Statea Advanced Monitoring System* Division August 198B
En^trcnmental Prcteetloa Las "Vcgaa, Nevada ,
' '?
! ,
-------
SURFACE-BASED ELECTRICAL SURVEYS
FOR
INJECTION WELL FLUIDS
by
Roy B. Evans, Ph.D.
Environmental Research Center
University of Nevada-Las Vegas
and
Pieter Hoekstra, Ph.D.
The Earth Technology Corporation
and
Aldo T. Mazzella, Ph.D.
Environmental Monitoring Systems Laboratory
Cooperative Agreement
CR 812189-01
for
Aldo T. Mazzella, Project Officer
Advanced Monitoring Systems Division
Environmental Monitoring Sytems Laboratory
U.S. Environmental Protection Agency
Las Vegas, Nevada
August 1988
-------
NOTICE
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under cooperative
agreement number CR 812189-01 to the Environmental Research Center of the
University of Nevada-Las Vegas. It has been subject to the Agency's peer
and administrative review, and it has been approved for publication as an
EPA document.
Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
ABSTRACT
Injected fluids present difficult electrical targets. Detection of
target objects, formations, or strata by electrical geophysical terhniques
depends on the contrast in some electrical property between the target of
interest and the material within which it is imbedded. Electrical
conductivity is the principal property of interest for most electrical
methods. The conductivity of injected fluids is usually roughly
proportional to their total dissolved solids content. Injection usually
takes place at depths of a few thousand feet, beneath confining strata
which are often highly conductive, such as shale. Often the injection zone
already contains connate water high in total dissolved solids and hence of
high conductivity. In such situations, with properly constructed and
operated injection wells, surface electrical methods probably will not be
very useful for directly monitoring injected fluids because the contrast
between the injected fluids and the connate water of the injected stratum,
as well as the conductive overburden, would not be great. However, in
cases where wells are improperly constructed and operated or cases in which
fractures connect the injection zones with overlying fresh water aquifers,
movement of the highly conductive injected fluids into the overlying fresh
water aquifers could create a conductive electrical target which, contrasts
sharply with the surrounding material. In such cases, electrical methods
may indeed be useful in detecting and mapping the contaminated zones.
The important criteria for evaluating the applicability of various
electrical methods to exploration objectives are the practical depth of
exploration, lateral resolution, vertical resolution, sensitivity to
geologic noise, and survey productivity.
For cases where conductive contamination lies very near the surface
(<30 meters depth), continuous, frequency-domain electromagnetic induction
is likely to be useful in mapping the areal extent of the brines. D.C.
resistivity may be successful at slightly greater depths, although cultural
interferences (power lines, pipelines, fences) are likely to cause
problems. Neither of these methods are likely to produce useful information
at dppths approaching a typical injection horizon (thousands of meters).
Because of its relative insensitivity to near-surface geological noise
and its ability to achieve substantial depths of exploration with
relatively small loop arrays, time-domain electromagnetic induction (TDEM)
appears to offer the greatest chance of success of the methods discussed
here at depths approaching injection horizons. However, even TDEM will
need a relatively large conductivity contrast (factor of 2 or greater)
between the target and surrounding background to do successful mapping. To
be detectable, the target must have lateral dimensions which are a
substantial fraction of the depth of exploration; mapping requires targets
which have dimensions greater than the depth of exploration.
11
-------
This report was submitted in fulfillment of Cooperative Agreement No.
CR 810550-01 by the Environmental Research Center of the University of
Nevada-Las Vegas under the sponsorship of the U.S. Environmental Protection
Agency. This report covers a period from July 1985 to July 1988, and work
was completed as of August 1988.
IV
-------
CONTENTS
Page
Notice i i
Abstract " i i i '
List of Figures vii
List of Tables ix
CHAPTER I: INTRODUCTION
1.0 Purpose and Scope of the Document 1
2.0 . UIC Regulations 2
3.0 Injection Wells 5
3.1 Numbers and Types of Injection Wells 5
3.2 Geological and Chemical Characteristics 12
3.2.1 Available Information 12
3.2.2 Hydrogeology 12
3.2.3 Hydrogeology of Class I Wells 13
3.2.4 Hydrogeology of Class II Wells 15
3.3 Volumes and Characteristics of Injected Fluids 18
3.3.1 Class I Wells: Injected Fluids 18
3.3.2 Class II Wells: Injected Fluids 18
4.0 Pollutant Pathways 24
CHAPTER II: ELECTRICAL PROPERTIES OF FLUIDS, ROCKS AND GEOLOGIC
FORMATIONS 29
CHAPTER III: PHYSICS OF RESISTIVITY MAPPING
1.0 General 37
2.0 Direct Current (D.C.) Resistivity Method 39
3.0 Electromagnetic Induction Methods 43
3.1 General 43
3.2 Continuous Electromagnetic Systems 47
3.3 Transient Electromagnetic Systems 52
4.0 MagnetoTelluric (MT), Audio-MagnetoTelluric (AMT), and Controlled-
Source MagnetoTelluric (CSAMT) Methods 63
5.0 Induced Pollarization Method 68
CHAPTER IV: APPLICATIONS AND CASE HISTORIES
1.0 Criteria for Evaluation of Electrical Methods 75
1.1 Lateral Resolution 75
1.2 Vertical Resolution 75
1.3 Sensitivity to Geologic Noise 77
2.0 Electromagnetic Induction Methods 77
2.1 Frequency Domain Methods 79
2.1.1 Instrumentation 79
2.1.2 Operational Advantages and Disadvantages 80
2.1.3 Lateral Resolution 84
2.1.4 Geologic Noise 86
2.1.5 Case Histories: Fixed Frequency Magnetic Induction 86
-------
CONTENTS (Continued)
Page
2.2 Time Domain (TDEM) Soundings : 92
2.2.1 Instrumentation 92
2.2.2 Operational Advantages and Disadvantages 92
2.2.3 Lateral Resolution 95
2.2.4 Sensitivity to Geologic Noise 95
2.2.5 Case Histories: Time-Domain Methods .............. 97
3.0 CSAMT Method 104
3.1 Instrumentation 104
3.1.1 Operational Advantages and Disadvantages 104
3.1.2 Geologic Noise 107
3.1.3 Lateral Resolution 107
3.2 Case History: CSAMT 109
4.0 Direct Current (D.C.) Method 113
4.1 Instrumentation 113
4.1.1 Operational Advantages and Disadvantages 113
4.1.2 Lateral Resolution 113
4.1.3 Vertical Resolution 115
4.1.4 Sensitivity to Geologic Noise 115
4.2 Case Histories: D.C. Methods 115
CHAPTER V: SUMMARY AND CONCLUSIONS 120
REFERENCES 123
APPENDICES
Appendix 1: 1984 EPA Injection Well Inventory 126
Appendix 2: Class I Well Injection Zone Characteristics 129
Appendix 3: Confining Zone Characteristics of Class I
Injection Wells 130
-------
FIGURES
Page
1.1 Ideal injection well and site 3
1.2 Injection well totals by EPA region 7
1.3 Injection well totals by state 8
1.4 Injection well totals by well type 9
1.5 Texas injection wells by well type 10
1.6 Pennsylvania injection wells by well type 11
1.7 Average hydrogeological characteristics of Class I injection wells
in the Great Lakes region 14
1.8 Average hydrogeological characteristics of Class I injection wells
in the Gulf Coast region 16
1.9 Idealized cross-section of Anadarko Well No. "A" No. 1,(Class II
injection well 25
1.10 Idealized cross-section of Sun Well No. 6-4L (Class II injection
wel 1) 26
1.11 Migration pathways for injected fluids 27
2.1 Electrically equivalent concentrations of a sodium chloride
solution as a function of resistivity (or conductivity) and
temperature (from Keys and MacCary, 1972) 30
2.2 Approximate resistivity of granular aquifers versus prosity for
several water salinities (from Guyod, 1966) 32
2.3 Electrical conductivities corresponding to Figure 1.8 35
3.1 Geometry of current flow in electrical methods 38
3.2 Current flow within a layered medium (electric charges arise at
resistivity boundaries) 40
3.3 Charge formation in response to an impressed electric field .... 41
3.4 Two-layer apparent resistivity curves for MT, TDEM, and DC 42
3.5 Schematics of Wenner and Schlumberger arrays 44
3.6 Schlumberger array two-layer master curves of apparent
resistivity versus electrode spacing 45
3.7 Waveforms and transmitter-receiver geometry for electromagnetic
induction • 46
3.8 Operation of electromagnetic induction in a layered earth 48
3.9 Schematic of horizontally-layered earth 51
3.10 Electromagnetic induction response versus depth for horizontal
and vertical dipoles 53
3.11 Transmitter-receiver array used in TDEM soundings (the array
is moved for each sounding) 54
3.12 TDEM waveforms 55
3.13 TDEM eddy current propagation in a homogeneous earth 56
3.14 Eddy current intensity as a function of depth below the surface. 58
3.15 Migration of the location of maximum eddy current intensity as
a function of time 59
3.16 Behavior of EMF due to vertical magnetic field (BJ and
horizontal field (Bx) on a profile through the center of the
transmitter loop . .*. 61
3.17 Computed two-layer apparent resistivity curves over resistive
and a conductive lower layer 62
3.18 Signal sources for magnetotelluric methods 64
VI 1
-------
Figures (continued)
Figure Page
3.19 Propagation of electric and magnetic fields in magnetotelluric
methods 55
3.20 Skin depth of electromagnetic plane waves as a function of
frequency and earth resistivity 67
3.21 Induced polarization effect: Transmitter current and voltage
waveforms in induced polarization 69
3.22 Commonly used array types for DC resistivity and IP 70
3.23 Transmitter current and receiver voltage.waveforms in IP field
measurements 72
4.1 Examples of hypothetical geoelectrtc sections which could be
associated with injection wells > 74
4.2 Geologic noise obscuring the mapping of a freshwater-saltwater
interface 78
4.3 Geonics EM-31 conductivity meter vertical dipole mode 82
4.4 Geonics EM-34-3 conductivity meter horizontal dipole mode 83
4.5 Eddy current intensity as a function of depth for electro-
magnetic induction 87
4.6 Location of industrial complex, conductive contaminant plume,
and survey line for EM measurements in Henderson, NV 89
4.7 Apparent conductivity values from Geonics EM-31 along survey
line in Henderson, Nevada •... 90
4.8 Effect of topography on apparent conductivity values from EM-34
survey near Falmouth, Massachusetts 91
4.9 Sewage plume as outlined by survey with Geonics EM-34 near
Falmouth, Massachusetts 93
4.10 TDEM eddy current intensity as a function of depth 96
4.11 Geologic cross-section and borehole induction logs, Haskell
County, Oklahoma 98
4.12 Computer modeling studies of detectability of thin pay zone
with TDEM, Mt, and DC resistivity 99
4.13 Layout of TDEM survey and locations of drilled wells, Haskell
County, Oklahoma, showing typical TDEM apparent resistivity
curves over locations with gas saturation, brine saturation,
and dry holes 102
4.14 TDEM apparent resistivity curves along survey line in Siberian
Basin, showing oil-water contact 103
4.15 Plan view of TDEM loops and nearby test wells, Osage County, OK 105
4.16 Typical geoelectric cross-section derived from TDEM soundings,
Osage County, Oklahoma 106
4.17 CSAMT geologic noise created by charge formation along
discontinuities 108
4.18 Schematic layout for CSAMT survey 110
4.19 CSAMT psuedo-section, Lincoln County, Oklahoma Ill
4.20 CSAMT vertical psuedo-section, Line 8, resistivity versus depth 112
4.21 Geologic cross-section, Woburn, Illinois, consolidated oil field 116
4.22 Plan view of Woburn Site, with apparent resistivity contours .. 117
4.23 Apparent resistivity depth soundings, Woburn, Illinois 119
vm
-------
TABLES
Table Page
1.1 1984 EPA Injection Well Inventory 6
1.2 Hydrogeological characteristics of Class I injection wells 17
1.3 Hydrogeological characteristics of Class II injection wells 19
1.4 Waste characteristics of 108 Class I injection wells active
in 1983 in the United States 20
1.5 Hazardous waste stream components and concentrations in Class I
injection wells in'the United States in 1983 21
1.6 Analyses of oil field brines (in ppm) 22
2.1 Contributions of various ions to fluid conductivity 32
2.2 Resistivities of sediments 33
4.1 Relationship between resistivity and measured EMF ....* 76
4.2 Commercially available frequency-domain EM systems 81
4.3 Approximate effective depth of investigations for Geonics
EM-31 and EM-34 85
4.4 Characteristics of transient systems 94
4.5 Maximum decrease in apparent resistivity and measured voltage
caused by actual increases in conductance for TDEM, MT, and DC . 101
4.6 Partial list of commercially available DC equipment 114
5.1 Summary table for surface-based electrical surveys 122
IX
-------
SURFACE-BASED ELECTRICAL SURVEYS FOR INJECTION-WELL FLUIDS
I. INTRODUCTION
1.0 Purpose and Scope of the Document
This document surveys available and potentially usefu.l surface-based
electrical geophysical methods for mapping contaminant fluids from
injection wells. The goal has been to present a "snapshot in time" of the
presently available technology in this field. The approach has been to
survey current literature to determine the extent to which surface-based
electrical geophysical methods have been used to map plumes of contaminant
fluids from injection wells, and the success of these efforts to date.
Emphasis has been placed on the definition of the problem: injection wells
and injected fluids as electrical targets.
In 1974, Congress mandated the EPA (Safe Drinking Water Act, Part C)
to promulgate regulations to protect underground sources of drinking water
(USDW's) from improper injection practices. As a response, injection wells
were regulated by EPA in 1980 in the Underground Injection Control
Regulations which today are codified in 40 CFR, Parts 124, 144, 145, 146,
and 147. Because injection takes place at depths of hundreds or thousands
of feet below the surface, direct monitoring of injected fluid movement has
usually been difficult or impossible. The construction of monitoring wells
for the purpose would be very expensive, and such monitoring wells would
themselves present potential conduits for undesirable fluid movement.
Monitoring of the injection process is usually limited to such data as
injection pressures and volumes and the chemical characteristics of the
injected fluids. Some investigators have also used nearby wells to measure
changes in pore pressure as injection occurs. Monitoring wells in
surficial aquifers have also been used. Electrical geophysical
measurements at the surface have also been suggested as a possible means of
monitoring injected fluid movement; surface-based electrical measurements,
if feasible, would be much less expensive than the construction of
monitoring wells. The feasibility of electrical measurements is still
unresolved as of the preparation of this document.
To help understand the kinds of ground-water contamination problems
which can be created by injection wells and the potential uses of
surface-based electrical geophysical methods in mapping these problems,
background material is presented on the Underground Injection Control
Regulations; on the types, numbers, and geographical distribution of
injection wells; on the volumes and chemical/electrical character! tics of
injected fluids; on pathways of migration for injected fluids; and on
injected fluids as electrical targets. With this background, the
1
-------
surface-based electrical geophysical technology currently available for
monitoring injected fluids is reviewed.
2.0 UIC Regulations
Underground injection began to be widely used about 50 years ago in
secondary petroleum recovery. In the 1930's underground injection became a
common alternative to surface disposal of produced brines from petroleum
wells. Disposal of industrial wastes began in the 1950's as a method to
isolate difficult-to-treat wastes by placing them into -deep formations
where they would presumably remain through geologic time. Construction of
injection wells grew rapidly. EPA estimated the total number of injection
wells for all purposes at 500,000 in 1979; an inventory in 1984 placed the
total number at approximately 246,000, with oil and gas secondary recovery
operations accounting for about 167,000, roughly 68 percent of the total.
Hydrogeologic concepts behind the process are simple. In several
areas of the United States, the basement rock is covered by up to 20,000
feet of sedimentary rocks, which have been deposited over millions of years
and have remained relatively undisturbed. These rocks were stratified as
they were deposited, and the strata vary widely in composition, structure,
permeability, and porosity. They also contain water whose quality varies
considerably with depth. Generally, the total dissolved solids (TDS)
concentrations increase with depth. To place TDS concentrations in
perspective, the health-based national (secondary) drinking water standard
for TDS is 500 mg/1; California classes waters with TDS > 2100 mg/1 as
"unsuitable under most conditions" for irrigation (McKee, 1963). The
Underground Injection Control Regulations protect ground water with TDS
< 10,000 mg/1 since such water could conceivably be rendered potable
through treatment. Seawater usually has TDS of about 35,000 mg/1. In
comparison, brines associated with oil and gas production generally contain
30,000 to 100,000 mg/1 TDS. The fact that there are large differences
between the composition of surficial and deep water indicates that the
various impermeable strata act as barriers to the upward movement of the
deep saline water. Sedimentary rocks with adequate permeability,
thickness, depth, and area! extent provide the most satisfactory injection
zones. The location of such thick sedimentary sequences is one important
factor in determining where deep well injection can occur.
