October 1972
Environmental Protection Technology Series
Feasibility Study of
Electromagnetic Subsurface Profiling
I*
I
\
LU
O
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This x-?ork provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-T2-082
October 1972
FEASIBILITY STUDY OF ELECTROMAGNETIC SUBSURFACE PROFILING
By
Rexford M. Morey
Walter S. Harrington, Jr.
Contract No. 68-01-0062
Project 1102^ GRF
Project Officer
Allyn St. C. Richardson
EPA - Region I
John F. Kennedy Bldg.
Boston, Massachusetts 02203
Prepared for
OFFICE OF RESEARCH AJD MONITORING
oS. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, B.C. 20^60
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, B.C. 20402 - Price $1.25
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection Agency and approved for
publication. Approval does not signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency, nor does mention of trade names or com-
mercial products constitute endorsement or recommendation for use.
11
-------
ABSTRACT
A study was made of a unique radar system which produces a continuous profile of sub-
surface conditions showing depth and location of geological formations and buried utilities.
Information is obtained by sending electromagnetic pulses into the earth and then
receiving the reflected pulses from interfaces and objects. The unit travels at 3 mph, and
can detect interfaces directly below it to depths of 10 feet in clay and 25 feet in sand. Depth
of penetration is governed by conductivity and dielectric constant. Water content in-
fluences these soil parameters; an increase in water content decreases penetration. The
penetrability of the soil determines the maximum depth at which pipes can be detected. A
break in the pipe can be detected by the saturated soil around the break. Limits of
penetration have not been reached; work is being done to determine empirical standards of
system performance on a wide variety of soils. Since better information yields better cost
estimates for designing sewage collection systems, the advantages of the radar system are
apparent.
This report was submitted in fulfillment of Project Number 120/25-11024 GRF, Contract 68-
01-0062 under the sponsorship of the Environmental Protection Agency.
in
-------
CONTENTS
Section
I Conclusions
II Recommendations 3
III Introduction "
IV Description of ESP System 7
V Objectives ^
VI Geological Considerations 17
VII Underground Pipes and Utilities 33
VIII Appendix 47
IX Acknowledgements 67
X References 69
XI Glossary '1
-------
FIGURES
Page
1 Frequency Spectrum of a Time-Limited Signal (Gaussian) 7
2 Example of Subsurface Profile 8
3 Block Diagram of ESP System 10
4 ESP System in Operation 11
5 Interior of Truck 11
6 Antenna Being Towed by Hand 12
7 Antenna Unit Operating in Rough Terrain 12
8 Typical ESP Graph 13
9 Test Pit 18
10 Set-Up for Antenna Pattern Measurements 18
11 Side to Side Antenna Pattern 19
12 Front to Back Pattern 19
13 Example of Two Pipe Targets 20
14 Textural Classification Chart 21
15 Map of Unlithified Materials 23
16 Coaxial Transmission Line 24
17 The Effects of Interfaces on a Video Pulse Travelling Down a 25
Coaxial Line
18 Block Diagram of Measurement Set-Up for Pulse and Time Domain 25
Reflectometry Measurements
19 Example of Data Using Pulse Method 27
20 Example of Data From TDK Measurement 27
21 Block Diagram for Attenuation Measurement 28
22 Example of Data From Attenuation Measurement 29
VI
-------
FIGURES (Contd)
Page
23 Dielectric Constant vs % Saturation 30
24 Comparison of Depth Scales for Various Samples 31
25 1 Inch Diameter Calibration Pipe at 3 Feet 6 Inches Deep in Sand 32
26 1 Inch Diameter Calibration Pipe at 6 Feet - 3 Inches Deep in Sand 32
27 1 Inch Diameter Calibration Pipe at 9 Feet 6 Inches Deep in Sand 32
28 Location and Depth of Various Pipe Samples 33
29 1/2 Inch Diameter Copper Pipe 34
30 5 Inch Diameter Steel Pipe 34
31 2 1/2 Inch Diameter Steel Pipe 34
32 5 Inch Diameter Orangeburg Pipe 34
33 15 Inch Diameter Vetrified Clay Pipe 35
34 4 Inch Diameter PVC Pipe 35
35 7 Inch Diameter Transite Pipe 35
36 8 Inch Diameter Steel Pipe 35
37 15 Inch Diameter Concrete Pipe 36
38 7 Inch Diameter Transite Pipe 36
39 12 Inch Diameter Steel Pipe 36
40 Test Pit with Pipes 37
41 4 Inch PVC Pipe 27 Inches Deep 38
42 8 Inch Transite Pipe 3 Feet 8 Inches Deep 39
43 8 Inch Transite Pipe with Leak at One End 40
44 Parallel Scan Along Transite Pipe Before and After Being Damaged 40
45 Cross Scan Over Transite Pipe Before and After Being Damaged 41
vn
-------
FIGURES (Contd)
Page
46 Example of Buried Utilities in Stratified Sand and Gravel 42
47 Three Scans Taken at 10 Foot Intervals Across a City Street 43
48 Intersection in Business District of a City (Lawrence, Massachusetts) 44
49 Example of Utility Mapping 45
50 Delta From Groveland Test Site 48
51 Data From Groveland Test Site 49
52 Data From Groveland Test Site (Scan is Perpendicular to Scan 50
in Figure 51)
53 Geological Cross Section in Glacial Delta (See Figure 50) 51
54 Geological Cross Section Figure 51 Interbedded Sand and Silt 52
55 Data From Northfield Test Site 54
56 Cut Bank at Northfield Test Site 55
57 Data From Methuen Test Site (Till) 56
58 Data From Billerica Test Site (Till) 57
59 Data From Burlington Test Site (Till) 57
60 Data From Groveland Test Site (Till) 58
61 Cambridge Test Site (Clay) 59
62 Data From Northfield Test Site (Clay) 60
63 Route 3 and 62 Test Site (Bedrock) 62
64 End of Route 3 Test Site (Bedrock) 63
65 Burlington Test Site (Bedrock) 64
66 Northfield Test Site (Bedrock) 65
via
-------
TABLES
No. Page
1 Properties of Selected Soil Samples 26
2 Comparison of Till Samples 55
IX
-------
SECTION I
CONCLUSIONS
1. At present, engineers and municipalities have no systematic, comprehensive method
of obtaining continuous underground information to pin-point the location of
bedrock and utilities. The cost of laying new sewer pipe is approximately
$100,000/mile; it soars when extensive bedrock, gas-pipes, telephone wires, etc., are en-
countered and have to be avoided or repaired if damaged. A method to provide better
subsurface information would solve the major problem of estimating the cost of in-
stalling new sewer pipe. Another facet of the same problem is that there is no method
of locating "lost" or broken pipe in an extant sewer system. Public works crews must
dig and refill holes in an inefficient quest of faulty water and sewer lines.
2. Present methods of determining the quantity of bedrock consist of borings and
seismic soundings. Borings do not provide enough information at a reasonable cost,
and seismic soundings do not work well at the depths where most utilities are located.
When they even exist, maps and plans showing the location of buried utilities depend
upon surface references that may have been moved or obliterated during subsequent
construction.
3. A radar system developed by Geophysical Survey Systems, Inc., presents a non-
destructive method of profiling the earth. It transmits an electromagnetic pulse into
the earth, and receives and analyzes return pulses which have been reflected from in-
terfaces. The depth of the medium can be determined from the reflected pulse. It is
also possible in theory to identify materials from this pulse. Depth of penetration of
the signal is governed by the conductivity and the dielectric constant of the medium.
As the conductivity increases, attenuation of the signal increases; and therefore,
penetration by the pulse decreases. The effect of water content of four samples
was measured in the laboratory. The percent saturation with water in a silty sand
sample was increased from 2% to 20% and the effective penetration was reduced by
4. Initially, maximum penetration of the pulse through sand in field studies was 25-feet;
penetration through varved clay was 5-feet. Tests conducted after this program was
completed indicate penetration limits have been expanded to 60-feet through sand,
and 10-feet through varved clay.
Laboratory work with metal and non-metal pipes indicates that metal pipes can be
located readily at these same depths. If the non-metal pipe is filled, it is as easily
detected as the metal pipes at the same depths through the same materials.
The system detected the saturated zone of soil around the break in a transite pipe
buried 3 feet-8 inches in a soil composed of 98% sand, and 2% silt.
5. The system at present is utilized in soil and bedrock surveys, and underground utility
mapping. Soil and bedrock surveys consist of a visual inspection by a geologist, an
ESP survey along the proposed trench route, an initial interpretation of the data to
direct a limited'boring program for material identification and to calibrate the data,
-------
and a final interpretation and report. Underground utility mapping includes gridding
the street or intersection on a 10-foot grid pattern, taking the ESP survey along the
grid lines, logging all visible indications such as manholes and gates, checking all
available utility maps, and plotting all this information plus ESP survey results on a
drawing.
