Introduction
With the publication of conductivity soundings near
a landfill at Canada Forces Base (CFB) Borden, McNeill (1980) introduced the
use of EM to map sites of environmental contamination. Over 150 cases published since then (Taylor,
2000) attest to the popularity that has been gained by this technique.
EM senses contamination that is present as
conductive pore fluid. Common
contaminants are leachate from municipal waste, and wastes from agriculture,
chemical manufacturing, mining and metallurgy, and petroleum production.
Taylor (1999) described a dual-geometry EM system
that geometrically sounds conductivity simultaneously to two distinct
depths. The simultaneous soundings
provide information about geological layering and the depth to contaminated
material.
Many sites contain shallowly buried metallic
containers. The distribution of metal
helps to define undocumented sites, and such containers may require special
treatment during site remediation. With
simultaneous measurement of two distinct responses to metallic objects, the
dual-geometry system enhances the capability of EM for this aspect of
environmental site mapping.
Dual-geometry
EM
The dual-geometry EM system incorporates a
transmitter coil with horizontal windings, a horizontal receiver coil and a
vertical receiver coil. As shown in
figure 1, the coils form both the horizontal co-planar geometry (HCP) and the
perpendicular geometry (PRP).
Figure 1: Dual-geometry EM
system
For geometric sounding, the dual-geometry system is
designed to operate at low induction number, so the depth of sounding is a
function of the separation between the transmitter and each receiver. Figure 2 shows the cumulative response with
depth for HCP and PRP. Based on these
curves, it is generally accepted that the HCP depth of sounding is 1.5
coil-separations, and the PRP depth of sounding is 0.6 coil-separations. Data in the examples that follow were
acquired with the DUALEM-4 (D-4), the dual-geometry EM instrument with 4-m coil
separation.
Figure 2: Dual-geometry depths
of sounding.
Metallic objects yield responses that are complex,
but typically detectable if the dimensions of the object are at least a
significant fraction of the coil separation, and if the object is within 1
coil-separation of the EM instrument.
HCP and PRP responses for such objects contrast in
amplitude and symmetry. In combination,
the responses enable a more accurate interpretation of the type and location of
an object.
Delineation of
Groundwater Contamination
The landfill at CFB Borden received ash, wood,
debris and some food waste between 1940 and 1976. The landfill sits on about 20 m of sandy soil, which decreases in
thickness to about 10 m at the northern edge of the area shown in figure
3. An aquiclude of clay underlies the
sandy soil.
Contours of chloride concentration in test wells,
from Sweeney (1983), indicate the extent of groundwater contamination. The plume manifests itself in induction logs
as a horizon with conductivity between 20- and 45-mS/m. The depth to the horizon is 6.5 m in
(borehole) BH 1.
The data of figure 4 were recorded at 1-s intervals
on a traverse of the plume. The D-4 was
carried at hip height, and at a speed of about 1.1 m/s along the gravel road
north of the landfill.
Figure 3: Chloride in
groundwater.
Figure 4: Traverse of
groundwater plume.
As dry, sandy soil extends more than 3 m below the
surface, PRP values are low, except for modest increases when the traverse
crosses the core of the plume (at 105 s), and a stream (at 410 s). The deeper-sounding HCP maps the location
and character of the plume; conductivities increase up to 10 mS/m over the core
of the contamination.
Detection of Buried Metal
The 200-L (55-gallon) steel drum is encountered
frequently during the course of environmental investigations. A drum of this type served as the target for
a series of test measurements of HCP and PRP response.
Figure 5: Guidelines for drum
depth.
Figure 6: Guidelines and WMU
drums.
The measurements were made in electrically resistive
surroundings. The drum was moved past
the D-4 along profiles 10 m in length, with measurements taken at 0.25-m
intervals. The distance from the D-4 to
the profiles ranged from 0.5 m to 4 m, in 0.25-m increments.
Figure 5 presents the PRP and HCP amplitudes for
each of the profiles. The amplitudes
are plotted against the distance between the D-4 and the profiles, on which the
top of the drum was positioned. With
increasing distance, the profile amplitudes decrease rapidly, as extrema
attenuate and broaden.
The logarithms of the amplitudes show excellent
correlation with the distance between the D-4 and the drum. The correlation is linear within the ranges
of distance where the profiles have similar shape and character. The linear relationships suggest the use of
simple guidelines, such as those of figure 5, for the interpretation of drum
depth.
Buried drums at Western Michigan University (WMU)
provide realistic targets for evaluating the interpretation of depth by
amplitude. The targets are described
(after W. Sauck, personal communication) in table 1.
Table 1: Description of WMU drums.
|
Target description |
Depth to Top (m) |
|
Drum, long axis aligned
with survey line |
0.5 |
|
Drum, long axis upright |
0.5 |
|
Drum, long axis transverse
to survey line |
0.5 |
|
Drum, long axis transverse
to survey line |
1.0 |
|
Drum, long axis transverse
to survey line |
2.0 |
|
3 crushed drums, random
orientation |
3.0 |
Two passes at walking speed were made over the
targets. For the first pass the D-4 was
carried at hip height, so the distance between the instrument and each target
was about 0.9 m greater than the nominal depth of burial. For the second pass, the D-4 was carried at
upper shin height, which made the distance between the instrument and each
target about 0.4 m greater than the nominal depth of burial.
Figure 6 shows the measured PRP and HCP amplitudes
plotted against the apparent distances between the targets and the D-4. The measured amplitudes cluster around the
guidelines for interpreting depth. The
guidelines are broader for distances where the sampling density at walking
speed can result in underestimation of amplitude.
The adherence of the measured amplitudes to the
guidelines indicate that a geophysicist can use D-4 data to interpret the depth
to the top of a drum, buried at depths to 3 m, to an accuracy better than 0.3
m.
Summary
Since 1980, EM has been used to map hundreds of
sites of environmental contamination.
EM maps contamination that is present as conductive pore fluid, or that
is associated with shallowly buried metallic containers.
At many sites, geological layering controls the
location and migration of contamination.
Geometric EM soundings yield simple and reliable information about
conductivity and depth. Dual-geometry
EM is especially useful, as it sounds conductivity simultaneously to two
distinct depths.
Dual-geometry EM produces two profiles over metallic
containers that are distinct in amplitude and symmetry, improving the depth of
detection and the interpretability of the responses from these significant
targets.
References
McNeill, J.D., 1980, Electromagnetic terrain
conductivity measurement at low induction numbers: Geonics Ltd., Technical Note
TN-6.
Sweeney, S.J., 1983, Concentration, distribution and
time variations of a contaminant plume in an unconfined sand aquifer: B.Sc.
Thesis, University of Waterloo.
Taylor, R.S., 1999, Development and applications of
geometric-sounding electromagnetic (G-SEM) systems: Society of Exploration
Geophysicists 1999 Technical Program Expanded Abstracts, Sixty-Ninth Annual
Meeting.
Taylor, R.S., 2000, An annotated bibliography of
applications of geo-conductivity meters: www.dualem.com/abib.html.
Acknowledgements
The author expresses his thanks to the staff of CFB
Borden and the Environmental Geophysics Facility at the University of Waterloo
for information about and access to the CFB Borden Test Site, to the staff and
students of the WMU Institute for Water Sciences for assistance with data
acquisition at the Asylum Lake Property, and to Dr. Scott Holladay of
Geosensors Inc. for helpful suggestions regarding data calibration.