The engineering of injection wells was based on oil-field technology
and was further developed by industrial firms to dispose of specific waste
streams. A typical injection well is several thousand feet deep and
injects wastes directly into highly saline permeable injection zones. As
shown schematically in Figure 1.1, the well consists of concentric pipes.
The outer pipe or surface casing ideally extends below the base of any USDW
and is cemented back to the surface. Two pipes extend to the injection
zone, the long string casing which is also usually cemented back to the
surface, and the injection tubing, placed within the long string. Waste is
injected through the tubing and perforations in the bottom of the
long-string casing. The space between the tubing and the casing (called
the annulus) is closed off at the bottom by a device called a packer, which
keeps injected fluids from backing up into the annulus. This annular space
2
-------
Gauge on injection tubing
Gauge on tubing-casing annuluse^o r^g;
" Water table ~^"^
Land surface
400-
Feet
800-H
1200H
1600—1
2000—f
2400—1
2800—1
3200—1
3600—|
4000—J
4400H
Surface casing»»-J
Annulusi
[positive pressure)
Cement
Long-string jjj
Injection tubing
Perforations
| Surficial aquifer (USDW)
| Confined aquifer (USDW)
Saline aquifer >10,000 IDS
P Saline aquifer <30,000 TDS
| Confining zone
Injection zone >30,000 TDS
Figure 1.1. Ideal injection well and site.
3
-------
is typically filled with an inert, pressurized fluid. The inert fluid is
kept at a higher pressure than the injection pressure in the tubing to
prevent escape of the waste into the annulus if a leak should occur.
Capping the well is the well head, which contains valves and gauges to
control and to monitor injection.
Underground injection came under Federal control in 1974 with the
enactment of the Safe Drinking Water Act; Part C of the Act established the
Underground Injection Control (UIC) Program to protect underground sources
of drinking water (USDW's). The law required the Environmental Protection-
Agency to establish minimum standards and technical requirements which the
States were to adopt before assuming primary enforcement responsibility.
(primacy). The UIC regulations which were adopted in 1980 have the
following chief provisions:
They defined underground sources of drinking water (USDW's) as
the resource to be protected; the definition includes all
aquifers containing water with less than 10,000 mg/1 TDS.
They categorize injection wells into five classes:
Class I wells inject hazardous and non-hazardous waste below
the deepest USDW.
Class II wells are associated with oil and gas production and
with hydrocarbon storage and comprise the majority of all U.S.
injection wells.
Class III includes special process wells used in conjunction
with solution mining of minerals, in situ gasification of oil
shale, coal, etc.
Class IV wells inject hazardous wastes into or above USDW's.
These wells were banned by the 1984 amendments to the Resource
Conservation and Recovery Act of 1976 (RCRA), Section 7010, and
by the UIC regulations (40 CFR, Part 144.13).
Class V wells are non-hazardous waste injection wells that do
not fit into the other four classifications: recharge wells,
air conditioner return flow wells, drainage wells, etc.
Minimum technical requirements were established to ensure that
injected fluids will go into the proper horizon and remain
there.
The technical requirements include the following:
siting in areas which are free of faults and abandoned wells
and which possess adequate confining zones;
construction requirements for casings, tubing and packer,
cementing, logging, and testing;
-------
operation to prevent fracturing of the injection zone;
monitoring which includes periodic testing of the integrity of
the well and reporting; and
plugging and abandonment procedures which include demonstration
of financial responsibility.
For a State to have a federally-approved UIC program and to assume
primacy, it must meet minimum regulatory standards. The 1980 amendments to
the SDWA created Section 1425. States can be delegated primary
responsibility for oil and gas injection wells if they can show that their
program is effective in protecting USDWs and includes permitting,
surveillance, reporting, and enforcement. The EPA oversees State programs
to ensure that the standards are implemented. EPA has to follow the same
standards in States where the Agency implements the UIC program. As of
January 9, 1987, 33 States had been approved for full primacy an J 6 for
partial primacy. EPA is implementing a full program for 18 States and a
partial program in 6.
3.0 Injection Wells
3.1 Numbers and Types of Injection Wells
In 1983 the EPA conducted a nationwide inventory of injection wells,
utilizing a computerized system called FURS (Federal UIC Reporting System).
FURS contains data on owner, operator, location, well status, general
comments, date of construction or conversion, volume injected, injection
pressure, and construction features. As of November 1983, the total number
of injection wells in the inventory was approximately 240,000; this figure
included wells in the U.S. territories of Puerto Rico, the Virgin Islands,
Guam, the Pacific Islands Trust Territory, and American Samoa. For the
purposes of this report, only wells in the 50 states will be discussed.
Table 1.1 summarizes the results of this inventory, showing the number of
injection wells known to FURS as of November 1983 by class and EPA region
for the 50 United States (Fairchild, 1984). Regions VI, VII, and V had the
largest numbers of wells in this inventory and have had the largest numbers
historically (Fairchild, 1984). Appendix 1 presents a table of the
distribution of injection wells by class, State, and EPA region. Figures
1.2, 1.3, and 1.4 graphically summarize the data in Appendix 1. Figure 1.2
shows that EPA Regions V, VI, and VII together contained 70 percent of the
approximately 246,000 injection wells in the 1983 EPA inventory. Figure
1.3 shows that 6 states — Texas, Kansas, Pennsylvania, Illinois, Oklahoma,
and California--together contained about 73 percent of the national total.
Figure 1.4 shows that the most numerous well types nationwide were Class 2
(68 percent of the total), Class 5 (19 percent), and Class 3 (14 percent).
However, the contributions of the different well types to the totals vary
considerably in individual states, as shown by Figures 1.5 and 1.6. Figure
1.5 ^hows oil and gas injection wells (Class 2-60 percent) and '.'>lution
mining wells (Class 3-39 percent) predominating in Texas. Figure 1.6 shows
-------
Table 1.1. 1984 EPA INJECTION WELL INVENTORY
EPA
REGION
I
II
III
IV
V
VI
VII
VIII
XI
X
Class
I
0
23
12
127
129
264
59
5
39
1
Class
II
0
3859
4444
5936
27395
71195
31323
7943
14374
161
Class
III
0
249
41
10
112
31080
394
991
490
0
Class
IV
0
0
17
20
0
1
0
2
27
11
Class
V
0
3651
18827
5149
6001
595
1978
24
43
9114
Regional
Totals
302
7782
23341
11242
33637
103135
33754
; 8965
14973
9287
Totals
659 166630
33367
78
45684
246418
-------
Region VI (42%)
Region V (14%)
Region IV (5%)
Region III (9%)
Region II (3%)
Region I (0%)
Region X (4%)
Region IX (6%)
Region VIII (4%)
Region VII (14%)
Figure 1.2. Injection well totals by EPA region.
-------
KS
PA (8%)
IL (8%
32%
OK (6%
All other states (27%)
CA (6%)
Figure 1.3. Injection well totals by state.
8
-------
Class 2 (65%
Class 4 « 1%)
Class 1 « 1%)
Class 5 (19%)
Class 3 (14%)
Figure 1.4. Injection well totals by we)) type.
9
-------
Class II (60%)
Class I,!V,V
• « 1%)
Class III (39%)
Figure 1.5. Texas injection wells by well type.
10
-------
Class IV «1%)
Class II (20%)
Class I,III «1%)
Class V (80%)
Figure 1.6. Pennsylvania injection wells by well type.
11
-------
20 percent of the Pennsylvania state total in oil and gas injection wells
and 80 percent in Class 5 ("other").
The numbers in the 1983 EPA inventory should be regarded as
preliminary and indicative only of the relative geographic distribution and
the relative numbers of wells in the various classes. For example, Table
1.1 (from the 1983 inventory) shows a nationwide total of 670 Hass 1
(waste disposal) wells; in a May 1985 report to Congress on injection of
hazardous waste, EPA identified 252 existing Class 1 wells (EPA, 1985). In
particular, the numbers of Class 5 wells in the 1983 survey are suspect
since the various states had not yet completed their own inventories of
Class 5.
3.2 Geologic and Chemical Characteristics
3.2.1 Available Information
Fairchild (1984) reviewed available information on injection wells in
1984 and concluded that nationwide characteristics for the five classes of
injection wells could not be identified. One reason is the lack of
information; comprehensive data collection on injection well numbers, uses,
and characteristics did not begin until the late 1960's. Organized data
for Class II wells is probably most complete; state files contain
considerable information on Class I wells particularly on their, operation
and maintenance, but such data are not readily accessible. Data on Class
III wells are scarce because such wel.ls are few in number, many are
experimental, and their data are regarded as proprietary. Since states
were not required to inventory and assess Class V wells until 3 years after
achieving primacy, data on their numbers and characteristics are nearly
nonexistent.
Completeness of available information depends on the particular
characteristic of interest. Much has been published about recommended or
proper injection well construction, but there are relatively few published
case studies of actual installations for each of the five classes. Data on
general waste characteristics are available for particular industries, and
chemical analyses of injected wastes can be found in state files. The
literature contains little geological information except on large scales.
Applications approved under the UIC program contain limited hydrological
information such as location and depth of water wells, monitoring wells,
piezometric gradients, and depth to the base of fresh water.
3.2.2 Hydrogeology
As mentioned above, the literature contains several discussions of
recommended injection well construction practices but relatively little
about actual case studies. The recent report made by the EPA to Congress
on Class I wells summarizes characteristics of 252 waste disposal wells
(EPA, 1985). Fairchild (1984) summarized available information on wells of
all types. Rocks are generally classified by their origin as igneous,
metamorphic, or sedimentary. While nearly all rock types may occasionally
have the proper engineering characteristics to serve as injection zones,
12
-------
sedimentary rocks are most likely to have sufficient porosity,
permeability, thickness, and area! extent to permit the rock to act as a
liquid storage reservoir at safe injection pressures. Sedimentary rocks
with sufficient porosity and permeability to accept relatively large
volumes of fluids in the unfractured state include sandstones, limestones,
and dolomites. Naturally fractured limestones and shales may provide
satisfactory injection horizons, since oil and gas are commonly produced
from such strata. Impermeable confining strata must overlie and underlie
the injection zone to prevent the vertical migration of injected fluids.
Unfractured shale, clay, salt, anhydrite, gypsum, marl, and bentonite often
provide good seals against the upward flow of oil and gas. Most
sedimentary rocks with appropriate characteristics for injection wells were
deposited in marine environments and contain saline water at depths below
the present level of fresh water circulation.
Folding and fracturing of sedimentary strata must also be considered
in the evaluation of an injection well site. Structural geology on both the
regional and local scale influences subsurface liquid flow, the engineering
properties of rocks, the location and distribution of minerals, and
earthquakes. Simple folds are either synclines (downward folds) or
anticlines (upward folds). Synclinal basins are particularly attractive as
injection well sites because they contain relatively thick sequences of
saltwater bearing sedimentary rocks, and because the subsurface geology of
these basins is usually well-known from oil and gas exploration.
3.2.3 Hydrogeology of Class I Wells
The EPA 1985 report to Congress on hazardous waste disposal with
injection wells summarized the pertinent geologic characteristics of 252
Class I wells (EPA, 1985); summaries of this information are given in
Appendices 2 and 3. The EPA found that, nationwide, most of the injection
wells are constructed to inject wastes into sand and sandstone formations
(76 percent); fewer are constructed to inject into limestone or dolomite
formations (14.3 percent) and sandstone shale formations (9.7 percent).
The EPA found that the most common confining zones were composed of shale
(42.7 percent), sandstone shale (20.8 percent) and limestone shale (10.0
percent). The rest of the confining zones (for the 252 Class I wells) are
composed of silt, clay, dolomite, and other impermeable materials.
The greatest concentrations of Class I wells are found in the Great
Lakes and Gulf Coast areas. The EPA found that Class I injection wells in
the Great Lakes vicinity typically inject into the Mt. Simon sandstones
with an average thickness of about 600 feet or into dolomite at an average
depth of about 2,500 feet. In the Great Lakes area, confining zones of
shale with some limestones, dolomite, or siltstone average about 630 feet
thick. The average separation between the lowermost USDW and the injection
zone in the Great Lakes was found to be about 2,300 feet (EPA, 1985).
Figure 1.7 shows these average Great Lakes Class I injection well regional
characteristics in an idealized cross-section.
13
-------
feet
Base of fresh water
2500 feet
pooooooooooooooooooooooo
oooooooooooooooooooooooc
oooooooooo ooooooooc
OOOOOOOOOOSHALE OOOOOOOOO
oooooooooooooooooooooooc
JOOOOOOOOOOOOOOOOOOOOOOO
t
Confining zone
630 feet
; SAND. SANDSTONE. ;
LIMESTONE, DOLOMITE
Injection zone
600 feet
I
Figure 1.7.
Average hydrogeological
wells in the Great Lakes
characteristics
region.
of Class I injection
14
-------
In the Gulf Coast area, the injection zones were typically sand or
sandstone averaging about 500 feet thick and lying at an average depth of
about 4,600 feet. The Gulf Coast confining zones were mainly shale with
some marl or clay and averaged about 1,000 feet thick. The average
separation between the lowermost USDW and the injection zone is about 3,300
feet in the Gulf Coast (EPA, 1985). Figure 1.8 shows the averaqe Gulf
Coast Class I injection well characteristics as an idealized cross-section.
The EPA found a nationwide average for Class I injection well
thicknesses of about 560 feet, under a confining zone with an average
thickness of about 930 feet. The nationwide average depth to the top of
the injection zone was about 4,100 feet, and the nationwide average
separation of the lowermost USDW from the top of the injection zone was
about 2,900 feet (EPA, 1985). Table 1.2 summarizes the Class I
hydrogeological characteristics for the Great Lakes and Gulf Coast regions
and the nationwide (by well) averages.
3.2.4 Hydrogeology of Class II Wells
About 68 percent of all injection wells fall into Class II, as
indicated by Figure 1.4. Class II includes wells injecting fluids (1)
brought to the surface in the process of conventional oil or natural gas
production; (2) to enhance recovery of oil or natural gas; o-r (3) for
storage of hydrocarbons which are liquids at STP. Because of the large
numbers of Class II wells (over 166,000), they probably constitute the most
important injection category in terms of their impact on groundwater and
USDW's. In a 1974 EPA national survey, Keely (Scalf, 1983) ranked Class II
brine injection wells as the sixth and seventh worst causes of ground-water
pollution in the northwest and southeast, respectively (Scalf et al, 1973).
In a 1980 prioritized listing, EPA ranked petroleum exploration and
development as the second worst specific source of ground-water
contamination in the U.S. (Miller, 1980).
Perhaps because of the large numbers of Class II wells, national
summary statistics of their hydrogeological characteristics are not
available in the literature. While these data are probably available in
State and petroleum company files, compilation of summary statistics would
be a very large task. Fairchild was unable to present nationwide statistics
in her survey (Fairchild, 1984). She did describe practices in several
wellfields across the country. In most Class II wells, fluids are injected
into the "pay zones" of producing fields. In general, such practices occur
or will occur in most (if not all) petroleum and natural gas producing
fields in the country. In many cases, fields which are exhausted or nearly
exhausted are used for storage of gas or petroleum products produced
elsewhere. In her 1984 summary, Fairchild describes a gas storage operation
in the Richfield formation in Upper Peninsula Michigan; the Richfield had
been a producing formation in previous years.
15
-------
I
1300 feet
Base of fresh water
4600 feet
oooooooooooooooooooooooc
oooooooooooooooooooooooc
)OOOOOOOOOOOOOOOOOOOOOCO
OOOOOOOOOOSHALEOOOOOOOOOC
3OOOOOOOOOOOOOOOOOOOOOOO
DOOOOOOOOOOOOOOOOOOOOOOO
5OOOOOOOOOOOOOOOOOOOOOOO
t
Confining zone 1000 feet
SAND.SANDSTONE
Injection zone
500 feet
Figure 1.8. Average hjdrogeological characteristics
wells in the Gulf Coast region.
of Class I injection
16
-------
TABLE 1.2. HYDROGEOLOGICAL CHARACTERISTICS OF CLASS I INJECTION WELLS
Characteristic Great Lakes Region
Nationwide Average
Gulf Coast Region by woll
Principal
Injection zone
Injection zone
thickness (ft)
Injection zone
depths (ft)
Principal
confining zone
rock types
Confining zone
thickness (ft)
Separation from
lowermost USDW
Sand, sandstone;
limestone, dolomite
600
2500
Shale
630
2300
Sand, sandstone
500
4600
Shale
1000
3300
560
4100
' 930
2900
17
-------
Table 1.3 presents hydrogeologic characteristics from some of the
cases described in Fairchild's summary. In the absence of summary
statistics for Class II injection well hydrogeology, Table 1.3 will be
assumed to be generally representative of characteristics found nationally.
