PB84-167600
Toxic Area Delineation by Canine Olfaction
Municipal Environmental
Research Lab.-Cincinnati, Edison, NJ
1984
>t^jr ;f..?T U.t -L-AT>.'.f
32E3SSESEE35ZS3HES33I
ULS. Oc?srfe£f?t of Ce&sftsre
»,. f^ss^ Tecfensd EsJsf^Ssa
-------
EPA-600/D-84-074
1984
TOXIC AREA DELINEATION BY CANINE OLFACTION
by
L.D. Arner
Biosensors, Inc.
Westmoreland, NY
H. Masters
U.S. Environmental Protection Agency
Edison, NJ
G.R. Johnson
Guardian Training Academy
Windsor, Ontario
and
H.S. Skovronek
Environmental Services
Morris Plains, NJ
Contract Number 68-03-3069
Project Officer
Hugh E. Masters
Oil and Hazardous Materials Spills Branch
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory (Cincinnati)
Edison, New Jersey 08837
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
TECHNICAL REPORT DATA
rcaJ lourvfiiuni o>. the male ln-lutf completing)
1 REPORT NO
EPA-6OO/D-84 - 07A
•4. TITLE AND SUBTITLE
TOXIC AREA DELI N: AT I ON BY vJANINE OL FACTION
7. AUTHOR(S)
L. D. Arner, H. Master, G.R. Johnson, M.S. Skovronek
9. PERFORMING ORGANIZATION AME AND ADDRESS
IT Corporation
312 Directors Drive
Knoxville. TN 37923
1?. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory-Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati , Ohio 45268
3. RECIPIINT'S ACCESSION-NO.
.JPB814-1.6.Z6.QCL
5. REPORT DATE
1984
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
tO. PROGRAM ELEMbNT NO.
CBRD1A
11. CONTRACT/GRANT NO.
68-03-3069
13. TYPE OF REPORT AND PERIOD COVERED
Paper
TV SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Hugh Masters (201) 321-6678
16. ABSTRACT
The ability of animals to respond to pollutants 1n their environment is
a well-known phenomenon. Uhiie this capability has be?n used to trace
people and to find explosives and narcotics, there has been little effort
to apply this 'talent* to environmental problems. The idea of using dogs
to detect or locate sources of pollution was conceived several years ago by
one of the authors (M.S. Skovronek). while the use of animals will
(probably) never replace sophisticated instrumentation, source monitoring
by animals offe. s a rapid, economical means of screening suspect locations
for specific pollutants. Recently, a program was initiated by the U.S.
Environmental Protection Agency to explore the application of nx>nitoring by
animals to expedite sampling programs at hazardous spills or waste sites.
This paper describes a feasibility study in which a dog/handler team
was used to locate low concentrations of a hazardous substance (i.e.,
trichlorophenol and toluene) hidden in a field, thus suggesting that a dog
can be trained to locate such materials on industrial sites, abandoned
landfills, etc. The use of a dog/handler tean to uncover simulated
hazardous wastes infiltrating into buildings such as might be encountered
with groundwater leakage, seepage frora storage tanks, etc. will also be
described. Lastly, the use of dogs to assist wo: »ers at a hazardous site
In delineating the contaminated area will be discussed.
To a limited extent, the use of state-of-the-art portable gas/vapor
detection Instruments at waste sites will be compared with the application
of this new 'Instrument". The experience with and the /Jpa*r>
UNCLASSIFI-J)
21. NO. OF PAGES
15
22. PRICE
EPA Form 2220-1 (9-71)
-------
NOTICE
This document, has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
TUXIC AREA DELINEATION
BY CANINE OLFACT10N
L.D. Arner (1), H. Masters (2),
G.R. Johnson (3), and H.S. Skovronek (4)
ABSTRACT
The ability of animals to respond to pollutants in their environment is
a well-known phenomenon. While this capability has been used to trace
people and to find explosives and narcotics, there has been little effort.
to apply this "talent" to environmental problems. The idea of using dogs
to detect or locate sources of pollution was conceived several years ago by
one of the authors (H.S. Skovronek). While the use of animals will
(probably) never replace sophisticated instrumentation, source monitoring
by animals offers a rapid, economical means of screening suspect locations
for specific pollutants. Recently, a program was initiated by the U.S.
