In the WS-CWS model, event detection is defined as a signal from the CWS that is indicative of a
possible contamination incident. This signal could be a pattern of unusual water quality, a cluster
of unusual consumer complaints, or unusual symptoms picked up by a public health surveillance
program. Although public health surveillance systems have their own event detection algorithms,
these do not exist, or are not widely deployed, for water quality and consumer complaints. Thus,
there is a need to develop event detection software (EDS). The most important function of the EDS
is to filter out the anomalies that normally occur, or which have known causes, and signal only those
events that are likely to be possible contamination incidents. In short, the purpose of the EDS is to
reduce the false-positive rate without missing potential events.
Although the EDS can indicate a possible contamination threat, it cannot indicate a credible threat
that requires response actions to protect public health. Furthermore, the human element cannot be
removed from the credibility determination step. However, it is possible to develop a tool to support
officials in decision making by guiding them through the evaluation process and aiding in the
synthesis of information necessary to make timely and appropriate response decisions. Ultimately,
the decision will always rely on human judgment and evaluation of incomplete information.
However, this decision tool can be a great aid in the process, and might substantially reduce the time
to make critical response decisions.
August 2005 A-5
-------
U.S. EPA Office of Water Early Warning Systems
Consequence Management
Early detection of a contamination incident is only beneficial in minimizing public health or
economic impacts if effective response decisions also can be made in a timely manner. A
consequence management protocol will provide a decisionmaking framework that governs when,
how, what, and who will be involved in making decisions in response to contamination threat
warnings to minimize the response timeline and implement operational or public health response
actions appropriately. Availability of a robust and tested consequence management protocol will
be a key factor in initiation of monitoring and surveillance activities.
The systematic approach for assessing credibility in response to contamination threat warnings
ensures that all available information is analyzed in a timely and efficient manner to minimize both
false alarms and over-response to a trigger that has not been determined to be credible. While the
system architecture will identify an integrated set of contamination threat warnings that provide
input for consequence management decisions, the consequence management protocol will be
independent and able to inform response decisions and containment strategies for existing triggers
and contamination threat warnings in the absence of a formal CWS.
Data Management
A CWS is not merely a collection of monitors and equipment placed throughout a water system to
alert of intrusion. Fundamentally, it is an exercise in information management. Different
information streams must be managed, analyzed, and interpreted in a timely manner to recognize
potential contamination incidents in time to respond effectively.
The elements of the WS-CWS present a rich topography of information needs to be collected,
integrated, and analyzed to make response decisions. For each CWS component, EPA and partners
in the WS initiative would identify the requirements of the technology employed to analyze data and
distinguish a potential incident from the established baseline.
Based on CWS components identified through WS system architecture and program design, data
may need to be extracted from a combination of SCADA systems, laboratory information
management systems (LIMS), consumer complaint surveillance systems, security monitoring
systems, and public health surveillance systems in a manner that allows for timely decision making
and response.
Next Steps
WaterSentinel has been proposed as a demonstration project to commence in Fiscal Year 2006. EPA
will launch this project by building on existing efforts. Throughout Fiscal Year 2005, EPA will
continue to work with the water sector on activities that could lay the groundwork for WaterSentinel.
Such activities could include the design of a model contamination warning system, analysis of
contaminants that could be effectively monitored for a timely response, the development of
consequence management protocols for response to a potential incident, and research into
technologies that could be candidates for deployment.
August 2005 A-6
-------
U.S. EPA Office of Water Early Warning Systems
APPENDIX B
Agencies Involved in Early Warning Systems
The following are brief descriptions of agencies involved in research, evaluation, or development
of early warning systems.
Federal Agencies (Research and Programs)
The U.S. Environmental Protection Agency
EPA plays a lead role in protecting the water sector. In National Strategy for Homeland Security
(July 2002), EPA is designated as responsible for protecting our national water supply. Also, under
the Homeland Security Presidential Directive-9 (January 2004), EPA is one of the Federal Agencies
responsible for agriculture, food, and water security to "develop robust, comprehensive, and fully
coordinated surveillance and monitoring systems,... that provide early detection and awareness of
disease, pest, or poisonous agents."
To spearhead the research efforts on water security, EPA formed the National Homeland Security
Research Center (NHSRC) in the Office of Research and Development (ORD). EPA has initiated
many efforts related to early warning systems, including the following:
• Drafting the Water Security Research and Technical Support Action Plan (Action Plan), which
provides the basis for this project.
• Initiation of an NHSRC's Water Awareness Technology Evaluation Research Security
(WATERS) Center to conduct research projects including an evaluation of various sensor
technologies and data acquisition systems.
• Use of EPA's Environmental Technology Verification (ETV) Program to provide credible
performance data for commercial-ready environmental technologies including homeland security
technologies (e.g., early warning systems).
• Sponsoring efforts by the American Society of Civil Engineers to develop Guidance for
Designing Online Contamination Monitoring System.
• Forming the Distribution System Research Consortium (DSRC) in June 2003 to provide a forum
for information exchange on a diverse water distribution system security topics including early
warning systems research (e.g., sensor, field studies, sensor placement).
• Forming the Threat Ensemble Vulnerability Assessment (TEVA) Research Program for
Drinking Water Distribution System Security. One of the products is the help in designing early
warning systems and for evaluating strategies for locating sensors in the distribution system.
• Developing various guidance on emergencies and early warning systems.
Department of Homeland Security
The Department of Homeland Security was established in 2002 in order to protect the nation from
terrorist attacks.209 As part of this mission, DHS has the responsibility to protect the nation's
drinking water. DHS has established the Ready Campaign for the purpose of educating the public
on a continuous basis so that communities will be ready in case of an emergency scenario.210 The
2004 Homeland Security Presidential Directive/HSPD-9 also requires DHS to protect the agriculture
and food systems from attacks, disasters, and other emergencies.211 Part of this directive's
August 2005 B-l
-------
U.S. EPA Office of Water Early Warning Systems
Awareness and Warning section is to enhance intelligence operations in securing water assets, such
as implementing effective monitoring and detecting capabilities. The directive also orders further
research and development of detection, prevention, characterization, response, and recovery
countermeasures. Furthermore, DHS, in conjunction with EPA and other agencies, is involved
reviewing the available information on technology performance.212 DHS has also provided grants
to projects involved with the research and development of technologies that could be used for early
warning systems in water, such as the SCADA systems.213
Department of Defense
The danger of nuclear, biological, and chemical arms has been a concern of the Department of
Defense (DOD) even before the 9-11 terrorist attacks. Programs such as the Joint Service Agent
Water Monitor Program and other detector and sensor programs were formed since the late 1990s.214
Now, under E2.1.20. Food and Water Security of the Antiterrorism Program from Directive number
2000.12 of 2003, the Department of Defense is obligated to protect food and water sources from
disruption and contamination or other terrorist attacks. DOD must fulfill this requirement by taking
action to detect, prevent, and mitigate the effects from intentional contaminations of food and water
sources.215 DOD is thus continuously supportive of the research and development of more advanced
water contamination detective devices such as detectors that are faster, lighter, and smaller and can
be used in-field.216 DOD strives to achieve this by funding organizations, such as SNL, that are
developing EWS technology. Finally, DOD contributes to the development of EWS through its
research and development programs, such as the Defense Advanced Research Projects Agency
(DARPA), further described below. A list of other DOD-funded R&D organizations is available
at http://www.dtic.mil/ird/websites/orgsites.html.
Defense Advanced Research Projects Agency
DARPA is the main research and development organization under DOD and strives for superior
military technology.217 For water security measures, DAPRA is looking for fast, highly sensitive,
and highly specific biosensor systems.218 The four thrust areas of the Biosensors Technologies
Program are, (1) Mass-Based Identification Technologies, (2) Surface-Based Identification
Technologies, (3) Nucleic Acid-Based Identification Technologies, and (4) Breath Analysis-Based
Identification Technologies.219 The Biosensors Technologies Program is conducted in collaboration
with 20 universities and laboratories. As mentioned in this report, the DARPA Chemical and
Biological Sensors Standards Study presents methods for evaluating sensors by capturing the
performance trade-offs between sensitivity, probability of correct detection, false-positive rate, and
response time.
Naval Research Laboratory
The Naval Research Laboratory (NRL) is the corporate research laboratory for the Navy and Marine
Corps. The focus of this laboratory encompasses scientific and technology research and
development particularly for maritime applications as well as atmospheric and space sciences and
technologies.220 NRL has been researching and developing methods to detect contaminants in the
environment. In collaboration with GeoCenters, Inc., these efforts have resulted in the development
of a method to detect radioactive material such as uranium on a microchip.221 Under the sponsorship
of the Office of Naval Research, NRL also developed a field test method in potable water for
August 2005 B-2
-------
U.S. EPA Office of Water Early Warning Systems
cyanide that did not require laboratory testing.222 NRL also developed a microchip for detecting
cyanide in vapor and drinking water,223 in a collaborative effort with GeoCenters, Inc., the New
Mexico State University, and the University of California. NRL is currently developing a high-
sensitivity biosensor for monitoring airborne and water-borne contaminants in the environment. The
force amplified biological sensor can detect the force between DNA and antigen molecules.224 Other
advancements include the development of RAPTOR™ and the Bead ARray Counter (BARC) chip,
as discussed in this report. RAPTOR™ is a portable, rapid, automatic fiuorometric assay system
for monitoring biological agents, toxins, explosives, and chemical contaminants. The BARC chip
consists of an array of DNA spots immobilized on a surface,225 detecting hybridized sample DNA
using magnetic beads.226 The development of BARC was sponsored by DARPA and ONR.227
Edgewood Army Proving Grounds
Although Edgewood Army Proving Grounds was originally founded as a chemical weapons
research, development, and testing facility, Edgewood now focuses on chemical weapon defensive
measures, and responds to the Army Chemical and Biological Defense Command, which oversees
the Army's nonmedical chemical and biological defense activities.228 Army Edgewood Joint Service
Agent Water Monitor Program is actively looking at sensors for distribution using the following
proven classes of detection technologies: conventional, optical techniques, polymers/materials,
assays, and sentinel species. Additionally, it is examining new areas of MEMS (Micro-Electro-
Mechanical Systems) and MOEMS (Micro Optical Electro Mechanical Systems), which rely on a
concept proven in one of the other classes. Edgewood is collaborating with EPA in order to
heighten homeland security efforts.229
U.S. Geological Survey
Water quality has been a large focus for the U.S. Geological Survey. USGS has evaluated many
different real-time continuous water quality monitoring stations.230 In addition, USGS was part of
workshops such as the ILSI monitored Early Warning Monitoring to Detect Hazardous Events in
Water Supplied in 1999,231 as well as the 2004 National Monitoring Conference, "Building and
Sustaining Successful Monitoring Programs" by the National Water Quality Monitoring Council.232
USGS has also supported EWS projects such as the development of the Efficient Hydrologic Tracer-
test Design Program, which provides scenario simulations of a release event so that water systems
can test their preparedness.233
National Laboratories
Sandia National Laboratories
Since well before 2001, Sandia National Laboratories' (SNL) Chem/Bio Program has been involved
in the development of advanced sensor technologies for rapid detection of chemical and biological
warfare agents.234 Most of these technologies are pre-commercial and in various stages of
prototyping, including the uChemLab and Micromachined Acoustic Chemical Sensor. However,
SNL believes that once mature, its technologies should offer cost-effective monitoring alternatives.
SNL's specific water sensor development activities include (Wayne Einfield, SNL, personal
communication) the following:
August 2005 B-3
-------
U.S. EPA Office of Water Early Warning Systems
1) Adaptation of gas-phase juChemLab for detection of trihalomethanes and hydrolysis products
of chemical agents in water. Researchers are building a "front-end" for an existing chip-based
gas chromatograph. The system is initially being developed for analysis of trihalomethanes in
water as well as hydrolysis products of chemical agents. They anticipate the completion of a
field-ready prototype in 2005.
2) Use of liquid-phase juChemLabfor the continuous online detection ofbiotoxins in water. This
is probably the most mature of the SNL portfolio of technologies. A proteomics-based analysis
scheme utilizes microfluidics-based capillary-zone and capillary-gel analyses for biotoxin
separation coupled with laser induced fluorescence detection, all in a hand-held, field portable
package. SNL is in final discussions with two major cooperative research and development
agreement (CRADA) partners and a successful agreement will launch an aggressive project to
optimize and test this system for use as a real-time device for water distribution system
monitoring.
3) Preconcentration of bacterial species in water using insulative dielectrophoresis and
microfabricated fixtures. This project is aimed at the development of a preconcentrator for
biological species in water that offers promise to separate various types of bacteria based on
their mobility in an electric field. Researchers have demonstrated the ability to differentiate
between live and dead E. coli cells as well as the ability to differentiate between bacteria and
inert particles in the water matrix.