6. The system is presently used commercially for subsurface investigations in the New
England area. The main application of the impulse radar system is in the location
and identification of buried or submerged materials, either man-made or natural. Its
unique abilities make it an invaluable tool in any field which requires probing the
earth.
-------
SECTION II
RECOMMENDATIONS
This project was limited to gajjiejlng^^^cinfpjmatio.n_of system performance on spilst
bedrockjjand underground utilities. Limited laboratory and field studies were performed
under ideal conditions. Field tests should be continued in a wide variety of soils to deter-
mine the system's capacities. Due to modifications and improvements of the antenna, dep-
ths of penetration in all media are constantly expanding; limits have not yet been reached.
Although it is known that the radar will penetrate certain materials, the degree of in-
fluence of characteristics such as the number of soil layers, or the water content should be
established. Resistivity measurements are being made in the field with the radar to deter-
mine empirical guidelines which would then serve to evaluate the ability of the system to
perform under different circumstances in various locations. Laboratory analysis of soils
should be continued to determine the correlation between electrical parameters of a
medium and penetration limits of the system for a particular medium and for combinations
of materials. An in-depth investigation of how materials affect broadband video pulses
must be performed. In order to enhance data presentation and to facilitate interpretation
of the ESP data, it is recommended that computer processing of the data be developed.
Finally, data handling should be automated as much as possible.
It is now necessary to use borings in conjunction with the radar system for material iden-
tification and system calibration. It is recommended that this need for borings be
eliminated. Further investigations should be conducted to prove that media can be iden-
tified by knowing the incident and reflected video pulses. These investigations should study
the pulse characteristics, computer analysis, and development of the theory associated with
the transmission of the pulse through a multi-layer medium.
A full-scale demonstration is recommended. The costs and problems encountered by one
municipality building a collection system by means of standard subsurface investigation
techniques would be monitored. ESP surveys would be performed for another cooperating
municipality having essentially equivalent problems. Final construction costs for both
municipalities would be compared.
-------
SECTION III
INTRODUCTION
The present _sewagejp_Qll_ecliQ_n system in the United States represents a tremendous un-
seen investment, and the cost of putting additional sewer pipe into service will increase
this investment to about $82 billion by 1980. The U. S. Department of Commerce predicted
that 2.94 billion linear feet of residential sewer pipe would be in the ground by 1970. This
represents an investment of $53 billion. By 1980, the prediction is that 3.73 billion linear
feet of residential sewer pipe would be in service to handle the flow from 178 million
residents. It becomes economically important to have a thorough understanding of sub-
surface conditons for the efficient design of subsurface sewage collection systems.
Ignorance of detailed subsurface conditions is the major problem in planning and
estimating the cost of installing sewer pipe. These costs are primarily affected by the
geology and the buried utilities along the excavation path.
Presently available methods of mapping the subsurface conditions along a proposed tren-
ching route are not adequate in most cases. For excavation work, the material of primary
interest is bedrock, since the result of errors in estimating the quantity of bedrock can be
very expensive. The most common subsurface investigation technique is boring a hole in
the ground. Rods of various types may be either hammered or drilled into the ground. The
operator counts the number of hammer blows to determine the density of the material. If
drilling is used, the material is brought to the surface for evaluation. In both methods soil
samples are sometimes taken for laboratory analysis. Borings give information only at the
probed points and do not differentiate between bedrock and large boulders. Depending
upon the nature of the investigation, holes are placed every 50 or 100 feet and occasionally
as far apart as every 300 feet. Interpolations of data are then made between these holes to
form estimated profiles of subsurface conditions.
Seismic (sound) profiling is a relatively deep acoustical sounding technique that does not
work well at shallow depths of 10 to 20 feet because of the very short propagation times in-
volved. Sound waves generated at the surface of the ground are reflected from various
layers in the subsurface and are detected by geophones stuck in the ground. The geophones
convert the sound into an electrical signal for recording. The initial seismic energy is
generated by explosives buried in the ground, or by striking the ground with a sledge ham-
mer, or by inducing vibrations in the ground with a mechanical device. Continuous sub-
surface profiles cannot be developed by land seismic surveying.
The location and identification ©f underground utilities, such as water, gas, and telephone,
is necessary for the design and installation of new sewer collection systems. Presently, this
information is retained «n Maps and plans maintained by the various utility companies and
the cities and towms. In many cases the plans do not exist or are inadequate for actually
locating the utility m tihe street.
Geophysical Survey Systems, Inc. has developed a unique radar system for shallow sub-
surface exploration. This system is called Electromagnetic Subsurface Profiling (ESP). The
system produces a continuous profile of subsurface conditions showing the depth and
location of geological formations and buried utilities. This report demonstrates and
evaluates the ESP system's ability to detect difficult excavation conditions that affect
-------
sewer construction costs. The report is three-fold in nature. First, a brief description of the
theory and operation of the ESP system is presented. Second, subsurface geology as it af-
fects the characteristics and capabilities of the ESP system is discussed. Finally, buried
utility location and identification capabilities of the ESP system are outlined.
Geophysical Survey Systems, Inc. obtains subsurface information by sending elec-
tromagnetic waves into the earth from an antenna moving across the ground. As the
energy penetrates the ground it is reflected in part by various geological interfaces and ob-
jects in the ground. These reflections are picked up by the antenna and detected by a sen-
sitive receiver for recording on magnetic tape. The magnetic tapes on which the data are
stored are then processed and the data are converted into a graph. These graphs are one of
the main tools employed by the Company's geophysicists and geologists for determining
subsurface conditions and for producing continuous subsurface profiles.
-------
SECTION IV
DESCRIPTION OF THE ESP SYSTEM
This section will describe the principles, theory, and operation procedures of the Electro-
magnetic Subsurface Profiling (ESP) system.
The ESP System is a broadband video pulse radar. This video pulse radar system emits a
pulse that is approximately Gaussian in shape (see Figure la). It is possible to express this
time domain video pulse through its fourier transform as a frequency spectrum (see Figure
Ib). The reflection of each of the spectral components and hence the video pulse is deter-
mined by the dielectric constant and conductivity of the medium from which it is reflected.
This is because the reflected pulse must satisfy the boundary conditions at each interface.
These boundary conditions are derived from Maxwell's equations (Ref. 1).
-3-2-1 01 23
TIME (t)
a. Time-Limited Signal
0 «•
*> = 2*rf Radian
Frequency o>
b. Frequency Spectrum
Figure 1. Frequency Spectrum of a Time-Limited Signal (Gaussian)
If the video pulse passes through multiple media it is partially reflected and partially tran-
smitted at each interface (see Figure 2). It is the reflected pulse that the radar system
developed by Geophysical Survey Systems, Inc. detects, records, and analyzes.
Video pulse propagation through naturally occurring media is a complicated phenomenon.
Physical and chemical properties of a medium affect the dielectric constant and con-
ductivity. The dielectric constant and conductivity in turn influence video pulse shape and
propagation. It is with these two parameters that the theory and experimentation are con-
cerned. The dieJectric constant determines the velocity at which the video pulse travels
through the medium. If the velocity aed time of travel are known, the thickness of the
-------
material can be determined. The conductivity of a medium causes heat loss of energy in the
pulse and determines the attenuation or loss of energy as the pulse propagates through the
medium. These two parameters depend on the density and moisture content of the medium.
Section VI illustrates how these parameters vary for different soils and moisture content.
LEGEND
Incident Pulse (Pulse Transmitted
by Radar)
Reflected Pulse From Interface
Transmitted Pulse Into Ground
Distance
Time
Dielectric Constant
Conductivity
z and t
Figure 2. Example of Subsurface Profile
-------
The time-limited signal (video pulse) shown in Figure la is assumed to be Gaussian in form,
which is close to the form it has in practice since there is a finite .rise and decay time. The
Gaussian pulse f(t) is represented mathematically by the relation:
f(t) = e'l/ "ul (1)
where ti is the width of the pulse. The pulse given by (1) can be decomposed into a
frequency spectrum F (6)) where
F(o>) =/ fWe^dt (2)
«/ _oo
If f(t) is given by (1), then the corresponding spectrum F (o>) can be found from (2)
A narrow time or video pulse is associated with wide frequency spectrum. For example, if
the time pulse is in the order of one nanosecond (10-;) seconds) in half width, then the
frequency spectrum extends from zero to about 350 MHz. The response R (o>) of a particular
material to a single frequency can be calculated. The actual reflected signal in the time
domain is then obtained by superposition. Thus the time response r (t) which corresponds to
R (<«>) is
r (t) = 1/R (o>) F (w) e ~JG>t do> m
Z "i.00 V '
where F (tu) is the frequency spectrum of the pulse. The relevant quantity for subsurface
exploration is the coefficient of reflection R (o>). For a two-layered medium it is
R =k2'kl (5)
k2+kl
o
=
kl =W I'l-J**!0! k2 = G)y262
where
H = magnetic constant
( = electric (or dielectric) constant
-------
The ESP System is composed of a video pulse transmitter, a receiver, an antenna, a tape
recorder, a graphic recorder, and special control electronics as depicted in Figure 3. The
present unit is packaged in the rear of a four-wheel drive Chevrolet "Carryall" van (see
Figures 4 and 5). The essence of the ESP subsurface investigation method is the analysis of
recorded reflections of radar signals directed into the ground.