Environmental Protection Agency to explore the application of monitoring by
animals to expedite sampling programs at hazardous spills or waste sites.
This paper describes a feasibility study in which a dog/handler team
was used to locate low concentrations of a hazardous substance (i.e.,
trichlorophenol and toluene) hidden in a field, thus suggesting that a dog
can be trained t locate such materials on industrial sites, abandoned
landfills, etc. The "se of a dog/handler team to uncover simulated
hazardous wastes infiltrating into buildings such as might be encountered
with grourdwater leakage, seepage from storage tanks, etc. will also be
described. Lastly, the use of dogs to assist workers at a hazardous site
in delineating the contaminated area will be discussed.
To a limited extent, the use of state-of-the-art portable gas/vapor
detection instruments at waste sites will be compared with the application
of this new "instrument". The experience with and the inherent detection
potential of canines will be reviewed and new directions explored.
(1) Biosensors, Inc., Westmoreland, NY
(2) U.S. Environmental Protection Agency, Edison, NJ
(3) Guardian Training Academy, Windsor, Ont.
(4) Environmental Services, Morris Plains, NJ
* This paper has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency.
-------
BACKGROUND
With the passage of the Resource Conservation and Recovery Act (RCRA)
and the Comprehensive Environmental Response, Compensation, and Liability
Act of 1980 (CERCLA,also known as "Superfund"), the nation's attention has
shifted strongly to the search for and the cleanup of hazardous waste
sites. The location of such sites is, however, far from easy. Many sites
evolved over the years as industrial and municipal authorities sought to
dispose of their wastes in the safest and most cost-effective manner.
Disposal in dumps, landfills, and "the back 40" were all common practices—
with no realization that a legacy of toxic wastes was also being left
behind to leach into groundwater, surface water, and air. And to such
practices, unfortunately, must be added the intentional uncaring disposal
of wastes, even those known to be toxic, in blatantly unsuitable sites.
The need for controlled disposal is now better understood, and there is
a current appreciation of the risks that such past disposal practices may
have created for us and for our progeny. A massive program is underway
both by regulatory agencies and by industry to locate and to delineate such
sites and, soon, to undertake corrective action where needed and
economically achievable.
In addition to the problems created by the inappropriate disposal of
wastes now known to be toxic or hazardous, society must also face the
consequences of the releases or discharges into the environment that result
from transportation and industrial accidents and disasters. A key factor
in assessing the impact of such incidents is delineating the dispersal
-------
of the contamination quickly so that cont?;nment and cleanup operations can
be started.
Sophisticated instrumentation continues to be developed to locate and
identify pollutants in the environment. These instruments are capable of
measuring lower and lower concentrations and of producing reliable results
much more rapidly than in past years. The cost, complexity, and very
sensitivity of modern instruments often makes them less than ideal for
rapid deployment and for use in preliminary field measurements of a
"screening" nature. To some extent, industrial health-monitoring
instrumentation can fill the gap and provide quick field measurements,
although often not at the very low concentrations now demanded to protect
field workers and neighboring residents at hazardous waste release and
spill sites. The search for techniques and methods to assist field workers
in delineating-contaminated areas continues.
Another approach has been available for many years—and, indeed, widely
used in the environmental area—albeit from a different viewpoint: the
responses of animals to pollutants. Much of the biological testing used to
evaluate the hazard of wastes and specific compounds depends on the effects
on living organisms. All are familiar with LC-50 and LD-50 tests in which
the lethality of materials to mice, rats, or other animals is determined.
Bacterial and enzyme-inhibitor work has also been undertaken. Another key
factor in animal tests is the behavior of the species at lower, non-lethal
doses. Over the years, the activity of fish, either in aquaria or in
tethered cages in a river, has been extensively used as a measure of water
quality. Specific activities such as swimming, coughing, gill movement.
-------
etc., have all been monitored. Indeed, there are typical biorr.onitoring
requirements in many National Pollutant Discharge Elimination System (NPDES)
permits.