4) Microfabricated electroanalysis system for the detection of inorganics in water. This project
is aimed at the detection of various electroactive species in water (e.g., lead, cadmium, arsenic)
and utilizes a microfabricated multi-electrode array to measure various species by anodic
stripping voltametry. SNL researchers have a table-top prototype that is being optimized for
lead analysis, however, they anticipate that the functionality of this instrument could be
expanded as one of a suite of early warning sensors.
5) SNL, CH2M Hill (Colorado), and Tenix Investments (Australia) have signed an agreement that
calls for an online water monitoring prototype, based on uChemLab, to be developed and testing
to begin by June 2005.235 The first phase of testing will focus on detecting ricin and botulinum
toxin. The development team eventually also hopes to address viruses, bacteria, and parasites.
Lawrence Livermore National Laboratory
Lawrence Livermore National Laboratory (LLNL) is a national laboratory operated by the
University of California for U.S. Department of Energy. Although LLNL was founded as a nuclear
weapons design laboratory, it has broadened its field of work to include energy, biomedicine, and
the environment.236 LLNL has been developing a wide range of technologies for sensors.237 As
discussed in this report, LLNL utilized Luminex® technology to develop an Autonomous Pathogen
Detection System (APDS). The system has an automated sample preparation module based on
sequential injection analysis (SIA). Also discussed in this report is the "hand-held nucleic acid
analyzer"(HANAA) based on real-time PCR (TaqMan) developed by LLNL.
August 2005 B-4
-------
U.S. EPA Office of Water Early Warning Systems
Oak Ridge National Laboratory
The Oak Ridge National Laboratory (ORNL) is a multiprogram science and technology laboratory
managed by UT-Battelle, LLC for the DOE.238 ORNL conducts research and development in the
areas of energy, environment, and national security. In addition, the laboratory produces isotopes,
manages information and technical programs, and provides research and technical assistance to other
organizations.239 ORNL is one of the leading laboratories in developing biosensors and
bioreporters.240 With regard to EWS in water, ORNL has developed the Large-Scale Water Supply
Sentinel, a device that analyzes the characteristics of algae photosynthesis, in response to concerns
for military and municipal water safety.241 ORNL also licensed the VeriScan™ 3000 System,
produced by Protiveris.242 In addition, ORNL is associated with organizations such as the Center
for Advanced Biomedical Photonics, and the Advanced Biomedical Sciences and Technology
Group.243 ORNL researchers are collaborating with Gary Sayler, University of Tennessee on
biosensor and nanotechnology projects (Gary Sayler, U. of Tennessee, personal communication).
Pacific Northwest National Laboratory
The Pacific Northwest National Laboratory (PNNL) is a national laboratory operated by Batelle for
DOE. PNNL conducts research and development and supports education in the areas of
environment, energy, health and national security, and economics.244 Homeland security has been
a focus of PNNL even before 9/11. In the area of chemical, nuclear, and biological weapon
detection, PNNL has contributed to the development of sensor and measurement technology,
electronic (including controls) and system integration application requirements.245 The Sensors and
Electronics branch of PNNL is developing its electronics and systems in the areas of biological
sensors, chemical sensors, physical property sensors, nuclear radiation sensors, and macro property
measurement.246 In order to supplement the rapid detection of biological threats, PNNL has
developed the Biodetection Enabling Analyte Delivery System (BEADS). BEADS automated
technology that isolates bacteria, spores, viruses, and their DNA from water, air, or dirt samples
without requiring any manual preparations of samples.247
Idaho National Laboratory
On February 1, 2005, the Idaho National Engineering and Environmental Laboratory and Argonne
National Laboratory-West became the Idaho National Laboratory (INL248). EPA-NHSRC and INL
have an interagency agreement to develop and produce a next generation prototype of the
Ultrafiltration Concentration (UC) device previously developed by NHSRC and other stakeholders.
The UC bench top device concentrates microbial pathogens within a 100 L municipal drinking water
sample into a 250-mL volume in approximately a 2-hour time frame (400-fold concentration). INL
hopes to use the bench top UC system, which has been tested at the NHSRC in Cincinnati, to
redesign/repackage and automate the components such that the new device can be operated in the
field as a near-commercial, or field prototype system
August 2005 B-5
-------
U.S. EPA Office of Water Early Warning Systems
Other Agencies/Organizations
Pittsburgh Water & Sewer Authority
The Pittsburgh Water & Sewer Authority is one of the organizations that the Water Environment
Research Foundation and EPA are funding in order to identify, screen, and treat contaminants to
ensure water security. For example, the Authority has conducted verification tests on many of the
newer technologies for rapid detection in drinking water. It is also determining the risks and toxicity
associated with biological, chemical, and radiological agents introduced to sewers, analyzing the fate
and transport as well as treatment methods in the wastewater treatment facility, and developing
emergency operating/containment procedures.249
National Academy of Sciences: Water Science and Technology Board
The Water Science and Technology Board (WSTB) was organized in 1982 by the National Research
Council to provide a focal point for studies related to water quality and water resources.250 Projects
administered by the WSTB related to EWS include "Public Water Supply Distribution Systems:
Assessing and Reducing Risks",251 and the "Review of EPA Homeland Security Efforts: Panel on
Water System Security Research."252
American Water Works Association Research Foundation
The American Water Works Association Research Foundation (AwwaRF) is an international
nonprofit organization that sponsors research efforts for its subscribing member organizations to
provide safe and affordable drinking water.253 AwwaRF has been sponsoring many water security
projects with the support of governmental agencies such as CDC and EPA, national and international
research foundations, city and state water departments, and universities. These projects cover a wide
range of water security topics including assessment of technologies, online monitoring and early
warning systems, communication, assessing microbial contamination, and disaster response.
AwwaRF projects relevant to early warning systems for drinking water have focused on the early
detection of pathogens, chemicals, radioactivity, and biotoxins so that utilities can appropriately
respond in case of an intentional contamination of the water system. As terrorist concerns heighten,
the ability to quickly detect and identify contaminants is vital. One of the major focuses of the
AwwaRF projects is to develop a portable hand-held water monitor capable of real-time detection
of all harmful agents and implement Supervisory Control and Data Acquisition (SCAD A) systems
to facilitate and advance monitoring and communication. Projects are developing and advancing
methods to correctly identify E. coli and the different stages of Cryptosporidium parvum using
technology such as the Multi-Angle Light Scattering (MALS). Other projects are focusing on
strategy planning, from preventing intrusions into the drinking water systems, selecting water
sampling locations and methods, and protecting utility SCADA equipment, to emergency
management planning. Finally, some projects are focused beyond water treatment and distribution
systems. For example, point-of-use drinking water devices may be useful as a short-term emergency
response option. AwwaRF Project Summaries are presented in Appendix D.
August 2005 B-6
-------
U.S. EPA Office of Water Early Warning Systems
Water Environment Research Foundation
The Water and Environment Research Foundation (WERF), created in 1989, funds wastewater and
water quality research.254 The subscribers of the Foundation are utilities, municipalities,
corporations and industrial organizations, all with a common interest in promoting research and
development in water quality science and technology. The vast majority of WERF's research
embraces a broad spectrum of scientific and technical disciplines (e.g., microbiology, wastewater
ecology, toxicology, biological sciences, environmental engineering, and instrumentation) as well
social/behavioral sciences (e.g., communications and public perception).
Prior to 9/11, WERF has undertaken studies on sensors with a view to study influent toxicity
monitoring and process control. Since 9/11, WERF, under the aegis of a security grant from EPA,
has been carrying out projects in the areas of (i) chemical, biological, and radioactive contamination
events (accidental or purposeful), (ii) design of expert support systems for tracking anomalies in
wastewater characteristics, (iii) GIS-based modeling of contaminants travel in piped conveyance
systems, (iv) design of intelligent sensors for online, real-time upset early warning devices (UEWD)
for tracking chemical and biological contaminants, and (iv) cyber security related to process control
systems at water/wastewater utilities. WERF undertook a review of UEWDs in Yr 2000 that
recommended that fundamental studies would be needed to articulate the links between how an
influent event (the source) leads to an intermediate biochemical/physiochemical response (the cause)
that results in an observable disruption in the treatment process (the effect). In Yr 2004, WERF
sponsored sensor technology studies for water quality monitoring using fiber-optic biosensors (for
rapid pathogen detection), bioluminescent biosensors (for toxicity-screening) and x-ray fluorescence
spectroscopy (for waterborne metals). WERF also organized a sensor workshop (August 30-31,
2005) to prioritize research and development needs for rapid online contaminant monitoring.
American Society of Civil Engineers - Water Infrastructure Security Enhancements (WISE) -
Standards Committee
EPA has contracted WISE to produce a guidance document to assist water utility companies to
design and implement online contamination monitoring systems to detect intentional contamination
events.
Rutgers University
With the support of EPA, Rutgers University has been involved in the research on EWSs. For
example, they recently sponsored an annual Workshop on Advanced Technologies in Real-Time
Monitoring and Modeling for Drinking Water Safety and Security.255
August 2005 B-7
-------
[this page intentionally left blank]
-------
U.S. EPA Office of Water Early Warning Systems
APPENDIX C
List and Criteria for Selecting Products and Technologies
The goal of this EWS document is to report on the state-of-the-art techniques for detecting
contaminants, specifically chemical/microbial/radiological agents in drinking water distribution
systems. To focus on the most promising products/technologies in this relatively new area, the
following criteria were developed for including technologies/products in this document. The list of
technologies is included at the end of this appendix.
The following three categories of technology development were designated:
• Available now (being used or could be used by water utilities)
• Potentially adaptable technology used (but needs additional steps to address specific challenges
for use with water distribution system)
• Emerging technologies that may be applicable
The overriding criterion for categories is the focus on field portable (carried into field) and online
technologies (not benchtop). Also, because an analysis of market viability was beyond the scope
of the document, technologies/products are presented regardless of whether they would be at some
point cost effective for the water utility industry. However, some costly technologies/products may
not be considered because the manufacturers already determined that their products are too
expensive for the water utilities market and thus have not actively developed or adapted their
products for water detection.
Category 1. Available now:
Criteria:
• used/available now for water
• may be validated for water, perhaps for distribution conditions
Elaboration on Criteria:
• Portable and online products that are on the market and are specifically marketed for drinking
water distribution systems. This includes basic water quality online monitors, since they may
be adapted to provide early warning for CBW agents. This also includes toxicity kits that are
marketed to water utilities. This will cover the online technologies listed in the ASCE Guidance
and include additional portable kits and devices.
Category 2. Potentially adaptable technology used, but needs additional steps to address specific
challenges for use with water distribution system:
Criteria:
• demonstrated bench top version for water with ongoing work on field portable version (carried
into the field)
• portable products for water, but have some hurdles (sample concentration, chlorine removal)
August 2005 C-l
-------
U.S. EPA Office of Water Early Warning Systems
• available for detecting CBW agents (in use for other media such as food or air ) and can be
adapted to water (accessory equipment)
• technology used in similar application (source water)
Elaboration on Criteria:
Several technologies that are in use for applications other than drinking water testing could be used
for drinking water if additional sample preparation steps are taken.
• Systems that would require removal of chlorine residuals - Cell- and organism-based
biomonitors, are presented. Their adaptation for drinking water relies on the development of
methods for removing chlorine residuals. These methods are in the research stage, but are highly
desired, so should be forthcoming.
• Concentration of sample volume - Some technologies, such as portable PCR systems, could
be used with drinking water samples if used in conjunction with methods for concentrating large
sample volumes into reaction-size volumes. These products are presented with the expectation
that sample concentration techniques are forthcoming.
• Vaporization/Volatilization - Some of the portable and online vapor phase detectors that are
on the market for detecting CBW agents and are used by first responders can read water samples
if used with accessory equipment. There are existing methods/equipment for vaporization and
volatilization. This technology would not be suggested for microbial detection. Specific vapor
phase sampling products will be presented with the note that they have not been validated for
use with water samples. Companies are interested in adapting existing products for use with
water.