Graphic
Recorder
Power
Supply
1
1
Tape
Recorder
Pulse
Transmitter
1
Receiver
Transmit-
Receive
Selector
Ground Surface
Transmitted
Pulse
Target
Relfected Pulse
Figure 3. Block Diagram of ESP System
10
-------
Figure 4. ESP System in Operation
Figure 5. Interiot of Truck
11
-------
Once an area for subsurface investigation has been established, the lightweight antenna
unit is towed over the ground by truck or by hand (see Figures 6 and 7). The antenna tran-
smits a radar pulse into the ground every three inches of travel. A portion of the radar
pulse is reflected from the interfaces of soil, rock and other objects and received by the an-
tenna. Radar reflections from the interfaces are governed by the differential in the dielec-
tric constant and conductivity of the materials. A cable connects the antenna with
receiving and recording instruments in the truck. The radar pulses are displayed in real
time on a screen and are recorded on magnetic tape. By recording a signal return every
three inches, a continuous profile is developed showing utilities and subsurface strata.
Figure 6. Antenna Being Towed by Hand
Figure 7. Antenna Unit Operating in Rough Terrain
12
-------
A typical ESP graph is shown in Figure 8. Interpretation of these graphs requires a great
deal of skill and experience; however, a few pointers on interpretation will be given. The
graphic presentation of a vertical cross-section of earth is called a "profile" The "profile"
has been produced by printing high signal levels in black and no signal white. Intermediate
signals are in the gray range. The surface of the ground is the top horizontal line. The
horizontal scale in feet is dependent upon the speed of the antenna across the ground. The
horizontal scale in the example is approximately 40-feet per inch. The vertical scale is a
time scale, which is converted into a depth scale with a knowledge of the velocity of
propagation in the particular material. The main feature of the data is the display of dark
bands which extend throughout the profile at varying depths. Note that these dark bands
are displayed in groups of three closely related bands. Each group of three banded lines is
the reflection from an interface between two materials. The triple band is a characteristic
of the ESP System and is caused by oscillations in the reflection of the pulse. This
oscillation or banding limits the ability of the system to discriminate closely spaced in-
terfaces. In the sample of data shown, for example, the total span of the three banded
signal is 2-feet. This means that as the spacing between two interfaces approaches 2-feet,
the reflections from these interfaces will begin to superimpose. The lower interface will not
be completely masked since portions of the oscillations will show up in the data. However,
it will be difficult to locate the actual depth of the second interface.
• Ground Surface
DISTANCE (FEET)
180 200
220
240 260
14-
Figure 8. Typical ESP Graph
13
-------
Note also in Figure 8 that there are 4 dark bands just beneath the surface running the
length of the data. These bands are caused by the reflections from the surface and in-
terfaces immediately below the surface. They cause difficulty in tracing interfaces up to
the surface; however, there is information contained in this zone. Strong reflections from
pipes can be picked out of this zone easily, and discontinuities in the surface which may be
caused by trenching or changes from one surface material to another are easily identified.
14
-------
SECTION V
OBJECTIVES
Geophysical Survey Systems, Inc. (GSS) has been contracted by the Environmental
Protection Agency (EPA) to provide the necessary personnel, materials, and facilities to
demonstrate and evaluate GSS's unique radar geophysical measurement system for shallow
subsurface exploration. GSS, Inc. provided and maintained, at its own expense, the Elec-
tromagnetic Subsurface Profiling System.
The objectives of this evaluation are:
1. To determine the signature or picture characteristics of the geological formations
most frequently enountered in sewage collection system construction, and the
relevant anomalies such as buried pipe, cables, difficult excavation conditions, etc.,
that affect sewer construction cost.
2. To determine the eccentricity possibilities of the measurement system; that is, to
determine the possible effects of the shape and/or composition (absorption/reflection)
of a buried material on signature curves and other elements as encountered in the
course of the demonstration. The intent is to outline possible causes of departure and
degree of departure from the fundamental signature curves.
3. To make a "first-cut" estimate of the numbers and kinds of interfaces which can be
identified at one time before the signature curves overlap.
4. To establish the depth limitations of the measurement system under common
geological conditions and with the most frequently encountered obstacles.
5. To compare the curves of broken or crushed sewer pipe with intact pipe with all other
variables remaining constant.
6. To develop a utilitarian method of logging and recording results for use by field
crews. The intent is to have a well thought out method and technique for which the
system can be used in a full-scale demonstration. As part of this task, possible sites
for full-scale demonstration project are suggested.
To meet these objectives, the project is divided into two major parts; (1) geological con-
siderations and (2) an underground pipe and utility study. Laboratory measurements and
field measurements are performed during the investigation of each of the two parts of the
project.
-------
SECTION VI
GEOLOGICAL CONSIDERATIONS
Of the geological formations encountered in the construction of sewage collection
systems, the one presenting the most difficult excavation conditions is bedrock. In order to
study bedrock with the ESP System, an associated study of the overburden must be in-
cluded. In addition, the most important part of any study of buried pipes is a study of the
medium in which the pipe is buried. Therefore, most of the data gathered in this section
concern the soil components of the overburden.
To evaluate the ESP System and its ability to detect geological phenomenon and other sub-
surface conditions, it is necessary to determine the radiation characteristics of the antenna
and to determine the response of the system to various soil materials and conditions. To ac-
complish this end, the study program was organized as follows:
1. Measurements of the antenna's field of view were made in a controlled situation. The
results, of this experiment provide an insight into the printout characteristics, or
signature, of various underground targets, the effects that side targets can have on
the recorded information, and the limitations on the system's ability to detect sloping
surfaces.
2. Data were collected in various natural situations using the ESP System. These data,
when printed out and studied, provide information regarding the capabilities and
limitations of the system. They also provide signature information for various
geological phenomena.
3. Soil samples were collected from all the sites investigated with the ESP System.
These samples were analyzed in the laboratory to determine soil properties that could
affect system operation. In addition to grain size and moisture content deter-
minations, electrical characteristics of the samples were measured under naturalistic
conditions. Water content was varied in some samples to determine the effect of
water on system characteristics.
During the initial phases of the project, a test pit was constructed (see Figure 9) for making
laboratory measurements. This pit was originally used to reconstruct various soil and
geological situations. This approach to geological studies became impractical. More time
was spent filling the pit or changing the material in the pit than in collecting data. In ad-
dition, the situations created were artificial and the data collected not likely to bear any
relationship to field conditions. It was decided to use the test pit as a laboratory standard
or control and to use it extensively in the buried pipe studies.
Antenna pattern measurements were taken in order to understand the reflections received
by the system from broad irregular surfaces such as those encountered in natural
geological situations. A 12-inch diameter steel pipe installed in the test pit was selected as
the target for these measurements, since the pipe presents a long, narrow, polarized reflec-
tor to the antenna. The strength of the reflected signal from the pipe for various beam
angles was recorded along a line normal to the target axis (see Figure 10). These
measurements were taken for both the front-to-back plane and the side-to-side plane. The
results of these measurements are shown in Figures 11 and 12. These patterns differ from
the conventional free-space radiation patterns of an antenna. They are a measure of the
-------
two-way reflection sensitivity, or the field of view, of the antenna. This information is
essential in determining how normal printout signatures can be affected by changes in
reflector surface shape or slope, and how the signatures are affected by subsurface per-
turbations within the field of view.
l"O.D.x 10'Long
Tranilte Pipe
12" 0 D.x
10' Long
Steel Pipe —
\
Washed Concrete Sand
Granite
Coane Gravel
A
Figure 9. Test Pit
Antenna
Orientation For Side
To Side Pattern
Antenna
Orientation for Front
To Back Pattern
Beam Angle
Target
Figure 10. Set-Up for Antenna Pattern Measurements
18
-------
Left
Facing FWD
Relative Field Strength
Right
Facing FWD
80°
TOP
70°
706050403020100 10203040506070
60°
60°
50°
30°
30°
20°
10° go 10°
20°
Figure 11. Side to Side Antenna Pattern
BACK
60°
Relative Field Strength
70 60 50 40 30 20 10
FRONT
10 20 30 40 50 60 70
50°
40°
40°
30°
30°
20°
10° 0 10°
20°
Figure 12. Front to Back Pattern
19
-------
Note that the front-to-back pattern of the antenna is wide; therefore, as the antenna ap-
proaches targets, the targets come into the field of view when the line of sight between the
target and the antenna is approximately 45° The reflection from this target prints out at a
depth that is equal to the "slant range" or distance between the target and the antenna.
This distance is greater than the actual depth of the target. As the antenna approaches the
target, the depth printout approaches the actual depth. This can be seen more clearly by
studying the sample printout of a pipe target shown in Figure 13. The "tails" on the pipe
signature are the reflections from the target when the antenna is at a 45° angle to the
target.