3.3 Volumes and Characteristics of Injected Fluids
3.3.1 Class I Wells: Injected Fluids
The EPA compiled data on waste characteristics of 108 hazardous waste
injection wells for the year 1983 (EPA, 1985). Utilizing their annual flow
volumes and waste concentrations, the EPA found that during 1983 the 108
wells disposed of 228,021,900 gallons of non-aqueous waste with 6.2 billion
gallons of water. These figures would yield an average dissolved solids
concentration of about 35,500 mg/1, with an average injected volume of
about 59 million gallons per well. The composition of this waste is given
in Table 1.4. Table 1.5 lists the individual waste components classified
as either acids, heavy metals, organics, or hazardous inorganics.
3.3.2 Class II Wells: Injected Fluids
Class II wells used for brine disposal constitute a large percentage
of all disposal wells. In 1970, 3.5 billion barrels of oil were produced
in the United States, and most of the associated brine was reinjected
through approximately 7,400 saltwater disposal wells. Half of these wells
were in Texas, and another 1,500 existed in Oklahoma (Scalf et al., 1973).
For 1980, it was estimated that 11 billion barrels of brine (equivalent to
140,000,000 tons of salt) would be injected back into the ground (Miller,
1980). (This corresponds to an approximate average brine concentration of
55,000 to 60,000 mg/1.)
From a water pollution standpoint, brine waters are of concern because
of the presence of dissolved ions. Those dissolved ions which are most
commonly present in greater than trace amounts are sodium, calcium,
magnesium, potassium, barium, strontium, ferrous iron, ferric iron,
chloride, sulfate, sulfide, bromide, and bicarbonate. Dissolved gases in
these waters often include carbon dioxide, hydrogen sulfide, and methane
(Collins, 1974). The chlorides of sodium, calcium, and magnesium comprise
95 percent or more of the dissolved salts in most brines. Table 1.6
contains a summary of some brine analyses from various locations in the
United States. An analysis of ocean waters is included for purposes of
comparison (EPA, 1977). Brine waters from different strata vary
considerably in their dissolved chemical constituents, and this makes
difficult the identification of a water from a particular strata. As some
petroleum wells become older, the produced fluids may be more than 99
percent brine. Produced brines differ in concentration but usually consist
primarily of sodium chloride in concentrations ranging from 5,000 to
400,000 mg/1 and averaging 50,000 mg/1, 2.5 times the NaCl concentration in
seawater.
18
-------
TABLE 1.3. HYDROGEOLOGICAL CHARACTERISTICS OF CLASS II INJECTION WELLS
Characteristic
Pennsylvania
Michigan
gas storage
Michigan
Oil Production
Oklahoma
saltwater disposal %
waterflood well
(Anadarko "A" No. 1)
Princ ipa1
injection zone
rock types
Devonian sands
Stray sandstone
(Mississippian)
Richfield formation
(dolomite pay zones)
Runnymede silt
Injection zone
thickness (ft)
12
12
67
Injection zone
depths (ft)
600 Co 2000
1300
4613
1407
Principal
confining zone
rock types
Shale
Gypsum, limestone
Shale
Confining zone
th ickness (ft)
1000
Separation from
lowermost USDW
-------
TABLE 1.4. "WASTE CHARACTERISTICS OF 108 CLASS I INJECTION!
ACTIVE IN 1983 IN THE UNITED STATES
Waste type
Acids
Heavy metals
Organics
Hazardous
organics
Non-hazardous
organics
Other
Gallons
44140900
1517600
39674500
89600
118679700
22964600
(EPA 1985 )
Percent of
total
gallons
20.26
.7
17.4
.04
52.04
9.91
Pounds
367250000
12626100
3300900,00
745800
987410000
191070000
Well
count
35
19
17
4
50
33
Total
(non-aqueous)
227066900
100.35
1889191900
Actual total 228021900
(minus overlaps, e.g. "organic acids")
20
-------
TABLE 1.5. HAZARDOUS WASTE STREAM COMPONENTS AND CONCENTRATIONS
IN CLASS I INJECTION WELLS IN THE UNITED STATES IN 1983
I
Waste stream Waste
type components
Acids Hydrochloric acid
Sulfuric acid
Nitric acid
Formic acid
Acid, unspecified
Heavy metals Chromium
Nickel
Metals, unspecified
Metal hydroxides,
unspecified
ncidence of
injection
by wells
15
6
2
2
12
1 1
5
2
1
Average
concentration
(ag/1)
78573
43000
75000
75000
44900
1 .4
600
5500
1000
Organics
Total organic carbon
(TOO
Phenol
Oil
Organic acids
Isopropyl alcohol
Organic cyanide
Forma 1dehyde
Acteophenone
Urea "N"
Chlorinated organis
Formic acid
Organic peroxides
Pentachlorophenol
Acetone
Ni t r i le
Methacrylonitrile
Ethylene chloride
24
22
6
3
3
3
2
2
2
2
2
2
2
2
1
1
1
1 141 3
805
3062
10000
1775
400
15000
650
1250
35000
75000
4950
7 .6
650
700
22
970
Hazardous
Selenium
. 3
Organics
Cyanide
391
21
-------
TABLE 1.6. ANALYSES OF OH. FIELD BRINES (in ppm)
Field
Kawkawlin
Ml
Seminole
OK
Glenn
OX
Nikkei
KA
Yates
IX
Monument
Shelby
MT
Frannie Dome
WY
Grass Creek
WY
Edison
CA
Ventura Avt .
CA
•p
Formation Chloride
Dundee
limestone 161,200
Wilcox
sand 89 , 990
Arbuckle
limestone 101,715
Hunton
limestone 76,797
Sanandres
dolomite 2,518
Gray burg
limestone 6,630
Madison
limestone 1,179
Tenaleep
sand 27
Frontier
sand 256
Upper Duff
sand 79
Pico Repetto
sand 14,212
Car- Blear- Mag-
Sulfate bonate bonate Sodium Potassium* Calcium nesium Misc. Total
155 - 60 66,280 - 25,740 4,670 258,105
515 - 65 44,020 - 9,460 1,990 146,060
120 - 60 50,345 - 10,160 2,120 164,520
207 - 61 40,284 - 5,440 1,790 124,579
2,135 - - 1,624 - 587 288 7,445
160 - 1,740 3,735 - 515 365 13,145
659 71 1,270 1,322 - 143 66 3,388
2,303 0 691 51 - 760 240 4,022
6 1,211 - 1,087 - 52 2,565
4 29 648 299 - 17 1 962
59 - 1,846 8,607 - 729 242 26,091
(Continued)
-------
TABLE 1.6. ANALYSES Of OIL HELD BRINCS (in ppm)
field formation Chloride
Bay City Salina
HI dolomite 403,207
Sharon Sere a
P* sand 6,740
Evnna City 3d Venango
PA sand 88,820
Bell Hun Salt
WV sand 64,9)0
Reed City Msrston
MI dolomite 156,225
B-ifbank Bart lesvi 1 le
CiK aand 107,895
East fe»as Woodbine
U sand 40,958
Vent urn Pico-Re-
CA petto sand 14,212
1 rjit^ale fairhaven
'_ A s«nd 245
-;>"r. i;i Ha'"rs
:ft;,\n } - 19,i'0
Car- Bicar- Mag-
Sulfate bonate bonate Sodium Potassium* Calcium nesium Misc. Total
0 - 1,208 21 - 206,300 7,300 642,798
0 0 250 3,440* 700 160 11,290
180 0 40 - 38,660* 13,790 2,220 143,710
0 0 90 - 30,450* 7,580 1,620 104,670
265 - 20 59,080 - 28,440 5,155 249,185
35 50,000 - 14,340 1,875 174,145
278 - 569 - 24,540* 1,388 282 67,649
59 - 1,846 8,607 - 729 242 Nllj.-66;
51:170;
re,Al:160 26,091
11 60 2,235 - 1,024" 0 10 3,595
2,700 70 - 10,710 - 420 1,300 35,000
'in loot brine analyses Na and K are deter.nined together and reported as No. For those cases, however, where the alkali content was
listed *-.. Va * K, tho vsluns are tubulated between the to and K columns. A similar uncertainty exists in the case of some of the
r'-p'">rl*d >ali*»ri ofC03 and HCOj concunt rut ionn.
-------
Brine disposal wells in oil fields range considerably in injection
rate and depth. Injection pressures vary from a vacuum to 2,000 psi, and
flow rates range from 35 to 150 gpm per well. One well in the Arbuckle
formation is reported to have taken water at an average rate of 17,730
barrels per day (750,000 gpd) for periods of 600 days. If the operation is
one of combined discharge-recharge, the well depth is the same as the
producing oil formation. However, if it is desired to merely disnose of
the brine, the well depth is dictated by the depth of the formation
suitable for injection.
As an example of a salt water disposal operation, Fairchild (1984)
cited the East Texas Salt Water Disposal Company of Kilgore, Texas. This
company was organized in January 1942 and was subscribed to by about 250
large and small operators in the East Texas oil field. In ,1958 there were
82 injection wells operating in the East Texas oil field which returned 165
million barrels to the reservoir. 60 of these wells, operated by the East
Texas Salt Water Disposal Company, returned 138 million barrels of brine to
the reservoir, an average of roughly 2 million barrels/year per well.
Figures 1.9 and 1.10 show schematically the significant geometrical
factors for two Class II wells in Oklahoma, as given by their permit
applications to the Oklahoma Corporation Commission (Fairchild, 1984).
Figure 1.9 shows a Class II well injecting brine into a 67-foot thick silt
formation at a depth of about 1,400 feet. In this well, the Cprporation
Commission limited injection pressures to 300 psi and the brine' injection
rates to 300 barrels/day. Brine TDS concentrations are reported as about
60,000 mg/1 (95 percent NaCl). Figure 1.10 shows brine injection into a
114-foot thick lime formation at a depth of about 3,900 feet. Here
injection pressures are limited in the permit to 2,000 psi, and the
injection rates are limited to 2,000 barrels/day. Brine TDS concentrations
are reported as about 166,000 mg/1 (92 percent NaCl). In both cases, the
principal confining zones are shale.
4.0 Pollutant Pathways
Some possible causes of leakage and subsequent contamination from
injection well operations include: (a) corrosion of injection well casing;
(b) fracturing of reservoir and confining units; (c) poor cementation
between the well casing and borehole; and (d) conduits to overlying USDW's
(abandoned wells and existing fractures in the confining zones). Figure
1.11 schematically depicts these various pollutant pathways and suggests
the geometry of their attendant injected fluid migration.
However, in cases where wells are improperly constructed and operated
or in cases in which fractures connect the injection zones with overlying
fresh water aquifers, movement of the highly conductive injected fluids
into the overlying fresh water aquifers could create a conductive
electrical target which contrasts sharply with the surrounding material.
In such cases, electrical methods may indeed be useful in detecting and
mapping the contaminated zones. Unfortunately, the literature does not
contain case studies delineating the actual geometries and conductivities
of such contamination incidents.
24
-------
Fresh water
IDS 400 mg/
T
500 feet
SHALE AND RED BEOS
Base of fresh water
(10,000 mg/l IDS)
RED BED]
1407 feet
oooooooooooooooooooooooo
OOOOOOOOOCSHALEOCOCOOOOO
oooooooooooooooo o ooooooo
Injection zone
(Runneymede silt)
Injection fluid TDS
= 60,000 mg/l
(~ 95% NaCI)
67 feet
Texas county, Oklahoma
Injection pressure -• 300 psig
Injection rate » 300 BWPD
injection zone porosity •* 10°'o
Figure 1.9. Idealized cross-section
injection we!1).
of Anadarko Uell No. "A" No. 1 (Class II
25
-------
oooooooooooooooooooooooc
>OOOOOOOOOOOOOOOOOOOOOOO'
5OOOOOOOOOOOOOOOOOOOOOOO
OOOOOOOOOcSHALE'OOOOOGOOOC
oooooooooooooooooooooooc
3OOOOOOOOOOOOOOOOOOOOOOO
2OOOOOOOOOOOOOOOOOOOOOOO
/ ° / ^ 1 C I ~ I ^ 1 Z i -* / 0 / ~" / O i
/I/I/I *te/l/*/f/l/f/
( 0 /.° / SHALE AND SAND ~ 3 ,
(-> / O / w /;://-) . ij ; ^ / ~ / ^ / a / <
0 /° /°/o /C /O /O /C /^ /0/C
o/o/o/o/i'/o/o/o/^/o/c
2/n/Q./o/-/^/o/o / - / o / o
T/a/c/^/-/C/G/0/X/0/0
? /°/0/ X ~ o ° ° /o/°/0
ililmjiiwilili,
III! l+l III!
oooooooooooooooooooooooo
oooooooooooooooooooooooc
>ooooooooooooooooooooooo<
IOOOOOOOOO SHALE3OOOOOOOOO
3OOOOOOOOOOOOOOOOOOOOOOOI
(OOOOOOOOOOOOOOOOOOOOOOOC
oooooooooooooooooooooooc
• ••••LIME (Hunton-Viola)»««««
•••••••••••••••••••••A*
'
; 200 feet '
/
Base of fresh water
no. 000 mg/l IDS)
1060
1670
• 3880 f<
1
k
3et
r
Injection zone
/
Injection fluid IDS 114 feet
= 165,984 mg/l
(-92% NaCI)
Sun well no. 6-46
Pontotoc county, Oklahoma
Max. injection pressure = 2000 psi
Max. injection rate = 2000 B/D
Injection zone porosity = 11%
Figure 1.10.
Idealized
well).
cross-section of Sun Well No. 6-41 (Class II injection
26
-------
ISJ
Injection well
Abandoned well
Underground source
of drinking water
Fractures
Casing seepage
Injection zone
Figure 1.11. Migration pathways for injected fluids.
-------
II. ELECTRICAL PROPERTIES OF FLUIDS, ROCKS, AND GEOLOGIC FORMATIONS
The preceding section discussed injection wells as a water quality
problem and attempted to present some general information on their
hydrogeology, on the volumes and chemical properties of the injected
fluids, and en the potential pathways for migration of injected fluids. To
understand injected fluids as electrical targets, we now need to discuss
the effects of soils, rocks, and their pore fluids on the electrical
conductivity of subsurface strata.
The bulk resistivity of a rock formation depends on the resistivity of
the contained fluids, the resistivity of the rock itself, and on the
geometry in which the two phases coexist. In clays and shales (and in
metallic mineral deposits) the rock itself is conducting. In clean aquifers
(i.e., aquifers with no clay or shale component), however, the resistivity
of the rock matrix (e.g., sand) is usually so large as to have no influence
on the conduction of current.
Electrical resistivity is defined by Ohm's law:
->
->
J • E/p
where J is the current density in amperes/meter2, E is the electric
field in volts/meter, and p is the resistivity in Ohm-meters. A
homogeneous material of resistivity p, cross-section A, and length 2(see
diagram below) has a total resistance along its length of
-^
0
X
^-
^ (
R =» p«/A (i)
Now consider a non-conducting rock which has a porosity
-------
F = (pt/pj = a/(tpm) (ii)
where a and m are constants and F is known as the "formation factor." This
equation implies that, in a clean formation with constant porosity, the
bulk resistivity p (which can be measured with surface and borehole
geophysical methods) is linearly related to the resistivity of the pore
water. Alternatively, if the pore water is unchanged, it relates changes in
bulk resistivity to changes in porosity. The exponent m is sometimes
called the cementation index and has a value in the range from 1 to 3
depending on the rock type. Usually, larger values of m are associated with
"tighter" rocks and low porosities (Greenhouse, 1982; McNeil!, 1980).
For poorly consolidated aquifers, Archie's Law is often, taken as
F = 0.62
-------
0.3 0.5
Resistivity (ohm-m)
2 5 10 20 50
100 200
I I
20000 10000 5000
2000 1000 500
200
•ill I li
100 50
-100
- 110
- 120
- 130
1-140
03
f—
C
03
03
03
03
03
Q
03
CO
l_
03
Q.
E
03
Conductivity (pmhos/cm)
Figure 2.1.
Electrically equivalent concentrations of a sodium chloride
solution as a function of resistivity (or conductivity) and
temperature (from Keys and MacCary, 1972).
30
-------
3000
10 20
Porosity, $ (%)
Figure 2.2.
