Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and
Resistivity
1
Clinton M. Wood, PhD
Michelle L. Bernhardt, PhD
Tim Moody
Behdad Mofarraj
Non-Intrusive 3D Levee Imaging Workshop
St. Louis, MO; 11 November, 2016
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Presentation Outline
2
1) Motivation for Work2) Data available at Mel-Price Wood River Levee3) Data Collection and results at the Mel-Price Wood
River Levee1) Ohmmapper resistivity 2) Multi-channel Analysis of Surface Waves (MASW)
4) Pitfalls associated with inversion problems5) Final Thoughts
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Levees in the US
3
ASCE Report Card (2013)
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Levees in the US
4
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
The Problem$
5
1) Limited funding to assess the estimated 100,000 miles of levees• Currently only about 15% of the nation’s levees are in the
National Levee Database - Over 22% of those levees are rated as unacceptable
• Only about 37% are documented in FEMA’s Midterm Levee Inventory
2) Limited funding for necessary or cautionary repairs• ASCE estimates over $100 billion is needed to repair and
rehabilitate the US levee system• Only $415 million is allocated for the entire flood control
program annually
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
The Approach
6
Develop and refine a rapid and non-destructive assessment procedure which can cost effectively address both problems1) Geophysical field testing2) Statistical analysis of data to determine most effective
methods3) Probabilistic framework to assess performance
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Mel-Price Wood River Levee Section
7
~4 km Section of Levee
Mississippi River Lock and Dam #26
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Geotechnical Information Mel-Price Wood River Levee
8
CPT Sounding
Geotech Boring
Hand Probe
Legacy Boring
Centerline of Levee
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Mel-Price Wood River Levee Data Collection
9
A combination of geophysical methods were used with the goal of determining both soil type and stiffness of the levee material
A. Resistivity testing via a Geometrics Ohmmapper Capacitively Coupled Resistivity Instrument
B. Surface wave testing (MASW) via landstreamer, 4.5 Hz geophones, and sledgehammer source
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Mel-Price Wood River Levee Data CollectionResistivity
10
Testing Parameters1) Ohmmapper TR5 system with five receivers was used.2) Dipole length of 5 meters with a rope length of 2.5 meters3) Dipole length of 10 meters with rope lengths of 5, 20, and 40 meters
CPT Sounding
Geotech Boring
Ohmmapper
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Resistivity Testing on Mel-Price Wood River Levee
11
Receivers
Transmitter
GPS Unit
Laptop
Dipole Cable
Dipole Length (5-10 m)
Rope Length (2.5-40 m)
Receivers
Dipole Cables
Transmitter
Controller
Rope Length
Transmitter Dipole Length
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 12
Resistivity Processing
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Resistivity Processing for Mel-Price Wood River Levee
13
CPT Sounding
Geotech Boring
Ohmmapper
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 14
Ohmmapper
Resistivity ProcessingFour separate passes along Levee to obtain full Apparent Resistivity
Dipole length (m)
Rope length(m)
5m 2.5m
10m 5m
10m 40m
10m 20m
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 15
Ohmmapper
Resistivity Processing
Inversion completed using Res2D software
Bottom Clay layer
Sand Core
Clay cap
Much more processing to come
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Mel-Price Wood River Levee Data CollectionSurface Wave Testing (MASW)
16
Testing Parameters1) Geostuff landstreamer with 24, 4.5 Hz vertical geophones setout with a 2 m spacing between geophones (total array length of 46 m).2) Measurement spacing between 25-50 meters depending on line3) Sledgehammer source with source locations of 5, 10, 20 meters from the first geophone and 3-5 shots per location
CPT Sounding
Geotech Boring
MASW
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Surface wave Testing on Mel-Price Wood River Levee
17
Receivers (24 total)
Laptop and Geode seismograph
Array Length (46 m)
Source Locations
Vertical Geophone on Landstreamer
Strike Plate
Sledgehammer (3-5 Averages)
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 18
CPT Sounding
Geotech Boring
MASW
Surface Wave Testing (MASW)
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity
Frequency (Hz)
Ph
ase
Vel
oci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
5m offset
10m offset
20m offset
Frequency (Hz)
Ph
ase
Vel
oci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
5m offset
10m offset
20m offset
19
Dispersion Processing
Frequency Domain Beamformer Method
Combined with multiple source offsets
Surface Wave Testing (MASW)
MASW
Three Source Locations
Frequency (Hz)
Ph
ase
Vel
oci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
Frequency (Hz)
Phas
e V
eloci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
Frequency (Hz)
Ph
ase
Vel
oci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
Frequency (Hz)
Phas
e V
eloci
ty (
m/s
ec)
10 20 30 40 50 60 70 800
100
200
300
400
500
5m offset
10m offset
20m offset
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 20
Nearfield Effects
Frequency (Hz)
Ph
ase
Vel
oci
ty (
m/s
ec)
10 20 30 40 500
200
400
600
800
1000
5m Offset
Frequency (Hz)
Ph
ase
Velo
cit
y (
m/s
ec)
10 20 30 40 500
200
400
600
800
1000
10m Offset
Frequency (Hz)
Ph
ase
Velo
cit
y (
m/s
ec)
10 20 30 40 500
200
400
600
800
1000
20m Offset
Frequency (Hz)
Phase
Velo
cit
y (
m/s
ec)
10 20 30 40 500
200
400
600
800
1000
40m Offset
5 10 15 20 25 30 35 40 45 50
0
100
200
300
400
500
600
700
800
900
1000
Frequency (Hz)
Phase V
elo
city (
m/s
ec)
5m offset
10m offset
20m offset
40m offset
Passive HRFK
5% Error30% Error 15% Error
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 21
Surface Wave Testing (MASW)
Unfortunately the surface wave results are still under construction for the Mel-Price Wood River Levee
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 22
Potential Pitfalls and Limitations in Inversion Process
0.1 1 10 1000
50
100
150
200
250
300
350
400
Wavelength (m)
Ph
as
e V
elo
cit
y (
m/s
)
Mean with +/- 1
3-D Theoretical Fit
0 200 400 600
0
5
10
15
20
25
30
35
40
Shear wave Velocity (m/s)
De
pth
(m
)
0.1 1 10 1000
50
100
150
200
250
300
350
400
Wavelength (m)
Ph
as
e V
elo
cit
y (
m/s
)
Mean with +/- 1
3-D Theoretical Fit
0 200 400 600
0
5
10
15
20
25
30
35
40
Shear wave Velocity (m/s)
De
pth
(m
)
Dispersion Curve Shear wave velocity Profile
Apparent Resistivity True Resistivity
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 23
Inversion Process
(Foti 2000)
b) Calculate theoretical dispersion curve (DC) for system (forward problem)
c) Compare theoretical DC to experimental DC acquired in field (misfit function)
d) Revise layers (i.e., thickness, Vs, etc.) until satisfactory fit is achieved (backward problem)
a) Assume a system of linear elastic layers over a half-space (H, r, Vs & Vp)
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 24
Inversion Challenges
• Nonlinear– Relationship between the data space (Vr vs. freq or wavelength) and the
model space (Vs vs. depth) is nonlinear
• Ill-Posed
– Attempting to recover 4 model parameters (H, r, Vs & Vp) indirectly from two data parameters (Vr, freq)
• Mixed-Determined– The model solution for deeper layers is dependent on the model
solution for shallower layers
• Result…Non-unique Solution!– Many models can fit the experimental data “equally well”
– The choice of layering parameterization has a HUGE impact on the ability to recover the “true” layered model
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 25
How do many 2D and Pseudo 2D methods solve the inversion problem?
Distance along Line
Depth
The use of lots of unconstrained layering in the inversion models can lead to
1) Unrealistic layering that does not make sense geologically and geotechnically2) Smearing of layer properties at interfaces making it difficult to recover true
properties
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 26
No evidence of inversion/LVL in dispersion data
Is this geologically or geotechnicallyreasonable?
Example of recovering unrealistic geotechnical properties
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 27
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.456 - 0.28]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.414 - 0.232]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
= 2.0
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.333 - 0.162]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
= 2.0
= 3.0
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.171 - 0.102]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
= 2.0
= 3.0
= 3.5
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.135 - 0.084]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
= 2.0
= 3.0
= 3.5
= 3.5*
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.028 - 0.026]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Shear Wave Velocity (m/s)
Depth
(m
)
0 1000 2000 3000
0
50
100
150
200
250
300
Shear Wave Velocity (m/s)
0 1000 2000 3000
0
50
100
150
200
250
300
= 1.2
= 1.5
= 2.0
= 3.0
= 3.5
= 3.5*
= 5.0
Solution
Frequency (Hz)
Rayle
igh P
hase V
elo
city (
m/s
)
[0.248 - 0.248]
100
101
0
500
1000
1500
Frequency (Hz)
100
101
0
500
1000
1500
Experimental DC
Disp. Misfit Range for
Best 50 Models
Depth
Model25 Layers
Depth
Model12 Layers
Depth
Model5 Layers
From Teague and Cox 2016
Depth
Model3 Layers
Example of smearing layer properties
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 28
1. Pseudo 2D methods such as Resistivity and MASW can be powerful tools to rapidly evaluation geotechnical infrastructure. However, care must be taken from the data collection to the data process to insure valuable results are obtained and not just fancy color contour plots.
2. Pseudo 2D methods still have ways to go in the inversion process to be able quickly determine realistic layering and material parameters.
Final Thoughts
Pseudo 2D Imaging of the Mel-Price Wood River Levee via MASW and Resistivity 29
Acknowledgements• USDOT and MarTREC
– This material is based upon work supported by the U.S. Department of Transportation under Grant Award Number DTRT13-G-UTC50. The work was conducted through the Maritime Transportation Research and Education Center at the University of Arkansas. This work reflects the views of the authors, who are responsible for the facts and the accuracy of the information presented herein. This document is disseminated under the sponsorship of the U.S. Department of Transportation’s University Transportation Centers Program, in the interest of information exchange. The U.S. Government assumes no liability for the contents or use thereof.