Category 3. Emerging technologies
Criteria:
• considered technologies defined as promising, given funding/grants from organization working
directly on EWS technologies (AwwaRF, EPA, DHS, DOD)
• appears in literature numerous times (could be various researchers) at recognized Detection
Conferences. Efforts from company, university, national laboratory, or government, and/or are
now being licensed to company for development/testing of prototype or product
• proof-of-concept has been demonstrated and the technology is still being aggressively pursued
• could be utilized for drinking water sampling with further development
Elaboration on Criteria:
The following conference abstracts were examined to determine which technologies are current hot
topics:
• 2004 Biodetection Technologies, Washington DC, June 2004256
• Detection Technologies, Arlington, VA, December 2003257
• Research, Technologies and Applications in Biodefense, Washington, DC, August 2003258
• 2003 Biodetection Technologies, Arlington, VA, June 2003259
August 2005 C-2
-------
U.S. EPA Office of Water Early Warning Systems
• Biosensing Research and Development, World Technology Evaluation Center, Inc. National
Institutes of Health, December 2002260
• Detection Technologies, Arlington, Va, December 2002261
• Workshops on Advanced Technologies in Real-Time Monitoring and Modeling for Drinking
Water Safety and Security, Rutgers CIMIC, Newark, NJ, June and December 2002262
• BioMEMS & Biomedical NANOtech World 2002, Columbus, Ohio, September 2002263
• Biodetection Technologies, Alexandria, VA, May 2002264
Additionally, Army Edgewood Joint Service Agent Water Monitor Program is actively looking at
sensors for distribution systems using the following proven classes of detection technologies:
conventional, optical techniques, polymers/materials, assays, and sentinel species. It is also
examining new areas of MEMS (Micro-Electro-Mechanical Systems) and MOEMS (Micro Optical
Electro Mechanical Systems), which rely on a concept proven in one of the other classes.
What was not covered:
• Benchtop products for drinking water analysis.
• Portable products designed specifically for clinical samples.
• Sporadic research papers that have not gained wide attention.
• Technologies/products that were either only conceptual or were not currently envisioned to
apply to water.
In the report, technologies are classified based on the categories above (e.g., available, potentially
adaptable, emerging) and details are provided on the level of verification, proof of concept,
pilot/field tested, or round robin, if the information was available. Note that for most products,
except where noted, manufacture's claims have not been evaluated by independent sources and
products mentioned are not endorsed by EPA. The following is a list of technology products by
the type of contaminant detected (e.g., chemical, microbial, or radiological), the status of the
technology development (e.g., available now, potentially adaptable, or emerging), and the chapter
that it appears in the document; the assay type, and the primary contaminant monitored.
The tables presented in Chapter 9 that compare detectors to desired EWS characteristics (Exhibits
9-2, 9-4, 9-6, and 9-8) are based on information obtained from references cited in the main text in
Chapters 5-8 where the technologies are discussed. In some cases information is incomplete because
the information was not available on the public part of the company website. In cases where there
is no vendor website or the product has not yet been released, information was also not available.
August 2005 C-3
-------
U.S. EPA Office of Water
Early Warning Systems
List of Technologies and Techniques
Product
Company or Developer
Technology Status
Chapter
GENERAL WATER QUALITY
Series CIS Water Quality
Monitoring
Sentinal™
Six-Cense™
Model 1055 Solu Comp II
Analyzer
AquaTrend panel
TOC Process Analyzer
Model A 15/B-2-1
Model 5500
DataSonde 4a
Model Troll 9000
Signet Model 8710
Model 6000 continuous monitor
STIP-Scan
Analytical Technology Inc.
Clarion Systems
Dascore
Emerson
Hach
Hach
Analytical Technology Inc
GLI International
Hydrolab
In-Situ
Signet
YSI
STIP Isco GmbH
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Potentially Adaptable
5 & 9
5 & 9
5 & 9
5 & 9
5 & 9
9
9
9
9
9
9
9
5
CHEMICAL
Quick™ tests
AS 75 arsenic test kit
As-Top Water test kit
PDV 6000 portable analyzer
Nano-Band™ Explorer
CHEMetrics VVR
1919 SMART 2 Colorimeter
Mini-Analyst Model 942-032
AQUAfast® IV AQ4000
Thermo Orion Model 9606
Cyanide Electrode
Cyanide Electrode CN 501 with
the Reference Electrode R503D
and Ion Pocket Meter 340i
Scentograph™ CMS 500
Scentograph™ CMS200
Industrial Test Systems, Inc
Peters Engineering (Austria)
Envitop Ltd. (Oulu, Finland)
Monitoring Technologies
International Pty. Ltd. (Perth,
Western Australia)
TraceDetect (Seattle,
Washington)
CHEMetrics
LaMotte Company (Chesterton,
MD)
Orbeco-Hellige (Farmingdale,
NY)
Thermo Orion (Beverly, MA)
Thermo Orion (Beverly, MA)
WTW Measurement Systems (Ft.
Myers, FL)
Inficon
Inficon
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
6.2.1 & 9
6.2.1 &9
6.2.1 & 9
6.2.1 & 9
6.2.1 & 9
6.2.2 & 9
6.2.2 & 9
6.2.2 & 9
6.2. 2 & 9
6.2.2& 9
6.2.2 & 9
6.2.3 & 9
6.2.3 & 9
August 2005
C-4
-------
U.S. EPA Office of Water
Early Warning Systems
List of Technologies and Techniques
(continued)
Product
CT-1128
HAPSITE®
Field Enzyme Test
Aquanox™
Eclox™
Tox Screen
Tox Screen II
ToxTrak™
Bio Tox™ Flash
Polytox™
Microtox®
DeltaTox®
microMAX-TOX Screen
MosselMonitor®
Bio-Sensor®
LuminoTox
MitoScan
IQ-Toxicity Test™
Daphina Toximeter
Algae Toximeter
Fish Toximeter
Fish and Daphnia Toximeter
Lumitox®
HazMatID™
X-ray fluorescence
SABRE 4000
HAZMATCAD™
Cyranose 320®
Nosechip™
Clam Biomonitoring
Transgenic zebrafish
Fish Biomonitoring System
Company or Developer
Constellation Technology
Corporation with Agilent's
(5973NMSD)
Inficon
Severn Trent
Randox Laboratories
Severn Trent
Check Light, Ltd
Check Light, Ltd
Hach Company
Hidex Oy
Interlab Supply, Ltd
Strategic Diagnostics Inc.
Strategic Diagnostics Inc.
SYSTEM Sri. (Italy)
Delta Consult
Biological Monitoring Inc.
Lab Bell inc.
Harvard BioScience, Inc
Aqua Survey
bbe moldaenke, Germany
bbe moldaenke, Germany
bbe moldaenke, Germany
bbe moldaenke, Germany
Lumitox Gulf L.C.
SensIR
ITN
Smiths Detection
Microsensor Systems Inc.
(Bowling Green, KY)
Cyrano™ -Smiths Detection
Cyrano™-Smiths Detection
U. North Texas-EPA
Great Lakes WATER Inst.
US Army Center for
Environmental Health Research
Technology Status
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Emerging
Emerging
Emerging
Chapter
6.2.3 & 9
6.2.3 & 9
6.2.4 & 9
6.2.4 & 9
6.2.4 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5 & 9
6.2.5
6.3.1
6.3.1
6.3.2 & 9
6.S.2& 9
6.S.2& 9
6.3.2 & 9
6.3.2 & 9
6.S.2& 9
6.3.3 & 9
6.3.4
6.3.5 & 9
6.3.V& 9
6.3.8 & 9
6.3.S& 9
6.4.1
6.4.1
6.4.1
August 2005
C-5
-------
U.S. EPA Office of Water
Early Warning Systems
List of Technologies and Techniques
(continued)
Product
SOS Cytosensor
Portable Cell-Based Biosensor
Portable Neuronal
Microelectrode Array
Dicast®
Fiber optic
MicroDMx™
SAW based sensor
Micromachined Acoustic
Chemical Sensor
Micro-ChemLab CB™
S-CAD
Surface enhanced raman
Company or Developer
Adlyfe Inc.
Gregory Kovacs at Stanford
University
U.S. Naval Research Laboratory
Optical Security Sensing (Optech
Ventures LLC)
Great Lakes WATER Inst.
Sionex
PNNL
SNL
SNL
Science Applications
International Corporation
Real-Time Analyzers
Technology Status
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Chapter
6.4.2
6.4.2
6.4.2
6.4.3
6.4.3
6.4.4
6.4.5
6.4.5
6.4.5
6.4.5
6.4.6
MICROBIAL
Bio-HAZ™
SMART™ Tickets
Bio Threat Alert (BTA)
BADD
RAMP
AMSALite™
Continuous Flow ATP Detector
WaterGiene™
Profile™ -1 (using Filtravette™)
Microcyte Aqua® and Microcyte
Field®
Micro-Flow Imaging
Bio Sentry
Light scattering technology
RAPTOR™
xMAP® / Automated Pathogen
Detection System (APDS)
EAI Corporation
New Horizons Diagnostics
Tetracore
ADVNT
Response Biomedical
Corporation
Antimicrobial Specialists and
Associates Inc.
BioTrace International
Charm Sciences Inc.
New Horizons Diagnostic Corp.
BioDetect
Brightwell Technologies
LXT/JMAR
Rustek Inc.
Research International (Naval
RL)
Luminex and LLNL
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Available
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
7.2.1 & 9
7.2.1 & 9
7.2.1 & 9
7.2.1 & 9
7.2.1 & 9
7.2.2 & 9
7.2.2& 9
7.2.2& 9
7.2. 2 & 9
7.2.3
7.2.3
7.2.4 & 9
7.2.4& 9
7.3.1 & 9
7.S.2& 9
August 2005
C-6
-------
U.S. EPA Office of Water
Early Warning Systems
List of Technologies and Techniques
(continued)
Product
RapiScreen™
Bio Flash™
Smart Cycler® XC System
HANAA
TIGER
RAZOR
Ruggedized Advanced Pathogen
Identification Device (RAPID)
Bio-Seeq™
PathAlert™
BOSS
Spreeta™
M1M
Meso Scale cartridge reader
Quantitative lateral flow assay
(QLFA)
Qdot™
Upconverting Phosphor
Technology™
DynaBeads®
BEADS
Doodlebug
Sen-Z
MAGIChip
Bead Array Counter BARC
GeneChip
VeriScan™ 3000
Bio -Alloy™
"electronic taste chip"
Molecularly Imprinted Polymers
Magneto elastic Sensors
Company or Developer
Celsis-Lumac
Innovative Biosensors
Cepheid
Cepheid
Ibis
Idaho Technologies
Idaho Technologies
Smiths Detection
Invitrogen
Georgia Tech
Nomadics
BioVeris
Meso Scale Defense
NASA
Quantum Dot Co. /EPA research
project
SRI International-OraSure
Technologies
Dynal
PNNL
Biopraxis
CombiMatrix
Argonne/DARPA
Naval Research Lab
Affymetrix
Protiveris
latroQuest Corporation
University of Austin John T.
McDevitt)
Grimes Group
Technology Status
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Emerging
Chapter
7.3.3 & 9
7.3.4
7.3.5 & 9
7.3.5 & 9
7.3.5 & 9
7.3.5 & 9
7.3.5 & 9
7.3.5 & 9
7.3.5
7.3.6
7.S.7& 9
7.3.8
7.3.8
7.4.1
7.4.2
7.4.2
7.4.3
7.4.4
7.4.5
7.4.6
7.4.7
7.4.8
7.4.9
7.4.10
7.4.11
7.4.12
7.4.13
7.4.14
RADIOLOGICAL
SSS-33-5FT
SSS-33DHC
Technical Associates
Technical Associates
Available
Available
8.2 & 9
8.2& 9
August 2005
C-7
-------
U.S. EPA Office of Water
Early Warning Systems
List of Technologies and Techniques
(continued)
Product
SSS-33DHC-4
SSS-33M8
MEDA-5T
3710 RLS Sampler
LEMS-600
OLM-100 Online Liquid
Monitoring System
ILM-100
GammaShark™
Online real-time alpha radiation
detection instrument
Groundwater radiation detector
Thermo Alpha Monitor
Company or Developer
Technical Associates
Technical Associates
Technical Associates
Teldyne Isco
Canberra
Canberra
Canberra
Clarion Systems
DOE, now Los Alamos National
Laboratory
PNNL
Thermo Power Corp.
Technology Status
Available
Available
Available
Available
Potentially Adaptable
Potentially Adaptable
Potentially Adaptable
Emerging
Emerging
Emerging
Emerging
Chapter
8.2& 9
8.2 & 9
8.2 & 9
8.2& 9
8.3 & 9
8.3 & 9
8.3 & 9
8.4
8.4
8.4
8.4
August 2005
C-8
-------
^ §
s- o
2 to
a. •£
"^
o
OB
GC «
Ml ^
.E e
e «
'— U
>>
*
-d
B
S
s
GC
B
«« .2
o -g
91 "On
"S fi
Q 5
^^ O
o
o
"S
3f
OB
B
^H
x.