The approaches described above may be considered as "passive"
monitoring; the investigator merely observes the behavior of a species in a
contaminated environment and compares that behavior to that which would be
expected in the creature's normal, uncontaminated environment. The utility
of this approach is well-documented. The opportunity now exists to move
forward bj applying more sophisticated "active" monitoring using higher
creatures, specifically, the dog.
Everyone is familiar with the use of the dog in locating explosives,
narcotics, lost children, criminals, etc. No instrumental method has
approached the olfactory (i.e., sniffing) ability of the canine in these
pursuits. For several years, it has been our opinion that a dog could be
trained to track and locate specific pollutants or classes 01 pollutants in
a similar manner. Some very encouraging results have been reported by
researchers working at the do§ training industry interface. For example,
using procedures evolved from formalized man-tracking techniques, Glen
Johnson used dogs to successfully traverse a 94-mile natural gas pipeline
and to uncover 150 leaks, many of which went undetected when sophisticated
instrumentation was brought in. Leaks were detected in sections buried
18-ft deep, as well as in 12-ft elevated segments.
In another experiment, dogs were trained to find and to differentiate
between nitrogen, helium and Freon 12 with 100% rellability.2 In a
"Johnson, Glen R. Tracking Dog Ther-y and Methods. Arner Publications,
Inc., Westmoreland, NY,1977. pp. 15-21.
^Johnson, Glen R. Odorless Gas Detection by Domestic Canines. Off-lead 6:
18-19, 1977.
-------
little-known study, Johnson also useo dogs to locate leaks of electrical
fluid "below 10 in. of concrete covered by 8 in. of asphalt" in New York
City's streets.3 _ .
The little understood olfactory ability of the dog, coupled with its
ability to work with man, can be harnessed, by qualified trainers and
handlers, and used in the location of hazardous wastes. The overall goal
of this program i; to demonstrate that dogs, properly trained and properly
handled, can be used to assist field respose personnel -in several ways.
Specifically, the canine's extreme olfactory sensitivity can aid in
screening areas, both within and outside of structures, where specific
pollutants may be present. For example, the dog's responses may show that
a building contains .1 source of BTX (benzene, toluene, xylenes), such as
may leak from storage tanks or arise from the seepage of contaminated
groundwater. The trained dog can potentially delineate the perimeter of a
site contaminated by hazardous wastes or the area contaminated by movement
of spilled material, either in the water, soil, or air. This knowlege will
assist field workers in determining where safety equipment must be worn and
can also save costs by reducing the number of samples that must be
collected and analyzed—often at considerable cost and delay—to define the
extent of a contaminated site.
Hypothetically, a trained dog could (at this time), have gone into
Times Beach, MU or the Ironbound section of Newark, NJ and quickly outlined
the area contaminated with dioxin. Sampling teams would have then taken
^Johnson, Glen K. New York Experiment. Off-lead, 10: 10-13, 1981.
^McCartney, William. Olfaction and Odours. Springer-Verlag, New York, New
York, 1968. pp. 15-70.
-------
samples to confirm and quantify the level of contamination at the sites.
This paper covers only the first, or feasibility, phase of a study in which
the goal is to'demonstrate that the use of dogs to locate pollutants and to
define contaminated areas is viable and can be applied to hazardous waste
situations.
TECHNICAL PROGRAM
Many aspects of the study cannot currently be approached from a rigid
scientific approach since there is no full understanding of the phenomenon
of the dog's olfactory ability. Other compromises are also necessary in
the interest of speed and economy. To compare canine detection with
conventional field methods, demonstrations will be conducted in parallel
with rapid on-site methods of chemical analysis that use seme of the most
sophisticated portable equipment available today.
The goal of this feasibility study is to train dogs to discriminate,
above background levels certain hazardous substances. Sensitized
specifically to toluene and 2,4,6-trichlorophenol (TCP), a more volatile
and a less volatile species, the dog will either move toward the source,
presumably along a line of increasing concentration, or will alert h^s
handler that he is at or close to the source. In many ways, this training
parallels the work on detection of narcotics or explosives.
In the case of TCP. the dog will be trained NOT to move to the source
of an odor but, instead, to stop and alert his handler so that the dog need
not approach or enter an area of high pollutant concentration. TCP was
selected as an indicator for dioxin since it is usually present in the
synthetic scheme that yields dioxin, but does not present the extreme
hazards associated with dioxin.
