Performance Verification Testing Rapid Toxicity Monitoring and Detection Systems Overview and Analysis

Performance Verification Testing
Rapid Toxicity Monitoring and Detection Systems
Overview and Analysis
Background
As part of its role in protecting the nation's drinking water supplies and systems from acts of terrorism, the United States
Environmental Protection Agency (EPA) has supported the testing and evaluation of rapid toxicity monitoring and detection
systems. EPA's efforts in drinking water system security focus on physical security, including cyber security, destruction and
disruption of services to both water and wastewater systems, and the threat of delivering chemical, biological, and
radiological contaminants into water systems.
An integral part of the Agency's approach to meeting its mission is testing and evaluating the performance of commercially
available technologies that may be useful for monitoring and detecting contaminants; as early warning systems; for mitigation
and treatment; and, for providing alternative water supplies. The testing of rapid toxicity monitoring and detection systems
was conducted under the auspices of the EPA's Environmental Technology Verification Program (ETV). The seven vendors,
who voluntarily participated, represent eight technologies: Aqua Survey, Inc's IQ Toxicity Test™; Strategic Diagnostics Inc's
Deltatox® and Microtox®; Severn Trent Services' Eclox; InterLab Supply, Ltd's PolyTox™; Hach Company's ToxTrak™;
CheckLight, Ltd's ToxScreen; and Hidex Oy's BioTox™.
Prior to the inception of testing, a water security stakeholder group was formed to help guide testing and evaluation efforts.
The vendors had not previously tested their technology's ability to detect the presence of several of the contaminants included
in this effort. Despite the lack of vendor testing and evaluation of some of the contaminants, they were in wide agreement
with the approach. This was in part due to the fact that the stakeholder group strongly advised that technologies be challenged
using contaminants that required testing in special facilities which typical water utilities do not have access. This document
provides a broad summary of the data generated in the test. The reader is encouraged to obtain copies of each report from the
ETV web site (www.epa. gov/etv) for more detailed individual technology information.
Principle of Operation and Use
The rapid toxicity technologies are all bioassays designed to use living organisms or enzymatic reactions as toxicity indicators
either directly or in combination with reagents. These organisms or reactions produce a baseline level of light or use dissolved
oxygen at a steady rate in uncontaminated water. When subjected to water containing toxics, these indicator organisms should
ideally react by exhibiting a reduced light intensity or dissolved oxygen uptake rate.
Rapid toxicity technologies are intended to serve as quick screening tools for determining whether water being tested contains
contaminants at concentrations harmful to human health or the environment. Until recently, these technologies have been
used for water monitoring at facilities that discharge effluents, such as industrial production facilities, to monitor the potential
ecological impact of the effluents. A potential new use for this type of technology is to monitor for early signs of intentional
acts of chemical contamination on drinking water systems. The rapid toxicity monitors are not intended to be used to identify
a specific contaminant or to measure the amount of contaminant. Their envisioned role is to provide a relatively quick
analysis of a water sample to determine whether a contaminant exists in a quantity that could pose a risk to consumers. Once
the sample is prepared for analysis, these technologies typically can be used to generate a result in less than an hour. By
comparison, standard EPA methods for measuring acute toxicity take between 24 and 96 hours to generate a result.
Performance Verification Test Design
The objective of the verification testing was to rigorously test the ability of each rapid toxicity technology to detect selected
toxic threat contaminants (excluding biological warfare agents) in water and to determine their susceptibility to commonly
occurring but potentially interfering compounds and matrices. Stakeholders from the water utility industry and others,
including vendors, helped design the test. The target contaminants were chosen as representatives of five different classes of
potential contaminants: industrial chemicals, pesticides, pharmaceuticals, nerve agents, and biotoxins. The five toxic
industrial chemicals and four chemical warfare agents are aldicarb, colchicine, cyanide, dicrotophos, thallium sulfate,
botulinum toxin, ricin, soman, and VX. Testing also included five potentially interfering elements (aluminum, copper, iron,
manganese, and zinc) and two samples that were each treated using two different disinfection processes (chlorination and
chloramination). Vendors participated during testing and reviewed results as did quality assurance managers from EPA and
the testing organization. Test plans and reports were also reviewed by technical peer reviewers external to EPA and the
testing organization.