•d S
B Ml
h a
O *
o ^^
'— 'S
B — -9
0 0
^» fi_J
Project Description
«
.P*
H
^
0
o
(N
^J
oo
rt
5
I
c<3
o
OS
tu
2
o
o
s
Will document successful online monitoring
tu
O
0
en
d
o
"ti
ra
O
*ft
PH
s
S o
w o
^
o
operations as well as problems encountered to
identify operational, maintenance, and calibration
requirements. Will focus on existing problems whi
S
s
be
pH
;-H
0
."tn
d
o
s
d
J^
0<
in'
60
^
s
•g _^
•Ji, .2
O ^-H
£ ^
60
a
includes:
• Poor design of sample-handling systems (pluggi
of sample lines)
• Uncertainty in results and poorly implemented
quality assurance
• Inappropriate application of the instruments,
resulting in misleading information.
0
o
(N
a
'
(N tu
O J3
O in
(N rt
~~-^ ^
•-H OH
T3 60
a S
cS \S
^, rt • -
S -^ S
w -2 .2 ^
yl|l
el> cd ^"
_, _d £>0 i>
ra ^3 ^, Q)
PH O O en
a>
a>
d -9 o y
g d ^ g*
^ f* d en
p to rj cd a>
0 -9 | g S ^
« "i 5 •- g 1=
t~? o ^ "^ £ ^
&: 0 ca S OH •<
60 2
S ri _ tu _^ ^
Developed early warning and source water
monitoring systems for real-time contaminant
monitoring. These systems will allow operators to
predict water quality events that may tax subsequer
treatment processes. This study showed that an
integrated approach that considers all parts of the
water delivery system should be adopted in designi
and operating an early warning system. A general-
purpose, one-dimensional spill model for use in riv
that was developed and applied to the Ohio River
could be easily adapted for use in a wide range of
rivers. Additionally, a systematic method for
designing and operating early warning systems thai
considers the highly variable, probabilistic nature o
many aspects of the system was developed and
demonstrated.
tu 60
TH Q 0 (N
w a c« ' rt Si
<+H ^ tu O ^te
0 60 > 5 m
a -9 '5 7, S
.60 g .a 5 £
D fS £ ?S 60
-------
e
tfe
e
o
s
o
S3 ^>
P
•< _2
X. *U
^ fl> O
^ 5 J ^
i •§.£ 1
Z s x. •«
W ° c
SSl^l
-*H +^ *"
^ a5 «
GC ,*
.S G
e rt
i> ***
>> I
•S ^
^ ^~*
w *
•S |
§^
2 c
S o
S a
S
GC
GC
e
«« .2
o -g
•2 "H.
Q I
O
o
"S
OD
°£
5U
^.
c
hH
^d oS
S3 W3
c^ c^
Contractor
Project Man
Project Description
fl
ca
tl ca
-rf t/!
tu g
S cr
^ <^
a a>
.2 »
N cd
•< O
t/!
0< ^-N
O >^
Foundation and (
PROAQUA (Hal
CRS-PROAQUA recognized Ihe drinking water
induslry's need for a comprehensive resource
^
o> un
*JJ ^~i
^ 0
Q W
a
Q
[/}
[/}
's
S
0
o
Amhersl
^
accurately depicl palhogen occurrence and variabili
in relation lo various sources wilhin watersheds
'1^
5 *
•^ 'T3
C/} o>
E? "2
o ta
•tn |>.
o
S cS
Projecl Manager
Linda Reekie
during and after wealher, hydrologic, or land-use
evenls.
-------
bo
5B
•s
s
o
«
2
H
•< at
'— *u
/^ fl> O
^ 5 J ^
i ^^ 1
Z s x. •«
ttj « o "S
E S ^ §
5? It^
~^N W5 ^
GC ^
Ml ^
.S e
!> ^*
^ i
•s ^
^ ^>
w *
•S |
&^
2 e
S o
E a
!<«
in
e
o "-C
•2 "a.
53 e
Q 1
o
o
"S
3f
a
HH
•d «
e oe
b c
° ^
5« +j
X. y
•^ fl>
o 'o°
0 £
Project Description
^
N
a
"o
t/}
Tj" ,c
0 S
0 &
Q 0
(N 0
^H H
•*
o
0
(N
fl
^
tu O
a '5i
PointSoui
Technolo
And
.E^ 5?
S 60 hJ
'S .S •<
This project was looking to develop a more i
friendly real-time continuous online monitor
method. The multi-angle light scattering (M
cS
13 o
S -£
o i>
d fl to
•2 ^ -g
S O o
"> 1 °
pq H S
1 fc
^ "5
§ 5
.•S tw
Metropol
District o
60
fl
'3
^H
CS
* $
c3
tu tu
t/}
^H ^
3 o
tu fl
•° .2
O T3
™ '§"
tu ^
.5* o
a «
•S o
tu o
S
^
S
o
*: o
&H (N
^, t^
o *
<*
t-^
fTl
Californi;
contamination outbreaks since it enabled the
i-H
60
a
cd
tj
o1
OH
*^
a g
•2 &
-C ^— i
u *s
« T3
estimation of the Cryptosporidium oocyst de
limit. MALS can also discriminate between
a
M
ffi
1>
Is
S
£
physical states of oocysts, including oocysts
^
g
O
treated with ozone, heat treated, or excysted
3
untreated oocysts. The technology tested in
^ I
o >-,
project has the potential to provide an impac
drinking water industry through distribution
monitoring, treatment optimization, end-user
60
rt
o
protection and monitoring, and influent mon:
and source water selection.
CO
Q
-------
•3
tit
a
o
s
o
S3 ^£
P
•< at
'— *u
^ fl> O
^ 5 J> <$
2^2 i
Z s •— «
W ° c
£f!l
•^ 5^5 ^
in K*
.S e
!> ^*
^ i
•s ^
^ ^>
w *
•S |
&^
2 e
S o
S a
s
in
GC
o
-2
Q
o
es
3f
"5
§J
e
HH
•d
e
^
o
cs
•—
"S
o
O
e
'•C
"H,
S
o
O
S
§0
e
5
o
•••*
o
£
Project Description
^p
a
c/5
While some unconventional options have been
a
13
o
S
o
O
ca
•e
ca
'3
5-n
O
O
examined in connection with specific contaminants
ca
a
o
a
a
0
0
fl
^
fl
"o
[/}
ft
0
*^
o
H
t/}
1
60
^
fe
^
•3
^^
i-I
60
ca
ca
S
•^
i>
o1
OH
and regulations, future regulatory scenarios have not
1
o
M
o
ft
^
IT)
O
o
(N
en
•.•H
n
^
been taken into account. Unconventional approaches
a
o
a
OH
0>
0
t/!
may turn out to be more cost effective and feasible in
t—
*
meeting new stringent standards. This project will
compare conventional and unconventional water
treatment and distribution providing quality water to
customers, including POU and point-of-entry (POE)
devices, small neighborhood systems, and bottled
water. Will include capital costs, operational and
maintenance costs, ability to meet health standards,
and aesthetic quality goals. Will also consider risk
associated with the options, long-term reliability of
the options, and practical engineering considerations.
Will assume plausible stringent future regulatory
scenarios and consumer demands for high quality and
aesthetically pleasing drinking water. See also
Project #2924,
-------
bo
5B
-5 _
A) M
•o-B
£ 3
JS S
o S
S- O
1 •=
Pi 2
« S
^Q S
w f*s
11
•< .2
X. *U
/^ fl> O
^ 5 J ^
£•§•£ §
z s x. •«
W ° c
SSl^l
-*H +^ *"
^ a5 «
^^ ^^
GC ,*
.S e
Summary of Early Warn
Sponsored by America
e
«« .2
o *-g
** "ft
Q |
O
o
3f
"1
^.
e
M
L
•d S
S3 W3
^ ^
o .2
+j 5
^* "o
o 'o5
^ SM
e
o
Project Descripti
7:
N
O -i-
^ o
i O
Q M
O .a
O to O
o a> ^
O-) T3
1 I «,
1> -^ cd i>
1 1| ! i
S w 6 £ <
s «
* | | o § .1" 'B «§ „ s 1 . i ^|
en r^T ^ ^ "~^ ^ ^ I-^H* "Hi W -<-> ^ "S
^ -5 ^*^ ^ ' — ' P^ d ^H ^^ CH ^ M ^ ^J ^ D
•^ O^^ | §'.2'3-S'0 |^'o SH to ^
"g'l^S^^.S g g i-i^l g S-^S
iiilliilliliflfi!!
sgxl§^H^||»lr1^^
1 11^11-^1^1 pl?l ll
SlsI§1gll.g^IltSfc§lll
c-
r-
r-
^ o 5^
O f . c/}
x«s 2 §
3 2 S "o
h to 60 0
iJ 0 rt "S
Trt M S t/}
> ™ ^ °
1* •< 2 •<
a>
.^
3
O 1>
o ^o _
rj ,_, g
5^ ^ ^
"Z,
a
£
a>
fl
OH
fl
X
0<
i-H
60
CS
a
CS
> ^
ti ^
TO U O
.-? 51 S
S a
^H '* a> X
0> rf en SH ti
& § a § '3
'.S T3 'fl B s
> '> o » g
0 T3 0 8 g
1£* ^ §
11 develop and evaluate new and i
terns to rapidly detect chemicals 1
sses), radioactivity, pathogens, or
ter in order to help make these sy
ble for use by the drinking water
•j"1 E? a a a
&: £"o * '>
,_
fO ^
i>
S £ ^
o> p> i>
I.^
H 60
\S > 'in
S f> 0
cti
5 w S
-------
bo
e
o
s
o
fc
« S
3 S
H
•< _2
X. *U
/^ fl> O
^ 5 J ^
i ^^ 1
Z s x. •«
W ° c
£f!l
•^ 5^5 ^
GC ^
Ml ^
.S e
!> ^*
^ i
•s ^
^ ^>
w *
•S |
&^
2 e
S o
a a
!<«
GC
e
«« .2
OJ «
•^ CX
S"1
Q_
Q
y
o
53
.Sf
^
>
HH
Jj
S3 W)
C
II
CS i)
X. y
o 'o°
U Cu
Project Description
«
+^
N
d
T3
0>
JU
0
g
0
0
tu ^
"° o
0 o
H (N
~ !-H
<: B
, qn ^ .?
"^ ' " O C*
OH ™ - \"
PL] d .p '-P '-P
C3 -^H TO TO
!-; & s a a „
° '3 B .S .S S
S '-S > u o is
G G 'g d d o
OH OH iD O O fS
cs
OH*
In"
bO
<+H c3
o s
|| tL
'3 .S °" -3
£ b OH ^
o a
Will develop and evaluate new and innovative
systems to rapidly detect chemicals (individual
classes), radioactivity, pathogens, or biotoxins i
water in order to help make these systems more
viable for use by the drinking water community
to "">
q^ -t-^ OO
° s s
d 60
.2 •< £
Q ^ r~j
t> 'd 2
Q 0 [/>
H
3 S B
ft o S
CS -S K?
OH CQ l>
d
T3
0>
t>
"ft
g
0
0
^ S" a
o S a §
a> o 2 HH
^ g: S £ «
"o -S o « ^
Jj ^H _O "^? M
ts o 15 g • -
pq ^ _j £ S
TJ ' a>
ta ri g ^ !-H — i
Some of the difficulties in developing a univers
method include the physical differences betwee
pathogen groups, the need to concentrate large-
volume water samples to detect low target
concentrations of certain pathogen groups, the r
to remove co -concentrated inhibitors from the
sample, and the problem of standardizing a cult
independent endpoint detection method. This p
will design a DNA microarray as the endpoint
detector that simultaneously targets several wel
characterized virulence genes such as E. coli
O157:H7 and Cryptosporidium parvum.
<,- « £ "
^D ^ ^
Q cfi
'g o -^ ,u -9
fl *». >^ o X rt
0 Z? 60 S -P
0>
"ft
g
0
0
0>
"° o
0 o
H (N
< 1
^ o i>
PH "S
02 cJ '^ ^
" ^ o "« t/1
PH LJ ^ PH ri
l > O g
§ <; "d ffi o o
OH 60 ^) O O S
•S °!