Approximate resistivity of granular aquifers versus prosit/ for
several water salinities (from Guyod, 1966).
31
-------
TABLE 2.1: Contributions of Various Ions to
Fluid Conductivity
(after Greenhouse, 1982; McNeil!, 1980)
ION RATIO OF ION CONDUCTIVITY
TO CONDUCTIVITY-OF EQUAL
CONCENTRATION OF NaCl
Ca*2 0.95
Mg*2 2.00
K* 1.00
S04-2 0.50
1 0.27
1.26
0.53
32
-------
TABLE 2.2: Resistivities of Sediments
(after Telford, 1976)
Rock Type
Resistivity Range (ohm-m)
Consolidated Shales
Argil!ites
Conglomerates
Sandstones
Limestones
Dolomite
Unconsolidated wet clay
Marls
Clays
Alluvium and sands
Oil sands
20 -
10 -
2 x
1 -
50 -
3.5
20
3 -
1 -
10 -
4 -
2 x
8 x
10*
102
103 - 104
6.4 x
107
x 102
70
100
800
800
108
- 5 x 103
33
-------
In general, aquifer conductivity is directly proportional to the
electrolyte concentrations in the pore water and takes place through the
moisture-filled pores and passages which are contained within the (usually
insulating) aquifer matrix. The conductivity is determined for both rocks
and soils by
(1) porosity (shape and size of pores, number, size, and sivape of
interconnecting passages);
(2) moisture content (the extent to which the pore passages are filled
with water);
(3) concentration of dissolved electrolytes in the pore water (for us, a
• function of the salts present and of the pollutant concentration);
(4) amount and composition of clays and shales in the aquifer matrix.
With the foregoing theoretical considerations, the kind of highly
simplified cross-sectional schematics of injection wells presented earlier
in Figures 1.7 - 1.10 can be interpreted in terms of geoelectric sections:
vertical sections of the earth composed of horizontal layers of different
resistivities. Assume a "typical" Class I injection well has been
operating for 10 years and has been injecting about 60 million gallons per
year of wastes with a IDS concentration of about 40,000 mg/1. Furthermore,
assume that the porosity of the injection zone and the overburden above the
confining zone is approximately 10 percent and that the IDS concentrations
of fluids in the entire vertical column under the base of fresh water are
about 10,000 mg/1. Also assume that the overlying (confining zone) shale
has a resistivity which is about the midpoint of the range given by Table
2.2, about 200 ohm-meters (or a conductivity of about 5 millimhos/meter).
Then, by using Equation (iii) for Archie's Law and a rough relationship
between TDS concentrations and injected fluid conductivity (Equation v),
the conductivity of the cylinder of injected fluids can be estimated to be
about 65 millimhos/meter. The overburden (assumed to be free of clays)
would have a conductivity of about 16 mill imhos/meter. The TDS
concentration of the freshwater aquifer is assumed to be about 350 mg/1
which corresponds to a conductivity of about 0.6 millimhos/meter when a
porosity of about 10 percent is assumed. The diameter of the cylinder of
injected fluids will be about 450 feet after 1 year of operation, and about
1,400 feet after 10 years of operation. This geometry is indicated in
Figure 2.3 with these resistivity values converted to conductivities in
mi 11imhos/meter.
Figure 2.3 shows an injected cylinder about 450 feet in diameter and
about 500 feet thick, covered by a confining zone 1000 feet thick and an
overburden of another 3600 feet. The injected cylinder has a conductivity
about 4 times that of the rest of the injected zone and 13 times larger
than the confining layer. The overburden, up to the base of fresh water,
will have a conductivity about one-fourth of the injected cylinder This
geometry is not promising for the application of surface-based electrical
methods. A target which is about 4 times more conductive than its
34
-------
surroundings is buried at a depth approximately 10 times its diameter. In
a situation where the surface was free of interferences such as buried
pipes, fences, powerlines, buildings, and scrap metal, the injected
cylinder may be detectable. However, the use of surface methods to map the
outlines of this target would probably not be successful. However, a
series of annual measurements commencing prior to injection probably would
be able to detect a change in apparent resistivities over time. For
example, the results of computer model calculations indicate orvly a 14%
increase in apparent resistivity would be observed for the model in Figure
2.3 if the entire injection layer conductivity increased from 16 to 65
mhos/meter. These results are for a Wenner array with an electrode spacing
of 6000 feet. The apparent resistivity changes for the finite three
dimensional injection zone of Figure 2.3 would be less than this 14%. The
overall Wenner array electrode spread in the above example would be 18,000
feet. Such large spreads have poor lateral resolution for detecting deep
structure boundaries. In addition, changes in the near surface geology
would likely be encountered over distances of 18,000 feet and the
associated near surface variations in conductivity would easily nask the
detection of the deeper structures. However, a series of repeated
measurements commencing prior to injection probably would be able to detect
a change in apparent resistivity over time. These changes could then be
interpreted to give information on the injection zone.
35
-------
DOOOOOOOOOOOOOOOOOOOOOOO
>ooooooooooooooooooooooo
oooooooooooooooooooooooo
)OOOOOOOOOSHALEOOOOOOOOOO
oooooooooooooooooooooooc
oooooooooooooooooooooooc
DOOOOOOOOOOOOOOOOOOOOOOO
(IDS =» 350 mg/l)
Fresh water aquifer
T
1300 feet
;S 0.6 millimhos/meter)
1.
Base of fresh water
4600 feet
(IDS* 10,000 mg/l)
over burden
(cr=16 millimhos/meter)
(0s 10%)
1000 feet
Confining zone
(o~s5 millimhos/meter)
t
0 1
I
Injection zone
(TDS310,000 mg/l) 500 feet
tcr = 16 millimhos/meter) X
Cylinder of injected fluids
(IDS* 40,000 mg/l)
Cylinder of injected fluids
(CT265 millimhos/meter)
450 feet
Figure 2.3. Electrical conductivities corresponding to Figure 1.8,
36
-------
III. PHYSICS OF RESISTIVITY MAPPING
1.0 GENERAL
Measurement of the electrical properties of the subsurface with
geophysical tools employed from the surface is one of the oldest
geophysical techniques. Electrical techniques figure prominently in
ground-water contamination and salinity mapping because the resistivity of
formations is highly dependent on the concentration of dissolved solids in
ground water. Other physical techniques that measure such properties as
seismic velocity, density, and magnetic susceptibility are barely
influenced by dissolved solids in ground water and are, therefore,
ineffective in mapping changes in ground-water quality or contamination.
They are effective for other exploration objectives. In the past 60 years,
electromagnetic methods have been used principally for exploration for
mineral ores where exploration targets generally have had very high
conductivity values compared to the rocks hosting the ore. The objectives
in development of techniques consisted of detecting anomalous responses
from smaller and smaller ore targets at greater and greater depths.
The objectives in ground-water investigations are somewhat different.
First, the resistivity contrast between formations with different
concentrations of dissolved solids in ground water is often considerably
less than is encountered in minerals exploration. In the last 5 years,
equipment manufacturers have begun to address seriously the requirements of
geophysics for ground-water objectives. This trend has been helped along
by the increased use of electrical methods for hydrocarbon exploration in
frontier areas such as volcanic-covered terrain, the Rocky Mountain
overthrust, and for permafrost mapping in Alaska.
Although the use of .electrical techniques for ore detection has
different objectives than their use in ground-water exploration, much has
been learned in the last 60 years about electrical techniques. This
knowledge forms a firm foundation for applying electrical methods to
ground-water investigations.
The electrical resistivity of the earth is measured by determining the
response of the earth to current flow. The various methods in use differ
mainly in the manner currents are generated. Figure 3.1 schematically
illustrates the geometry of current flow in the subsurface for several
methods discussed in this report.
In both electromagnetic induction and magneto-telluric (MT) type
methods, the current flow in a horizontally-layered earth is horizontal and
has no vertical component. In the electromagnetic induction methods the
currents are horizontally-closed rings concentric about the transmitter,
for both horizontal and vertical co-planar coils (Kaufman et a/., 1983).
The current flows in horizontal sheets for the MT-type methods. In the
direct current (D.C.) method there is a vertical component of current flow,
so that currents will cross the boundaries of layers with different
resistivities.
37
-------
Method
Schematic
Comments
Magnetic
induction
Transmitter is a local
source which generates
a time-varying primary
magnetic field which
induces eddy currents
in the ground.
MT
AMT
CSAMT
Ez
Magnetotelluric f y
- type * »»
Ey
Ex
£
Transmitter is a
distant source (radio
stations, geomagnetic
storms).
Direct
current
Current flows from
one electrode to
the other.
Figure 3.1. Geometry of current flow in electrical methods
38
-------
Whenever a current crosses a boundary of change in electrical
resistivity, electrical charges will arise (Kaufman, 1985). Figure 3.2,
for example, shows the formation of charges in a horizontally-layered
medium for the DC method (Alpin, 1947). This phenomenon is called
electrostatic induction and is observed in any conductive medium regardless
of its resistivity. The phenomenon is further illustrated in Figure 3.3
with a conductive body embedded in a uniform earth. Under the influence of
an electric field, the positive and negative charges residing inside the
conductor move in opposite directions. As a consequence of this movement,
electric charges develop on both sides of the conductor. These charges in
turn create a secondary electric field (Coloumb's Law) inside the
conductive medium. For a perfect conductor, the electric charges
distribute themselves in such a way that the total electric field inside
the conductor will disappear. These charges form on boundaries of lateral
resistivity variations in all electrical methods and on horizontal layers
in the D.C. method. This principle of electrostatic induction holds for
steady-state as well as time-varying electric fields and for perfect and
finite conductors (Kaufman, 1985; Alpin, 1947).
It is common to all electrical methods to define apparent resistivity
and to transform potentials and voltages measured into apparent
resistivities. The purpose of this transformation is to obtain a good
visualization of how the behavior of the voltages observed over the study
area differs from the behavior, for identical system parameters, over a
hypothetical earth which is uniform in resistivity with depth. A common
definition for D.C. resistivity is:
where pa is the apparent resistivity and p1 .is the true resistivity of the
top layer (overburden). E(obs) and l(unp^) are the observed voltage and
the computed voltage over for a uniform earth having resistivity pr
respectively.
In the various systems of electrical resistivity measurements,
different system parameters are varied to increase effective exploration
depth: frequency in controlled source audiomagnetotel lurics (CSAMT),
electrode spacing in D.C., and time in transient electromagnetic induction
(TDEM) . Figure 3.4 shows two-layer apparent resistivity curves for
different techniques. It is evident from all these curves that the
geoelectric section (resistivity layering) can be visualized. When the
resistivity of the second layer is greater than that of the upper layer, pa
increases when effective exploration depth is increased (or decreases when
pjp^ < 1). The definition of apparent resistivity therefore encompasses
all of the different methods of electrical resistivity mapping. It is an
important unifying factor.
2.0 Direct Current (D.C.) Resistivity Method
In the D.C. method of prospecting, current is injected into the earth
by galvanic contacts (steel probes). The usual practice consists of
39
-------
' /;.'•"•''':-- •' ''•^^•Vr'i/?-'•*•' J
Figure 3.2. Current flow within a layered medium (Electric charges arise at
resistivity boundaries).
40
-------
Surface
I
E (Primary)
P, °r
Secondary) _J
Conductor at depth
iqufe 3.3.
Charge formation in response to a-i i"-pr
-------
10-2
104 103 TO2 101 10° TO"1 10 2
Frequency (Hz)
10-3
10 1 10° -"O1 102 103 104 105
L
Figure 3.4. Two-layer apparent resistivity curves for MT, IDEM, and DC.
42
-------
driving current flow through the earth between two electrodes, I,, and 1^
(Figure 3.2). The current flow causes potential differences to exist at
different points along the surface. From the measurements of these
potential differences the electrical resistivity of the earth can be
derived.
Because the D.C. method is the oldest electrical technique and has a
history of 60 years of practice and improvements, there are more tieatises
and case histories published for this method than for the other methods of
electrical resistivity mapping. Probably the most complete discussion on
the technique is by Van Nostrand and Cook, 1966.
The depth of exploration in the-D.C. method is mainly a function of
the spacing between the potential electrodes and current electrodes. The
electrodes can be arranged in different arrays. Two common arrays are
shown in Figure 3.5 (Schlumberger and Wenner). The distance between
current and voltage electrodes is designated by L in the Schlumberger array
shown in Figure 3.5. Figure 3.6 shows apparent resistivity curve, for a
two-layered earth as a function of the L-spacing. For example, if the
objective of an investigation were to detect the presence of a conductive
layer with a resistivity p2 under a resistive layer of resistivity pr and
if /5,/p1 = 0.20, then, at an L-spacing of about 2h,, the apparent
resistivity pa would be about 75 percent of the resistivity of the surface
layer. This3 is about the limit of detection in the presence ,of normal
geologic noise. Thus, exploration depth in the Schlumberger array is
approximately one-fourth of the spacing between current electrodes (2L).
The relatively large arrays (array length is approximately ten times depth
of interest) required for depth investigations limit the D.C. method in its
ability to resolve lateral variations at depth. A third array, however,
the dipole-dipole array, has slightly better lateral resolution at depth.
Examples of the use of vertical soundings and horizontal profiling are
given in Chapter 4, Section 4.2.
3.0 Electromagnetic Induction Methods
3.1 General
In electromagnetic induction, eddy currents are induced in the earth
by a time-varying magnetic field. This primary magnetic field is generated
by a local source, most often a coil of wire or a grounded line through
which a time-varying electrical current is driven. Figure 3.7 illustrates
several geometries used by commercially available systems. All these
systems have in common a relatively small separation between transmitter
and receiver. In fact, the separation distance between transmitter and
receiver is about equal to the effective exploration depth (for flat-lying
conductors over a layered earth).
Figure 3.7 also illustrates common current waveforms driven through
the transmitter loop. In one type of system, the alternating current driven
through the transmitter loop has a harmonic frequency waveform The
transmitter continuously emits a primary magnetic field. In other types of
systems, the current waveform has a 50 percent duty cycle.
43
-------
Schlumberger Array
a is held constant
L increases
///////
/ / / / / /////////
^ a
•>
p, L
Current
electrode
Wenner Array
equal spacing between electrodes
^•^
I
f
\
Potential
electrodes
a
a
a
1
Current
electrode
Figure 3.5. Schematics of Wenner and Schlumberger arrays.
44
-------
1000
100- —
10- —
0.001
o.i--
0.01--
0.001
0.01
L/hi
100
1000
3.6. Schlumberger array two-layer master curves of
apparent resistivity versus electrode spacing.
45
-------
Transmitter Receiver
Geometry
Current Waveform
sinusoid
0-
'7777
Examples
(Frequency sepaTation
ana coil orientation
may be varied)
Geonics EM-31
Geonics EM-34
MAX-MIN II
Slingram
sinusoid
Geoprobe
s//7
\
Geonics EM-37
Geonics EM-42
UTEM
Sirotem
Figure 3.7. Waveforms and transmitter-receiver geometry for electromagnetic induction.
-------
The eddy currents induced in the earth have similar geometry both for
the alternating harmonic frequency and for the 50 percent duty cycle
waveform. The eddy currents in a horizontally-layered earth consist of
closed rings concentric about the transmitter. The pattern of current flow
is the same for both horizontal loops (vertical magnetic dipoles) and
vertical loops (horizontal magnetic dipoles). The eddy current intensity
is a function of earth resistivity (Kaufman et al., 1983).
3.2 Continuous Electromagnetic Induction
Figure 3.8 further illustrates the operation of the systems. An
alternating current is driven through a transmitter coil, TX. The
time-varying magnetic field arising from the currents in the transmitter
coil induces very small currents in the subsurface. These currents generate
a secondary magnetic field, HS, which is sensed together with the primary
magnetic field, Hp, by a receiver coil. The receiver coil is separated
some distance from the transmitter coil. These systems were originally
developed to detect anomalies caused by highly conductive ore bodies.
Rigorous interpretation of data from layered-earth situations is difficult.
In a rigorous and complete solution, the secondary field is related by a
complicated function to earth resistivity, intercoil spacing (s), and
operating frequency (f). However, it has been shown that for certain
ranges of frequencies and intercoil spacings, the secondary field is
related to earth conductivity by approximations which are relatively simple
functions (Kaufman and Keller, 1983). These simplifications are
incorporated in the design of the Geonics EM31 and EM34 terrain
conductivity meters. The physical concepts of their operation are briefly
discussed below.