6
-------
The dog will be trained, or conditioned to move in an indicated direction
(toward a suspected source) ONLY until he first detects the specific
material (TCP). He will then stop, indicate his observation to the
handler, and withdraw. This process r.hould be repeated from several
different directions until the perimeter of the contaminated area has been
delineated. In a more sophisticated adaptation, the dog will be taught to
move along the perimeter of the contaminated area at the distance where he
is only able to receive the first olfactory indication of the compound. In
this way, the handler or an accompanying observer can actually map the
Canine Detection Limit (CDL) perimeter of the source. Once it is
established whether the CUL is higher or lower than the detection limit
achievable by field instrumentation, samples will be taken or monitors
installed at or near this perimeter to protect workers and neighboring
residents.
EXPERIMENTAL PROCEDURES
The training of the dogs and the familiarization of the handlers with
the dogs' responses is essentially a developmental process. Technically,
the dog is sensitized or trained to recognize a specific material or a
class of materials using what is called an inductive or positive
reinforcement mode in which the dog's natural tendencies are emphasized and
allowed to "reinforce" the desired response.
The dog is sensitized by exposure to different materials that have been
impregnated with or contain small amounts of toluene or TCP. This
technique is used to help the dog "target" on a specific material or
chemical while teaching him to ignore all other odors. The dog is then
conditioned to select, for example, the toluene-impregnated items from
7
-------
among several different blanks. The amount of chemical used to impregnate
the targets can be reduced, but no attempt is made at this stage to achieve
the COL. Once the dog is consistently selecting impregnated targets in a
confined area, the dog's prior training in field tracking is brought to
bear. (The dogs used for this program all have previously demonstrated a
high proficiency in following human tracks through brush, woods, fields,
etc.) The dog's purpose for search is now transferred from a person to the
chemical and the dog is directed to seek o'jt a planted target. Each time
the dog successfully finds the planted target, he is vigorously rewarded.
The training next takes two slightly different directions in order to
address interior and exterior searches. Working outside and subject to the
weather makes it very difficult to estimate the actual concentration (ppm)
that is sufficient to catch the dog's attention. Several experiments will
be carried out with the dog approaching fresh sample targets from different
wind directions until he indicates discovery and moves toward the source.
Measurements using conventional techniques will be made along these same
vectors in order to estimate the level at which the dog first responds
confidently.
When the dog is consistently able to detect surface samples, samples
will be buried to depths of 2 ft. Similar holes will also be dug but
refilled to overcome the dog's ability to detect or recognize freshly dug
earth. With buried samples, there will be some time delay in the
appearance of "peak" sensitivity, which factor results from the gradual
diffusion or permeation of the vapor from its initial depth to the
surface.
When the dog consistently recovers buried samples from the maximum
8
-------
depth, a demonstration will be arranged for EPA observers. The
demonstration will consist of several small targets both located on the
surface and buried by EPA personnel at varous depths from zero to 2 ft. In
addition, there will be one or more "blank" holes. The dog will be brought
to the site approximately 24 hr after emplacement of the samples. The
handler will be given only a general direction and a distance limit (ca.
200 ft) within which to search. By allowing the dog to cast about and
sniff freely, the dog/handler team will attempt to locate all of the
samples without identifying any false targets. Ideally, the test will be
carried out early in the day when scent detection seems to be easiest.
The area will immediately be monitored after the dog has completed his
test to determine whether conventional instrumentation could have uncovered
the samples and to assess the time and effort that such monitoring would
require.
Throughout the study and particularly for the demonstrations, records
will be maintained of local temperature, relative humidity, approximate
wind speed, wind direction, and ground temperature.
At the same time that the field training is proceeding, the interior
search program—using toluene as an indicator for BTX, gasoline, etc.--will
also be underway. Once the dog has been trained to select the designated
chemical and move toward its source, samples will be hidden in buildings
and the dog encouraged to find them. As the dog becomes more proficient,
the size of the target (i.e., the quantity of toluene) will be reduced
until the dog's ability to detect the material fails. 8y knowing the
dimensions of a room in which the sample is placed and adjusting the sample
size, an ambient concentration can be approximated—assuming (1) no air
exchange and (2) uniform diffusion of the chemical.
