active and passive seismic as an indicator of large ...dschmitt/papers/ardeshir_etal_active... ·...
TRANSCRIPT
ORIGINAL PAPER
Active and Passive Seismic as an Indicator of LargeEquipment Interactions with the Oil Sand
A. D. Sharifabadi • T. G. Joseph • D. R. Schmitt
Received: 4 January 2010 / Accepted: 16 May 2010 / Published online: 30 May 2010
� Springer Science+Business Media B.V. 2010
Abstract The strain softening of oil sand in the
underfoot of ultra class mobile mining equipment, due
to the loading action of large mobile mining equipment
such as trucks and shovels, yields a highly unstable
condition for the operation of this ultra-class equipment.
Soft ground conditions in oil sand, due to the low
stiffness of the material a condition especially present in
the summer, can cause high rack, pitch, and roll in
trucks, leading to fatigue failure in structural compo-
nents. For shovels, poor ground stability can cause twists
in car bodies and undercarriages, resulting in major
damages. Track and shovel frame failures due to this
instability result in high maintenance costs. The authors
carried out a geophysical study of the oil sand in order to
evaluate the ground conditions under large mobile
mining equipment. A geophysical investigation per-
formed in summer 2001 encountered 6–8 m of thick soft
material, commensurate with very low velocities,
caused by the loosening of the surface material by
heavy mining machinery and excavation; and a transi-
tion zone of up to 25–26 m depth approaching the in situ
oil sand below. The depth of the oil sand zones can be
calculated by using the refraction analysis technique.
Spectral Analysis of Surface Waves was used to
estimate the ground stiffness. A new technique is
proposed to evaluate the changing ground stiffness
during the use of ultra-class mobile mining equipment.
Keywords Oil sand � Surface wave �Large mobile mining equipment � Ground stiffness
1 Introduction
Trucks and shovels dynamically load the ground during
the normal course of operations. The degree of impact
on the ground depends very much on its initial condition.
Soft ground, such as clay or oil sand, exhibits visco-
elastic–plastic behaviour, which on immediate unload-
ing, results in considerable permanent deformation. The
deformation is somewhat alleviated after a lengthy
period of relaxation. Joseph et al. (2003) reported that
after only a few operation cycles, trucks and shovels
operating on soft ground become less stable.
The oil sand deposits of northern Alberta are mined
using ultra-class scale ([325 tonne capacity trucks
and [46 m3 capacity shovels) bulk handling equip-
ment in an open-pit setting. Oil sand has specific
and unusual properties that allow it to behave like
A. D. Sharifabadi (&)
Geotechnic, AMEC Earth and Environmental, Calgary,
AB, Canada
e-mail: [email protected]
T. G. Joseph
School of Mining Engineering, University of Alberta,
Edmonton, AB, Canada
e-mail: [email protected]
D. R. Schmitt
Department of Physics, University of Alberta, Edmonton,
AB, Canada
e-mail: [email protected]
123
Geotech Geol Eng (2010) 28:727–743
DOI 10.1007/s10706-010-9335-3
sandstone in winter and weak clay in summer as the
temperature varies from -30 to ?30�C, respectively.
Standard refraction analysis techniques were used
to demarcate the oil sand zones with different
stiffness values (see Fig. 1). Lower velocities near
the surface indicated the higher degree of disturbance
caused by operating machineries; creating a loose or
mobile zone. Higher velocities at deeper layers were
commensurate with a more compact and undisturbed
nature of at-depth material, described as ‘‘transition’’
and deeper still ‘‘in situ’’ zones. It has been reported
that after only a few cycles, trucks and shovels
operating on progressively softening ground become
less stable (Joseph et al. 2003; Welz and Schmitt
2002).
The objective of this study was to investigate oil
sand behaviour in terms of ground stiffness or shear
modulus due to the operational loading by mining
equipment. Another important objective was to
develop a better understanding of the in situ proper-
ties of oil sands. Geophysical methods incorporating
passive seismic data analysis were used to understand
the interaction between the ground and heavy mining
equipment. The mining equipment was passively
recorded as the ground motion source and correlated
with ground reactions to dynamic loading from large
mobile mining equipment.