0> 2
g ffi -a u;
1 1 -o S" S
<; S ^ cs g
^H >^~l I^-H H
g s 1 1 S g
£ § 1 « > 8 8
gu-§ s^'a g1^
t>0> s g
S *3 2
td O S
'7^ ° O
^ M -r^
« ° " ^
^4 ^ o> o
S S Q *
-------
5B
-5 _
A) M
•o-B
£ •§
JS S
o S
H o
H
•< _2
X. *U
fl> O
£•§•£ §
Z S :- «
W ° C
&l!i
•^ 5^5 ^
O5 K*
.S e
!> ^*
^ i
•s ^
^ ^>
w *
•S |
&^
2 e
S o
S a
s
in
GC
fi
<_ so
o ^
dJ ^"i
** D.
Q S
W Q
O
o
"S
.Sf
'"S
^
e
HH
1*
•d S
S3 W3
C3 S3
X. *•
o ^
-S s
0 ^
c3 ^
X. o
G .2,
0 0
y£
Project Description
«
N
fl
•5
tu
&
0
0
_Q ^~l
o
0 0
H (N
S
o
• 1
• ^3
Q o>
o ™ ^
L" & r* ^
a5 '" 3 °
d ° ^
£2 Vjlj jU
c<3 c3 ^ ~Q
OH OH ffl S
tu
^ !-H
0 d S
o S ^
•^H ^^ TO t/1
D -i-H <5 HH
^ cn ffi
^ S t> cd
fe > W ^
S "S g1 .s
O '^
tu ^
fl 'o1 . 'S
Will screen three to five different water extractio
methods for biological agent surrogates. This pr
will develop a portable water monitor, preferabl}
hand-held, that is capable of real-time detection t
harmful substances, while providing no false-
positives.
VH
Extraction Methods
Early /Real-Time
Warning Systems fo:
Biological Agents A
#2908
_u
i
'3
m >
0 -<
(N G
--- o
2 &
, i-H
a o
0 » f a
'is -S " -^
rf ^ ts
T3 O ^ ^
fl -° 'o1 §
rt TO ^H ^H
60 i-J OH fe
^H -t-J
Captures lessons learned and provides a forum f(
information exchange on vulnerability assessmei
conducted by large drinking water utilities.
Results From the
Water Utility
Vulnerability
Assessment Lessons
Learned Study #290'
U-
.„ D
o a
o -rf
fN >-,
.S "3
0 «-> JH
Q ^ r~ | (— 1
Q .2 ca S
^ S ^ S
S
.fa
tu
& VH
S &
O tu
O OH
cf ™~
'•go"
U N ^
•^ -H tu
c3 ca *
^H tU
"Nf O t/}
• cS
ri ^3
^H ri
60 M
S in"
'*3 tu
"3 60
a «
0 ^
O 2 ™
g 0 I
•2 «-> s
60 OH 51
qH
O O5
(—1 u
o ;-S «>
Provide utilities with a security-oriented evaluati
water service provision options that they may be
considering now or in the future. Also helps util
evaluate and plan short-term emergency responst
options using bottled water, POU, and other
approaches (see also Project #2761).
!-H
M f^
Security Implication
of Innovative and
Unconventional Wa1
Provision Options
#2924
-------
bo
5B
-5 _
A) M
•o-B
£ ^
JS S
o S
H o
H
•< _2
X. *U
/^ fl> O
^ 5 J ^
i ^^ 1
Z s :- «
W ° C
&l!i
•^ 5^5 ^
O5 K*
.= e
& ^
i> ***
>> i
•s ^
^ ^~*
w *
•S |
&^
a I
s a
s
GC
GC
O
a,
o
oj
•^
SS
Q
o
es
3f
"&r
-j
>•
•d
e
^
o
^
x.
"S
o
O
e
'•C
fl^
"H,
S
o
O
o
DC
e
,2
^^
'o5
PM
e
'•C
.&
X
Q
1
'o5
«
N
a
"o
tu
"Pn
0
0
JO
o
H
OH
W
i-H
a
e
cS
OH
a
60
.3
"3
S
o
r )
*^s
s
c/5
Emergency management planning consists of four
tu"
^
o
&
t/}
tu
»H
t/}
t/}
Q
o
(N
1> en
•3 a
O «
(N .«
^j C
d ^
od
-3
^^
In"
M
^
ca
tu
o1
OH
principal functions: planning, crisis management,
consequence management, and mitigation. Better
.•^
^
T3 '-g
„ o
" s
0 .S
O en
tU $
OH 02
tu
."tn
^3
tu
li
en
en
od
1
preparation will minimize damage and financial
losses, reduce recovery time, and improve public
tu
cd
^ cs
££
a <*
'a +2
_ca \3
S ^
credibility. The ability to respond and recover is
greatly enhanced if a water utility has developed in
conjunction with stakeholders a disaster response,
T-*
recovery, and business continuity plan and conductec
routine training.
0
(N
a
o
ju
"ft
B
o
O
•§ Q
5^
tU JH
"3 ^
> ^
jo
C3
r^
cS
O
&
tu
04
60
•^
ca
0
•j-^
cS
&
1
60
intrusion. Also creates reasonable scenarios outlinin
m
(N
X
g
t/}
t/2
t/}
d
o o>
'Zl '^
a ™
.S -S
s ^
03 ^
£H li
O en
In"
bO
od
d
od
0>
o1
OH
the consequences of purposeful contamination to the
drinking water system. Develops recommendations
od
od
S
ca
u.
to reduce the vulnerability of critical elements and
sub-elements of the distribution system.
oo
Q
-------
bo
5B
-5 _
^ M
•o-B
£ -3
A S
o S
H o
H
•< _2
'— *U
/^ fl> O
^ 5 J ^
i •§.£ 1
Z S x. •«
W ° c
SSl^l
-*H +^ *"
^ a5 «
O5 K*
.= e
c ^
i> ***
>> i
•s ^
^ ^>
•s I
&^
a I
s a
s
GC
GC
e
"3 B
« ^
C^ fl
Q §
O
IM
o
Ml
1
e
HH
S3 W3
^ ^
. 23
o .2
•** *2
o «
x. y
•^ flj
fl «"^
0 0
^ 3
PH
Project Description
«
•^
N
a
1
it
g
0
0
ti
H B
cd en
O hS
, — ,
<:
Supervisory control and data acquisition (SCAD.
systems are increasingly susceptible to hacker
intrusions. The Gas Technology Institute (GTI)
!-H
<2 ™
*t I
;-H M
0 j/j
C/} <1^
^ Q
U ^ ^o
^ "^ C^
0 !/! =tfc
o M
^ 0<
t/2 PH
H W
S-H
bO
cd
PH ^
S ^
1> ^
••— i PH
O cd
^H ^H
OH PH
d>
r^H
recommended in 1999 that the gas industry adop'
Digital Encryption Standard (DBS), the Rivest,
60
O
rt
Gas Tec!
Institute
a
T-) 01
ri t3
Shamir, Adelman (RSA) public key algorithm, ai
the Diffie-Hillman number generating algorithm
^ a
§ g
1 1
1 *
a S
Chicago
of Water
suite of algorithms. This cyber-attack protection
includes preventing assailants from studying the
"2
N
'r^
system by listening to the communications, from
altering authorized commands to lead to unautho
T3
fl
operations, and to originate unauthorized comma
It is also possible to prevent a single insider from
performing an unauthorized operation.
1
"ft
g
0
0
^
"o1 -S
OH S
2
"3 -i-1
_Q PH
shortening the extraction step. This project will 1
upon research being done by the DOD on the Joi
+^
bO
13
u
"S 00
o ^
"o g!
ffl ffl
Service Agent Water Monitor (JSAWM).
-------
60
U _
A) M
•o-B
£ ^
A S
o S
H o
H
•< _2
X. *U
/^ fl> O
^ 5 J <$
i ^^ 1
Z S x. •«
W ° c
SSl^l
-*H +^ S"
^ a5 «
O5 K*
.= e
e «
i> ***
^ i
•s ^
^ ^>
•s I
&^
a I
s a
s
GC
GC
e
«« .2
o •**
OJ «
•^ ex
^ c
Q §
O
o
"S
3f
'S
V
e
HH
•d «
S3 W3
S3 ^
^ S
O £
+* ^2
u «
a +1
'— o
C — -9
O O
y *•«
Project Description
«
*j
H
T3
tu
t>
"ft
g
0
0
tu
0 o
H (N
Q
OH
Q
O
Q
O
i-H
fl
a
OH
&
a
t/2
Q
•^ ^
cu -~*
^ tu
> ^
tu '3
£ ^
This project will determine the feasibility of using
POU drinking water devices in the event of a
60
^
M .
rt
D 0
t/J -i-H
^u r*
'o ^
ri ^
<£ ^
i-H
60
fl S
cd t/i
^ cd
^ ffi
0 CS
tu ^
'o1 .2
£ S
bioterrorism attack to help identify the contaminant,
contaminant dissemination, and the public health
impacts. POU devices exist that remove
f
o
0> pH
^< o
pq -^
cd
bo ' — ' d
"w ^ S
en o cd
1> (H -^
en O H
•< S o
contaminants and have their performance certified by
credible testing regimes administered by independent
organizations, but have not been tested and certified
r^
S
tu
t/}
c/3
fl
0
s
0
43 oo
Q *
for their effectiveness in reducing chemical or
biological sabotage agents. Several studies have
researched extraction of bacteria from granular
activated carbon cores, and will constitute a starting
point for this project. Some standard methods utilize
fiber filters as a means to concentrate microorganisms
(e.g., Cryptosporidium) when large water volumes
have to be sampled. The intent of this study is to
better understand possible patterns of delivery of the
contaminant in the distribution system.
T.J
tu
tu
"ft
g
0
0
^ t-
o
0 o
H (N
^
OH
w
i-H
tu
fl
a
OH
^
a
^
• ^
.^
OH
g
"o
13
^
The data collected from sampling and monitoring
networks in distribution systems are often of limited
h
tu
ri ^
a ^
ri ri
"£ ">
t/} O
tU IH
•< 5
>,
^
i-H
60
a d
§ «
^ 60
tj ^
^ ^
'o1 fl
£ j
use for detecting and diagnosing important changes in
water quality. This project will develop methods and
tools to help utilities scientifically evaluate existing
60
fl
60 "rt 0
•S 5-H t^
% .a £
rt 05
CX3 M "^H
^ Cd !_
•i— 1 ^H ffj
rj O 03
ai ^
sampling plans and improve them. This project will
also include the development of procedures and
algorithms to address multiple purposes and benefits.
t—
o
*
S
S
t/!
-------
bo
e
tfe
5B
•s
s
o
9)
2
H
•< _2
X. *U
fl> O
£ *
~ ^
Q -d
% s
^ e
fe '-C
,® G
^ 2
+i —
58 ^
O5 K*
.S e
i> ^*
^ i
•s ^
^ ^">
w *
•S |
§^
2 e
S o
E a
BOJ
in
o
a,
e
«« .2
OJ «
•^ CX
S"1
M O
O
o
"S
3f
a
HH
•d %
S3 W!)
c^ c3
s a
Is
c^ ,^
S- 0
E.2,
0 0
^£
Project Description
«
N
fl
T3
o>
t>
"ft
g
0
0
.0 p"-
o
0 o
H >
o g o a § si
!> G3 PH _O rt ^
O s^ ^ ^ 2 ^
o s S £ -a ts -9
tt! g fl 2 cs «> |-J
« 0 .2 ^ 60 o' X
[-'-Jr'lOrH!~( LLlJ-J
o B so hS o £ S
[/} ^H
T3 --H CD 1>
Research is needed to explore data-processing
methods that can distinguish normal variability in
distribution system conditions from patterns relate
to unacceptable water quality trends or events. Th
project will develop a general data processing
approach to assist water quality managers and wat<
system operators detect abnormal patterns in onlin
monitoring data.
•g o g
Kj S 4> tj~}
60 -a f> s
•S O c/3 g}
1 £ I s>
° ^H "5 "^
1 til
Q •< Q S
T3
0>
0>
"ft
g
0
0
1>
"° o
0 o
H
g ^^S'g^
^) W OH tt! OH N
Will develop a system of integrated, hierarchical
models that not only describe the fundamental
processes that determine water quality in pipe
systems, but also simulate the temporal and spatial
behavior of water quality in the complete network.
0 0
X '^
S 5
°s 'C °°
£ Q *
d
^ fli ffi
O cd ^
^ P> r/}
-------
bo
U _
A) M
•o-B
£ -3
A S
o S
H o
« «
^ ^s,
M
•< _2
X. *U
/^ fl> O
^ 5 J ^
i ^^ 1
Z s :- «
W ° C
SSl^l
-*H +^ S"
^ a5 «
GC ,*
.S e
c ^
i> ***
^ 1
•S ^
^ ^>
•s I
&^
2 c
s 2
S a
s
GC
GC
e
•« .2
« ^
C^ fl
Q §
O
o
"S
'If
«
e
HH
•d «
S3 W3
b C
° W
"5 S
W ^^
c — ^
0 0
PH
Project Description
V
N
a
1
"B<
g
0
0
tu
,0 ^D
0
0 o
H (N
"o j~-
S w
tu g»
^ g a
P* S 2
i-H ^ Ctf
tU g T3
G '> §
CS fl O
OH W tin
•3
^a
i-H
W)
a
i5 i
*• o rt
<; o -S
Jj -^i ^
S °
w £ S1
tu
tu £
Will identify, organize, prioritize, and describe
most probable electronic security threats; risks
associated with the various security weaknesses
,_ ^
^ t/>
M T3 S
^ ri tu
2 tS to
| 1 |
^*~l "t^ Cd
• a & S
§ a 1
0 ° 5
I/! O <,
T3 tU
tu ^
N 2
available technology that can prevent unauthori:
and willful attacks; secure and stable infrastruct
.2
;tn
•T^
a
U-
T3 ^
' — " ^
« iJ
^S K^7
-' ^
^H
-------
bo
e
o
s
o
fc
H
•< _2
X. *U
/^ fl> O
all?