The secondary magnetic field H can be separated into two components:
(1) the component HSQ in quadrature (out-of-phase) with the primary magnetic
field H-, and (2) a component H in phase with Hp. Thus, the total magnetic
field, HT, at the receiver can be written as:
gr -ci - gsnogso (i)
HP np Hp
where subscripts I and Q refer to in-phase and quadrature components,
respectively, and i ~ J-l.
Typically, values for (Hs,/Hp) and (Hso/Hp) for the EM31 and EM34 are on
the order of 10"4. It is, therefore, very difficult to measure the first
term in Equation (i) accurately because the small value of (HSI/H ) caused
by currents in the earth must be measured in the presence of a v'alue lO1*
times larger which is caused by currents in the transmitter. For that
reason, only the term (Hso/Hp) is used to measure earth conductivity. That
term generally can be determined with an accuracy of 1 part in 104.
A well known characteristic of a homogeneous half-space is the
electrical skin depth 6, which is defined as the distance in the half-space
that a propagating place wave has travelled when its amplitude has been
attenuated to 1/e of the amplitude at the surface. The skin depth is given
47
-------
Figure 3.8. Operation of electromagnetic induction in a layered earth.
48
-------
by
6 = 72/ufioa
where
w =
f = frequency
H0 = permeability of free space
s = intercoil spacing (meters)
The ratio s/6, the intercoil spacing (s) divided by the skin depth 6,
is defined as the induction number B, whereupon
B = s/S - Jun
-------
The apparent conductivity which the instrument reads is then defined
by
oa = 4 H
-------
Figure 3.9. Schematic of horizontally-layered earth.
51
-------
a. - conductivity of the 7-th layer
R2(.i) = geoelectric factor1 of the ;-th layer
The geoelectric factor, Rz(i), is~a function of the depth and thickness of
the /-tl layer. The function RZ is readily available in mathematical and
graphical form. Figure 3.10 shows R(i) for both horizontal (RH) and
vertical (Rv) magnetic dipoles.
Commercially available systems operating on the principles outlined
above are now widely used in ground-water investigations. In general, their
effective depth of exploration is limited to about 30 meters (100 feet).
Their advantages, limitations, and applications are discussed in Chapter 4,
Section 2.1.
3.3 Transient Electromagnetic Systems
In this section, transient electromagnetic induction systems using 50
percent duty cycle waveforms are considered. Transient or time-domain
electromagnetic induction (TDEM) is a magnetic induction method, and in
many aspects its principles are similar to frequency-domain electromagnetic
induction profiling with the EM-31 and EM-34. A TDEM system consists of a
transmitter and a receiver. A common array, known as central loop
configuration, is shown in Figure 3.11. The transmitter is a square loop
of insulated wire laid on the earth surface. A multi-turn air coil wound
on a rigid mandrel (about 1 meter diameter) is employed in the center as a
receiver. The dimensions of the transmitter loops are a function of the
required exploration depth; in practice these dimensions have been varied
from 50m x 50m for about 100m exploration depth to 500m x 500m for
exploration depths in excess of 1000m. The array is moved for each
sounding, as shown in Figure 3.11.
A common current waveform driven through the transmitter loop is
shown in Figure 3.12a. The waveform consists of equal periods of time-on
and time-off. A primary magnetic field is associated with the current
driven through the transmitter loop (Kaufman and Keller, 1983). During the
rapid current turn-off ramp, this primary magnetic field is time-variant,
and in accordance with Faraday's Law, there will be electromagnetic
induction during the current ramp turn-off (Figure 3.12b). The
electromagnetic induction causes eddy currents to flow in the subsurface.
The intensity of these eddy currents at a certain time and depth is mainly
a function of earth resistivity; this fact is the basis for using TDEM to
derive subsurface resistivity layering.
In a horizontally-layered earth, the eddy currents are horizontal
closed rings concentric about the center of the transmitter loop. A
schematic illustration of eddy current distribution is shown in Figures
3.13a and 3.13b. Immediately after turn-off, the currents are concentrated
near th- surface (Figure 3.13a). With increasing time, currents are
induced at greater depth, and Figure 3.13b illustrates the eddy current
distribution at later times when currents have migrated. Thus, in a
horizontally-layered earth, there is no vertical component of current flow.
52
-------
2.0 Z
Figure 3.10. Electromagnetic induction response versus depth for horizonta
and vertical dipoles.
53
-------
Figure 3.11. Transmitter-receiver array used in IDEM soundings (the array is moved for each sounding)
-------
+-•
c
0>
l_
3
u
-*- Time-on •*•
/^ ^
/ \
/ \
/ \
•a*
'fi
(0<
cc
•*- Time-off -*• -*- Time-on -*•'
^
i
*^
Induced
electromo
force
/
;
/
i
i
\
i
b) Induced electromotive force caused by current turn-off
(EMF caused by turn-on not used.)
Measurement
during
time-off
T3 ® ^
o en .»
c) Secondary magnetic field caused by eddy currents
Figure 3.12. IDEM waveforms.
55
-------
(a)
(b)
Figure 3.13. IDEM eddy current propagation in a homogeneous earth.
56
-------
The time-varying eddy currents in turn cause a time-variant secondary
magnetic field (Figure 3.12c) which is measured as an electromotive force
(emf) in the receiver coil as a function of time.
Because a visualization of eddy current behavior and distribution is
critical to understanding IDEM soundings, one other dfsplay is given.
Figure 3.14 shows schematically the distribution of eddy current intensity
as a function of depth at several times. This figure shows that:
a) At an early time, t , currents are concentrated in the upper
layers. Consequentfy, a measurement of the emf at early time
will be sensitive primarily to the electrical resistivity of
the upper layers.
b) With increasing time the current intensity migrates to greater
depth. At later times the emf measured by the receiver will
progressively be more influenced by the resistivities of deeper
layers.
c) Thus, exploration depth is mainly a function of time rather
than of transmitter-receiver separation [as is the case for
direct current soundings and frequency-domain EM measurements
(EM-31 and EM-34)].
d) After a certain time, e.g., t-, most of the current density in
the near surface layers is small. In fact, for a
one-dimensional model, the intensity of eddy currents at late
time is independent of near surface conductivities. This means
that the electrical resistivities of near surface layers have
less influence on the measured emf at later times; measurements
. at later times have become transparent to surface layers
• (Newman et a/., 1986). This is of practical importance since
it reduces sensitivity to surface variations in resistivity,
often the main cause of poor data quality in other electrical
prospecting methods. Field measurements have shown, however,
that cultural effects, e.g., pipes near the surface, affect the
late time response.
Finally, Figure 3.15 is included to provide an approximate
visualization to correlate exploration depth to time of measurement. It
shows the distance from the center of the transmitter loop to the location
of maximum current intensity as a function of time. The rate of current
migration can be seen to be a strong function of earth resistivity. In
resistive earth, currents migrate quickly away from the center; in
conductive earth, this migration is much slower.
The secondary magnetic field caused by the eddy currents in the earth
have vertical and horizontal components that vary with time and along the
surface. Although in principle the resistivity layering can be derived
from both horizontal and vertical fields at any place inside or outside the
transmitter loop, it is common to use only the emf caused by the vertical
magnetic field in the center of the loop.
57
-------
Eddy current intensity
c
-------
o
c
03
«-•
CO
5
to
cc
02
Figure 3.15. Migration of the location of maximum eddy current intensity as a
function of time.
59
-------
Thus, in IDEM the time rate of change of the vertical magnetic field is
measured in the center of the loop. That field falls off rapidly with
.time, generally decreasing by four orders of magnitude over the time range
measured. One transient decay recorded over a few tens of milliseconds
contains information about resistivity layering over a significant depth
range. It is common to "stack" or to average several hundred transient
decays to improve signal-to-noise ratio. This stacking of transient decays
generally allows operation in urban environments in the presence of 60 Hz
noise. Stacking of several hundred transient decays requires a few
seconds, and multiple data sets are, therefore, quickly and easily
obtained.
Another transmitter-receiver array used in TDEM consists of a grounded
line transmitter and a receiver offset often several kilometers from the
grounded line.
There are several operational advantages to using measurements from the
center of the loop to derive resistivity layering. These advantages can be
understood from Figure 3.16 which shows the behavior of the vertical (z)
and horizontal (x) components in profiles in the center of the loop.
Figure 3.16 also shows that the profile of the z-component is relatively
flat about the center. Thus, errors in surveying the center are generally
negligible. Some distance away from the center, the z-component can change
rapidly with distance, and it will be necessary to survey the location
carefully. Figure 3.16 shows that the horizontal component is zero in the
center of the loop. Unwanted interference by horizontal components in
vertical field measurements caused by receiver misalignment is minimized.
Center loop measurements therefore allow quality control in data
acquisition which would be much more difficult to realize at other receiver
positions. In fact, TDEM measurements in the center of a loop can be
acquired w.ith a high degree of accuracy and repeatability with minimum
surveying requirements and other quality control procedures. Several data
sets can be recorded within a short time period to ascertain repeatability.
The processing of TDEM data proceeds in ways similar to those used in
other electrical prospecting methods. The emf measured as a function of
time is converted to an apparent resistivity. Figure 3.17 compares a
two-layer apparent resistivity curve from a Schlumberger D.C. sounding
with a similar curve from a TDEM sounding. In the Schlumberger sounding,
exploration depth is increased by increasing the L-spacing; at large
L-spacings, the apparent resistivity p approaches the true resistivity of
the lower layer,' /?2, at later times. In TDEM, exploration depth is mainly
a function of time, and pa approaches p1 at early time and approaches p2 at
later times. The spacing between transmitter and receiver remains
unchanged. In TDEM as in Schlumberger soundings the apparent resistivities
are entered into inversion programs which find a solution for the
resistivity layering that best fits the observed apparent resistivity
curve. These programs run on minicomputers, and inversions are generally
performed in the field for quick data turn-around and for quality control.
60
-------
1500
1000
500
o
-500
-1000
-1500
-2000
Bz at 2.2 msec after
transmitter turn-off
W 200
100
150
200
Bx at 2.2 msec after
transmitter turn-off
Center ot Loop
Figure 3.16. Behavior of EMF due to vertical magnetic field (B ) and horizontal field
(BJ on a profile through the center of the transmitter loop.
-------
10
£
i
E
o io2-
t;
-------
4.0 MagnetoTelluric (MT), Audio- MagnetoTelluric (AMI), and
Control! ed-Source Audio-MagnetoTelluric (CSAMT) Methods
Several techniques can be grouped under these methods. They go under
the namps of very low frequency (VLF)-radiohm, Controlled Source Audio
Magnetolellurics (CSAMT), MagnetoTellurics (MT), and Audio-MagnetoTellurics
(AMI) (Strangway et a?., 1973; Hoekstra 1978).
These systems have in common a general behavior of fields used in the
measurements. Common to all of these methods is a distant source for the
electromagnetic field. Figure 3.18 schematically illustrates the sources
of the various methods. In VLF-radiohm the sources are radio stations used
by the navies of the world operating in the VLF band (10 kHz to 30 kHz).
In CSAMT the sources are grounded line transmitters installed a
considerable distance from the measurement location. In MT the sources are
natural geomagnetic activity and thunderstorms. The operational frequency
bands of these methods are also shown in Figure 3.18.
The methods which have thus far seen most use in ground-water
exploration are the two methods using control led-sources, VLF and CSAMT.
An important reason is that use of controlled sources does not require the
extensive signal processing required in the use of natural sources (MT,
AMT) . The electromagnetic field emitted from the distant controlled source
has three field components at the surface of a horizontally-layered earth,
at a location far away from the source: the vertical and horizontal
components of the electric field, Ez and E , respectively, and a horizontal
magnetic field, H . Over a horizontafly-layered earth, there is no
vertical component Sf the magnetic field.
The wave propagates nearly vertically into the earth, and in the earth
the wave has two field components, EX and H (Figure 3.19). These
components are attenuated in the earth and are reflected from
discontinuities in electrical resistivity. At the surface, the quantity
(E/H ) is measured. This quantity is called surface impedance, Z, and is
related to the electrical resistivity of a homogeneous earth by:
H
y
where:
cj = angular frequency in radians/second;
li * magnetic susceptibility, in henrys/meter;
p " resistivity in ohm-meters;
/ « y - 1 .
For a heterogeneous earth, apparent resistivity is defined on the basis of
Ul by
63
-------
Method Source of Electromagnetic Waves Frequency Range
Magnetotelluric
CTl
Controlled-Source
Audio Magnetotelluric
Very Low Frequency
Thunder Storm
Ez
10~3Hz to 300 Hz
Ex
/ / / /^/ / / / /
Hy
Ez
1 Hz to 10 KHz
Radio
Transmitter
/ / /
Ez
/•/
.Ex 10 KHz to 30 KHz
Hy
Figure 3.18. Signal sources for magnetotelluric methods.
-------
Hy(0)
Uniform half-space
Hy(z)
Figure 3.19. Propagation of electric and magnetic fields in magnetotell
methods.
uric
65
-------
U/J
The simple definition of apparent resistivity and surface impedance
is valid for a horizontally-stratified earth. In the presence of lateral
resistivity changes, the surface impedance, Z, must be described by a
tensor, and the interpretation of field data becomes complex. Even though
the source of the waves is distant, -the evidence is that a measurement of
Pa derived from (Ex/H ) is a local determination of earth resistivity. The
components i and*HY are influenced by many factors such as path of
propagation between source (transmitter) and measurement station and
atmospheric effects. However, these non-local factors influence both EX
and H equally. The dominant effect is in E due to the local subsurface
electfical resistivities, while H is onfy slightly affected by the
properties of the subsurface in ythe absence of lateral resistivity
variations. The influence of non-local factors is eliminated in measuring
the ratio (E/Hy).
It is important as background for later discussions to realize that in
a horizontally-layered medium, the current flows in horizontal sheets.
There is no vertical component of current flow, and, therefore, no charges
are induced at the interface between horizontal layers of different
resistivities.
The wave propagating into the earth is attenuated. This attenuation
determines to a large extent the effective exploration depth. It is common
to correlate exploration depth with the skin depth of a plane
electromagnetic wave travelling into the earth. Effective exploration
depth is usually about one-half to two-thirds of the skin depth. Skin
depth is defined as the distance which causes an attenuation in amplitude
of 1/e (e = 2.7183). Skin depth is related to the resistivity of the earth
and frequency by:
d =
where
5 = skin depth in meters.
In Figure 3.20, 5 is given as a function of frequency. It is evident
that soundings to different depths of exploration can be obtained by
measuring at different frequencies. This commonly is done in CSAMT by
stepping the source and the receiver synchronously through a range of
discrete frequencies. Since the natural noise spectrum caused by
geomagnetic storms also contains energy over a large frequency range, a
sounding can also be made in MT. However, a major computational step is
involved in decomposing the spectrum measured as a function of tine to a
frequency-domain spectrum. The cost in acquiring and processing MT data
generally prohibits its use in ground-water investigations. Moreover, the
66
-------
E
_C
*-"
Q
c
CO
10
10
Frequency, Hz
Figure 3.20. Skin depth of electromagnetic plane waves as a function of frequency and earth resistivity.
-------
upper frequency range used in MT measurements is about 200 Hz to 300 Hz,
generally too low for effective use in ground-water investigations. CSAMT
data acquisition and processing do not" measure the surface impedance as a
tensor and therefore cannot deal effectively with lateral resistivity
variations. However, lateral resistivity variations can be estimated by
measuring at a large number of stations.
5.0 Induced Polarization Method
The principles of all the previous methods for measuring the electrical
properties of the subsurface are based on Ohm's Law, as given by
R = (V/I)
where
R = resistance (ohms)
V = potential difference (volts), and
I = electrical current (amperes)
This equation implies that the impression of a current, I, instantly
results in a potential difference, V.
There are, however, subsurface conditions where there is a significant
time lag between the impressed current and the resulting potential
difference. This phenomenon is called induced polarization (IP) and is
further illustrated in Figure 3.21. Figure 3.21 shows that the potential
difference V , arising from an impressed current I terminated at time t ,
drops instantaneously to a value Vs but subsequently decays gradually to
zero. The ratio (VS/V ) is called ttie chargeabil ity, M, and is an important
quantity measured in induced polarization.
The main objective of IP surveys has been the detection of disseminated
sulfide minerals. The magnitude of IP response generally increases with
the amount of mineralization. Also, extensive surveys have been conducted
to detect "halos" of disseminated sulfides that may or may not exist above
oil deposits.
Although the presence of disseminated sulfides is a dominant cause of
IP etfects, there are several other important causes (Olhoeft, 1985) such
as processes associated with ion exchange in clay minerals,
oxidation-reduction reactions, and clay-organic reactions (Ogilvy et a/.,
1972). Therefore, these causes must be taken into consideration in the
data interpretation.