2 Site Characterization, Equipment Selection
and Source
Alberta’s oil sand deposits are the biggest oil sand
reserve in the world and represent a major source of
oil. Canada’s crude bitumen exists entirely within the
province of Alberta and is found in sand and
carbonate sedimentary formations in three regions:
the Athabasca, Cold Lake, and Peace River oil sand
areas.
Syncrude Canada Limited is the world’s largest
producer of synthetic crude oil from oil sands and the
largest single source producer in Canada. Syncrude’s
operations are located just outside Fort McMurray in
the Athabasca Oil Sands.
Summer and winter refraction data measuring the
seismic properties of in-situoil sand was collected
from North Mine; near Mildred Lake on lease 17, and
the Aurora mine about 35 km to the north. The three
bitumen grades present at the Syncrude mines, 6, 8,
and 13.5%; were considered as part of this experi-
mental seismic refraction survey.
The summer experiment gathered seismic surfaces
wave using 14 Hz geophones at 1 m spacing. This
geophones spacing was chosen to avoid spatial aliasing
of wave lengths as small as 2 m (Nyquist wavenumber
k = 2p/kmin = p). This small wavelength contains
V1 = 380 m/s h1 = 4 m
V2 = 790 m/s h2 = 8 m
V3 = 1500 m/s
Mobile or soft zone
Approaching in situ zone
Transition zone
Fig. 1 Ground condition and corresponding velocity profile (Sharif-Abadi 2006)
728 Geotech Geol Eng (2010) 28:727–743
123
information about the shear wave velocity of the first 1
to 2/3 m. The sampling was collected by a 240-channel
geometrics geode seismic recorder at a sampling rate of
0.250 ms, providing a Nyquest frequency of 2,000 Hz
for refraction data. The passive seismic configuration
used was 4 ms sampling was conducted for a 65 s
record in order to obtain a complete picture of a truck
passing through a linear array of seventy-two 14 Hz
geophone sensors. The refraction source was an
accelerated weight drop for passive seismic and used
for both summer and winter data collection.
Passive sources are any sources that cause ground
motion over which there is little or no control, such as
road traffic, wind causing movement of objects, or
people walking. In this study the motion of mining
equipment such as trucks or shovels was the passive
source for ground motion.
3 Summer Refraction Data Acquisition
and Analysis
Seventy-two vertical 14 Hz geophones were placed
along a single line at 1 m spacing and, at an ambient
temperature of 28�C in an active mining area. The
source used was an accelerated weight drop and the
sampling rate was 0.250 ms. Data was recorded using
a Geometrics system and subsequently processed
using the Vista-Windows Seismic Processing soft-
ware. The passive record sensed high levels of noise
due to the heavy equipment operation. This noise was
filtered to remove the background influence of
motors, pumps, and tire treads from the gross loading
action of the equipment.
The data acquired showed high attenuation com-
mensurate with the expected greater fluidic nature of
the oil sand material at warmer temperatures. An
analysis of the refraction data showed that the
attenuation generally increased with the bitumen
content (increasing % grade by volume). Figures 2, 3,
and 4 illustrate examples of the seismic offset time
shot for operations on 6, 8, and 13.5% oil sand and
the corresponding refraction analysis. Figure 5 shows
the AGC (automatic gain control) plot of the unscaled
plot of the 13.5% oil sand with high level attenuation.
Table 1 shows the overall results of the refraction
data analysis by percentage bitumen content.
Forward and reverse refraction methods were used
to estimate the zones thickness in both directions. The
method employs refraction of wave arrivals from
shots offset in opposing directions targeting a given
receiver, having left the refracting surface from
effectively the same lateral location. The refractor
velocity and depth below the receiver location can be
determined for any receiver that records forward and
reverse refracted arrivals.
During refraction field tests, refraction data was
shot to a line of geophones from both ends and the
first arrival p-waves from both directions was
identified. Figure 6 illustrates how the forward and
reverse method allowing identification of arrival
times. It is obvious that the transition zone is
evident from the first 10 m of geophone response,
where the material properties vary considerably. The
truck travel route, Fig. 7, likely caused the variance
for the last 7 m of geophones (65 through 72),
where the actual route diverged from the geophone
line.