Q^^ c
Z S x. •«
W ° c
SSl^l
-*H +^ S"
^ a5 «
GC K*
.= e
e rt
i> ***
^ i
•s ^
^ ^>
•s I
&^
a I
s a
s
GC
GC
e
«•- 2
o '•£
a> >-«
^^ ^^
Q 5
^^ O
o
o
"S
•-c"
IK
e
HH
1*
•d 5
S3 W3
b c
O ^
•^ ^5
c$
x. y
•^ flj
C — -S
0 0
U ^"
Project Description
«
+^
N
T;) -S g
" « 13
1) On 5-n
'Hn J* »H
r- T3 .0
n o> ^
O 0> ^H
0 fl 0
of — O
o § S
H (N &
tu
.. S
^ o -^
O ^H
(2 o ^
• ^
3
^N OH
O
C/3
a .2
O t/}
-^ t/}
t/} -^
rf 8 JS
"3 s "S
TO H Q
J^ 0 K?
O O ^
ca cs
40 t!
/ill provide a broad and substantial knowledg'
)r utilities about the impact of a toxicological
n a distribution network, and cost-effective
f- q3 o
^
S
C tu to
° ^ X
ft 5: ^
60 g O
2 'g J
.2 « 'C
OJ M ^
Q 60 Q
cS
S
Q
i.s
< S
i-H
tu
60
CS
ti cd
cS J^
S J3
^ ffl
•2^i "fl
O cd
OH U.
Q
CS
pproaches for detecting and mitigating such a1
a
oo
o
*
-&*
"§
Q
t/!
tu
fl
'g
tu
tu
T3
tu
JO
0
H
£
•< "o
^ g
, J t/}
pfi 0>
I— (
0> !-H
ri i^
a ^
S-H"
U
c3
:3 -^H
S "S
02
"t^
o> ^
O cd
OH PH
tM
&0 0
p3
1 1
/ill integrate research on sensors and early wa
^sterns to enhance benefits and cost-effectiver
^ »
*H £
(2 S
q-H ^U
r- t/}
S o?
60 M
P d
t-H .3
"o ^
^ ^
5-H '
60 X
tu 15
S TO'
^ P-]
tu
S
rt
•3
ie research. Will consist initially of nine suboi
;udies:
"is t/i
}_,
-H
^
t>o
cS ^
^ ^
O 60
3 tJJ
t/! t/!
tu
fl
rt
Online Monitoring Using a UV-Probe
Organic Contaminant Detection Using an Oi
• •
m
o
m
i
B
t/!
X
&
&
t/!
HPLC and GC Analysis
Combined Daphnia and Zebrafish Monitor
Detection of Cholinesterase-Inhibiting
• •
"s
o .
Contaminants
Feasibility Study on the Use of Chemical-O]
Sensors for the Detection of Organic
•
2
"s
Micro-Pollutants
Annual Reporting on Developments in the F
of Sensors and Detection Techniques Data
•
tu
|o ?
60 0
Handling for Detection Systems Containing
or Multiple Sensors, and Procedures for Foil
on Alarms
T3
a
Strategies for Network Clean-up After
Contamination
Exploring the Opportunities for Co-funding
Collaboration in Sensor Development and
•
Implementation
-------
bo
5B
•s
s
o
«
2
H
•< _2
X. *U
/^ fl> O
^ 5 J ^
9^^ i
a.
o.
fe '-c
,® G
^ S
b ^^
v
*
in
.S e
!> ^*
^ i
•s ^
^ ^>
w *
•S |
&^
2 e
S o
E a
BOJ
in
O
-M
Q
o
"S
Ml
°£
CU
>•
e
•d
e
Contractor a
e
'•C
.&
^
&
Q
, ^
"5
'o°
e
'•C
"H,
S
o
O
V
DC
Project Mana
«
N
a
|
a>
a>
a>
0
H
a
.u
u
g
a>
a
e
cS
OH
O
Virginia Polytechn
'^
S
^
•^^
o>
'o1
T3
to
^
•^H
rt
o
&
c3
^H
d>
T3
a
3
, i
r^
^
1U
-^
o
5
o
OH
to
•^
0
^
5^
^
^_
^
Institute & State
.a
13
cr
0
cd
^
a>
g
"a
o
d
•^
1
"ft
g
o
o
!-H
a>
g
o
3
S
60
a>
rt
0
a
o
60
a>
rt
d
o
S
'^
o
[/}
^
University
0
r1H
CXD
a>
?
u
; —
g
en
bO
^
cS
a>
a
0
S
2
•3
.g
'cS
"ft
g
0
O
^H
u
g
O
M
S
O
cd
•3
i-H
a>
60
a
cS
tj
o1
OH
_^H
0
^
60
a
^H
>
0
u
[/}
a>
|-S
S
r^^
N
g
•3
g
a
cS
a>
60
c«
a>
rt
"a
O
>
'S
Q
Williams
u
'C
M
aj
1
0
0
T3
a
£H
istributic
T3
s
g
60
0
60
<
5
Q
x
cs
s
a
f— i
a>
cd
£
ndwater*
3
o
60
§
a>
5
60
1
T?
W
0
g
t/!
-------
U.S. EPA Office of Water Early Warning Systems
ENDNOTES (including website references)
1 http://cfpub.epa.gov/safewater/watersecurity/home.cfm?program id=9
http://www.epa.gov/safewater/watersecurity/pubs/action plan final.pdf
2 http://www.whitehouse.gov/news/releases/2004/02/20040203-2.html
3 http://www.whitehouse.gov/homeland/book/
4 http://www.whitehouse.gov/news/releases/2004/02/20040203-2.html
5 http://www.epa.gov/ordnhsrc/index.htm
6 http://www.epa.gov/etv/
7 http://www.epa.gov/ordnhsrc/news/news031005.htm
8 http://www.ewrinstitute.org/wisesc.html
9 http://www.epa.gov/safewater/security/index.html
http://cfpub.epa.gov/safewater/watersecurity/home.cfm7program id=8#response toolbox
10 http://cfpub.epa.gov/safewater/watersecurity/home.cfm7program id=9
http://www.epa.gov/safewater/watersecurity/pubs/action plan final.pdf
11 http://www.epa.gov/ordnhsrc/pubs/fsTTEP031005.pdf
12 http://www.awwa.org/conferences/congress/
13 http://www.awwa.org/education/seminars/index.cfm?SemID=47
14 http://www.ewrinstitute.org/wisesc.html
15 http://www.infocastinc.com/tech/rapid.html
16 http://www.who.int/csr/delibepidemics/en/chapter3.pdf
http://www.who.int/csr/delibepidemics/biochemguide/en/index.html
17 http://www.verdeit.com/VPages/SpiralDev.htm
18 http: //www. aoac. org
19 http://www.stowa-nn.ihe.nl/Summary.pdf
20 http://www.epa.gov/ORD/NRMRL/wswrd/distrib.htmtfTable%202.0%20Proposed%20DSS
21 http://www.hydrarms.com/brochurepdf.pdf
22 http://www.waterindustry.org/Water-Fact/Hach-l.htm
23 http://www.tswg.gov/tswg/news/2004TSWGReviewBookHTML/ip pl8.htm
August 2005 E-l
-------
U.S. EPA Office of Water Early Warning Systems
24 http://www.nsf.gov/eng/general/sensors/vanbrie.ppt
25 http://www.epa.gov/NHSRC/pubs/fsPureSenseCrada060204.pdf
26 http://www.7t.dk/company/default.asp
27 www.waterisac.org
29 http://www.hach.com
30 http://www. hach.com/hc/search. product, details. invoker/PackagingCode=69 5 0000/NewLinkLabel=
Hach+Event+Monitor+Trigger+System/PREVIOUS BREADCRUMB ID-HC SEARCH
KEYWORD/SESSIONID|Bmd5TURZME56SWlaMlZsYzNSUFZVOUNWREV4TVE9PUNEazV
31 http://www.instrument.org/Newsletter%20Articles/Summer%202003/Innovative%20Monitoring.pdf
http://www.swig.org.uk/Nick%20Sutherland.pdf
32 http://www.emersonprocess.com/raihome/liquid/articles/06-14B-2004.asp
http://www.emersonprocess.com/raihome/documents/Liq Brochure 91-6030 200408.pdf
http://www.emersonprocess.com/raihome/liquid/products/Model 105 5.asp
33 http://www.afcintl.com/water3.htm
34
http://www.clarionsensing.com/howitworks.shtml
35 http://www.elscolab.be/e/Stipscan%20iem%202003%20iulyl.pdf
http://www.elscolab.be/e/stipscan.pdf
http://www.stateoftheart.it/STIP-Buoy-SCAN.htm
http://www.baumpub.com/publications/arc/cep 04mav/avensys.htm
http://www.epa.gov/watersecuritv/guide/chemicalsensortotalorganiccarbonanalyzer.html