9
-------
When reliability has been established for interior searching, a formal
demonstration will also be carried out by having ERA personnel hide various
toluene-containing samples in several test rooms of a building. The dog
will be brought to each room to (1) determine whether toluene is present by
barking or by performing some other means of communication with his hanu'.sr
and (2) finding the source, which may be an open bottle or a spot on the
floor that had been impregnated with a small amount of toluene to simulate
seepage, as may occur from contaminated groundwater.
The third phase of the study, called the perimeter search, is expected
to be the most difficult since it requires the dog to inhibit his natural
desire to move from less concentrated material to more concentrated. It
will be necessary, instead, to teach the dog to stop as soon as he detects
the specific target odor. The handler will have to learn to "read" the
rather limited signal that the dog will be giving at first detection. The
dog will then withdraw and approach the site repeatedly from different
directions until a full perimeter has been delineated, all at the CDL.
This methodology was selected for three reasons: (1) if successful, it will
allow a contaminated area to be delineated rapidly so that the public can
be excluded and fif.ld workers can be alerted to the need for protective
equipment, (2) it will minimize the number of monitoring stations or
screening samples that must be taken to delineate the area quantitatively,
and (3) it protects both dog and handler from all but the minimum exposure
to the target material, which in practice could be highly toxic or
hazardous materials.
A demonstration of the dog's perirreter-delineation ability will be
carried out at a wel1-documented hazardous waste site where conventional
10
-------
instrumental delineation of the perimeter has been conducted. The
dog/handler team will know only the general area of contamination and will
develop a perimeter on its own. When necessary, the handler will be given
the le--els found at certain points to assist in planning the approach.
Should the dog/handler team stray toward more contaminated areas, it will
be warned off.. Still and TV film coverage will be obtained to record the
team's efforts. .When the perimeter derived by the dog/handler team is more
remote from the source than the results of the earlier instrumental
analyses, additional analyses will be taken by EPA personnel to attempt to
correlate the divergent results.
CONCLUSLUNS
Carefully selected and trained dogs, together with competent handlers,
are capable of following the "scent" of chemicals to their source, even
when those chemicals are present at extremely low airborne concentrations,
are non-volatile by conventional scientific standards, and are mixed with
other chemicals, soils, etc.
The uog's ability to detect and "home in" on chemical scents appears to
far exceed the capability of the most sensitive field monitoring equipment,
in the areas of sensitivity, speed, and overall cost-effectiveness.
Innovative training coupled with the amazing sensitivity of the dog's
olfactory senses allow contaminated areas to be delineated, at least
qualitatively, more rapidly and n»re completely than can be done using
existing instrumentation.
This preliminary or feasibility project demonstrates the utility of
competently trained dog/handler teams as an adjunct to the Agency's current
resources for locating and delineating hazardous wastes in the environment.
11
-------
Additional work is needed to verify the utility of dog/handler teams
for emergency response needs. Regionally located dog/handler teams should
be trained and made available to the Agency on an as-needed, contractual
basis for response to hazardous site release and spill emergency cleai jp
situations.
Thtre is reason to believe that the dog does quant Hate the level of
substance reaching his olfactory nerves. Because of changing conditions,
it is extremely difficult to quantify the dog's sensing ability under field
conditions, '-tore carefully controlled experiments are needed and should be
carrieo out in a regulated environment to determine whether the dog's
sensory ability can be quantified so that results obtained using the dog
can attain legal credibility.
ACKNOWLEDGEMENTS
The work on which this paper is based was performed under Contract No.
E3-7-14-9-MF for IT Corporation as a subcontract under Contract No.
68-03-3609 with the U.S. Environmental Protection Agency's Oil and
Hazardous Materials Spills Branch, Municipal Environmental
Research Laboratory, Edison, NJ. The encouragement of Mr. Michael
Gruenfeld (USEPA) and the cooperation of his staff in carrying out the
parallel field analytical work were a great he'p in conducting this
investigation. Of course, the major factor in the successful completion of
this project was the almost daily training efforts by Joyce Arner and her
dogs, "Justa", and Yellar", and by Melvin Manor and his dog, "Niner".
12
------- |