4 Determination of Ground Stiffness by Using
a Geophysical Technique with Summer Data
The velocity of oil sand is needed in the inversion
method to estimate the ground stiffness profile. An
inversion method was used on the summer data to
estimate the ground stiffness profile. This approach
may be used for ground and equipment performance
modeling during different seasons of the year. The
relative ground modulus changes during mining
activity can potentially be determined. The prelimin-
ary study of passive data revealed that the ground
modulus drops after a few truck runs (as explained in
the passive seismic data analysis).
Characterizing the near subsurface using geophys-
ical methods has been of great interest in recent years
since these methods are cheaper and faster than
conventional drilling and borehole logging. Being
able to determine a modulus-depth profile without the
aid of boreholes via surface wave geophysics is seen
as being highly advantageous. The most powerful
tools for evaluating the subsurface are spectral
analysis of surface waves (SASW) via non-intrusive
Rayleigh waves (Matthews et al. 1996, Haegeman
and Van-Impe 1999).
Knowing the properties of soil layers and having a
profile are the key factors in analyzing an overlying
dynamic loading source. SASW is currently used
Geotech Geol Eng (2010) 28:727–743 729
123
primarily for the evaluation of subsurface wave
velocity profiles.
The results obtained from this method represent of
the average properties of a relatively large mass of
soil. This method can be very cost-effective for
investigations (Haegeman and Van-Impe 1999 and
Ganji et al. 1998). An alternative technique for
determining ground stiffness is continuous surface
wave (CSW) analysis Continuous Surface Wave
analysis relies on the propagation properties of
vertically polarized seismic surface waves, where
the penetration depth by a surface wave is dependent
on the wavelength and frequency (Moxhay et al.
2001).
Shear wave velocity is a factor in identifying the
shear strength of a given formation. Rayleigh wave
dispersion has been used as a method for evaluating
the shear modulus of near-surface materials. Using
Rayleigh waves to obtain the shear wave velocity has
two steps; (a) finding the dispersion relationship for
the Rayleigh wave, and (b) applying the inverse
procedure to convert the dispersion curve to the shear
wave velocity versus the depth (Beaty and Schmitt
2003).
4.1 Determine Stiffness with Rayleigh Waves
One technique uses the surface dispersion curve to
determine the near-surface stiffness. A tool that has
increasingly been used to evaluate the shear modulus
of near-surface materials is Rayleigh surface-wave
dispersion (Stokoe and Nazarian 1985; Rix et al.
Fig. 2 Shot gather (time-
offset) in the 6% bitumen
oil sand and p-wave
velocities
730 Geotech Geol Eng (2010) 28:727–743
123
2001; Louie 2001). Usually the dispersion curves for
Rayleigh waves are used to examine the variability of
near-surface properties up to 15 m deep Rayleigh
waves travel along the earth-air interface and usually
contain more energy than body waves and are able to
provide substantial information on shear wave veloc-
ity, identifying structures of the near-surface (Beaty
2000; Ewing et al. 1957).
The dispersion nature of Rayleigh waves and the
velocity depend on its frequency. The Rayleigh wave
phase velocities depend primarily on the shear-wave
velocity structure of the near-surface materials.
Lower frequencies, or longer wavelengths, have deep
penetration, which is appropriate for determining the
depth stiffness (these phenomena contain information
about the deep layers). On the other hand, high
frequencies or lower wavelength Rayleigh waves are
appropriate for determining the upper layer stiffness.
A rule of thumb is that the depth of the relation of
Rayleigh waves is one-half of the wavelength.
Dispersion curves show the velocity of the wave at
each wavelength or frequency. These can be inverted
to obtain the shear wave velocity profiles of an area.