IP effects are measured by experimental arrays similar to those used in
D.C. resistivity measurements. Many different arrays are in use (Figure
3.22). They all have in common the fact that effective exploration depth
is approximately one-fourth of the distance between current and voltage
electrodes. For investigating the spreading of contamination at the
injection horizon of injection wells (typically 2,000 feet or greater),
large arrays would be required. Such large arrays have been used in Figure
68
-------
0
Square wave current waveform
v
Voltage drop
Figure 3.21. Induced polarization effect: Transmitter current and voltage
waveforms in induced polarization.
69
-------
Pole-Dipole
>10a
i. 1 «
a , a or na,
Pole-Pole
>10a
h H
Schlumberger
Wenner
.
Figure 3.22. Commonly used array types for DC resistivity and IP.
70
-------
geothermal exploration but are considered impractical in investigations
over oil and gas fields. The size of the field and the presence of culture
(pipelines, power lines) limit the area and size-of the array which can be
used.
IP data can be acquired in time domain or in frequency domain.
Figure 3.23 shows transmitter current waveform for time-domain operation; a
synchronous receiver measures the resulting voltage waveform. Typically,
the receiver measures the ratio of voltages at one or more centered
intervals of time (Vv V2, VJ. The decay time in IP is on the order of a
second and, therefore, much larger than decay times used in TDEM soundings.
«
In freqency domain IP, square waves are used, and the transmitter and
receiver are sequentially stepped through a range of frequencies. In
summary, IP is in operation similar to resistivity measurements. The
difference is that in IP, in addition to resistivity, information is
obtained about the time lag between impressed current and the resulting
potential difference. The acquisition of field data is similar.
The IP method is expected to have very limited applications to
investigations of contamination from injection wells. The reasons for this
are as follows:
a) IP measurements, like D.C. resistivity measurements, require
large arrays (6,000 to 8,000 feet) to investigate contamination
at the depths of injection horizons.
b) It is unlikely that significant changes in IP are associated
with brine contamination. IP effects may be associated with
organic 1iquids.
c) Injection wells are typically located in areas with significant
cultural interferences (power lines, pipe lines, fences), and
determining IP effects with any accuracy in such areas will
generally be very difficult.
d) For investigating brine leakage, no additional information is
expected to be derived from measurements of IP compared to
those which are obtained with D.C. resistivity.
e) IP measurements require careful decoupling of EM effects caused
by pipe lines, fences, and power lines. In operations near
producing oil fields, this may also be a serious limitation.
71
-------
IP Field Measurements
ro
Transmit Lt
+v
Receive Lt
Time Domain
Single Pulse
Multi Pulse
Phase Angle
Sine Wave
Frequency Domain
Square
Wave
Dual
Frequency
Figure 3.23. Transmitter current and receiver voltage waveforms in IP field measurements.
-------
IV. APPLICATIONS AND CASE HISTORIES
In Chapter 3 the physical principles of several geophysfcal methods
were discussed. All of these methods have, at one time or another, been
applied to investigations of ground-water contamination. However, in the
majority of these previous investigations, the contamination was detected
at shallow depths. There are few case histories dealing with leaknqe from
injection horizons or from wells at depths of several thousand feet where
injection frequently occurs. The quality of information obtained and the
settings in which each technique may be applied differ with different
methods. The various electrical exploration tools are far from equivalent.
«
The success of geophysical surveys is critically dependent on the
clear definition of exploration objectives. Geophysics is. no panacea; it
can accomplish certain tasks but not others. In the best of circumstances,
it can map the target of interest to a certain degree of accuracy. It
might well be that the required accuracy is greater than the accuracy
obtainable with surface geophysics. In such cases the use of geophysics
must be rejected.
There may often be more than one exploration objective; for example,
the mapping of contamination in two different aquifers at different depths
may be desired. Experience has shown that two exploration objectives can
seldom be achieved with one set of survey specifications. Better results
are often obtained by surveying for one objective at a time.
The objectives of surveys using electrical geophysical methods can
best be stated in terms of geoelectric sections. A geoelectric section
displays the variation in electrical resistivity laterally and vertically
across a vertical plane passed through some line on the surface. Examples
of geoelectric sections are shown in Figures 4. la and 4.1b where the
variation in resistivity is caused by lithology and brine leakage from a
disposal well. The detectability of the contamination will vary with
different sections and methods. Detectabil ity can often be evaluated from
modeling investigations.
To evaluate the applicability of various electrical methods to certain
exploration objectives, evaluation criteria are required. The important
criteria are as follows:
o lateral resolution
o vertical resolution
o sensitivity to geologic noise
o survey productivity (cost, ease of operation, personnel
qualifications)
73
-------
= 2 ohm-m ••
Figure 4.1. Examples of hypothetical geoelectric
associated with injection wells.
74
sections which could be
-------
The first three criteria can be evaluated from physical principles
discussed in Chapter 3. The fourth criterion is dependent not only on the
particular geophysical method in question but also on site-specific
factors: terrain, vegetation, and cultural interferences such as power
lines, fences, pipelines, and nearby radio-frequency transmitting stations.
1.0 Criteria for Evaluation of Electrical Methods
1.1 Lateral Resolution
Lateral resolution is defined as the accuracy to which lateral changes
in electrical resistivity can be mapped and interpreted. For example,-in
the geoelectric section of Figure 4.la and 4.1b, the distance which the
brine contamination has progressed horizontally is important. Moreover,
the concentration of brine in the ground water can be expected to decrease
gradually with increasing distance from the well. An ability to assess
this gradual decrease would be beneficial; this ability is related to the
lateral resolution of resistivity mapping.
The lateral resolution of different electrical methods is mainly
determined by such factors as:
o the separation between source (transmitter) and receiver
o the components of the electromagnetic field which are measured
(There is a difference between the lateral resolution of methods
which measure electric fields or a ratio of electric and magnetic
fields, and the resolution of methods which measure only magnetic
field components.)
o desired depth of exploration
While it is often possible to observe lateral resistivity changes in
the data, it is more difficult to determine the geoelectric section where
rapid lateral changes exist. This is again related to a strong tendency to
interpret data by one-dimensional models. Cost-effective 2- and
3-dimensional models are still not common in electrical prospecting.
1.2 Vertical Resolution
Vertical resolution is defined as the accuracy with which vertical
changes in resistivity with depth can be resolved. Vertical resolution is
mainly determined by two factors:
o the relationship between the electrical resistivity of the
subsurface and the measured voltages or electromotive forces. In
Table 4.1, this relationship is summarized for several
electromagnetic methods. The TDEM method with small
transmitter-receiver separation has the highest sensitivity to
earth resistivities. This high sensitivity is generally .ealized
in soundings in the center of the transmitter loop.
75
-------
TABLE 4.1. Relationship Between Resistivity and Measured EMF (or voltage) for
Various Electrical Survey Methods
Electrical Method
Direct Current
MT; CSAMT; VLF
Frequency Domain EM
Relationship
p = KV
p = K(emf)1/2
P = K(emf)
(low induction numbers)
Time Domain EM
(late stage, small transmitter-
receiver separation)
Time Domain EM
(early stage, small transmitter-
receiver separation)
p = K(emf)-3/2
p = K(emf)
The constants denoted by K are different values for each method,
76
-------
o the accuracy to which the field can be measured. Accuracy is
partly determined by minimum detectable signal and by geologic
noise.
1.3 Sensitivity to Geologic Noise
Geologic noise is defined as changes, in signal which are caused by
geologic conditions and which obscure the exploration objectives or
targets. Very often, geologic noise is the limiting factor in electrical
and electromagnetic surveys. The schematic geological and geoelectric
section of Figure 4.2 illustrates the nature of geologic noise. This
section is a simplification of the geologic setting of the Floridian
aquifer system. The limestone aquifer is overlain by Pleistocene deposits
of sands, clays, and organic material of various thicknesses,.
The Floridian aquifer is believed to have been infiltrated at one time
by saline water during high sea levels. Since the lowering of the sea
level, saltwater is gradually being replaced by freshwater. However, a
freshwater-saltwater interface occurs at some depth in the Floridian
aquifer, and the thickness between the top of the limestone and the
freshwater-saltwater interface represents the quantity of potable water. An
objective of electrical geophysical surveys could be the mapping of that
freshwater-saltwater interface. Geologic noise in this geologic setting
would be introduced by variations in soil types and thickness of sediments
above the freshwater-saltwater interface.
Another objective of geophysical surveys could be the delineation of
recharge areas. Recharge into the Floridian aquifer is likely to occur
where the Pleistocene sediments are dominantly coarse-grained or absent.
When the geophysical objective is to map the soil types and thickness of
the overburden, geologic noise could be introduced by variation in the
freshwater-saltwater interface, particularly when it would occur relatively
near the surface.
Geologic noise is, thus, a function of the exploration objective. By
clearly defining the exploration objective and the sources of geologic
noise, optimum geophysical tools can be selected to reduce that noise to
the maximum possible extent.
2.0 Electromagnetic Induction Methods
Electromagnetic induction methods as discussed here include both
continuous or frequency-domain EM systems and transient or time-domain
77
-------
oo
Surface Depression
Fic:re 4.2. Geologic noise obscuring the mapping of a freshwater-saltwater interface,
Geologic noise is caused by heterogeneity of the layer thickness.
-------
(IDEM) systems. Applications of frequency-domain systems to injection
well problems have been limited to cases where leaks in the upper reaches
of the well systems create contamination targets at shallow depths or where
surface brine pits have left near-surface saline targets. IDEM has also
been useful in defining large areas of contamination at considerable
depths.
2.1 Frequency-Domain Methods
2.1.1 . Instrumentation
The most common configuration of frequency-domain systems consists of
two air coils separated by discrete distances; one coil functions as a
transmitter and the other functions as a receiver. A cable typically
connects the transmitter and receiver and provides synchronization. The
first series of these instruments were used in the mining industry
(Slingram, Max-Min). In these instruments both in-phase and quadrature
phase components are measured. It was discussed in Chapter III that the
in-phase component which is caused by earth eddy currents must be measured
in the presence of a primary magnetic field caused by currents in the
transmitter. This is accomplished by electronically subtracting the
contribution of the primary field from the measured in-phase emf. Since
this primary field (at a constant frequency) is only a function of
distance, this can be done for each discrete distance. This requires that
these distances must be accurately chained in field work. Since the primary
magnetic field is proportional to (1/r3) where r is the
transmitter-receiver separation, this chaining must be accurate to 0.3
percent to obtain an accuracy of 1 percent in the measured in-phase
component. However, in-phase components with amplitudes of r percent of
the primary field or greater are generally only observed over mineral ore
at the frequencies employed. Measurement of earth conductivities using
in-phase currents would require 0.001 percent accuracy in measuring the
in-phase current to achieve 1 percent accuracy in measured apparent
conductivities. For this reason, the quadrature-phase component is
generally used in ground-water investigations where the contrast between
target and background is usually less pronounced than in minerals
exploration. The in-phase component or amplitude can be used when spacing,
frequency, and conductivity are such that the induction numbers are not
smal1.
Another problem in applying the systems used in mining to ground-water
investigations is the difficulty of interpretation. In general, the
relationship between measured emf and subsurface resistivity layering is
complex and inconvenient for routine interpretations. Interpretation
procedures and the relationship between measured emf and earth conductivity
can be considerably simplified by operating over certain ranges of
transmitter-receiver separation and frequency called low induction numbers.
The EM-31 and EM-34 are designed to operate at low induction numbers.
The exploration depth of frequency-domain systems is mainly a 'unction
of transmitter-receiver separation. A practical limit for maximum
exploration depth for the EM34 is about 30 meters, although highly
79
-------
conductive anomalies can be detected at greater depth. Increasing
exploration depth requires increasing transmitter-receiver separation and
transmitter output power.
The limited exploration depth (about 30 meters) of these
frequency-domain systems restricts their use in injection well leakage
investigations where targets will usually be more deeply buried. Thr method
can perhaps be used to detect leakage at shallow depths but would be
incapable of detecting targets at the depth of typical injection horizons,
several thousand feet. A -more likely class of targets for frequency-domain
systems would be leakage from temporary holding ponds.
Commercially available instrumentation for frequency-domain methods is
listed in Table 4.2. In ground-water contamination investigations, the
main instruments used are the Geonics EM-31 and the EM-34-3. The
configurations of the EM-31 and the EM 34-3 are sketched in Figures 4.3 and
4.4. An important reason for their popularity is their operation at low
induction numbers (Chapter III, section 3.2), resulting in a linear
relationship between apparent conductivity and the emf caused by the
quadrature field, which in turn allows the use of simple interpretation
procedures. In the EM-31, a rigid fiberglass boom connects the transmitter
and receiver, and an electronic bucking voltage cancels the primary field.
In the EM 34-3, the transmitter and receiver are connected by a cable. The
bucking voltage, equal to the primary field, is used to determine exact
transmitter-receiver separations.
The other instruments listed in Table 4.2 are mainly used in mineral
exploration to detect conductive ore targets. These instruments generally
do not operate at low induction numbers and require complicated functions
to relate instrument response to earth conductivity.
2.1.2 Operational Advantages and Disadvantages
Frequency domain systems have their application in detecting leakage
from injection wells which is manifested in the upper 30 meters beneath the
surface. For such shallow targets, frequency-domain EM and D.C. resistivity
are the principal electrical methods. However, there is little or no
application for these methods for targets at the depth of injection of
waste fluids (several thousand feet).
80
-------
Table 4.2. ComercU 1 ly Available Frequency-Domain EH Systems
CD
Systen
Geon i c s
CH-31
(iconics
(M-34-3
Max-Hsn
Syntex
Genie
Waveform frequency of Transmitter-
Frequency of Receiver
Separat ion
sinusoid 9800 Hz 3.66
s inusoid 6400 Hi 10 m
1600 Hi 20 m
400 Hi 40 m
unusoid 222 - 3555 Hz 8 m - 250 m
sinusoid 37 - 3037 Hi 6m- 300 m
(effective to
Component
Measured
quadrature
quadrature
quadrature
quadrature
in-phase
quadrature
in-phase,
quadrature,
Approximate
Effective Depth
of Explorat ion
(meters)
7
8
16
32
180
120
Ccxments
In-phase component electron-
ically cancelled; main area
of application in shallow
conductivity mjpping.
Bucking voltage, equal to the
primary field, is used to
determine exact separation.
Distances must be slope-
cha ined.
Distances are electronically
determined; no interconnecting
cable necessary between trans-
mitter and receiver.
-------
DO
rv>
Tx
ZZD
Receiver
Transmitter
Figure 4.3. Geonics EM-31 conductivity meter vertical dipole mode.
-------
CD
CO
Tx
Transmitter
Rx
Receiver
Figure 4.4. Geonics EM-34-3 conductivity meter horizontal dipole mode.
-------
Effective depth of exploration for EM is also about equal to the
intercoil spacing (See Table 4.3). There are substantial differences in
the lateral resolutions of the different arrays, however.
For detection and mapping of shallow ground-water contamination, there
are some inherent operational advantages in the use of the Geonics EM-31
and EM-34:
o no galvanic contact with the ground is required. Measurements can
be made over highly resistive upper layers, such as frozen ground,
dry sand surfaces, and asphalt.
o the instrumentation is lightweight and portable so that high survey
productivities can be achieved.
o the distance between transmitter and receiver need not be chained.
The in-phase component is used to derive the correct distance.
Disadvantages are:
o the eddy current intensity in the earth and the resulting emf
decreases with increasing earth resistivity. It is difficult to
measure earth conductivities less than 0.001 mho/m (resi.stivities>
1000 ohm-m). In practice this is not a major drawback in
ground-water studies.
o with commercial instrumentation the method is mainly applicable to
profiling: measuring lateral or areal variation at certain
effective exploration depths. Surveying at several effective
exploration depths yields some information about resistivity
variation as a function of depth, but the few data points in the
vertical direction make the interpretation of resistivity layering
difficult.
2.1.3 Lateral Resolution
The lateral resolution of the EM methods compares favorably with the
direct current method. To achieve the same detectability with D.C.
resistivity, by using the Wenner array, as with EM with a two-layered earth
would require an a-spacing of about 2h1 or a total electrode spread of 4hr
Although there is not a one-to-one correlation between the array size and
the lateral resolution of the array, D.C. resistivity methods tend to have
a somewhat lower lateral resolution than EM methods. The dipole-dipole
array provides slightly better lateral resolution than the Wenner array.