36 EPA Press Release May 19, 2004.
http://yosemite.epa.gov/opa/admpress.nsf/blab9f485b098972852562e7004dc686/754a0739ba5bc9c9
85256e99005a9f60!OpenDocument
37 http://www.epa.gov/ORD/NRMRL/wswrd/distrib.htmtfTable%202.0%20Proposed%20DSS
38 Phone conversation with Dr. Ryan James, Battelle Laboratory Advance Monitoring Systems, Center
Verification Test Coordinator, 19 August 2004; Interviewer - Stanley States
39 Presentation by K.L. King, Event Monitor for Water Plant or Distribution System Monitoring; Hach
Homeland Security Technologies; AWWA Water Security Congress; April 25-21, 2004
40 http://www.inficonvocmonitoring.com/downloads/pdf/haps-smart.pdf
http://www.inficonvocmonitoring.com/downloads/pdf/situprobe.pdf
http://www.inficonvocmonitoring.com/downloads/pdf/Scentograph%20CMS200%20Brochure%20-
%20Screen.pdf
41 http://www.contech.com/Chemical Detection Products.htm
42 http://www.awwa.org/education/seminars/index.cfm?SemID=47
August 2005 E-2
-------
U.S. EPA Office of Water Early Warning Systems
43 http://www.epa.gov/etv/pdfs/vrvs/01 vr eclox.pdf
http://www.quotec.ch/services/qeclox.htm
http://www.wateronline.com/content/Downloads/SoftwareDesc.asp?DocID={13C2390F-C9Bl-4362-
A1FO-4C3BDDD04B97}
44 http://www.randox.com/products.asp
45 http://www.epa.gov/etv/verifications/vcenterl-27.html
46 http://www.checklight.co.il/pdf/manuals/ToxScreen-II%20manual.pdf
47
http://www.epa.gov/etv/pdfs/vrvs/01 vr toxscreen.pdf
48 http://www.hidex. com/index. php?a=4&b=12&c= 12
49 http://www.epa.gov/etv/pdfs/vrvs/01 vr biotox.pdf
50 http://www.azurenv.com/dtox.htm
51 http://www.epa.gov/etvprgrm/pdfs/vrvs/01 vr deltatox.pdf
http://www.epa.gov/etv/pdfs/vrvs/01 vr microtox.pdf
52 http://www.epa.gov/etv/pdfs/vrvs/01 vr toxtrak.pdf
53 http://polvseed.com/html/polytox.htm
54 http://www.epa.gov/etv/pdfs/vrvs/01 vr polytox.pdf
55 http://www.mosselmonitor.nl/
56 http://www.biomon.com/biosenso.html
57 http://www.sparksdesigns.co.uk/biopapers04/papers/bsl71.pdf
http://abstracts.co.allenpress.com/pweb/pwc2004/document/?ID=42895
http://www.alga.cz/mk/papers/bios 02.pdf
http://www.lab-bell.com/main.jsp?c=/content/gestiondeseaux en.html&g=left produits en.html&l=en
http://www.lab-bell.com/main.jsp?c=/news/new.jsp&n id=30&l=en
58 http://www.alga.cz/mk/papers/bios 02.pdf
59 http://www.mitoscan.com/Applications.htm
http://www.mitoscan.com/technol.htm
60 http://www.detect-water-terrorism.com/
61 http://www.epa.gov/etv/pdfs/vrvs/01 vr aqua survey.pdf
62 http://www.bbe-moldaenke.de/
63 http://www.bbe-moldaenke.de/
64 http://www.bbe-moldaenke.de/
August 2005 E-3
-------
U.S. EPA Office of Water Early Warning Systems
65 http://www.lumitox.com/bioassay.html
http://www.dewailly.com/LUMITOX/lumitox.html
http://www.bioinfo.com/dinoflag.html
66 http://www.smithsdetection.com/PressRelease.asp?autonum=25&bhcp=l
67 http://www.sensir.com/newsensir/Brochure/ExtractIR%20Product%20Note.pdf
68 http://www.hazmatid.com/
69 http://www.itnes.com/
70 http://cfpub.epa.gov/ncer abstracts/index.cfm/fuseaction/display.abstractDetail/abstract/7477/report/O
71 http://www.etgrisorse.com/pubblicazioni/contamination.PDF
http://pms.aesaeion.com/ionpro/Products/theory
72 http://www.chemistry.0rg/portal/a/c/s/l/feature ent.html?id=7635201a690al Id7f2al6ed9fe800100
73 http://www.emedicine.com/emerg/topic924.htmtfsection~ion mobility spectroscopy
74 http://www. smithsdetection.com/prodcat. asp ?prodarea=Life+sciences&bhcp=l
75 http://www.healthtech.com/2003/mfl/index.asp
76 http://www.memsnet.org/mems/what-is.html
77 http://www.memsnet.org/mems/beginner/
78 http://www.biochipnet.com/EntranceFrameset.htm
79 http://www.nano.gov/index.html
80 http://www3.sympatico.ca/colin.kydd.campbell/
http://www.sensorsmag.com/articles/1000/68/main.shtml
81 http://www.microsensorsystems.com/index.html
http://www.army-technology.com/contractors/nbc/microsensor systems/
http://media.msanet.com/NA/USA/PortableInstruments/ToxicGasandOxygenIndicators/Hazmatcad
AndPlus/H azmatcadProductAnnounce.pdf
http://www.raeco.com/products/toxicagents/hazmatcad.pdf#search='HAZMATCAD'
82 http://hld.sbccom.army.mil/downloads/reports/hazmatcad detectors addl info.pdf
83 http://www.smithsdetection.com/PressRelease.asp?autonum=12&bhcp=l
84 http://www.ias.unt.edu/~jallen/littlemiami/Clam Page.html
http://www.ias.unt.edu/~jallen/clampage.html
85 http://www.uwm.edu/Dept/GLWI/cws/projects/carvan.html
86 http://usacehr.detrick.army.mil/aeam/Methods/Fish Bio/
August 2005 E-4
-------
U.S. EPA Office of Water Early Warning Systems
87 http://www.adlyfe.com/adlyfe/home.html
88 http://www.nrl.navy.mil/content.php7P-04REVIEW118
89 http://www.uwm.edu/Dept/GLWI/cws/
90 http://www.optech-ventures.com/products.htm
http://www.intopsys.com/markets brochures/Continuous-CableFOCSensor.pdf
91 http://www.sionex.com/technology/index.htm
92 http://www.saic.com/products/security/pdf/S-CAD.pdf
http://www.saic.com/products/security/s-cad/
93 http://www.sandia.gov/media/acoustic.htm
http://www.isa.org/Content/ContentGroups/InTech2/Features/20012/2001 October/Surface
acoustic waves to the mission control/Dangerous chemicals in acoustic wave sensorsandNum82
17; future.htm
94 http://www.sandia.gov/mstc/technologies/microsensors/flexural.html
95 http://www.sensorsmag.com/articles/1000/68/index.htm
96 http://www.mdl.sandia.gov/mstc/documents/uchembrochure.pdf
http://www.ca.sandia.gov/chembio/factsheets/chemlab chemdetector.pdf
http://www.sandia.gov/water/projects/ChemLab.htm
http://www.sandia.gov/water/FactSheets/WIFS SensorDevNew.pdf
http://www.ca.sandia.gov/chembio/tech projects/detection/chemlab gas.html
http://www.ca.sandia.gov/chembio/tech projects/detection/chemlab liquid.html
http://www.mdl.sandia.gov/mstc/technologies/microsensors/chem.html
http://www.oit.doe.gov/sens cont/pdfs/annual 0602/robinson.pdf
http://www.ca.sandia.gov/chembio/tech projects/detection/factsheets/chemlab-bio-detector2.pdf
http://www.ca.sandia.gov/chembio/microfluidics/index.html
http://www.ca.sandia.gov/news/2003-news/031110news.html
http://www.ca.sandia.gov/chembio/tech projects/detection/factsheets/famebrochure.pdf
97 http://www.nanodetex.com/index.html
98 http://www.ca.sandia.gov/news/2004 news/120704Mercury.html
99 http://www.technet.pnl.gov/sensors/chemical/projects/es4cwsen.stm
http://www.technet.pnl.gov/sensors/chemical/projects/es4selct.stm
http://www.technet.pnl.gov/sensors/chemical/projects/ES4SurfWvSnsr.stm
http://availabletechnologies.pnl.gov/chemmaterials/chem.stm
100 http://cfpub.epa.gov/ncer abstracts/index.cfm/fuseaction/display.abstractDetail/abstract/7487/report/O
101 http://www.medical-test.com/productl 19/product info.html
102 http://www.pall.com/OEM 4154.asp
August 2005 E-5
-------
U.S. EPA Office of Water Early Warning Systems
103 http://www.idmscorp.com/pregnancytest.html
http://www.qdots.com/live/upload documents/wQDVOct03 pg8-9.pdf
104 http://www.tetracore.com/products/domestic.html
105 http://www.nhdiag.com/index.htm
106 http://www.eaicorp.com/products sea bh.htm
107
http://www.baddbox.com/
108 http://www.osborn-scientific.com/PDF/Positive test for terror toxins in Iraq.htm
109 http://www.responsebio.com/pdf/summaryanthrax aug02.pdf
110 http://user.fundy.net/pjwhalen/adenosinetriphosphate.html
Biosensors & Food Safety Diagnostics (Paul S. Satoh Neogen Corporation March 1, 2004)
111 http://www.celsis.com/products/pdfs/cels0150.pdf
112 http://www.amsainc.com/atp.asp
113 http://www.amsainc.com/atp-numbers.asp
114 http://www.charm.com/pdf/400-6505-503-300-01 WaterG.pdf
115 http://www.biotrace.com/content.php?hID=2&nhID=16&pID=16
116 http://www.geneq.com/catalog/en/profile-l.htm
117 http://www.bio.umass.edu/micro/immunology/facs542/facsprin.htm
118 http://pcfcij.dbs.aber.ac.uk/aberinst/mcytmain.html
http://www.biodetect.biz/products/mc.shtml
http://www.biodetect.biz/products/MC.pdf
http://www.biodetect.biz/applications/app303.pdf
119 http://www.brightwelltech.com/pdf files/Micro-Flow Imaging.pdf
http://www.brightwelltech.com/applications/app notes/wtp.php
120 "JMAR Technologies, Inc. Plans Launch of Laser-Based Early-Warning System to Detect
Microorganisms in Water Supplies" JMAR Technologies, Inc. Press Release.
http://biz.yahoo.com/bw/040621/215346 l.html;
121 ETV Technology Profile: On-Line Turbidimeters http://www.epa.gov/etv/pdfs/techprofile/01 turbid.pdf
122 AwwaRF #2720: Continuous Monitoring Method for Crytpotsporidium (abstract from website)
http://www. awwarf.com/research/TopicsAndProj ects/projectSnapshot.aspx?pn=2720
123 http://www.connect.org/members/april.htm
http://www.awa.asn.au/news&info/news/26janQ3.asp
August 2005 E-6
-------
U.S. EPA Office of Water Early Warning Systems
124 "JMAR Technologies, Inc. Plans Launch of Laser-Based Early Warning System to Detect
Microorganisms in Water Supplies" JMAR Technologies, Inc. Press Release.
125 http://www.shu.ac.uk/scis/artificial intelligence/IntelM ALLS.html
126
http://www.shu.ac.uk/scis/artificial intelligence/biospeckle.html
127 http://www.nrl.navy.mil/pressRelease.php?Y=2004&R=26-04r
http://www.resrchintl.com/pdf/spie wqm.pdf
http://www.resrchintl.com/product bibliography source.htm
http://www.resrchintl.com/raptor.html
http://www.resrchintl.com/pdf/raptor 2%20.pdf
128 http://www.luminexcorp.com/01 xMAPTechnology/08 Tutorials/How xmap works
http://www.luminexcorp.com/01 xMAPTechnologv/02 Applications/01 index.html
129 http://www.ncbi.nlm.nih.gov/entrez/querv.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list
uids-15228315
130 http://www.celsis.com/products/pdfs/cels0158.pdf
http://www.celsis.com/products/products dairy.php
131 http://www.vectech.com/newsletters/2003/November Newsletter.pdf
132 http://www.innovativebiosensors.com/overview.htm
http://www.innovativebiosensors.com/tech.htm
133 http://www.idahotec.com/rapid/index.html
134 http://www.idahotech.com/pdfs/RAPID pdfs/ETV%20Report-RAPID-short-release.pdf
135 http://www.idahotech.com/pdfs/RAPID pdfs/SocietyScopeV6.3.pdf
136 http://www.defenseindustrvdaily.com/2005/05/ibaids-a-step-forward-for-bioweapon-detection/index.php
137 http://www.idahotec.com/razor/index.html
http://stm2.nrl.navy.mil/~lwhitman/pdfs/nrlrev2001 BARC.pdf
138 http://www.smithsdetection.com/product.asp?product=Bio%2DSeeq&prodgroup=Bio%2DSeeq&
prodcat=Biological+Agent+Detection&prodarea=Trace+detection&division=Detection
139 http://www.the-scientist.com/asp/Registration/login.asp?redir=http://www.the-scientist.com/yr2003/
may/lcprofile 030505.html
140 http://www.wrenwray.com/images/pdf/CEPHEIDA.PDF
141 http://news.monevcentral.msn.com/ticker/sigdev.asp?Symbol=CPHD&PageNum=l
142 http://www.chem.agilent.com/Scripts/PCol.asp?lPage=50
http://www.securitvpronews.com/news/securitynews/spn-45-20050407InvitrogenandAgilent
TechnologiesToCoMarketPathAlertDetectionSystem.html
August 2005 E-7
-------
U.S. EPA Office of Water Early Warning Systems
143 http://www.ibisrna.com/
http://www.robodesign.com/tiger2.shtml
144 http://www.micro.uiuc.edu/boss/bossframes.htm
http://www.darpa.mil/mto/optocenters/presentations/cheng.pdf#search='Georgia%20Tech%20BOSS
%20sensor%20system'
145 http://asl.chemistry.gatech.edu/research ir-sensors-frame.html
http://gtresearchnews.gatech.edu/newsrelease/ESMART.html
http://asl.chemistry.gatech.edu/pdf-files/conference%20abstracts/Mizaikoff SIcon 03 1301.pdf#