One method to get dispersion curves from Rayleigh
waves is to carry out Tau–p (s–p) transformation on
the data, followed by a one-dimensional Fourier
transformation along the s direction. s–p transforms
essentially carry the sum of the amplitudes along a
line in the offset-time domain (seismic data) with
intercept time s and slope p called the ‘‘slowness’’
(slowness is the inverse of velocity). This sum will
map onto a point (s–p). This is a simple, well-known
linear wave field transformation that takes an input
data set in the time-offset domain and transforms it
Fig. 3 Shot gather (time-
offset) in the 8% bitumen
oil sand and p-wave
velocities
Geotech Geol Eng (2010) 28:727–743 731
123
into a new data set in the intercept time—slowness
domain. In this case, the input data set, p, is the
observed seismogram wavefield. In S, the trans-
formed wavefield, p is the horizontal slowness and sis the time intercept.
Sðs;PÞ ¼Zþ1
�1
Pðsþ px; xÞdx ð1Þ
A simple way to look at the transform is to think of
each point in the s–p plane being the sum of all the
points in the t–x plane lying along a straight line with
a slope of p and a time-axis intercept of s. The
seismogram is decomposed into plane wave elements
(McMechan et al. 1982; McMechan and Yedlin 1981;
Louie 2001).
After the dispersion curve has been attained, it
must be inverted to a shear wave velocity profile.
Figure 8 shows the sequence of obtaining the
dispersion curve for the seismic data (Beaty 2000,
Matthews et al. 2000).
For inversion, the p-wave velocity and the thick-
ness of the layer are needed.
Obtaining the phase velocity dispersion curve
from Rayleigh waves requires two steps:
1. Isolate the Rayleigh wave from the other arrivals
on the seismograph (this step can be performed
by windowing the surface data in either the
offset-time or frequency-wave number domain
(f–k)).
2. Extract the dispersion curve. Different methods
can be used:
Fig. 4 Shot gather (time-
offset) in the 13.5%
bitumen oil sand and
p-wave velocities (high
level attenuation)
732 Geotech Geol Eng (2010) 28:727–743
123
Fig. 5 AGC scaled plot for
the 13.5% bitumen oil sand
Table 1 P-wave velocity
and thickness of summer
operating condition oil sand
Oil sand depth descriptor Oil sand grade (%) P-wave (m/s) Thickness (m)
First zone (loose) 6 1,200 2.25
Second zone (in situ) 1,500 Continuous
First zone (loose) 8 410 9
Second zone (transition) 1,100 28
Third zone (in situ) 1,950 Continuous
First zone (loose) 13.5 380 6.7
Second zone (transition) 790 26
Third zone (in situ) 1,500 Continuous
0
0.02
0.04
0.06
0.08
0.1
0.12
0 10 20 30 40 50 60 70 80
Offset - Geophones (m)
Tim
e (S
ec)
y = -0.002x + 0.1507R2 = 0.9974
y = -0.0011x + 0.0977R2 = 0.9993
y = 0.0021xR2 = 0.995
y = 0.0011x + 0.0136R2 = 0.9996
Reverse Forward
Fig. 6 Forward and reverse
method for summer
refraction data (p-wave)
Geotech Geol Eng (2010) 28:727–743 733
123
• Exploit the phase information of the Fourier
transformation of the surface wave (used in
earthquake seismology).
• Use dispersion information to locate the f–k
spectrum or frequency-slowness representing
the data (used for an array of multiple
geophones):
• Perform a s–p transformation on the data
followed by one-dimensional Fourier
transformation along s direction.
• Pick the peaks associated with the surface
wave energy in the f–k domain.
4.2 Dispersion Curve for Summer Data
Determination of the s–p transform of the surface
wave dataset is the first step in the method described
in McMechan and Yedlin (1981) for obtaining
dispersion curves from a multichannel seismogram.
Both s–p and f–p transformations were carried out on
the summer data (see Fig. 9).
In the t-offset domain followed by the s–p
transform, the f–p display and a reconstruction of
the data and the data reconstructed from the s–p have
preserved the information contained in the data well.
The difference between the original data set and
reconstruction will show if the original data set is
aliased or if the initial value for s and p was not
correct. Darker areas on the f–p map indicated a
concentration of energy related directly to the
dispersion curves.
