United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
P.O. Box 15027
Las Vegas NV89114
August 1933
Research and Development
vvEPA
Geophysical Techniques
for Sensing Buried Wastes
and Waste Migration
-------
ERRATA
GEOPHYSICAL TECHNIQUES FOR SENSING BURIED WASTES
AND WASTE MIGRATION
1/27/84
Page No. Correction
-^"6 Paragraph two - before "The field investigation may not ..."
insert "Because".
^34 Paragraph three, last line - delete last word "Specific" and
change heading on following line to "Specific Geophysical
Methods".
*^43 Correct equations as follows:
V.--S-
WJr
D = cT = VmT
2l£? 2
where Vm = velocity in material «
c » a constant, the velocity of light (3 x 10 m/sec)
er » relative dielectric constant
T * two-way travel time In naao-seconds
(1 nano-second (ns) = 10" seconds)
51 Line one - change "115" to "15".
Paragraph three, line two - change "pat" to "part".
^Une two - change "41" to "41A".
three - change "41" to "41B".
nine - change "a" to "an".
Paragraph -five, third line up from bottom - delete "(lower
and Insert "layer at depth. In this case, the apparent
resistivity will be lower".
Correct equations as follows:
(2IIA) (V/I)
pa » apparent resistivity (ohm-meters or ohm-feet)
A • "A" spacing (meters or feet)
V • potential (volts)
I - current (amperes)
-------
ERRATA (Continued)
Page No. Correction
Third line up from bottom — change “polt” to “plot”.
L.—t30 Paragraph five, line one — add hyphens after “two” and “three”.
L.-132 Paragraph three, line four — add hyphens after “two” and “three”.
‘190 Bottom line — delete everything after “used”.
Delete lower heading for Table 9.
‘i 8 Add attached page )98a.
Add attached page 203a.
Paragraph three, line three — change “area, playground.” to
“as a playground.”
‘—114 Last paragraph — change “111(a)” to “112(A)” and “ 111(b)” to
“112(B)”.
Last paragraph, line two — change “conservation” to
“conservative”.
‘120 ‘laragraph two, line eight — change “depostis” to “deposits”.
Paragraph five, line three — change “16 meter” to “15 meter”.
-------
GEOPHYSICAL TECHNIQUES FOR SENSING BURIED
WASTES AND WASTE MIGRATION
by
Richard C. Benson,
Robert A. Glaccum, Michael R. Noel
TECHNOS, Inc.
3333 N.W. 21st. Street
Miami, FL 33142
Subcontractors to Lockheed Engineering
and Management Services Company, Inc.
P.O. Box 15207
Las Vegas, Nevada 89114
Contract No. 68-03-3050
Project Officer
J. Jeffrey van Ee
Advanced Monitoring Systems Division
U. S. Environmental Protection Agency
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
NOTICE
The information in this document has been funded wholly or
in part by the United States Environmental Protection Agency
under contract number 68-03-3050 to Lockheed Engineering and
Management Services Company, Inc., it has been subjected to the
Agency’s peer and administrative review, and it has been approved
for publication. The contents reflect the views and policies of
the Agency. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.
ii
-------
ABSTRACT
Descriptions of the use of six geophysical techniques are
presented to provide a broad understanding of their application
to sensing buried wastes and waste migration. Technical language
is avoided as much as possible so that those with limited technical
background can acquire a general understanding of current tech-
niques sufficient to define project requirements, select profes-
sional support, and monitor and direct field progrms.
Emphasis on cost—effective investigations at hazardous waste
sites requires an integrated, phased approach: (1) preliminary
site assessment involving the use of aerial photography, on—site
inspections, and readily available information to approximate
site boundaries and locations of waste concentrations, as well as
probable site geology; (2) geophysical surveys to pinpoint buried
wastes, estimate quantities, and delineate plumes of conductive
contaminants in groundwater: and (3) confirmation of groundwater
contamination through monitoring well networks designed on the
basis of plumes and subsurface stratigraphy defined by the
geophysical surveys.
The six geophysical techniques described include metal detec-
tion, magnetometry, ground penetrating radar, electromagnetics,
resistivity, and seismic refraction. Metal detectors and, magneto-
meters are useful in locating buried wastes. Ground penetrating
radar can define the boundaries of buried tranches and other
subsurface disturbances. Electromagnetic and resistivity methods
can help define plumes of contaminants in groundwater. Resistivity
and seismic techniques are useful in determining geological
stratigraphy.
Simple metal detectors respond to changes in electrical
conductivity caused by the presence of metallic objects, both
ferrous and nonferrous. Magnetometers detect perturbations in the
earth’s geomagnetic field caused by buried ferromagnetic objects
such as drums, tools, or scrap metal. They sense ferrous objects
at greater depths than metal detectors and can locate objects
even in the presence of interferences created, for instance,
by fences.
A ground—penetrating radar system radiates short—duration
electromagnetic pulses into the ground from an antenna near the
iii
-------
surface. These pulses are reflected from interfaces in the earth
(such as trench boundaries) and picked up by the receiver section
of the antenna. Electromagnetic conductance measuring devices
yield a signal proportional to the conductivity of the earth
between the transmitter and receiver coils. Many contaminants
will produce an increase or decrease over the background con-
ductivity and thus can be detected and mapped. The resistivity
method measures the electrical resistivity of the geohydrologic
section which includes the soil, rock, and groundwater and provides
a tool to evaluate contaminant plumes and locate buried wastes.
Seismic refraction techniques can determine the thickness and
depth of geologic layers and the travel time or velocity of
seismic waves within the layers, thus revealing variations in
site conditions.
This document was submitted by Technos, Inc., in fulfillment
of Contract No. 68-03-3050 to Lockheed Engineering and Management
Services Company, Inc., under the sponsorship of the U.S.
Environmental Protection Agency. This report covers a period
from August 1, 1981, to December 31, 1982, and work was completed
as of December 31, 1982.
iv
-------
CONTENTS
Page
Abstract
Figures .
Tables. .
Acknowl edgetnent
I
II
III
Iv
V
VI
VII
VIII
Ix
x
XI
• . S S • S
• S • S • • S
Introduction
The Field Investigation
Evaluation of Subsurface
Ground Penetrating Radar
Electromagnetics (EM).
Resistivity . . .
Seismic Refraction
Metal Detection (MD)
Magnetometer . . .
Applications . . .
Closing Conunents .
• . iii
• S V
• xiv
• xv
1
.. 6
• 18
• 38
• 63
• 91
117
142
163
189
• 224
Bibliography . . . . . . . . . • . • . . • • . • . . . .
233
S • S S S
• . S S •
• S S S S
Conditions
(GPR)
S S S S S
S S S S S
• S S S S
S S S S S
S S S S S
V
-------
FIGURES
Number Page
1. Level of understanding versus level of effort of
hazardous waste site investigation 3
2. Factors affecting the subsurface investigation plan 7
3. Regional, local and detail aspects of a hazardous
waste site may all play a role in site investigation.... 10
4. Buried stream channel may direct hazardous waste flow... 11
5. Fractured rock can direct hazardous waste flow 11
6. Cross section of dissolved limestone karst area show-
ing potential for rapid transport of ground water
contamination to nearby stream 13
7. Distribution of karst areas in the U.S. (Ref. Davies
USGS). ........ . . .. ... ..... .... .. ....... . . 13
8. Uniform “layer cake 1 ’ soils. . . . . . . . . . . . . 19
9 . A complex so i. 1 ho r 1. Z on . . 2 0
10. Solution—eroded limestone.... . . ... .. . . .. . . . . . 21
11. For monitoring wells are the minimum required by RCRA... 23
12. Sample from drilling and monitor wells is only
representative of the immediate area . ........ 23
13. Ratio of overall site area to target area is often
large...... 24
14. Probability of detecting a target using a rectangular
gridandrandomly locatedborings............... 25
15. Simplified comparison of the volume sampled by
geophysical and dril lingmethods 30
vi
-------
FIGURES (Continued)
Number Page
16. Simplified example of the volume sampled by continuous
g eophys i cal meas urenient . . . . . 30
17. Continuous measurement will provide greater resolution
thanlimitedstationmeasurements. .......... 31
18. Three—dimensional perspective view of geophysical
electrical conductivity data from parallel transects
across a hazardous waste site 32
19. Isopleth map of geophysical electrical conductivity
data shown in Figure 18. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
20. Three modes of using remote sensing (geophysical)
methods . . . .......... . . . . . .. . . . . . 33
21. Block diagram of ground penetrating radar system.
Radar waves are reflected from soil/rock interface....... 41
22. Photograph of radar system equipme nt showing four
antenna sizes to right................................... 42
23. Radar record showing irregular clay horizon 44
24. Example of single radar waveform and resulting
graphic record . . . . . . . . . . . . . 47
25. Vehicle—towedradarantenna 48
26. Hand—towed radar antenna in limited—access area 50
27. Radarprofileoverburiedpipe .... 52
28. Real—time processing eliminates steady—state noise 55
29. Interpretation of radar data results in geologic
cross section 57
30. Location and boundaries of trenches may be obtained
frompare llelradartraverses... ...... 59
31. Soil profile showing two soil horizons and the edge of
a Paleo sink—hole . . . . . . 60
32. Radar profile of granite outcrop showing fracture
zones...... . . . . . . . . . 61
33. Example of radar traverse over trench 62
vii
-------
FIGURES (Continued)
Number
34. Block diagram showing EM principle of operations........ 66
35. Small hand—held EM system used in soil survey...... 67
36. Shallow EM system used in continuous record mode 68
37. Deep EM system used for station measurements 69
38. Truck—mounted EM system provides continuous
conductivity data to 15 meters depth 70
39. Range of electrical conductivities in natural soil and
rock . .... ,......., i.... . ... ....... 72
40. Continuous EM measurement (A) provides greater
resolution than limited station measurements (B). 74
41. Continuous EM measurement provides greater
resolution than limited station measurements 75
42. EM soundings are obtained by discrete station
measurements 76
43. Eleven parallel, continuously recorded EM profiles . 81
44. Three—dimensional perspective computer plot of EM
data showninFigure 43 82
45. Computer-generated isopleth plot of EM data shown in
Figures 43 and 44 •1•* 83
46. Sounding data yields vertical electric section which
can be related to geohydrologic section.. ... .. . ...... .... 85
47. Continuous EM data (bottom) is calibrated by three
borings (top) 87
48. EM method was used to map widespread contamination of
ground water caused by free flowing brackish well * 89
49. Computer plot of EM conductivity data, obtained over
a buried waste site. 90
viii
-------
FIGURES (Continued)
Number Page
50. Range of resistivities in commonly—occurring soils and
rocks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
51. Typical field setup for resistivity sounding (clay
cap at Love Canal)..... . . . . . . . . . . . . . . . . . . . ......... ...... 95
52. Diagram showing basic concept of resistivity
measurement . . . . . . 96
53. Threecommonelectrodearrangements.. . . 97
54. Increased electrode spacing samples greater depth and
volume of earth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
55. Profile measurements are accomplished by fixing the
electrode spacing and moving the entire array . 99
56. Resistivity profile across glacial clays and gravels..... 101
57. Isopleth resistivity map of profiling data . 102
58. Resistivity sounding curve showing two—layer system 103
59. Flow diagram showing steps in processing and interpre—
pretation of resistivity data 106
60. Two—layer master curves used to interpret Wenner
sounding data. . . . . . . . . . . . . . . . . . . . 108
61. Geoelectric cross section derived from seven resistivity
soundings... . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 110
62. P. three—dimensional or fence diagram may be constructed
frommu ltip leresistivitysoundings............. . 110
63. Field sounding curve over a four-layer geologic section.. 113
64. Correlation of resistivity sounding results to a
driller’s 10g....... . . . . .. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
65. Cross section of leachate plume based upon specific
conductancefroml974welldata.......................... 115
ix
-------
FIGURES (Continued)
Number Page
66. Isopleths of resistivity profiling data showing extent
of landfill plume. •.... ........ •. ......... . 116
67. Field layout of a 12—channel seismograph showing the
path of direct and refracted seismic waves in a two—
layer soil / r o c]c system. . . . . . . . . . . . . . . . . . . . . . . 1 20
68. A portable six-channel seismic refraction system
in use 123
69. A typical seismic waveform from a single geophone. 124
70. Recordingfromal2—channelseismograph........ . 125
71. Time/distance plot for a simple two—layer structure 126
72. Use of forward and reverse seismic lines is necessary
to determine true velocities and depths with dipping
horizon . . . . . . . . . . . . . . . . . . . . . . . 128
73. Time/distance plot shows scatter caused by non—uniform
soil conditions 131
74. Geologic section interpreted from seismic data........... 133
75. Flow diagram showing steps in processing and inter-
pretation of seismic refraction data...... . . ... 134
76. Time/distance plot showing lateral velocity change 135
77. Time/distance plot of field data showing three layer
geologic system. . . . . . . . . . . . . . . 138
78. Interpreted seismic data (Figure 77) compared to
driller’s log 139
79. Time/distance plot of field data showing forward and
reverse seismic refraction data..... ... .. . ... . . 140
80. Geologic section resulting from interpretation of seismic
data (Figure79) 141
x
-------
FIGURES (Continued)
Tumber Page
81. Industrial pipe/cable locator............... ....... 144
82. Typical treasure hunter type metal detector with large
search coil 142
83. Specialized metal detector system in use... . 146
84. Specialized metal detector system with large search
coil . . . . 147
85. Truck—mounted metal detector system provides rapid site
coverage over large areas....... . . .. 148
86. Simplified block diagram of a pipe/cable type metal
detector system * . . . . . . . 149
87. Approximate detection ranges for common targets 152
88. Continuously-recorded metal detector data over a trench
with buried drums . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
89. Three—dimensional perspective view of metal detector
data from parallel survey lines over a single trench 160
90. Plan view map of burial tench boundries based upon metal
detectordatainFigure89................. ...... 161
91. Perspective view of metal detector data from parallel
survey lines shows a complex burial site.... ...... 162
92. Distortions in the earth’s magnetic field due to
concentrations in natural soil iron oxides (left) and
buried iron debries (right) 165
93. Station measurements of a magnetic anomaly caused by a
buried steel drum. . . . . . . . . . . . . . . . . . . . . . . . 166
94. High sensitivity (0.1 ganmia) total field proton
magnetometer being used for station measurements . 168
xi
-------
FIGURES (Continued)
Number Page
95 . Fluxgat e gradiomet er . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
96. Fluxgate gradiometer . . . . . . . . . . . . . . . . . . 170
97 . Fluxgate gradionteter . . 171
98. Simplified block diagram of a magnetometer . 172
99. Comparison of total field and gradient measurements 174
100. Total field magnetometer response (in gammas) for
differenttargetdistanceandntass ..178
101. Magnetometer response will vary considerably depending
upon traverse location and direction with respect to
the target......... 179
102. Diagram of magnetic anomaly over burial trench...... 182
103. A single magnetic profile line showing a wide range
of magnetic anomalies 184
104. Simple contour map of magnetic anomalies shows relative
concentrationofburieddrums ...... . 185
105. Three—dimensional perspective view of magnetic profiles
over a trench containing buried drums. 186
106. Radar record over three buried 55—gallon steel drums...... 199
107. Technical resources and tools which may be applied
to subsurface investigations at hazardous
waste sites. 201
108. Data.from a single seismic refraction line and
resistivity sounding . . . . . . . . . . . . . . . . 208
109. Comparison of data obtained by auger, seismic refraction
and resistivity methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
xii
-------
FIGURES (Continued)
Number Page
110. Three—dimensional perspective view of EM data showing
spatial extent and magnitude of conductivity anomaly 21].
111. Contour plot of EM conductivity anomaly in figure 110
showing extent of buried contaminants 212
112. Metal detector and magnetometer data over a single
trenchcontainingburieddrums............. 215
113. Radar traverse across same burial trench as in figure
111 217
114. Mapping of leachate plume using resistivity methods
shows changes in plume over four—year period............. 219
115. Isopleth map of EM conductivity data at a hazardous
waste site shows a plume (shaded area) leaving the
site and considerable variation in surrounding site
conditions •.... 221
116. Three—dimensional perspective view of EM data
shown in Figure 115 . . . . • • • . • . . .... . . 222
117. Technical resources and tools which may be applied to
subsurface investigations at hazardous waste sites 225
118. Cost comparison curve for hazardous waste site investi-
gation using monitor wells alone versus an integrated
systems approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
xiii
-------
TABLES
Number Page
1. Applications of Geophysical Methods to Hazardous
Waste Sites...... . •....... ......... .•.......... .. •....... 5
2. Approximate Conductivities, Dielectric Constants, and
TravelTimeforVariousEarthMaterials.................. 45
3. Range of Velocities for Compressional Waves in Soil
& Rock.. . . . . . . . . . . . . . . . . . 119
4. Summary of Magnetometer Characteristics 176
5. Characteristics of the Six Geophysical Methods 190
6. Typical Applications of the Six Geophysical Methods 191
7. Susceptibility of Geophysical Methods to “NOISE”.. . 192
8. Comparison of Resistivity and Electromagnetic
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
9. Comparison of Metal Detector and Magnetometer Methods.... 197
xiv
-------
ACKNOWLEDGEMENTS
This document was prepared by Techrios, Inc., under Contract
No. 62—03-3050 to Lockheed Engineering and Management Services
Company, Inc., from the U.S. Environmental Protection Agency,
Environmental Monitoring Systems Laboratory, Las Vegas, Nevada.
The document was written by Richard Benson, Robert Glaccurn
and Michael Noel of Technos, Inc.
A group of independent reviewers, experts in the field of
remote—sensing geophysics, provided valuable input and guidance.
The authors wish to express their thanks for the contributions
from these reviewers:
Sheldon Breiner GeoMetrics
Frank Frischlcnecht U.S.G.S.-Denver
J.D. McNeill Geonics
Gary Olboeft U.S.G.S. Denver
Phil Rornig Colorado School of Mines
Gerald Sandness Battelle Laboratories
A.A. Zohdy U.S.G.S.-Denver
Also acknowledged are the coordination efforts of
Dr. Eric Waither, Lockheed Engineering and Management Services
Company, Inc., project manager.
The assistance and cooperation extended by numerous other
individuals who were contacted on matters related to this
document is gratefully acknowledged.
All illustrations and tables were provided through the
courtesy of Technos, Inc. except for Figures 7, 24, 28, 56, 60,
61 and 94, for which credit is given in the figure caption.
xv
-------
SECTION I
INTPOD(JCTIOFSJ
Rackg round
Traditional approaches to subsurface field investigations
at hazardous waste sites (HWS) often have been inadequate. This
is evidenced by the increased number of papers, conference topics,
and research projects devoted to the problem, and by the tighten-
ing of ground water regulations. Traditionally, site investiga-
tion for contaminants has relied upon (1) drilling to obtain
information on the natural setting, (2) monitor wells for gather-
ing water samples, and (3) laboratory analysis of soil and water
samples. This approach has evolved over many years, and is
often regarded as the standard analytical approach. However,
there are numerous pitfalls associated with this direct sampling
approach, which can result in an incomplete or even erroneous
understanding of site conditions.
tn the designing of monitor well networks, the p]acement of
wells has been done mainly by educated guesswork. The accuracy
and effectiveness of such an approach is heavily dependent upon
the assumption that subsurface conditions are uniform, and that
regional trends hold true for the local setting. Howev.er, these
assumptions are frequently invalid, resulting in non—representa-
tive locations for monitor well placement. If an attempt is made
to improve accuracy by installing additional wells, the project
may he thrown off schedule, and costs will increase. Such delays
are often unacceptable in rapid assessments required at HWS. At
certain sites, there are also increased safety risks associated
with drilling into unknown buried materials.
The accuracy of results can he affected by other errors
such as contamination introduced during drilling operations, well
construction, and sampling or preservation procedures. These
errors have been neglected for a number of years, hut recently
have been recognized as creating a major problem. Since the
number of spatial samples is typically limited, at many sites no
more than five monitor wells or clusters are drilled. If data
from just one of the five wells or clusters contains a signifi-
cant error, 20% of the total raw data is in error.
Some measurements are difficult or impossible to obtain by
conventional methods. For example, since the detection of
1
-------
contaminant movement through the unsaturated zone, and the
determination of detailed local flow patterns and directions are
both difficult, solutions are often derived hy conjecture. The
investigator’s inability to make some of the desired measure-
ments, and the enhanced possibility of error, as well as the
potential for increased risk, lost time and high direct costs,
often result in low levels of confidence in traditional methods
of field investigations.
Figure 1 shows a hypothetical curve representing the level
of information developed for a PWS investigation versus the
effort involved. Many investigations result in an unexpectedly
low level of accuracy. However, this need not always he the
case. There are two ways for investigators to obtain results
with optimal levels of accuracy: they can add more money, time,
sample stations, etc. , or they can adopt an integrated systems
approach. The latter is safer and more cost—effective and this
document describes the techniques used in such an approach.
fluring the past decade, xtensive development in remote—
sensing geophysical equipment, portable Field instrumentation,
field methods, analytical techniques and related computer
processing has resulted in a striking improvement in our
capability to assess hazardous waste sites. Further, many of
these improved methods allow measurement of parameters in the
field and rapid site characterization, sometimes with continuous
data acquisition at traverse speeds up to several miles per hour.
Some of these geophysical methods offer a direct mear s of
detecting contaminant plumes nd flow directions in both the
saturated and unsaturated zones. Others offer a way to obtain
detailed information ahout subsurface soil and rock conditions.
This capability to rapidly characterize subsurface conditions
without disturbing the site (much like nondestructive testing
used in many production facilities and test laboratories) offers
the benefits of lower cost and less risk, and provides better
overall understanding of complex site conditions.
Once a spatial characterization of the site is made by
these methods, an optimal direct sampling plan may he designed
to:
o Minimize the number of drilling sites;
o Locate drilling and monitor wells at representative sites;
o Reduce risk associated with drilling into unknowns;
o Reduce overall project time and costs;
o Provide improved accuracy and confidence levels.
In brief, geophysical methods provide a means of rapid
reconnaissance to characterize the site; drilling and monitor
wells are then used to provide specific quantitative data
from discrete stations, which have been located so as to he
2
-------
representative of site conditions. The drilling and monitor
wells are no longer used for expensive hit—or—miss reconnaissance
sampling. In sum, the systems approach using geophysical methods
utilizes cost—effective means for improving our understanding of
site conditions (Figure 1).
Geophysics has already been successfuJ.ly applied to many
F1WS investigations. Notable examples include Love Canal in New
York, “alley of the Drums in Kentucky, and the 5Rth Street
Landfill in Florida. At these sites, and numerous others
throughout the country, geophysics has been used to define
plumes, locate buried drums, detect boundaries of burial
trenches, and determine geological settings. The synergistic
features of an integrated systems approach combining traditiona]
and contemporary geophysical methods have resulted in enhanced
quality, safety, and cost—effectiveness in investigations at
numerous HWS.
ON SITE OBSERVATION
LITERATURE SEARCH
NEAR TOTAL LEVEL
OF UNDERSTANDING
LEVEL OF EFFORT
Figure 1.
Level of understanding versus level of effort for
hazardous waste site investigation.
100%
1’
OPTIMUM LEVEL OF UNDERSTANDING
FOR LEVEL OF EFFORT
C D
z
i-0
ZO
U W
0i•-
_1Cl)
w
>
LU
-J
INITIAL RECONNIASSANCE
3
-------
Objective of this Document
This document is primarily intended for management and
administrative personnel responsible for investigation and
assessment of hazardous waste sites. Although it is not
intended as a “how to” book, it does provide a basic understand-
ing of the technology and some field procedures for those who
may he involved in making recommendations or assisting with such
field activities. It has been assumed that the reader’s
technical background in the fields discussed in this document
may be limited. Accordingly, use of detailed theory and formulas
has been minimized. Technical language has been avoided as much
as possible. In some cases, the more technically—qualified reader
may recognize some imprecision; the authors recognize thic
possibility hut have opted to favor the general reader in an
effort to achieve better communication. The document should
provide the reader with a general understanding of current tech-
niques, together with sufficient background to allow him to proceed
to define project requirements, select professional support, and
monitor and direct field programs.
The scope of this document has been limited to the
description and use of six geophysical techniques. They are:
o Ground Penetrating Radar
o Electromagnetics
o Resistivity
o Seismic Refraction
o Metal Detection
o Magnetometry
These six techniques were selected because they are
regularly used and have been proven effective for hazardous
waste site assessments. Within each of the six techniques,
discussion has been further limited to equipment and methodology
meeting the same criteria. The application of geophysics to
hazardous waste site assessments is a relatively new field (less
than 5 years old). There are only a few qualified practitioners,
and a limited number of proven methods, equipment, etc., regularly
in use. With some types of geophysical equipment, there may he
only one manufacturer supplying the entire field. The methods
discussed offer a capability for in—situ measurements, and often
complement each other technically. Table ] shows some possible
applications of each of these methods to 1WS assessments.
The primary tasks to which these methods can he applied
include:
o Mapping of natural geohydrologic features;
o Mapping of conductive leachates and contaminant plumes
(landfill leachates, acids, bases);
£1
-------
0 Location and boundary definition of buried trenches;
o Location and definition of buried metaflic objects (drums,
pipes, tanks).
Organization
The document is organized as follows: Section II focuses
on factors that influence the planning and execution of a Ht S
assessment. Section III contains an overview of some of the
limitations of traditional approaches, together with an intro-
duction to concepts of applying geophysical techniques for RWS
evaluations. Sections I” through I X discuss use of six remote—
sensing geophysical techniques in particular, and Sections X and
XI summarize capabilities and limitations of the six methods,
concluding with a presentation of case studies.
TARLF’ 1.
APPLICATIOI’JS OF OFOPHYSICAL MF.THOPS TO
F-1AZAPflOITS WASTP SITES
APPUC Irn0N
Map ir of Ge& ydto!o lc
rea t . es
I
—
-
‘. appl g c Co ductlvo
Leac ates ar.d Ccrtta-i1n ii
P).jmes tex. La lifl,. AcI !s.
Paseal
2
I
I
—
—
-
Locaflon ar4 Oour. ary
L eft -tuon of 8 irted
Tr” cle3 with Metal
I
I
2
2
2
2
Loea! on erd boundary
I fIntti n of jriel
r.erc ea withaut Metal
I
1
2
2
—
—
Lcca:loi a d Doftntto.,
o’ B .rie Metallic Obtocts
‘ex. CriD ancoI
2
2
—
—
1
I
1. Primary method — Indicates the most effective method
2. Secondary Method — Indicates an alternate approach
I
/
5
-------
SECTION II
THE FIELP It%1VRSTI(’ ATION
Background
A proper evaluation of a hazardous waste site must include
consideration of a large number of variables, many of them
unknown and potentially interacting in a complex manner. A site
investigation program must define conditions to the necessary
level of accuracy, and meet project schedule and cost
constraints.
This section presents some of the variables that the project
manager may encounter and address in his project plan. It is
important to appreciate the magnitude and scope of various
factors in planning, field investigation and final analysis. -
field investigation may not provide all of the answers in a
particular site assessment, it is necessary to have a good
understanding of the variables in order to evaluate available
data, identify missing information and evaluate its relative
significance to the project.
Objectives
Figure 2 shows some of the many factors which may influence
the planning and execution of a HWS assessment. The three
primary objectives usually involved in subsurface investigations
at HWS are shown in the center of the figure and include:
o Location of buried waste materials, including the
resolution of quantity and type;
o Determination of the presence of plumes and the direction,
rate of movement, and distribution of contaminants;
o Characterization of the natural geohydrologic conditions,
and manmade factors which will influence these conditions.
Location of Rurjed Materials——includes establishing the
boundaries of trenches, as well as their depth and volume. The
investigator will wish to assess the contents of a trench or
burial site. For example, he may ask these questions: Were the
materials bulk—dumped or containerized? Are there drums present?
where are the drums located within the site, and how many are
there? Knowledge of the precise boundaries of burial sites is
F;
-------
COST Geologic 1 Hydrotog
( FEATURES
Biologic Anomoul
Culverts Burled )
\ lleS,Flfl>/
7 US
SOCIAL
( WASTES
Public OpInion
Types Form j
Pollilcel Pressure \DlsPOsOuI Chs>.
Press Hostilily
I SITE ASSESSMENT
( Wosie Lo o1ion j
Plume Oct hiiflon J
( AcTIONO \
LEGAL
REQUIREMENTS
WASTES WITHIN J
Access
Proprietary Dote
Legal Aclleri
/
LOGISTICS
Site LocatIon
c r.FocIlltussK v
< RVEYING
Figure 7. F’actors afI cting l he cubstirface investigation plan.
-------
important for safety considerations, as well as for quantifying the
contents for remedial action planning. For example, placement of
a monitor well in a trench may puncture containers within the
burial site resulting in explosion, fire, or release of toxic
fumes. An existing seal between the trench contents and the
surrounding soils or rock may also be breached. Prilling in
areas with soluble rocks such as limestone, could lead to rapid
movement of the contaminants into underlying aquifers.
Determination of the Presence of Contaminant Plumes——and
their flow direction and movement rate is commonly required.
Often the first question is whether leakage from the 1- 1WS is
occurring. If the existence of a plume is confirmed, it will he
necessary to establish its direction and extent. Rut if there
are only a few monitor wells, the data they provide may not he
truly representative of site conditions, and thus may lead to
incorrect conclusions. A typical example: If the local ground
water flow does not coincide with the regional flow, the monitor
wells may he incorrectly p]aced, and may fail to indicate the
presence of an existing plume. aeophysica monitoring of plume
location and dynamics may avoid these prohlem .
Characterization of Subsurface Conditions——is usually a
major portion of the field investigation. t most sites, the
local soil and rock types, the depth of the water table, and the
direction of ground water flow will strongly influence movement
of contaminants; therefore, these factors must he well defined,
as must the potential of the contaminants to he retarded by
soils. Natural anomalies within the geohydrologic section must he
taken into consideration, as well as surface drainage, sewers,
and buried utilities, all of which can affect both surface and
ground water flow around a HWS.
Factors to he Considered
The ten factors shown in the perimeter of Figure 2 may
impact on a specific site assessment. The following discussion
reviews some of these variables.
I ) Natural Site Conditions
o surface features can he easily observed; they include
the natural setting, vegetation, topography,
geomorphology, physiography, and cultural development;
o subsurface features including soils, rock, hydrologic,
chemical and biologic conditions cannot he seen;
therefore these subsurface conditions are difficult to
evaluate.
If subsurface conditions were as uniform as layer cake, the
assessment would be relatively straightforward. However, in most
field situations, this will not he the case. For example, a small
R
-------
change of a few percent in sand/clay content can change hydraulic
permeability by a factor of in. Therefore, the site investigator
must he alert to small variations which will cause significant
hut unsuspected errors. The subtle changes in subsurface
conditions are by far the most difficult to detect.
Daily or seasonal effects of temperature and precipitation
will influence contaminant stability and migration. Areas of
ground water recharge are important because contaminants may
easily enter a ground water system. Subtle variations in
permeability will permit preferential ground water flow
directions and rates unsupported by regional conditions, surface
observations or limited borings.
Assessment of natural site conditions requires that the site
he considered at various dimensional scales (Figure 3). While a
specific waste site may be only an acre in size, its contamina-
tion may he spread over many tens of acres. Its impact upon the
surrounding area can depend upon its regional setting, including
geology, vegetation, population, water supply, rivers, lakes, and
seasonal factors. Insight into the character of the local
setting can be derived from knowledge of the broader regional
picture, therefore it is commonly necessary to plan the investi-
gation to include an area considerably greater than the HW
itself. This will, provide an overview, which will enable the
local site conditions to he more rapidly and accurately evaluated.
(Contaminant transport by ground water and the geohydrologic
factors controlling it do not stop at property lines.
An analogy can he drawn to the use of a camera’s telescopic
zoom lens to zoom in from an overall view to a close—up of the
finer details. Cmitting the broad overview can result in a
number of critical gaps in information about the setti.ng. Here
are some examples:
Figure 4 shows a hazardous waste site situated over an old
buried stream channel. In many cases such channels act as
preferential pathways for movement of contaminants because of
their increased permeability. Understanding that a regional area
contains buried stream channels, and knowing where they may he
located, will he a significant aid in assessing the local
situation.
Figure 5 shows a hazardous waste site in a soil overlying a
massive hard rock, such as granite. Within the massive rock
itself, little if any water flow occurs. However, these rocks are
often fractured, increasing the overall permeability of the bulk
rock. To maximize the yield of potable water, wells are drilled
to intersect such fractures. The same fractures may also hecome
conduits for contaminants to move into the bedrock and ground
water system. The investigator must he aware of the regional
geologic setting, the secondary porosity of the granite due to
9
-------
fracturing, and the extraction of drinking water from these
localized fractures. The directions of major fracture trends u y
also he known by local drillers or geologists.
Regiona’ Setting
Local Setting
Direct Human Exposure
Vlø Skin Contact
and inhalation
ndrect Hu ncn Exposure
Deposition on Crops
ngestion in Animols
[ n
\ 0 WeIi
Leaclioli Movement in Soil and Groundwater
Wo te Site
Deposition
Surface Runall
Figure 3. Regional, local and detail aspects of a hazardous
waste site may all play a role in site investigation.
Gaseous and
ne Part ici
Septic Tank
In
-------
\
Buried Channel
Figure 4.
Buried stream channel may dir ct hazardous waste flow.
PLAN VIEW CROSS SECTION
Figure 5. Fractured rock can direct hazardous waste flow.
11
Hozordous
Waste Site
PLAN VIEW CROSS SECTION
-------
Figure 6 illustrates the influence that the presence of
dissolved limestone (karst) may have on a HWS. Soluble rocks
such as limestone and slowly dissolved by natural waters. Caves
and surface collapse result from this dissolution and associated
erosion. Waterfilled caverns are often part of the ground water
system under such conditions. In addition to these major water
conduits, the existence of many smaller fractures leads to in-
creased permeability. These site conditions are referred to as
karst, and susceptible areas are well—known on a regional basis.
Rapid and direct communication may occur between the HWS and
local and regional ground water at such a site. Figure 7 shows
the regional ground water at such a site. Figure 7 shows the
distribution of soluble rock area in the United States. Accord-
ing to Davies of the U.S. Geological Society Survey, 15% of the
United States has limestone or other soluble rock at or near the
surface. IE other forms of activity (pseudo—karst) anc] mining
activites are included, up to 54% of the U.S. area is included.
2) Cultural Features
Cultural development and modifications can also affect
the HWS. Paved areas and drainage systems concentrate surface
waters. Trenches for buried pipes, sewer lines, telephone
cables, and other utilities are often back—filled with materials
which are more loosely packed, or more permeable than the natural
soil and rock. These pathways are potential conduits for the
rapid movement of contaminants, which have been observed follow-
ing such pathways. The existence of canals and the pumping of
ground water may influence migration of contaminants over the
surface and into ground water. In addition, leaks from many
pipes or tanks are sources of pollution.
3) Hazardous Waste: Types, Forms, and Methods of Disposal
The Resource Conservation and Recovery Act (RCRA)
defines a hazardous waste as that which can cause substantial
damage to health or environment when improperly managed. The
definition of hazardous wastes includes four characteristics:
o Ignitability
o Corrosivity
o Reactivity
o Toxicity
To determine if wastes may cause or potentially cause
“substantial” hazard to human health or the environment, other
factors can be considered:
o degree of toxicity
o concentrations
o potential to migrate into the environment
o potential to bioaccumulate
o possibility of improper management
12
-------
Figure 6. Cross section of karst area showing potential for rapid
transport of ground water contamination to ne trby
stream.
Figure 7. Distribution of karst areas in the U.S.
(Ref. Davies USGS).
13
-------
o quantities of the waste
o historic records of human health and environmental
damage
The site investigator must be prepared for and deal with
virtually any material and condition. These materials may be
hydrocarbons, organic complexes, herbicides, pesticides, toxic
gases, heavy metals, explosives, fly ash, sludge, specific or
complex industrial processing wastes, or radioactive substances.
They may be in the form of gases, liquids, powders, sludges, and
solids——alone or mixed with general debris.
Waste materials may be buried or found as surface or
subsurface leaks and spills. They may leak or evaporate from
impoundments, or may simply be left abandoned on the surface. In
some cases, materials have been disposed of in rivers or estuaries.
Some will mix readily with water and some will not. Many
components at a HWS will migrate rapidly through the unsaturated
and saturated zones. Others will move more slowly and can be
attenuated by various chemical and biological mechanisms during
transport. Denser contaminants will sink more rapidly due to
their weight, while hydrocarbons, which are lighter, will float
on water.
The type of hazardous waste, its method of disposal, and
its behavior in the environment are quite varied. The
investigator must be aware of these factors when considering his
technical approach, so that he may select an optimal combination
of technologies.
4) Interaction of Wastes and the Natural Setting
Many contaminants will move by advection along with ground
water flow arid become dispersed by mechanical dispersion and
molecular diffusion. These processes cause spreading of the
contaminant, not only along the line of flow but also transverse
to the flow path. It is possible for some contaminants
(conservative parameters) to travel for long distances and to
spread out over large areas as a result of these processes. One
such plume had migrated more than 8 miles in 35 years and had
contaminated the ground water in over 10 square miles.
Regional and local ground water flow will often differ
due to influences of pumping, presence of canals, lakes and
impoundments, and runoff, as well as local changes in soil and
rock permeability. Flow rates are commonly estimated based on
soil or rock permeabilities shown in reference literature or from
laboratory tests; however, these usually indicate lower permeability
than is observed in the field. Furthermore, permeabilities are
commonly referenced to water as the pore fluid, whereas
permeabilities based upon specific chemicals are often found to
14
-------
differ significantly. Some chemicals have been observed to
migrate up to 10 times faster than ground water. In addition
to influencing flow rates, certain chemicals may react: directly
with the natural soil or clay liner materials to i.ncrc ase their
permeability, as for example by desiccation cracking of these
materials. Accordingly, it is important to know not only the
type of contaminant, hut also the kinds of materials within
which the contaminant is contained or is migrating.
Physical and chemical characteristics of soil and rock, and
biologic factors will significantly influence the transport and
attenuation of many hazardous materials. Some contaminants will
be retarded or attenuated as they move into the soil or with the
ground water. Materials such as heavy metals and PCB’s are often
readily attenuated by absorption or adsorption. If these
materials pass through clays or natural organics (mucks, peats),
they are attenuated more than by passing through clean sand.
Clean fine sands will provide greater attenuation than will more
permeable materials, such as coarse sands or highly permeabl.e
limestone. This increased attenuation can significantly retard
the extent: of contamination.
Low pH materials (acids) will he neutralized by the car-
bonates of natural limestone and by the buffering effect of sea
water. A chemically reducing environment will tend to immobilize
or retard the movement of heavy metals, while an acid environment
will allow metals to move more freely. In many cases bacterial
action will significantly reduce hydrocarbon fuel oil concen-
trations.
The many ways in which contaminants can interact, be absorbed
or released from the soils, and migrate through the unsaturated
and saturated zones will influence the technical approach selected
for subsurface investigation.
5) Site Surveying and Positioning
It is critical to any field work that the investigator
be able to position himself and the data with adequate accuracy.
Further, it is important that other investigators be able to
return to any location with the desired level of accuracy, so
that data obtained from various investigations may be compared.
On the other hand, it makes no sense to spend time and money on
plotting a survey grid with a high degree of accuracy if such
accuracy is unneccessary.
The level of precision and accuracy necessary will depend on
whether general reconnaissance or detailed work is in order, and
how the information is to he subsequently used. In some cases,
plus or minus 10 feet will be more than adequate; in others, a
tolerance of a few inches will be required. In some cases only a
random walk search with no pre—search survey grid laid in may be
15
-------
needed to look for buried drums. If a trench with drums is
detected, its boundary may then be surveyed and mapped.
When survey grids are required, they may he laid down
relative to some on—site reference points, arid may be paced off,
or located by tape measure. If it is required that the data then
he located by tape measure and reference points, a professional
and survey crew should be engaged.
6) Safety Aspects
FIWS operations require that suitable health and safety
precautions be met. The site must he initially characterized so
that a suitable health/safety plan can be implemented. Continu-
ous atmospheri.c monitoring may be required if digging, drilling,
or drum sampling activity is taking place. Decontamination of
both materials and personnel must he considered in on—site
operations. Special training, equipment and standards must he
utilized for field activities. Crews dressed in heavy, hot
protective clothing will certainly work at decreased efficiency.
Therefore, increased time and costs of site surveys will result
from increased levels of safety requirements.
Another safety problem to he considered is the risk of
drilling at a HWS without prior site characterization. ‘As the
number of drilled holes increases, the probability of acciden-
tally hitting a target such as a buried drum of hazardous waste
also rises. The methods discussed in this document may be used
before drilling or backhoe work to characterize the site and
minimize the likelihood of accidents and liability.
7) Logistics
A HWS may he located in a heavily-populated area (Love Canal,
New York) or in a pasture many miles from the nearest town (Denney
Farm Site, Missouri). Support facilities and their access are
important to [ -IWS investigations. The program should identify
facilities such as airports, hotels and restaurants, as well as
safety support such as hospitals, fire and police facilities.
Water for drinking and decontamination purposes may have to be
brought to the site by tank truck and special disposal arrange-
ments, for wash—down water and other disposables, must he made.
Weather conditions will obviously affect personnel comfort
and efficiency and may also influence technical work. If
possible, field work should be scheduled in periods of good
weather; if not, allowances should he made for unusual weather
conditions, particularly if respirators, self—contained breathing
apparatus (SCBA) , or cumbersome protective clothing must he worn.
Site access is often a critical aspect of field work. Steep
slopes, heavy vegetation or wet ground can inhibit movement of
both personnel arid equipment. Working access is sometimes
16
-------
required riot only at the immediate site but also in the
surrounding area, for tracing off—site contamination, or
obtaining background reference data.
8) Legal
Suitable personnel ID’s, procedures, and common courtesy are
called for in dealing with people. In matters which will involve
legal proceedings, the site investigator should consult with the
project legal staff before beginning the job. While initial
reconnaissance work may require only routine documentation,
subsequent investigations may require extensive documentation
for legal purposes. If the site investigation data are to he
used in court, attention to proper documentation, traceability of
samples, calibration and analysis will be important. There may
be legal obstacles impeding site access, due to liability
considerations and/or legal actions in process.
9) Social
Social and political, aspects of HWS investigations are
worthy of attention. Will local residents, special. interest
groups, agencies or industries be hostile in any way to the
presence of a field team? Is a “low profile” of activities
required to avoid unnecessarily alarming people? Are press
statements necessary and should a sp cific person be assigned
this responsibility? Will private citizens come in contact with
the UWS operation? What safety measures may he required for
nearby residents or passersby?
10) Economics
It is essential to develop a technical plan and budget
which are compatible and which meet the objectives. However,
some flexibility should be incorporated into both the technical
program and the budget, because the complexity of HWS assess-
ments will, not generally allow a detailed technical plan, arid
unforeseen variations are bound to occur.
Summary
It should be apparent that the HWS investigation is a complex
problem because of the many variables involved. Generally the
simple first—order approximations of site conditions will be
addressed first and then information will be upgraded as budgets
and time permit. The skills, tools and effort brought to bear
should be focused to bring about a rapid convergence of information
and results, to avoid (or at least minimize) some of the pitfalls
of the traditional approach. Practically speaking, one cannot
expect to obtain results that are 100% accurate; the investigator
must understand the possible effects of the variables involved
and he must be able to judge when he is close enough to the
project objectives.
17
-------
SECTION III
EVALUATION OF SUBSURFACE COMDITIONS
Background
In any subsurface assessment, the investigator hopes to
find a simple “layer cake” system of uniform, flat-lying soil and
rock strata (Figure 8). In the real world, however, conditions
are not so simple. Major variations in the soil, and rock
profiles occur in both horizontal and vertical directions
(Figures 9 and 10). These spatial variations can range from
macroscopic to microscopic, arid they all can affect HWS
conditions.
Subsurface variations are controlled by the stratigraphy
and structure of the geologic deposits and formations. Although
the individual geologic formations may be homogeneous, an entire
section may be heterogeneous because of differences in hydraulic
permeability between layers. Structural features such as joints,
fractures, folds and faults also influence the direction and
speed of water movement within the bedrock. Even mineral
composition or grain size substantially influence water seepage,
arid therefore the movement of contaminants. This geologic
heterogeneity can have a profound effect on the interrelationship
between regional and local ground water flow systems.
Both natural and man-induced factors can affect
subsurface conditions. Increased precipitation will lead to
greater leachate production from landfills, provide rapid
transport of contaminants by surface water, and may also provide
a benefit by diluting contaminants. Hydrocarbons or other light
materials will float on top of ground water, in the form of a
surface lens. In shallow water table conditions, elevation of
the water table by heavy rainfall may cause the movement of an
otherwise immobile contaminant. Man—induced fluctuations include
pumping of ground water for agricultural, industrial or drinking
purposes. Nearby pumping will often influence the direction and
rate of local ground water and contaminant flow. Accurate
measurement of subsurface conditions becomes more difficult as
both natural and man—induced variables increase site complexity.
18
-------
Figure A.
Uniform “layer cake”
19
Mtt. ’A
AER!C P4APt .t XJCO
-
‘Z
44
I
‘ t
uV’
• ¼ ’ • ‘
soils.
-------
A
0
4,
I
Figure Q. A complex soil horizon.
-------
Figure 10. solution—eroded limestone. The overhurden has been removed from
the limestone so it can he mined.
1
*
4. .
fit
r
‘S.
r
4
It
.
— 4
-------
Limitations and Requirements of Spatial Sampling
In monitoring a HWS, the new Resource Conservation and
Recovery Act (RCRA) standards require the installation of at
Least. one upgrarlient well and three downqradient wells to monitor
ground water contamination (Figure 11). No reference is made to
well locations or their relation to uniformity of site conditions.
The soil sample obtained from drilling, or the water sample
from a monitor well, is representative only of the immediate
surroundings from which it came, as shown in Figure 12. If the
subsurface geologic and hydrologic parameters are highly variable,
serious omissions and errors often result from the interpretation
of a limited number of sampling points. Clearly, in order to
reach a high level of accuracy, a statistically valid sampling
program must he implemented which anticipates the possibility of
site variability.
An insight to the number of discrete samples that are
required for site definition can be obtained by considering
detection probability curves. Figure 13 shows a burial site
which is 1/10 of the total site area. (The size arid location of
the target area are usually unknown.) Based upon detection
probability calculations, the number of samples or borings
required to achieve various detection probabilities at this site
is shown in Figure 14. More than ten holes are required for a
uniform grid search pattern and more than 40 are needed for a
random search pattern, in order to achieve a probability of
detection approaching 100%. If an error is made in estimating
the target size, and it is in fact smaller than assumed 1 a much
lower detection probability will result. A series of “misses” in
a drilling program will obviously lead to an erroneous conclusion
as to the presence or absence of a target.
With a smaller target, such as with a fracture system a few
inches in width, the As/At ratio (site to target area ratio)
increases significantly, and assessment by drilling becomes
almost impossible. Typical As/At ratios at various hazardous
waste sites may range from less than 10 to more than 1,000. As
this ratio increases, the search problem can rapidly become
comparable to “looking for a needle in a haystack”.
The above example discusses only the problem of hitting the
target, which requires only one contact: it does not address the
problem of definition of the target’s shape. Additional
sampling will be required to establish the spatial extent of the
target and to define its perimeter. As the shape of the site
becomes more complex, or if it is made up of several smaller
sites, or if the project requirements dictate that detailed
boundaries be established, the number of borings needed will
increase greatly. It is obvious that to achieve a good
22
-------
statistical evaluation of complex site conditions would require
test holes to he placed in a close-order grid, which would
reduce the site to “Swiss cheese”.
Figure 11. Four monitoring wells are the minimum required by RCRA.
DIRECT. SAMPLING (DRILLING)
DIRECT SAMPLING (WATER SAMPLING)
Figure 12. SaTnple from drilling and monitor wells is only
representative of the immediate area.
Regional Groundwater Flow
I-
I
23
-------
Area of Target
Figure 13. Ratio of overall site area to target area is often
large. Target area may represent a plume or burial
site. Smaller targets become more c9ifficult to find.
Other Errors Associated with Monitor Wells
In addition to the potential errors introduced from place-
ment and the number of borinas, samples and monitor wells,
factors such as poor quality monitor well construction, improper
sampling or preservation of samples, and imprecise analytical
methods may lead to other errors in site assessrPent. Improperly
sealed well screen intervals may produce unrepresentative
samples, and create cross—contamination problems. Even with
properly installed wells, the collection of representative
water samples is not easily accomplished. Furthermore, the
water chemistry within the monitor well will often change with
time making the result time—dependent. Depending upon the
type of sampling device used and the extent of wel] development,
the chemical parameter measured can present a distorted view
of site conditions. The effort to collect a representative
ground water sample is futile if the chemical composition
changes between the time of collection and the time of analysis,
due to improper sample preservation and storage.
Area of Site As
As
At
24
-------
20
Number of Boreholes
Figure 14.
Probability of detecting a target using a rectangular
grid and randomly located borings. Data for A /At
Patio of in.
— — — —
/
/
/
/
I
/
I
I
Grid Search
I00
80
60
C
0
4-
C )
C.)
4-
C,
0
40
>1
0
0
°20
0
Random
Search
0
I
‘0
I
30
40
50
-------
Errors associated with the location arid sampling of monitor
wells are not well understood and are commonly ignored, and the
resulting interpretations and calculations are based upon
assumptions of ideal conditions. While the resultina calcula-
tions may appear convincing, if they are based upon conditions
which are not representative, they can lead to significant
errors.
Safety and Risk Factors
An important factor in UWS assessment is the risk associ-
ated with drilling for monitor wells arid exploratory holes in
unknown conditions at hazardous waste sites. As the number
of holes needed to define a problem area increases, so does
the possibility of penetrating buried containers or trenches 1
and exposing field crews to toxic fumes and liquids. In
extreme cases 1 the detonation of explosive materials and
fire may result. In addition, there is the risk of cross—
contamination which can result from a drill penetrating a
natural or man—made seal. The hole may then act as a seepaae
route, possibly releasing contaminants into more permeable
soil horizons or fractured rock, arid ultimately into contact
with ground water.
An Alternative Approach——Geophysical Methods
The previous discussion has been centered around the
traditional approach to characterizing subsurface conditions
by drilling arid soil sampling, monitor wells, and water sample
analysis. The general philosophy of the discussion applies to
any direct sampling discrete measurements. In making such
measurements, it has been pointed out, a statistically signifi-
cant number of samples must be obtained to assure a reasonable
level of definition and accuracy The deficiencies of using the
traditional approach alone have recently been identified in
technical seminars arid literature. What is needed is a means
to optimize the approach to site assessment, to maximize the
benefits obtained in exchange for invested dollars, while
reaching a reasonable level of technical accuracy. More cost—
effective reconnaissance techniques are needed, which provide
rapid, continuous spatial coverage, and reduce the risk of
contamination and other hazards associated with conventional
drilling programs alone. Using remote—sensing geophysical
methods in an integrated systems approach is a proven way of
achieving these objectives. Such an approach does not eliminate
drilling and monitor wells——nor can it hope to compensate for
poorly constructed wells——but it does provide us with an
improved understanding of site conditions, and helps us to
place monitor wells in the right locations to be representative
of site conditions.
26
-------
What is Remote Sensing and Geophysics?
Our eyes are remote—sensing devices they produce irnaqes
of the spatial variations of electromagnetic energy within the
visible portion of the spectrum. The term remote sensina is
usually associated with aerial photography or complex computer
images derived from Landsat satellites. However, there are many
other types of remote sensing.
Some forms of remote sensing produce a representative measure—
merit, rather than an image. Temperature, for example, may be
measured directly by a common thermometer or by a remote infrared
sensor. Geophysical measurements are remote—sensing methods,
because by their nature they respond to changes in physical
and/or chemical parameters at a distance.
Geophysical measurement systems cover an extremely wide
range of techniques, applied to such fields as space
exploration, earthquake monitoring, and mineral exploration.
One of the more familiar geophysical methods is the seismic
reflection technique used by all major oil companies for select-
ing sites for exploratory oil wells. An “acoustic” signal is
generated near the surface, and travels thousands of feet into
the subsurface. Reflections of these signals are returned
from various rock interfaces and are recorded to produce a
geologic profile or cross section of the area. An examination
of this cross section then reveals to the trained geologist
the most promising locations to drill for oil, and gas.
Direct and Indirect Measurements
A soil or rock sample from a drill rig may be examined
visually and analytically for physical and chemical properties.
If, for example, the drilling log locates rock at a depth of 10
feet below the surface, it has determined the depth to top of
rock at that specific location. Such an observation is a direct
observation or measurement. Many other strategies are available
for determining depth to rock by less direct means. For example,
a probe could be driven into the ground and the force or number
of hammer blows could be measured. When we hit the rock, we
might expect that the measurement of force or blow counts would
increase, indicating the top of the rock. Such a determination
would be an indirect measurement of the top of the rock. We
have actually measured force or blow count, and have used it as
an indicator of contact with the rock. A geophysical method
would accomplish the same goal by measurement of some physical
or electrical property difference between the soil and rock.
The measurements made by the seismic refraction method,
for example, will yield an indirect measurement of depth to
27
-------
bedrock, which is based on a number of measurements and sub-
sequent calculations. However, the seismic method also
provides us with the seismic wave velocity of the rock (travel
time) which is in fact a direct measurement. Thus the ceo—
physical methods can provide both indirect and direct
measurement of subsurface properties.
The terms “direct measurement” and “indirect measurement”
used in this document are open to some interpretation. The use
of these terms is not intended to he definitive, hut merely a
convenient device to distinguish between two broad categories
of measurements.
In—Situ Measurements
Some change or degradation in sample properties occurs
when a sample is removed from its natural setting. In—situ
measurements provide a means of in—place, non—destructive
measurement and sometimes offer a more reliable mode of measure-
ment than methods which require removal of a sample. Geophysical
techniques provide the capabilities for such in-situ measurement
of various physical and electrical properties under certain
conditions. Besides these benefits, they provide a means of
characterizing site conditions so that the danger of drilling
into unforeseen hazards may be avoided.
Spatial Measurements
It has been previously established that when the number
of borings or monitor wells is limited, results may not be
representative of site conditions, and that in many cases, to
examine site details adequately would have made “Swiss cheese”
of the site. In general, data obtained from borings or monitor
wells comes from discrete depths. Unlike such discrete sarnplinq,
which yields only limited spatial and volumetric information,
geophysical methods measure a much larcer volume of the sub--
surface, thereby increasing the volume sampled for a given
measurement (Figure 15). This larger volume integrates any
variations within the sample, and provides an “average picture”
of subsurface conditions.
This aspect of geophysical measurement has both advantages
and disadvantages. One advantage is that a larger volume of the
subsurface is sampled with each measurement; a disadvantage is
that if a feature or anomaly is small, it may not he detected
in this larger volume. In practice then, there is a trade—off
between the two methods: on the one hand, the possibility of
obtaining better resolution through the use of direct sampling
by drilling (a large number of samples is required); on the
other, the more representative results provided by indirect
sampling with geophysics. When combined in an optimal manner,
28
-------
the two methods complement one another to produce a highly
accurate subsurface investigation. By using geophysical
methods for locating anomalous and non—anomalous zones then
converging on the critical areas with direct sampling, the
survey can proceed rapidly to completion.
Most traditional geophysical techniques assess subsurface
conditions by station measurements; however, some contemporary
techniques can measure subsurface parameters continuously along
survey lines (Figure 16). While theoretically the number of
station measurements could be increased to achieve a density
sufficient to yield the resolution of continuous measurements, to
do so would be impractical in many cases for technical and
economic reasons.
Although the continuous methods referred to in this
document are typically limited to a depth of 15 meters or less,
they are still to be preferred when applicable, as they enable
site coverage to approach 100%. In addition, they offer
significant benefits when applied to sites which are highly
variable because they provide a continuity of subsurface
information which is not practically obtainable from station
measurements. Continuous geophysical methods can be applied at
traverse speeds of 1 to 5 mph, resulting in a cost-effective
approach for relatively shallow survey work. In order to
illustrate the benef its of continuous measurements, a comparison
of station measurements and continuous measurements is discussed
below.
The lower data set in Figure 17 reveals the highly
variable nature of a site as recorded by a continuous spatial
measurement technique. The upper data plot shows the loss of
information and misleading interpretations that can result
from a limited number of station samples and interpolating
between sample points. As can be seen, a limited number of
measurements can result in distorted data. By increasing the
number of station measurements, greater resolution and accuracy
is attained.
Sampling of spatially varying data may be accurately
accomplished by discrete as well as continuous measurements. If
the size of the smallest feature in the data that will he of
interest can be established 1 a survey can be designed to obtain
adequate data from discrete station measurements. To accomplish
this requires that an estimate be made before the survey is
carried out. If our estimate is in error, our data will also be
in error. To minimize the possibility of making such errors, to
achieve maximum resolution, and to minimize project costs,
continuous methods should be employed whenever possible,
particularly when a small sample interval is required or site
conditions are suspected of being highly variable.
29
-------
Volume of Typical
Volume of Drilling
or Water Sampling
Geophysical Measurement
Pigur I .
Sirnp1i€i d comparicon of the volume sampled by
geophysical and drilling methods.
Pigure l . Simplified example of the vo]ume sampled by
continuous geophysical measurement.
3fl
-------
STATION
MEASUREMENTS
DISCRETE SAMPLING VS CONTINUOUS MEASUREMENTS
Figure 17.
Continuous measurement will provide greater
resolution than limited station measurements.
By running closely-spaced parallel survey lines with a
continuous method, changes in subsurface parameters (electrical
conductivity, for example) can be mapped, with the high—
resolution response showing subtle details of site conditions.
Results may be presented in the form of 3—dimensional figures or
isopleth maps (Figures 18 & 19). With detailed data of this
types the confidence level in site assessment is high because the
results are “continuous” and they show details which would he
missed by other approaches. Attempting to obtain this level of
data detail by drilling would have been unrealistic.
The data in Figures 18 & 19 are for the same site, and
not only show where an anomaly occurs, but may also provide
some idea of its size. With this information, the investigator
can converge rapidly upon unusual conditions and proceed to
drill and sample at discrete points in a logical manner
independent of drilling grids or statistical methods. Often,
no more than 3 to 6 direct samplings are needed to obtain a
high—accuracy assessment of a HWS, once it has been characterized
by such geophysical data.
CONTINUOUS
MEASUREMENTS
31
-------
Airborne 1 Surface, and Downhole Methods
Three possible ways of using the remote-sensing methods are
illustrated in Figure 20:
o Airborne or satellite remote sensing clearly has merits, in
terms of spatial coverage per unit time and cost, hut
relatively poor resolution of local details. It provides
little, if any, subsurface data other than that which is
derived by interpretation.
o Surface nethods yield less spatial coverage per unit time
hut can significantly’ improve resolution while providing
subsurface information. A three—dimensional °picture”
can often be generated using special measurement
techniques. An inherent limitation of all surface
geophysical methods is that their resolution (ability to
detect a small feature) decreases with depth.
o Downhole or hole-to—hole methods (lowering various sensors
Three—dimensional perspective view of geophysical
electrical conductivity data from parallel transects
across a hazardous waste site.
Figure 18.
32
-------
Figure 19. Isopleth map of geophysical electrical conductivity
data shown in Figure 18.
A AIRB NE OH SATELLITE 0 SuRFACE REVOlt SENSING
REMOTE SENSING
Figure 20. Three modes of using remote sensing (geophysical)
methods.
down boreholes) will improve vertical resolution over sur-
face methods, but the volume sampled is usually limited to
the area immediate].y around the boring, or that between two
borings, and the cost per unit area is high. However, if
holes are already in place, or if they are to be drilled
for other purposes, the overall cost can he reduced. The
major benefit of downhole methods is that detailed high-
resolution information may be acquired at significant
depths.
All three approaches—-airborne, surface, and downhole——have
a place in subsurface investigation. However, this document will
be limited to a discussion of six selected surface geophysical
methods.
C DOWNWOLE REMOTE SENSING
33
-------
Since each geophysical method measures a different sub-
surface parameter, the information obtained from one sensor is
often complemented by that from another. The synergistic use of
multiple geophysical techniques will often serve to enhance data
interpretations. Those familiar with traditional well logging
will recognize this concept, as multiple logs are commonly
obtained for this purpose.
It should be noted that the performance of any geophysical
technique depends on its specific application and site conditions.
r o single method, therefore, should be expected to solve all site
evaluation problems. Furthermore, geophysical technology is not
in itself a panacea; its successful application is dependent upon
integrating the geophysical data with other sources of information.
This must be done by persons with training and experience in the
methodology, as well as the engineering and earth sciences.
An Integrated Systems Approach
In order to effectively utilize the benefits of both direct
sampling and remote sensing techniques, an integrated systems
approach is needed. The surface geophysical methods are
generally used as reconnaissance tools to cover an area rapidly,
searching for anomalous, conditions. After these areas have been
identified 1 the locations for drillinq and monitor wells can be
selected, with a high degree of confidence in their being
located in the right places to produce a representative sampling
of site conditions. Analyses of soil and water samples from
such wells provide the necessary quantitative measurements of
subsurface parameters. This approach creates much greater
confidence in the final data interpretation with fewer wells and
overall cost savings. ow the drilling operations are no longer
being used for hit—or—miss reconnaissance, but rather as
specific quantitative tools. The geophysical methods make rapid
site coverage possible, followed by prompt convergence on
potential problem areas with direct sampling.
Even if monitor wells have a].ready been installed, geophysical
surveys can still provide significant benefits. The location of
existing monitor wells relative to problem areas can be assessed,
thus providing a means of evaluating the validity of data already
acquired. If additional wells are needed to fill gaps_in the
overall site coverage, they can he precisely placed.
5peC’Ff C Geophysical Niethods
The following sections of this document describe several
surface geophysical methods and their applications. The six
geophysical methods which have been selected for presentation
are: Ground Penetrating Radar, Electromagnetics, Resistivity,
Seismic Refraction, Metal Detection and Magnetometry. It is only
in the past five to ten years that the geophysical methods have
34
-------
been extensively applied to shallow investigations, and only
within the last five years have they been applied to hazardous
waste site assessments. The results have been impressive and
these methods are now rapidly gaining acceptance, often playing a
pivotal role in hazardous waste site investigations.
Geophysical methods themselves have been used for many
decades in the fields of exploration for oil, gas and minerals.
The methods applied to these programs are highly developed and
have been applied with great success for many years. Some geo-
physical methods have been found to be effective in engineering
geology and ground water investigations, and their use has become
widespread in these fields. While the basic concepts of the
methods in all of these applications are similar, unfortunately
the equipment and specifics, as applied to the deeper or larger
oil, gas and mineral deposits, are not necessarily applicable to
hazardous waste site investigations which require shallow, high—
resolution surveys.
On the other hand, the seismic refraction and resistivity
approaches used in the engineering geology and hydrologic fields
are, in fact, directly applicable to hazardous waste sites. The
electromagnetics and ground penetrating radar techniques discussed
in this document are relatively new, and both have been readily
adapted to hazardous waste site work. Metal detecting is not
often considered to be a geophysical technique, although its
principles of equipment operation are similar to the others.
Treasure hunters, the armed services, and public utilities have
used metal detectors to locate treasure, ordnance, and buried
pipes and cables; the metal—detecting equipment and technology
used at HWS has been developed from these applications. While
the magnetometer does see extensive use in the fields of geology
and geophysics, its primary use in the area of concern to this
document is in finding ferrous metal objects: pipe/cable
location, survey stake location, searching for lost aircraft and
sunken ships, and archeology.
There are a number of new geophysical methods, and quite a
few older ones, which in principle may be applicable to hazardous
waste sute investigations. However, this document discusses
only those methods which have met the following criteria:
o they are regularly used for hazardous waste site assess-
ment;
o they have proven capability in hazardous waste site
assessment;
o they are suitable for broad application to the problems
typically found at hazardous waste sites.
35
-------
The six selected geophysical methods are briefly described
below, arid are then described in further detail in sections IV
through IX.
o Ground penetrating radar is a reflection technique using
highfrequency radio waves, which are bounced off subsur-
face features. The picture—like presentation associated
with the radar method is highly useful to evaluate
details of subsurface conditions. Ground penetrating
radar is used to detect natural geoliydrologic conditions
and the presence of both natural and man—made anomalous
conditions. Of the contemporary geophysical methods,
radar is one of the most effective and impressive; it
offers the capability of continuous profiling information
at speeds up to several miles per hour. Its performance,
however, is highly site—specific and is ].imi.ted to
investigation at shallow depths.
o Electromagrietics allows measurement of subsurface
electrical conductivities. Much as the chemist can
measure the specific conductance of a water sample, the
electromagnetic method can measure the conductivity of
the subsurface including the water contained in the
soil and rock. Measurements can be made as station
measurements or as continuous profiling measurements.
Because of the capability of making continuous profile
measurements, the method enables subsurface details to
be mapped effectively. The method provides the means
of mapping contaminant plumes, locating trenches and
buried waste, and identifying buried utility lines. It
is one of the more powerful methods now being applied
at hazardous waste sites.
o Resistivity is a traditional geophysical method by
which measurements of subsurface electrical resistivity
may be made. The method is somewhat analogous to the
electromagnetic method and the data is related. Resis-
tivity measurements must he made by station measurements
and they can provide effective sounding, or vertical
information, as to the depth and thickness of the sub-
surface layers. The method is also effective for profile
measurements in the horizontal plane.
o Seismic refraction is also a traditional method, in
that it has been extensively applied to shallow investi-
gations. The method involves transmission of seismic
waves into the ground, and by measurements of the
travel time of the waves, the thicknesses arid depths of
geological layers can be established. The method can
be applied to the location and definition of burial
36
-------
pits and trenches, as well as to providing information
on the natural geohydrologic setting.
o Metal Detector-—s metal detector will respond to both
ferrous and non—ferrous metal objects. The metal detector
can provide information as to drum location, as well as
tank, pipe, and utility cable locations at or near a waste
site.
o Magnetometry, the magnetic method, as discussed in this
document, applies to the location of buried ferrous
metals such as drums. By detecting anomalies in the
earth’s magnetic field caused by ferrous objects, the
magnetometer provides a means of locating such objects.
The magnetometer will respond only to ferrous metal, such
as iron or steel; it does not respond to non—ferrous
metals, such as copper, lead and brass.
Table 1 summarizes some applications of these six geo—
physical methods to hazardous waste site assessments.
37
-------
SECTION IV
GROUND PENETRATING RADAR (GPR)*
Introduction
Ground penetrating radar (GPR) uses high frequency radio
waves to acquire subsurface information. From a small antenna
which is moved slowly across the surface of the ground, energy
is radiated downward into the subsurface, then reflected hack
to the receiving antenna, where variations in the return signal
are continuously recorded; this produces a continuous cross—
sectional “picture ’ or profile of shallow subsurface conditions.
These responses are caused by radar wave reflections from
interfaces of materials having different electrical properties.
Such reflections are often associated with natural geohydrologic
conditions such as bedding, cementation, moisture and clay
content, voids, fractures, and intrusions, as well as man-made
objects. The radar method has been used at numerous HWS to
evaluate na .ural soil and rock conditions, as well as to detect
buried wastes.
Radar responds to changes in soil and rock conditions.
An interface between two soil or rock layers having suffi-
ciently different electrical properties will show up in the
radar profile. Buried pipes and other discrete objects will
also be detected.
Depth of penetration is highly site—specific, being
dependent upon the properties of the site’s soil and rock.
The method is limited in depth by attenuation, primarily
due to the higher electrical conductivity of subsurface
materials. Generally, better overall penetration is achieved
in dry, sandy or rocky areas; poorer results are obtained in
*GPR has been called by various names: ground piercing radar,
ground probing radar and subsurface impulse radar. It is
also known as an electromagnetic method (which in fact it is);
however, since there are many other methods which are also
electromagnetic, the term GPR has come into common use today,
and will be used herein.
38
-------
moist, clayey or conductive soils. However, many times data
can be obtained from a considerable depth in saturated
materials, if the specific conductance of the pore fluid is
sufficiently low. Radar penetration from one to 10 meters is
common.
The continuous nature of the radar method offers a number
of advantages over some of the other geophysical methods. The
continuous vertical profile produced by radar permits much more
data to be gathered along a traverse, thereby providing a sub-
stantial increase in detail. The high speed of data acquisition
permits many lines to be run across a site, and in some cases,
total site coverage is economically feasible. Reconnaissance
work or coverage of large areas can he accomplished using a
vehicle to tow the radar antenna at speeds up to 8 KPH. Very
high resolution work or work in areas where vehicles cannot
travel can be accomplished by towing the antenna by hand at
much slower speeds. Resolution ranges from centimeters to
several meters depending upon the antenna (frequency) used.
Initial in—field analysis of the data is permitted by
the picture—like quality of the radar results. Despite its
simple graphic format, there are many pitfalls in the use of
radar, and experienced personnel are required for its operation
and for the interpretation of radar data.
Radar has effectively mapped soil layers, depth of bedrock,
buried stream channels, rock fractures, and cavities in natural
settings.
Radar applications to UWS assessments include:
o Evaluation of the natural soil and geologic conditions;
o Location and delineation of buried waste materials,
including both bulk and drummed wastes;
o Location and delineation of contaminant plume areas;
o Location and mapping of buried utilities (both metallic
and non—metallic).
Principles and Equipment
The radar system discussed in this document is a
commercially-available impulse radar system. Continuous
wave (CW) or other impulse systems exist, but they are
generally one of a kind, being experimental instruments,
and are not discussed here.
39
-------
Figure 21 shows a simplified block diagram of a radar
system. The system consists of a control unit, antenna,
graphic recorder and an optional magnetic tape recorder
(Figure 22). In operation, the electronics are typically
mounted in a vehicle. The antenna is connected by a cable
and is mounted or towed behind the vehicle, or may be towed
by hand. System power is usually supplied by a small gasoline
generator. Various antennas may he used with the system to
optimize the survey results for individual site conditions and
specific requirements.
40
-------
GRAPHIC RECORDER
I I
I I
[ I I I I\.
I I
SOIL.
1JIIIROCK !tIl IiTTflT Th
I I I I I I I I I I I I I I II
Figure 21. Block diagram of ground penetrating radar system.
Radar waves are reflected from soil/rock interface.
ANTENNA CONTROLLER
— —— —
GROUND SURFACE
-------
Figure 22.
Photograph of radar system equipment showing four antenna sizes
to right. Larger size antennas are lower frequency.
r
I ’ ,. ’
-------
The impulse radar transmits electromagnetic pulses of short
duration into the ground from a broad—band antenna. The antenna
is usually in close proximity to the surface of the ground.
Pulses radiated from the antenna are reflected from various
interfaces within the subsurface and are picked up by the
receiver section of the antenna. They are then returned to the
control unit for processing and display. The radar data can be
recorded by a graphic recorder and/or a magnetic tape recorder.
The graphic recorder provides a picture—like display of the radar
data (Figure 23). Radar reflections will be returned from any
natural or man—made object which has a contrast in its dielectric
properties. Reflections from deeper targets will appear lower on
the graphic display.
The time the electromagnetic pulse takes to travel from the
antenna to the buried object and back to the antenna is propor-
tional to the depth of the buried interface or object. This time
is called two-way travel time an9 is dependent on the dielectric
properties of the media through which the pulse travels. These
dielectric properties are in turn a complex function of the
composition and moisture content of the subsurface soil and rock
materials. Table 2 shows the range of dielectric values,
velocities and two-way travel times for various natural
materials. In almost all cases, the moisture content has the
greatest influence, because water has a very high relative
dielectric value compared to common soils anl rock. The greater
the amount of water saturation, the lower the radar velocity, as
given by:
Vm
Accordingly, the lower the velocity, the lower the object will
appear in the radar record. Depth is calculated from this
velocity using:
VmT
D= — =
r2
where Vm = velocity in material
C = a constant, the velocity of light (3 x 10 m/sec)
= relative dielectric constant
7 = two—way travel time in nanoseconds
(1 nanosecond (ns) = i0 9 seconds)
Depth of penetration is a function of the radar signal.
attenuation within the subsurface media. This attenuation
consists of electrical losses, scattering losses and spreading
losses. Since spreading losses are inherent in the radar
systems, they are constant and will not be considered further.
43
-------
•1 .— SURFACE
2 METERS
I FINE
QUARTZ
SAND
CL AY
LOAM
Figure 23. Radar record showing irregular clay horizon.
Electrical and scattering losses, however, are highly dependent
on site conditions.
The primary factors controlling electrical attenuation of
radar are the electrical conductivity of the soil/rock system
and the radar frequency. An increase in either subsurface
conductivity or the radar frequency will result in greater
attenuation of the radar signal. The frequency of the radar
may be varied by changing antennas. Unfortunately, the conduc-
tivity of the subsurface cannot be varied. High conductivities
due to dissolved salts from natural sources or contamination
will cause strong attenuation of the radar signal.
An increase in the water content of dry soil or rock can
also increase its electrical conductivity greatly. Similarily,
an increase in clay content will usually increase conductivity.
However, water or clay content alone will not always seriously
degrade radar performance. Experience has shown that penetra-
tions of more than ten meters can be obtained in water—saturated
sands where conductivity is low. Furthermore, the radar method
has been used to profile bottom and sediment conditions through
ice and fresh water.
44
-------
TABLE 2.
APPROXIMATE CONDUCTIVITIES, DIELECTRIC CONSTANTS, AND TRAVEL
TIME FOR VARIOUS EARTH MATERIALS (Modified from
Geophysical Survey Systems, Inc.)
Approximate Dielectric
Constant, r
Two-way Travel Time
Nanoseconds/Meter
(one nanosecond = 10 sec)
Approximate Conductivity
Material O(mho/m)
A
Ui
Air
0
1
6.6
Fresh Water
1O 4 to 3 x i0 2
59
Fresh Water Ice
10—4 to io2
4
13
Permafrost
b . . ’ 5 to 10—2
4 to 11
13 to 15
Granite
1O to10
5.6 to 8
18.7
Dry Sand
io ” to i0
4 to 6
13 to 16
Sand, Saturated
(Fresh Water)
i0 ’ to 102
30
36
Silt, Saturated
(Fresh_Water)
io— to icr 2
10
21
Clay. Saturated
(Fresh Water)
icr’ to 1
8 to 25
18.6 to 23
Average “Dirt”
io ” to 10—2
16
23 to 30
-------
Resolution of the radar profile can be increased by increas-
ing the frequency of the radar. A change in frequency is
accomplished by selecting the appropriate antenna; antennas of
higher frequency and shorter wavelength (500 to 900 M [ -Iz)
provide resolution of a few centimeters 1 but are unable to
penetrate the ground very far, due to increased losses at these
higher frequencies. Lower—frequency antennas (80 to 125 M Hz)
are capable of working to greater depths and of operating in
poor soil conditions, but lack the resolution to define features
smaller than about one meter in size.
Radar reflections from a single interface generally result
in a set of multiple black bands on the graphic display (see
Figure 24). This type of response is inherent in the impulse
method. Generally the location of an interface is picked at one
of the white lines between the black bands. Occasionally, these
multiple bands can obscure information if two interfaces are
close together. If necessary, special processing techniques
originally developed for seismic exploration can he employed to
help alleviate this problem (see processing section).
Factors to be Considered for Field Use
During field operations, the radar system electronics are
usually mounted in a van or other suitable vehicle, with the
antenna towed behind (Figure 25). Vehicle-mounted or towed
configurations can be used to acquire data at speeds up to 8 KPH.
These speeds may be used for reconnaissance surveys, where sub-
surface details are not of interest. This permits much larger
areas to be covered in relatively shorter periods of time.
If there are access problems, the antenna may also be hand-towed
over the site (Figure 26). With cable of sufficient length,
the electronics may be located up to 300 meters from the antenna.
If necessary, extremely high lateral resolution may be
obtained by slowly towing the antenna by hand across the site.
Speeds as slow as 0.5 km/h are commonly used: This allows a
greater number of radar signals to be transmitted and received
per unit distance. At a traverse speed of 1 km/h, radar
sampling density may yield up to 187 samples/meter (at 1 mi/h,
35 samples per foot).
In operation an appropriate time window (range) of the
system is selected. The range is measured in units of nano-
seconds (1 nanosecond = l0 seconds). An estimate of the
radar wave’s travel time (velocity) is made based upon what
is known about site conditions. A time window is then chosen
which will usually provide coverage to a depth which is slightly
greater than the depth of interest.
46
-------
SIGNAL
AMPU fl. E
— -.—o— +
HORIZONTAL
TRAVEL —
RECORDER PRINT
THRESHOLDS
TRANSUITTED PULSE - ‘•‘l•
SURFACE ii{
- , . DE?
OR
— INTERFACE_......*.
S 1GNAL
. 4. ht(k 4 444
W *i4
!!
Figure 24.
1) SKETCH OF A SINGLE
PULSE AND REFLECTIONS
AS SEEN BY THE RECEIVER
2 EXAMPLE OF PROFILE INFORMATION
AS DISPLAYED BY THE
GRAPHIC RECORDER
Example of single radar waveform and resulting
graphic record. (from Geophysical Survey
System, Inc. Manual).
Project requirements and site conditions will dictate which
antenna will be used. Generally the requirement for attaining
adequate penetration depth will be the major factor in
determining the appropriate antenna. Once adequate radar
penetration is achieved, the resolution requirements may
then be considered. Generally results obtained with 250—500
MHz antennas are excellent for delineation of soil horizons,
soil/rock surfaces, soil piping, buried trenches and other
shallow and smaller targets. Attenuation caused by subsurface
conditions may require the use of lower—frequency antennas.
In these cases, the 80 MHz—125 MHz frequency antennas can
be used at the expense of some resolution.
-4
F
TI
i
47
-------
a
4tt .
¾ —
Figure 25.
Vehicle—towed raaar antenna.
-------
Field operations are normally conducted using a single
antenna structure which contains both transmitter and receiver
(monostatic mode).
Multiple antennas may be deployed side by side to cover a
greater area with one pass. Two antennas of different
frequencies may be used to provide optimum benefits from low
and high frequencies. In another configuration, two antennas
may be used in the bistatic mode, one being used as the
transmitter and the other as the receiver. This method can
often minimize unwanted surface noise and can be used to detect
small vertical fractures. While many optional configurations
are possible to help solve particular site problems, the use
of a single antenna (as in Figures 25 and 26) is the most
common and cost—effective approach to most HWS problems.
Quality Control
The radar system measures two—way travel time from the
transmitter antenna to a reflecting surface and back to the
receiver antenna. Calibration of the radar system and data
requires a two—step process:
o First, the total time window (range) set by the
operator must be accurately determined.
o Second 1 the electromagnetic velocity (travel time) of
the local soil/rock condition must be determined.
After completing these two steps, the radar data may then be
calibrated for depths to particular features.
The time window (range) which has been picked for the
survey is calibrated by use of a pulse generator in the field.
This cjenerator is used to produce a series of time marks on the
graphic display, measured in nanoseconds. These pulses are
counted to determine the total time range of the radar (see
Figure 27). A calibration curve can be made up for each radar
system.
In order to precisely relate travel time to actual depth
units, the velocity (or two—way travel time per unit distance)
must be determined for the particular soil or rock found at the
site. Table 2 shows that a wide range of two—way travel times
occurs for natural materials, ranging from 6 to 40 nanoseconds
per meter (ns/m).
Various levels of accuracy in determining travel time can
be used. These may range from first order estimates to
precisely measured on—site values. Often, accurate depth
determination may be relatively unimportant, and only the
relative spatial changes may he of interest. A practical
49
-------
-
:
Figure 26. Hand—towed radar antenna in limited—access area.
rr
U i
0
-
I
-------
first approximation is to use 115 ns/meter (5 ns/ft.) in clean,
dry, unsaturated soils and 30 ns/meter (10 ns/ft.) in silty,
clayey and saturated soils. These two numbers are easy to
remember and lend themselves to quick mental calculations.
More refined estimates can be made if necessary, as one gains
experience at a site.
Using the depth of a known target (trenches, drilling
logs, road cuts or buried pipes/road culverts can provide a
radar target of known depth), a radar record taken over the
known target, and a time scale provided by the pulse generator
will provide information as shown in Figure 22. From these
data a two—way travel time can he accurately determined at the
given target location. While this approach may give accurate
calibration at the specific site, the assumption must be made
that conditions in other areas to be surveyed are the same as
in the calibration areas. If they are riot, errors will occur
in determining depths.
If significant changes in soil type or moisture content
occur with depth, travel time will not he the same throughout
the vertical radar profile, and the vertical radar depth scale
may be non-linear. Such a condition is common, and occurs
whenever an unsaturated zone exists over a saturated zone.
Noise
Sources of unwanted noise which can degrade radar data can
be grouped as follows:
1) System noise;
2) Overhead reflections due to power lines, trees, etc.
(unshielded antennas only);
3) Noise due to surface factors such as ditches, metal,
etc.;
4) Noise due to natural subsurface features or buried
trash;
5) External electromagnetic noise from radio transmitters.
Of these factors, system noise is the most common problem.
Steady—state noise may be introduced by improper cable placement.
Locating antennas too close to the metal vehicle from which they
are towed will also cause noise problems. (Such noise can he
minimized, but not always eliminated, by system adjustments.)
Lower-frequency antennas are not shielded on their top
surfaces and, therefore, receive radar reflections from overhead
objects such as tree branches, power lines, and buildings. Such
a reflection can be identified by an experienced operator by
means of the characteristic signal associated with its very low
two—way travel time in air. Once identified, such signals can he
ignored in the analysis of the data.
51
-------
Feet
15 10 5 .
—10
S •% : W L S- .
—20
—30 t
- C
C
0
‘S
0
—60
f - 70
___ -80
Figure 27. Radar profile over buried pipe. This profile was
obtained for calibration purposes over a pipe of
known 10—foot depth. The time calibration shows that
fifty nanoseconds were required for the radar wave
to travel from the antenna to the pipe and back to
the antenna. This two-way travel time results in
an average 5 nanoseconds/foot travel time or 0.2
feet/nanosecond velocity.
—
52
-------
Surface noise may be generated by pieces of metal lying on
the ground, which can cause a reverberation or ringing of the
radar signal throughout the record. While smaller objects such
as nails do not ordinarily cause problems, an object as small as
a wire coat hanger can create a substantial problem. An effort
should be made to remove such debris from the immediate area of
the radar antenna path.
Small topographic variations may cause some variations in
the data. Crossing a small ditch, for example, can introduce a
band of noise in the data. Radar records acquired in areas
having appreciable clay concentrations at the surface will often
have a smeared or distorted appearance, which may mask useful
information in the data. In addition, some natural geologic
settings will result in apparent noisy data caused by scattering
from a large number of natural boulders. If radio transmitters
are in use nearby, their radiated signal will occasionally cause
significant noise to appear on the graphic recorL
Data Format, Processing, Interpretation and Presentation
The radar data may be generated in three general formats:
1. Individual waveforms (Figure 24);
2. Graphic record (picture—like record, Figure 23);
3. Data storage on magnetic tape media.
The individual radar waveforms may be observed directly on
an oscilloscope in real time. Details of the signal may he
observed and evaluated to select the time window (range) arid
optimize the radar signal. A graphic recorder may be used to
print a copy of the data in the field for quality control and
initial qualitative analysis. This graphic format is typically
used for final display of the radar data. Radar data may be
recorded on magnetic tape or other media. These magnetic
records provide an archive copy of the data, permit the operator
to play back data to optimize data quality, and provide a
signal input to a computer system for processing options.
Various forms of processing may be applied to radar data
to improve its interpretation or presentation. A limited
amount of processing may he done in real time. Time—variable
gain may he applied to radar data, so that a proper amount of
gain is applied to both shallow and deep targets to improve
overall data quality. The graphic recorder may also be adjusted
to improve the visual quality of the data. Analog filtering
of the radar waveforms is possible to eliminate unwanted
high-frequency and/or low—frequency components (noise) which
may obscure useful radar data. The horizontal and vertical
scales of the radar data may be varied to obtain an optimal
visual presentation of the data. The vertical scale of graphic
53
-------
data will usually be much less than the horizontal scale (see
Figure 23). At present, real-time processing can be accom-
plished with a built-in microprocessor to remove steady—state
background noise from the radar profile (see Figure 28).
Replaying radar data which has been recorded in the field
on magnetic tape can provide a number of processing options:
1. Data may be played back in the same manner as described
in real—time processing. The advantage of a post—survey
playback is that various options may be tried to further
optimize the quality of the data.
2. Various analog and digital filtering techniques may be
applied to remove background noise, clutter, or steady-
state systems noise.
3. Discrete waveform analysis methods may be applied to
extract subtle information from the data.
4. Computer processing of the entire profile may include:
a. Averaged waveforms to enhance trends;
b. Deconvolution to remove multiple bands from the
graphic record;
c. Evaluation of the data in the frequency domain.
The process of deconvolution can remove the multiple signal
which is an inherent pattof the radar process. Since these
multiple bands may obscure fine details in the graphic record,
deconvolution can improve resolution in the graphic data and
also aid an inexperienced interpreter.
Individual waveforms (Figure 24) may be analyzed via
computer in order to determine very subtle conditions or details
obscured by larger amplitude or lower-frequency signals. Such
details are often lost in the graphic recorder’s output, due
to the resolution and saturation limits of the recorder. In
addition, processing on digital oscilloscopes or computers can
yield time measurements, thereby permitting accurate deterrnina—
tioris of travel time, or conversely, aiding in the determination
of dielectric constants.
Because of the large volume of information produced by the
radar method, data processing procedures can he time consuming
and may require very specialized computers; they, therefore, are
costly. The processed data may not yield an amount of new
information commensurate with the level of costs incurred. The
essential technical information can often be detected in raw
data by an experienced professional. Many times, processing
algorithms may improve data in the manner desired, but may remove
other information. On the other hand, the improved appearance of
processed data is often useful for presentation to lay personnel
or to publications.
54
-------
I i ifilkilUl
I lIlt I ii liii I,
I II I
hI. I !I :11 :I:: I I Il l::
iIiI)i : i i IIiI Il II 111!
3E cIF.. ?ROC 5SIC liii I:ILIIUI1III IIILIIILIIIII
—
I: 1;
I ill II I j i II 1 11 1 II I I III I I
i i
I,,::
I it, I I I t’II’?I’ I lII’II IIIII III
I I 1 Ip 111111
IIIIiII!iI:fIItIIII! III I III Ilt I:
‘ i i
‘ III II I Iii II 1 11111 1 1
.1.1 ‘till
‘I ll II I Ij I lti lt I p 11 1111 IIIUI 1i 1 1,Ii,iI ii l iIiI I ’ lJ
j illii ii I I I Ij I I ,I 1 I’H’I IIHtIIII:III lIIIIII 1!I:Il
II’ ‘uII I j It Ij II! 1 ( Iij ljii IIIIhI; 1t!:;ltI I1’j’I( IIfIIiiI hItIIh;IIIIh 1 I 11It IIIIlIIII
II II l 1 1 1 1 Iji,
I .1.
I’1I,iiIIIIIIl1IIhIssItiII;: II II
Figure 28. Real—time processing eliminates steady—state noise
bands. Care must be exercised because valuable
may also be removed. (from
geologic information
Geophysical survey systems, Inc.)
AFTER PROC SSI G
c .
7.,.
‘1
t t
.. . . , s. . ’i .’s’L.,.ii. • ..lb ’I fl S
4:
55
-------
The primary reasons for the popularity of the radar method
are its continuous “picture—like” format, high resolution, and
spatial presentation of anomalous features.
Radar data is generally interpreted visually from the
printed graphic record. A rapid qualitative analysis can be
made in the field from this raw data. The continuity of sou l
rock layers and anomalous conditions can often be quickly
located and evaluated. The use of calibration procedures and
correlation to direct data such as drill logs will permit more
quantitative assessment of depth measurement. Relating the
graphic data to the local setting can be accomplished by maricing
the radar record during the survey at regularly spaced station
marks along the traverse route. Variations in lateral traverse
speeds can also be corrected or minimized by referring to
these station marks. The radar data is referenced to the
surface topography, and the interpretation must take this into
account if the topography along a profile line changes.
Radar results are often presented as raw or processed
radar profiles, as printed by the graphic recorder. Schematic
interpretations of these profiles may be made if the site has a
relatively complex setting (Figure 29). In instances where
spatial trends are important, anomalies of interest can be
extracted from the profile data and may be plotted on a map of
the site. This may be accomplished manually or by computer
processing. For example, the presence of burial trenches may be
revealed by several parallel radar profiles across a site; a plan
view of this data will show the location and areal extent of the
trenches (See Figure 30).
Summary
In areas where sufficient ground penetration is achieved,
the radar method provides a powerful assessment tool. Of the
geophysical methods discussed in this document, radar offers the
highest resolution. The method provides continuous spatial sam-
pling and can be carried out very rapidly at traverse speeds from
0.5 to 8 KPH. Its continuous graphic format permits rapid semi—
quantitative interpretation for in—field analysis.
Radar performance is highly site-specific. Depth of pene-
tration is primarily dependent upon soil propert fluids which
influence electrical conductivity. In the wide range of natural
soil/rock conditions found throughout the United States, GPR
penetration varies from less than a meter to more than 30 meters.
Typical maximum penetrations at any given site are 1 to 10
meters.
56
-------
30
60 90 120
HORIZONTAL DISTANCE IN METERS
150 180
(I ,
z
z
0
Figure 29. Interpretation of radar data results in geologic
cross section.
Interpretation of radar data is relatively straightforward
if site conditions are simple and a strong dielectric contrast
exists between the features of interest and the surrounding soil.
As subsurface conditions increase in complexity, interpretation
of the data becomes difficult, and more elaborate interpretation
and processing may he necessary. The high quality of the radar
data shown in Figure 23 is not commonly obtained in the field;
however, experienced interpreters are usually able to cope with
field data of lower quality.
A radar system is a complex instrument. The results of a
radar survey are dependent on many interacting system controls,
various field procedures, site conditions, and interpretation.
Therefore, the successful application of the radar method re-
quires personnel with an understanding of electronics, physics,
57
-------
and basic earth sciences. The more complex the site problem, the
greater the amount of training and experience required.
Capabilities
o The radar method provides continuous data along a
traverse line, producing a picture—like display in real
time.
o Traverse speeds range from 0.5 to 2 km/h for detailed
studies and up to 8 km/h for lower-resolution
reconnaissance work.
o The graphic record can often be interpreted in the
field
o The method provides very high resolution from a few
centimeters to 1 meter, depending upon the frequency
used
o System optimization to local site conditions can he
accomplished by changing antennas (frequency); high
frequency provides the best resolution ; lower
frequency provides deeper penetration.
o P pproximate depths and relative depths are easily
established using simple assumptions arid interpretation
techniques.
o The method may be used in fresh water and through ice
to obtain profiles of depth and sediments.
o A wide variety of processing techniques may be applied
to radar data to aid interpretation and presentation.
Limitations
o Depth of penetration is very site—specific and limited
by the electrical conductivity of pore fluids and clay
minerals.
o Depth of penetration is commonly less than 10 meters.
In extreme soil conditions, effective penetration may he
less than 1 meter.
o Both the instrumentation and technique are sophisticated
arid, therefore, require experienced personnel for
operation.
o Interpretation of raw data may be very difficult under
some conditions.
o Semi—quantitative and quantitative assessments require
considerable care to avoid numerous interpretation
pitfal is.
o Processing of data may be required in some cases;
however, costs will he increased, and processing may
remove some of the important data.
o Depth calibration requires careful on—site work and if
site conditions change the depth calibration wil he
affected. Further, the depth scale is often nonlinear.
o The data can be affected by a variety of sources of
noise.
58
-------
I
Figure 30. Location and boundaries of trenches may he obtained
from parallel radar traverses. Radar is often able
to detect the disturbed soil associated with burial
sites.
Exa mp 1 e s
Radar Assessment of Natural Setting
Figure 31 is an unprocessed radar record of a soil profile
containing three layers in a karst area. The radar range window,
set by the operator, was limited to about 3.5 meters for this
particular survey, although radar penetration exceeded 6 meters
in the area. Clean sand on the surface is under].ain by an
organic/iron—cemented sand (spodic) layer which, in turn, is
underlain by a clay loam (argillic) horizon. The feature in the
upper left corner of the record represents the edge of an ancient
59
-------
0 30 meters
Horizontal Scale
•0
•1
0
0
a
0
0
0
2
.3
edge
Figure 31. Soil profile showing two soil horizons and the
of a Palec sinkhole.
sink-hole which has long since been filled by the surface sands.
No depression or other evidence of this sink—hole feature was
observed at the surface. The significance of this profile is
that it presents an understanding of the soil structure and
irregularities, which will influence the movement of contaminants
or leachate through the shallow ground water system. For
example, in the soil section represented here, a surface spill
or contaminant introduced near the right-hand side of the
record might he expected to follow the clay horizon down dip
to the left. Pockets of contaminant material may he perched
in the low areas in the clay surface. Entering the old sink-
hole area, contaminants could quickly enter the ground water
system via a permeable recharge pathway.
60
-------
Bedrock/Fracture Evaluation
Figure 32 is a radar profile showing a soil profile overlying
granite bedrock. The soil/rock interface is revealed, as well as
the zone of weathered rock. Fractures within the bedrock may also
be located, as a consequence of increased moisture and clay
content relative to the massive granite. Such fractures can
permit the rapid migration of contaminating fluids into the
ground water system. (Radar penetration can he substantial--ten
to thirty meters-—in massive dry igneous rocks.)
Location of Trenches
Figure 33 is a radar profile which was run perpendicular to a
long burial trench known to contain steel drums. The trench
boundaries can be seen in the radar data. Multiple parallel
passes across the trench provided data for mapping the trench
boundaries. While radar could, in fact, detect a single 55—gallon
drum by itself, no discrete drums can be identified in this
particular profile.
Soil - [
Zone of .—.
Weathered
Rock
Rock Outcrop
1 I
[ o 5 meters
I Horizontal Scale ]
Figure 32. Radar profile of granite outcrop showing fracture
zones.
U
4 4
1 ’ Fracture
Zones in
Granite
0
2
-S
3
0
A *
b
-I
(I )
.5
--
61
-------
TRENCH
c1 .. rIr - ‘-. 7. - — i.., •.•.,•%-.
Figure 33. Example of radar traverse over trench.
a)
I-
C)
C)
—2
62
-------
SECTION V
ELECTROMAGNETICS (EM) *
Introduction
The electromagnetic (EM) method provides a means of
measuring the electrical conductivity of subsurface soil,
rock and ground water. Electrical conductivity is a Function
of the type of soil and rock, its porosity, its permeability
and the fluids which fill the pore space. In most cases the
conductivity (specific conductance) of the pore fluids will
dominate the measurement. Accordingly, the EM method is
applicable both to assessment of natural geohydrologic
conditions and to mapping of many types of contaminant plumes.
Additionally, trench boundaries, buried wastes and drums,
as well as metallic utility lines can he located with EM
techniques.
Natural variations in subsurface conductivity may be
caused by changes in soil moisture content, ground water
specific conductance, depth of soil cover over rock, and
thickness of soil and rock layers. Changes in basic soil or
rock types, and structural features such as fractures or
*The term electromagnetic has been used in contemporary
literature as a descriptive term for other geophysical methods,
including GPR and metal detectors which are based on
electromagnetic principles. However, this document will use
electromagnetic (EM) to specifically imply the measurement of
subsurface conductivities by low—frequency electromagnetic
induction. This is in keeping with the traditional use of the
term in the geophysical industry from which the EM methods
originated. While the authors recognize that there are many
electromagnetic systems and manufacturers, the discussion in
this section is based solely on instruments which are
calibrated to read in alectrical conductivity units and
which have been effectively and extensively used at hazardous
waste sites. There is only one manufacturer of such instruments
at the time of this writing.
63
-------
voids may also produce changes in conductivity. Localized deposits
of natural organics, clay, sand, gravel, or saltrich zones will
also affect subsurface conductivi.ty.
Many contaminants will produce an increase in free ion
concentration when introduced into the soil or ground water
systems. This increase over background conductivity enables
detection and mapping of contaminated soil and ground water at
EIWS, landfills and impoundments. Large amounts of organic fluids
such as diesel fuel can displace the normal soil moisture, causing
a decrease in conductivity which may also be mapped, although
this is not commonly done. The mapping of a plume will usually
define the local flow direction of contaminants. Contaminant
migration rates can be established by comparing measurements
taken at different times.
The absolute values of conductivity for geologic materials
(and contaminants) are not necessarily diagnostic in themselves,
but the variations in conductivity, laterally and with depth, are
significant. It is these variations which enable the investigator
to rapidly find anomalous conditions.
Since the EM method does not require ground contact,
measurements may be made quite rapidly. Lateral variations in
conductivity can be detected and mapped by a field technique
called profiling. Profiling measurements may be made to depths
ranging from 0.75 to 60 meters. Instrumentation and field
procedures have been developed recently which make it possible to
obtain continuous EM profiling data to a depth of 15 meters. The
data is recorded using strip chart and magnetic tape recorders.
This continuous measurement allows increased rates of data
acquisition and improved resolution for mapping small geohydrologic
features. Further, recorded data enhanced by computer processing
has proved invaluable in the evaluation of complex hazardous
waste sites. The excellent lateral resolution ohtain& from EM
profiling data has been used o advantage in efforts to outline
closely—spaced burial pits, to reveal the migration of contaminants
into the surrounding soil, or to delineate fracture patterns.
Vertical variations in conductivity can also he detected by
the EM method. A station measurement technique called sounding
is employed for this purpose. Data can be acquired from depths
ranging from 0.75 to 60 meters. This ranqe of depth is achieved
by combining results from a variety of EM instruments, each
requiring different field application techniques. Other EM
systems are capable of sounding to depths of 1000 feet or more,
but have not yet been used at HWS and are not adaptable to
continuous measurements.
64
-------
Profiling is the most cost—effective use of the EM method.
Continuous profiling can be used in many applications to
increase resolution, data density and permit total site
coverage at critical sites.
At UWS, applications of EM can provide:
o Assessment of natural qeohydroloqic conditions;
o Locating and mapping of burial trenches and pit s
containing drums and/or bulk wastes;
o Locating and mapping of plume boundaries;
o Determination of flow direction in both unsaturated
and saturated zones;
o Rate of plume movement by comparing measurements taken
at different times;
o Locating and mapping of utility pipes and cah].es which
may affect other geophysical. measurements, or whose
trench may provide a permeable pathway for contaminant
flow.
Principles and Equipment
Although there is available a wide variety of EM equipment,
most of it is intended for geophysical exploratJ.on of mineral
deposits. These units have not been used at 1-IWS and do not
provide a simple conductivity reading. This document discusses
only those instruments which are designed and calibrated to read
directly in units of conductivity.
The basic principle of operation of the electromagnetic
method is shown in Figure 34. The transmitter coil radiates an
electromagnetic field which induces eddy currents in the earth
below the instrument. Each of these eddy current loops, in turn,
generates a secondary electromagnetic field which is propor-
tional to the magnitude of the current flowing within that
loop. A part of the secondary magnetic field from each loop
is intercepted by the receiver coil and produces an output
voltage which (within limits) is linearly related to subsurface
conductivity. This reading is a bulk measurement of conduc-
tivity; the cumulative response to subsurface conditions
ranging all the way from the surface to the effective depth of
the instrument.
The sampling depth of EM equipment is related to the
instrument’s coil spacing. Instruments with coil spacings of
1, 4, 10, 20 and 40 meters are commercially available: Figures
35, 36, 37 and 38 show several EM units and field configurati.orts.
The nominal sampling depth of an EM system is taken to be
approximately 1.5 times the coil spacing. Accordingly, the
65
-------
fly FIELD
Coil
INDUCED
LOOPS
GROUND SURFACE
SECONDARY FIELDS
FROM CURRENT LOOPS
SENSED BY
RECEIVER COIL
Figure 34. Block diagram s1 owing EM principle of operations.
66
-------
Figure 35.
Small hand—held EM system used in soil survey.
Depth ranqe: 0.75 to 1.5 meters (Denney Farm
Site).
67
-------
Figure 36. Shallow EM system used in continuous record mode.
Depth range: 3 to 6 meters. Man on left is
carrying recorder (Love Canal).
68
‘p
-------
37. Deep EM system used for station measurements. Depth range: 7—1/2
to 60 meters depending upon coil spacing and orientation selected.
t
0
fr
I
— i
\ P 1 lb
I
M
—
S
Figure
-------
— . r I ‘ 1
“ f t
—3
0
‘r
. . ‘. I
.4 -
..F.
r
Figure 38. Truck—mounted EM system provides continuous
to 15 meters depth (Love Canal).
conductivity data
-------
nominal depth of response for the coil spacinas given above is
1.5, 6, 15, 30 end 60 meters.
The conductivity value resultino from an EM instrument
is a conposite, and represents the combiner 1 effects of the
thickness of soil or rock layers, their depths, and the specific
conductivities of the material.s. The instrument readinq
represents the combination of these effects, extendincj from the
surface to the arbitrary depth range of the instrument. The
resultino values are influenced more strongly by sha] low
materials than by deeper layers, and this must he taken into
consideration when interpreting the data. Conductivity
conditions from the surface to the instrument’s nominal depth
range contribute about 75% of the instrument’s response.
However, contributions from highly conductive materials lying
at greater depths may have a significant effect on the reading.
EM instruments are calibrated to read subsurFace conduc-
tivity in millimbos per meter (mm/rn). These units are related
to resistivity units in the followinc manner:
1000/(mil limhos/meter) = 1 ohm-meter
1000/(rni]limhos/meter) = 3.28 ohm—feet
1 millirnho/meter = 1 siemen
The advantage of using millimhos/meter is that the common
range of resistivities from 1 to 1000 ohm—meters is covered by
the range of conductivities from 1000 to 1 millimbos/meter.
This makes conversion of units relatively easy.
Most soil and rock minerals, when dry, have very low
conductivities (Figure 39). On rare occasions, conductive
minerals like nagnetite, graphite and pyrite occur in sufficient
concentrations to greatly increase natural suhsur ace
conductivity. Most often, conductivity is overwhelmingly
influenced by water content and the following soil/rock
parameters:
o The porosity and permeability of the material;
o The extent to which the pore space is saturated
o The concentration of dissolved electrolytes and colloids
in the pore fluids;
o The temperature and phase state (i.e., liquid or ice) of
the pore wat r.
A unique conductivity value cannot he assigned to a particular
material, because the interrelationships of soil composition,
structure and pore fluids are highly variable in nature.
71
-------
Conductivity (millimhos/meter)
tO’ 10’
Clay and Marl / / / // / /
Loom I L l// I I
Top Soil VJLI
Clayey Soils ‘ I A
Sandy Soils ______
Loose Sands k / / / ‘ 1 / /
River Sand and Gravel _____ ___
Glacial Till 1/i//IJ/IIA
Chalk ____
Limestones //f/fJ
Sandstones V i) i / / iii
Basalt yi4 ____ ____
Crystalline Rocks ______________
Figure 3 . Ranq of e1ectrir 1 Cr)n41LIrtivIti ”s in n t-iir i qr ii ci rock.
(Modified 1 ’r C i1 icy ef a! .)
-------
In areas surrounding iWS, contaminants may escape into the
soil and the ground water system. In many ca5es, these fluids
contribute large amounts of e]ectrolytes and colloids to both tl’e
unsaturated and saturated zones. In either case, the around
conductivity may be greatly affected, sometimes increasing by one
to three orders of magnitude above background values. However, if
the natural variations in subsurface conductivity are very low,
contaminant plumes of only 10 to 20 percent above hackqrounc1 may
be mapped.
In the case of spills involving heavy non-polar, organic
fluids such as diesel oiL, the normal soil moisture may he
displaced, or a sizeahie pool of oil may develop at the water
table. In these cases, subsurface conductivities may decrease
causing a negative EM anomaly. (A negative anomaly will occur
only if substantial quantities of non—conductive contaminants
are present.
Factors To Be Considered for Field Use
Profiling-—is accomplished by making fixed—depth FM
measurements along a traverse line (see Figure 40). Profiling
data has traditionally been obtained from discrete station
measurements along the traverse line (see Figure 37); recently,
continuous data has been collected at depths up to 15 meters with
a truck—mounted system (see Figure 38).
Profilina provides an effective means of mapping lateral
changes in subsurface conditions and is the primary EM field
technique. The continuity of the information obtained is
invaluable in resolving details of complex subsurface features
along traverse lines.
Two examples of profiling data are shown in Figure 41.
The first profile (a) shows data plotted from a field log—the
station interval was 30 meters. The second (b) shows the same
survey line, continuously recorded. It can he seen that the
continuously—recorded data provides a more accurate represen-
tation of local variations.
Sounding——is accomp].ished by making conductivity measure-
ments to various depths at a given location (see Figure 42).
EM soundings will provide information on major vertical changes
related to natural conditions or contamination. The method is
generally limited to resolving 2 or 3 soil/rock layers. As
soundings are always accomplished by using station measurements,
more field time and quantitative analysis of the data is required
than with the profiling method.
A number of different field techniques can be used to obtain
sounding information. Within its depth limitation, a single EM
instrument can be used for soundings. Simple qualitative
73
-------
Station 2 3
Measurement
B
A _________ Continuous
Measurement Jf \
____ Surface
Figure 40. Continuous EM measurernerth (A) provifles qreater resolution than
iirnitprl station measurements (R).
-------
(A)
DISCRETE SAMPLING VS CONTINUOUS MEASUREMENTS
Figure 41
Continuous EM measurement: provides qrea er resolution than
li.mited station measurements.
STATION
MEASUREMENTS
(B)
U’
CONTINUOUS
MEASUREMENTS
-------
Figure 42.
EM soundings are obtained by discrete station
measurements. Maximum depth is dependent upon coil
spacing and orientation selected. (for clarity
only three depths shown).
76
60 meters
-------
information can be rapidly obtained with a single instrument by
reorienting the coils 90 degrees. Another instrument (Figure 36)
enables sounding information to he obtained by changing coil
spacing and orientation. Additional information can he obtained
by combining sounding data from several EM instruments having
different depth ranges.
EM instruments are calibrated by the manufacturer to
measure the absolute conductivity over a uniform section of
earth; however, the earth is rarely uniform. For example, in
a layered earth,, each layer may have a different conductivity;
the resulting instrument reading will be some intermediate value
depending on the thickness of each of the. layers, their depths,
and the specific conductivities of the individual materials.
The instrument reading is then the result of the cumulative
contributions of all the layers from the surface to the depth
range of the instrument. A strict solution for this function
would require knowledge of the thicknesses of the layers and
their respective conductivities. Hence, a unique interpretation
of subsurface conditions generally cannot be obtained ¶rom EM
sounding data alone; it must he supported by drillinc data or
other geologic information.
Generally, the most cost—effective approach is to use
profiling to locate anomalous features. Subsequent analysis,
using soundings at selected areas, can assist in a semi-
quantitative depth evaluation of anomalies and background
conditions.
When planning an investigation, careful consideration
should he given to the selection of an EM system to match
site requirements. Some factors which will influence survey
planning are:
o Basic objective(s) of the survey;
o Total area to be covered;
o Depth(s) of profile data needed;
o Site coverage density and resolution requirements;
o Sounding requirements;
o Computer processing requirements;
o Site access;
o Possible cultural interferences which may inhthit
or restrict results.
Final choice of the system(s) to he used wiU he most
dependent on depth and resolution requirements. For example, to
detect a plume from a landfill in a shallow (less than 6 meters)
aquifer, the continuous “6—meter-depth” EM system would he ideal,
offering both the correct depth range and high resolution in the
profiling mode. Generally, spatial coverage using parallel lines
spaced at 15 to 30 meters or more has been adequate for most
landfill evaluations where continuous profiles are possible.
77
-------
Quality Control
EM instruments are calibrated over a massive rock outcrop
used as a geologic “standard” by the manufacturer. After
calibration, the instrument.s will, generally retain their accuracy
for long periods. However, a secondary standard area should he
established by the user for periodic recalibration. On large
projects a local standard site may be established in the field.
This will provide a reference base station, to check “drift” in
the instrument’s performance and to permit correlation between
instruments.
While precision (repeatability) can be easily checked simply
by comparing the instrument to a standard site, accuracy
(closeness to the truth) is much more difficult to establish and
maintain.
EM instruments are often used to obtain relative measure-
ments. For these applications, maintenance of absolute accuracy
is not critical; however, the precision of the instrument can he
important. For example, in the initial mapping of the spatial
extent of a contaminant plume, a moderate level of precision is
necessary. If the same site is to he resurveyed annually to
detect small changes in plume growth and movement, a very high
level of precision is necessary. If the objective of the suri ey
is to obtain quantitative results from the EM data, for
correlation to other measureable parameters (e.g., specific
conductance), the accuracy of the measurement becomes critical.
Under these conditions, proper steps should be taken to assure
good instrument calibration. This is particularly important
when performing surveys in areas of low conductivity, where the
accuracy error can he significant.
The dynamic range of EM instruments varies from 1 to 1000
mm/rn. At the lower concluctjvties, near 1 mm/rn and less, it is
difficult to induce sufficient current in the ground to produce
a detectable response, hence readings may become unreliable.
At conductivity values greater than about 100 mm/rn, the received
signal is no longer linearly proportional to subsurface conduc—
tivites, and corrections must be applied to the data, if it
is to be used for quantitative purposes.
Noise
EM systems are susceptible to signal interference from a
variety of sources, originating both above the ground and below.
Electromagnetic noise may be caused by nearby power lines,
powerful radio transmitters, and atmospheric conditions. At some
sites shallow EM surveys can be carried out in the immediate
vicinity of power lines; at others, conditions may he so had that
measurements are impossible. Generally, deeper measurements
using larger coil spacings will be more susceptible to noise than
shallower measurements. In addition to other forms of electro—
78
-------
magnetic noise, instrument responses from subsurface or surface
metal may make it difficult to obtain a valid measurement. For
instance, piles of drums, nearby vehicles, fences or railroad
tracks can act as targets and produce an unwanted response.
Within a range of 1.5 to 2 times the coil spacing, these large
items may influence the data. Sma].l items of metallic trash
usually create no problem. Buried pipes and cables will cause
very large EM anomalies. However, because of their
characteristic response 1 they can be recognized, and then either
ignored or filtered out of the data. Unfortunately, near such
buried objects, important information of lesser magnitude is
often lost.
EM surveys have been successfully carried out in scrap iron
yards over construction debris fill. The acquisition of a large
amount of data by station measurements and continuous measure-
ments, and the use of special field techniques and computer
processing permitted the location and delineation of contaminated
grounc’ waters. While the total effort was time—consuming and
costly, useful results were obtained under extremely difficult
conditions.
Data Format, Processing, Interpretation and Presentation
EM data can he recorded in the field in several formats:
o Field notebooks;
o Strip chart records;
o Magnetic tape.
EM system output, whetherprofiling or sounding, may be
taken directly from the instrument and recorded in a field ]oq.
Continuous profile data must be recorded on a strip chart
recorder or on magnetic media. When recorders are used some means
of noting survey marks and comments must he provided.
Corrections may be applied to EM data for:
o Accuracy (calibration);
o Drift (precision);
o Spatial variations (due to changes in speed while
recording continuous data);
o Scale changes (necessary to provide adequate
resolution);
o Nonlinearities (associated with high conductivity
values).
These corrections may be applied manually or by computer;
however, raw uncorrected data may he adequate for a given
problem.
79
-------
Profiling Data——A simple profile line nay be drawn from
station field data (Figure 41 1)or a raw strip chart record may
he used (Figure 410), EM profile data is commonly acquired from
a series of parallel traverses across the site and recorded in
the strip chart or magnetic tape format. Because of the large
quantity produced, the data usually must he handled by computer.
The data may be presented as single profile lines, stacked
profile lines, as a three—dimensional, composite view of the data
set, or as anisopleth map. Examples of these different formats
are given in Figures 43, 44, and 45. These figures represent
changes in subsurface conductivities due to varying amounts of
soil moisture related to fractured bedrock. Major trends can be
located in the series of stacked profile lines (Figure 43).
Trends and other details are often better understood through the
use of a three-dimensional perspective plot of the data (Figure 44).
This format gives the viewer a complete graphic picture of the
data at a glance. The isopleth plot (Figure 45) presents these
features in a manner which facilitates accurate location and
determination of size.
Besides handling large amounts of data and creating the
presentation shown above, computer processing can be applied
to achieve a variety of results. For example, filtering may
be applied to remove small unwanted spatial features in order
to emphasize t1 e major characteristics of the plume. In addition,
cultural noise from buried pipes or cables can be removed from
the data to clean up the presentation. Subtle features which
might otherwise have been overlooked have been enhanced and
identified by processing.
Although computer processing is generally applied to
continuous profile data acquired on strip charts or magnetic
tape, it is also applicable to high—density discrete station
measurements.
The most common use of profile data is to locate anomalous
conditions. The spatial relationships of relative values are
noted, enabling the user to locate and follow trends over the
site (see Figures 43, 44 and 45). Drilling sites or other
measurements may then be precisely located. Two profile lines,
run at different effective depths, will provide semi-quantitative
information on the relative conductivities of shallow and deeper
layers. Such information is invaluable in assessing the three-
dimensional nature of site conditions. Contour plots can be
used to accurately determine the spatial extent and direction
of flow, as well as to make an estimate of the magnitude of
contamination. In addition, if complete sets of data are
obtained on two different occasions, the rate of movement can
he established by direct in-situ measurement.
80
-------
SOUTH
P
I I I I U
NORTH
Figure 43. Eleven parallel, continuously recoriec EM profiles.
This format shows the extreme variability of con-
ductivity values across the site anci locates fracture
trends in underlying rock.
1400’
I I I
81
-------
Figure 44. Three—dimensional perspective computer plot f
EM data shown in Figure 43. This format allows
the interpreter to quickly grasp the spatial
and amplitude relationshops in the data. Subtle
as well as major trends are emphasized.
Sounding data——is always acquired from a number of discrete
station readings. A simple qualitative assessment may he
obtained using a single instrument. It is fairly easy to
establish the conductivity of near—surface conditions relative
to deeper conditions.
More quantitative evaluations of the vertical layering of
the subsurface can be obtained from detailed sounding measure-
ments. One or more instruments may be required. Field
data can be compared to calculated conductivities derived from
the EM response equations, using estimated layer parameters.
This iterative process will converge on a model of the vertical
section. This approach will not necessarily yield a unique
solution; however, if good geologic information is available
for the area, a unique solution may be obtained.
NORTH
SOUTH
82
-------
NORTH
Figure 45.
Computer—generated isopleth plot of EM data shown in Figures 43
and 44. Plan view isopleth plots are effective in determining
exact location of features.
SOUTH
-------
The EM sounding method can rarely identify more than 2 or 3
layers with reasonable confidence. The greater the contrast in
the conductivity values of each layer, the better the results.
Often, the more detailed resistivity soundinq method is used to
complement EM profiling data.
The results of a sounding analysis are usually presented
as a vertical section, in which the conductivity layers are
identified as a function of depth. The analyst may he able to
correlate these layers to geohydrologic units believed to exist
at the site (Figure 46).
Summary
Although the EM technique can he used for profiling or
sounding, profiling is the most effective use of the EM method.
Profiling makes possible the rapid mapping of subsurface
conductivity changes, and the location, delineation and
assessment of spatial variables resulting from chariqes in the
natural setting or from many contaminants.
EM is a very effective reconnaissance tool. The use of
qualitative non—recorded data can provide initial interpretation
in the field. If site conditions are complex, the use of a
high—density survey grid, continuously—recording instruments,
and computer processing may be necessary, in order to properly
evaluate subsurface conditions. When continuously—recording
instruments are used, total site coverage is feasible. More
quantitative information can he obtained by using conductivity
data from different depth ranges. At p esent, three different
syster s must be used to acquire data from 0.75 meters to 60
meters. Very often, however, data from two standard depths, e.g.
6 and 15 meters, is adequate to furnish depth information.
Capabilities
o The EM profile method permits rapid data acquisition,
resulting in high-density and high-resolution surveys.
o Profiling data may be acquired from various
discrete depths, ranging from 0.75 meters to 60 meters.
o Continuously—recording instruments (to 15 meter depth)
can increase survey speed, density and resolution
permitting total site coverage, if required.
o EM reads directly in conductivity units (mm/rn)
permitting use of raw data in the field, and correlation
to specific conductance of ground water samples.
o EM can map local arid general changes in the natural
geohydrologic setting.
o EM can detect and measure the boundaries of a
conductivity plume.
o Direction of plume flow can be determined from an EM
conductivity map.
84
-------
Depth in Meters
0 ____________ —Water Table
Cloy
20mm/rn
10 — — — I I I
I F I
I 1 1
I I I
I I I
I I I
F I — Pothbte
tO mm/rn 11 I I Aquifer
20 — — I I
1 I I
i r I
1 I 1
I- I
I I
I. I I
1 1 T
I I ii
I I IL
too mm/rn - Contcminated
30 - _ 1 LJ Aquifer
__ ___ - 1_
Figure 46. Sounding data yields vertical electric section whici
can he related to geohydrologic section.
85
-------
o EM measurements taken at different times can provide the
means to compute movement rates of conservative
contaminants.
o EM can detect and map burial pits and trenches of both
bulk arid drummed wastes.
o EM can detect and map the location of buried metallic
utility lines.
Limitations
o EM has less sounding (vertical) resolution than the
resistivity method, due to its limited number of depth
intervals.
o The acquisition of data from depths of 0.75 to 60 meters
requires the use of three different EM systems.
o Continuous data can be obtained only to depths up to
approximately 15 meters.
o 1 n EM measurement is influenced by the shallower
materials more than the deeper ones; this must be
considered when evaluating the data.
o EM measurements become non—linear in zones of very high
conductivity.
o The EM method is susceptible to noise from a number of
sources, including natural atmospheric noise, powerlines,
radio transmitters, buried metallic trash, pipes, cables,
nearby fences, vehicles and buildings.
Examples
Buried Natural Organics
The understanding of natural geologic/hydrologic conditions
and the location of permeable migration routes in the soi]. and
rock are important considerations in an evaluation of a HWS.
Many constituents of soils, such as natural organic deposits,
have strong sorption properties. Their presence and extent can
be of major concern in evaluating potential migration.
Figure 47 shows the thickness of natural organics (peats)
over an eroded limestone bedrock. Because of the relatively
high conductivity of peat compared to that of limestone, the
conductivity reading was primarily a function of the thickness
of the peat. Three borings were made to the top of rock,
which provided a means of correlating the EM data to peat
thickness. The EM data was then calibrated from the boring
data, and higher values of conductivity could be related to
greater thicknesses of peat. Figure 47 shows the results of the
three borings and the calibrated EM data which represents the
approximate profile thickness of the peats.
The use of EM measurements combined with borings to
86
-------
7
,
1/
/
/
Profile of Thickness of Organics Based on
Three Borings
2
4-
4,
%s_4.
C
-C
4-
0.
0
8
10-
2
4-
4,
C
-C
4-
0.
4,
0
a
tO
Thickness of Organics Based Upon
Continuous Measurements
Figure 47. Continuous EM data (bottom) is ca1ibrate by three
borings (top). Results show thickness of natural
orcianics (Peats).
Position (in feet)
87
-------
calibrate the data enabled a large site to be mapped in a
relatively short time with a reasonable degree of accuracy
and at a much higher level of evaluation of the spatial
variables.
Contamination from Flowing Abandoned Well
Since 1945, water from an artesian well tapping the top of
the Floridari Aquifer had been flowing continuously at a rate of
about 5500 cubic meters per day, allowing water high in total
dissolved solids to spill onto the surface of the ground.
Exploratory monitor wells had been established downqradient
from the source.
Analysis of water samples indicated the presence of elevated
concentrations of chloride (1150 ppm) over the full thickness of
the aquifer (approximately 15 meters) arid about 1.7 kilometers
from the artesian source. The areal and vertical extent of the
plume was measured using EM methods. The EM method was selected
because it could provide very rapid profi]e measurements, with a
reasonable number of stations for the large area (70 square
kilometers) which the plume was suspected to occupy.
This study revealed that the plume was about 12 kilometers
long and two kilometers wide (Figure 48). These results show
that this area of the aquifer is highly permeable, as the plume
had traveled over 12 kilometers in a period of 35 years-—an
average of about 1 meter/day. This is due to the highly porous
nature of the limestone and the general lack of sand ir fi1ling,
allowing polluted water to travel faster with much less filtering
than in other more sandy regimes of the aquifer.
Buried Bulk Wastes and Drums
Many burial sites were believed to exist in a certain area.
The EM technique was selected to provide a rapid reconnaissance
in order to locate possible trenches. Determination of the
extent of contaminant migration into the surrounding soil was
also of interest at this site. Twelve parallel survey lines, 120
meters long, spaced 15 meters apart were established in the
area. The survey lines were oriented approximately perpendicular
to the suspected trenches. These lines were traversed using a
shallow (6 meter) EM system, with its output continuously
recorded on a strip chart. The data was entered into a computer
system for spatial corrections, smoothing and plotting. To
provide perspective, a three—dimensional view of the data set was
developed, as shown in Figure 49.
-------
Figure 48.
Source
EM method was used to map widespread contamination of
ground water caused by free flowing brackish well.
Over 10 square miles have been contaminated.
‘V
9 Mite
89
-------
These results indicate that a large contrast exists
between t} e relatively high conductivities of the waste material
and that of the surrounding natural dry soil. Moreover, analysis
of the processed data indicates that conductivity highs can he
correlated from line to lines revealing the linear extent of a
series of narrow trenches. The data also reveals that the
trenches are relatively close together. The fact that no
obvious high EM values exist in the area surrounding these
trenches indicates that the soil is relatively tight
essentially containing the fluid wastesin the trench area.
Figure 49. Computer plot of EM conductivity data, obtained
over a buried waste site. The linear patterns
of conductivity highs indicate buried trenches.
(One linear trend is shaded to show trench.)
View: NW
90
-------
SECTION VI
RESISTIVITY
Introduction
The resistivity method is used to measure the electrical
resistivity of the geohydrologic section which includes the
soil, rock and ground water. P ccordingly, the method may be
used to assess lateral changes and vertical cross sections of
the natural geohydroloaic settings. In addition, it can he
used to evaluate contaminant plumes and locate rjpt wastes
at hazardous waste sites.
Pkpplication of the method requires that an electrical current
be injected into the ground by a pair of surface electrodes.
The resulting potential field (voltage) is measured at the
surface between a second pair of electrodes. The subsurface
resistivity can be calculated by knowing the electrode separation
and geometry of the electrode positions, applied current, and
measured voltage. (Resistivity is the reciprocal of conduc-
tivity, the parameter directly measured by the EM technique.)
In general, most soil arid rock minerals are electrical.
insulators (highly resistive); hence the flow of current is
conducted primarily through the moisture—filled pore spaces
within the soil and rock. Therefore, the resistivity of soils
and rocks is predominantly controlled by the porosity and
permeability of the system, the amount of pore water, and the
concentration of dissolved solids in the pore water.
The resistivity technique may be used for “profiling” or
“sounding”. Profiling provides a means of mapping lateral
changes in subsurface electrical properties. This field
technique is well suited to the delineation of contaminant
plumes and the detection and location of changes in natural
geohydrologic conditions. Sounding provides a means of
determining the vertical changes in subsurface electrical
properties. Interpretation of sounding data provides the
depth and thickness of subsurface layers having different
resistivities. Commonly up to 4 layers may he resolved with
this technique.
91
-------
Applications of the resistivity method at hazardous waste
sites include:
o Locating and mapping contaminant plumes;
o Establishing direction and rate of flow of contaminant
plumes;
o Defining burial sites by
— locating trenches,
— defining trench boundaries,
— determining the depths of trenches.
o Defining natural geohydrologic conditions such as
— depth to water table or to water—bearing horizons,
— depth to bedrock, thickness of soil, etc.
Principles and Equipment
Most dry mineral components of soil arid rock are highly
resistive except for a few metallic ore minerals. Under most
circumstances, t.he amount of soil/rock moisture dominates the
measurement greatly reducing the resistivity va1u . Current
flow is essentially electrolytic, being conducted by water
contained within pores and cracks. A few minerals like clays
actually contribute to conduction. En general, soils and rocks
become less resistive as:
o Moisture or water content increases;
o Porosity and permeability of the formation increases;
o Dissolved solid and colloid (electrolyte) content
increases;
o Temperature increases (a minor factor, except in areas
of permafrost).
Figure 50 illustrates the range of resistivity found in
commonly—occurring soils and rocks. Very dry sand, gravel or
rock as encountered in arid or semi—arid areas will have very
high resistivity. As the empty pore spaces fill with water,
resistivity will drop. Conversely, the resistivity of earth
materials which occur below the water table but lack pore space
(such as massive granite and limestone) will be relatively high
and will be primarily controlled by current conduction along
cracks and fissures in the formation. Clayey soils and shale
layers generally have low resistivity values, due to their
inherent moisture and clay mineral content. In all cases, an
increase in the electrolyte, total dissolved solids (TDS) or
specific conductance of the system will cause a marked increase
in current conduction and a corresponding drop in resistivity.
This fact makes resistivity an excellent technique for the
detection and mapping of conductive contaminant plumes.
92
-------
Clay and Marl 1/f/i//i
Loam ____
Top Soil
Clayoy Soils ____
Sandy Soils _______
Loose Sands
River Sand and Gravel ______ ____
Glacial Till _______ _______ _____
Chalk ______
Limest ones _______
Sandstones _____ _______
Basalt _____
Crystalline Rocks _____________________
Fiqure 50. Ranqe of resistivit.i.es in c’ommonly—ocrilrrincl soils nr1 rocks.
(Modifir l AFter Culley et al.).
I 10’
Resistivity (ohm-meters)
102 io 4
Ice
1A
LJ
(f//I/I)
( li//I l
V/ I/f
V7_/11(11
vi )J/i Il/I
v/ I
yiii ii / ç I/I
-------
It is important to note that no geologic unit or plume has
a unique or characteristic resistivity value. Its measured
resistivity is dependent on the natural soil and rock present,
the relative amount of moisture, and its specific conductance.
However, the natural resistivity value of a particular formation
or unit may remain within a small range for a given area.
Figure 51 shows typical field equipment for making
resistivity measurements; Figure 52 is a schematic diagram
showing the basic principles of operation. The resistivity
method is inherently limited to station measurements, since
electrodes must be in physical and electrical contact with the
ground. This requirement makes the resistivity method slower
than a non—contract method such as EM.
Many different types of electrode spacing arrays may he used
to make resistivity measurements; the more commonly used include
Weriner, Schiumberger, and dipole—dipole (Ficure 53). Due to its
simple electrical geometry, the Weriner array wil]. he used as an
example in the remainder of this section; however, its use is not
necessariy recommended for all site conditions. The choice of
array will depend upon project objectives and site conditions and
should he made by an experienced geophysicist.
Using the Wenner array (Figure 53), potential electrodes
are centered on a line between the current electrodes; an
equal spacing between electrodes is maintained. These “A”
spacings used during F-IWS evaluation commonly range from 0.3
meter to more than 100 meters. The depth of measurement is
related to the “A” spacing and may vary depending upon the
geohydrology.
Current is injected into the ground by the two outer
electrodes which are connected by cables to a DC or low—frequency
AC current source. (If true DC is used, special non—polarizing
electrodes must be used.) The distribution of current within the
earth is influenced by the relative resistivity of subsurface
features. For example, homogeneous subsurface conditions will
have the uniform current flow distribution shown in Figure 54a
and will yield a resistivity value characteristic of the sampled
section. On the other hand, Figure 54 shows a case where the
electrodes spacing has been increased, an4 c 3 rre j ,. distribution
is pulled downward by a low—resistivity & r 1 €’h n that of the
surface layer, due to the influence of the lower resistivity
material at depth.
The current flow within the subsurface produces an electric
field with lines of equal potential, perpendicular to the lines
of current (Figure 52). The potential field is measured by a
voltmeter at the two inner electrodes.
94
-------
Figure 51.
Typical field setup for resistivity sounding (clay cap at Love Canal).
“A” spacing between electrodes is one meter at beginning of sounding.
j
a
-n
n
‘+0
U ’
. 1
.4
I i ‘/. k l w
r
4
a
k.
a —
- . 4
S
Ar
-------
Current
Voltage
Figure 52. Diagram showing basic
measurement.
concept of resistivity
Apparent resistivity va]ues using the Wenner array are
calculated from the measured voltage and current and the spacing
between electrodes as shown in the following equation:
pa =(2TrA) )(v/I)
where pa
A
V
I
= apparent resistivity (ohm—meters or ohm—feet)
= “A” spacing (meters or feet)
= potential (volts)
= current (ampers)
Current
Source
urrent Meter
C2
/
Current Flow
Through Earth
96
-------
A A A
.—• .-•--• •-•-•---•-• . ...__.__...__.-,
C P P C
Wenner Electrode Arrangement
C P P C
A M 1 N B
Schiumberger Electrode Arrangement
C C P P
I I
I I
1••• - - x..-...-. - -
Axial Bipole - Bipole Electrode Arrangement
Figure 53. Three common electrode arrangements. C - designates
a current electrode. P — designates a potential
electrode.
97
-------
Small Electrode Spacing
Large Electrode Spacing
Figure 54. Increased electrode spacing samples greater depth
and volume of earth.
The apparent resistivities are usually calculated and plotted
as the measurements are made, perrnittina immediate quality
control of the measurement in the field.
For acquisition of relatively shallow data, typically
required at many HWS, the lower—power, self-contained resistivity
units are quite satisfactory. Their transmitter is capable oE
obtaining data to about 50 to 100 meters, using self-contained
rechargeable batteries. Very deep surveys will require
higher—power transmitters usinq generators. Much of the newer
equipment utilizes electronic signal enhancement to improve the
signal-to—noise ratio and to allow measurement of lower voltages.
Cables of specific length to fit the selected “A” soacirig
are advantageous for extensive profiling surveys to speed data
acquisition. Sounding measurements require a wide range of cable
lengths; therefore, wire on reels is normally used.
Steel stakes are commonly used for electrodes and are driven
into the ground to a depth of about 10 to 30 centimeters. Longer
or multiple electrodes may be needed in dry sandy areas to
provide better electrical contact with the ground. Water or
salty water is also used to increase the effective electrode
contact with dry soils.
Factors to be Considered for Field Use
Profiling——Profiling is the technique of making resistivity
measurements with a fixed electrode spacing. Electrode “A”
spacing should be 1 to 2 times the depth of interest. The
fixed—spacing electrode array is moved to a number of different
2. Low
98
-------
STATION
0
STATION 2
•1+
STATION 3
D
4.
STATION 4
D
3
Figure 55.
Profile measnrements are accompitshe by fixinq the electrode spacing
and movinq the entire array. The dist tnce between stations, 1), is
dictated by the lateral resolution desired.
-------
locations to obtain data over the entire area of interest (Figure
55). Since depth of influence remains constant from one station
to the next, profiling measures lateral changes in resistivity.
Such changes permit the detection and mapping of anomalous
spatial features over the area surveyed. The method may be
modified to include measurements at more than one depth, thereby
providing additional information on lateral variations with
depth.
By making many profiling measurements along a traverse,
a profile of subsurface resistivity may be obtained (Figure
56). Multiple profile lines or numerous profile stations may be
placed at intervals across the area of interest to generate a
contour map (Figure 57). The lateral resolution requirements
of the investigation will dictate the distance between successive
profiling stations.
Sounding——The sounding technique measures vertical changes
in the geologic section. A series of resistivity measurements is
made, each with successively larger electrode spacings (Figure
54). As the “A” spacing is increased, the depth of sampling at
the sounding station also increases. The maximum •A” spacing
should he at least 3 to 4 times the depth of interest in order to
permit adequate characterization of deeper layers. Therefore,
the overall array length including current electrodes will be 9
to 12 times the depth of interest. With such long arrays, the
operator must insure that adequate space is available at the site
and that it is relatively clear of buried pipes and fences.
Successive electrode spacings should be equally spaced on a
logarithmic scale with a minimum of three per decade, although
six are recommended. Commonly, more measurements are used to
evaluate noise and provide reasonable quality in the data. For a
typical sounding, 12 to 16 separate measurements may he made over
an “A” spacing range of 0.3 meter to 100 meters.
The resulting data is plotted on log/log graph paper with
apparent resistivity versus electrode “A” spacing (Figure 58).
This graph can be visually interpreted for qualitative trends, or
compared to master curves to determine layer thicknesses, depths
and true resistivities. Computer processing may be applied to
achieve quantitative results, as obtained by master curves or to
analyze more complex data.
Although resistivity sounding methods are intended for use
in uniformly layered geological conditions, useful data may often
be obtained from the complex subsurface conditions often found at
HWS.
With both profiling and sounding techniques, inhomogene-
ities in the near—surface soils may induce noise in the data.
100
-------
Some surfac conditions may limit or preclude use of the
resistivity method. Dry surface material having extremely
high resistivity will make injection of the current difficult
and require special field procedures. In areas with paved
surfaces such as asphalt and concrete roads or parking lots,
electrode contact may not he possible.
0
100
HORIZONTAL DISTANCE. IN METERS
700 300 400
500
Figure 56.
Resistivity profile across glacial clays and gravels.
(from Zohdy, 1964).
Survey objectives will determine whether profiling or
sounding data is required. For example, profiling should he
used for mapping contaminant plumes. Because profiling is a
faster field technique, a larger number of stations may be
occupied with the higher density providing better lateral
resolution. The selection of the proper “A” spacing for the
profiling survey may be determined from several initial sound-
ings in the area of the suspected plume.
0
200
P00
101
-------
A SPACING
STATION SPACING 10 METERS
= 10 METERS ISOPLETH INTERVAL
20 OHM METERS
STATION
LOCATION
Figure 57.
Isopleth resistivity map of profiling data.
A potential safety hazard exists during the operation of
the resistivity unit; substantial amounts of current and voltage
are present at the current electrodes during the time the
transmitter is energized. Field procedures must he designed
to insure that none of the field crew are in contact with the
electrodes during this period. An experienced field crew will
not have problems, but if persons unfamiliar with the techniques
are involved with the field operations, additional caution
should be exercised.
Quality Control
Considering the length of the wire cables, their connections
to the stakes and the stakes’ contact with the ground, there are
a number of possibilities for poor electrical contact and sl- ’ort
circuits in the resistivity array. These conditions can he
monitored by observing instrument readings and trends in the
data.
.
.
I
I
102
-------
(I)
LU
I —
LU
I
>-
I—
>
I-
U)
U)
0 W
(*) a:
I—
z
LU
a:
0
1000
500
200
100
50
20
I0
I 2 5 10 20 50 100 200
ELECTRODE “A” SPACING (IN METERS)
500 1000
Figure 58.
Resistivity souruling curve showinq two—layer system.
-------
Apparent resistivities should be calculated and plotted
during field acquisition as a means of quality control. Sounding
curves should he smooth, and jumps in the data should not occur
(Figure 58). Note that the one circled data point at 80 foot “A”
spacing lies off the main trend of the data and hence is
disregarded in the plot and analysis. Profiling data should also
show a general trend in the data from one station to the next
(Figure 57). Note that there is a general trend in the data from
one point to the next which provides confidence in the quality of
the data, but the circled data point to the upper right lies off
the main trend. This point will he ignored in the final analysis
as it is caused by some source of noise at the specific station.
However, abrupt changes do commonly occur in both sounding and
profiling data; these may be unwanted “noise” due to near—surface
inhomogeneities or electrode contact problems or, in the case of a
profile line, may indicate a real change in geohydrology.
Experienced interpreters can often evaluate these problems in the
field and take corrective action if it is required.
The resistivity instrument can be calibrated using standard
resistors. Calibration is particularly important if the data is
to be compared to resistivity measurements from other instruments
or other parameters, such as specific conductance of water
samples.
Noise
Noise from several sources may affect resistivity
measurements.
Equipment—related noise may occur due to coupling between
wires or between reels of long cable arrays. Poor electrical
contact between the ground and electrodes will also produce
noisy data. Exceeding the depth capability (power and receiver
sensitivity) of the resistivity instrumentation will also yield
poor data at very large electrode spacings. In most cases,
however, experienced field personnel will he able to mitigate
such problems.
Cultural noise caused by stray currents, potential fields
and electromagnetic energy can interfere with the resistivity
measurement. This interference can be caused by nearby power
lines and man—induced ground currents. The influence of nearby
fences, railroad tracks and buried metallic pipes and cables
can “short” or strongly distort current flow. These effects
of proximity to metallic structures must be evaluated by
experienced personnel.
Natural sources of electrical noise include earth currents
and spontaneous potential (SP). Most modern instruments are
designed to cope with such noise problems.
104
-------
Poor electrode contact with the earth, and local variations
in shallow subsurface conditions near the electrodes can produce
significant scatter in the data. Decreasing the spacing between
stations, using appropriate field arrays and usinq averaging
.echniques can minimize the influence of these variations.
Data Format, Processing, Interpretation and Presentation
The resistivity instrument measures the voltage and current
between the electrodes. This value is converted to apparent
resistivity, using the proper equation for the type of array
being used and for the specific el.ectrode spacing. (A hand
calculator will speed this calculation.) This value is then
logged and usually plotted on a graph in the field for quality
control. Additional steps to he taken in processing and
interpretation depend upon the application of the data, and
whether data was obtained from profiling or sounding. The flow
diagram in Figure 59 outlines typical steps in the processing and
interpretation of resistivity data as discussed below.
Profiling——The calculated apparent resistivity values from
many profile stations can be plotted as profile lines (Figure
56). A contour map (Figure 57) can be developed from many
profile stations. These stations can he along a straight line
or randomly located over the area. Profile lines and contour
maps can then he used to locate geologic variations or
contaminant plumes. The apparent resistivity values are
typically used because the primary objective is to use the
data for location purposes.
Sounding——Calculated apparent resistivities are plotted
against “A” spacing on log/log graph paper for each station
(Figure 58). These data points define a curve which can he used
qualitativeLy and quantitatively to determine vertical changes in
resistivity. Relative trends and semi—quantitative analysis are
often immediately obvious to the experienced interpreter from
such a plot.
Subsequent analysis requires analytical techniques which
provide a means of modeling. Two approaches or a combination may
be employed. A Forward Model produces a resistivity sounding
curve from a specified geologic section whose resistivities
are known. The analysis can be carried out by making an estimate
of the geoelectric properties and calculating a forward model
(sounding curve). These results are then compared to the
field data. Iteration of the above process occurs until a
reasonable match between the model and the field data is found.
An Inverse Model provides the geoelectric cross section from
the field data; in this case the solution is not unique in
that a number of possible combinations exist that will fit
the field data. In this case, knowledge of the geohydrologic
105
-------
Flow diagram showing steps in processing and interpretation of
resistivity data.
Profile Line
Profile Measurements
Voltage Current
Electrode Spacing
I-
0
a.’
Sounding Measurements
Voltage Current
Electrode Spacing
,
/
/
/
/
“A” Spacing
— ___
Figure 59.
-------
sections must be employed to select the most likely situation.
Both of these methods require a small computer (the forward
method can be carried out on a small hand—held programmable
calculator).
The computer programs offer a more effective means of
processing, by allowing considerable interaction and iteration.
The computer is used only as a processing tool; it is not a means
of obtaining a final answer without interaction.
Another approach is to use computer—generated master curves
which are compared to the field data. The use of master curves
can produce quantitative depth information and true resistivities
for up to three— to four—layer simple geologic systems. A master
curve for a uniform two-layer geologic section is shown in Figure
60. This curve covers both high resistivities under]ain by lower
values (lower half of the curves) and low surface values underlain
by high resistivity values (upper half of the set of curves).
A number of ratios between the two—layer resistivtties are
provided which make up the family of two—layer curves. Often
the field data will not fit the set of master curves exactly
and interpolation and extrapolation must be carried out.
Although simple geologic conditions may he easily interpreted,
generally the use of master curves and computer analysis
requires considerable knowledge of the overall methodology
and geohydrology.
A number of shortcut interpretation procedures exist
such as Barnes layer and cumulative methods. These methods
are commonly employed because of their simplicity, and because
they avoid the use of the more complicated procedures of master
curves or computer processing. It should be recognized that
these shortcut methods are only approximations and may provide
erroneous data under some conditions; therefore their use is
not generally recommended.
The resistivity sounding results will indicate the
number of geologic layers present as well as their depth and
thickness, but only those layers that are sufficiently thick
and have adequate contrast in their electrical properties
will be detected. Data from a number of soundings can he
used to create a cross section plot of resistivities (Figure
61), called a geoelectric section. This figure shows a two—
dimensional representation of the data. Single soundings can
be represented in a similar manner to provide one—dimensional
data. If soundings are available over a large area or along
somewhat perpendicular traverse lines, the resulting data can
be shown as a three-dimensional section or as a fence diagram
(Figure 62)
107
-------
1 1 1=
40
A
A
Q
20- - —
7- 4-
5.
20
05 -
04
03-
‘32 -
40 -
23
‘5
I C
7
5
25
20
I 25
Io
010
C C ?
Figure 60. Two—layer master curves used to interpret Wenner
sounding data. (from Orellana — Mooney master
curves for V.E.s.).
108
065
015 -
0 5
0 25 -
005
TWO-LAYER CURVES
-------
Sumxriary
The resistivity method provides a means of measuring one
of the electrical properties o the geohydrologic section
including soil 1 rock and ground water. These measurements may
he used to assess latera]. changes and vertical cross sections
of the natural geohydrologic settings. Since the resistivity
of soils and rocks is predominantly controlled by porosity,
permeability, amount of water, and concentration of dissolved
solids in the water, the method provides a tool to evaluate
contaminant plumes arid to locate buried wastes at hazardous
waste sites.
The resistivity technique may he used for “profilinq’ or
“sounding”. Profiling provides a means of mapping lateral changes
in subsurface electrical properties. This field technique is
well suited to the delineation of contami.nant plumes and to
the detecti.on and location of changes in natural geohydrologic
conditions. Profile ].ines and contour maps can he used to
locate geologic variations or contaminant plumes. The apparent
resistivity values are typically used, because the primary
objective is to use the data for location purposes. Relative
trends and semi—quantitative analyses are often irnmerliately
obvious to the experienced interpreter from a plot of sounding
data.
Sounding provides a means of determining the vertical
changes in subsurface electrical properties. Interpretation
of sounding data provides the depth and thickness of subsurface
layers having different resistivities. Commonly, 3 to 4 layers
may be resolved with this technique. The resistivity sounding
method is in general a more effective method than the EM sound-
ing method described. The analysis of resistivity sounding
data requires that the interpreter he knowledgeable about tile
resistivity method, the conditions under which the data were
obtained, the geohydrologic conditions, as well as the specific
techniques, computer models 1 or curve matching.
The operator must insure that adequate space is available
at the site and that it is relatively clear of buried pipes
and fences. Finding sufficient space for a long profile array
with an overall length three to six times the depth of interest 1
or a sounding array with an overall length nine to twelve times
times the depth of interest can sometimes he a problem.
Although resistivity sounding methods are primarily
intended for use in uniformly layered geological conditions,
useful data may be obtained from the complex subsurface condi-
tions often found at HWS. With both profiling and sounding
techniques, irihomogeneities in the near—surface soils may
introduce noise in the data. Some surface conditions such as
log
-------
C - )
B
4OGOOHS-CdttER$
LT AND CLAY
9Q-1200flS-MCtENS
GRAVELLY CLZV
300-4200HU-SCt(RS
SMC M S GRAVEL
Figure 61.
Figure 62.
¶ Sounding Location
Geoelectric cross section derived from seven
resistivity soundings (from Zohdy 1 1964).
A’
“ I
MA
MA
MA
C
-a
50 c
A three-dimensional or fence diagram may be constructed
from multiple resistivity soundings 4
U ,
4
0
EXPL &NATiON
0 50 100 M(T(RS
I 1 I - 1
V
Loeclion of Scundng
110
-------
dry surface materials, concrete roads or parking lots may
preclude the use of the resistivity method.
The resistivity method is inherently limited to station
measurements, since electrodes must he in physical and electrical
contact with the ground. This requirement makes the resistivity
method slower than a non—contact method such as EM.
Capabilities
o Resistivity profiling techniques can be used to detect
and map contaminant plumes and changes in geohydrology.
o Resistivity sounding me.thods can estimate the depth,
thickness and resistivity of subsurface layers, or depth
to the water table.
o Both profiling and sounding data can be evaluated
qualitatively or semi—quantitatively in the field.
o Resistivity values can he used to identify the probable
geologic composition of a layer or to estimate the
specific conductance of a plume
o Depth to bottom of landfills and large burial sites can
sometimes be estimated.
Limitations
o The sounding technique requires that site conditions he
relatively homogeneous laterally.
o The method is susceptible to noise caused by nearby
fences, pipes and geologic scatter, which may interfere
with usefulness of the data.
o Quantitative interpretation requires the use of master
curves and/or computer programs, and experience in their
use.
Examples
Determining Depth and Thickness of a Clay Layer for a Proposed
Disposal Site
The depth and thickness of a clay layer was required in an
assessment of natural site conditions. Several resistivity
soundings were conducted over the area of interest, using “A”
spacings from 1 to 300 feet (0.3 to 100 meters).
Figure 63 shows one of the sounding curves. A visual,
in-the—field qualitative evaluation of the curve indi.cates that
three, and possibly four, resistivity layers are present. The
relative resistivities are
(1) very high at the surface,
111
-------
(2) extremely low near an “A” spacing of 70 to 100 feet (21
to 30 meters),and
(3) increasing beyond 100 foot (30 meters) “A” spacing.
A general knowledge of the local, geology suggested that
these layers were dry quartz sand (which was o1 served at the
surface), massive clay, and ].imestone bedrock. An inter-
mediate layer, possibly a sand—to-clay transition, existed
between the high—resistivity surface layer and the low—
resistivity layer.
Computer analysis using an inverse resistivity model
provided th results shown in Figure 64. An intermediate layer
was identified as a transition zone of clayey sand between the
surface sand and the top of the clay.
Mapping of Landfill. Leachate Plume
A leachate plume was known to exist at a large landfill,
based on samples from existing monitor wells (Figure 65).
However, its lateral extent was unknown, as was its maximum
distance from the landfill. The landfill was situated in an
unconfined limestone aquifer with a shallow water table.
Several soundings were made initially in the area of the land-
fill to determine the approximate depth of the leachate plume
within the aquifer. It was found to lie between the surface
and a depth of 20 meters. Electrode “A” spacinqs of 5 and 15
meters were selected to map lateral changes in resistivity
around the landfill. The 5—meter spacing would indicate the
existence of a shallow plume and provide a measure of the
variations due to shallow geohydrologic conditions. The 15-
meter spacing was selected to provide a reasonable average
measure of the main core of the plume. The area was somewhat
developed, which made the location of long profile lines
difficult; therefore, profiling stations were placed wherever
sufficient space was available. The data for each station was
plotted and contoured in map form (Figure 66). A large plume
extending two kilometers downgradient from the site was mapped.
The shallow data shows considerable variation in the plume,
due to the influence of many near—surface variables. The
deeper data shows a more uniform plume pattern, as the plume
is less influenced by the surface variables.
Existing monitor well data was evaluated based upon this
new information and additional monitor well sites were selected
to provide a more extended chemical analysis of the plume.
112
-------
I-
I’
I
0
>.
I-
>
I-
(I )
U)
‘ Ii
I-
U i
a.
a.
I - .
I - . .
(A l
to
5 10
A SPACING
(IN FEET)
Figure 63.
Field sounding curve over a four—layer geoloqic section.
-------
Resistivity
Sounding Drillhole
__________ _______ _________ — —0 FEET
Gray Sand
13,000
Tan Sand
Sand
____________ — Tan Sand w/Clay
?6500 Red Sandy Clay —10
Clcyey Sand
Dark Tan
Sandy Clay
Tan Plastic Clay
18 Gray Plastic Clay
Massive Clay
—30
End of Hole
?:300 —4°
Limestone
Figure 64. Correlation of resistivity sounding results to
a driller’s log. (Resistivity values are in ohni—ft).
114
-------
DISPOSAL SITE
A’
SPECIFIC CONDUCTANCE
(uJ mho/cm)
MONITOR WELL LOCATION
Figure 65.
Cross section of leachate plume based upon specific conductance
from 1974 well data (‘ocations shown in Fioure 66).
A
V
V V V
SEA
LEVEL
40’ .
60 ’.
REGIONAL
GROUND
WATER
FLOW
r
1o0,
-------
A
Figure 66.
Isopleths of resistivity profiling data showing
extent of landfill plume. Values in ohm—feet.
SHALLOW MEASUREMENTS
OF POLLLUTANT PLUME
(0-15’)
340)
34
DEEP MEASUREMENTS
OF POLLUTANT PLUME
(0-45’)
116
-------
SECTION VII
SEISMIC REFRACTIO1’
Introduction
Seismic refraction techniques are used to determine the
thickness and depth of geologic layers and the travel time or
velocity of seismic waves within the layers. Seismic refraction
methods are often used to map depths to specific horizons such as
bedrock, clay layers, and water table. In addition to mapping
natural features, other secondary applications of the seismic
method include the location and definition of burial pits and
trenches at HWS.
Seismic waves transmitted into the subsurface travel at
different velocities in various types of soil, and rock and are
refracted (or bent) at the interfaces between layers. This
refraction affects their path of travel. An array of geophones
on the surface measures the travel time of the seismic waves from
the source to the geophones at a number of spacings. The time
required for the wave to coniplete this path is measured,
permitting a determination to he made of the number of layers,
the thicknesses of the layers and their depths, as well as the
seismic velocity of each layer. The wave velocity in each layer
is directly related to its material properties such as density
and hardness.
A seismic source, geophones, and a seismograph are required
to make the measurements. The seismic source may he a simple
sledge hammer with which to strike the ground. Explosives arid
any other seismic sources may be utilized for deeper or special
applications. Geophones implanted in the surface of the ground
translate the received vibrations of seismic energy into an
electrical signal. This signal is displayed on the seismograph,
permitting measurement of the arrival time of the seismic wave.
Since the seismic method measures small ground vibrations, it. is
inherently susceptible to vibration noise from a variety o
natural and cultural sources.
At HWS, seismic refraction can be used to define natural
geohydrologic conditions, including thickness and depth of soil
and rock layers, their composition and physical properties, and
117
-------
depth to bedrock or water table. It can also he used for the
detection and location of anomalous features, such s pits and
trenches, and for evaluation of the depth of burial sites or
landfills. (In contrast to seismic refraction, the reflection
technique, which is common in petroleum exploration, has not
been applied to HWS. This is primarily because the method
cannot be effectively utilized at depths of less than 20
meters.)
Principles and Equipment
Although a number of elastic waves are inherently associ-
ated with the method, conventional seismic refraction methods
that have been employed at HWS are concerned only with the
compressional wave (primary or P—wave). The compressional
wave is also the first to arrive which makes its identification
relatively easy.
These waves move through subsurface layers. The density
of a layer and its elastic properties determine the speed or
velocity at which the seismic wave will travel through the
layer. The porosity, mineral composition, and water content of
the layer affect both its density and elasticity. Table 3
lists a range of compressional wave velocities in common
geologic materials. It can be seen from these tables that the
seismic velocities for different types of soil and rock overlap,
so knowing the velocities of these layers alone does not permit
a unique determination of their composition. However, if this
knowledge is combined with geologic information, it can he used
intelligently to identify geoloqic strata.
In general, velocity values are greater for:
o dense rocks than light rocks.
o older rocks than younger rocks.
o igneous rocks than sedimentary rocks.
o solid rocks than rocks with cracks or fractures.
o unweathered rocks than weathered rocks.
o consolidated sediments than unconsolidated sediments.
o water—saturated unconsolidated sediments than dry
unconsolidated sediments.
o wet soils than dry soils.
Figure 67 shows a schematic view of a 12—channel seismic
system in use and the compressional waves traveling through a
two-layered system of soil over bedrock. A seismic source
produces seismic waves which travel in all directions into the
ground The seismic refraction method, however, is concerned only
with the waves shown in Figure 67. One of these waves, the
118
-------
TABLE 3. RANGE OF VELOCITIES FOR COMPRESSIONAL WAVFS IN SOIL AND ROCK
(After -jakosicy, 1950)
Material Vetocity (meters/sec)
Weathered surface material 305 - 610
Gravel or dry sand 465 - 915
Sand (Wet) 610 —1,830
Sandstone 1,830 - 3,970
I — .
‘ 0
Shale 2,750 —4,270
Chalk 1,830 — 3,970
Limestone 2,140 - 6. 100
Salt 4,270 — 5,190
Granite 4,380 — 5,800
Metamorphic rocks 3,050 - 7,020
-------
Figure 67. Field layout of a 12—channel seismoqraph showing the path of direct and
refracted seismic waves in a two—layer soi 1./rock system.
Seismograph
I ’ . ,
0
-------
direct wave, travels parallel to the surface of the ground.
A seismic sensor (geophone) detects the direct wave as it moves
along the surface layer. The time of travel along this path is
related to the distance between the sensor arid the source and
the material composing the layer.
If a denser layer with a higher velocity, such as bedrock,
exists below the surface soils, some of the seismic waves will
be bent or refracted as they enter the bedrock. This phenomenon
is similar to the refraction of light rays when light passes
from air into water and is described by Snell’s law. One of
these refracted waves, crossing the interface at a critical angle,
will move parallel to the top of the bedrock at the higher
velocity of the bedrock. The seismic wave travelling along
this interface will continually release energy back into the
upper layer by refraction. These waves may then he detected in
the surface at various distances from the source (Figure 67).
Beyond a certain distance (called the critical di.stance),
the refracted wave will arrive at a geophone before the direct
wave. This happens even though the refraction path is longer,
because a sufficient portion of the wave’s path occurs in the
higher velocity bedrock. Measurement of these first arrival
times arid their distances from the source permits calculation
of layer velocities, thicknesses and bedrock depth. Application
of the seismic method is generally limited to resolving three
to four layers.
The preceding concepts are based upon the fundamental
assumptions that:
1. Seismic velocities of geologic laye.rs must increase
with depth. This requirement is generally met at most
sites.
2. Layers must he of sufficient thickness to permit
detection.
3. Seismic velocities of layers must he sufficiently
different to permit resolution of individual layers.
There is no way to establish from the seismic data alone whether
a hidden layer (due to 1 & 2 above) is present; therefore,
correlation to a boring J.og or geologic knowledge of the site
must be used to provide a cross check. IE such data is not
available, the interpreter must take this into consideration in
evaluating the data.
Variations in the thickness of the shallow soil zone,
inhomogeneities within a layer, or irregularities between layers
will often produce geologic scatter or anomalies in the data.
This data scatter is useful information, revealing some of the
natural variability of the site. For example, a zone containing a
number of large boulders in a glacial till deposit will yield
121
-------
inconsistent arrival times, due to variable seismic velocities
between the boulders and the clay matrix. An extremely
irregular bedrock surface as is often encountered i.n karst
limestone terrain, likewise, will produce scatter in the seismic
data.
The seismic refraction technique uses the equipment shown
in Figure 68. The seismic source is often a simple ten—pound
sledge hammer or drop weight which strikes the ground, generat-
ing a seismic impulse. Explosives and a variety of other
excitation sources are also used for the greater energy levels
required for information at deeper layers.
Seismic waves are detected by geophones implanted in the
surface of the ground at various distances from the source.
The geophone converts the seismic waves mechanical vibration
into an electrical signal in a manner similar to that of a
microphone. This signal is carried by cable to the seismograph.
The seismograph is an instrument which electronically
amplifies and then displays the received seismic signal from
the geophone. The display may he a cathode ray tube, a sinale—
channel strip chart (Figure 69), or a thermal printer, commonly
used on multi-channel systems (Figure 70). The identification
and measurement of the arrival time of the first wave from the
seismic source is obtained from this presentation. The time
is measured in milliseconds, with zero time or start of trace
initiated by the source, which provides a trigger signal to the
seismograph.
Travel time is plotted against source—to—geophone distance
producing a time/distances (T/D) plot (Figure 71).
o The number of line segments indicates the number of
layers.
o The slope of each line segment is inversely
proportional to the seismic velocity in the
corresponding layer.
o Break points in the plot (critical distance, X ) are
used with the velocities to calculate layer depth.
Factors to be Considered for Field Use
The seismic line must he centered over the required infor-
mation area and overall line length must he three to five times
the maximum depth of interest. Resolution is determined by the
geophone spacing. Spacings of 3 to 15 meters are commonly
used; however, closer spacings may he necessary for very high
resolution of shallow geologic sections.
122
-------
Figure 68. A portable six—channel seismic refraction system in
use. A sle1gehammer is being used as a source.
123
. I
-------
ci )
1
0.
E
ci ,
>
(0
a)
I I I I I I 1 1 1 I
0 20 40 60 80 100 120 140 160 180 200 220
Time in Milhseconds
Figure 69. A typical seismic waveform from a single geophone.
Arrow marks arrival of first compressional wave.
Repetition of seismic refraction lines along a traverse
wilL reveal lateral variations. Resultina data can he used to
indicate trends of dipping layers and to detect anomalous
conditions, such as fractures and disturbed zones.
The general concepts presented so far have been for a simple
two—layer case with no dip. Since the presence of a dipping
layer may not be known, it is accepted practice to run both
forward and reverse lines to obtain true velocities and depths
if the geologic beds are not horizontal. A reverse line is
simply a second set of seismic measurements with the source
located at the opposite end of the same line of geophories
(Figure 72).
A modification of the classic seismic refraction method
will provide a rapid means of high—resolution profiling.
This profiling approach employs a single geophone and a fixed
spacing; this array is moved across the site. The distance
the station is moved each time is usually short (2 to 10 meters)
to provide good resolution of small features. Typically,
velocity anomalies (low velocities) will occur as the array
crosses a disturbed soil zone such as a trench.
A single channel seismograph is the simplest seismic
instrument arid is used with a single geophone and usually a
hammer source. The geophone is usually placed at a fixed
location arid the hammer is struck at regularly increasing
distances from the geophone. First wave arrival times are
identified in the instrument display, logged in a field hook,
and immediately plotted on a T/D plot. The single waveform
will approximate the one shown in Figure 69.
124
-------
Geophone
2
3
4
5
I a
t%J
Ui 7
8
9
10
11
12
Figure 70. Recording from a 12—channel seismoqraph. All 12 channels were
recorded sim’]ltaneously from a sinqie hammer impact.
0 50 100 150 200 250 300 350 400 450 500
Time in Milhseconds
-------
“r’ / “ill,,
Depth 0
+
/ / ‘‘ /
Layer I
/1/114 ‘ I/ IIfIIII lII/ IlI II/ IIIII//f 1/11/fl
Layer 2
Source to Geophone Distance
Layer I
Layer 2
Velocity of Layer I I/Slope of L I
Velocity of Layer 2= I/Slope of L2
Critical Distance
D Xc\/V2-VI
2 VV2’Vl
(For Two Horizontal Layers)
Time/distance pLot for a simp1 e two—layer structure.
i_s used to calculate depth of Layer.)
(Equat i_on shown
-S
•1
E
I-
0
a
I-
L2
I — .
0\
Ll=
L2
V2=
xc=
Fiqure 71.
-------
Multichannel seismographs increase the rate of data
acquisition using an array of 6, 12, 24 or more geophones
(Figure 67). All geophone signals are recorded simultaneously
after initiation from a single hammer blow (Figure 70). The
display of simultaneous waveforms enables the operator to
measure arrival time by noting trends in the composite data
set. This is especially useful in noisy areas. More
sophisticated instruments commonly incorporate a considerable
amount of control over gain and filtering of the signals,
which is of great use on difficult or “noisy” sites.
Since the seismic method measures ground vibration, it is
inherently sensitive to noise from a variety oE sources.
Signal enhancement is a significant aid when workinq in noisy
areas and with smaller energy sources. Enhancement capability
is available in most single and multi-channel systems. Enhance-
ment is accomplished by adding a number of seismic signals
from repeated hammer bLows. The coherent seismic signal is
increased in direct proportion to the number of blows, while
random noise in the seismic signal is increased only by the
square root of the number of blows. This causes the seismic
signal to “grow” out of the noise level, permitting operation
in noisier environments and at greater liamrner-to—geophone
spacings. The overall results provide a more accurate measure-
ment of the first arrival time.
Depending on site conditions, a hammer is useful for obtain-
ing seismic data to depths of 10 to 15 meters, while a 250-
kilogram (500-pound) drop weight is required for depths of 50
to 100 meters. A more powerful seismic source is necessary to
obtain deeper data or for work in noisy areas. Many sources are
available for meeting specialized needs. If,the use of explosives
or projectile sources is contemplated, the project manager must
consider the safety hazards inherent in such methods, as well as
their impact on the hazardous site itself, and the response from
the surrounding neighborhood. Local laws, insurance requirements
and the increase in project cost associated with compliance may
also restrict the use of explosives.
Quality Control
Quality control can be achieved in several ways:
o A check of the seismic signal and noise conditions on
the instrument display will verify the proper function-
irkg of geophones and trigger cables and the correct
setting of the instrument.
o In cases where paper records are not made, arrival time
picks made from the electronic display should be
immediately plotted on a T/D graph in the field.
Problems with improper picks are often discovered by
early inspection of these plots.
127
-------
Source
Figure 72. Use of forward and reverse seismic lines is necessary
to determine true velocities and depths with dipping
horizon.
Forward Line
Source
Reverse Line
128
-------
o If the data is to be used for legal purposes, or if it
must be reviewed by persons other than the field party
chief, a hard copy of the data must he made. Multi—
channel systems provide a much hetter means of present-
ing the data than do single—channel units (compare
Figures 69 and 70). The individual traces of the single—
channel system would have to be clipped and pasted
together and would provide a much less acceptable—looking
and workable record. For simple, smaller surveys,
however, the single-channel units are quite satisfactory
if used by experienced personnel.
o Background or off—site data is often required for
correlation to known geologic information and to
establish clean background data. This information is
also useful as a reference for evaluating complex site
conditions.
o Boring 1 oys should be obtained to minimize the
possibility that low velocity (hidden layers) or thin
beds will remain undetected.
o Electronic calibration of the timing circuits of the
seismograph may be made in the laboratory. However, this
is rarely necessary because these timing circuits are
crystal-controlled and have inherently low drift. Normal
annual factory maintenance will include such calibration.
o The seismic system may also be run at a standard base
station for periodic check of the instrument operation.
Noise
Seismic signals are strongly affected by ground vibration
noise; less so by geologic scatter. In addition, the subjective
pick of first arrival time can contribute a few mi].liseconds of
error.
Unwanted vibrations which affect the seismic signal at the
geophone may be caused by:
o Strong winds which move nearby trees 7
o Sounds of airplanes;
o Surface sources, such as moving vehicles on nearby high-
ways and railroads;
o Field crews walking near geophones;
o Nearby blasting or operation of heavy construction
equipment.
Geologic scatter may be caused by lateral variation in layer
composition or an irregular interface between layers. Such
scatter can complicate interpretation of the T/D pIet, hut is
also a valuable indicator of site conditions (see Figure 73).
Examples include:
129
-------
o Variations in the thickness of the “soil zone”;
o Boulders in glacial clay or till;
o Zones of increased cementation in sandstone and
limestone;
o Lenses of sand in clay layers;
o Variations in saturated water content caused by perched
water tables;
o Irregular bedrock surfaces;
o Limestone containing numerous cavities.
Data Format, Processing, Interpretation and Presentation
First arrival times are usually measured from seismic
signals in the field arid are recorded in field logs and plots.
Waveforms may be recorded as hard copy by strip chart records,
oscillograph and thermal printers, or by magnetic media for
archives, subsequent playback and processing. T/D plots
permit calculation of layer parameters. The results may then
he interpreted to yield a geoloqic section of subsurface condi-
tions (Figure 74). Figure 75 shows this sequence of processing
and interpretation.
The processing procedure begins with determination of the
first arrival time to each geophone. The enhancement technique
may be employed to aid in recognizing the first arrivals on the
display. Multi—channel seismographs can also assist in identify-
ing this first wave by revealing a trend in the composite
received signals (Figure 70).
Once the arrival times are determined for each geophone ,
the time/distance plot is constructed. Straight line segments
are fitted to linear sections of the plot by least square
techniques. The number of segments and their slopes correspond
to the number of geologic layers and their velocities. These
velocities and critical distances (determined by breaks in the
line segment) are used to calculate the depth of the layer.
Forward and reverse data is needed to provide true velocities,
depth, and dip of each layer if layering is not horizontal.
Generally two-or three—layer systems can be analyzed in
the field by the use of nomograms and simple calculations. More
complicated sites having three to four layers with dip will
require a programmable calculator or a small computer to solve
the seismic equations.
Single refraction stations can he represented in a similar
manner to provide one—dimensional data. The results from a number
of refraction stations can he interpreted and combined into a
two—dimensional cross section, as shown in Figure 74. If
refraction lines are available over a large area or along
perpendicular traverse lines, the resulting data can be shown as
a three—dimensional section or as a fence diagram.
130
-------
t
w
E
I-
MODEL
00
0
‘,
High Velocity
0
600
0 a
0 0
1200
Figure 73.
Time/distance plot shows scatter caused by non-uniform
soil conditions. such conditions might he caused by
differences in local cementation, bolders, etc.
(velocities shown in meters/sec.)
0
Distance —
131
-------
Time/distance plots may show abnormal slopes or breaks.
Such plots reveal that the subsurface is not composed of
homogeneous layers of uniform thickness and velocity. Several
types of conditions are recognizable from their characteristic
T/D plots. For example, Figure 76 shows a T/D plot where a
seismic line passes over a deep burial trench. The degree of
geologic scatter in the data (Figure 73) is also a good indicator
of subsurface conditions.
Many possible pitfalls exist in the acquisition 1 processing,
and interpretation of seismic data. Solutions are not unique
and all interpretations must be based on some assumptions
about the site, together with independent information on the
geohydrological conditions at the site. Velocity inversions
and thin bed cases may not be detected. Dipping bedding can
cause considerable error in calculations; therefore, both
forward and reverse lines must be used.
Summary
The seismic refraction method can he used to aid in
riefinirig natural geohydrologic conditions, including thick-
ness and depth of soil and rock layers 1 and depth to bedrock
or water table. Generally two-or three—layer systems can he
analyzed in the field by the use of nomograms and simple
calculations. More complicated sites having three to four
layers with dip will require a programmable calculator or a
small computer to solve the seismic equations.
Since seismic wave velocity is directly related to the
material properties of the layer such as density and hardness,
lateral variations in composition or an irregular interface
between layers will show up as geologic scatter on a T/D plot.
This is a valuable indicator of variations in site conditions.
The analysis of this data requires that the interpreter he
knowledgeable about the method, the conditions under which the
data was obtained, and the geohydrologic conditions.
The seismic line must be three to five times the maximum
depth of interest. Lateral resolution in the data is determined
by the geophone spacing.
Depending on site conditions, a hammer source is useful
for obtaining seismic data to depths of 10 to 15 meters, while
a 500—pound drop weight is required for depths of 50 to 100
meters’ Explosives or projectile sources may he used to obtain
deeper data.
Since the seismic method measures small ground vibrations,
it is susceptible to vibration noise from a variety of natural
and cultural sources.
132
-------
0
10
20
‘a
Li.
z
4o
I i i
0
30
50
60
70
Figure 74.
Geologic section interpreted from seismic data
(seismic velocities in feet/second).
WEST
REVERSE
EAST
FORWARD
0
I0
20
30
40
50
60
70
133
-------
Single Channel I D Plots
I
I_______________ Manual Pick of _________
Arrival Times
Multi Channel ______________
_________________ — I Velocity I
______________ Table
Geophones F — v\J cjv’ rv rw
Geologic
Data
v vw
Computer
Magnetic Tape
—
100 __
Interpreted
Geologic Section
Figure 75. Flow diagram showing steps in processinq and interpretatinn of
seismic refraction data.
-------
T/D Plot
Model
I
1000
500
1000
Loose
Well-Cemented
Fill
Well-Cemented
Soils
Material
Soils
Time/distance plot showing lateral velocity change.
Such a plot could he ohtainec when the refraction
line crosses a burial trench. (velocities in
meters/sec.).
1000
500
0 Distance -
Figure 76.
135
-------
The seismic method is inherently a station measurement
because geophonos niust he implanted in the surface of the ground.
This makes the method relatively slow when compared to the other
continuous techniques.
Capabilities
o Seismic refraction measurements can provide depth and
thickness of subsurface geologic layers including depth
to rock and water table.
o Seismic velocity of the layers can he related to their
physical properties including composition, density arid
elasticity.
o Disturbed soil zones can often be detected and mapped,
permitting the location and delineation of burial zones
at WS.
o Depth to bottom of disposal areas and landfills may he
estimated without drilling.
Limitations
o Seismic data is gathered as a station measurement and
involves relatively slow field procedures compared to
continuous methods.
o Interpretation requires that site condition be relatively
uniform to obtain highly accurate resurts.
o The seismic method is very susceptible to vibration
noise.
Examples
Determining Depth and Thickness of Clay Layer for Proposed
Disposal Site
The depth and thickness of a clay layer was required in
an evaluation of natural site conditions. A number of seismic
stations were located throughout the area of interest.
Figure 77 shows one of the time/distance plots made during
this survey. Note that there is very little scatter in the
data forming each line segment; this indicates that the each
subsurface layer is relatively homogeneous. Both forward and
reverse lines were run to evaluate clip, which was found to be
insignificant. Since no explosives could be used on this site,
a ten pound sledge hammer was employed. This inherently limited
the maximum depth of the data arid the bottom of the clay was not
detected, hut a minimum thickness could be established.
Figure 78 shows the localized geologic section constructed
from this single refraction measurement The results of drilling
are also shown for correlation. The seismic data had excellent
136
-------
correlation with the major geologic changes, although it did not
detect the subtle changes in color which showed up in the horinq
logs.
Deterniininq Soil Thickness and Bedrock Depth at HWS
A burial trench containing drums of hazardous waste was
located in a karst area. The depth of the residual clayey soil
over bedrock was of prime concern in the evaluation of possible
contaminant migration away from the burial site into the
permeable bedrock. The burial trench was known to have been
dug by bulldozers and was believed to be less than 2 to 3
meters deep. Drilling was not used because of the risk of
opening up a pathway through the clay to the fractured
limestone.
Four seismic refraction stations, using forward and
reverse lines, were used to determine the depth of the lime-
stone bedrock around the burial trench perimeter. One of
these T/D plots is shown in Figure 79. There is consi erahle
geologic scatter in the data, which indicates that there are
inhomogeneities within the soil horizon and an irregular soil/
bedrock contact.
Analysis of the seismic data revealed a steeply dipping
limestone surface fro n one nd of the trench to the other
(shown in Figure 80). This was cross—verified by the seismic
refraction data taken on the opposite side of the trench, as
well as the data from both ends of the trench. This multiple
confirmation provided a high confidence level in the assessment
of depth to rock without the use of drilling. Depth to lime-
stone vari.ed from four meters to ten meters at the two seismic
stations off the ends of the trench. These results showed
that there was a minimum of two meters of soil between the
deepest trench bottom and the top of the rock; the risk of
rapid contaminant migration into the bedrock was judged to he
low based upon this data.
137
-------
(1)
V
C
0
C)
0
U,
E
0
E
I -
Figure 77.
Time/distance plot of field data showing three layer geologic system.
(only forward li.ne shown for clarity).
I- . .
Distance (in feet)
-------
Seismic Section Drillhole
—0 FEET
Gray Sand
V I 1117 F/S Tan Sand
______________ - Tan Sand w/Clay
Red SandyClay 10
V2:2222 F/S
Dark Tan
Sandy Clay
Tan Plastic Clay
- —20
Gray Plastic Clay
V3:4878 F/S
-30
End of Hole
—40
Figure 78. Interpreted seismic data (Figure 77) compared to
driller’s log.
139
-------
40
35
30
25
20
‘5
I0
5
0
C
0
C.)
E
E
Figure 79.
Distance (in feet)
‘l’ime/distance plot of fie1 data
reverse seismic refraction data.
showing forward and
0 0 20 30 40 50 60 70 80 90 100
100 90 80 70 60 50 40 30 20 10 0
X40
-------
East West
V o 0 Surface
Loess Layer / V 69 ‘s
I
/
Probable Outline of Burial Trench I
?
IC
V 2-2493
4 )
‘I-
Clayey Soil
hertBlock
4)
O 20-
Limestone
30- 30
Figure 80. Geologic section resulting from interpretation of
seismic data (Figure 79). Estimated outline of
trench is shown. (velocities in feet/secon I)
141
-------
SECTION VIII
METAL DETECTION (MD)
I ntroductiori
Metal detectors (MD) are designed to locate buried
metallic objects. They are commonly used by treasure hunters
searching for coins and by utility crews for locating buried
pipes and cables. In HWS investigations, MD are invaluable
for detecting buried drums and for delineating the boundaries
of trenches containing metallic drums.
Metal detectors can detect any kind of metallic material,
including both ferrous metals such as iron and steel and
non—ferrous metals, such as aluminum and copper. (In contrast,
another search device, the magnetometer, discussed in Section
IX, responds only to ferrous metals.)
Metal detectors have a relatively short detection range.
Small metal objects such as spray cans or quart-sized containers
can be detected at a distance of approximately 1 meter. Because
the response of a metal detector increases with the target’s
surface area, larger objects like 55-gallon drums may he detected
at depths of 1 to 3 meters. Massive piles of metallic materials
may be detected at depths of 3 to 6 meters.
The metal detector is a continuously—sensing instrument
which can provide total site coverage and which is well suited
for locating buried metal. Experience at HWS investigations has
shown that metal detectors can be effectively used to:
o Locate buried metallic containers of various sizes;
o Define boundaries of trenches containing metallic
containers;
o Locate buried metalLic storage tanks;
o Locate buried metallic pipes;
o Avoid buried utilites when drilling or trenching;
o Locate utility trenches which may provide a
permeable pathway for contaminants.
Principles and Equipment
A metal detector responds to the electrical conductivity of
metal targets, which is relatively high compared to normal levels
142
-------
of soil conductivity. These targets must, of course, be
within the range of the instrument to be detected.
There are many different types of metal detectors available
commercially. This document will consider three general classes
of equipment:
1. Pipeline/cable locators (Figure 81)
2. Conventional “treasure hunter” detectors (Figure 82)
3. Specialized detectors (Figures 83, 84 & 85)
Numerous pipeline/cable locator metal detectors are
commercially available. Besides being effective for locating
buried utility cables and pipes, they can be used to detect
larger buried targets such as 55—gallon drums, with the added
feature that they will not respond to small unwanted surface
targets like soda cans. This type of detector is corrirnonly used
by EPA Field Investigation Teams (Figure 81).
There is also a wide variety of “treasure hunter” metal
detectors on the market (Figure 82). While many of these units
are generally designed for locating small coin-sized objects,
some of them offer the option of larger sensor coils, which makes
them suitable for surveys at intermediate depths. Some of these
units are also capable of operating in areas where natural soil
conditions (such as large amounts of iron minerals) could
adversely affect the instrument’s performance.
Specialized detectors have been designed to deal with
unique problems. They are expensive, not commonly available,
and require an experienced operator. However, these units are
quite versatile: working to greater depths, covering a wider
area with each pass, producing continuously recorded data
(Figures 83 & 84), and operating from a vehicle when necessary
(Figure 85). They are invaluable for coping with special field
problems such as interference from natural soil conditions and
nearby man-made materials.
Figure 86 shows the principle of operation and the
functional parts of the typical pipe/cable detector shown in
Figure 81. The transmitter of a metal detector creates an
alternating magnetic field around the transmitter coil. A
balance condition must be achieved to cancel the effect of this
primary field at the receiver coil. Shown in Figure 86, the
balance or null is accomplished by orienting the planes of the
two coils perpendicular to one another. The primary field will
induce eddy currents in a metal target within range of the
instrument. These eddy currents, in turn, produce a secondary
field which interacts with the primary field to upset the
existing balance condition. The result will be an output on a
meter and/or an audio signal.
143
-------
Figure 81.
Industrial pipe/cable locator.
is in common use by EPA field investigation teams.
144
This type detector
r ;
‘ ‘:
£. .
4è1V .
‘..
1
, •1
: #
a.
-------
Figure 82. Typical treasure hunter type metal detector with
large search coil.
145
I
-------
[
p
C ’
. , . ; : % : . ...
rrr
• •,wA
- p
I
I
a - ’
1
t —j
*f
4
S
- a
rise
C .
Figure 83.
Specialized metal detector system in use. Operator on left
recorder and system electronics (Denney Farm site).
is carrying
-------
Figure 84. Specialized metal detector system with large search coil. Note strip chart
recorder on chest of operator (Savanna Army Depot).
- .4
-------
Figure 85.
Truck—mounted metal detector system provides rapid site coverage over large
areas.
I-I
—
-------
Figure 86. Simplified block diagram of a pipe/cable type metal
detector system. Primary field from transmitter is
distorted by buried metallic objects causing upset
of null at receiver coil.
/
Receiver
Ground Surface
Buried Drums
149
-------
Other types of metal detectors combine the transmitter arid
receiver coils into one sensor package, and they may respond to
the eddy currents generated in the target in different ways.
These eddy currents may he sensed directly by the receiver or
they may cause direct loading effects on the transmitter. A
discussion of the details o the various types of metal detectors
and their electronics is beyond the scope of this document.
Several factors influence metal detector response: the
properties of the target, the properties of the soil, and the
characteristics of the metal detector itself.
The target’s size and its depth of burial are the two most
important factors. The larger the surface area of the target,
the greater the eddy currents that may he induced, and the
greater the depth at which the target may be detected. (Response
is proportional to the cube of the area.)
For example, if all the steel. in a 55—qallon drum were
collapsed into a solid rod of approximately the same ].enath as
the drum, the rod would yield a very poor MD response. However,
because the steel mass in the drum is in the form of a thin
sheet, there is a considerably greater area of metal, in which
substantially greater eddy currents may he developed. Conse-
quently, a single 55-gallon drum is an ideal target and may he
detected at distances of 1 to 3 meters depending on the specific
equipment used.
The MD’s response to a target decreases at a rate equal to
reciprocal of its depth to the sixth power (1/depth 6 ). There-
fore, if the distance to the target is doubled, the MD response
will decrease by a factor of 64. Consequently, the MD is a
relatively nearfield c1evice it is generally restricted to
detecting small targets at relatively shallow depths or larger
targets at limited depths. Generally, most metal detectors are
incapable of responding to any targets, no matter how large,
at depths much greater than 6 meters.
Although the shape, orientation and composition of a target
will influence the MD response, these factors will have much less
influence than will the size and depth of the target. Target
deterioration, however, may have significant impact. Metallic
containers will corrode in natural soil conditions and this
corrosion can he accelerated by unusual conditions at the HWS.
If a container is corroded, its surface area will he signifi-
cantly reduced arid this, in turn will degrade the response of a
metal detector. Using average corrosion rates for steel in soil
and considering a range of drum metal thicknesses, the life of a
buried drum might range from 5 to 20 years under normal soil
conditions. Under adverse conditions, however, this corrosion
rate could be accelerated by a factor of 5 or more.
150
-------
High concentrations of natural iron—bearing minerals in the
soil will limit the performance of many metal detectors. Simi—
larily, high concentrations of salt water, acids and other highly
conductive fluids will also reduce the effectiveness of a metal
detector. Search operations conducted in an area with consider-
able metallic debris will range from very difficult to
impossible.
Iron minerals, conductive fluids, and metallic debris will
affect the MD in much the same way as a target. A false response
will he produced which may confuse the searcher or render the
search impossible. In the case of metallic debris, the successful
application of a MD will depend on the relative size of the
debris and its density. It is obvious that a metal detector
survey for buried drums could not he conducted if the surface
were to be totally covered with drum lids. In such a case, the
detector would respond to the drum lids and deeper targets would
be masked.
Some compensation for natural soil conditions, metall.ic
debris, and nearby metallic structures can be made by using
certain specialized equipment and modified field procedures.
These are described in the sections dealing with field usage and
noise.
Because the MD ’s response weakens rapidly with increasing
target distance, system gain and instrument stability are impor-
tant. Coil size is the only variable which can easily be modified
on some metaldetectors by the use of interchangable coils. The
influences on system response of target size and coil size are
shown in Figure 87. These data shows the detection ranges for
various common targets and a e presented for two coil sizes.
Since the equipment capabilities vary widely, this curve is
intended to provide only an approximate guide to metal detector
response.
Factors to be Considered for Field Use
Before beginning a metal detector survey, an estimate of the
types, sizes and depths of metallic targets should be made. Soil
conditions, metallic debris, fences, and the size of the search
area should he considered. Finally, the type of MD should be
selected to fit. the overall survey objectives. This may mean
selecting more than one MD to fulfill a project requirement.
The pipeline type of detector has often been used by EPA
for locating buried drums. As its name implies, it is also very
effective for locating buried utilities; as such, it is a dual—
purpose instrument. ecause it has an effective coil diameter of
about 1 meter (the distance between the transmitter and receiver
coils), it is useful in surveys for larger targets and greater
depths. Its larger effective coil size also makes it somewhat
151
-------
Coins
Soda Cans
Gallon
Containers
55 Gallon
Drums
Massive
Targets
Soda Cans
Gal Ion
Containers
55 Gallon
Drums
Massive
Targets
0
v // Il
I Meter Diameter Coil
V//I/Il/f I/I//I/f/ I
34 e
Depth (in meters)
Figure 87.
Approximate detection ranges for common targets.
Data is shown for two se trch coil, sizes. A wide
variation in detection range occurs because of the
many variables involved.
U
0
0.3 Meter Diameter Coil
I I
2
0
‘ . 4
U)
0
a)
I-
a
I-
0
N
U)
0
a’
I-
0
I-
3
Depth (in meters)
4 5 6
152
-------
insensitive to small pieces of metallic surface debris,
if the detector is elevated above the surface of the around, as
shown in Figure 81.
Metal detectors with coils of sma].ler diameter (typically
less than 0.3 meters), such as those used by treasure hunters,
(Figure 82) are much better suited to locating small targets at
limited depth. If the field problem is to locate smaller but
critical targets, such as individual quart cans of toxic
materials, this type of detector is a more logical choice. The
detector in Figure 82 is shown with a larger coil which can he
useful for detecting single 55—gallon drums at depths of up to 2
meters. These detectors may also be used to evaluate the extent
and influence of near-surface trash.
Specialized metal detectors, such as those shown in Figures
83, 84, and 85, are available to handle unusual site cortd.ttions
or to deal with unique project requirements. These specialized
metal detectors can provide:
o Increased depth range;
o CancelLation of interference caused by nearby fences,
etc;
o Compensation for unusually difficult soil conditions;
o Continuous recorded data;
o Larger coil configurations to provide greater width
of coverage on each pass;
o Vehicle—mounted configurations to cover very large areas;
o Improved semi-quantitative assessments (i.e., estimating
the depth and number of drums);
o Classification off targets.
One of the more significant applications of specialized
detectors is realized at sites with complex conditions. Here,
the MD output can be continuously recorded on a strip chart or
magnetic tape for later plotting to provide improved mapping and
analysis. Vehicle—mounted sensors are almost a necessity if the
area to be surveyed is very large.
A simple reconnaissance investigation for the purpose of
detecting the presence of drum-sized targets may satisfy many
site survey requirements. On the other hand, remedial action
may call for a detailed study to locate huria]. trenches with high
resolution and to provide estimates of the quantity of drums.
The lateral resolution capability of MD instruments often permits
the successful delineation of closely—spaced multiple trenches,
when other techniques fail. If site conditions are relatively
simple, a detailed assessment of burial site boundaries may he
carried out without recording the data. Stakes may be placed
around the burial boundaries as the survey progresses.
153
-------
More complex sites may require very detailed information
and a means of recording many target locations. For such
cases, a survey grid of parallel lines should be established
at an interval which will provide the resolution required.
Typically, such grids may employ spacings of from one meter to
more than 10 meters. (Survey stakes should obviously he made
of nonmetallic materials.) The coarser grid spacing might be
used first to locate larger burial sites, tanks or pipeline
crossings; then, selected areas of interest could be surveyed
at closer intervals for more complete coverage. Overlap of
survey lines may be required at some sites if smaller or discrete
targets are important. When data is recorded, the records must
he annotated with station locations so that the spot can be
located again or for the construction of a location map.
Depending on the size of the area to be covered, hand—
carried or vehicle-mounted systems may be used. The vehicle
systems can tow wide coils (typically up to 2 to 3 meters wide)
for greater coverage. In certain areas, the need for high
resolution may require a hand search to provide the necessary
details.
The effects of high iron content in the soil or numerous
small metallic fragments lying on or near the surface can be
minimized by elevating the search coil one to three feet above
the ground. This technique is applicable only if the targets are
large enough to be detected at the increased distance from the
search coil. Figure 84 shows a coil of 1 meter diameter being
used in this mode to locate massive concentrations of buried gas
canisters, where the surface of the ground is covered with small
pieces of metallic debris.
Quality Control
Metal detectors are usually not calibrated. They respond
in a relative way; i.e., closer or larger metallic targets
create a greater output level than smaller or more distant
ones. n experienced operator can usually make a reasonably
accurate estimate of target size and depth. However, any
attempt at detailed calibration will likely he useless, because
of the many variables involved. For example, hicalibrationu
curves relating MD meter response to a steel drum as a function
of distance may he accurate under a given test standard condition,
but unfortunately these curves are seldom valid, because of
the variability and complexity of actual field conditions.
Moreover, the operator cannot easily determine the difference
between a single drum located at medium depth and several
drums lying deeper. What he can report, however, is that
drum—sized targets are present in certain specific areas.
154
-------
Noise
The effectiveness of a metal detector is dependent upon
the relative magnitude of the target signal, the noise produced
by the surrounding soil, and other variables. The procedure
used to null a metal detector serves to cancel most of the
soil interference; however, some level of noise from soil
conditions may be present during a survey. As the target
response decreases and/or the noise level increases, the target
response will eventually he lost in the noise. While it is
true that the larger coils will yield better signals from
larger and deeper targets, they are also more susceptible to
soil effects and other electrical interference. However, the
larger coils can be raised, up to about 1 meter off the ground,
to minimize both the soil effects and the effects of metal
trash near the surface.
When the coil is carried too close to the ground, small
shallow targets may easily saturate the system to a full—scale
response. When this occurs, other targets, no matter how large,
cannot cause a further increase in response and will, therefore,
remain undetected.
It is important to understand that a metal detector radiates
a field in all directions (Figure 86). However, its most
sensitive zones are “focused” directly above and below the plane
of the sensor coils. This characteristic can he quite useful in
the field. The focused response characteristic of the MD will
allow the operator to work relatively near some metallic items,
so long as they are far enough to the side of the sensor coil.
In addition, the focused response provides good definition of
the edges of trenches containing buried drums.
The operator must exercise care to avoid interference from
nearby fences and vehicles, as well as from buildings and buried
pipes. For example, by running a survey line parallel or
oblique to one or more unknown pipelines, the operator can cause
invalid data to be produced. Certain welded fence materials and
the mesh used for concrete reinforcement will provide a very good
MD response, despite the fact that they are not solid metallic
surfaces.
Precaution must also be taken to remove metal from the
operator, or to minimize its effects. Steel—toed boots,
respirators and air bottles can all cause considerable problems
with noise.
155
-------
Data, Processing, Interpretation, and Presentation
Unrecorded Data—-Most commercial metal detectors have both
audio and meter indicators, with no provision for directly
recording the information output. Reconnaissance—level surveys
with relatively simple site conditions can be handled effec-
tively with these instruments. MD responses or target locations
can he noted in a field log or site map: stakes or paint marks
can be placed over target centers or around their boundaries as
the survey proceeds. This is the approach commonly used to
locate buried utilities. Since MD results are generally self—
evident, further analysis and processing are unnecessary.
An experienced operator using these simple field proce-
dures, commercial equipment, arid unrecorded data might he able
to provide the following:
o The location and delineation of buried metallic objects:
o A crude approximation of depth;
o A crude approximation of size of discrete targets.
Recorded Data-—Metal detectors with recording capability
should be considered if:
o There is a need for coverage of large areas:
o Complex distribution of burial areas exists;
o Semi-quantitative results are required;
o Difficult soil conditions prevail;
o Documentation is necessary.
The output from some specialized MD can be recorded directly
on strip charts for interaction in the field or on magnetic tape
for later playback and/or processing. Such recorded data is
invaluable in locating and mapping the boundaries of metal in
randomly-oriented trenches and burial pits. Data acquired along
grid or parallel survey lines can he assembled into an accurate
composite map of the site.
Both strip chart and magnetic tape data lend themselves
to computer graphics and processing. Corrections for profile
linearity and scale/range changes can he made Filtering may he
applied to all or part of the profile data. High—frequency
noise from small local targets such as soda cans may he removed
to improve the analysis and display of massive burial sites.
Finally, the results can be plotted in contour maps or as
3—dimensional views of the combined data set. Semi—quantitative
assessments, such as the determination of the number of drums
and their burial depth, are not easily accomplished using the
MD method alone. It is commonly necessary to use other tech-
niques in conjunction with the MD survey in order to derive
this expanded level of information.
156
-------
Summary
At HWS, metal detectors are primarily used to determine
the presence, location and definition of trench boundaries.
They can also be used to assist in the process of selecting a
site for drilling, so that metal.lic containers and underground
utilities are not accidentally struck during the drilling
operations. Buried tanks and pipes which may be sources of
leaks can be located; and in addition, the location of utilities
may serve to define areas representing more permeable passage—
ways in which contaminants may flow.
Metal detectors will detect any kind of metaUic material,
including ferrous metals, such as iron and steel, and non—ferrous
metals, such as aluminum and copper. (In contrast the magneto-
meter, discussed in Section IX, responds only to ferrous
metals.)
The metal detector is a continuously—sensing instrument
which can provide total site coverage, and which is well suited
for locating buried metal within its depth range. The ].ateral
resolution capability of M D instruments often permits the
successful delineation of closely—spaced multiple trenches when
other techniques fail.
Metal detectors have a relatively short range. They can
detect quart—sized containers at a distance of anproximately
one meter. The response of a metal detector increases with
the target’s surface area; therefore, larger objects like 55—
gallon drums may be detected at depths up to 3 meters, and
massive piles of metallic materials may he detected at depths
up to 6 meters. Specific performance is highly dej,endent upon
the type of metal detector used. Generally, most metal detectors
are incapable of responding to any targets, no matter how large,
at depths much greater than 6 meters.
An experienced operator can usually make a reasonably
accurate estimate of target size and depth. However, any
attempt at detailed calibration will likely be useless,
because of the many variables involved.
Metal detectors are very susceptible to noise caused by
some natural soil conditions, unwanted metallic debris, pipes,
fences, vehicles, buildings, etc.
There are many different types of metal detectors available
commercially, each with its own advantages and limitations. The
choice of a MD should be determined by the type of targets to
be located, their depth, the nature of the soil, the size of
the search area, site conditions and other project requirements.
157
-------
Capabilities
o MD respond to both ferrous and non—ferrous metals;
o They will detect single 55—gallon drums at. depths of
up to 1 to 3 meters;
o They will detect large masses of drums at depths of
up to 3 to 6 meters;
o MD provide a continuous response along a traverse line;
o A wide range of commercial equipment is available most
of which is relatively easy to use;
o MD provide very good definition of boundaries in burial
trenches and pits containing metal;
o Limited semi—quantitative information may be obtained
from the use of commercial detectors;
o Specialized equipment is available for recording data,
coping with unique site conditions, or obtaining semi—
quantitative information.
Limitations
o Metal detectors are inherently limited in depth
capability;
o They are susceptible to a wide range of noise including
that introduced by natural soil, metallic debris, pipes
and cables, and nearby fences and metal structures;
o The performance of many commercially—available detectors
is marginal for use at HWS;
o They are limited in providing quantitative data concern-
ing the number and depth of targets;
o Specialized MD instruments are uncommon and require
experienced personnel;
o CompLex site conditions will demand increased levels of
skill; special equipment; the recording, processing and
plotting of data; and experienced interpreters.
Examples
Selecting a Safe Drilling Site
When work is being done in a hazardous area, drill sites
should be positioned with care to avoid drilling into disposal
pits, drums, pipes or cables. Figure 82 shows a proposed drill
site 7x7 meters in area which was selected for a monitor well.
The site was immediately adjacent to an area known to contain
extremely hazardous buried materials. The area was surveyed
by two metal detectors to provide a high confidence level in
selecting the precise well location. A large coil system was
used to detect larger, deep targets; a small coil system was
used for the smaller, shallower targets. Data was not recorded;
targets were noted as the survey progressed. They were imme-
diately marked with wooden stakes, then reverified after
placement of the stakes to eliminate possible position errors.
158
-------
The use of a rope grid permitted a very rigorous and complete
survey of the site to be completed in a relatively short time.
After completion of this search procedure, the largest clear
area was again resurveyed to verify that it was, indeed, free
of metal. The exact well location was then positioned in the
center of this clear area.
Location of a Single Burial Trench
An area suspected of containing a trench with buried drums
was investigated using a specialized MD system with a recorded
data output. The first survey line was run perpendicular to
the obviously disturbed soil area and yielded the profile
shown in Figure 88. This profile shows a very strong response
over a distance of about 10 feet. Additional parallel MD
lines were run across the remaining disturbed area. These
profiles when plotted together (Figure 89) show a linear trench-
like feature composed of a strong central response and very
distinctive boundaries. Such results are characteristic of a
trench filled with a large number of steel drums. The same
profile data may also be presented in a plan view, as seen in
Figure 90. This format permits exact location of the edges of
the burial trench, and subsequent calculation of its area.
Using depth estimates from other data, the investigator can
calculate the volume of the trench and arrive at an estimate
of the number of buried drums.
Location of Multiple Burial Trenches and Pits
A large singular burial pit containinn canisters of extremely
hazardous materials was known to e*ist in a field (100 x 1 0
meters) that had been overgrown by bushes and weeds. Several
MD reconnaissance lines were run across the general area to
locate the hurial site. Within an hour, this unrecorded recon-
naissance survey revealed that the entire area was a complex
maze of metallic targets and, possibly, contained multiple huria]
sites. (At this point, the exact location of the large burial
pit was uncertain.) Many of the smaller targets could be
attributed to metal debris lying on or just below the surface of
the ground.
It was determined that a high—density recorded MD survey was
necessary Lo properly evaluate the site. A detailed survey was
designed using multiple parallel lines of 15 meter spacing.
A specialized detector system was selected which provided a
chart-recorded output and a search coil one meter in diameter.
The additional depth capacity of the large coil permitted it
to be elevated above the ground to minimize the effects of
small fragments of metallic debris at the surface (See Figure
84). The results of this survey produced the data shown in
Figure 91. This presentation clearly indicates the overall
complexity of the site. Although only one large burial trench
159
-------
was believed to exist on the site, this detailed MD survey
revealed the existence of smaller pits and trenches within the
survey area. This evaluation would have been very difficult,
or impossible, without the use of the specialized MD system anrR
recorded data.
Figure 88.
Continuously-recorded metal detector data over a
trench with buried drums. Note good resolution of
trench edges.
TECHNOS 11/C, If 1PM !
Figure 89.
Three—dimensional perspective view of metal detector
data from parallel survey lines over a single trench.
T RENCI-I
EDGE
TRENCH
EDGE
160
-------
Figure 90. Plan view map oE burial trench boundaries based upon metal
detector data in fiqiire l9.
T
N
-l
+
Q
+
N
4.
I-
C ,)
+
CL
-------
Figure 91.
Perspective view of metal detector data from parallel
survey lines shows a complex burial site. A large
trench containing metal can he seen in the left—hand
side of the data (shaded).
rECI/NOS INC, MIfiMI
162
-------
SECTION IX
MAGNETOMETER
Introduction
Magnetic measurements are commonly used to map regional
geologic structure and to explore for minerals. They are
also used to locate pipes and survey stakes or to map arche-
ological sites. They are commonly used at E WS to locate
buried drums and trenches.
A magnetometer measures the intensity of the earth’s
magnetic field. The presence of ferrous metals creates
variations in the local strength of that fields permitting
their detection. A magnetometer’s response is proporti.onal
to the mass of the ferrous target. Typically, a single drum
can be detected at distances up to 6 meters, while massive
piles of drums can be detected at distances up to 20 meters
or more.
Some magnetometers require the operator to stop and
take discrete measurements; other instruments permit the acqui-
sition of continuous data as the magnetometer is moved across
the site. This continuous coverage is much more suitable for
high resolution requirements and the mapping of extensive
areas.
The effectiveness of a magnetometer can he reduced or
totally inhibited by noise or interference from time—variable
changes in the earth’s field and spatial variations caused by
magnetic minerals in the soil, or iron and steel debris, ferrous
pipes, fences, buildings, and vehicles. Many of these problems
can be avoided by careful selection of instruments and field
techniques.
At HWS, magnetometers may be used to:
o Locate buried steel containers such as 55-c allon
drums;
o Define boundaries of trenches filled with ferrous
containers;
o Locate ferrous underground utilities, such as iron
pipes or tanks, and the permeable pathways often
associated with them;
163
-------
o Select drilling locations that are clear of buried
drums, underground utilities, and other obstructions.
Principles arid Equipment
A magnetometer measures the intensity of the earth’s
magnetic field. Variations in this field may be caused by
the natural distribution of iron oxides within the soil and
rock or by the presence of buried iron or steel objects. (The
magnetometer does not respond to nonferrous metals such as
aluminum, copper, tin, and brass.)
The earth’s magnetic field behaves much as if there were
a ]arge bar magnet embedded in the earth. Although the earth’s
field intensity varies considerably throughout the United
States 1 its average value is approximately 50,000 qammas.*
The angle of the magnetic field with respect to the earth’s
surface also varies. In the U.S., this anole of inclination
ranges approximately 60 to 75 degrees from the horizontal.
The intensity of the earth’s magnetic field changes daily
with sunspots arid ionospheric conditions which can cause large
and sometimes rapid variations. With time, these variations
produce unwanted signals (noise) and can substantially affect
magnetic measurements.
If the magnetic properties of the soil and rock were
perfectly uniform, there would he no local magnetic anomalies
however, a concentration of natural iron minerals, or a buried
iron object 1 will cause a local magnetic anomaly which can he
detected at the surface (Figure 92).
An example of a magnetic anomaly indication over buried
drums is shown in Figure 93; the exact shape of which may vary
considerably. Typical magnetic anomalies at -1WS will range
from one to hundreds of gammas for small discrete targets,
depending on their depth. Massive piles of buried drums will
result in anomalies of from 100 to 1000 gammas or more.
* The unit of magnetic measurement is the gamma. Recently, the
gamma unit has been renamed the Nano Tesla. At this time, most
instruments are still labeled in gammas as are specification
sheets, existing literature and field data; hence, all
references to magnetic data in this document are expressed in
gammas.
164
-------
Figure 92.
There is a wide variety of magnetometers available commer-
cially; two basic types commonly used at HWS are the fluxgate
and the proton magnetometer. Typical equipment is shown i ’
Figures 94 through 97. A simplified block diagram of a magneto-
meter is shown in Figure A. In a fluxgate magnetometer, the
sensor is an iron core which undergoes changes in magnetic
saturation level in response to variations in the earth’s
magnetic field; differences in saturation are proportional to
variations in field strength. The electronic signals produced
by these variations are amplified, then fed to an amplifier,
whose output drives a meter or a recorder.
The signaL output of a single element fluxqate maqne—
tometer is extremely sensitive to orientation. To overcome
this problem, two fluxgate elements can be rigidly mounted
together to form a gradiometer. This gradiometer measures the
gradient of a directional component of the earth’s magnetic
Distortions in the earth’s magnetic field due
to concentrations in natural soil iron oxides
Cleft) and buried iron debris (right).
165
-------
100
80
60
40
20
0
0
U-
a
4-
0
C
C
0
C
C)
‘U
E
E
‘ U
0•
-®
Figure 93.
Station measurements of a magnetic anomaly caused by a buried steel drim.
-------
field. The qradiometer configuration of the fluxgate magne-
tometer, one which measures the vertical component of the
field, is the instrument that is discussed in this document
(see Figures 95, 96, and 97).
In a proton magnetometer, an excitation voltage is applied
to a coil around a bottle containing a fluid such as kerosene.
The field produced reorients the protons in the fluid; when the
excitation voltage is removed, the spinning protons reorient to
line up with the earth’s magnetic field. By nuclear precession
they generate a signal, the frequency of which is proportional
to the strength of the field. The signal is amplified anc9
the precession frequency measured by the use of counter circuits.
The frequency is electronically translated into gammas and the
output is fed to a digital display, a digital memory, or a strip
chart recorder. Proton magnetometers measure the earth’s total
field intensity, and they are not sensitive to orientation.
However, the proton magnetometer will cease to function when it
is used in areas with very high magnetic gradients (above 5,000
gammas/meter) which may be found in junk yards or near steel
bridges, buildings, vehicles, etc.
Portable cesium magnetometers may have application to HWS
investigations as recognized by the authors, hut at the time of
this writing they are unaware of any such successful application
and have not included them.
All types of magnetometers can he used for taking station
measurements in the manner shown in Figures 93 and 94. The
operator stops, takes a reading, records it, and moves on to
the next station. A great many of these station measurements
are required to cover an area. Recent improvements in portable
proton magnetometers have incorporated a built-in microproces-
sor, so that the variables of time, station number, location,
and magnetic intensity can all he recorded in memory for later
playback into a printer or a portable computer for processing.
This new system enables station measurements to be made more
rapidly. However, the minimum sample time for these ground-
portable proton magnetometers still ranges between 2 and 4
seconds. Because it requires many station measurements to
cover a site adequately, the station—by-station approach is
not often used at HWS.
An alternative to the station measurement method is to use
a continuous measurement magnetometer system, as shown in
Figures 95, 96, and 97. These units provide continuous measure-
ment of the gradient of the magnetic field as the operator
moves along a traverse line, and they provide considerably more
detail than can be obtained by station measurements. By proper
selection of the spacing between survey lines, total site
coverage may be obtained at reasonable cost.
167
-------
Figure 94.
High sensitivity (0.1 gamma) total. fiei proton
magnetometer being used for station measurements.
(Photo courtesy GeoMetrics.)
16P
.. .
.i’S .
-------
Figure 95. Fluxgate gradiometer. A continuous—sensing low sensitivity (20 gammas/meter)
magnetometer for shallow search.
-------
Fluxgate gradiometer: a continuous—sensina high
sensitivity (1 gamma/meter) magnetometer (Valley
of the Drums, Kentucky).
170
. .ti
S
7.
Figure 96.
-------
- .4
Figure 97.
Fluxgate gradiometer.
A vehicle—mounted continuous-sensing
high sensitivity
(1 gamma/meter) magnetometer (Love Canal, New York).
-------
Figure 98.
Simplified block diagram of a magnetometer. A
magnetometer senses change in the earth’s
magnetic field due to buried iron drum.
Amplifiers
o v id
Counter
Circuits
Ground Surface
I
172
-------
Two types of magnetic measurements may be made. A total
field measurement is made with a proton magnetometer by observing
the value of the magnetic field at a selected point (Figure 99a).
A gradient measurement is the difference between measurements
taken at two different points (Figure 9gb).
Figures 99h and 99c show two methods of taking a gradient
measurement. A total field proton magnetometer can he used to
take two total field readings; the difference between them is
the gradient of the total magnetic field (Figure 99b). The
gradiometer is a magnetometer composed of two separate sensors,
either fluxgate or proton. Both sensors respond to the total
field at their respective locations; the difference between
them is obtained electronically to provide a gradient reading.
Although the gradient may be measured in any direction, it is
commonly the vertical gradient that has been measured at HWS,
as shown in Figure 99c. For purposes of this document, total
field measurements will imply measurement using a proton magnet-
ometer, while gradient measurements will be accomplished by
either proton or fluxgate systems.
Several factors influence the response of a magnetometer.
The mass of a buried target is one factor; it will affect the
magnetometer’s response in direct proportion to the amount of
ferrous metal present. The depth of the target is an even more
significant factor, as response varies by one over the distance
cubed (lid 3 ) for total field measurements; this means that the
response will decrease by a factor of 8 if the distance between
the target and the magnetometer is doubled. If a gradiometer
is used, the response falls off even faster, at the rate of one
over the distance to the fourth power (l/d 4 ). If sensors of
identical sensitivity are used, the total field system provides
the greater working range.
Another factor which will influence the response of a
magnetometer is the permanent macnetism of the target.
Ferrous objects will have two superimposed magnetic
values; one due to induced magnetism and one due to permanent
magnetism. The permanent magnetism of an object is like
that of a bar magnet. Its value may be many times that of
the induced magnetism, which may add to or reduce the result-
ing anomaly. As a result, the value of a magnetic anomaly
may vary over a wide range, making the quantitative analysis
of magnetic data difficult.
In addition, the target’s shape and orientation
together with its state of deterioration also affect the
magnetometer’s response. (See corrosion rates for drums, in
Section VIII . .)
173
-------
Figure 99.
)
Comparison of total field and gradient measurements.
A. Total field measurement.
B. Gradient measurement consisting of 2 total field
measurement
C. Gradient measurement made with 2 sensors simul-
taneous ly.
Some magnetometer characteristics are summarized in
Table 4. The maximum sensitivities (0.1 gamma for total
field measurements, and 0.05 gamma/meter for gradient
measurements) shown in the table are rarely required at
EWS. In fact, excessive sensitivity can be a severe
handicap and may even inhibit acquisition of usable field
data if the instrumentation does not have available the
necessary useful dynamic range.
B
174
-------
Factors To Be Considered For Field Use
Project objectives, site conditions, equipment and field
procedures will all inf].uence the technical results and the
cost of a search and location mission. Some aspects of the
project which should he considered are: site access, level
of search (reconnaissance or detail), estimate of mass or
quantity of targets, maximum depth of search, and safety
requirements. Equipment features, including sensitivity,
susceptibility to noise, capability for recording, total
field or gradient measurement, continuous or station measure-
ment, recording capability, and hand or vehicle mode of
operation must be r ’atched to the job.
Although a reconnnaissance investigation provides 1.ess
than total site coverage, it is often more than adequate to
provide a sampling of site conditions. Detai]ed surveys
with a high density of survey lines provide a greater deqree
of spatial resoluton. The level of site coverage may he
increased by degrees until total site coverage has been
obtained. The use of a continuous gradient magnetometer is
preferred, as it will provide the highest level of lateral
resolution and will minimize unwanted responses so that
anomalies may be more easily detected.
If a high—resolution survey is required over a large
area, the benefits of using continuous sensors (Figures
95 & 96) and vehicle—mounted systems (Figure 97) are self—
evident.
The most difficult magnetometer survey task will he to
quantify the depth and number of 55—gallon drums or other
ferrous targets. Theoretically, the total number of drums
may he calculated from the amplitude of the magneti.c anomaly
(Figure 100), and their location and depth may he obtained
from the shape and width of the anomaly (Figure 101). Iowever,
because of the number of variables associated with target, site
conditions, and calculations, such results should be considered
only approximations. Actual results may vary by a factor of
2 to 10. Factors such as the target’s magnetic properties.
its geometry, orientation, deterioration, and permanent
magnetism are not considered in the nomograph of Figure 100.
Furthermore, Figures 100 and 101 address only discrete targets
such as single 55—gallon drums; the effects of large numbers
of randomly distributed drums are not considered. Because of
these many variables, high levels of accuracy should not be
expected in evaluations of the depth and quantity of drums.
To he realistic, quantities and depths should be stated in
terms of a range of values.
175
-------
TABLE 4. SUMMARY OF’ MAGNETOMF TER CHARACTERISTICS
TOTAL FIELD MEASUREMENTS
GRADIENT MEASUREMENTS
Most Sensitive
Susceptible to Noise
Less Sensitive
Insensitive to Noise
Improved Location
Station Continuous
NA - Not applicable
1 - Commonly used mode of operation
2 - Maximum sample time for portable proton ground magnetometers
presently range from 2 to 4 seconds.
Station
Continuous
I-
-4
0 \
Typical Sensitivity
FLUXGATE
NA
NA
YES
YES
0.1 gammas/meter to
20 gammas/meter gradient
PROTON
YES’
NO 2
YES
NO
.1 gammas total field
.1 gammas/meter gradient
-------
The magnetometer shown in Figure 95 is extremely insensi-
tive to nearby fences, cars or buildings; it can provide an
effective reconnaissance survey for shallow 55-gajion drums,
but would be ineffective for deeply buried drums. In the
latter case, a magnetometer with greater sensitivity should be
selected (Figures 94 or 96). However, using an instrument
with greater sensitivity than necessry can lead to excessive
noise in the data, which, in turn, will make analysis difficult.
If a magnetometer sensor is carried too close to the
ground, it becomes susceptible to noise produced by variations
in the magnetic characteristics of the soil. Raising the
sensor 3 to 6 feet off the ground can reduce or eliminate this
noise (Figures 94 & 96), but at the same time it may appreciably
reduce the target signal. Therefore, to minimize noise, a
proper balance must be struck between instrument sensitivity
and operating height.
Quality Control
The precision (repeatability) of a magnetometer survey
may not be a matter for concern if the survey is conducted
over a short period of time and the results are not to be
compared with subsequent surveys. Errors may be present due
to changes in the earth’s field ocurring over the day of
measurement or during the time interval between surveys.
Total field measurements may be corrected for these time
variations by employing a reference base station magneto-
meter; changes in the earth’s field are removed by subtract-
ing fixed base station readings from the moving survey data.
Gradiometers do not require the use of a base station, as
they inherently eliminate time variations in the data.
Accuracy: If semi—quantitative or quantitative results
are not needed in a survey (such as when merely locating an
ob:ject or defining trench boundaries), accuracy is of little
concern. However, if estimates of the depth and num1 er of
drums are to be made, instrument calibration is very important.
Quantitative analysis requires that field data be fitted to a
model for interpretation; therefore, the values of the magnetic
anomalies must be sufficiently accurate to do so.
Proton magnetometer sensors are inherently calibrated,
as their operation is based on nuclear precession; only their
crystal—controlled counters may require occasional factory
calibration. On the other hand, fluxgate magnetometers are
not calibrated; they will require calibration if accurate
results are to he obtained. (A laboratory calibration of any
magnetometer can be accomplished by using a standard magnetic
field created by a set of coils carrying a known current.)
However, a much more practical approach is a reference magnet,
177
-------
500
I
Figure 100.
Total field magnetometer response (in gammas)
for different target distance and mass. Due
to the many uncertainties associated with mag-
netic anomalies, the estimates shown in this
graph may vary up to an order of magnitude.
(Modified from S. Breiner).
which is an invaluable aid in the field. It provides a quick
way to verify instrument operation and to perform an in—field
calibration. This reference magnet will itself require cali-
bration to absolute standards at periodic intervals.
Noise
Noise is any unwanted signal or response, and a large
signal-to—noise ratio is desirable. Noise may be caused by
time variations such as the natural changes in the earth’s
field and by spatial variations. Spatial noise may be asso-
ciated with changes in local soil conditions or produced by
passing over ferrous debris.
200
I00
Meters
178
-------
C
A
Buried Dru
B
PLAN VIEW SHOWING
TRAVERSE LOCATION
A A’
c
Figure 101. Magnetometer response will vary considerably
depending upon traverse location and direction
with respect to the target.
A’
C D’
179
-------
The effects of time changes in the earth’s field can be
eliminated from total field measurements by using a second
magnetometer as a base station. The time changes sensed by
the fixed base station are removed from the values obtained
by the roving search magnetometer. The result of this
process is a series of measurements sliowinq only the spatial
changes in the magnetic field. (A gradiometer accomplishes
this process automatically.)
By lifting the sensor up off the oround and carryinq it at
some distance above the surface, as shown in Figures 94 and 96,
the noise due to natura] soil arid rock variations and small
particles of metal debris can he minirnizerl. At the same time,
the increased target—to—sensor distance will not appreciably
reduce the instrument’s response if the target, for instance,
is a massive pile of 55—gallon drums. However, if the target
is only one steel drum, the sensor’s response may fall ofF
dranatically as a result of its increased distance from the
target. In this case, the advantage of reducing noise must he
weighed against the accompanying disadvantage of decreasing the
instrument’s sensitivity.
Cultural features can cause large unwanted anomalies in
magnetic data. For example, a buried pipe may he the cause
of a large magnetic anomaly, but it can often be identified as
such and be separated from other targets. However, if a single
55—gallon drum is buried next to a large iron pipe, the drum
probably will not he identified as a separate target and could
remain undetected.
Noise interference from personal effects and clothino
may also be a problem. The solution is to eliminate all ferrous
material from the operator’s person. Steel—toed hoots and
some respirators are sources of noise, but they may he required
safety measures at certain locations. Noise from this equip-
ment must be minimized by keepinq the sensor as far from the
operator as possible.
If the magnetometer is mounted on a vehicle, the sensor
must he as far from the vehicle as possible, and/or it must be
compensated for the presence of the vehicle. Some influence
can always be expected from the vehicle; to minimize its
effects, survey lines should be straight, and they all should
run in the same direction to eliminate the directional effects
of the vehicle and provide for simple visual analysis of the
field records.
Data Format, Processing, Interpretation and Presentation
A magnetometer’s output will depend upon the instrument
lAO
-------
used. Audio signals, analog meters, diqital numeric displays
and recorders, and strip chart recorders are all commonly
used.
Proton magnetometers provide a numerical value which
can be recorded in a field notebook or directly on a map.
Newer equipment has an internal memory which stores field
data. This data may be retrieved as a printed tabulation or
plot, or fed directly to a computer for processing and plotting.
Raw values may be plotted directly on a map or profile line to
provide in-field quality control and initial interpretation.
Final presentation of data is typically limited to profile
lines locating anomalies or contour maps showing the location
of buried material.
The simple hand-held magnetometer shown in Figure 95
provides only an audio signal to the operator. This type
of instrument has a continuous response and can he swept
from side to side across a traverse line. The audio response
indicates the presence of a target, which can he marked with
a non-ferrous stake as the survey progresses. Locations of
anomalies may be recorded in field notes, but no output from
the magnetometer is recorded or noted directly.
The location of an anomaly can he determined in this
manner with reasonable accuracy. However, the angle of the
earth’s field must he considered, as the target may he offset,
not necessarily lying under the largest portion of the anomaly
(Figure 102). Further, the shape of a magnetic anomaly can
he complex, and in the vicinity of the target, it may vary
from one traverse to another (Figure 101).
A magnetometer with continuous recording capabilities
can he used to produce a strip chart or a digital record of
the field data. Such magnetometers provide the field party
with a graphic profile of the data and assist in assessing
signal—to-noise ratio, anomaly shape, and target location;
such records, thereby, provide a means of exercising quality
control over field data. The raw records can be used in the
field to locate buried drums, to define boundary limits, and
to provide estimates of the depth and mass of targets. The
same records may be replotted into final profile lines with
corrections for instrument range changes and spatial position
variations.
A number of processing options may he carried out on
magnetic data. They include:
o Corrections for instrument drift;
o Corrections for changes in the earth’s field;
o Filtering to remove noise:
181
-------
o Enhancement or removal of surface targets, or deeper
targets as required.
Data can he interpreted quantitatively to provide anomaly
locations along a profile line or burial areas on a map. Semi-
quantitative data for depth and mass (number of drums) can he
obtained by the use of a model (see Figure 100) and calibrated
instruments. However, error factors of 2 to 10 may occur in
such calculations.
Direction of Earth’s Field
North
-
Figure 102. Diagram of magnetic anomaly over burial trench.
Note that the peak anomaly may not necessarily
lie over the center of the trench due to the
angle of the earth’s field.
Raw magnetic data, such as a strip chart record of a
profile lines may he sufficient for final presentation (Figure
103) Simple maps may be drawn to show the concentrations of
suspected buried drums (Figure 104). If high resolution data
is available, a map can be contoured to provide more detailed
information. A graphic presentation may be made by compiling
parallel profile lines into a three—dimensional image of the
magnetic data (Figure 105).
‘S
•.. -.. .-.
- I
182
-------
Summary
A magnetometer responds to the presence of buried ferrous
metals. At HWS, maqnetometers may he used to:
o Locate buried 55—gallon drums;
o Define boundaries of trenches filled with ferrous
containers;
o Locate ferrous underground utilities, such as iron
pipes or tanks, and the permeable pathways often
associated with them;
o Aid in selecting drilling locations that are clear
of buried drums, underground utilities, and other
obstructions.
While several factors influence the response of a magneto-
meter, the mass of a buried target and its depth are the most
important. A magnetometer’s response is directly proportional
to the mass of ferrus metal present and varies by one over
the distance cubed (1/a 3 ) for total field measurements. If a
gradiometer is used, the response falls off even faster, as
one over the distance to the fourth power (i/c94). With sensors
of equal sensitivity, the total field system provides the
greater working range. Typically, a single drum can he detected
at distances up to 6 meters, while massive piles of drums can
he detected at distances up to 20 meters or more. There is a
wide variety of magnetometers available commercially; specific
performance is highly dependent upon the type of magnetometer
and the field conditions. While the number of drums may be
calculated, such results should be considered only approxima-
tions because of the numhet of variables associated with
targets, site conditions and calculations. Actual results may
vary considerably.
A magnetometer with continuous recording capabilities can
be used to produce a strip chart of the field data, which is
helpful in assessing signal—to—noise ratio, anomaly shape, and
target location, and provides a means of exercising quality
control over field data. This continuous coverage is much more
suitable for high—resolution requirements and the mapping of
extensive areas.
The effectiveness of a magnetometer can he reduced or
totally inhibited by noise or interference from time—variable
changes in the earth’s field and spatial variations caused
by magnetic minerals in the soil, or iron and steel debris,
ferrous pipes, fences, buildings, and vehicles. Many of these
problems can he avoided by careful selection of instruments
and field techniques.
183
-------
MAJOR ANOMALIES
MAJOR ANOMALIES
NO ANOMALIES
Figure 103. A single magnetic profile line showing a wide range of magnetic anomalies.
-------
Figure 104. Simple contour map of maqnetic anomalies shows relative concentration
of buried drums (buried 55—gallon drums are inferred).
I - .
-------
Figure 105.
Three—dimensional perspective view of magnetic profiles over a trench
containing huried drums.
I- .
0
-------
Capabilities
o Magnetometers respond to ferrous metals (iron or
steel) only.
o Individual drums can he detected at depths up to
E meters.
o Large masses of drums can be detected at depths of
6 to 20 meters.
o Magnetometers can provide a greater depth range
than metal detectors.
o Interpretations of their data may he used to provide
estimates of the number and depth of buried drums.
o They can provide a continuous response along a
traverse line.
o They may be mounted on vehicles for coverge of a
large site.
Limitations
o In general, magnetometers are susceptible to noise
from many different sources, including steel fences,
vehicles, buildings, iron debris, natural soil minerals
and underground utilities.
o Low cost units are limited in depth range (but their
limitations make them insensitive to many of the above
sources of noise).
o Total field instruments are also sensitive to
fluctuations in the earth’s magnetic field which can
seriously affect data.
o Data is of limited use in determining the number and
depth of targets.
o Complex site conditions may require the use of highly
skilled operators, special equipment, and the recording
and processing of data, along with skilled interpreta-
tion.
Examples
Determination of Drum Distribution——
An old covered dump in a marshy area was suspected of
containing drums of hazardous waste. Location of the areas
containing major concentrations of drums was necessary in order
to undertake a ground water sampling program. The area (300 x
400 meters) was surveyed, using a continuously—recording
gradiometer. Approximately 70 lines spaced 10 meters apart
were used to cover the site. The 10—meter spacinci was selected
because only large groups of drums were of interest.
Magnetic data from a typical profile line is shown in
Figure 103. t’lote the many strong magnetic anomalies on the right
portion of the line, indicating the presence of numerous ferrous
187
-------
SECTION X
APPLICATIONS
Summary of the Six Geophysical Methods
This section presents a summary of the six geophysical
methods discussed in this document. The following tables
highlight features of each method and are intended as general
guidelines to provide the reader with a capsule summary of the
capabilities and limitations of the six methods, including
factors which may aff ect their measurements. These tables are
based on extensive field experience, arid they present data which
will he applicable in most cases. However, the reader should use
them as guidelines, recognizing that exceptions may occur because
of the wide range of site conditions and project objectives.
Table 5 — summarizes the primary technical characteristics of the
six methods, including: mode of measurement, depth of
penetration, relative resolution, and data format.
Table 6 — outlines primary (more suitable or more commonly used)
and secondary (less commonly used or less effective)
applications of each method.
Table 7 — lists sources of noise which may affect the performance
and utilization of each method.
No single method, whether traditional direct sampling or one
of the contemporary geophysical techniques, will solve all site
investigation problems. All the methods discussed are founded on
sound scientific principles and can be extremely effective in the
field; but any of them may fail, when improperly applied or when
applied to the wrong objective. The methods and approach
discussed in this document have been successfully applied to a
number of site investigation problems as outlined in Table 6. A
large number of sites have been evaluated throughout the United
States with a diverse set of both hazardous waste and
geohydrologic conditions. In addition, the approach has been
used repeatedly in evaluation of new disposal sites, and to
evaluate conditions after clean up or remedial action has
occurred. By selecting the most suitable methods, combining
methods, and utilizing the synergistic benefits of an integrated
189
-------
TABLE 5 CHARACTERISTICS OF THE SIX GEOPHYSiCAL METHODS
i— a
t O
a
I. Ground Penetrating
Radar (G m)
RESPONDS TO
ChANGE IN
Coetplas Dialecteic Constant of
soil. roa. pots flu4is. end
nan-made objecia
Bulk sleemic conductivity of soil.
rock and poro fluids (Pore fluids
tend to dominate)
Sulk Eiscoicel resistivity of soil.
rock and pa. (bids (Por. fluids
rend to dominate)
Setsrsio velocity of soil a rack
which is rstaied to density and
elastic properties.
MODE or
MEASUREMENT
Continuous Profile .4 km /br.
dsraii - 0km /hr. rsonnrtalssancs
(Ground contact not necessary)
Continuous Profilss to .5 to i s
newts depth. Station aoaiuro
meets to i i to 10 eetcrs depth.
lott ie sounding capability
(Ground contact not necessary)
Station Measursrssnts for
profiling or soundutg (Must
have ground contact)
Station Measuranisnis
(Must have ground contact)
DEPT h Or
PENET IAT J OI 5
Ons to isn meters typical-
highly site npocitio.
Limited by fluids and soils
with high aiecirlcai con-
ductivity and
by line grain natnstsis.
Depth controiind by nyalen
coil spacing S to 60 enrrrs
typicsi
Depth contoliad by cisc- 1
teda spacing. Limited by
space available for stay.
instunrent power and
sensiiivtty beeorsa ins-
portant at grestar dapth.
Depth limited by snap
length and anm y nourca.
Greatest of all aim
geophysical rsaihcda
Excellent iaierai
iuiion. Varticia
resoituion of? layers.
Thin layers may noi b.
date cied.
Good vertical resolu-
tion of 3 to 4 layers.
Thin layers may not
be dstacted.
Good vertioni resolu-
tion of 3 to4 isyere.
Seismic valacity must
increaas with depth —
thin spore may not be
detected.
Picture-iiko graphic diepiay.
Anstog taps
Digital tape
Numerical values of conductivity
from station meIsurerrenis.
Stripohert and/or magnetic
rcccrdcd data yields continuous
prolitirrg.
Numeric values of voltage
airreni end dinsnsions of ezra ,.
Can plot pro(iio or sounding
curves from raw data.
Numgric values of time end
distence. Can piot T/D graph
boa raw dets
S. Metal Detector (MDI
S. Magnetometer (MAG)
tiectical conductivity of Venous
and non-Venous metals
Msgnelio susceptiblli ty of
ferrous metals
Continuous (Ground contact
not neoensaryi
Cgntimaoue Taial Pieid a
gradient measusemests.
Many Instoamsets are iimitad
to station meesurmeecia.
(Ground contact not necessary)
Stogie 55 gal. dnrmssp to 2 Very good ability to
I meiers toonie targets
Maesiva piles Si gel. dnrers
up tab meters
Single 5$ gal. dmzmuP to 2 Good ability to
5 meters locate targeis
Massive piles 55 gsi. dnrms
up to 20 metars
Relutivo response from audiof
visual indicators (wsy record
data)
Non-quantitative response hoD
audio/visuai indicator..
QuantItative lnarrumertwrsvida
meter a digitsi display (may
record date)
I. Depth is else related to equipment capability.
hiETiiOfl
2. Eiecfrmnaomniics (EM)
3. Resistivity Sounding (Ru)
4. Seismic Refraction
RESOLUTION RAW a tm rORMAT
2. Depth is very dependent upon Instumeni used
-------
TABLE 6 TYPICAL APPLICATIONS or THE SIX GEOPHYSICAL METHODS
Application Radar EM Res Seismic MD MAC
NATURAL CONDITIONS -
Layer thickness and depth of soil and rock 1 2 1 I NA NA “
Mapping lateral anomaly locations 1 1 I I NA NA “
Determining vertical anomaly depths 1 2 1 1 NA NA
Very high resolution of lateral or vertical anomalous conditions I 1 2 2 NA NA
Depth to Water table 2 2 1 1 NA NA
SuB-suRFACE CONTAMINATION LEACHAITSIPLUMES -
Existence of contaminant (Reconnaissance Surveys) 2 • 1 1 NA NA NA
Mapping contamInant boundartaa 2 • I NA NA NA
DetertsIning Vertical extenrof contaminant 2 • 2 I NA NA NA
Quantify magnitude of contaminants NA 1 I NA NA NA
Determine flow direction 2 • I I NA NA NA
Flow rata usIng 2 measurements at different times NA I I NA NA NA
Detection of organics floating on water table 2 • V P NA NA NA
“0 DetectIon & Mapping oi contaminants within unsaturated zone 2 1 I NA NA NA
I -
LOCATION AND BOUNDARI1E OF BURIED WASTIZ -
Bulk Wastes I I I 2 NA NA
Non-Metallic containers I I I 2 NA NA
Metallic Containers
-ren’ous 2 1 NA NA I 1
-Non-ferrous 2 I NA NA 1 NA
Depthofburial 2 2 1 2 P 2
UTIL !TiIZ -
Location of pipes, cables, tanks I 1 NA 2 1
Identification of permeable pathways associated with loose
fill in utility trenchaa I I NA 2 i
Abandoned Well Casings NA NA NA NA I I
SAFE l Y -
Pro-drilling site clearance to avoid dnnna, breaching aenchas, etc. 1 I 2 NA
- Denotes primary use
2 - Denotes possible applicntiona, secondary use; howaver, In some special cases 2 may be the only effective epprcech due to circumstances.
NA — Not applicable
‘ LImIted applications
“ Not applicable in the context used in this document.
-------
TABLE 7. SUSCEPTIBILITY OF GEOPHYSICAL METHODS TO ‘NOISE’
this table shown the incc;ubility ol the QcophySlCel ncthc4s to uAriows tort-is ci 1%0i50 witch amp trmllucnco
li.IJ opelillon. renultlnf djt3 .nd sub ,a uOnt IniotprctSLIOfl.
ZOVRCt
or !iOi
etsi tr iry ZZI5;.ilc hID 11 1.0
Oui1e Pipes
will detect.
but rmay
alfici duwi
I
only U close
to pipe
I
onlj If survey
Is parallel end
close b
2
onI It sur vey
I s direct ij
over
I
any rmct3l
pipes
I
jiccl pipcs
only
MalSi fcncos
NA
I
anly It close
t o fence
P
only It sursey
line is p3r lic
& close to
fence
. A
2
If
r me3rb
1
sical tcncci
cnl#
Oueihcsd ‘.lr.u
ip c.uenJIn i
2
enl If
ummslilelded
nriennss
era i.ncd
1
l.A
l.A
NA
P
j Ot- WI racy
rc pcr td
Grourd Vitlutions
NA
hA
?A
I
1. 5
:A
l.ir orr.e Llcctio-
rtaqnctlc N tIs.
NA
2
2
NA
2
Ito 2
(Car iia LitId
Chan cd
Ground Curreni
and Votta s
NP.
NA
P
NA
NA
NA
iron, 2
only If
unyhi ld ed
uninnites
era used
z
(Wind rolsol
NA
•.A
hlet l Iron Buiktjogs. 2
Vemicicn. etc. oily if ncsrby
& urteftielded
ertoimnas
a, u ,ed
2
only if nearby
2
oft!? ii rc:rby
‘ A
P
only I I ncarby
2
crlp it ncy;by
Srsali Itotalite Oobrtj I 2
on Surface Near I
Surface (nails wire •
coa ther i sers l
NA
NA
NA
1
1
ferrous natal
only
I
Larq. MelallIc Debris 2
on Surface iaer
Sur1 cs fOrums. Drum
Covcrs, etc.) .
I
2 2 NA
I
I
1
I
fnryoue metal
only
Susceptible to noise boat I
Oround conieci/ Elecamods P NA I 2
problems
NA
A
- lcry SuecnpUbia
P MIn Problcm
NA - hot Appilcabi.
192
-------
systems approach, high levels of accuracy arid cost—
effectiveness can he achieved in subsurface investigations
of HWS. The following discussion will illustrate some of the
trade—offs or compromises that may be required in applying the
methodology to HWS investigations.
Detecting and Mapping Conductive Plumes
Table S compares the capabilities and limitations of the
EM arid resistivity methods for detecting and mapping conductive
plumes. Table 6 indicates that radar is a less effective or less
commonly used means of measuring contaminant plumes. This does
not mean that radar cannot he used to map the top of a shallow,
electrically—conductive plume; it can. Most of the time,
however, it will be more productive to use EM or resistivity for
that purpose. Even so, if information concerning shallow soils!
cementation and other variables is considered to he an important
factor in assessing the migration of contaminants, the radar
method might be used to complement EM or resistivity data.
Furthermore, both EM and resistivity may he rendered totally
ineffective by noise from a variety oE sources; for example, the
presence of nearby railroad tracks or buried pipes may be found
to make measurement impossible. In that case, the radar method
might be used with great success, where the other methods had
failed.
The final decision as to which method to use should be made
only by those with a comprehensive understanding of the entire
array of methodologies. In many cases the program should be
designed to be flexible with respect to the decision—making
process, so that a final determination can he made in the field,
after conditions have been examined first—hand.
In weighing the use of resistivity versus EM on a particular
project, it must be decided whether profiling or sounding data are
needed, and how much of either will he required to produce a
statistically valid measurement. The required level of detail,
quantification of results, and data format should be established
before work is begun.
The resistivity arid EM methods are compared in Table 8.
Both resistivity and EM are capable of vertical sounding;
however, the vertical resolution of the EM method is limited.
The depth to which sounding data can be obtained with resistivity
is virtually unlimited: depths of 100 meters or more are easily
obtained. EM, however, is limited to approximately 60 meters
depth, based upon the equipment discussed in this document.
Therefore, the resistivity sounding technique is the preferred
approach if detailed vertical information such as depth to
193
-------
TABLE 8. COMPARISON OF REStSTIVITY AND ELECTROMAGNETIC METHODS
RESISTIVITY rLrCTRc )MAGNE1’ICS -
Yes (limited number of depths available)
Depth of Sounding Measurement
Not Liuni ted
60 meters n .n imum with equipment
discussed
Profile Station Measurements
Yes
Yes - to 60 meters depth
Continuous Profile Measurement
No
Yes - to 15 meters depth and at speeds
up to 8 Km/hr
Relative Lateral Resolution
Good in Profile Mode
Good in profile mode with station
measurements. Excellent in continuous
profile mode.
Relative Speed of Measurement
Good
Very Rapid
Total Site Coverage
Not Generally Economical
Feasible at reasonable cost
Susceptible to Noise and
Burled Pipes/Cables
Yes
Yes (continuous measurement
aid identification of pipes nd cablc )
Electrode Contact Problem
Yes
No (operates through dry sands • concrete
blacktop. etc.)
Overall Length of Wenner Array
or Coil Separation for given
Sounding Depth
6 - 12 times Depth *
of Interest
Less than 2 times depth of interest *
Length of Array or Coil Typically 4.5 to 6 times depth* 2/3 Dcpth *
separation for given profile (Minimum of 3 times depth)
* Comparison of depths of resistivity and EM measurements are only approximation’, because of
differences in contributions from various (lcpLli inherent in each IIicllle)(k.
Yes
Vertical Sounding Capability
I- ’
0
-------
bedrock, depth to water table, or depth and thickness of the
soil/rock layers is required; or when data deeper than 60
meters is required.
Although both resist:.ivity and EM can he used For profiling
work, EM is limited to about five discrete profi).ing depths to
60 meters. This limitation of EM is more than made up for
by its capability for rapid measurements, and continuous
profile measurements at up to 15 meter depths. Continuous
profiling measurements have extremely high lateral resolution
and can he run at speeds from 1.5 to 8 km per hour, depending
on the detail required. The resistivity method is not capable
of producing these continuous profiling measurements, due to
the need to set electrodes in place to make contact with the
ground.
Both resistivity and EM measurements can miss a subsurface
feature or contaminant plume if the station or profile line is
in the wrong location; however, with the capability of continuous
EM profiling, a site can be covered by a number of lines with
very close spacing, to approach total site coverage. While
the high lateral resolution inherent in a continuous EM measure-
rnent can be approximated by a higher—density resistivity survey,
cost and time considerations do not make this a very practical
approach. EM measurements are preferred for profile work,
particularly where continuous sampling can be employed.
Both resistivity and EM methods are susceptible to noise
due to buried pipes, cables, fences and other metallic cultural
features. They are also susceptible under some conditions to
electromagnetic noise created by powerlines. Because the number
of resistivity stations is usually less than is used for EM,
it is difficult to assess if cultural noise is affecting a
particular resistivity station measurement. This leaves a
degree of uncertainty as to the validity of that data. When
using EM many more stations are used or continuous data is
acquired which aids in evaluation of noise interference.
When noise can he recognized, it is often possible to remove
it or take it into account in data interpretation.
The requirement of ground contact in the resistivity method
creates additional problems not encountered with the EM method.
Because electrodes must he driven into the ground, conducting a
resistivity survey over a concrete or blacktop surface or hard
soil can be a difficult task. If the surface material is
resistive (dry sand), the electric current will be difficult to
inject. Furthermore, the resistivity method is disproportion-
ately afEectecl by resistivity variations in the surface soils
near the electrodes.
195
-------
Another factor that influences the choice and applicability
of the methods and their spatial resolution is the physicaJ
length of resisi.tivity arrays, or the EM coil, separation, required
to make a measurement to a given depth. The overall resistivity
array length will typically he 6 to 12 times the depth of
interest with a Wenner configuration. Information to a depth of
20 meters will, therefore, require an array of 120 to 240 meters
in overaJ.1. length. Finding accessible space on a site to place
this long array may be difficult. Further, longer arrays are
more likely to be influenced by noise factors and electrode
contact problems. On the other hand, the overall length of the
EM coil separation will he less than two times the depth of
interest. Again, in the case of data to 20 meters depth, the
EM coil spacing will be between 20 and 40 meters, as compared
to 120 to 240 meters for resistivity.
Array length or coil spacing also determines the volume
of subsurface that is sampled; the resistivity method integrates
a larger voluine than does the EM method. The EM method will,
therefore, provide an improvement in lateral. spatial resolution,
as well as the capability to work in tighter quarters.
In summary, the resistivity method is the preferred tool
for obtaining vertical sounding information; the EM method
provides the better tool for profiling. If high—resolution
to depths of no more than 15 meters is require 1, EM is,
preferred over the resistivity method because of its continuous
profi].ing capabilities.
Each technique is susceptible to noise from a variety of
sources and there will he instances where one technique will
fail to function at a site due to noise, while the other
technique will function perfectly. For example, both methods
have been used successfully under high-voltage transmission
lines, and each method has at one time or another failed under
such conditions. To be successEul in carrying out a field
investigation under such conditions requires that both
resistivity and EM options he available to the field party.
Comparison of Methods to Detect Buried Metals
A comparison of metal detector and magnetometer techniques
for use at hazardous waste sites is shown in Table 9. A metal
detector will respond to both ferrous and non—ferrous metals,
while a magnetometer will respond only to ferrous metals.
Therefore, it is necessary to determine what metals may be
present in order to select the proper instrument.
The metal detector is a continuous-sensing device arid may be
used on continuous traverse lines, or may be swept from side to
196
-------
TABLE 9. COMPARISON OF METAL DETECTOR AND MAGNETOMETER METI-IODS
i.jr.ji..J1JLJ —M [ TJ’LL P .1 L&..-1 ttr i’ . .n.!i, . L. 1 Tfl L”
DETECTOR — M G TOf.1LTfl1 i _
Detects Yes Yes
Ferrous Metals
Detects
Non-ferrouz Metals
Yes
No
Responds to
Surface area
of target
Mass of target
Pro )1des continuous
coverage
Yes
Yes (some equipment
limited to station
mee surement)
Define boundaries of
buried materials
(lateral resolution)
Very good
Good
Depth of detection
Relatively shallow
Single Drum up to 3 meters
Massive piles of drt ms
up to 6 meters
Shallow to deep.
depending upon
sensitivity and
configuration
Single Drum up to
6 meters
Massive piles of
drums up to 20 meters
Noise prablerns
Susceptible to metallic pipes •
fences, vehicles and surface
trash, as well as some soil
conditions
Susceptible to ferrous
pipes • fences • vehicles
and surface trash as well
as some soil conditions
Ability to quantify data Very limited capability Limited estimates of
depth and quantity
197
-------
side to cover an area. Some magnetometers are also capable
of this continuous coverage, while many commonly available
magnetometers are limited to taking discrete station measure-
ments. Recent improvements in magnetometers allow fairly
rapid station measurements to be taken. However, for small,
discrete, critical targets, continuous magnetometer coverage
may be required to provide sufficient resolution and greater
probability of detection.
Metal detectors have relatively shallow depth-sensing
capability. A single 55—gallon steel drum may be detected at
depths up to 3 meters, while massive piles of steel drums may
be detected at depths up to 6 meters, depending upon equipment
sensitivity. Magnetometers can sense a single steel drum to a
depth up to 6 meters and a massive pile of steel drums to a
depth up to 20 meters.
Metal detectors provide reasonably good spatial resolution
to pinpoint the location of a target. Magnetometers, however,
do not provide the same level of definition of target location
because they are affected by the dip of the earth’s magnetic
field, and the shape of the magnetic anomaly is more complex.
Both metal detectors and magnetometers are highly
susceptible to interference from nearby metallic cultural
features such as pipes, fences, vehicles, metallic surface
debris and even some soil conditions. Any of these factors
can produce an erroneous response from the metal detector, a
response which may be incorrectly interpreted as a subsurface
target. Because metal detectors are relatively short—range
devices, they can be operated closer to such sources of
noise than can most magnetometers. Proton magnetometers are
susceptible to interference from high magnetic gradients and
nearby power lines, whereas fluxgate gradiometers do not
suffer from these shortcomings.
The metal detector outputs are usually qualitative and,
therefore, have limited capability to evaluate the size and
depth of targets. Magnetometers. because their output can
be calibrated and equations are available, can provide data
for estimating the depth and number of drums. Due to the
wide range of variables which may influence these instruments,
any estimate from metal detector or magnetometer data regarding
the depth and especially the quantity of drums should be
considered an approximation.
In summary, both the metal detector and magnetometer
respond to ferrous metals, but only the metal detector will
respond to non—ferrous metals as well. The metal detector
is normally limited to detecting metal objects lying at rela-
tively shallow depths, but the magnetometer can detect metallic
198
-------
objects buried at deeper levels. The MD is capable of pin-
pointing the location of a buried object with somewhat greater
accuracy than the magnetometer. Careful measurement and use of
combined data from both MD and magnetometer will aid in estima-
ting the depth and quantity of buried drums, and will often allow
reasonably accurate estimates to be made.
In planning metal detector and magnetometer surveys, an
estimate must be made of what it is the investigator is looking
for, and its estimated depth of burial. For example, looking for
single isolated drums at depths of 10 meters is an unreasonable
survey requirements due to the depth limitations of both instru-
ments. The area to be surveyed and the spatial resolution re-
quired should be considered in deciding how best to provide a
statistically valid measurement. If the objective of the program
is to provide a first-approximation assessment for large burial
trenches, the sampling grid spacing can be increased. In an
extremely critical situation where buried materials might inter-
fere with a drilling operation, the survey grid will be tightened
up so that overlap between the survey lines occurs, to provide a
measurement with a high margin of safety.
Use of GPR to Locate Buried Drums
Radar can be and has been used to locate buried steel drums
(Figure 106). However, if soil conditions are not favorable for
radar penetration, or if the relationship between the orientation
of the buried drums and the radar antenna is not optimal, or if
too much noise or too many subsurface reflections from other
sources are present in the data, the drum(s) will not be detected.
Furthermore, there are many sources (other than drums) which
produce hyperbolic reflections. The presence of a hyperbolar,
therefore, does not inherently imply the existence of drums. If
the site contains a buried pile of drums, the composite reflec-
tions will be very difficult to identify as drums. However, it
will certainly be possible to say that an anomalous condition
exists. Since there are other methods and instruments, such as
metal detectors and magnetometers, to detect buried drums with
much greater certainty, even when soil conditions are bad, it
would seem prudent to consider these two methods first. On the
other hand, if site conditions (such as proximity to a steel
building) make the use of a metal detector or magnetometer impos-
sible, radar does provide a secondary alternative. If the depth
of a drum, or depth to the top of a pile of buried drums is
required, radar may provide estimates with a high level of accur-
acy than could be derived from metal detector or magnetometer
data.
1 98a
-------
—
—
.
I
I
•
•
i
.
I
I
•
I
Figure 106. Radar record over three buried 55-gallon steel
drums.
199
Three Buried 55 Gaflon Drums
1 r
1 ‘
-------
Site Investigation Considerations
This section outlines some procedural steps and variations
in conducting a geophysical survey of a hazardous waste site.
Figure 107 shows some irriportanb resources and tools which may be
used efEectively for the HWS investigation.
While this document deals with only six of the contemporary
surface geophysics methods shown in the lower center portion of
Figure 107, background information, traditional methods 1 computer
and analytical capabilities, technical experience and
professional and technical personnel are all important factors.
Just as the geophysical methods provide synergistic support to
each other, all of these factors provide synergism to the
project. In addition to these technical components 1 other
factors affecting the hazardous waste site assessment plan are
shown in Figure 2. Factors such as budget, natural site
conditions, cultural features, hazardous waste type, access,
positioning, safety considerations, logistics, legal
requirements, social/economic consicierations——all will influence
the planning, execution and results of a hazardous waste site
investigation.
Some site investigations may be relatively straightforward 1
consisting of only a few days of on—site effort, while others
may be more complex, requiring many weeks. The followina
example (addressing the geophysical effort only) illustrates
some considerations which may be required for various levels
of effort.
Small or Simple Site Investigation
A simple site investigation may require only reconrtaisance—
level geophysics with little if any subsequent interaction.
Such an investigation (outlined below) might consist of three
phases: (1) planning, (2) field operations, and (3) analysis
and report.
The field investigation may consist of:
o Initial site characterization (establishing background
values and evaluating cultural features, noise, etc.)
o Establishing survey grids
o Data acquisition and quality control
o Direct sampling (nct required on all projects)
o Analysis and report.
A more extensive effort may be required for a large or complex
project 1 an example of which is shown below.
Major Site Investigation
Planning
o Establish objectives
o Review existing data (aerial photos)
200
-------
Technical resources and tools which may he appliecl to subsurEace investigations
at hazardous waste sites.
0
I - .
Figure 107.
-------
o Visit site
o Establish field survey requirements
— Coverage considerations
- Depth and resolution
— Determination of techniques to he employed
— Accessibility
— Safety
— Logistics and coordination
Field Operations
o Reconnaissance
- Safety considerations
— Site familiarization
- Initial site characterization (establish
background values, and evaluate cultural
features, noise, etc.)
— Check fit of plan to actual conditions
o Survey
— Establish survey grids
— Data acquisition and quality control
— Initial direct sampling (soil and/or water
samples)
— “Fill—in” data acquisition (geophysical!
direct) based upon preliminary findings
o Design and execution of major direct sampling, and
laboratory analysis program
o Review and integration of all data with analysis
Final Analysis and Report
It is essential to clearly establish and periodically
review the objectives of a project. Initial efforts in the
planning phase would include review of existing data as a
critical part of establishing the objectives, especially in
major site investigations. Generally, a considerable amount of
information is available and can he obtained from a variety of
sources, such as topographic maps, aerial photos, local USGS
office publications, and government soil publications. Many of
these readily available documents contain a wealth of information
and are worth the time it takes to obtain and review them.
A site visit provides information for use in optimizing the
field investigation plan and in safety planning. Those personnel
involved in planning the field operations should participate in
the site visit whenever possible.
The area to be included in the site investigation and the
density of site coverage are important factors to be considered
when establishing survey requirements. It is often imperative to
survey an area larger than the actual site itself in order to
202
-------
Some site investigations require that geophysical survey
lines be referenced tO the legal land boundaries, while other
investigations do not require establishing any survey base. For
example, the boundaries of a burial trench delineated by a metal
detector may be marked in the field without relating the trench
locationto legal land boundaries. The necessity to have a
geophysical survey grid tied into a formal land survey ususally
depends upon the technical and legal requirements of the project.
Even though a formal land survey may not be required, all
on—site work should be tied to local references, so that any part
of the survey can be checked or repeated, or an anomaly located,
with sufficient accuracy. - In many cases, however, the total
cumulative error in positioning can easily be large enough so
that the location of a single drum, or the definition of the edge
of a trench, may not be accurately indicated. This will require
that the geophysical team support any remedial action work, or
stake it out in the field directly, so that exact locations may
be obtained.
Sampling of spatially varying data may be accomplished by
discrete, as well as continuous measurements. If we can estab-
lish the size of the smallest feature in the data that will be of
interest, we can indeed design a survey so that adequate resolu-
tion can be obtained by discrete station measurements. Before
the survey, however, we will not often be able to accurately
estimate the minimum sample distance. Hence, if our estimate is
sufficient]y in error, our data will be in error. (Review con-
cepts of discrete and continuous measurements in Section III,
Figure 17.) To minimize the possibility of making such errors,
to achieve maximum resolution, and to minimize project costs,
continuous methods are recommended whenever possible.
Basic quality control for the field survey involves a number
of factors i c]uding:
o Checking that instrumentation is working properly;
o Assuring that the positioning of stations is sufficiently
accurate;
o Noting stations accurately on recorded data;
o Noting unusual conditions;
o Monitoring signal—to—noise ratios;
o Plotting data in the field;
0 Evaluating initial results for reasonableness.
Quite often, only the relative values from the geophy-
sical data are needed to identify a potential problem area.
In such cases we may not- be concerned with the repeatability
or accuracy of the measurements, but only with the finding
that an anomaly exists (e.g., ugh conductivity). From such
information we can further evaluate the anomaly by
20 3a
-------
establish background values, local trends, and develop the
overall “picture”. Coverage of an area much larger than the site
itself is common in plume measurements and is an important factor
to he considered in planning, logistics and budgeting.
The resolution requirement will determine the density of the
site coverage. A survey of insufficient density may miss desired
information, whereas an excessively high—density survey results
in unnecessary expenditures.
The accessibility of a site will affect the time it takes to
conduct a survey. Trees, brush, trash piles and property fences
can all cause serious delays for a survey team. If accessibility
is a problem, trees and brush can sometimes be removed, or
surveys can occasionally be designed around them. Permission to
enter the property should be acquired prior to on-site
activities to ensure optimum utilization oE the field party’s
time.
Providing for personal safety is always of utmost importance
when designing a field survey. A safe plan of investigation must
be preceded by a thorough evaluation of existing data and an
on-site reconnaissance to reveal safety hazards requiring
special attention. Depending on the outcome of this evaluation,
proper protective headgear, eyewear, footwear, clothing arid
respiratory equipment should he utilized. If safety precautions
are necessary, they will certainly slow down field operations,
particularly in hot weather. Decontamination procedures will
also add time to a field survey. Other factors to consider are
the safety hazards involved in use of the geophysical equipment,
such as the high voltage arid currents associated with resistivity
measurements, or the danger of injury from explosives, if they
are used in a seismic survey. None of the saEety hazards
associated with the use of geophysical equipment poses a serious
problem to an experienced field crew, hut they all may become
critical considerations if inexperienced persons are in the
field.
After the field requirements have been established, the
actual field operation begins with a reconnaissance effort
incorporating general familiarization with the site, a review of
safety considerations, and a check of the “fit” of the plan to
actual site conditions. The survey work can then begin.
Initially, the geophysical survey should establish both nearby
and background values from off the site to aid in understanding
the setting and to have values for comparison. In addition, the
potential for cultural interference must he evaluated. By means
of this initial reconnaissance, the on—site geophysical project
manager can usually obtain a quick overview of natural site
conditions and cultural activities which might affect subsequent
operations.
203
-------
resistivity sounding, drilling, trenching, or whatever
method is appropriate to the mission.
When more than one geophysical method is being used, and
if surveys are to be repeated at a later date, quality control
and repeatability of the measurements also becomes an important
consideration. In some cases it may he desirable to establish
a background test site so that all instruments on subsequent
surveys may be referenced to this standard. Then if changes
in seasonal conditions (temperature, snow cover, frost, rain-
fall, etc.) have occurred, the extent of their influence can
be evaluated and compensated for by corrections to instruments
or data. The level of quality control required at a site will
vary considerably depending upon project requirements. There
is little need for a high—level quality control program if the
project objectives are straightforward, such as a metal detector
survey to identify burial site boundaries. Overkill here will
simply add needless cost to the program. But when legitimately
required, an in—field quality control program should he estab-
lished, ensuring that only those procedures that are considered
necessary will he performed.
Correlation with existing data (drill logs, ground water
chemistry, etc.) may he done before, during, or after the
geophysical survey. Sufficient data may already be available
from existing monitor wells or soil analyses. In some cases,
additional direct sampling may he required in anomalous areas
identified by geophysical methods. Many times, however, the
geophysical survey will he carried out with little or no exist-
ing site information; in those instances, geophysical data can
be used effectively to locate monitor wells or other direct
sampling stations.
In some cases, technical, safety, and legal considerations
may preclude direct sampling and it is frequently necessary
to determine subsurface conditions without their benefit. For
example, the survey objective may be to define the location
and boundaries of buried materials without drilling because
of the high safety risk involved. When this situation occurs,
a systematic survey using multiple geophysical methods often
provides the investigator with the ability to semi-quantify
subsurface conditions without direct sampling. In many
instances, a highly accurate evaluation has been made without
direct sampling; in others, virtually no conclusions can be
drawn without further direct investigation. The results of
surveys made without actual direct sampling are, of course,
more speculative, and the limits of accuracy from such a survey
should he clearly recognized by all concerned. When direct
sampling is required, considerable care must be exercised to
avoid creating problems such as might result from drilling
204
-------
into hazardous waste or drums. Of course drilling operations
for the evaluation of natural subsurface conditions will
usually present no safety hazards.
Preliminary data analysis is oFten done in the field.
The analyses will often indicate the need for additional
measurements, and will aid in the selection of additional
geophysical methods. Final analysis will usually he done in
the office with the support of manual an(1 computer calculations.
Although data analysis can yield a solution from a
geophysical measurement, the solution is not unique. Other
information must be integrated to arrive at an assessment which
portrays, in geological rather than physical terms, actual site
conditions. This process requires a trained and experienced
interpreter. The more complex the site conditions or the overall
problem, the greater the level of skill that will he required.
In many cases, experienced personnel can accomplish this
i.ntegration and interpretation process quite rapidly and
eFfectively.
While experienced personnel may be able to interpret much
of the required information directly from the raw data, computer
processing can he a significant aid in its analysis. Processing
is also used to improve the presentation of the data, so it may
he better understood by persons not familiar with geophysics.
Computer processing generally is used to:
o Assist in handling larger quantities of data.
o Apply corrections to the data (e.g.,, for non—linearities
and calibration).
o Compensate continuous data for spatial non-linearities
due to variations in speed when traversing the survey
line.
o Remove cultural responses such as pipelines, etc. This
cleans up data which may otherwise he extremely complex
to interpret.
o Perform modeling calculations to aid interpretation of
the data (e.g., forward and inverse calculations for
resistivity sounding interpretations).
o Evaluate data amplitude, frequency, or phase.
o Process the data to improve technical aspects or visual
presentations.
o Correct or exaggerate vertical or horizontal scales to
present data in the best visual format.
o Plot stacked profile lines.
o Contour data.
o Create three—dimensional plots, with various viewing
directions and angles.
o Filter data to eliminate unwanted noise’.
205
-------
o Overlay multiple sets of data, such as maps, site plans,
geophysical data, etc.
o Analyze multiple data sets, for correlation and
statistical trends.
Finally, all existing and new field data is reviewed and
a final report written. Since there is generally no unique
solution to he found in a given set of geophysical data, the
mere acquisition of data (through the use of geophysical methods)
does not in itself provide the solution to a site assessment
problem. Interpretation of the data is required, and the
accuracy of the interpretation will depend on the training and
experience of the interpreter. The more complex the problem,
the greater are the demands on this skill.
Examples of Field Investigations
The following examples describe the use
oE geophysical methods at a variety of hazardous waste
sites. The examples are taken from actual site investi-
gations. These five cases include:
1) Investigation of Natural Setting Prior to Construc-
tion of a Disposal Site.
2) Mapping and Characterization of Bulk Buried Wastes.
3) Delineation of Trench Boundaries.
4) Mapping and Assessment of Landfill Leachate Plume.
5) Locating Monitor Wells at an Uncontrolled Hazardous
Waste Site.
Investigation of Natural Setting
In selecting a site for waste disposal 1 an evaluation was
required of a natural clay layer believed to exist over lime-
stone bedrock. Data on the lateral extent, continuity and
thickness of the clay layer was needed to evaluate the effectiv —
ness of the clay in preventing contamination of the underlying
limestone aquifer.
Seismic refraction and resistivity sounding surveys were
conducted over the area. Samples from several augured holes
were obtained to determine the physical and chemical properties
of the clay, and to correlate with the geophysical data.
Seismic refraction and resistivity produced a faster, more
reliable survey at lower cost than could have been achieved
with just monitoring wells. Borings could not be made through
the clay layer; for if improperly sealed, they would have
created potential pathways for the migration of contaminants.
Ten seismic refraction lines were located over the 25—acre
site to obtain adequate coverage. Since explosives were
prohibited at this site, a hammer was selected as a seismic
206
-------
source for tile shallow work. The use of a sledge hammer source
limited the seismic station lines to a length of approximately
30 meters, providing information to a depth of 10 to 12 meters
Five-foot. geophone spacings were used to obtain detailed
information regarding site variability. Initially, forward and
reverse seismic lines were made to check for possible dip oE the
underlying strata. From the first few seismic lines, it was
found that the bedding was horizontal and the remaining stations
were run without the reverse line, in order to reduce costs.
The seismic data in Figure 108(a) indicated a three—layer
system which was tentatively identified as sand, sandy clay and
massive clay (this data is shown in detail in Figures 77 and 7 fl.
The length of the seismic line was not sufficient to detect
the top of the limestone because a small hammer was used;
however, calculations provided the minimum depth to the top
of the limestooe, and a minimum thickness of the overlying
clay layer was obtained. The lack of geologic scatter in the
seismic data indicated that the materials were fairly uniform.
Five resistivity sounding stations overlapped the seismic
stations. Wenner array soundings were carried out to 100—meter
electrode, or “A” spacing, and sounding data was obtained to a
depth of “A” approximately 30 meters. The resistivity data
was of “text book” quality with virtually no geologic noise.
The interpreted resistivity data indicated a four-layer
system (Figures 63 & 64). The first three layers corresponded
to the sand, sandy clay and massive clay layers identified by
the seismic methdd. The fourth layer was identified as lime-
stone. The depth to top of limestone was calculated at 12
meters and the massive clay layer was determined to he about
8 meters thick; these results are summarized in Figure 109.
Tile geophysical survey results were confirmed by use of
five shaLlow auger borings which also provided samples for
].ahoratory mineral and permeability analysis. The auguring
log also provided detailed information about the color of
sand samples, which were not discernible in the seismic and
resistivity results. The borings were limited to 10 meters
so that the clay layer would not he penetrated. The auger
log revealed seven sediment layers, varying from dry sand to
clayey sand, sandy clay and massive plastic clay (Figure 109).
Major soil changes disclosed by the auger correlated well with
the seismic and resistivity results. As shown in Figure 78,
the top of the massive clay layer was established at 7.5 meters
depth by the seismic method. This was within 1/3 meter of the
depth established by the auger. The resistivity interpretation
identified clay material beginning at about 3 meters 1 and a
massive clay at 7.5 meters. The homogeneity and flat—lying
207
-------
U)
•0
C
0
U
a,
U)
E
I-
S
>‘
>
4-
U)
U)
a,
4-
C
a,
0.
0.
Figure 108.
Data from a single seismic refraction line anc
resistivity sounding.
Distance (in feet)
Electrode “A” Spacing (in feet)
208
-------
Result of Result of
Generalized Resistivity Seismic Result of
Geologic Section Sounding Refraction Auger Boring
FTC— —OFT
13.000 OHM-Ft 1100 Ft/Sec
Sand Tan Sand
Quartz Sand Dry Quartz Sand
- — ran Sand w/Trace Clay
10— -— — —10
7000 OHM-Ft
Red Sandy Cloy
Sandy Cloy Sandy Cloy
to _____— 2200 Ft/ Sec — —
Tan Sandy Cloy
Cloyey Sand Sandy Cloy
Tan Clayey Sand
20— —20
70 OHM-Ft —— -
0
Cloy
Gray Plastic
Massive Cloy 4900 Ft/Sec Clay
— 30
30 — Clay
Auger Boring
Terminated
Minimum Depth
40 — _______________ Calculated — 40
Limestone 350 OHM-Ft to Bottom of Cloy
Limestone 31 Feet
Figure 109. Comparison of 1at ohtai.ned by auger, seismic refraction and resistivity
methoris. Interpretation oF this data yields r1 qener 1ized qeo]oqjc s ction.
-------
strata of the subsurface structure permitted excellent
correlation between the seismic, resistivity, and auger data.
The combination of two geophysical methods provided good
correlation both vertically and laterally. With the overlap-
ping positions of the seismic and resistivity stations,
excellent spatial coverage of the site was achieved. The
geophysical data, together with the data from the five auger
borings, provided a very high level of confidence in the
estimates of the depth, thickness and continuity of the clay
layer.
Mapping and Characterization of Bulk Buried Wastes
A burial disposal site was thought to be located in an
open grassy area, 300 meters wide by 400 meters long, which
had been used aSea, playground. A small river hounded the
park on one side. Stream sediment. analysis had shown high
concentrations of organic compounds, but the extent and
depth oE the toxic wastes believed present were unknown.
A subsurface investigation using six monitor wells had
revealed trace levels of contaminants at the site. However,
the location of the burial site and the extent and type of
contamination were still unknown.
The objectives of the geophysical work were to locate
and map any buried disposal areas, to provide an estimate
of the depth and volume of contaminants, and to assess the
positions of the six monitor wells with respect to the buried
wastes. Initial reconnaissance surveys using ground penetrat-
ing radar and electromagnetics indicated that anomalous soil
conditions existed between several of the monitor well locations.
Parallel survey lines were established 15 meters apart
for electromagnetic and radar measurements. Data from both
measurements revealed the boundaries of the disposal area.
Figures 110 and ill show the boundaries based upon EM data.
Radar was unable to penetrate to the base of the contaminants
due to the conductive nature of the contaminant material.
Radar did indicate that the top of the waste was within one
meter of the surface. Electromagnetic soundings indicated
that the maximum depth of the buried material was probably
not more than 5 meters. The amplitude of the EM conductivity
data in Figure 110 provided an estimate of the quantity of
buried wastes.
The burial area was then surveyed with a magnetometer to
check for the presence of steel drums. A high sensitivity,
0.1 gamma total field magnetometer was found to be ineffective
due to the extremely high variation in magnetic susceptibility
of the soil and/or waste material. Since the objective of
the magnetometer survey was to look for the presence of
relatively shallow buried steel drums, a fluxgate gradiometer
210
-------
Figure 110.
Three—dimensional perspective view of EM data showing spatial extent
and magnitude off conductivity anomaly. Six borings have missed the
burial site.
• BORING LOCATION
-------
Figure 111. Contour plot of EM conductivity anomaly in
figure 110 showing extent of buried contaminants.
212
-------
magnetometer was used, at reduced sensitivity to minimize
the effects of the wide variations in magnetic response due
to soil or waste material. Only two distinctive magnetometer
anomalies were found which might have been single steel drums.
The magnetometer survey indicated that the buried materials
consisted primarily of hulk wastes with possibly a few scattered
drums.
Radar, EM and magnetic surveys were completed within
three additional days and revealed the detailed boundaries
of the burial zone. (The boundaries were identified and
marked in the field.) The Field data was subsequently computer
processed, and detailed maps (Figures 110 and ill) were produced
in less than a week. These figures demonstrate two different
forms of data presentation. The three—dimensional figure
gives the viewer a quick grasp of the extent and maqnitude of
subsurface conductivity. The contour map provides an accurate
means of locating contaminant boundaries.
Final analysis indicated:
1) The material was not containerized but was dumped in
hulk.
2) The existence of a few buried steel drums was possible.
3) The bulk material was highly conductive.
4) The top of the material was quite close to the surface
(less than 1 meter deep) arid the bottom of the material
was variable, but less than 5 meters deep.
5) An estimate of the volume of the waste was possible,
based upon the magnitude of EM data in Figure 110.
6) The material was a fly ash matrix containing other toxic
contaminants (identification was based on samples from
one drill ho].e).
Location and Delineation of Trench Boundaries
A number of burial sites containing an unknown quantity
of steel drums were reported by an eye witness to the burial.
The drums were suspected of containing highly toxic wastes. The
exact number of individual sites was unknown, but four possible
burial sites had been tentatively defined within an overall
area of a few hundred acres. These areas were identified on
aerial photographs which showed clearings in an otherwise
densely forested area.
The first project objective was to conduct an initial
survey to search the four possible target areas for the
presence of buried drums. This initial survey was carried
out using a commercial pipe/cable—locating metal detector (as
in Figure 81). The detector was selected on the basis that it
was to be handled by one person, and maneuvered through dense
underbrush. This particular equipment is a reasonable choice
for reconnaissance work when the drums are at shallow depth,
213
-------
or in sufficiently massive groups to enable detection at
greater depths. In order for a metal detector to detect steel
drums it must he passed directly over a single drum, or over
the edge of a trench containing a number of drums; therefore
each of the four prospective sites was given total site coverage
due to the suspected critical nature of the buried waste.
This reconnaissance metal detector data was not recorded;
instead, the operator simply made notes of his findings and
marked any target locations with survey flags. (This practice
is quite satifactory for reconnaissance surveys where only an
indication of the presence or absence of buried drums is
required.) The results of this search effort identified only
one burial site. Contaminants in the soil were characterized
on the site by an organic vapor analyzer and subsequently by
chemical analysis.
Once the burial site had been locate-I, additional infor-
mation was required to assess conditions and to provide a
basis for remedial action. Project objectives were:
o To establish trench boundaries, so that sampling and
monitor wells might he placed immediately outside the
trench area. A high level of confidence was necessary
in this phase of the geophysical work, as breaching
tlie trench, or penetrating a drum, could be extremely
dangerous.
o To determine the approximate dimensions and depth of
the trench in order to estimate the number of drums.
Field work consisted of establishing the trench boundaries
with three geophysical instruments: a specialized metal detector
with a large diameter coil, a calibrated fluxgate grac3iometer
magnetometer and a ground penetrating radar system. Metal
detector and magnetometer data are shown in Figure 112.
The special metal detector was capable of operation at full
sensitivity within a few feet of a chain link security fence.
This detector also provided better resolution of the trench edges
than the pipe/cable locator used in the initial reconnaissance
effort. The metal detector’s output was recorded for later
analysis. Figure 112(4) shows a single line of data from the
trench, illustrating the accuracy with which the boundaries may
be delineated. A composite data set of eleven traverses is shown
in Figure 112(0). The lateral boundaries of the burial trench
can he seen along each traverse, and the end of the trench can
be seen in the last two lines of metal detector data in the
foreground.
214
-------
(A)
Wall
Figure 112.
Metal detector and magnetometer data over a single
trench containing buried drums. (A) Single metal
detector traverse over trench. (B) Three—dimensional
perspective view of metal detector data from
parallel survey lines over trench. (C) Three—
dimensional perspective view of magnetic profiles
over trench.
Wall
(B)
(C)
215
-------
Since the metal detector responds to both ferrous and
non—Eerrous metals, a complete magnetic survey was also
carried out to confirm that the buried metals were primarily
ferrous. The magnetometer selected was an adjustable sensi-
tivity Eluxgate gradiometer, and the data was run along the
same transect lines as the metal detector and radar data, for
correlation purposes. Special field procedures and processing
were used to correct for the eEfects of the steel fence on the
magnetometer. The composite magnetometer data shown in Figure
112(c) confirms that the trench contains ferrous metal (steel
drums). The magnetometer data was not used to locate the
trench boundaries because of the offset between the magnetic
anomaly and the buried tar.get location, due to dip of the
earth’s magnetic field (see Figure 101).
The radar results outlining the trench were based upon
disturbed soil, Figure 113. The radar data provided a means
for estimating the depth of the trench from 2 to 2.5 meters.
From the metal detector, magnetometer and radar data:
o The boundaries of the trench were accurately estab-
lished;
o The contents of the trench were tentatively identified
as steel drums;
o The volume of the trench was calculated and the number
of drums was estimated.
The geophysical work conducted in the vicinity of the
burial site was carried out with protective suits and respira-
tors. In addition, decontamination procedures were used to
clean the equipment. These safety procedures increased the time
required to conduct the survey by a factor of five. Overheating
is a major problem for personnel using protective gear, and
magnetic surveys are made more difficult by the presence of
steel-toed hoots and other ferrous parts of protective clothing
and safety equipment.
Mapping and P ssessment of Landfill T eachate Plume
Monitor well data and resistivity surveys located a plume at
a 30—year--old landfill. (See example in Section VI and Figures
65 and 66.)
A few years after the resistivity survey was completed, a
new auxiliary well field was installed nearer the landfill, at a
distance of approximately 1—1/2 miles c9owngradient and in the
direction of ground water flow. After this well field had been
pumping intermittently for about two years, analysis of the water
from the well field showed increasing levels of ammonia. A newly
installed early-warning monitor well had failed to indicate the
presence of contaminants. The landfill had also been identified
by EPA as one of the major hazardous waste sites in the country
216
-------
SOUTH -10’ +10’
NORTH.
-o
-3’
-6’
-9,
Figure 113. Radar traverse across same burial trench as in
Figure 112. Although trench is full of drums,
individual drums cannot be identified in the
radar data.
because of its proximity to the local municipal well field
which supplied drinking water to a large city.
Before selecting the locations for additional wells, a
second geophysical survey was conducted to define the current
extent of the leachate plume and to aid in the interpretation
of existing monitor well data. This second resistivity survey
(four years after the first) was performed using the same
equipment and station locations. The time of year was the
same as in the previous study. The new survey area was extended
beyond the earlier survey area in order to cover more area
downgradierit. Results of this second survey are also shown in
- ! .r ‘ r ir I -,,- ‘ -
I I
217
-------
Figure 114. The following conclusions were drawn:
1) The plume had shifted direction and had migrated to the
northeast in response to a change in the local ground—water
gradient created by the newly—active auxiliary well field.
The plume had extended into the well field cone of
influence, creating the increases in ammonia observed in the
water analysis.
2) As may be seen in Figure 114, the “early—warning” well had
not detected any leachate, because the well was located in
an anomalous area relatively clear of leachate. A review of
the monitor well’s geologic log indicated the presence of
fine sand and clay, instead of the highly permeable
limestone which was typical of the area. The well had
inadvertently been located in a small zone of lower
permeability.
3) Comparison of the two sets of geophysical data revealed that
the lateral extent of the plume had increased by approxi-
mately one kilometer (0.6 miles) in four years, which is a
migration rate of about 0.5 meters (1.5 feet) per day,
roughly half the rate of the calculated regional ground
water flow. This migration rate, measured in this manner,
takes into account the combined effects of all the variables
influencing the rate of leachate migration.
4) The areal and vertical extent of the plume, determined by
resistivity, made possible a calculation of the total volume
of aquifer contamination.
5) The survey identified a number of other point sources which
were contributing to the contamination of the aquifer (not
shown in the data).
With a map of the leachate plume, a measure of how fast it
was migrating, an understanding of the factors influencing its
path, and an evaluation of previous water quality data, it
was possible to obtain a more accurate understanding of site
conditions. In addition, new monitor wells could be installed
with a high degrees of confidence, with their locations being
representative of ground water conditons.
Measured by resistivity or EM techniques, the plumes
shown in Figure 114 are representative of conservative chemical
parameters (e.g., chlorides, sodium, etc.). The outer contours
are near background values and represent a reasonable estimate
of the maximum extent of contamination by sanitary landfill
leachates. Many contaminants of a hazardous nature will remain
within these boundaries, not migrating as far as the conservative
parameters. Although there are some contaminants which do migrate
218
-------
LI ACTIVE
WELL
FIELD
Figure 114.
Mapping of leachate plume using resistivity methods shows changes in
plume over four—year period. (Shaded area represents 200 Ohm—ft
contour.)
(A)
(B)
‘.0
-------
faster than ground water, they are not commonly associated with
sanitary landfills in large enough quantities to make them a
dominant factor in plume behavior.
Locating Monitor Wells at an Uncontrolled Hazardous Waste Site
A ground water investigation was initiated at an uncontrolled
hazardous waste site to determine the direction and extent of any
contaminant migration. Existing information noted a regional
ground water gradient to the northeast. In addition, geologic
logs revealed a highly variable geohydrologic setting composed of
sand and gravel lenses within a clay matrix. The occurrence of
these more permeable sand and gravel depostts could significantly
influence the path of contaminant migration and the placement of
monitor wells.
Before locating any monitor wells, a detailed electromag-
netic (EM) conductivity survey was conducted to map contaminant
migration and to evaluate the natural setting. This data was
then used to direct the placement of monitor wells and provide
a guide for soil sampling.
The field data consisted of about 30 parallel profile lines,
1000 meters long, spaced 30 meters apart. Data was acquired
only around the perimeter of the site and not directly over it.
Using continuously-recorded data, the high density EM survey
was accomplished in less than a week. Continuous EM data was
obtained to 6 meter and icmeter depths using two EM systems.
Subsequently, the recorded data was digitized, computer—processed
and then plotted in both contour (Figure 115) and three-dimen—
nsional perspective (Figure 116) formats. The contoured data was
used to locate the monitor wells accurately. The three—dimen-
sional view aided in the interpretation of overall site condi-
tions. The EM data showed a high degree of natural variability
at the site. The data helped to identify clearly the existence
and extent of two plumes.
The size and extent of the main plume emanating from the
storage area is clearly seen in the center of the figure as a
conductivity high. The major plume appeared to migrate toward
the east—northeast, with minor lobes extending north, east and
south (see Figure 115). The highest conductivities occurred
around the northeast corner of the site where a disposal
impoundment was located. A minor plume which extended toward
the west (regionally upgradient) was probably caused by the
mounding of ground water within the elevated hazardous waste
site. Extensive background data, obtained outside the
immediate site, allowed a good statistical assessment of the
range of natural variations in conductivity. Once the maximum
range of natural variations was determined, any higher values
measured were attributed to contaminated ground water. The
220
-------
Isopleth map of EM conductivity data at a hazardous
waste site shows a plume (Shaded area) leaving
the site and considerable variation in surrounding
site conditions.
221
Figure 115.
-------
Figure 116.
outer boundaries of this contaminated water are delineated by
the shaded area in Figure 115.
The use of continuous recording EM technique and high
data density, together with the computer presentation,
permitted identification of the conductive plume boundaries
well into the “noise level” caused by variations in the
natural site conditions.
Besides identifying the contaminant plume, the EM data
had indicated that natural background conductivities were
extremely variable. State geologic references, plus well—
placed borings, revealed that these natural conductivity
fluctuations were associated with river—deposited lenses of
Three—dimensional perspective view of EM data
shown in Figure 115. Note plume in center of
plot and variability in conductivity due to
natural geohydrologic conditions.
222
-------
sand, gravel and clay. This variability is related to the
complex distribution of permeable sand and gravel deposits
(low conductivity) within the clay matrix (high conductivity).
These variations in sand and gravel content were due to
differential deposition of these materials in old buried
stream beds. The existence of such buried stream beds had
been established from local geologic literature. The two
sets of EM data (6 meter and 15 meter depths) revealed that
most of the variations in the sand and clay deposits lay
within 6 meters of the surface. Below this depth, conditions
were much more homogeneous.
Water flow and contaminant migration may be expected to
follow the most permeable routes (low conductivities) in this
shallow unconfined system. Therefore, higher trace levels of
contaminants might be expected in these more permeable zones,
and the EM lows could be used to place monitor wells beyond
the extent of the obvious conductive plume.
Four wells were located within the identified plume. A
fifth well was located upgradient in a low conductivity sand
lens for background determination. Augering confirmed the
existence of the sand and gravel deposits, as suggested in
the EM maps and geologic literature.
The plume boundary delineated by the EM method (Figure
115) represented the extent of transport of the conservative
ionic parameters. Quantitative data subsequently obtained
from the wells indicated that the boundary of the conductivity
plume approximated the 1 ppm level of the priority pollutants.
(This correlation is applicable only at this site and should
not be used as a rule of thumb.)
It has been recommended that the EM survey be repeated in
two years, to observe any changes in the plume, and to correlate
these changes with the longer-term quantitative well data.
Comparison of the two EM surveys made at different times will
show the absolute migration rate, and indicate whether the minor
lobes are developing into primary pathways.
223
-------
SECTION XI
CLOSING COMMENTS
Most geophysical methods have been in existence for many
years. With the exception of radar, the principles and early
applications of the methods discussed in this document can be
traced back to the 1930’s. In recent years, however, remark-
able advances in electronics have allowed geophysical measure-
ments to be carried out more effectively and, in some cases,
have helped bring about new technologies.
Most of the geophysical methods have evolved in the
mining and oil exploration industries, where the methods are
used to evaluate much deeper and larger targets than those at
HWS. The use of geophysical methods for evaluations of ground
water contamination and geotechnical investigations has become
widespread only in the past 5 to 10 years.
Although airborne remote-sensing and downhole geophysical
methods are viable investigative approaches at hazardous waste
sites, this document addresses only surface geophysics:
specifically, the six methods that have been successfully
applied to numerous hazardous waste site investigations.
These six surface methods are only part of the total geo-
physical technology which may be applicable to hazardous
waste site investigations (Figure 117). Further, geophysics
itself is only a small piece of the total subsurface investi-
gation systems approach. Drilling, analytical laboratory
methods, trained earth science personnel, sound project manage-
ment, and many other factors must be used in combination with
geophysical technologies in order to solve site investigation
problems.
Of the three remote—sensing geophysical approaches, only
surface geophysics is considered in this document, and in that
connection, only six methods are discussed. The criteria used
for the selection of the six methods were presented in Section
III and are repeated here:
o the methods are regularly used in HWS assessments;
o they have proven capability in HWS assessments;
o they are suitable for broad application to the problems
typically found at UWS.
224
-------
p . ,
4 -fl
Figure 117. Technical resources and tools which may be applied to subsurface investigations
at hazardous waste sites. The six methods discussed in this document represent
a small but important portion of the total technology which may be brought to
bear.
-------
There are many other geophysical mthods and approaches which
might be used at HWS, but at this time the authors believe that
only those included in this document have met the above cireria.
Pitfalls in Using Geophysical Methods
In the past, use of the geophysical methods in hydrology
and engineering geology has been cyclical. One reason for
this is that in gathering, processing, and interpreting the
data, many pitfalls await the unwary investigator. This
section will attempt to outline some of the areas in which
pitfalls commonly occur.
Many technical limitations of the methods have been
pointed out in this document, although the listing is by no
means all—inclusive. However, further discussion of technical
limitations, field problems and their pitfalls in data processing
and interpretation is well beyond the scope of this document.
Unfortunately, this information is not well documented and is
much more likely to be discovered by actual experience than by
exploring existing literature or texts. This is one reason that
trained professionals, with experience in the field, must be used
to carry out this work. Pitfalls related to the non-technical
aspects of geophysical work are just as important as those in
technical areas. Some of them are outlined below.
Non-technical Pitfalls in Using Geophysical Methods
With any technology (drilling, installing monitor wells,
analyzing water samples, or computer modeling ground water flow)
the user must realize that some problems’may arise. They can
generally be minimized by well—trained, experienced professional
personnel.
o Geophysical instruments are as sophisticated, but
unfortunately not as well known or accepted, as the
chemist’s gas chromatograph or mass spectrometer. They
are all based upon solid scientific principles, yet many
professionals consider geophysical tools to be mysterious
“black boxes”. Often these black boxes are expected to
provide the solution to a problem in some mysterious way.
Obviously, they cannot do this.
o Geophysical techniques alone do not generally provide
unique solutions to problems. By integrating other
knowledge with the geophysical results in a systematic
approach, unique solutions can be obtained. If the tools
are properly integrated, they can provide outstanding
results in terms of site assessment capability. If,
however, they are improperly used, the methods can lead
226
-------
to disappointment. The information obtained by
integrating geophysics into a program will simply help
the investigator arrive at a better answer faster. This
requires:
— trained personnel
— experienced personnel
— the right equipment
a total understanding of the problem
— the ability to integrate a variety of information into
the interpretation.
o Like the chemist using’ a mass spectrometer, the
geophysicist must be properly trained in order to obtain
accurate results. Not only should this training include
equipment operation, it also must encompass a background
in the earth sciences. Few universities offer courses in
geophysics; those that do are inclined to slant them
toward mineral and oil exploration. These applications
are sufficiently different from those discussed in this
text that direct transfer of technology and personnel is
not readily made. Civil engineering proqrams may also
mention some of the geophysical technologies in courses
on soil and rock mechanics, but little if any solid
training has been provided in these areas.
o Experience has shown that whoever plans to interpret the
data should be involved in the data acquisition. No
matter how well trained a person may be, an
interpretation can be grossly misleading without
first—hand familiarity with field conditions.
o There is oftep a tendency to use short-cut interpretation
procedures. Approximation methods are sometimes
acceptable if their limitations are clearly understood.
All too often, however, approximations are treated as
absolutes.
o Investigators often develop a dependence upon a single
technology. Individuals or firms may be familiar with
only a single geophysical method, which they tend to
promote and use extensively. Under such circumstances,
the technology may often be applied where its use is not
appropriate.
o Often when a specific technique is oversold, it is
expected (and often fails) to provide information under
conditions for which it was not designed. The potential
user must appreciate the limitations as well as the
benefits of a geophysical method and must apply that
method in the proper manner.
227
-------
o The use of equipment may be extended beyond its designed
capability. This practice may lead to poor results.
Selection of Professional Consultants to Perform Site
Investigations
After the project manager has defined his project
objectives, they may be achieved in the following ways:
(1) He can contract for data acquisition only, in which
case data interpretation must be done by someone else.
(2) He can contract for both data acquisiton and
interpretation in a single package.
(3) If he is certain of what needs to be done from a
geophysical point of view 1 he may write a contract
specifying the use of only a single method.
(4) Many HWS investigations will demand flexibility and may
require multiple methods to be used as determined in
the field. Such effort requires a very flexih 1e
contract and a consultant capable of providing the
appropriate geophysical services. The geophysical data
must be incorporated with the existing geological,
hydrological, biological and chemical data into a
composite understanding of site conditions.
There are firms which offer data acquisition only, at
reduced prices. The authors of this document do not believe
that this is an effective approach to the problem, because
the data acquisition is often accomplished by non—professionals
who may not be able to evaluate site conditions and to respond
accordingly.
There are many qualified persons and firms who are capable
of providing both data acquisition and interpretation with any
single seismic refraction or resistivity method discussed in
this text. On the other hand, in some of the newer or special-
ized areas (radar, continuous EM, specialized metal detectors
and magnetometry) there are only a few experienced persons
to be found in the entire U.S. In many cases, their expertise
covers only one method. However, if project requirements
can be defined with sufficient accuracy for one method to
yield the necessary information, then this can be a viable
approach.
Only a limited number of professional firms have the broad-
based capability and experience in applying the six methods
discussed herein. Even fewer have experience on hazardous waste
sites. The authors estimate that at the time of this writing
228
-------
there are fewer than ten firms in the U.S. which might offer
this inclusive capability. Clearly, this integrated systems
approach to carry out the HWS investigation with the flexibility
and synergism of multiple geophysical methods provides the
most cost—effective approach to the project. It is in this
area that the most effective HWS investigations have been
carried out to date. By use of this approach, the project
manager can construct a complete package including:
o flexibility;
o multiple geophysical methods;
o data acquisition;
o data interpretation;
o total integration of the project;
o a professionally trained and experienced staff.
Clearly, the project manager should determine if the firm
and individual(s) performing the services are experienced in such
investigations. Further, he should ascertain the qualifications
and training of personnel, and the types of equipment recommended
to perform the surveys. In most instances, the consultant should
own the equipment used for the survey. If not, chances are that
the firm and its field party do not have adequate experience. A
hazardous waste site is not a training ground.
Finally, the project manager must develop confidence in the
professional abilities of his qualified consultants. He should
utilize his consultant’s experience to optimize the outcome of
the field investigation. In addition, he should see that the
contract is flexible and that it allows for options in adjusting
the work to achieve optimal project objectives.
Cost Comparison: Systems vs. Traditional Approach
When making a cost comparison, one cannot evaluate geo-
physical field project costs versus the costs of drilling and
installing a monitor well only. The only reasonable way to make
a cost comparison is to look at the total project bottom—line
costs. If it will speed the basic understanding of site
conditions and improve the accuracy of an assessment,
the integrated systems approach which uses geophysics is the
low—cost approach to take. The systems approach requires less
drilling and fewer monitor wells, thereby minimizing the number
of chemical analyses necessary. The results are lower program
costs with a better understanding of site conditions achieved in
a shorter time.
229
-------
Figure 118 shows a cost comparison of a combined geo-
physical systems approach versus monitor wells only. The
curve shows cost versus number of wells. Monitor well costs
are based upon a total program cost of $3 000 to $10,000 per
well which includes drilling, installation of a quality monitor
well, well development, initial sampling, analysis of priority
pollutants, and a subsequent quarterly sample program over one
year, including supporting reports and project management.
All wells are assumed to be 10 meters deep. The costs are
conservative in that the direct costs and intangible risks of
drilling into highly hazardous areas are included.
Costs for the geophysical systems approach are based upon
a site survey coverage approaching 100%, plus three monitor
wells and all supporting efforts, reports and project manage-
ment as described above. One well is typically used for back-
ground measurement, and two are placed in representative
locations to quantify specific contaminants within a plume.
The geophysical systems approach becomes increasingly
cost—effective as the number of required monitor wells increases.
The more complex the site, and the greater the number of unknowns
and risks, the greater are the benefits to be derived from using
such a systems approach. The level of confidence in the results
of the site investigation will increase, and the risk that the
investigation will create a serious health, fire or explosion
hazard will be reduced.
Future Possibilities
A methodology for investigation of HWS exists today--there
is no need to wait for future developments to occur, although
some improvements in the six methods are expected. Each method
has potential for refinements in hardware, processing and
interpretation. For example, some improvement in the depth and
operation of radar in more difficult soil conditions can be
expected. Advances in real-time processing and plotting of radar
data are also under way. Seismic methods will be using processing
systems which provide higher resolution, thus permittino more
detailed investigations. Metal detectors will attain slightly
greater depth capabilities, and interpretive techniques for
both metal detector and magnetometer data are expected to
improve.
The greatest advancement will likely be the integration
of various hardware systems into a single sensing network.
ighspeed recording and processing will be applied to combined
data sets, as well as to individual sensor data.
230
-------
1000
0
0
o
( )
I .-.
o 0
C 100
01 0
o U ’
0 0
I—
C
10•
10 100 1000
NUMBER OF MONITOR WELLS
Figure 118. Cost comparison curve for hazardous waste site investigation using monitor
wells alone versus an integrated systems approach. (Overall project
accuracy and effectiveness are not considered in this data.)
Usng Monitor
Wells Only
Systems Geophysical Approach
-------
Other techniques will come into use. For example,
transient EM can provide increased sounding capability to
depths of 300 meters or more, and complex resistivity may
provide more diagnostic information than traditional resistivity
measurements. While neither of these methods has yet been
applied to hazardous waste sites, they and others may eventually
come to be used and may one day supplement the six existing
technologies described in this document.
Meanwhile, those six methods will continue to make a
contribution to hazardous waste site assessment.
232
-------
BI BL I0GRAPHY
Benson, R. C., and R. A. Glaccum. 1979a. Radar Surveys for
Geotechncal Site Assessment. In: Geophysical Methods in
Geotechnical Engineering, Specialty Session, American
Society of Civil Engineers, Atlanta, Georgia. pp. 161-178.
Benson, R. C., and R. A Glaccum. 1979b. Remote Assessment of
Pollutants in Soil and Groundwater. In: Proceedings of the
Hazardous Material Risk Assessment, Disposal and Management
Conference, Miami Beach, Florida. pp. 188—194.
Benson, R. C., and R. A. Glaccum. 1980. Site Assessment:
Improving Confidence Levels with Surface Remote Sensing.
In: Proceedings of National Conference on Managewent of
Uncontrolled Hazardous Waste Sites, Washington, D.C. pp.
59—6 5.
Benson, R. C., R. A. Glaccum, and P. Beam. 1981. Minimizing
Cost and Risk in Hazardous Waste Site Investigations Using
Geophysics. In: Proceedings of National Conference on
Management of Uncontrolled Hazardous Waste Sites,
Washington, D.C. pp. 84—88.
Breiner, S. 1973. Applications Manual for Portable
Magnetometers. GeoMetrics, Sunnyvale, California. 58 pp.
Evans, R. B. 1982. Currently Available Geophysical Methods for
Use in Hazardous Waste Site Investigations. American
Chemical Society Symposium Series, No. 204. pp. 93—115.
Evans, R. B., R. C. Benson, and J. Rizzo. 1982. Systematic
Hazardous Waste Site Assessments. In: Proceedings of
National Conference on Management of Uncontrolled Hazardous
Waste Sites, Washington, D.C. pp. 17—22.
Freeze, R. A, and J. A. Cherry. 1979. Groundwater. Prentice—
Hall, Inc., Englewood Cliffs, New Jersey. 604 pp.
Geraghty and Miller. 1978. Electrical Resistivity Evaluations
at Solid Waste Disposal Facilities. U.S. Environmental
Protection Agency, Washington, D.C. 94 pp.
233
-------
Geraghty, J. J. 1980. Evaluation of Hydrogeologic Conditions.
In: Proceedings of National Conference on Management of
Uncontrolled Hazardous Waste Sites, Washington. D.C. pp.
49—5 2.
Glaccum, R. A., R. C. Benson, and M. R. Noel. 1982. Improving
Accuracy and Cost—Effectiveness of Hazardous Waste Site
Investigations. Ground Water Monitoring Review, Summer,
1982. pp. 36—40.
Gilkeson, R. H., and K. Cartwright. 1982. The Application of
Surface Geophysical Methods in Monitoring Network Design.
In: Proceedings of Second National Symposium on Aquifer
Restoration and Ground Water Monitoring, Columbus, Ohio.
pp. 169—183.
Greenhouse, J. P., and D. D. Slaine. 1982. Case Studies of
Geophysical Contaminant Mapping at Several Waste Disposal
Sites. In: Proceedings of Second National Symposium on
Aquifer Restoration and Ground Water Monitoring, Columbus,
Ohio. pp. 299—315.
Griffiths, D. H., and R. F. King. 1969. Applied Geophysics for
Engineers and Geologists. Pergarnon Press, Elmsford, New
York. 223 pp.
Horton, K. A., R. M. Morey, L. Isaacson, and R. H. Beers. 1981.
The Complementary Nature of Geophysical Techniques for
Mapping Chemical Waste Disposal Sites: Impulse Radar and
Resistivity. In: Proceedings of National Conference on
Management of Uncontrolled Hazardous Waste Sites,
Washington, D.C. pp. 158-164.
Johnson, R. W., R. A. Glaccum, and R. Wojtasinski. 1980.
Application of Ground Penetrating Radar to Soil Survey. In:
Proceedings of Soil and Crop Science Society of Florida,
vol. 39. pp. 68—72.
Leis, P. G., and K. A. Sheedy. 1980. Detailed Location of
Inactive Disposal Sites. In: Proceedings of National
Conference on Management of Uncontrolled Hazardous Waste
Sites, Washington, D.C. pp. 116—118.
Lord, A. E. Jr., R. M. Koerner, and J. E. Brugger.. 1980. Use of
Electromagnetic Wave Methods to Locate Subsurface Anomalies.
In: Proceedings of National Conference on Management of
Uncontrolled Hazardous Waste Sites, Washington, D.C. pp.
119—124.
234
-------
McKown, G. L., and C. A. Sandriess. 1981. Computer—Enhanced
Geophysical Survey Techniques for Exploration of Hazardous
Waste Sites. In: Proceedings of National Conference on
Management of Uncontrolled Hazardous Waste Sites,
Washington, D.C. pp. 300—305.
McNeill, J. D. l980a. Electrical Conductivity of Soils and
Rocks. Technical ote TN—5. Geonics Limited, Ontario,
Canada. 22 pp.
McNeill, J. D. 1980b. Electromagnetic Terrain Conductivity
Measurement at Low Induction Numbers. Technical Note TN-6.
Geonics Limited, Ontario, Canada. 15 pp.
McNeill, J. D. 1982. Electromagnetic Resistivity Mapping of
Contaminant Plumes. In: Proceedings of National Conference
on Management of Uncontrolled Hazardous Waste Sites,
Washington, D.C. pp. 1—6.
Mooney, H. M. 1973. Handbook of Engineering Geophysics, Volume
1: Seismic. Bison Instruments, Minneapolis, Minnesota.
216 pp.
Mooney, H M. 1980. Handbook of Engineering Geophysics, Volume
2: Electrical Resistivity. Bison Instruments, Minneapolis,
Minnesota. 90 pp.
Noel, M. R., R. C. Benson, and R. A. Glaccurn. 1982. The Use of
Contemporary Geophysical Techniques to Aid Design of
Cost-Effective Monitoring Well Networks and Data Analysis.
In: Proceedings of Second National Symposium on Aquifer
Restoration and Ground Water Monitoring, Columbus, Ohio.
pp. 163—168.
Pease, R. W. Jr., and S. C. James. 1981. Integration of Remote
Sensing Techniques with Direct Environmental Sampling for
Investigating Abandoned Hazardous Waste Sites. In:
Proceedings of National Conference on Management of
Uncontrolled Hazardous Waste Sites, Washington, D.C. pp.
171—176.
Sgambat, J. P., and J. R. Stedinger. 1981. Confidence in
Ground—Water Monitoring. Ground Water Monitoring Review,
Spring, 1981. pp. 62-69.
Telford, W. M., L. P. Geldart, R. E. Sheriff, and D. A. Keys.
1976. Applied Geophysics. Cambridge University Press, New
York. 860 pp.
235
-------
White, R.M., and S. S. Brandwein. 1982. Application of
Geophysics to Hazardous Waste Investigations. In:
Proceedings of National Conference on Management of
Uncontrolled Hazardous Waste Sites, Washington, D.C. pp.
91—93’
Yaffe, H. J., N. L. Cichowicz, and P. J. Stoller. 1980. Remote
Sensing for Investigating Buried Drums and Subsurface
Contamination at Coventry, Rhode Island. In: Proceedings
of National Conference on Management of Uncontrolled
Hazardous Waste Sites, Washington, D.C. pp. 239—249.
Zohdy, A. A., G. P. Eaton, and D. R. Mahey. 1974. App1icatio’
of Surface Geophysics to Ground—Water Investigations.
Techniques of Water—Resources Investigations of the United
State Geological Survey, Chapter Dl. 116 pp.
Procedures Manual for Ground Water Monitoring at Solid Waste
Disposal Facilities. U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1977. 269 pp.
236
------- |