84
-------
TABLE 4.3 APPROXIMATE EFFECTIVE DEPTH OF INVESTIGATIONS FOR
GEONICS EM-31 AND EM-34
Instrument
EM 31
EM 34-3
Frequency
of Operation
9.8 khz
9.8 khz
6.4 khz
1.6 khz
0.4 khz
Transmitter -Receiver
Separation
3.66m (11.7 ft.)
3.66m (11.7 ft.)
10m (32 ft.)
20m (64 ft.)
40m (128 ft.)
Dipole
Orientation
Verticle
Horizontal
Horizontal
Horizontal
Horizontal
Effective
Exploration
Depth, Ft.
19
9.5
24
48
96
-------
2.1.4 Geologic Noise
Frequency-domain EM instruments are chiefly used for profiling. They
are used to measure lateral variations in earth conductivity within a
certain depth interval (fixed intercoil spacing). The methods are employed
to cover a large area quickly, and by far the greatest limitation to data
quality is sensitivity to geologic noise.
The degree of geologic noise is highly dependent on the geologic
setting and is one of the survey parameters most difficult to predict.
Variation in subsurface composition is often greatest in the upper 30
meters. This is the zone where rapid changes are most likely to occur in
soil types, water content, and organic content.
Sources of geologic noise can often be distinguished from the
exploration target if:
o the amplitude of the noise is less than the amplitude of the
response from the exploration target; or
o the spatial spectrum of the target includes different frequencies
than the spectrum of the noise
The cause for the relatively high sensitivity of EM to geologic noise
can perhaps be best understood from Figure 4.5. Figure 4.5 shows eddy
current intensity as a function of depth for horizontal, co-planar dipoles.
The intensity of eddy currents is high near the surface and decreases with
depth. For that reason variation in thicknesses and conductivities of
near-surface layers will have a proportionally greater influence on
instrument response than layers at greater depth.
2.1.5 Case Histories: Fixed-Frequency Electromagnetic Induction
The use of the Geonics EM-31 and EM-34 for injection well
investigations is restricted by the limited exploration depth (<30 meters).
They may be employed for detecting leakage from injection wells where the
leaks are manifested near the surface or for investigating contamination
emanating from temporary surface holding ponds. Two case histories have
been selected to illustrate the application of these devices to shallow
targets. Neither case history is from an injection well, but both
illustrate the use of the EM-31 and EM-34 for detection and mapping of
shallow, conductive contamination.
86
-------
Eddy current intensity
CD
O
co
-o
13
O
O)
03
_Q
Q.
0)
Q
Figure 4.5. Eddy current intensity as a function of depth for
electromagnetic induction (horizontal, co-planar dipoles).
87
-------
2.1.5a Henderson, Nevada.
Reference:
"Study of Subsurface Contamination with Geophysical Monitoring Methods at
Henders;n, Nevada," E.G. Walther, D. LaBrecque, and D.D. Weber, Lockheed
Engineering and Management Services Company, Inc., and R.B. Evans and J.J.
van Ee, EMSL-LV, EPA, Proceedings of the National Conference on Management
of Uncontrolled Hazardous Waste Sites, 1983, Washington, D.C.
Figure 4.6 shows the layout of the industrial complex and a survey
line along which measurements were made with an EM-31 which was oriented in
the vertical dipole position. The objective of the survey was to detect
contamination leakage plumes. The field observations are shown in a profile
in Figure 4.7. The plume is identified by an increase in conductivity above
100 mill imhos/meter. When conductivities exceed about 70 mmho/m, a
correction for nqn-1inearity must be applied to observations with the
EM-31. Figure 4.7 shows measured and corrected apparent conductivities. It
is evident that the contamination along the profile is above background.
Noise in the data is caused by cultural interferences (fences, highways),
but the signal caused by the plume is several times greater than the
interference signal.
2.1.5b Falmouth, Massachusetts
Reference:
Olhoeft, Gary R., and D.R. Capron, "The Use of Geophysics in Hazardous
Waste Investigations," Open File Report, U.S. Geological Survey, Denver,
CO, 1986.
The second case history involves an EM-34 survey near Falmouth,
Massachusetts, conducted to detect ground-water contamination from a sewage
lagoon. The near-surface geology of the site consists of clean sands and
gravels (glacial outwash) several hundred feet thick. The water table
occurs at a depth of about 20 feet. The topography is rolling sand dunes
with elevation differences of about 50 feet.
Since the objective of the investigation was to map ground-water
contamination, the apparent conductivity of the ground water was to be
determined. The relatively dry sands above the water table have very low
electrical conductivity.
Figure 4.8 shows the influence of topography along a profile with
topographic relief. The relatively small changes in conductivity of the
ground water are masked by topographic relief. Topographic relief in this
case is the source of geologic noise; unless the survey data are corrected
88
-------
'-D
O
West
350
250
-
•*—*
>
«—•
u
J 150
c
0)
k.
(O
n.
OL
50
0
50
100
East
A ^ A.Asy^>.-r-v.
•vx/
150 200 250
Station Number (feet)
300
350
Original Data
Corrected Data (for non-linear response of instrument at high conductivity values)
Figure 4.7. Apparent conductivity values from Geonics EM-31 along survey line in Henderson, Nevada.
-------
Topographical relief removed
Topographical relief
Spacing (m)
CTa = 0
1 G,
+
CT2
G
2
cra = o
"2 G
2
G
r
(elevation)
B
7 m
Topography
1 I
CT, = 0
7 m
T
0*2 = 30 mmhos/m
Contamination plume.
;
-------
for topography, the contaminant plume cannot be delineated from the profile
data.
Also shown in this figure is the EM profile corrected for topographic
relief. The changes in ground-water contamination are now measurable. A
map shov/ing the survey line and the outline of the contaminant plumn is
shown in Figure 4.9.
The two case histories illustrate the applications and limitations of
the. fixed-frequency EM methods for ground-water and brine leakage
investigations. These limitations are as follows:
o limited exploration depth (<75 feet); and
o relatively high sensitivity to geologic noise.
In some situations the data may be corrected for geologic noise; the
Falmouth site was such a situation. In other cases, the signals caused by
contamination may greatly exceed noise, as was the case with the Henderson
site. In general, it appears that more manipulation of the data (noise
reduction) would improve the use of the method. The major advantages of
the method are its low cost, rapid data collection, and ease of operation.
The principal disadvantage of fixed-frequency EM for investigating
injection wells is the limited exploration depth.
2.2 Time Domain (TDEM) Soundings
2.2.1 Instrumentation
The instrumentation available for TDEM soundings is listed in Table
4.4. TDEM 'equipment is primarily used for two purposes:
o to detect highly conductive zones associated with mineral ores, and
o to perform soundings (that is, to determine subsurface resistivity
layering).
In Table 4.4 a judgment has been made about the performance of the
equipment for soundings. There is little information available in the
literature to back up that judgment.
2.2.2 Operational Advantages and Disadvantages
The advantages of TDEM soundings are as follows:
o no galvanic contacts (probes) are required. This is particularly
attractive for the relatively deep exploration targets associated
with injection wel1s.
92
-------
-3000
1000 -500 0 500 1000
Meters, East
Figure 4.9. Sewage plume as outlined by survey with Geonics EM-34 near
Falmouth, Massachusetts.
93
-------
TABLE 4.4 CHARACY .STICS OF TRANSIENT SOUNDING SYSTEMS
03
System
Geonics EM-42
Geonics EM 37-3
Mineral Control
Instrumentation
Ply., Ltd.,
SIROIEM
Zonge GDP- 12
Lamontagne
UTEM-II
Current
Wave-
form
bipolar
50% duty
cycle
bipolar
50% duty
bipolar
50% duty
cycle
bipolar
50% duty
cycle
triangular
Maximum
Current
40A
40A
10A
or
20A
27A
or
43A
6A
Turn-off Measurement
Time for Time
450m x 450m Loop Range
200/iS 890/iS - 3.3
400ps 89/is - 79ms
100/;s 50/is - 161ms
ns* 120fis - 900ms
ns 25/u - 25ms
Comments
separate receiver and
transmitter; computer-
controlled digital
acquisition
separate receiver and
transmitter; data logger
available
single unit for single
loop operation; built-in
logger; quite portable
separate receiver and
transmitter; built-in
data logger
separate receiver and
transmitter; built-in
data logger
ns' (not specified)
-------
o measurements require limited space and can be conducted in an
environment with high ambient electrical noise (powerlines and
radio stations)
o procedures for data acquisition are highly automated
The disadvantages of IDEM soundings are:
o the method generally cannot be applied to exploration in the upper
50 feet
o equipment is expensive, and equipment operation and interpretations
are complex and require highly-trained technical personnel
2.2.3 Lateral Resolution
IDEM soundings map subsurface resistivity layering with good lateral
resolution. The physical reasons for high lateral resolution are as
follows:
o exploration depth is mainly a function of time rather than spacing
between source and receiver
o a local source for the electromagnetic field is used
Although hard data are not available, lateral resolution is
approximately equal to exploration depth. Thus, for a target at an
exploration depth of 1,000 feet, a measurement is probably representative
of an area within a 500-foot radius of the measurement site.
2.2.4 Sensitivity to Geologic Noise
TDEM soundings have a lower sensitivity to geologic noise than any
other method. This fact can be conceptually explained from Figure 4.10
where the eddy current intensity as a function of depth is schematically
illustrated at different times after turn-off. With increasing time the
maximum current intensity occurs at increasingly greater depth; for
example, at time t2 the current intensity in the upper layers is small. A
measurement at time t, will, therefore, be relatively insensitive to
near-surface layers. The geologic noise from the upper layers will be
small in measurements at later times. High data quality can be realized
with TDEM because of the low influence of geologic noise; this is one of
the major advantages of TDEM. Some of the case histories are used to
illustrate this point.
95
-------
Eddy current intensity
"to"
T current
r~
\ .
Time to t<\ 12
Figure 4.10. IDEM eddy current Intensity as a function of depth
96
-------
2.2.5 Case Histories: Time Domain Methods
IDEM has been used for exploration targets at depths varying from 15
meters to about 3,000 meters. This is clearly the depth of interest in
investigations of problems related to leakage from injection wells or to
migration at the injection horizon. Two published case histories are
reviewed, one with a mapping target at about 500 meters depth and another
where contamination from injected fluids was evident at shallow depths.
2.2.5a TDEM Survey in Haskell County, Oklahoma
Reference:
Wightman, W.E., A.A. Kaufman, and P. Hoekstra, "Mapping Gas-Water Contacts
in Shallow Producing Formations with Transient EM," presented at 53rd
Annual Meeting, Society of Exploration Geophysics, Las Vegas, Nevada
(1983).
In Class II injection wells, the produced brines are often reinjected
into the pay zone. In this case history, the objective was to map the
extent of the producing zone and the contact between hydrocarbon saturation
and brine saturation in this zone; this is an objective closely related to
well siting and completion.
Figure 4.11 shows a geologic section of the study area with four
induction logs. These logs show that where the producing horizon is
saturated with brine, formation resistivities of 2 to 3 ohm-meters are
encountered; where the producing horizon is saturated with hydrocarbons,
formation resistivities in excess of 50 ohm-meters are found. The
formation resistivities above and below the producing horizons vary between
15 ohm-meters and 30 ohm-meters.
Thus, in traversing the contact between hydrocarbon and brine
saturation, a large change in resistivity occurs in a relatively thin layer
(20 meters) at a depth of about 500 meters. It is common in all electrical
methods to express the electrical property of a thin layer by its
conductance, the ratio of its thickness to its resistivity. Thus, the pay
zone sa'urated might have a conductance of about 10 mhos when saturated
with brine and a conductance of less than 1 mho when saturated with
hydrocarbons.
The first step in the investigation was to make an evaluation with
computer modeling of the feasibility of detecting the contact between the
brine and the hydrocarbons by TDEM from the surface. For that purpose the
geoelectric section of Figure 4.11 is simplified to the model shown in
Figure 4.12, consisting of a thin layer embedded in a half-space with a
97
-------
w
feet
1000-4-305
SERVICE
#1 Blankenship
SUN
#1 Whisenhunt
SERVICE
#1 Bryan
Sea
Level
CD
-1000 305
1000 meters
Figure 4.11. Geologic cross-section and borehole induction logs, Haskell County, Oklahoma
-------
ro
Q_
0.80
0.25
100
0.25
0.01
1000
AB/2 (meters)
Frequency (hz)
sec
1/2
10,000
Q.
CO
Q_
1.0-
0.63-
0.47-
0.25-
10, C
MT
S2 - 60-| 30-7
1 /
^--^ Ji*—^'**"
\t^ X"
1 , 1
300 100 1.0 0.
01
S2 - 60^ 30-i 15-7 0
Figure 4.12*. Computer modeling studies of detectabil ity of thin pay zone with
IDEM, Mt, and DC Resistivity.
99
-------
uniform resistivity of 20 ohm-meters. The conductance of the thin layer
is varied in the model. In Figure 4.12 the results of computer modeling
studies for three electrical methods are shown (IDEM, MT, and D.C.). The
model curves show that there is a range of time, frequency, and spacing
where tie change in conductance in the producing zone causes a decrease in
apparent resistivities. Two important conclusions can be derived from
these models:
0 The change in apparent resistivity for Che IDEM and MT methods is
considerably larger than for the D.C. method. Table 4.5 lists the
maximum decrease in apparent resistivity caused by changes in the
conductance.
o Impractically large electrode separations are required in D.C.
soundings to measure the decrease in apparent resistivity. For
IDEM soundings, measurements are made within a 500-meter square
loop.
The layout of the IDEM survey and the locations of drilled wells are
shown in Figure 4.13. Square loops 500 meters on a side were laid across
the gas field, and measurements were made in the loop centers. Typical
apparent resistivity curves over locations with gas saturation, brine
saturation, and over dry holes (low porosity) are superimposed on Figure
4.13. The characteristic decrease caused by brine is clearly visible in
the curves of Loops No. 1, 4, and 10 in Figure 4.13. From the behavior of
the curves of the various soundings, a contour map of producing horizons
was made, outlining areas where either gas or brine saturation occurred or
where porosity of the normally high porosity sandstone layer was reduced.
This map was consistent with logs from four available wells and agreed with
logs from wells which were subsequently drilled.
Similar case histories are available from the Russian literature. An
example from the Siberian Basin is shown in Figure 4.14. As the oil-water
contact is traversed, the apparent resistivity curve changes shape
(Rabinovich, 1973).
The implications of these case histories for investigations of
injection wells appear to be the following:
o TDEM soundings can determine the depth and water quality of high
porosity layers in sedimentary basins.
o TDEM soundings have been used to determine the areal extent of
these high porosity, brine-saturated layers.
100
-------
TABLE 4.5 MAXIMUM DECREASE IN APPARENT RESISTIVITY AND MEASURED
VOLTAGE CAUSED BY ACTUAL INCREASES IN CONDUCTANCE FOR
TDEM, MT, AND DC
Conductance
15
30
60
Maximum Decrease in For Various Methods
TDEM
20%
45%
55%
MT
19%
37%
53%
DC
13%
20%
36%
101
-------
Transmitter Loop
\ Sounding Curve
t
8
11
~V~ Booch Producer, gas only
Booch Producer, gas/water
Non Producer, water only
.A_
L/ /. ] Area of low porosity
f ~) Area of brine/gas saturation
r~. ,1 Area of gas saturation
-N-
V
\
scale
1 km
Figure 4.13. Layout of IDEM survey and locations of drilled wells, Haskell County, Oklahoma, showing typical TDEM
apparent resistivity curves over locations with gas saturation, brine saturation, and dry holes.
-------
80 93 75
81 74 91 90 73
72
54
1000 -
JL -m
100
OM.M
Square root of time
o
u>
80 93 75
81 74 91 90 73
Brine
72
54
tion number _
\A/dl number
2120-j
2160 -
2200 -
11 8 -5
GasyL^lIL=^
4 1
" ~ -^
Figure 4.14.
IDEM apparent resistivity curves along survey line in Siberian Basin,
showing oil-water contact. Distance from station 80 to 54 is 16 km.
-------
2.2.5b IDEM Survey in Osage County, Oklahoma
Reference:
Fitterman, D.V., P.V. Raab, and F.C. Frischknecht. "Detection of Brine
Contamination from Injection Wells Using Transient Electromagnetic
Soundings/ EPA, pre-issue copy, January, 1986.
IDEM test surveys were conducted to determine brine leakage from
injection wells in Osage County, Oklahoma. In the test area vegetation
damage was a visual surface indicator of brine contamination.
The producing horizons are sandstone layers in the upper Pennsylvania
sediments, and brine is reinjected into these producing horizons. The
producing horizons occur at various depths ranging from 200 feet to 1,000
feet. In some parts of Oklahoma, such as Texas County, large fresh water
aquifers such as the Ogallala Formation occur above injection horizons, and
the possibility for contamination exists. The Ogallala Formation is,
however, absent in Osage County.