4.3 Inversion
One technique employed by Beaty (2000) uses the
surface dispersion curve to determine the near-
surface stiffness. To obtain a reasonable S-wave
profile, good estimates of the values for P-wave and
density profiles should be sufficient. The forward
modeling technique was used to obtained theoretical
dispersion curves in an elastic, layered medium and
followed by the outline inversion technique to obtain
the velocity profile from the measured dispersion
curve. Beaty (2000) used a matrix propagator method
to obtain theoretical dispersion curves for Rayleigh
waves in a vertically varying medium consisting of a
set of n homogeneous layers overlying a homoge-
neous half-space. To invert the dispersion curve data
for the S-wave velocity profile, the forward
model was incorporated into a simulated annealing
algorithm.
After the dispersion curve has been attained, it
must be inverted to a shear wave velocity profile.
Even though both summer and winter data was
aliased, the summer data showed less alias from the
s–p graph. Therefore, an inversion was carried out on
the summer refraction data using a simulated anneal-
ing algorithm (Beaty 2000).
The p-wave and S-wave velocities are related to
the elastic constants through the relations
Vp ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffikþ 2l
q
s
Fig. 7 Trucks route
Detection of motion on the ground surface
Dispersion curve: Phase velocity of Rayleigh wave versus frequency
Variation of shear wave velocities with depth
Small strain stiffness profile (G0 versus depth)
Processing
Inversion
G0=ρ.Vs2
Acquisition
Fig. 8 Sequence of obtaining shear wave velocities
734 Geotech Geol Eng (2010) 28:727–743
123
Vs ¼ffiffiffilq
r
where q, density; l, shear modulus; k, Lame’s first
parameter. The input for the model is P-wave velocity
profile, density, layer thickness.
Table 2 illustrates the results of the inversion and
calculation of elastic modules based on the shear
modulus. The elastic modulus results are very similar
to the previous triaxial test results presented by
Joseph et al. (2003).
5 Summer Passive Data
Ninety-six geophones, 72 in one line and 24 in a
perpendicular decay line were set-up for passive and
refraction data acquisition (see Fig. 10). The distance
Slowness (s/m)
Freq
uenc
y (H
z)
f - p Transform
0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
0
20
40
60
80
100
120
140
t - offset
20 40 60
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
τ - p
0.01 0.02
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
t - offset (reconstructed)
20 40 60
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Fig. 9 Summer data
f–p and s–p graphs
Geotech Geol Eng (2010) 28:727–743 735
123
between each geophone was set at 1 m, with trucks
passing within 1–2 m of the line. Three site locations
of varying bitumen content and corresponding initial
softness were evaluated.
The sum of the amplitudes in line A of 72
geophones for trucks passing was calculated for each
location. Figure 11 shows that the sum of the
amplitudes increased as the truck velocity increased.
The same phenomenon as in the winter data was
seen; the sum of the amplitudes increased at the
beginning and then suddenly decreased after a
number of cycles (see Fig. 12).
6 Winter Refraction Data
In the case of winter data acquisition, where the
ambient temperature was -38�C, the top layer of oil
sand was frozen to a depth of several metres, yielding
higher velocities compared to those beneath. No
reliable P-wave refraction was discernable from this
data, and the spacing of 3 parallel lines of 24
geophones at 3-m intervals (Fig. 14) was not enough
to attain refraction from the stiff ground. Spatial
aliasing was seen in all data sets and was probably
caused by insufficient distance between geophones
and the thin frozen layer of oil sand at the surface.
Noise was a problem due to the presence of active
heavy machinery. The high noise levels were present
even in the geophones furthest from the shot point.
The surface waves were very strong in this frequency
Fig. 10 Summer test set-up
Table 2 Presents the inversion results
Changing oil
sand property
Oil sand
grade (%)
P-wave velocity
(m/s)
Thickness
(m)
S-wave velocity
(m/s)
Shear modulus
(MPa)
Elastic modulus
(MPa)
First zone (mobile/soft) 13.5 380 6.7 137 37 94
Second zone (transition) 790 26 236 120 301
Third zone (in situ) 1,500 Infinity 248 136 339
y = 1E+08x - 2E+08
R2 = 0.9072
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
7 7.5 8 8.5 9 9.5
Velocity (m/s)
Am
plit
ud
e *1
0E9
(Vo
lt)
Fig. 11 Sum of the amplitude versus truck velocity for an
unloaded truck
736 Geotech Geol Eng (2010) 28:727–743
123
range, as shown in Fig. 13. However, the direct
waves gave a range of velocities for the frozen layer
of 1,500–1,700 m/s. Figure 13 illustrates the direct
wave of a winter data set.
7 Passive Data Acquisition
Winter data was acquired using a 3-line array of
geophones and the moving trucks as a source of
vibration. The purpose of the test was to (a) evaluate
the interaction between the ground and mobile
mining equipment under firm ground conditions
and, (b) to show that seismic equipment and survey
techniques could monitor interactions. Figure 14
shows the winter test set-up (an advance over the
original summer test set-up used in Fig. 7) using 3
parallel lines, each having 24 vertical 14 Hz geo-
phones placed with an instrument spacing of 3 m,
giving a test array of 72 m length, similar to the
0
2
4
6
8
10
12
14
16
18
0 10 20 30 40 50 60 70 80
Geophone #
Am
plit
ud
e *1
0E6
(Vo
lt)
Last Truck Passing
First Truck Passing
Fig. 12 Changing sum of
amplitudes in each
geophone
Fig. 13 Winter surface and
refraction data
Geotech Geol Eng (2010) 28:727–743 737
123
original summer configuration. The geophone acqui-
sition system was set at a 250 Hz acquisition rate.
The only source of vibration was set as a controlled
single truck passing through the array, so that a few
seconds before the truck reached the array, the
geophones would be manually activated to start
recording data. As expected, the middle line B had
greater amplitude compared to that of lines A and C
(see Figs. 15, 16) for a single run. Figure 16
illustrates the maximum amplitude of each geophone
for different runs. The results also showed that the
rear of the truck put more energy into the ground due
to the truck’s approximately 2/3:1/3 rear to front axle
load distribution.
7.1 Winter Passive Data Analyses
Figure 17 shows plot of winter passive data; the truck
velocity is determined by the slope of the inclined
line. Figure 18 illustrates the velocity of the truck for
the entire test (trucks passing 50 times through the
line). As was expected, the passive data shows that a
truck traveling at a slow velocity has a higher band of
amplitude than that of a faster truck yielding a
reduced band (Ws [ Wf) (see Fig. 19).
The geophones’ amplitude from units in the same
column was calculated, (e.g., 1, 25 and 49) giving an
accurate correlation between the velocity of the truck,
and the sum of the amplitudes. Figure 20 shows the
Fig. 14 Test set-up for
passive seismic winter data
acquisition
Fig. 15 Passive data shot acquired for a single run of truck
738 Geotech Geol Eng (2010) 28:727–743
123
correlation between truck velocity and the sum of the
amplitudes.
From Fig. 20 it can be concluded that trucks with
higher velocity transfer more amplitude or energy
into the ground. The time between two consecutive
runs is termed the relaxation time. The relaxation
time for 50 runs performed during the test was not
constant. Since oil sand in summer behaves like a
viscous elastic material, one important parameter
related to the degree of permanent ground deforma-
tion relative to the next truck passing over the same
ground is the relaxation time. The tests yields four
levels of time gaps between runs (see Fig. 21),
although the expected effect on ground performance
was not thought to be significant for the stiffer ground
in winter.
The extent of ground deformation can be estimated
by normalizing the sum of the amplitude across the
0
2
4
6
8
10
12
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69
Geophone (#)
Am
plit
ud
e (1
04 )
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
Line B (Geophone # 25 to 48)
Line A (Geophone #1 to 24)
Line C (Geophone # 49 to 72)
Fig. 16 Max amplitude for each geophone for 50 runs of truck
Fig. 17 Truck velocity
from passive data
Geotech Geol Eng (2010) 28:727–743 739
123
width (from the summing up procedure) by the sum
of total amplitude for the same run. The more
amplitude or energy transferred into the ground, the
lower the ground stiffness. Less energy corresponds
to harder ground conditions. A study compiled from
each of the 4 relaxation groups shown in Fig. 21,
proved that after a number of runs for each group, the
total energy first increased, corresponding to the
progressively softening ground. However, with addi-
tional runs, the energy suddenly leveled off to a
constant value. This lead to a tentative conclusion
that after several runs the ground reached a stable
condition and higher stiffness in winter condition (see
Fig. 22). Unlike results of energy increases being
parallel for all groups, this condition was certainly
more predominant in later runs of the test (group 4)
than in group 2. Figure 23 is normalization of Fig. 23
with the maximum amplitude in each column of
geophones (percentage of the maximum amplitude
0
2
4
6
8
10
12
1 6 11 16 21 26 31 36 41 46
# Run
Vel
oci
ty(m
/s)
Fig. 18 Simple truck velocity for winter passive data
Fig. 19 Slow truck
velocity versus fast truck
velocity
740 Geotech Geol Eng (2010) 28:727–743
123
value) and different lines representing progressive
relative ground stiffness changes by run and in turn
grossly representing parameters that may affect the
degree of softening: truck velocity, ground profile,
and relaxation time.
To evaluate onboard the truck’s onboard informa-
tion system’s (VIMS) ability to predict poor ground
conditions, the seismic-data-acquisition system was
set at 250 Hz (1/250 Hz = 4 Ms). However, this
setting was mismatched with the OEM on-board data-
acquisition system (VIMS) at 1 Hz, so making a
correlation between the two sets was difficult.
y = 2E+07x + 1E+08R2 = 0.882
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5
Velocity (m/s)
Am
plit
ud
*10
E8
(Vo
lt)
Sum 1 Sum 2 Sum 23 Sum 24Sum 3
Total Sum
1 24
25 49
50 72
Fig. 20 Sums of
amplitudes for each array
position
0:00:00
0:14:24
0:28:48
0:43:12
0:57:36
1:12:00
1:26:24
1:40:48
0 10 20 30 40 50
# Run
Tim
e (M
in)
Group 2 Group 3 Group 4Group 1
h:min:s
Fig. 21 Relaxation time between runs
Group Two (4-20)
0
5
10
15
20
25
21161161
Geophones
Am
plit
ud
e *1
0E6
(Vo
lt)
Group 2Group 3
Fig. 22 Sum of geophones
columns amplitude versus
relaxation group
Geotech Geol Eng (2010) 28:727–743 741
123
Studying the trends of the rack, pitch, and roll
(Joseph and Welz 2003) from the predominant truck
motion did however provided some correlation. The
difference in energy was determined via the sum of
the truck’s front-to-back energy (amplitudes) using
the seismic data. Figure 24 shows that the rack
yielded a reasonable shape correlation, indicating an
influence on the ground.
Figure 25 provides a closer look at the passive
data. Here, the front-to-back motion is evident from
the curves caused by front and back of truck tire due
to the undulated ground and poor ground condition.
However, devising an algorithm to separate the front
and back motion or to distinguish the path of the
curve apex was difficult for several reasons. The time
of the front-to-back motion depends on the velocity,
which may be extracted from the passive data, but
only as an average value. Also, the start time is
manually activated, so that distinguishing the exact
time of passing adjacent to any given geophone was
made very difficult in post processing.
8 Conclusion
The collected data reveals that the near surface
material loosen by heavy mining machinery and
excavation and that it has lower velocity. The deeper
material is more compact and undisturbed due to a
change in Bitumen content of the oil sand.
In summer, the p wave velocity of loose oil sand
material is between 300 and 800 m/s where the p
wave velocity of in situ oil sand is between 1,500 and
2,000 m/s. The shear modulus of the soften oil sand
(13.5%) due to the mobile mining equipment activ-
ities varies from 30 to 140 MPa. The elastic modulus
of the 13.5% oil sand was shown to be predictable
from passive seismic analysis at approximately
90 MPa for lose material and 400 MPa for in situ
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
12611161
Geophones
Per
cen
t (%
)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49
Sum 1 Sum 2 Sum 23 Sum 24Sum 3
1 24
25 49
50 72
Fig. 23 Comparing the amplitude normalization for different runs
-8
-6
-4
-2
0
2
4
6
1.0 3.0 5.0 7.0 9.0 11.0
Time (Sec)
Rac
k ca
lcu
late
d f
rom
sei
smic
am
plit
ud
e (*
10E
6)
-20
-15
-10
-5
0
5
10
15
20
Rac
k ca
lcu
late
d f
rom
VIM
S d
ata
(kP
a)
Seismic Truck
Fig. 24 Comparing the seismic data and VIMS data
742 Geotech Geol Eng (2010) 28:727–743
123
oil sand. In winter, the p wave velocity of the frozen
oil sand layer is close to in situ oil sand at
approximately 2,000 m/s.
There is a correlation between the loading action
of a truck and the information collected by the trucks
onboard information system measurements of rack,
pitch, roll, and bounce. The ground softening due to
mobile mining equipment can be correlated to
seismic reaction.
Acknowledgments This study was financially supported by
James Progithin International Ltd and the University of
Alberta. The authors are grateful to the significant help of
Dr. Kristen Beaty and Marek Welz to this research.
References
Beaty KS (2000, Fall) Determination of near surface variability
using Raleigh waves. MSc Thesis, Department of Physics,
University of Alberta, Edmonton, 208 pp
Beaty KS, Schmitt DR (2003) Repeatability of multimode
Raleigh-wave dispersion studies. Geophysics 68(3):782–
790
Ewing WM, Jardetzky WS, Press F (1957) Elastic waves in
layered media. McGraw-Hill, New York
Ganji V, Gucunski N, Nazarian S (1998) Automated inversion
procedure for spectral analysis of surface waves. J Geo-
tech Geoenviron Eng 124(8):757–770
Haegeman W, Van-Impe WF (1999) Characterization of dis-
posal sites from surface wave measurements. J Environ
Eng Geophys 4(1):27–33
Joseph TG, Welz M (2003) Mechanical action and geophysical
reaction: equipment—oil sand interactions. In: Proceed-
ings of CIM coal and soft rock workshop, Saskatoon,
October 3–5, 10 pp
Joseph TG, Sharif-Abadi AD, Shi N (2003) A broken material
approach to modelling oil sand under dynamic load. In:
CD Proceedings of computer applications in the minerals
industries (CAMI), Calgary, September 8–10, 12 pp
Louie NJ (2001) Faster, better: shear-wave velocity to 100
metres depth from refraction micrometre arrays. Bull
Seismol Soc Am 91(2):347–364
Matthews MC, Hope VS, Clayton CRI (1996) The use of
surface waves in the determination of ground stiffness
profiles. In: Proceeding of the institution of civil engi-
neering—geotechnical engineering, vol 119, 2, pp 84–95
Matthews MC, Clayton CRI, Own Y (2000) The use of field
geophysical techniques to determine geotechnical stiff-
ness parameters. In: Proceeding of the institution of civil
engineering—geotechnical engineering, vol 143, 1, pp
31–42
McMechan GA, Yedlin MJ (1981) Analysis of dispersive
waves by wave-field transformation. Geophysics 46:869–
874
McMechan GA, Clayton RW, Mooney WD (1982) Application
of wave field continuation to the inversion of refraction
data. J Geophys Res 87(B2):927–935
Moxhay AL, Tinsley RD, Sutton JA (2001) Monitoring of soil
stiffness during ground improvement using seismic sur-
face waves. Ground Eng Mag 34(1):34–37
Rix GJ, Lai CG, Foti S (2001) Simultaneous measurement of
surface wave dispersion and attenuation curves. Geotech
Test J 24:350–358
Sharif-Abadi AD (2006) Cyclic performance of soft ground.
PhD thesis, University of Alberta, 267 pp
Stokoe KH, Nazarian S (1985) Use of Rayleigh waves in liq-
uefaction studies. In Woods RD (ed) Measurement and
use of shear wave velocity for evaluating dynamic soil
properties: Am Soc Civ Eng, pp 1–17
Welz M, Schmitt DR (2002) Passive seismic as an indicator of
larger equipment interaction with oil sand. In: 104th CIM
AGM, Vancouver April 27–May 1
Fig. 25 Closer view of the
passive data
Geotech Geol Eng (2010) 28:727–743 743
123