-------
The contaminant-spiked drinking water samples that were evaluated by each technology were prepared from a single drinking
water source. The drinking water samples were dechlorinated and then spiked with various concentrations of contaminants or
interferents. Individual solutions containing each contaminant or potential interferent were prepared and analyzed. Samples
containing a contaminant were prepared at a lethal dose concentration, estimated for a 154 lb person ingesting about 8 oz of
water, then diluted multiple times (by a factor of 10 each time) to determine the sensitivity of the technology to each
contaminant at several concentration levels. Samples containing the interferents were analyzed at a single concentration. Each
technology also analyzed contaminant-free drinking water samples from two disinfection processes (chlorination and
chloramination) to determine if commonly occurring disinfection byproducts would affect performance. Endpoints and
precision, toxicity threshold (for each contaminant), false positive/negative responses, ease of use, and sample throughput
were evaluated. Samples used in the verification test included spiked drinking water test samples and quality control samples
including method blank and positive and negative control samples.
Performance and Results
The rapid toxicity detection systems varied in their responses to specific toxics introduced into the water samples. All eight
technologies responded to cyanide and thallium sulfate; six of the eight identified toxicity in response to the presence of
aldicarb and dicrotophos; five of the eight responded to colchicine and ricin; three to VX, only two to botulinum toxin; and
one to soman. Results were considered false negative if the response to the lethal dose concentration of each contaminant was
not significantly different from the negative control. Table 1 gives the number of technologies that exhibited false negative
results for each contaminant.
Cations that are likely present in water sources were evaluated as potential interferences in this test, including: copper, zinc,
iron, and aluminum. In addition, chlorinated water samples (from either chlorination or chloramination processes) without any
spiked contaminants or interferents were also analyzed. The interferents tested were considered to interfere if the percent
inhibition detected with the interferent alone was likely to cause an error (in the positive or negative direction) when
determining the toxicity of an unknown water sample. Table 1 also gives the number of technologies susceptible to each
potential interferent. The potential for responses that falsely indicate the presence or absence of toxic contaminants due to the
presence of background interferents or byproducts of chlorination and chloramination processes or both must be carefully
considered and accounted for by potential users. The three systems that were not affected by chloramination by-products were
also the least sensitive to the contaminants of interest. Conversely, those technologies that were sensitive to the widest variety
and concentrations of contaminants were also inhibited by the commonly occurring interferents.
It is important to simultaneously consider the response of each technology to the contaminants of interest and the interferents
to fully gauge and understand performance. The fact that there were numerous false positive and negative results suggest that
these technologies must be used with care by water utility managers. The prudent water utility manager would use these types
of technologies in conjunction with other analytical methods that can confirm the presence or absence of a toxic contaminant,
rather than as stand-alone early warning devices. Another option is for the water utility manager to fully characterize the
performance of the technology using a range of samples that represent the typical distribution system water quality. It would
then be necessary to fully characterize the chemical and physical characteristics of the water using standard methods to
sufficiently understand the rapid toxicity monitoring system.
The data generated in this verification test are not indicative of performance for every possible combination of contamination,
water source, and potential interferent, but the data do represent a reasonable snapshot of performance. It is important to note
that users should fully understand typical performance in their water systems for the most accurate interpretation of positive
and negative results. Users of these technologies are encouraged to carefully review the verification reports and verification
statements posted on the EPA ETV Web site (www.epa. gov/etv ) and to discuss performance data generated for their water
systems with vendors.
For further information:
Grace Robiou, USEPA, Water Security Division, Office of Water, Washington, DC - 202-564-6319 or robiou.grace@epa.gov
Jonathan Herrmann - National Homeland Security Research Center, Office of Research and Development - 513-569-7839 or
herrmann. i onathan@epa. gov

-------
Table 1. Number of Technologies Exhibiting False Negative Responses and Susceptibility to
Interferents
Contaminant
(Lethal Dose - mg/L)
Technologies Giving a False
Negative Response at the Lethal
Dose
Potential Interferent
(Concentration-mg/L)
Technologies
Susceptible to
Interference
Aldicarb (280)
2
Aluminum (0.36)
1
Colchicine (240)
3
Copper (0.65)
4
Cyanide (250)
0
Iron (0.069)
2
Dicrotophos (1,400)
2
Manganese (0.26)
1
Thallium Sulfate (2,400)
0
Zinc (3.5)
4
Botulinum Toxin (0.3)
6
Chloramination
5
Ricin (15)
3
Chlorination
3
Soman (0.2)
7


VX (0.2)
5

-------