Location of
Pipe
Pipe Signature
Pipe Signature
Location of
Pipe
14-
16-
Figure 13. Example of Two Pipe Targets
On the other hand, the width of the beam or the width of the field of view in the side-to-side
plane of the antenna is quite narrow, and objects or targets that are on the side of a par-
ticular survey path will not be recorded. Also, if light-reflection theory is assumed, plane
surfaces such as interfaces will reflect energy from that portion of the beam that is normal
to that surface. As the slope of this surface approaches 45°, and gets steeper than 45°, the
ray of energy from the antenna at the normal angle to that surface is getting considerably
weaker. Therefore, surfaces making an angle greater than 50° with the ground surface will
not be detected by the system.
20
-------
Once the impracticality of the test pit for geological studies was determined and the
radiation characteristics of the antenna were measured, a program of field studies was un-
dertaken. These studies were confined to areas subjected to continental glaciation in the
recent geologic past. Several basic types of material are found in and around such areas.
These materials are gravel, sand, silt, clay, glacial till, artificial fill, and bedrock.
Textural classification of mixtures of the above materials in this report will be made in ac-
cordance with the Triangular Classification Chart developed by one of the divisions of the
U. S. Corps of Engineers (see Figure 14).
100
Size Limits
Sand 2.0 to 0.05 mm.
Silt 0.05 to 0.005 mm.
Clay....Less than 0.005 mm.
Percent SILT
Figure 14. Textural Classification Chart
21
-------
The nature of unlithified deposits and their relationship to bedrock in non-glaciated areas
varies considerably. Most of the materials are found in both glaciated and non-glaciated
areas; the ESP System's operation in materials from non-glaciated areas will not differ
much from the materials tested except where different chemical properties result in gross
differences in the electrical properties of the soils. For example, silty-sand in a non-
glaciated region could be expected to have the same system charecteristics as the silty-sand
in a glaciated area, but would be quite different if the sample had a high salt content. A
highly generalized map of the major deposits of unlithified materials found in the con-
tiguous U. S. is shown in Figure 15. Selected characteristics of most deposits are briefly
discussed.
Wind-Blown Deposits - These include both silt and sand; high concentrations of
salts exist in some areas of these deposits.
Saprolite - This generally consists of massive clay. The contact with un-
weathered bedrock is generally gradational.
Coastal Plain Deposits These are generally sand, silt, or clay. The water
table is high and swamps are numerous, particularly near the coast.
Desert Deposits These are generally dry sand, silt, or clay. Caliche may be
present in some areas.
Basin Deposits - These are generally dry silt and clay in low areas surrounded by
sand and gravel. Deposits and groundwater may be highly mineralized. Caliche
is common in some areas.
Alluvium - Materials present are variable. The water table is high in many areas;
it may be highly mineralized.
Lake Deposits - These are generally silt and clay. Deposits and groundwater may
be highly mineralized.
Field investigations of each of these non-bedrock materials, as well as bedrock, were con-
ducted in natural situations. These field results are included in Appendix A. Laboratory in-
vestigations of these materials were also conducted and the results of the laboratory tests
were compared to the field results.
Results of the investigations allow a prediction of the effectiveness of ESP methods in most
glaciated areas. In areas where the overburden materials are sand or silty sand, depth of
penetration is excellent, being as much as 25 to 30 feet in good conditions, and as much as
10-feet where lossy conditions exist. System resolution is 2- to 3-feet depending on the
saturation of the material in the area. The number of interfaces penetrated is high; exam-
ples in the data show as many as 6 interfaces. The data received in areas of sand and gravel
deposits clearly show the geological formation of the area. The patterns are distinctive and
these areas are thus easy to classify.
Where the overburden is till, penetration depth is dependent on attenuation charac-
teristics. Attenuation is high in till with high clay content, and depth of penetration is
correspondingly low. The presence of clay in the overburden greatly decreases the
capacities of the system, although some information can be collected. The operation of the
system depends on the percent of clay in the material and the moisture content of the
material.
22
-------
it
-m
[•v;-j Alluvium
^| Saprolite
JQ Lake deposits
|;V§] Glacial deposits
Pt-^ Desert deposits
E§jj Basin deposits
|55| Coastal Plain deposits
^3 Wind-blown deposits
i | Deposits variable - generally thin
Limit of glaciation
v j""'-
****«:
Figure IS. Map of Unlithified Materials
-------
The location of bedrock with the ESP System is a complex and difficult task; however, suc-
cess has been experienced using the following technique:
1. The ESP data are recorded.
2. An evaluation of the site is performed by an experienced geologist.
3. The ESP data are examined for one of the characteristic signatures of
bedrock.
4. One or two of these areas are tested with a boring for material, iden-
tification.
5. The interpretation is completed using the boring information as a guide.
Laboratory measurements were made on the soil samples collected in the field. To deter-
mine the ESP System's response to specific soil materials, one must know the electrical
properties of these materials as they appear to the transmitted pulse. To measure these
properties, a standard coaxial transmission line is used. This coaxial line, shown in Figure
16, is filled with the material sample to be measured and is packed to the desired density.
The nature of the coaxial line lends itself nicely to pulse measurements and Time Domain
Reflectometry (TDR) (Ref. 2). Figure 17 illustrates the effects on a pulse as it propagates
along the transmission line. At each interface part of the pulse is reflected and part is tran-
smitted. Figure 18 is a block diagram of the test set-ups used. Results of these
measurements and the physical properties of each sample are presented in Table I. These
results provide a record of the data collected on a particular sample and can be used in the
calculation of various electrical properties of the samples. From these laboratory
measurements, the velocity of propagation and the amount of attenuation are directly
determined. The dielectric constant and conductivity are then calculated.
Figure 16. Coaxial Transmission Line
24
-------
INPUT
Interface -
f
jC,
Coaxial Transmission Line
OUTPUT
AIR
AIR
Soil Sample
Input
•S 3
1
:s "2 •
Transmission
Transmission
Reflexion
on f\
A
Reflexion
A
Distance & Time -
Figure 17. The Effects of Interfaces on a Video Pulse Travelling Down a Coaxial Line
Figure 18. Block Diagram of Measurement Set-Up for Pulse
And Time Domain Reflectometry Measurements
25
-------
0016-1
0016-2
0016- 15
0016 -15a
0016 • 25
0016-25*
0016 - 27
0016 -27a
0016-30
0016 - 36
0016 - 37
0016 • 37a
0016 • 37b
i 0016-38
SAND
SAND
SILTYSAND
SILTY SAND
SAND
SAND
SILTY SAND
SILTY SAND
SILTY SAND
SILTY CLAY
SAND
SAND
SAND
SAND
UNIT. WT.
109.4
96.7
99.5
72.4
89.03
90.0
91.29
90.6
87.2
72.67
97.6
96.7
105.0
94.14
MOISTURE
CONTENT
ft
2.36
5.39
.55
7.65
3.326
1.87
4.85
12.37
25.22
2.488
5.99
4.1
%
SATURATION
12.2
20.0
2.19
20.12
10.2
6.1
15.6
36.5
52.3
9.5
22.4
14.3
SAND
*b
98
98
48
48
82
82
74
74
68
1
100
100
100
100
SILT
%
2
2
36
36
18
18
26
26
30
56
CLAY
X
16
16
2
43
DIZLECTRIC CONSTANT
METHODS
PULSE
4.0
5.7
7.0
4.7
2.4
3.1
4.6
6.7
4.8
6.8
2.6
4.5
TDR
3.7
5.5
62
4.7
3.1
6.9
10.0
4.8
6.3
4.5
ATTENUA-
TION
db
2.9
8.5
15.8
2.7
1.8
2.9
7.3
2.8
9.6
2,0
3.2
1.8
Table 1. Properties of Selected Soil Samples
The ESP System detects dielectric interfaces, that is, the interface between two materials,
by detecting the change in dielectric constants as it passes through these materials.
As a first approximation, the dielectric constant of the sample material is calculated from
the two-way transit time through the material in the transmission line. This can be
measured two ways. It can be measured by using the pulse method shown in Figure 18. This
measurement results in a curve as shown in Figure 19. Three distinct pulses can be seen on
this trace. The first and largest pulse is the incidence or transmitted pulse. The second
pulse is the first reflection or the reflection of the pulse from the beginning of the material
packed in the sample tube. The third pulse is the reflection of the pulse at the end of the
sample tube. The time between the first and second reflection pulse is the two-way transit
time through the material and is measured between the peaks of the pulse. This time is ac-
curate to within plus or minus one nanosecond. The dielectric constants presented in Table
1 are calculated using this time and the sample length in the following equation:
where:
c=velocity of light, 3 x 108 meters/sec
t =time between pulse peaks
L= sample length in meters
26
-------
Reflected Pulse From
Air Sample li
NOTE: Incident and Reflected Pulses are
Approximately Gaussian in Shape.
Reflected Pulse
From Air Sample
0 5 10 15 20 25 30 35 40 45 50
TIME (NANOSECONDS)
Figure 19. Example of Data Using Pulse Method
The second method to determine the two-way transit time uses time domain reflectometry.
The equipment set-up for this method is also shown in Figure 18. This is a standard
technique for determining the. impedance of a transmission line (Ref. 2). The transmission
line (see Figure 16) used to hold the sample is a 50 ohm impedance air dielectric line. The
resulting waveform from a test of this line without a sample in it, that is, simply as an air
line, would be a straight line trace indicating that there are no impedance discontinuities
along the line. However, when the line is packed with a sorl sample, the dielectric constant
of the soil sample is different from that of the air and causes a change in the impedance of
the transmission line. In the example shown in Figure 20, the soil-filled coaxial line has a
lower impedance than the air-filled line. The second discontinuity occurs at the point where
the sample ends and the line is again 50 ohms. The distance between these discontinuities is
the transit time in the sample material.
6 8 10 12 14 16 18 20
TIME (NANOSECONDS)
Figure 20. Example of Data From TDK Measurement
27
-------
The velocity of propagation of the pulse through the samples is calculated from this transit
time. Knowing the length of the transmission line and the transit time one can compute the
velocity of propagation directly, since velocity equals length (or distance) divided by time.
This velocity of propagation for each geological material is used to calculate a scale factor
for determining the depth scale on the ESP profiles.
Both pulse and TDK measurements were made on all the samples tested, and the dielectric
constant and velocity of propagation were calculated from both methods, thus providing a
cross-check on the measurements.
Attenuation is a measure of the amount of energy that is dissipated or lost as the pulse
passes through a material. Figure 21 is a block diagram of the test set-up to measure the
pulse distortion and attenuation through the sample, using an insertion technique. The
pulse shape before it enters the material is found by connecting the pulse generator direc-
tly to the sampling head of the oscilloscope and recording this pulse on the XY recorder.
The sample-filled coaxial line is then placed between the generator and the sampling head
and the pulse shape is recorded, thus giving the shape of the pulse after it has passed
through the material. A sample of this data is shown in Figure 22. The difference between
the peak level of the two pulses is a measure of the energy lost in passing through the
material. The results of the attenuation measurements are listed in Table 1
Pulse
Generator
Coaxial Line
With
Sample
Sampling
Head
Oscilloscope
X-Y
Recorder
Figure 21. Block Diagram for Attenuation Measurement
28
-------
NOTE: Input and Output Pulses are
Approximately Gaussian in Shape.
6 8 10 12 14 16 18 20
TIME (NANOSECONDS)
Figure 22. Example of Data From Attenuation Measurement
By comparing the results of the attenuation measurements for the different materials with
the ESP data collected in the field, it is possible to estimate the depth of penetration in
each particular material.
All collected soil samples were subjected to- a grain size analysis using standard seive
analysis techniques. Where necessary, hydrometer measurements were made to determine
the content of silt and clay. Moisture content was measured on all samples, and where
possible, in-place density measurements were made. From these data the percent
saturation was calculated. This information was collected in order to have a classification
cross reference to the electrical measurements made on these samples.
Table 1 includes a summary of the physical characteristics of the samples measured in the
laboratory. Soil moisture has a great effect on dielectrical constant and the other related
electrical characteristics. All initial measurements were made on the soil samples as they
were received from the field.
To gain further insight into the effect of moisture on these electrical characteristics, the
moisture content of selected samples was adjusted and additional electrical measurements
were made. The moisture content of three sand samples was increased to approximately
5% by adding water; two samples were dried. In an attempt to determine the relationship
between moisture and dielectric constant, the values for these two characteristics were
plotted against each other. The results were scattered and no conclusions could be drawn.
This is understandable since electrically the soils are a composite material made up of the
soil grains, air and water. In a given sample the variables are the air and water. The dielec-
tric constant of a particular sample is determined by the percentages of the three com-
ponents on a volumetric basis and so varies with each sample. The moisture content
measurements, on the other hand, are based on the dry samplw weight and so do not
provide a good basis for direct comparison of results.
29
-------
To overcome this disparity, the percent saturation was calculated for each of the samples.
This is a measure of the percentage of the air voids that are filled with water, and is
therefore a measure of the water content of the sample on the volumetric basis. The
calculated dielectric constant of the samples were then plotted against percent saturation
and the results are shown in Figure 23. Note that the data follow a predictable trend.
s
Assumed Mean Value
for Sand and Silty
iui oaiiu aim oiuy .-. .
Sand w/< 5% Clay \
15 20 25 30
% Saturation
35
40
Figure 23. Dielectric Constant vs. % Saturation
The indications from the test data are that the dielectric constants for dry sand samples
with up to 30% silt are all within the range of 1.5 and 3.0. As the water content of the
material increases the dielectric constant increases, with all the materials showing the
same effect. A zone is marked off on the dielectric constant-saturation curve that includes
a spread of plus or minus 10% in scale factor. The mean value of the dielectric constants is
also plotted. Based on available data, if the dielectric constant indicated by the mean curve
for a given percent saturation were used to determine the scale factor on the ESP profiles,
the maximum error would not exceed ± 10%.
30
-------
Samples taken from the borings (which will normally be done on all jobs) can be analyzed
for sand, silt and clay content, and for percent saturation. A reasonable scale for the area
then can be selected. More testing is required in cases where the clay content of the soil is
up to 20%, but the indications are that the same regularity of results will exist and that a
method for scaling can be developed for these materials as well.
Examination of the attenuation values shown in Table 1 indicates that this characteristic
is not as easy to classify as the dielectric constant data. However, the attenuation levels for
sand and silty sand are 1.5 to 4.0 db, while the attenuation level for materials with high
clay content is between 8.0 to 15.0 db. It can be seen that this characteristic is borne out in
the field investigations where depth of penetration in clay materials is limited to depths of
approximately 5 feet.
Figure 24 shows the 0 to 10-foot scale to be used on the profiles of the ESP System for the
different materials evaluated. There is a large variation in the scale between samples of
dry sands and samples with high clay content. This is not a problem since large areas to be
investigated with this system would consist of materials of like characteristics, and when
materials with much different scale factors are present they would also exhibit very large
dielectric constant contrast. As is discussed in the evaluation of the dielectric constant
data, a method is available for determining the correct scale factor in the field. As a
calibration test, scans were taken over pipes at known depths in one of the test sites. In or-
der to have the pipe installed at the known depth but still have an undisturbed overburden,
a site was selected with a cut bank running parallel to the scan path. The pipe was buried
under the scan path through the cut bank at a given depth. Scans were taken with the pipe
at three different depths. The printouts of these tests are shown in Figures 25, 26 and 27.
The actual depth of the pipe is shown in the figure, as well as the depth that would have
been determined by the relevant scale factor.
Siliv Sand
Sample
OOld-27
Sand
Sample
00lt>-l
Sand
Sample
Silli Sand
Sample
OOK.-.W
Sill} Cla\
Sumpk
OOI«v3t.
Figure 24. Comparison of Depth Scales foi Various Samples
31
-------
00
3' - 6"
6' - 3"
10-
Figiue 25. 1" Diameter Calibration
Pipe 3'-6" Deep in Sand
Figure 26. 1" Diameter Calibration
Pipe 6'-3" Deep in Sand
Figure 27. 1" Diameter Calibration
Pipe 9'-6" Deep in Sand
-------
SECTION VII
UNDERGROUND PIPES AND UTILITIES
The ESP System can be used to locate underground pipes and utilities. Laboratory tests
were performed to gain an understanding of the ESP System's capabilities for detecting
various types of pipe.
A series of laboratory tests determined the system's response to different pipe materials. A
trench approximately 120 feet long was dug and selected pipe samples were buried at
various depths (see Figure 28). The material in which the pipes were buried is a clayey till,
similar to Sample 0016-15 (see Table.l). As was discussed in the section on geological in-
vestigations, this and other materials with high clay content decreases the system's
penetration ability. This make's clayey material a good medium to test performance on
pipes since it will be possible to bury pipes at or below the penetration limits. ESP
measurements were taken across the top of each pipe and the data were recorded. Figures
29 through 39 are the data printouts of these measurements.
The following conclusions were drawn:
1. Pipes at depths less than 2-feet are easily detected with this system. The reflected
signal is superimposed on the surface reflection oscillations; however, the pipe
reflection is easy to pick out.
2. The maximum depth at which metal pipes can be detected in this material is
probably not more than 5-feet. Maximum detection depths in other materials will
be the same as in the geological tests, that is, 10-feet for tills with negligible clay
content, and up to 25-feet in sand.
3. The non-metalic pipes do not return a strong reflection and in some cases are im-
possible to detect. Concrete pipe cannot be considered a non-metalic pipe in this
study since the pipe contains metal reinforcing wires. The results of the plastic
pipe measurements prompted further studies on plastic pipe.
= . i ; 2 Ji .
•K .-i1- .:n?h & .V -^ •••> ?«* -IV '-.
x C - '"-. ?. f-i S -f. ~. ** 3 V - > 't '* £ 'T x £ 'i x 2 « - o * x H x " a 5e
(PROFILt VltW)
I II
Figure 28. Location and Depth of Various Pipe Samples
33
-------
Center of Pipe
Center of Pipe
2' - 2"
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 29. W Diameter Figure 30. 5" Diameter
Copper Pipe Steel Pipe
Center of Pipe
Center of Pipe
4'-8'
Pipe not detected
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 31. 2V4" Diameter Figure 32. 5" Diameter
Steel Pipe Orangebuig Pipe
34
-------
Center of Pipe Center of Pipe
*
w
4' - 3"
5- m
6-
*>A
•*
^ '^i (Pipe Not
Detected)
7-
8-
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 33. IS" Diameter Figure 34. 4" Diameter
Vetrified day Pipe PVC Pipe
Center of Pipe
Center of Pipe
4'-6
(Pipe
Not
Detected) 5_
7-
8-
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 35. 7" Diameter Figure 36. 8" Diameter
TranritePipe Steel pipe
-------
w
6'-ID"
(Pipe
Not
Detected) g ^
Center of Pipe
Center of Pipe
NOTE: Dimension of side of figure is
measured depth to top of pipe.
(Pipe Not
Detected)
Figure 37. 15" Diameter Figure 38. 7" Diameter
Concrete Kpe Transite Pipe
Center of Pipe
6-
7-
8-
r
8'
1
NOTE: Dimension of side Figure
is Measured depth to top of pipe.
(Pipe Not
Detected)
Figure 39. 12" Diameter
Steel Pipe
36
-------
Two pipes were installed in the test pit (see Figure 40). One of these, a 12-inch diameter
steel pipe, was installed as a target in the study of the antenna system, and as a model
from which the normal signature of a pipe could be determined. To further investigate the
problems of detecting non-metallic pipe, a specially designed PVC pipe was installed in the
test pit. This pipe was fitted with end caps and hose fittings, which allowed the pipe to be
filled with water or drained as desired. It was reasoned that the poor results on non-
metallic pipe in the test trench was caused by an insufficient difference between the dielec-
tric constant of the pipe material and the surrounding soil. In actual conditions these pipes
may contain water or a fluid with a very high percentage of water. If the ESP System can-
not detect the pipe, then it should detect the contained fluid. ESP measurements were made
across the top of this pipe under both the filled and empty conditions. Figure 41a is the
reflection of the empty pipe and Figure 41b is the reflection of the pipe filled with water.
The PVC pipe filled with fluid was easily detected by the ESP System. Another experiment
included attaching a thin wire along the length of an empty PVC pipe. The ESP
measurements showed that the pipe with the wire was detected, while the PVC pipe
without the wire was not detected.
-1" Steel Pipe
-4" PVC Pipe
-13" Steel Pipe
20'
-30'-
Figute 40. Test Pit with Pipes
-------
Center of Pipe
^^w^^^^wp«w»iv(iipp
a. PVC Pipe
Empty
b. Water Filled
PVC Pipe
NOTE: Dimension of side of Figure is
Measured Depth to top of Pipe.
Figure 41. 4" PVC Pipe 27" Deep
An experiment was conducted to determine the capability of the system to detect a
'damaged pipe. The pipe used in the experiment was an 8-inch transite pipe fitted with hose
connections to allow filling and draining it with water. Because transite pipe is non-
metallic it is difficult to detect when empty. Its dielectric constant is close to that of the
surrounding materials and so does not give rise to a strong reflection. This is shown in
Figure 42a which is a crosscut over the dry transite pipe. The known location of the pipe is
indicated on the figure; there is a weak reflection. This would make the detection of a
damaged, unused transite pipe very difficult; however, the situation for this pipe filled with
water or with an organic material is quite different. In Figure 42b, which is a crosscut over
the same path as in Figure 42a, the water-filled pipe reflects a strong signal.
-------
Center of Pipe
a. Transite Pipe
Empty
b. Water Filled
Transite Pipe
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 42. 8" Transite Pipe 3* - 8" Deep
In the course of this experiment one of the end caps in the pipe had developed a slow leak.
Figure 43 is a scan along the length of the water-filled pipe. The data show a strong reflec-
tion below the right-hand end of the pipe indicating a zone of saturated soil.
To obtain a damaged pipe situation, the pipe was drained and a small hole was augered
from the ground surface to the top of the pipe. The auger was used to obtain a minimum
disturbance of the material over the pipe. The pipe was then broken using a steel bar and a
sledge hammer. The augered hole was backfilled and water was run into the pipe through
the fill hose. Figures 44 and 45 compare the parallel and cross scans of the pipe before and
after it was damaged. Note the differences in each scan caused by the saturated area below
the pipe.
39
-------
Position of
Pipe
Results of
Leak
Figure 43. 8" Tiansite Pipe with Leak at One End
. Position i Position i- _
" -I ofPipe r H ofPipe H
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Result of
Broken
Pipe Saturating
Soil Below It.
a. Pipe Note Broken
b. Pipe Broken
Figure 44. Parallel Scan Along Transite Pipe Before and
After Being Damaged
-------
Center of Pipe
Center of Pipe
Result of Pipe
Saturating Soil
Below It.
a. Pipe Not Broken
b. Pipe Broken
NOTE: Dimension of side of figure is
measured depth to top of pipe.
Figure 45. Cross Scan Over Tranrite Pipe Before and After Being Damaged.
This experiment indicates that the system can detect damaged pipe by detecting the
saturated zones in the vicinity of the break.
Field investigations were conducted to determine the capabilities of the ESP System to
locate underground utilities in city streets. Samples of data collected in some of these in-
vestigations are included in this report and are shown in Figures 46 through 48. Figure 47 is
an example of data collected in an area with a high utility congestion. The three scans were
taken at 10' intervals across a city street. This is an example of the ESP System's ability to
discriminate closely spaced pipes. Figure 48 is an example of data taken where trolley car
tracks have been paved over and forgotten.
41
-------
Pipe Connecting
#1
Pipe #5
Pipe#l
Pipe #4
-y mm
100 120 140 160 180 200
100 120
DISTANCE (FEET)
Figure 46. Example of Buried Utilities in Stratified Sand and Gravel
-------
UTILITIES
#1 #2
#3 #4
0 10 20 30 40
40
DISTANCE (FEET)
Figure 47. Three Scans Taken at 10' Intervals Across a City Street
-------
Manhole
Asphalt Covered
Trolley Tracks
Pipes
Pipes
40 SO 60
DISTANCE (FEET)
i
70
i
80
Figure 48. Intersection in Business District of a City (Lawrence, Mass.)
Many practical applications exist for the ESP System in locating utilities and other man-
installed underground systems. It can be used when a system such as a sewage collection
system is to be installed in the street and it is important to know the location and depth of
water and gas services and storm drains that cross the trench path. In this situation a scan
can be taken along the trench path; the pipes crossing the trench path can be located from
the data. They can be shown either in a drawing or can be painted on the road surface. A
second application exists where a sewer line is to be installed through an intersection con-
taining a high congestion of utilities and manholes. For this investigation of utility
location it is best to mark off a grid in the intersection and take data scans along the lines
in the grid. The data are then evaluated in the laboratory and a map of the actual linear
system locations can be constructed. Figure 49 is an example of the results of this type of
utility location investigation.
44
-------
60" MDC W*t*l
CENTRE
I J K
b
""¥f
1
E
B ?
B C I t
I J X L
6
""'Vf
(t-
1
5
,
*> ,
=.
T
5
C*. Co. Nip/
LBC4tlon -/
41 5 d 7 t
r
Figure 49. Example of Utility Mapping
-------
SECTION VIII
APPENDIX
This appendix describes the field investigations of geological materials as they occur
naturally. Laboratory investigations of these materials were also conducted and the results
of the two investigations are compared.
A number of potential study areas were evaluated by the following criteria:
1. The area should have a minimum combination of materials and the material to be
studied should be the major component.
2. Some method of physical examination should be possible along the path surveyed
with the ESP System. The methods employed, listed in order of preference, are:
a. excavation and visual examination;
b. boring with a split spoon sampler; and
c. examination of a cut zone parallel to the ESP survey path.
Sand
A site was selected in Groveland, Massachusetts, where sand could be studied by both the
ESP System and by trenching. The area is a delta, common in glacial areas. The portion of
the delta studied consisted of fine sand interbedded with minor amounts of both coarser
and finer materials. Selected areas of the survey paths were trenched, examined, and map-
ped. Figures 50, 51, and 52 are data printouts from different areas in this site. The survey
line shown in Figure 50 was selected for trenching because of the interesting feature in the
geological deposits shown at Stake 2. A trench, varying in depth from 5-feet to 8-feet was
dug by backhoe along this line, and the distribution of the various materials was mapped
(see Figure 53). Samples of different materials were collected for laboratory analysis. The
major portion of the sand-bedding in this survey line is much like sample 0016-25, which
was taken from another area on the same site. The electrical characteristics shown in
Table 1 for sample 0016-25 apply to this area.