search='evanescent%20field%20sensor'
146 http://www.cpac.washington.edu/~campbell/projects/spr.html
http://www.photonics.com/spectra/features/XQ/ASP/artabid.745/QX/read.htm
147
http://www.ee.washington.edu/research/denise/www/Lab/files/mike spr final.ppt
148 http://www.bitc.unh.edu/annual.reports/2004BITCfactsheet.pdf
149 http://www.aigproducts.com/surface plasmon resonance/spr considering.htm
http://www.aigproducts.com/surface plasmon resonance/spr evaluation module.htm
http://www.nomadics.com/products/spr3/
150 http://www.aigproducts.com/surface plasmon resonance/spr.htm
151 http://www.stanford.edu/~bohuang/Research/Anal%20Chem%202002.pdf
152 http://www.bioveris.com/technology.htm
http://www.mesoscaledefense.com/technology/ecl/diagram.htm
http://www.mesoscaledefense.com/technology/ecl/walkthrough.htm
http://www.sbs-archi.org/02pres/Umek.pdf
153 http://www.bioveris.com/products services/life sciences/instrumentation/mlmanalyzer.htm
154 http://www.biospace.com/news story.cfm?StoryID=17104620&full=l
http://us.diagnostics.roche.com/press room/2003/072403.htm
155 http://www.mesoscaledefense.com/coming soon.htm
156 http://spaceresearch.nasa.gov/general info/homeplanet.html
157 http://www.qdots.com/live/upload documents/wQDVOct03 pg8-9.pdf
158 http://www.qdots.com/live/render/content.asp?id=47
159 http://www.qdots.com/live/render/content.asp?id=87
http://www.bio-itworld.com/archive/021804/horizons dot.html
August 2005 E-8
-------
U.S. EPA Office of Water Early Warning Systems
160 http://www.sciencenews.org/articles/20030215/boblO.asp
http://www.sciencedaily.com/releases/2002/ll/021127071742.htm
http://www.eurekalert.org/pub releases/2004-06/uosc-qds061404.php
http://www.smalltimes.com/document display.cfm?document id=3811
http://www.llnl.gov/str/Lee.html
161 http://www.epa.gov/OGWDW/methods/current.html
162 http://www.bravurafilms.com/proiects/projectrep/phosphors.html
http://www.orasure. com/products/default. asp?cid=10&subx=4&sec=3
http://www.sri.com/news/releases/02-17-98.html
http://www.sri.com/rd/chembio.html
163 http://www. orasure.com/products/prodsubarea. asp ?cid=l&pid=126&sec=3&subsec=4
164 http://www.sri.com/rd/chembio.html
165 http://www.nanobioconvergence.org/speakers.aspx?ID=33
http://www.orasure.com/products/prodsubarea.asp ?cid=2&pid=126&sec=3&subsec=4
166 http://www.dynal.net/
167 http://www.technet.pnl.gov/sensors/biological/projects/ES4BEADS-Svs.stm
http://www.pnl.gov/breakthroughs/win-spr02/special3.htmltfbiothreat
168 http://www.chem.vt.edu/chem-ed/spec/vib/raman.html
http://en.wikipedia.org/wiki/Raman spectroscopy
169
http://www.iwaponline.com/wio/2003/04/wio200304WFOOHHE8UR.htm
170 http://www.deltanu.com/companyinfo.htm
http://www.chemimage.com/products/
171 http://www.combimatrix.com/news NBCKing5Aug04.htm
http://www.combimatrix.com/products biothreat.htm
172 http://www.promega.com/geneticidproc/ussympl Iproc/content/llewellyn.pdf
http://www.eurekalert.org/features/doe/2003-ll/dnl-lft031804.php
173 http://www.foresight.org/conferences/MNT8/Abstracts/Colton/
http://stm2.nrl.navy.mil/~lwhitman/pdfs/nrlrev2001 BARC.pdf
174 http://stm2.nrl.navy.mil/~lwhitman/pdfs/Rife Sensors Actuators A published.pdf
175 http://www.gwu.edu/~physics/colloq/miller.htm
176 http://www.affymetrix.com/technology/manufacturing/index.affx
177 http://www.dsls.usra.edu/meetings/bio2003/pdf/Biosensors/2149Stahl.pdf
178 http://www.sciencemag.org/feature/e-market/benchtop/biochips3 10 18 O2.shl
August 2005 E-9
-------
U.S. EPA Office of Water Early Warning Systems
179 http://www.protiveris.com/new/products folder/veriscansystem.html
http://www.buscom.com/nanobio2003/session3c.html
180 http://pharmalicensing.com/news/headlines/1070473459 3fce20f35628c
181
http://www.memsnet.org/mems/what-is.html
182 http://www.iatroquest.com/En.htm
http://www.iatroquest.net/Linked%20Docs/IatroQuest%20Corporation%20Fact%20Sheet%2004
0217-1334.pdf
183 http://www.cm.utexas.edu/mcdevitt/ET Broch.pdf
http://www.cm.utexas.edu/mcdevitt/tastechip.htm
184 http://www.silsoe.cranfield.ac.uk/staff/apturner.htm
http://www.cranfield.ac.uk/ibst/ccst/mips/sensors.htm
185
http://www.scs.uiuc.edu/suslick/pdf/pressclippings/naturebiotech.0902-884.pdf
186 http://www.tekes.fi/ohjelmat/diagnostiikka/diag esitvkset/turner.pdf
187 http://www.ee.psu.edu/grimes/sensors/
188 http://www.isco.com/WebProductFiles/Product Literature/201/Specialty Samplers?3710RLS
Radionuclide Sampler.PDF
189 http://www.epa.gov/watersecurity/guide/radiationdetectionequipmentformonitoringwaterassets.html
U.S. EPA, Radiation Detection Equipment for Monitoring Water Assets, Water and Wastewater Security
Product Guide
190 http://www.tech-associates.com/dept/sales/product-info/sss-33-5ft.html
Drinking Water Rad-Safety Monitor Model #SSS-33-5FT. Updated 1/31/2002.
191 http://www.tech-associates.com/dept/sales/product-info/meda-5t.html
Updated in 2002.
192 http://www.isco.com/WebProductFiles/Product Literature/20 I/Specialty Samplers?3710RLS
Radionuclide Sampler.PDF
193
http://www.tech-associates.com/dept/sales/product-info/sss-33dhc.html
194 http://www.tech-associates.com/dept/sales/product-info/sss-33m8.html
195 http://www.canberra.com/products/802.asp
196 http://www.canberra.com/products/803.asp
197 http://www.canberra.com/products/801.asp
198 http://www.clarionsensing.com/home.shtml
199 http://apps.em.doe.gov/ost/pubs/itsrs/itsr312.pdf
August 2005 E-10
-------
U.S. EPA Office of Water Early Warning Systems
200 http://www.cpeo.org/techtree/ttdescript/alpharad.htm Last updated 10/2002.
201 http://www.technet.pnl.gov/sensors/nuclear/projects/ES4Tc-99.stm
202 http://www.epa.gov/watersecuritv/guide/radiationdetectionequipmentformonitoringwaterassets.html
203 http://www-emtd.lanl.gov/TD/WasteCharacterization/LiquidAlphaMonitor.html
204 http://www.cfsan.fda.gov/~dms/fssupd72.htmltfgrants
205
http ://www.j ohnmorris .com. au/html/Mantech/titrasip .htm
206 http://www.hach.com/hc/search.product.details.invoker/PackagingCode=6950000/NewLinkLabel=
Hach+Event+Monitor+Trigger+System/PREVIOUS BREADCRUMB ID-HC SEARCH
KEYWORD/SESSIONID|B3hNRFkwTlRRNE56RTROekVtWiNWbGMzUk5Sdz09QWxKYVZ6RQ=H
207 FBI/CDC June 2002 Evaluation of Hand-Held Immunoassays for Bacillus anthracis and Yersinia pestis
208 http://www.isco.com/WebProductFiles/Product Literature/201/Specialty Samplers/3710RLS Radio
nuclide Sampler.PDF
209 http://www.dhs.gov/dhspublic/theme homel.jsp
210 http://www.dhs.gov/dhspublic/display?theme=36
211 http://www.globalsecurity.org/securitv/librarv/policv/national/hspd-9.htm
212 http://www.epa.gov/etv/homeland/
213 http://www.dhs.gov/dhspublic/displav?theme=38&content=4014&print=true
http://www.fcw.com/fcw/articles/2004/0419/web-scada-04-19-04.asp
http://www.dhs.gov/dhspublic/interapp/editorial/editorial O359.xml
214 http://www.fas.org/man/dod-101/army/docs/astmp98/sec3k.htm
215 http://www.dtic.mil/whs/directives/corres/pdf2/d200012p.pdf
216 http://www.defenselink.mil/news/Sep2004/n09032004 2004090304.html
217 http://www.darpa.mil/index.html
218 http://www.darpa.mil/dso/thrust/biosci/biostech.htm
219 http://www.darpa.mil/dso/thrust/biosci/biosensor/enabtech.html
http://www.darpa.mil/mto/People/PMs/carrano dtec.html
http://www.darpa.mil/dso/thrust/biosci/biostech.htm
http://www.darpa.mil/dso/thrust/biosci/biosensor/enabtech.html
http://www.darpa.mil/dso/thrust/biosci/biosensor/auburn d.html
http://www.darpa.mil/dso/thrust/biosci/biosensor/auburn i.html
http://www.darpa.mil/dso/thrust/biosci/biosensor/sandia.html
http://www.darpa.mil/dso/thrust/biosci/biosensor/rush med.html
http://www.darpa.mil/dso/thrust/biosci/biosensor/argonne.html
August 2005 E-ll
-------
U.S. EPA Office of Water Early Warning Systems
http://www.darpa.mil/body/procurements/old procurements/ian2000/mtojanOO.html
220 http://www.nrl.navy.mil/content.php7P-ABOUTNRL
221 http://pubs.rsc.org/ei/CC/2000/b003185m.pdf
222 http://www.nrl.navy.mil/content.php7P-04REVIEW115
223 http://www3.interscience.wiley.com/cgi-bin/abstract/107061018/ABSTRACT
224 http://www.nrl.navy.mil/pao/pressRelease.php?Y=1996&R=26-96r
225 http://www.foresight.org/conferences/MNT8/Abstracts/Colton/
http://stm2.nrl.navy.mil/~lwhitman/pdfs/nrlrev2001 BARC.pdf
226 http://stm2.nrl.navy.mil/~lwhitman/pdfs/Rife Sensors Actuators A published.pdf
227 http://stm2.nrl.navy.mil/~lwhitman/pdfs/nrlrev2001 BARC.pdf
228 http://www.globalsecurity.org/wmd/facilitv/edgewood.htm
229 http://www.epa.gov/ordnhsrc/
230 http://water.usgs.gov/wicp/acwi/monitoring/conference/2004/conference agenda links/power points
etc/06 ConcurrentSessionD/90 Rml5 Vowinkel.pdf
231 http://www.ilsi.org/publications/pubslist.cfm?pubentityid=13&publicationid=268
232 http://water.usgs.gov/wicp/acwi/monitoring/conference/2004/conference web agenda.html
233 http://water.usgs.gov/ogw/karst/kig2002/msf development.html
234 http://www.mdl.sandia.gov/mstc/technologies/microsensors/techinfo.html
http://www.sandia.gov/sensor/noframes.htm
http://www.ca.sandia.gov/chembio/news center/ST2002v4no3.pdf
235 http://www.ca.sandia.gov/news/2004 news/120704Mercury.html
236 http://www.llnl.gov/llnl/001index/02about-index.html
237 http://www.llnl.gov/sensor technology/SensorTech contents.html
238 http://www.ornl.gov/ornlhome/about.shtml
239 http://www.ornl.gov/ornlhome/about.shtml
240 http://www.ornl.gov/info/ornlreview/rev29 3/text/biosens.htm
241 http://www.ornl.gov/sci/engineering science technology/sms/Hardv%20Fact%20Sheets/
Countermeasures.pdf
242 http://pharmalicensing.com/news/headlines/1070473459 3fce20f35628c
August 2005 E-12
-------
U.S. EPA Office of Water Early Warning Systems
243 http://www.ornl.gov/sci/biosensors/
http://www.ornl.gov/adm/tted/LifeSciencesTechnologies/LifeSciencesTechnologyList.htm
244 http://www.pnl.gov/main/welcome/
245 http://www.pnl.gov/main/sectors/nsd/%20homeland.pdf
246 http://www.technet.pnl.gov/sensors/
247 http://www.pnl.gov/main/sectors/nsd/%20homeland.pdf
248 http://www.inl.gov/index.shtml
249 http://www.amsa-cleanwater.org/meetings/04winter/ppt/ppt/30%20--%20FRI%20-%20Reinhardt,
%20Glen/AMSA%20-%20Security%20Panel%20Discussion%20-%20GR.pps
250 http://www7.nationalacademies.org/wstb/index.html
251 http://www4.nationalacademies.org/webcr.nsf/CommitteeDisplay/WSTB-U-04-06-A7OpenDocument
252 http://www4.nas.edu/webcr.nsf/5c50571a75df494485256a95007a091e/5d3beab7fa3bb8bc85256dOb
00705acf?OpenDocument
253 http://www.awwarf.org/theFoundation/
254 WERF's website: www.werf.org
255 http://cimic.rutgers.edu/epa-workshop.html
256 http://www.knowledgepress.com/
257 http://www.knowledgepress.com/events/7011409 p.pdf
258 http://www.healthtech.com/2003/btr/
259 http://www.knowledgepress.com/events/12111105 p.pdf
260 http://www.wtec.org/biosensing/proceedings/
261 http://www.knowledgepress.com/events/7191716 p.pdf
262 http://cimic.rutgers.edu/epa-workshop.html
http://cimic.rutgers.edu/workshop2.html
263 http://www.healthtech.com/2002/bms/abstracts/symposium3.htm
264 http://www.knowledgepress.eom/events/l 1071420 p.pdf
August 2005 E-13
-------
&EPA
United States
Environmental Protection
Agency
Office of Research and Development
National Homeland Security Research Center
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
EPA/600/R-05/156
December 2005
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |