Artículo en PDF
How to cite
Complete issue
More information about this article
Journal's homepage in redalyc.org
Sistema de Información Científica
Red de Revistas Científicas de América Latina y el Caribe, España y Portugal
Rev. Int. Contam. Ambie. 30 (3) 317-329, 2014
ELECTROMAGNETIC METHODS APPLICATION FOR CHARACTERIZING A SITE
CONTAMINATED BY LEACHATES
Salvador Isidro BELMONTE-JIMÉNEZ
1
*, Alberto BORTOLOTTI-VILLALOBOS
1
,
José Óscar CAMPOS-ENRÍQUEZ
2
, Marco Antonio PÉREZ-FLORES
3
, Omar DELGADO-RODRÍGUEZ
4
and María de los Ángeles LADRÓN DE GUEVARA-TORRES
1
1
Centro Interdisciplinario de Investigación para el Desarollo Integral Regional-Oaxaca, Instituto Politécnico
Nacional, Hornos 1003, Col. Noche Buena, Santa Cruz Xoxocotlán, Oaxaca, México. C.P. 71230
2
Instituto de Geofsica, Universidad Nacional Autónoma de México, Coyoacán, 04510, México D.F., México
3
Departamento de Geo±ísica de Exploración, Centro de Investigación Científca y Educación Superior de En
-
senada, Carr. Ensenada-Tijuana, 3918, Zona Playitas, Ensenada, Baja Cali±ornia, México. C.P. 22860
4
Instituto Mexicano del Petroleo, Eje Central Lázaro Cárdenas Norte 152, San Bartolo Atepehuacan, 07730,
México DF., México
* Autor de correspondencia; salvador.belmonte.j@gmail.com
(Recibido diciembre 2013; aceptado junio 2014)
Key words: contamination, geophysical methods VLF and LIN, electric conductivity, landfll
ABSTRACT
Two electromagnetic geophysical methods, very low ±requency (VLF) and low induction
number coils (EM-LIN) were used to obtain the response to the presence o± leachates
±rom a waste disposal site used ±or more than 24 years, covering an area o± 0.16 km
2
. This
landfll is located in ±ractured shale and sandstone associated with the Oaxaca Fault. The
study was per±ormed on six profles, ±our o± which were common to both methods, with
lengths o± 325, 320, 300 and 645 m, in two others only the VLF method was used. The
interpretation o± VLF data using the Hjelt and Karous flter resulted in current density
sections. The current density variation was assumed to indicate the presence o± ±ractures,
along which the infltration o± leachate takes place. The interpretation o± EM-LIN data
provided two-dimensional models showing the distribution o± the conductivity o± the
subsoil. The integration o± these results shows a main conductive anomalous zone in the
southeastern part o± the landfll that increases in thickness towards the middle and with
a depth up to 30-40 m. Correlation with natural sur±ace runo±± enables to in±er that the
conductive anomalous body indicates the presence o± leachates. Both electromagnetic
methods provided a good response in fractured zones.
Palabras clave: contaminación, métodos geo±ísicos VLF y LIN, conductividad eléctrica, tiradero de basura
RESUMEN
Se utilizaron dos métodos geo±ísicos electromagnéticos, ±recuencia muy baja (VLF, por
sus siglas en inglés) y bobinas a bajo número de inducción (EM-LIN, por sus siglas en
inglés), para estudiar la presencia de lixiviados de un sitio de disposición de residuos
sólidos municipales, el cual opera desde hace más de 24 años y cuya superfcie es de
0.16 km
2
. El tiradero se encuentra en un medio cuyo ±racturamiento está asociado con la
Falla Oaxaca. El basamento del tiradero está constituido por lutita y arenisca. El estudio
S.I. Belmonte-Jiménez
et al.
318
se realizó en seis perfles de los cuales cuatro Fueron comunes para ambos métodos, con
longitudes de 325, 320, 300 y 645 m, y dos más, únicamente usando el método VL±. La
interpretación de los datos VL± utilizando el fltro de Karous y Hjelt dio como resultado
secciones que indican alta densidad de corriente interpretadas como asociadas a Fracturas
que Favorecen la infltración de los lixiviados. La interpretación de los datos de EM-LIN
brinda modelos bidimensionales que muestran la distribución de la conductividad del sub
-
suelo. La integración de estos resultados indica una zona anómala conductora principal en
la parte sureste del tiradero que, de acuerdo a su Forma geométrica, se incrementa hacia la
parte central del mismo y se observa hasta una proFundidad de entre 30 y 40 m, coincidiendo
con la zona por donde ocurre el escurrimiento superfcial natural a lo largo del cual ²uyen
los lixiviados, y que han aprovechado algunas Fracturas para infltrase, infriéndose que
las zonas anómalas detectadas son debidas a la presencia de estos contaminantes. Ambos
métodos electromagnéticos han proporcionado buena respuesta en un medio fracturado.
INTRODUCTION
A study was conducted in a landfll, located
about 15 km to the south oF Oaxaca city, along
the road From Oaxaca City to Puerto Angel, in
the District of Zaachila (
Fig. 1
), where more than
500 tons oF wastes per day are deposited. This site
represents an important environmental problem
because of the presence of leachates that seep into
the ground, contaminating the surFace water and
the local aquiFer. Due to the need For Fresh water,
people dig shallow wells near this dump, risking
health problems as in other places (i.e., Adepelumi
et al
. 2005, Samsudin
et al
. 2006). The possibilities
oF contamination oF the aquiFer in the landfll area
oF Oaxaca city are high, since according to studies
conducted to evaluate vulnerability to ground
-
water contamination (Aragón
et al.
2006), the
transit Factor oF the landfll infltration ranges From
0.172 × 10
–6
to 34.546 × 10
–6
/s i.e., over the maxi-
mum value, 3 × 10
10
/s, established by the Mexican
OFfcial Standard (NOM-083-SEMARNAT-2003 in
SEMARNAT 2013). This large transit ensures that
leachates infltrate the aquiFer with a signifcant
contaminant charge.
In this situation, any action taken to diminish the
pollution caused by leachates will likely support the
economic development oF the region. ±ortunately,
geophysics plays an important role by helping to
determine the geometry, spatial extent, and depth oF
a contamination plume (i.e., Soupios
et al.
2007).
Geophysics has been used to explore the underground
and determine iF the leachates originated in landflls
reached the depth level where the aquiFer is Found
(i.e., Busquets and Casas 1995, Mondelli
et al
. 2007).
Among others advantages, geophysical methods are
fast and are non-destructive, less expensive than di-
rect methods and provide an overview oF the study
area. Some authors like Cossu
et al.
(1991) and
Soupios
et al
. (2007) report that geophysical methods
are favorable for investigation of dumps because
they usually contain materials characterized by a
high electrical conductivity that can be detected by
geoelectrical methods. Karlik (2001) recommended
direct current (DC) methods and VLF as tools for
mapping groundwater contamination and to deter
-
mine the extension of the contamination plume.
In this paper we report on the use oF two electro
-
magnetic methods, specifcally, VL± and EM-LIN
(working at low induction numbers similar to EM-
34, EM-31 and EM-38 equipments), to evaluate
their response in a Fractured zone aFFected by leach
-
ates. In the EM-LIN case, measurements were taken
in both horizontal and vertical loops modalities for
the three separations between the coils allowed by
the EM-34. This is equivalent to six measurements
For every site, increasing the inFormation obtained
From the ground. VL± measurements were taken
at three diFFerent Frequencies in order to select the
96°41’0"W
96°41’0"W
96°41’30"W
96°41’30"W
96°42’0"W
96°42’0"W
96°42’30"W
96°42’30"W
96°43’0"W
96°43’0"W
96°43’30"W
96°43’30"W
16°56’30"N
16°56’30"N
16°56’0"N
16°56’0"N
16°55’30"N
16°55’30"N
16°55’0"N
16°55’0"N
16°54’30"N
16°54’30"N
00
.5
11
.5
0.25
Km
Scale 1:25 000
Legend
North
Well
Rivers
Landfill
Urban area
Contours
Roads
Fig
.
1.
Location map oF the landfll oF the Oaxaca city
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
319
ground station with the best coupling with the un
-
derground structure.
Hydrogeology of the studied area
The garbage landfll is located in theAtoyac River
basin.TherunoFFscoeFfcientsrangebetween10to20%
and they are Favored by the regional topographical
slope. The drainage network consists oF a basin system
through which the superfcial runoFFs ±ow intermit
-
tently. In this system the landfll plays an important
role, since the leachates move permanently to the
surFace drainage. The number oF wells in the area is
low and the main groundwater ±ow direction is toward
the southwest (Belmonte
et al.
2005). The aquiFer is
Free, with the water table located at depths between 6
to 10 m in consolidated and strongly Fractured rocks,
since due their origin to the tectonic activity oF the
Oaxaca ²ault that controls the regional groundwater
±ow. There is secondary permeability in some areas.
Electromagnetic prospection at low induction
numbers
The EM-LIN method uses two loops, one as
source and the other as receiver. The source circu-
lates a current along the coiled wire at a Frequency
depending on the distance between source and re
-
ceiver (
Table I
). The source induces a magnetic feld
with the same Frequency into the ground. IF both
loops are over a whole space, the measurement is
predictable and is named primary feld (
Hp
). When
a halF space is present, a secondary feld will appear
(
Hs
). For a homogeneous half space,
Hs
is known
and the rate between both felds is (McNeill 1980):
4
2
0
s
iw
H
H
P
S
μ
=
σ
(1)
Where:
H
S
= Secondary magnetic feld on the receiver.
H
P
= Primary magnetic feld on the receiver.
w
= 2
π
f.
f
= ²requency in Hz.
µ
0
= ²ree-space permeability.
σ
= Ground conductivity oF the homogeneous halF
space (mS/m).
s
= Distance between coils in m.
i
= (–1)
1/2
When the Earth is an inhomogeneous half-space,
the conductivity in (1) becomes an apparent one.
Using this equipment (EM-34, EM-31, and EM-38
From Geonics and GEM-5 From Geoplex), we can
obtain the ground apparent conductivity From the
Following equation:
2
0
4
ws
H
H
P
S
a
μ
=
σ
(2)
Using vertical or horizontal coplanar loops, we
can get profles oF apparent conductivity at diFFerent
separations (corresponding to several penetration
depths). Such profles give us a very broad idea oF
the conductivity distribution in the ground.
In order to obtain the depth, geometry and true
conductivity oF the buried body, it is necessary to invert
the apparent conductivities. In a two-dimensional Earth
(2D), this problem can be posed in terms of the integral
equation as the scattering equations. However, we use
another integral equation that relates the magnetic felds
measured by the receptor and the ground conductivity
(Gómez-Treviño 1987). Applying the approximation
For low conductivity contrasts, the equation becomes
(Pérez-²lores 1995, Pérez-²lores
et al.
2001):
()
()
()
dxdz
z
x
x
x
z
x
F
x
x
x
x
ff
i
z
x
x
a
,
,
,
,
,
0
2
1
2
1
2
1
=
σσ
(3)
Where
σ
a
is the apparent conductivity obtained
from (2) that depends on the source (
x
1
) and re-
ceiver (
x
2
) positions. IF we assume a halF-space
discretized by a grid oF rectangular prisms with
center coordinates (
x, z
), and every prism having
a constant conductivity (s), then
F
constitutes a
weighting Function. This means that every mea
-
surement is a kind oF a volumetric average oF the
product oF s’s by the corresponding value oF
F
. This
function depends on the source and receiver posi-
tions, the characteristics of the grid (dimensions)
as well as on the magnetic felds induced by the
loops. When the loops are horizontally coplanar,
this Function is:
TABLE I.
ESTIMATED EXPLORATION DEPTH ACCORD-
ING THE COIL SEPARATIONS AND FREQUENCY
USED
Coil
separation
(m)
Used
Frequencies
(Hz)
Estimated exploration depth (m)
Horizontal
dipole
Vertical
dipole
10
6400
7.5
15
20
2600
15
30
40
600
30
60
S.I. Belmonte-Jiménez
et al.
320
()
()
()
()
[]
()
[]
dy
z
y
x
x
z
y
x
x
x
x
x
x
y
x
x
z
x
F
+
+
+
+
+
=
3
2
2
2
2
3
2
2
2
1
2
1
2
2
1
,
,
,
(4)
The prisms have infnite extension in the
y
direc-
tion. This integral has an analytic solution (Pérez-
Flores 1995, Pérez-Flores
et al.
2001).
For vertical coplanar loops, equation (4) is more
complex:
()()
()
()
()
[
]
dy
x
z
y
x
E
x
z
y
x
E
x
z
y
x
E
x
z
y
x
E
x
x
z
x
F
y
y
x
x
+
=
2
1
2
1
2
1
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
(5)
Where
E
x
and
E
y
are the electric felds in the
x
and
y
directions for both, source (
x
1
) and receiver
(
x
2
). The expressions are developed and explained in
Pérez-Flores (1995) and Pérez-Flores
et al.
(2001).
Equation (3) can be expressed as a linear equa
-
tion system:
F
=
a
σσ
(6)
Where
is a vector with the apparent conductivi
-
ties,
F
is the weighting matrix and
is the vector
with the true conductivities o± the ground (unknown).
This linear system can be solved in many ways. In
this work we used quadratic programming (Gill
et
al.
1986) and smoothing ±actors ±or the numerical
stabilization.
VLF prospecting
VLF instruments are lightweight and portable, and
they can be used to study large areas quite quickly
(Liu
et al
. 2006). The VLF method is based on the
use o± very low ±requency radio waves (in the range
o± 15 to 30 kHz) ±or exploration o± ±ractured zones,
groundwater contamination and minerals (Jeng
et
al.
2004, Drahor 2006, Dutta
et al
. 2006, Ganerod
et al
. 2006, Zlotnicki
et al.
2006, Kaya
et al
. 2007).
It helps to determine the electrical characteristics of
the underground and shallow rocks (Hutchinson and
Barta 2002). There are VLF stations transmitting, for
marine communication primary purposes, electro
-
magnetic signals traveling between the ionosphere
and the Earth’s surface.
The signal emitted by the antennas around the
world can be captured in the feld by means o± VLF
instruments, and according to the basic electromag-
netic theory, at long distances ±rom the source, the
wave±orm approaches a plane wave (Zlotnicki
et al
.
2006). There is a relation o± primary magnetic feld
(
Hp
) and magnetic secondary feld (
Hs
) created by a
conductive body that acts as a second source (Kaya
et al
. 2007). This means that electric currents in the
conducting body (e.g., a ±racture) are generated when
radio waves (EM feld) pass through it, creating an
-
other magnetic feld (
Hs
).
The presence o± ±aults and ±ractures in a hy
-
drogeological system contributes to increase the
hydraulic conductivity and porosity acting, in turn,
as structures controlling groundwater ²ow (Sharma
and Baranwal 2005, Adepelumi
et al
. 2006). Thus,
±ractures may also become pre±erential pathways
±or the ²ow o± leachate, signifcantly increasing the
electrical conductivity o± the subsur±ace (Mondelli
et al
. 2007, Soupios
et al
. 2007). Depth of penetra-
tion depends largely on ground conductivity, but
according to Oskooit and Pedersen (2005) this is
less than 100 m.
The resulting vector from the sum of
Hp
and
Hs
produces a time-varying elliptically polarized feld.
This elliptical shape has two components with the
same ±requency, but di±±erent amplitude and phase.
The in-phase amplitude
Hp
is the real component,
while the out o± phase
Hp
is the imaginary component
or quadrature component (Eze
et al
. 2004).
The electromagnetic feld equation ±or a conduc
-
tive medium can be represented by the Helmholtz
equation derived ±rom the Maxwell equations:
=
H
E
w
i
H
E
μ
2
σ
(7)
Where
E
and
H
are respectively the electric and
magnetic felds,
σ
(mS/m) the conductivity, μ per
-
meability (Henry / m) and ω the angular ±requency.
In contrast, both the tilt angle (θ) and ellipticity (e)
are calculated using the ±ormula proposed by Smith
and Ward (1974, see also Sharma and Baranwal
2005, and Dutta
et al.
2006). Once simplifed, they
are expressed as:
)
cos
(
tan
1
sen
H
H
p
s
=
(8)
Where
Hp
is the primary feld,
Hs
is the secondary
feld,
ϕ
is the change o± phase between
Hp
and
Hs
.
Hs
is tilting, and
α
represents the angle above
Hp
due to the coupling between the transmitter and the
underground structure. Then, it is defned
Hs
sen
α
= Δ
Hy,
thus equation (8) becomes.
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
321
)
cos
(
tan
1
p
y
H
H
=
(9)
Where Δ
H
y
cos
ϕ
= real component or in-phase of
the
Hs
feld.
The tangent of the tilt angle is proportional to the
Hs
real component, which is measured in the verti
-
cal direction. Therefore, the measurement of the tilt
angle is very similar to measuring the real component
(in-phase) of
Hs
in the vertical direction.
VLF data can be enhanced by applying fltering
procedures. The flter application is essential to obtain
a reasonable correlation between the anomaly and the
structure. The flters are designed to decrease noise
±rom the EM signal. Fraser, as well as Karous and
Hjelt flters, are two methods widely used in VLF
data processing.
Fraser flter is a low-pass ±unction ±or estimating
the average o± tilt angle measurements produced by
a subsur±ace conductor. In a linear sequence o± tilt
angle measurements
M
1
,
M
2
,
M
3
, .
..,
M
n
, the Fraser
flter F
1
is expressed as:
)
(
)
(
2
1
4
3
1
M
M
M
M
F
+
+
=
(10)
The frst value
F
1
is located between
M
2
and
M
3
positions, the second value between
M
3
and
M
4
, and so
on. There are several studies using Fraser flter (Ade
-
pelumi
et al
. 2005, Cossu
et al
. 1991, Liu
et al
. 2006,
Monteiro-Santos
et al.
2006, Zlotnicki
et al.
2006).
Karous and Hjelt (1983) developed a statisti-
cal linear ±ilter, based on Fraser´s one, which
provides a profle o± current density vs. depth (
H
0
)
and is derived from the magnitude of the vertical
component o± the magnetic feld (
Hz
) in a specifc
position. These authors used linear fltering ±or the
analysis o± VLF dip-angle data in an extension o±
the Fraser flter.
They describe the magnetic feld, arising ±rom a
subsurface 2D current distribution, assumed in a thin
horizontal sheet o± varying current density situated
everywhere at a depth equal to the distance between
the measurement stations. Their technique involves
fltering the same data set ±or various depths and gives
an idea about the conductivity with depth, since high
current density corresponds to good conductors. This
technique has ±ound wide popularity as it provides
a simple, readily implemented scheme ±or semi-
quantitative analysis and target visualization. The
apparent current density pseudo section should
provide a pictorial indication of the depths of
various current concentrations and hence the spatial
distribution o± subsur±ace geologic ±eatures (Ogilvy
and Lee 1991).
When we measure over conductors, the in-phase
part o± the equivalent current distribution has only
positive values. Negative parts on both sides of
the conductor can be caused either by the length o± the
flter or by a decrease o± current density due to current
gathering which is not present in 2D structures (Nab
-
ighian 1982). In the simplest ±orm, the Fraser flter is:
)
2
/
(
205
.
0
323
.
0
446
.
1
446
.
1
323
.
0
205
.
0
)
2
/
(
2
3
2
1
0
1
2
x
I
H
H
H
H
H
H
x
I
z
a
a
+
+
+
=
(11)
Where Δ
z
is the assumed thickness o± the cur
-
rent sheet,
I
a
, is the current density,
x
the distance
between data points and also the depth to the current
sheet. The
H
2
through
H
3
values are the normalized
vertical magnetic feld anomalies at each set o± six
points. The location o± the calculated current density
is assumed at the geometrical center of the six data
points (Sundararajan
et al.
2007).
METHODOLOGY
Measurements were conducted along six profles
with the EM-LIN and VLF methods; ±our profles
are common for both methods, respective lengths of
325, 320, 300 and 645 m (
Fig. 2
). Three profles were
made in the southern side, one on the northern side
o± the landfll with E-W direction. Two additional
Garbage Area
plant
Processing
Tank
Profile 1
Dirt road
Azucenas
Street
Natural drain
Profile 2
Profile 3
Profile 6
Profile 4
Profile 5
Perimeter of the
landfill
Los Pinos street
Main access
Guard
house
Fig.
2.
Location o± the studied profles. Profles 1 to 4 were
measured with EM-LIN, 5 and 6 profles only with VLF
S.I. Belmonte-Jiménez
et al.
322
VLF profles were made: one in the middle part o± the
landfll with E-W direction. The other one on the
western side, with a N-S direction: lengths respec
-
tively o± 415 and 275 m. In all cases the measure
-
ments were made every 5 m.
Structural ±eatures ±rom rock outcrops, ±racture
degree and its pre±erred direction were also measured.
For the topographic survey, anAshtech Promark GPS
post processing equipment was used.
The VLF survey was done with a Scintrex
equipment with transmitting signals ±rom three sta
-
tions: NAU located in Aguada, Puerto Rico, with a
±requency o± 28.5 kHz; NPM located in Lualualei,
Hawaii, with a ±requency o± 23.4 kHz and NSS lo
-
cated inAnnapolis, Maryland, USA, with a ±requency
o± 21.4 kHz. For the respective data processing, the
so±tware KHFFILT-2006 was used (prepared by Pirt
-
tijärvi 2004), to apply the Karous and Hjelt (1983)
and Fraser (1969) fltering to VLF data.
The control source electromagnetic data were
measured with an EM-LIN equipment (Geonics),
consisting of transmitter and receiver coils; per-
±orming measurements with three di±±erent intercoil
spacing (10, 20 and 40 m) at di±±erent ±requencies
(6400, 2600, and 600 Hz) respectively, and in two
±orms: horizontal coplanar coils (vertical magnetic
dipole) and vertical coplanar coils (horizontal mag-
netic dipole). Data processing for the electromagnetic
coils method were per±ormed using the so±tware
CICEM35-2006. This program applies the theory
developed by Pérez-Flores (1995) and Pérez-Flores
et
al.
(2001) ±or the EM data inversion, which considers
the measurements obtained with EM-34 equipment
as a weighted average o± the earth conductivity
distribution.
ANALYSIS AND DISCUSSION OF THE
RESULTS
The electromagnetic LIN array comprised several
source-receiver separations. For horizontal and verti-
cal coplanar loops, we obtained ±our pseudo-sections
(images) o± the apparent conductivity (mS/m) along
the profiles comprising several studied depths.
Figure 3
shows, as an example, ±or profle 1, ±or a
vertical dipole (
Fig. 3a
) as well as ±or a horizontal
dipole (
Fig. 3b
), the behavior of electrical conductiv-
ity along the profle, ±or di±±erent depths.
The six data sets ±or profle 1 are di±±erent because
every array looks at the ground in a di±±erent way
(
Fig. 3
). However, they show a general behavior
that can give us a very broad idea o± the conductivity
variation with depth and horizontally. The observed
conductors are assumed to be related with conduc
-
tive materials like contaminant ²uids contained in
the porous media. In order to merge those six data
sets to get one conductivity model or image o± the
ground conductivity, we joint inverted them into a
model whose response fts the six data sets very well
(Pérez-Flores
et al.
2001). In
fgure 4
are shown the
conductivity images or models ±or the ±our EM-LIN
profles located in
fgure 2
.
Regarding the VLF data, only the in-phase com
-
ponent was processed.
Figure 5
shows the behavior
o± the in-phase and quadrature curves ±or profle 1 in
order to illustrate the data collected in feld.
The interpretation o± VLF data was based on flter
-
ing procedures ±ollowing Fraser (1969) and Karous
and Hjelt (1983) methods widely used by other authors
(i.e., Benson
et al
. 1997, Sundararajan
et al
. 2007).
Fraser flter turns the crossing points into peak signals
that enhance the conductive structures. In
fgure 6
for
profle 1, we can appreciate the results o± Fraser flter,
showing the percentage (%) in the y-axis. In particular
a signifcant negative value can be observed at position
125 m due to changes in the underground conductivity.
Karous and Hjelt (1983) fltering was used to obtain
current density pseudo-sections (mA/cm
2
) for the six
VLF profles (
Fig. 7
). Processed data are presented
with iso-contour lines o± current density. Low current
density values correspond to high resistivity values.
Profile 1. Vertical dipole
a
b
Profile 1. Horizontal dipole
400
350
300
250
200
150
100
50
0
0
100
20
03
00
400
Distance (m)
Conductivity (mS/m)
DV10
DV20
DV40
400
350
300
250
200
150
100
50
0
0
100
20
03
00
400
Distance (m)
Conductivity (mS/m)
DH10
DH20
DH40
Fig.
3.
Behavior o± electrical conductivity along profle 1. a)
vertical dipole, and b) horizontal dipole
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
323
By correlating the EM-LIN conductivity image
with the corresponding VLF current density pseudo-
section, it is possible to construct a geological-
geophysical model for every pro±le. In the following
interpretation, only anomalies well de±ned by sev
-
eral neighboring assignment points were taken into
05
0
100
150
200
250
300
Distance (m)
60
100
140
180
220
260
300
340
Profile 1
Depth
–15 m
–30 m
–60 m
NORTH
SW
NE
Conductivity (mS/m)
05
01
00
15
02
00
25
03
00
5
20
35
50
65
80
95
–15 m
–30 m
–60 m
NORTH
SW
NE
Conductivity (mS/m)
Distance (m)
Depth
Profile 2
100
120
140
160
180
200
220
240
260
280
Conductivity (mS/m)
Distance (m)
NE
Depth
–15 m
–30 m
–60 m
SW
05
0
100
150
200
250
300
NORTH
Distance (m)
NE
50
70
90
110
130
150
170
190
210
230
Conductivity (mS/m)
Depth
–15 m
–30 m
–60 m
SW
05
0
100
150
200
250
300
NORTH
Profile 3
Profile 4
Fig.
4.
2D sections obtained by data inversion. See
Fig.
2
for location of pro±les
S.I. Belmonte-Jiménez
et al.
324
account (i.e., anomalies defned by several contigu
-
ous points along the horizontal and along the verti-
cal). Also anomalies located at the upper end points
were not considered (i.e., only the middle portion).
In general there is a Fair-to-good agreement between
conductivity zones imaged by the EM-LIN inversion,
and the areas oF high current density obtained From
the VL± data. Each oF the high conductivity zones
interpreted and reported below is supported by both
data sets types, as well as by the presence at surFace
of leachates.
Along profle 1 (
Fig. 8
), located near the southern
perimeter oF the landfll, two conductive zones are ob
-
served at the central part (at 200 m, and around 300 m),
which are the most important in terms oF spatial dis
-
tribution (see
Fig. 4a
). The anomaly at 200 m is also
represented as current density anomalies (
Fig. 7c
),
thereby it is interpreted as a Fracture zone. In the VL±
section a narrow anomaly is observed around 150 m,
which is also assumed a Fracture zone. Both zones
lie below a local topographic depression (see profles
2 and 3), and below the surfcial leachate ²ow. The
topography is irregular, the geology is represented by
alternations of sandstone-fractured shale.
The geological-geophysical model associated
with profle 2 (
Fig. 9
), which is located outside the
landfll, to the south oF profle 1, shows three major
conductive zones (between 0 and 50 m, between 100
and150m,andbetween200and300m)(
Fig. 4b
). Ac-
cording to VL±, there is a Fracture zone between 225
and 275 m (i.e., correlating with the third conductive
zone) (
Fig. 7d
). Another fracture zone might be lo-
cated around 175 m. The third conductive anomaly
is associated with the presence oF Fractures with N-S
80
60
40
20
0
05
0
100
Distance (m)
Profile 1 VLF
150
20
02
50
30
03
50
–20
–40
–60
–80
Phase
Quadrature
Fig.
5.
Behavior oF the in-phase and quadrature oF profle 1
50
Profile 1
0
05
0
100
150
Distance (m)
20
02
50
Real component
300
–50
Response
–100
–150
Fig.
6.
Profle 1 where the real component or phase is processed
by the ±raser ±ilter
–10
100
200
300
40
05
00
600
Depth (m)
Profile 4
NORTH
a)
–20
–30
–10
100
200
300
40
05
00
600
Depth (m)
Profile 6
b)
–20
–30
–10
100
200
300
40
05
00
600
Depth (m)
Profile 1
c)
–20
–30
–10
100
200
300
40
05
00
600
Depth (m)
Profile 2
d)
–20
–30
–10
100
200
300
40
05
00
600
Depth (m)
Profile 3
e)
–20
–30
–10
100
–60
Equivalent current density (mA/cm
2
)
–50
–40
–30
–20 –1
00
10
200
300
40
05
00
600
Depth (m)
Profile 5
f)
–20
–30
Fig.
7.
a, b, c, d, e and F. VL± profles processed with Karous-
Hjelt flter. The scale oF values represents the current
density (mA/cm
2
), associating positive values to con-
ductive areas because of the content of leachate seeping
into the fractured medium
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
325
direction (and in a lesser proportion with an E-W
orientation), through which leachates are infltrat
-
ing. The anomalies have a lesser vertical extension.
The topography is not Fat. There is a topographic
local depression at the eastern half, and the ground
comprises an alternation o± shale and sandstone with
a sur±ace layer o± the same materials but altered and
weathered.
Profle 3 (
Fig. 10
), located outside the landfll,
to the south o± profle 2, presents only a minor
±eature that can be associated with the presence o±
leachates at the subsurface (
Fig. 4c
; around 200 m).
This indicates that the contaminant plume does not
continue southwards, or it has deepened and it can
not be sensed due to the limited penetrating power
o± the two used methods. VL² indicates the pos
-
sible existence o± three ±ractures zones (between 50
and 100 m, around 200 m, and another centered in
250 m;
Fig. 7e
). The last ±racture would be located
below the surfcial leachates Fow, and would point
to the infltration o± leachate into the ground through
it, and can be interpreted as a leachate plume.
Geologically, this profle consists o± a thin layer
o± sand-shale rock ±ragments and then alternation o±
shale and sandstones.
Figure 11
shows the geological-geophysical
model ±or profle 4, which is the longest and located
in the northern portion o± the landfll in an area where
Simbology
N
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
Profile 2
Distance (m)
Deph (m)
0
20
40
0
200
100
300
325
Layer of alterated rocks fragments
(Sandstone and Shale)
NE
SW
VLF
VLF
VLF
Fig.
8.
Geological model along profle 1, with superposed leachate presence in±erred by LIN and
±ractures by VL² EM methods.
Profile 1
Distance (m)
Depth (m)
0
20
40
02
00
100
30
0325
SW
NE
Compacted backfill with
fragmented rocks
(Sandstone and Shale)
Simbology
N
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
Fig.
9.
Geological model along profle 2, with superposed leachate presence in±erred by LIN and
±ractures by VL² EM methods.
Profile 3
Distance (m)
Depth (m)
0
20
40
02
00
100
30
0325
SW
NE
Layer of altered
rocks fragments
(Sandstone and Shale)
Simbology
N
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
Fig.
10.
Geological model along profle 3, with superposed leachate presence in±erred by LIN and
±ractures by VL² EM methods
S.I. Belmonte-Jiménez
et al.
326
Profile 4
Distance (m)
0
20
40
200
0
SW
NE
Layer of clay filled garbage
60
400
600
645
Simbology
N
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
Depth (m)
Fig.
11.
Geological model along profle 4, with superposed leachate presence inFerred by LIN and
Fractures by VL± EM methods
Profile 5
Distance (m)
Depth (m)
0
20
40
02
75
NW
SE
Layer of clay filled garbage
projected
N
100
200
250
Simbology
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
Fig.
12.
Geological model along profle 5, with superposed leachate presence inFerred by LIN and
Fractures by VL± EM methods
Profile 6
Distance (m)
0
20
40
0
100
400
415
SW
NE
Depth (m)
200
300
350
Simbology
Shale
Sandstone
Leachate plume
Fractured zone
Sandstone and Shale
fragmented rocks
N
Layer of clay filled garbage
Fig.
13.
Geological model along profle 6, with superposed leachate presence inFerred by LIN and
Fractures by VL± EM methods
in the past waste was deposited but not anymore. Two
anomalies are mapped at 100, extending From 250
and towards 300 m (
Fig. 4d
). Two relative anomalies
are observed between 375 and 425 m. VL± indicates
fractures centered at 200 and at 400 m (
Fig. 4a
), cor-
relating with the second conductive zone. At feld,
leachates were observed percolating through these
fractures.
The geological-geophysical model oF the profle
5 is presented in
fgure 12
, where two zones with
high current density are interpreted as Fractures (at
85 m, and one centered at 200 m;
Fig. 7f
). Onto this
profle is projected the location oF the main body oF
garbage with a clay flling in the middle oF it, and
with leachates observed at its surFace and whose
spatial distribution was obtained From VL± and EM-
LIN measurements along profles 1, 2, 3, 4, and 6.
The geology corresponds to an alternation oF shale-
sandstone. This profle is located in the western part
oF the landfll.
Figure 13
corresponds to the geological-geophys
-
ical model oF the profle 6. This profle was conducted
at the center oF the landfll, so almost halF oF it is
over a layer oF garbage with a thickness From 10 to
15 m oF depth. VL± indicates the presence oF Four
possible Fracture zones (at 65 m, at 150 m, at 225
m, and between 300 and 400 m) (
Fig. 7b
). It sounds
logic to suppose that through the two easternmost
Fractures leachates percolates, since they are located
below the local topographic depression where the
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
327
maximum thickness of wastes is located. The geol
-
ogy is represented by an alternation of fractured shale
and sandstone.
The analysis of all the models shows that the
conductive areas are related to the presence of leach-
ates and may suggest a continuous Fow of them
through the land±ll of Oaxaca city, and also that the
subsurface layers are being impregnated with this
contaminant Fow through the fractures up to depths
of more than 30 m.
Figure 14
shows interpolated images of apparent
conductivity measured with the EM-LIN considering
the respective four pro±les, in the con±guration-
modality of horizontal coplanar coils (vertical di
-
pole). It can be observed that at different depths, the
contaminant plume is larger in the N-S preferential
direction of the leachate Fow. We can observe that
the high conductivity anomaly correlates with the
land±ll limits. ²ractures inferred from VL² data are
superposed. Two fracture systems can be observed
to run with a NW-SE direction.
CONCLUSIONS
The Oaxaca city garbage land±ll is located in a
fractured zone where the VL² and EM-LIN electro
-
magnetic geophysical methods were applied to study
the presence of leachates in the subsoil. Accord-
ingly, the preferential Fow of leachates is through
the set of NW-SE fractures. We de±ned their spatial
distribution. The data obtained from EM-LIN coils
show high conductivity zones, interpreted to be due
to the presence of leachates. From the information
Fig.
14
. Perspective at different depths for the EM coils processed data, corresponding to the horizontal (A) and vertical (B, C, and
D) dipole mode
AB
CD
3.0 3.7
4.5
5.5
6.7
8.2
10
12
15
18
22
746200
746300
746400
746500
746600
746700
746200
746300
746400
746500
746600
746700
1872800
1872700
1872600
1872500
1872800
1872700
1872600
1872500
Vertical dipoles
Intercoil spacing 10 m
Exploration depth = 15 m
Vertical dipoles
Intercoil spacing 40 m
Exploration depth = 60 m
Vertical dipoles
Intercoil spacing 20 m
Exploration depth = 30 m
Horizontal dipoles
Intercoil spacing 10 m
Exploration depth = 7.5 m
ρ
a
, Ohm.m
S.I. Belmonte-Jiménez
et al.
328
obtained from the VLF method (in-phase component)
about 13 fractured zones were identiFed along the
studied proFles, favoring the leachate percolation
through the fractured zone. In general, there is a
good correlation between high conductivity areas
obtained with EM-LIN and those high current densi
-
ties obtained with VL±.
Considering that the water table of the surround
-
ing areas is located at a depth less than 15 m, there
are high possibilities that the aquifer could be con
-
taminated. In the outskirts of the landFll there is a
semi-conFned aquifer, however, the fractures favor
leachates ²owing inside the area, allowing the leach
-
ate to inFltrate to greater depths. There are anomalous
conductive areas with depths varying from 60 to 35 m
in the case of EM-LIN and VLF method of coils
respectively.
±inally, it is considered that the use of these two
methods (VLF and EM-LIN) for the detection of
leachates in the subsoil is suitable as a technique for
assessment of contaminated areas. It is faster than
other geophysical methods and not invasive.
REFERENCES
Adepelumi A.A., Ako B.D., Afolabi O. y Arubayi J.B.
(2005). Delineation of contamination plume around
oxidation sewage-ponds in Southwestern Nigeria.
Environ. Geol. 48, 1137-1146.
Adepelumi A.A., Yi M. J., Kim J.H., Ako B.D. y Son J.S.
(2006). Integration of surface geophysical methods for
fracture detection in crystalline bedrocks of southwest
-
ern Nigeria. Hydrogeol. J. 14, 1284-1306.
Aragón S. M., Belmonte J. S. I. y Navarro M. S (2006).
Vulnerabilidad del acuífero del tiradero municipal de
la ciudad de Oaxaca a la contaminación subterránea.
Memorias. XV Congreso Nacional de Ingeniería Sani-
taria y Ciencias Ambientales. Guadalajara, Jal. 24 al
26 de mayo, 2006. CD-ROM.
Belmonte-Jiménez S.I., Campos-Enríquez J.O. yAlatorre-
Zamora M.A. (2005). Vulnerability to contamination
of the Zaachila aquifer, Oaxaca, Mexico. Geofís. Int.
44, 283-300.
Benson A.K., Payne K.L. y Stubben M. A. (1997). Map
-
ping groundwater contamination using dc resistivity
and VL± geophysical methods - A case study. Geo
-
physics 62, 80-86.
Busquets E. y Casas,A. (1995). Caracterización de vertede
-
ros y detección de penachos contaminantes mediante
la utilización de métodos geofísicos. Física de la Tierra
(Servicio
de Publicaciones de la Universidad de Com-
plutense) 7, 207-226.
Cossu R., Ranieri G., Marchisio M., Sambuelli L., Godio
A. y Motzo G.M. (1991). Geophysical methods in sur
-
veying old landFlls. Proceedings of the Contaminated
soil ‘90 Third International KfK/TNO Conference.
Karlsruhe, ±ederal Republic of Germany. December
10-14. I, 575-582.
Drahor M.G. (2006). Integrated geophysical studies in
the upper part of Sardis archaeological site, Turkey. J.
Appl. Geophys. 59, 205-223.
Dutta S., Krishnamurthy N.S.,Arora T., Rao V.A.,Ahmed
S., y Baltassat J.M. (2006). Localization of water bear
-
ing fractured zones in a hard rock area using integrated
geophysical techniques in Andhra Pradesh, India.
Hydrogeol. J. 14, 760-766.
Eze C.L., Mamah L.I., y Israel-Cookey C. (2004). Very low
frequency electromagnetic (VL±-EM) response from a
lead sulphide lode in theAbakaliki lead/zinc Feld, Nige
-
ria. Int. J.Appl. Earth Obs. Geoinformation 5, 159-163.
±raser D.C. (1969). Contouring of VL±-EM data. Geo
-
physics 34, 958-967.
Ganerød G. V., Rønning J.S., Dalsegg E., Elvebakk H.,
Holmøy K., Nilsen B. y BraathenA. (2006). Compari
-
son of geophysical methods for sub-surface mapping of
faults and fracture zones in a section of the Viggja road
tunnel, Norway. Bull. Eng. Geol. Environ. 65, 231-243.
Gill P.H.S., Murray W., Saunders M. y Wright M. (1986).
User’s guide for LSSOL: A package for constrained
linear least-square and quadratic programming. Stan
-
ford University. Technical Report SOL-886-1. EUA.
Gomez-Trevino E. (1987). Nonlinear integral equations
for electromagnetic inverse problems. Geophysics 52,
1297-1302.
Hutchinson P.J. y Barta L.S. (2002). VL± surveying to
delineate longwall mine-induced fractures. Leading
Edge 21, 491-493.
Jeng Y., Lin M.J. y Chen C.S. (2004).Avery low frequen
-
cy-electromagnetic study of the geo-environmental
hazardous areas in Taiwan. Environ. Geol. 46, 784-795.
Karlik G. y Kaya M. A. (2001). Investigation of ground
-
water contamination using electric and electromagnetic
methods at an open waste-disposal site: A case study
from Isparta, Turkey. Environ. Geol. 40, 725-731.
Karous M. y Hjelt S.E. (1983). Linear Fltering of VL±
dip-angle measurements. Geophys. Prospect. 31,
782-794.
Kaya M.A., Özürlan G. y Şengül E. (2007). Delineation of
soil and groundwater contamination using geophysical
methods at a waste disposal site in Çanakkale, Turkey.
Environ. Monit. Assess. 135, 441-446.
Liu H., Liu J., Yu C., Ye J. y Zeng Q. (2006). Integrated
geological and geophysical exploration for concealed
ores beneath cover in the Chaihulanzi goldFeld, north
-
ern China. Geophys. Prospect. 54, 605-621.
EM METHODS APPLICATION FOR CHARACTERIZING A CONTAMINATED SITE
329
McNeill J.D. (1980). Electromagnetic terrain conductivity
measurements at low induction numbers. Technical
Note TN-6, Geonics Ltd. Mississauga, Canada.
Mondelli G., Giacheti H., Boscov M., Elis V. y Hamada
J. (2007). Geoenvironmental site investigation using
different techniques in a municipal solid waste disposal
site in Brazil. Environ. Geol. 52, 871-887.
Monteiro-Santos F.A., MateusA., Figueiras J. y Gonçalves
M.A. (2006). Mapping groundwater contamination
around a land±ll facility using the VLF-EM method -A
case study. J. Appl. Geophys. 60, 115-125.
Nabighian M.N. (1982). A review of time-domain
electromagnetic exploration. In: Proceedings of the
International Symposium of Applied Geophysics in
Tropical Regions (J. Seixas Louren
ç
o, L. Rijo, Eds).
Conselho Nacional de Desenvolvimento Cientí±co e
Tecnológico.
Belem, Brazil, September l-8.
Ogilvy R.D. y Lee A.C. (1991). Interpretation of vlf-em
in-phase data using current density pseudosections.
Geophys. Prospect. 39, 567-580.
Oskooi B. y Pedersen L.B. (2005). Comparison between
VLF and RMT methods. A combined tool for mapping
conductivity changes in the sedimentary cover. J.Appl.
Geophys. 57, 227-241.
Pérez-Flores M.A. (1995). Inversión rápida en 2-D de
datos de resistividad, magnetotelúricos y electromag
-
néticos de fuente controlada a bajos números de induc
-
ción. Ph. D. Thesis. Centro de Investigación Cientí±ca
y de Educación Superior de Ensenada. Baja California,
México. 352 pp.
Pérez-FloresM.A.,Méndez-DelgadoS.yGómez-TreviñoE.
(2001). Imaging low-frequency and dc electromagnetic
±elds using a simple linear approximation. Geophysics
66, 1067-1081.
Pirttijärvi M. (2004). KHF±lt program. A geophysical
software for Karous-Hjelt and Fraser filtering on
geophysical VLF (very-low-frequency) data. Geopys
-
ics Division, Department of Geosciences, University
of Oulu, Finland.
Samsudin A., E. B., Zurairi W., Hamzah U. (2006).
Mapping of contamination plumes at municipal
solid waste disponsal sites using geoelectric imaging
technique: case studies in Malaysia. J. Spat. Hydrol.
6, 13-22.
SEMARNAT (2003). Norma O±cial Mexicana NOM-
083-SEMARNAT-2003. Especi±caciones de protec
-
ción ambiental para la selección del sitio, diseño,
construcción y operación, monitoreo, clausura y
obras complementarias de un sitio de disposición ±nal
de residuos sólidos urbanos y
de manejo especial.
Secretaría de MedioAmbiente, Recursos Naturales y
Pesca. Diario O±cial de la Federación. 20 de octubre
de 2004.
Sharma S.P. y Baranwal V.C. (2005). Delineation of
groundwater-bearing fracture zones in a hard rock
area integrating very low frequency electromagnetic
and resistivity data. J. Appl. Geophys. 57, 155-166.
Smith B.D. y Ward S.H. (1974). On the computation
of polarization ellipse parameters. Geophysics 39,
867-869.
Soupios P., Papadopoulos N., Papadopoulos I., Kouli M.,
Vallianatos F., Sarris A. y Manios T. (2007). Applica
-
tion of integrated methods in mapping waste disposal
areas. Environ. Geol. 53, 661-675.
Sundararajan N., Nandakumar G., Narsimha M., Ramam
K. y Srinivas Y. (2007). VES and VLF - An applica
-
tion to groundwater exploration, Khammam, India.
Leading Edge 26, 708-716.
Zlotnicki J., Vargemezis G., Mille A., Bruère F. y Ham
-
mouya G. (2006). State of the hydrothermal activity
of Soufrière of Guadeloupe volcano inferred by VLF
surveys. J. Appl. Geophys. 58, 265-279.
logo_pie_uaemex.mx