In studying the geological mapping of these material deposits, there are some interesting
features that can be used to evaluate the system's discrimination capability. There are both
thick and thin deposits of material as well as a number of lensing situations (the thinning
out of a stratum in one or more directions). The reflection oscillations in this material are 2
1/2-feet thick, meaning that the resolution of the system in this area is from 2 1/2 to 3-feet.
Most of the strata in Figure 50 are less than 2-feet thick. This, and the fact that lensing
exists, means that the picture printed out is a composite of the superimposed reflections of
the various interfaces. Although distinct individual interfaces of the materials cannot be
picked out of the data, the general characteristics show up well.
Figure 51 shows a printout of another survey line.from the same site. The depth of
penetration to the left of the printout is 2-feet, while 100-feet away at the right of the
printout the depth of penetration is greater than 20-feet. This line was also trenched and a
map of the materials is shown in Figure 54. Note the location of a stratum of material with
a high moisture content. The effect of moisture content can be seen by referring to Table 1
and by noting the results of tests conducted on samples 0016-25, 27, and 37. These samples
were tested at various moisture contents or percent saturation. The tests indicate that at-
47
-------
tcnuation increases with the percentage of water in the sample. The 8.2% moisture is
equivalent to approximately 25% saturation. In Figure 51 the effects of attenuation
became obvious. Above the critical layer the interfaces reflect strong, sharp bands, while
below the critical layer the bands are weaker and indistinct. Also note that as the scan
progresses from stake 3 to stake 9 reflections are received from more interfaces. At stake 4
there are reflections from approximately 4 interfaces. The reduction in the signal strength
caused by increasing number of interfaces can be seen by tracing the lower interface in the
printout from left to right.
Survey Stake Numbers (On a 50 -Ft. Spacing)
0 -
#1
1
#2
i
#3
i
#4
i
Figure 50. Delta From Groveland Test Site
-------
Survey Stake Numbers ( On a SO • Foot Spacing )
9- /.
-
*f
Figure 51. Data From Groveland Test Site
49
-------
Survey Stake Numbers (On a 50-Foot Spacing)
#9 *8 #7
Figure 52. Date From Cleveland Test Site (Scan is Perpen-
dicular to scan in Figure 51)
50
-------
Fine Sand Cemented with Iron Oxide
Numbers (On a 50' Spacing)
U
o.
§
Sand & Silt,
Occasional Pebbles
Compact
Black Fine
Sand
Sand with
Silt
Fine Sand
Interbeds
Fine Sand
Figure 53. Geological Cross Section in Glacial Delta (See Figure 50)
-------
Survey Stake Numbers (On a 50' Spacing)
o-i
u.
NOTE: Percentage Numbers Represent
Moisture content.
Figure 54. Geological Cross Section of Figure 51.
Interbedded Sand and Silt
Critical
Layer
Bottom of
Trench
52
-------
A second deltaic deposit was studied in the Northfield, Massachusetts area of the Con-
necticut River Valley. These deltaic deposits were laid down in a pre-glacial lake that oc-
cupied the valley in early post-glacial times. The deposits consist of pebble and cobble
gravel overlying pebbly sand. Figure 55 is an ESP profile from a scan in this area. Sample
0016-38 in Table lis sand from this site. The depth scale and resolution are the same as for
the Groveland, Mass., site. The topset and foreset beds are distinct and lensing is easily
picked out. Maximum depth of penetration is 25-feet. A photograph of a cut bank parallel
to the scan line is shown in Figure 56.
Glacial Till
Till is defined as non-sorted, non-stratified sediment carried or deposited by a glacier. Tills
are encountered constantly in glaciated regions and are therefore worthy of investigation.
They are mixtures of material of all grain sizes from clay to cobbles and boulders. As per-
centages of each can differ greatly from place to place, it is not possible to categorize the
system's operational characteristics in "till" Four sites selected for investigation by the
ESP System contained till as the major unlithified material. Table 2 lists the locations of
these sites and compares the composition of the materials. Note the differences and
similarities in these till samples. For example, the till from Methuen and Billerica are very
similar, each having a high clay content. The till from Burlington and Groveland are also
quite similar and these have low clay content. Note the differences in the electrical proper-
ties of the two basic types of till. The important difference is the attenuation charac-
teristics. The sample having high clay content has an attenuation of 15.8 db, whereas the
sample having the low clay content had.an attenuation of 2.8 db. It would be expected that
depth of penetration in the Methuen and Billerica sites would be limited. This is borne out
in the samples of data from these two sites shown in Figures 57 and 58. Conversely, good
depth of penetration could be expected from the Burlington and Groveland sites. This is
also borne out as is shown in samples of the data from these areas shown in Figures 59 and
60. Note the confused character of the interfaces and compare them to the very orderly in-
terfaces from the deltaic regions (Figures 50 and 55). The data sample of glacial till in
Billerica was recorded along the top of a cut bank. A silt layer varying from 2 to 4 1/2-feet
deep could be seen in the bank. This layer can be seen in the data and its depth correlates
well with the calibration determined in the laboratory of the material from this site (0016-
15).
Clay
Two sites were selected to study the system's capabilities in clay deposits. A site in Cam-
bridge, Massachusetts was selected which had 4- to 8-feet of artificial fill, sand and gravel,
underlain by massive impermeable marine clay. The clay is from 3- to 14-feet thick and is
underlain by glacial till. Figure 61 is a printout of the data from one of the survey lines
taken at this site. Note that a well-defined, strong reflection is received from the top of the
clay deposit. Some weak reflections are received from below this interface indicating high
attenuation of the signal. The sample of the clay from this site (sample 0016-36) was tested
in the laboratory. Because of the difficulty in packing the clay into the electrical test sam-
ple tube, a realistic test sample density was not achieved and the sample was tested with an
inordinate quantity of air voids. For this reason the electrical characteristics shown in
Table I must be considered approximate and low. It can also be seen from the data that the
discrimination limit in clay is 18-inches.
53
-------
DISTANCE (FEET)
200 220 240 260
0-'
1-
i**K.V4V
J3S&G*
m m
Figure 55. Data from Noithfield Test Site
-------
Top of Cut Bank
Figure 56. Cut Bank at Northfield Test Site
SAMPLE
0016 • 14
0016-15
0016-23
0016-30
LOCATION
METHUEN, MA.
BILLERICA. MA.
BURLINGTON. MA.
GROVELAND. MA.
SAND
«
31
48
66
68
SILT
%
56
36
33
30
CLAY
«
13
16
2
2
MOISTURE
CONTENT
10
7.6
7
12.3
DIELEC-
TRIC
CONSTANT
11
6.9
ATTENUA-
TION i
db
16.8
-
2.8
Table 2. Comparison of Till Samples
-------
20
DISTANCE (FEET)
80 100
3-
l»nrp-
"•V y& --^
-5-S8 ?
j»v
•\-^
,*•'
4-
B
a
0 6-
I
8-
9-
M
ft
0.
Figure 57. Data From Methuen Test Site (Till)
-------
DISTANCE (FEET)
0 25 50 75 100
Figure 58. Data From Billerica Test Site (Till)
DISTANCE (FEET)
60 80
100
120
140
mW'fiif?
**» *., *-* p,f a**
i .J>
ftS
5-
6-
•'*££*
v Wff*^1 ,
I*l*r.Ji
#»,
!rf"»| '
. r^#.,»w
D 4i
J ^ wjy
^jfel
w
r-Si*N
t«^!M
^ Sw
• ™ l*v
fe^^>44
*S !^«K? -..^ &
iw'
(.SB
"111:.. 2
: • ^^mm^imM^mim
m
Figure 59. Data From Burlington Test Site (Till)
57
-------
DISTANCE (FEET)
60 80
s
* «**> 1 •*»• * ' *
-
Figure 60. Data From Groveland Test Site (Till)
-------
Clayev Till Clay
0
DISTANCE (FEET)
SO 100
Sand&
Fill
Figure 61. Cambridge Test Site (day)
A second site was investigated in the Northfield, Massachusetts area of the Connecticut
River Valley in the same general area as that studied for deltaic deposits. Figure 62 is a
data printout of a scan taken over lacustrine deposits consisting of thin, alternating beds
of silt and clay.
Test probes were made at selected points along the survey path to determine the depth to
the top of the clay beds. The reflection from the top of the clay beds is welf defined in the
data. Of particular interest in this scan is the reflection from an interface within the clay
beds. This is probably an unusually thick layer of silt. The printout of this interface is
another good example of signal attenuation through a lossy medium. Note that this is a
lens shaped deposit, being very thin at the edges and having a maximum thickness of ap-
proximately 5 1/2-fcet. Note that as the thickness of the deposit increases, the reflection of
the silt layer becomes weaker until at 5 1/2-feet it is barely discernible. This indicates that
the maximum depth of penetration into a varved clay deposit of this type is 5 1/2-feet. Sam-
ples of the clay deposit were tested in the laboratory but, because of the extreme difficulty
in handling this material, the electric measurements were not conclusive. The electrical
characteristics of this material are assumed t«> lie similar to those of sample 0016-36.
-------
DISTANCE (FEET)
Top of Clay
Sand Over Clay
250
H
uu
Sat Layer
In Gay
Figure 62. Data from Northfield Test Site (Clay)
-------
As can be seen from these data and from the data collected on tills, the presence of clay in
an area greatly decreases the capacities of the system, although some information can be
collected. The operation of the system is dependent upon a number of variables, the most
important of these being the percentage of clay in the material and the moisture content of
the material. More study of the system operation in clay media is necessary.
Bedrock
The most difficult materials to evaluate are the formations of igneous and metamorphic
rock. Whereas sedimentary rocks have a layered or a flat structure and therefore good
reflective characteristics, the igneous and metamorphic rock have highly irregular surfaces.
These irregularities scatter more energy than they reflect. This may account for the ab-
sence of reflected signal or the confused nature of the reflected signals from areas of
known bedrock. The detection of bedrock from the ESP data must be done by an ex-
perienced interpreter. At present the ESP survey must be coupled with a field investigation
performed by a trained geologist. After this investigation, the data from probable bedrock
areas are examined for the many signatures of bedrock. Figures 63 through 66 are examples
of data taken in known bedrock areas. These data were taken over paths where bedrock
shows in a parallel sidecut, or across the path above outcrops, or in areas where bedrock is
known to be and later verified by borings. Figure 63 is a sample of data that was taken
across the top of a rock surface and away from the rock. Where the rock is at the surface
there is an absence of well-defined reflections. As the rock gets deeper the sand interfaces
lens in with the rock surface. These help to identify the location of the rock surface. Figure
64 is a sample of data taken along a path parallel to a cut which showed bedrock. The
bedrock has a less definite interface reflection and is more defined by the lack of a signal.
Figure 65 is another sample of bedrock close to the surface. However, in this sample notice
the more definite reflections from the rock surface. Also note the characteristic lack of
signal below the rock surface. Figure 66 is another example of data taken along a path
where bedrock was close to the surface. Probes were made through the sand overburden to
the top of the bedrock. Note the confused type of reflection received from the bedrock and
also note the strong reflections received from the bedded sand overburden.
61
-------
Bedrock
Figure 63. Route 3 and 62 Test Site (Bedrock)
-------
DISTANCE (FEET)
40
I
**:: it ' \ •«*» J
.
^v..
Hgme64. End of Route 3 Test Site (Bedrock)
-------
DISTANCE (FEET)
0 20 40 60 80 100 120 140
o-;.
i i
Figure 65. Burlington Test Site (Bedrock)
64
-------
Gravel
Sand
Figure 66. Noithfield Test Site (Bedrock)
-------
SECTION IX
ACKNOWLEDGEMENTS
The laboratory and field studies, experimentation, analytical work and report
preparation was performed by a team from Geophysical Survey Systems, Inc. consisting of
Messrs. W.S. Harrington, Jr., K.J. Campbell, R.M. Morey, G.J. Bontaites, Jr., A.K. Drake,
and Ms. E. Bertram.
The support of the project by the Environmental Protection Agency, and the help provided
by Mr. Allyn St. C. Richardson, the Project Officer, and Mr. Francis J. Condon, Project
Manager Municipal Pollution Control Section, is acknowledged with sincere thanks.
67
-------
SECTION X
REFERENCES
1. Adler, Richard B.; Chu, Lan Jen; Fano, Robert M.; Electromagnetic Energy Tran-
smission and Radiation, pp 190, John Wiley & Sons, Inc., 1960.
2. Strickland, James A.; Time Domain Reflectometry Measurements, Tektronix,
Inc., 19.70.
3. Lundien, J. R.; Terrain Analysis by Electromagnetic Means, Technical Report
3-693, U. S. Army Waterways Experiment Station, 1966.
4. King, Ronald, W. P.; "The Transmission of Electromagnetic Waves and Pulses
into the Earth", Journal of Applied Physics, Vol. 39, No. 9 4444-4452, August 1968.
5. Needleman, S. M.; Molineux, C. E.; Earth Science Applied to Military Use of
Natural Terrain, AFCRL 69-0364, Air Force Cambridge Research Laboratories, L. G.
Hanscom Field, Bedford, Massachusetts, 1969.
6. Nikodem, H. J.; "Effects of Soil Layering on the Use of VHF Radio Waves for
Remote Terrain Analysis", Proceedings of the Fourth Symposium on Remote Sensing
of the Environment, University of Michigan, 1966.
7. Leighty, R. D.; "Remote Sensing for Engineering Investigation of Terrain",
Proceedings of the Fifth Symposium on Remote Sensing of Environment, pp 667-683,
1968.
69
-------
SECTION XI
GLOSSARY
Angle of Incidence; the angle between the normal to the surface at the point of incidence
and the line of propagation approaching the surface.
Attenuation; the reduction of the amplitude of an electrical signal.
Characteristic Impedance; the ratio of voltage to current that depends only on parameters
of the transmission line.
Conductivity; a measure of the ability of a material to conduct an electric charge, the
reciprocal of resistivity.
Dielectric Constant; the permittivity of a material.
Electromagnetic Wave; a wave propagating as a periodic disturbance of the elec-
tromagnetic field.
Impedance; symbol Z; a measure of the total opposition to current flow in an alternating
current circuit, equal to the ratio of the root mean square electromotive force in the circuit
to the root mean square current produced by it, and is usually represented in complex
notation as, R + jX, where R is the ohmic resistance and X is the reactance.
Permittivity; the ratio of electric flux density produced by an electric field in a medium, to
that produced in a vacuum by the same field. Also called "relative permittivity" or dielec-
tric constant.
Pulse; an abrupt change in voltage, either positive or negative with respect to a reference,
that conveys information. This change is characterized by a rise and decay of a finite
duration.
Pulse Width; the time in seconds of the duration of the pulse measured between the half-
power points of the pulse.
Radar; a mehtod of detecting distant objects and determining their position, velocity, or
other characteristics by analysis of very high frequency radio waves reflected from 'their
surface. RA(DIO) D(ETECTING) A(ND) R(ANGING).
Reflectivity; the ratio of the reflected electric field intensity at the interface or target
divided by the incident electric field intensity.
Signature Test; a test in which reflected power from the interface or target is measured as
the antenna is moved through a range of incidence angles.
Step Function; a mathematical function that rises in zero time to unity and remains at
unity for all time (a normalized B.C. voltage, for example).
Time Domain Reflectometry; a method of measuring reflections in the time domain from
discontinuities occurring on a transmission line.
71
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1. Report No.
2.
Feasibility Study of Electromagnetic Subsurface Profiling
7. Authors) Morey>RM
Harrington, W.S., Jr.
9. Organization
GEOPHYSICAL SURVEY SYSTEMS, INC.
16 Republic Road
No. Billerica, Mass. 01862
' .12. Sponsoring Organization ; ;
IS. Supplementary Notes
Environmental Protection Agency report
number EPA-E2-72-082, October 1972.
3. Accession No.
w
1 f /
S. Report Date
6.
8. Performing Organization
Report No.
10. Project No.
120/25-11024GRF
11. Contract/Grant No.
68-01-0062
13. Type of Report and
Period Covered
16. Abstract
A study was made of a unique radar system which produces a continuous profile of subsurface conditions
showing depth and location of geological formations and buried utilities. Information is obtained by sending electro-
magnetic pulses into the earth and then receiving the reflected pulses from interfaces and objects. The unit travels at
3 mph, and can detect interfaces directly below it to depths of 10 feet in clay and 25 feet in sand. Depth of penetra-
tion is governed by conductivity and dielectric constant. Water content influences these soil parameters; an increase
in water content decreases penetration. The penetrability of the soil determines the maximum depth at which pipes
can be detected. A break in the pipe can be detected by the saturated soil around the break. Limits of penetration
have not been reached; work is being done to determine empirical standards of system performance on a wide variety
of soils. Since better information yields better cost estimates for designing sewage collection systems, the advantages
of the radar system are apparent.
17a. Descriptors
*Remote Sensing, 'Subsurface Investigations, *Geologic Formations,*Underground Structures, *Excavation,
Radar, Profiles, Construction Costs, Engineers Estimates, Stratigraphy, Bedrock, Soils, Pipes, Electric Cables
/ 76. Identifiers
*Electromagnetic Profiling
17c. COWRR Field & Group Q7B, 06B, 08A, 08D, 08G
18. Availability
19. Security Class.
(Report)
20. Security Class.
(Page)
Abstractor
21. No. of
Pages
22. Price
'?', Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
•vt U.S. DEPARTMENT OF THE INTERIOR
" : WASHINGTON, D. C. 20240
Institution
WRSIC 102 (REV. JUNE 1971)
913.26!
------- |