For the TDEM survey, coincident loop soundings were used with 76-meter
or 152-meter loops. Figure 4.15 is a plan view of the loops and of nearby
test wells.
A typical geoelectric section derived from the TDEM soundings is shown
in Figure 4.16. In the vicinity of Well #1W, a substantial decrease in
resistivity (about a factor of 3) is observed. The decrease in resistivity
was interpreted to occur near the surface. The bottom contact of the low
resistivity zone was not derived.
This case history showed brine contamination from the injection well
to be localized. This is an example where the validity of one-dimensional
interpretation may have to be determined. TDEM surveys were capable of
detecting the presence of brines at depths of several hundred feet and were
also capable of mapping, at least near the surface, the areal extent of the
contamination.
3.0 CSAMT Method
3.1 Instrumentation
3.1.1 Operational Advantages and Disadvantages
Advantages are as follows:
o Many receiving stations can be serviced by the same CSAMT source.
This reduces survey line setup time.
104
-------
96»40'
O
tn
1 Well #5
T.D.
' 910- 277m
Section 9
I
Section 10 »
0
b
±
1000
=1
meters
4000
1
feet
i
Oklahoma
—i-"\
Study \
Area I
.__ ^J
96°37'30'
36°40' —I—
Figure 4.15. Plan view of IDEM loops and nearby test wells, Osage County, Oklahoma.
-------
B WEST
B' EAST
o
cr.
300
c 200
o
«-•
fl3
>
—
UJ
100
14500
T
39
4.4
1.8
#1W
13500
9500
T
15 500 11 500
6250 10250
T T T—
11
52
30 9.3
/.••v.'l.2 •':•'.••• 4.5'.••:'••.• 1.0'.A
/•:•'•.'•:••--. ••.'•:••. .• - .•:•'•."-.••. .--.'• ^
Resistivities in ohm-m
2.0
-400
-200
Distance (m)
0
Sounding Number 13250 Loopsize (250 ft by 250 ft)
Loop Center
Interpreted extent of brine contamination
200
Figure 4.16. Typical geoelectric cross-section derived from IDEM soundings, Osage County, Oklahoma.
-------
o Data collection is rapid.
Disadvantages of the method are as follows:
o The method is highly sensitive to geologic noise, and lateral
variation in resistivity can have a major influence on soundings.
o In CSAMT, the source must be located several kilometers from the
receiver (measurement site). This may create problems in obtaining
access to needed sites.
o Common data acquisition procedures in CSAMT involve obtaining data
with only one orientation of electrical dipoles. As a result, the
full tensor of the surface impedance is not measured and surface
impedance is treated as a scalar quantity.
3.1.2 Geologic Noise
The method is highly sensitive to geologic noise. An important reason
for this high sensitivity is that a ratio of an electric field to a
magnetic field is measured. The electric field is greatly affected by
near-surface resistivity variations. At boundaries between regions of
different electrical resistivity, electrical charges form. Because of the
usual variability of surface layer composition, the surface layer is likely
to have many such boundaries. These local charge formations distort the
electrical field and cause high sensitivity of the method to geologic
noise.
3.1.3 Lateral Resolution
Lateral resolution for mapping of features exposed near the surface
can be very high. Interpretation of the depth of targets may be very
difficult, however. The physical reasons are illustrated in Figure 4.17;
electrical charges form at the boundaries of lateral discontinuities in
resistivity and in turn greatly alter the local electric fields.
Consequently, apparent lateral resolution may be very high if a receiving
station is located near the discontinuity, as would be the case for Station
A in Finure 4.17. On the other hand, it will also cause distortion in a
sounding at B. In general the effect of electric charges will extend both
vertically and laterally. Although an anomaly may be measured at A, it may
not be possible to determine reliably the depth of the cause of the
anomaly.
Electrical charges form when electrical currents cross boundaries of
different resistivities. The formation of charges is, therefore, also
highly dependent on the direction of the electrical field, as illustrated
in Figure 4.17. This phenomenon often results in different apparent
resistivity curves based on the direction of measurement of electric fields
(or the orientation of survey lines).
107
-------
Station A
Station B
Es
Figure 4.17. CSAMT geologic noise created by charge formation along
di scontinui t ies.
108
-------
3.2 Case History: CSAMT Survey of Brine Contamination in Lincoln County,
Oklahoma
Reference:
Syed, T., K.I. Zonge, S. Figgins, and A.R. Anzzolin, "Application of the
Controlled Source Audio Magnetotellurics (CSAMT) Survey to Delineate Zones
of Ground-Water Contamination: A Case History," presented at the Second
National Conference and Exposition on Surface and Borehole Geophysical
Methods in Ground-Water Investigations, National Water Well Association,
Fort Worth, TX, February 12-14, 1985.
A CSAMT survey was conducted in Lincoln County in east-central
Oklahoma. Oil production in the study area began in the 1930's. Water
injection for secondary and saltwater disposal purposes began in the
1950's, and the cumulative volume of water injected through August 1982 was
approximately 75 million barrels.
The major objective of the ground-water contamination study included
the identification of high chloride concentrations in upper aquifers and an
investigation of the possibility that the source of these high chloride
concentrations could be oil field brines. CSAMT was part of an overall
exploration program consisting of drilling, ground-water sampling. Although
the hydrocarbon formations occur at depths of about 3,000 feet, the
freshwater/saltwater interface occurs at shallow depths from 150 feet to
450 feet. This was the interface mapped by CSAMT; the survey did not
determine whether the near-surface brines originated at deeper horizons.
A -schematic survey layout is given in Figure 4.18. A transmitter
dipole length of 5,000 feet was used and was positioned approximately 3 to
4 miles from the survey area. In the survey area, measurements were made at
intervals of 200 feet along the survey lines. The direction of the electric
field measurement was parallel to the survey lines. At each measurement
station, scalar apparent resistivities were determined at discrete
frequencies over a range from about 5,000 Hz to 22 Hz. The data were
presented in vertical and horizontal pseudo-sections where the apparent
resistivities measured along profiles are displayed and contoured. An
example of a vertical pseudo-section is given in Figure 4.19. Typical
background resistivities in the area are on the order of 10 ohm-meters, so
that skin depth at 5000 Hz is about 20 meters, and at 22 Hz skin depth is
about 340 meters.
The vertical pseudo-section of Figure 4.20 illustrates quite well some
of the problems in CSAMT interpretation. At Station -4.0, a near-vertical
anomaly of low resistivity, limited to this one station, is observed.
109
-------
400 Cycle Engine
Current Electrodes
Magnetic Coil
Note: Nol to scale
Figure 4.18. Schematic layout for CSAMT survey.
-------
Scalar apparent resistivity (ohm-m)
18 17 16 15 14 13 12 11 10 9 8 76
I 1 1 1 1 1 1 1 1 1 1 1 1 S 60 W
Shaded legend
r~n 1.39 -1.58
Figure 4.19. CSAMT Psuedo-section, Lincoln County, Oklahoma.
Ill
-------
Scalar apparent resistivity (ohm-m)
i
S 60 W h-
0 -1 -2 -3 -4 -5
H 1 1 1 1 1
N 60 E
Shaded legend
I 1 1.69 - 2.00
Figure 4.20. CSAMT vertical pseudo-sect ion, Line 8, resistivity versus depth
112
-------
Particularly at lower frequencies, the data approached background values at
Stations -3.0 and -5.0. Only at higher frequencies (shallow exploration
depths) is the anomaly of wider lateral extent. It is very likely that
CSAMT only mapped shallow brine occurrences in this example, and very
little information about resistivity layering as a function of depth was
obtained
4.0 Direct" Current (D.C.) Method
4.1 Instrumentation
Extensive instrumentation is available for D.C. methods, and this fact
reflects the 60-year history of the method. Table 4.6 gives a partial list
of available equipment.
4.1.1 Operational Advantages and Disadvantages
Advantages of the D.C. methods are as follows:
o The equipment is inexpensive and does not require extensive
personnel training.
o Interpretation procedures are well-developed. Forward modeling
curves (the Mooney Master Curves) for a layered earth and for two-
and three-dimensional situations are published and are readily
available.
o Forward and inverse modeling computer programs are readily
available.
o The method can be used for both depth soundings (determining
resistivity strata or layers) and profiling (determining lateral
variations in resistivity).
Disadvantages of the D.C. method are as follows:
o Large transmitter-receiver arrays are required to achieve the
exploration depths associated with injection wells.
o Contacts with the ground (electrodes) are required.
4.1.2 Lateral Resolution
Lateral resolution of the method is poor, mainly because of the large
arrays required for deeply-buried targets. In a typical Wenner or
Schlumberger array, the spacing between current and voltage electrodes is
approximately four times the required exploration depth. For mapping
lateral resistivity variation the dipole-dipole array can provide great
detail. However, two or three dimensional model computer calculations are
required to interpret these data and near-surface lateral conductivity
variations will still dominate and possibly mask the deeper structures.
113
-------
Table 4.6. Partial List of Commercially Available D.C. Equipment
Manufacturer
Bison
Bodenseewerk
Phoenix
Model
2350B
GGA-30
RV-1
Output
Quantity
Displayed
27T(V/I)
V,I
V I
Range
0-10 ohms
0-1000 mV
0-10 V
Sensitivity
1% of full scale
10 /W
0.5% of full scale
Accuracy
± 2%
± 1 /*V
± 1 %
CISCO RSPIP-10 2*(V/I) 0-2 megohms 0.1% of full scale ± 1%
Scintrex RSP-6 V.I 0-1000 mV 0.2mV ±0.5%
114
-------
4.1.3 Vertical Resolution
Vertical resolution is often limited by geologic noise in the data.
The fact that the electrodes are spaced over distances several times the
required exploration depth and the fact that they are moved for each
discrete depth measurement, can introduce much noise in a data set.
4.1.4 Sensitivity to Geologic
Noise
The method is highly sensitive to near-surface geologic noise. The
reason is that inhomogeneities in resistivity near the voltage electrodes
have a large effect on the measurement. Voltage electrodes must be moved
frequently in a vertical sounding, and this introduces another source of
noise. Noise which is caused by near-surface resistivity variation can also
be severe in profiling applications.
4.2 Case Histories: D.C. Methods
Reference:
Reed, P.C., K. Cartwright, and D. Osby. "Electrical Earth Resistivity
Surveys Near Brine Holding Ponds in Illinois," Illinois State Geological
Survey, 1981.
A relatively large number of case histories describing the mapping of
ground-water contamination with D.C. resistivity methods have been
published. These investigations have usually dealt with shallow leakage
from brine holding ponds rather than with contamination in deeper aquifers.
There are at least two reasons for this different focus: first, such
contamination sources have received greater attention historically; and
second, D.C. methods are better suited to investigations of such shallow
targets than they are to deeper targets.
A typical investigation using D.C. methods on a brine holding pond was
reported by Reed et al. in Illinois. Before disposal of brines through
reinjection into producing horizons became common practice, brine was
placed into holding ponds. In Illinois, where annual precipitation exceeds
evaporation, brine discharged into ponds may slowly enter ground water and
eventually may be discharged into streams. The purpose of the study by Reed
et al. was to test the usefulness of D.C. resistivity methods for mapping
brine migration from ponds.
The investigators reported the results of surveys at four sites; one
of these surveys is reviewed here. The ponds were constructed in an
overburden of glacial till, consisting of sands, silts, and clays. A
cross-section of the Woburn site is given in Figure 4.21, and the plan view
is given in Figure 4.22.
115
-------
A
A'
R26
Dike at pond
Creek
R28A
Gamma Log Configuration
Sand
I Slotted Interval
R3
R3A
Pond fluid level
Ground Surface
^s~~"""""—"~ ——
Wisconsinan loess
Weathered loosely
compacted sandy, silty til!
Hagaretown Member
Vandelia Till Member
Compact silty, sandy till
Smithboro Till Member
Kansas Banned Fm.
Compact sandy, silty till
silty till
777777
Pennsylvanian Bedrock
10 ft
100 ft
£
o
V-
-------
Holding pond
Abandoned pond
J978,
0
b
150 300 feet
r1955
Q.^Allx Approximate limit of unvegetated area
— 10 Contour interval showing apparent resistivity - 5 ohm-m
-- 580 Elevation contour (feet)
• Resistivity station
-#• Abandoned oil well
Figure 4.22. Plan view of Woburn Site, with apparent resistivity contours,
117
-------
In Figure 4.23, apparent resistivity curves measured as a function of
spacing for the Wenner array are shown. These data are typical of
near-surface D.C. resistivity soundings in that the curves are distorted
and would be difficult to interpret in terms of a horizontally-layered
earth. However, a continuous decrease in resistivity is readily observed in
going from sounding 11 to 12 to 28. The location map of Figure 4.22 gives
the apparent resistivities at an -"'a'^spacing of 10 feet. This map shows
that the brine contamination is more pronounced in a northern direction.
The depth of brine contamination was not evaluated in this
investigation. Effective exploration depth of a Wenner array with 10-foot
a-spacing is probably about 7 to 10 feet. The best way to determine depth
of contamination is to invert the curves of apparent resistivity versus
a-spacing. Where there are significant distortions from lateral variations
in resistivity, inverse modeling may be difficult or impossible.
The use of D.C. methods for investigations of leakage from injection
wells is expected to be limited. The method may be useful for measuring
shallow contamination, but the large arrays required for mapping of deeper
targets makes the method generally impractical for mapping injection well
contamination. Another problem is created by the rapid decrease in brine
contaminant concentration with increasing distance from the source; this
creates rapid lateral variations in resistivity, making inverse modeling
difficult.
118
-------
>-
*-<
'>
(/)
0>
cc
CO
Q.
O.
40-
--30-
E 20-
.C
~~ 10 H
0
Profile 11
Profile 12
Profile 28
0 20 40 60 80 0 20 40 60 80 0 20 40 60 80
"A" Spacing Depth (ft)
Figure 4.23. Apparent resistivity depth soundings, Woburn, Illinois.
-------
V. SUMMARY AND CONCLUSIONS
Injected fluids present difficult electrical targets. Detection of
target objects, formations, or strata by electrical geophysical techniques
depends on the contrast in some electrical property between the target of
interest and the material within which it is imbedded. Electrical
conductivity is the principal property of interest for most electrical
methods. While we do not fully understand the chemical interactions between
contaminants and aquifer matrices, the conductivity of injected fluids is
roughly proportional to their total dissolved solids content. Injection
usually takes place at depths of a few thousand feet, beneath confining
strata which are often highly conductive, such as shale. Often the
injection zone already contains connate water high in total dissolved
solids and hence of high conductivity. In such situations, with properly
constructed and operated injection wells, surface electrical methods
probably will not be very useful for directly monitoring injected fluids
because the contrast between the injected fluids and the connate water of
the injected stratum, as well as with the conductive overburden, would not
be great.
However, in cases where wells are improperly constructed and operated
or in cases in which fractures connect the injection zones with overlying
fresh water aquifers, movement of the highly conductive injected fluids
into the overlying fresh water aquifers could create a conductive
electrical target which contrasts sharply with the surrounding material. In
such cases, electrical methods may indeed be useful in detecting and
mapping the contaminated zones.
The success of geophysical surveys is critically dependent on the
clear definition of exploration objectives. The objectives of surveys using
electrical geophysical methods must be stated in terms of geoelectric
sections. A geoelectric section displays the variation in electrical
resistivity laterally and vertically across a vertical plane passed through
some line on the surface. The detectability of the contamination will vary
with different sections and methods. Detectabil i ty can often be evaluated
through the use of mathematical models.
The use of mathematical models is essential in interpreting electrical
geophysical data in terms of a geoelectric section composed of of
horizontal strata of different resistivities. Current practice usually
relies on 1-dimensional models for this interpretation. While it is often
possible to observe lateral resistivity changes in the data from electrical
geophysical surveys, the limited lateral resolution of electrical methods
makes the determination of the geoelectric section difficult where rapid
lateral changes exist. In such cases, relatively sophisticated
2-dimensional and 3-dimensional models are needed. However, the use of
these more advanced models is still not routine in electrical prospecting.
120
-------
The use of such models usually substantially increases interpretation costs
for routine surveys; they also require additional data which are often
unavailable.
The important criteria for evaluating the applicability of various
electricil methods to exploration objectives are as follows:
o practical depth of exploration
o lateral resolution
o vertical resolution
o sensitivity to geologic noise
o survey productivity (cost, ease of operation, personnel
qualifications)
Table 5.1 summarizes the discussions of Chapters 3 and 4 concerning
the capabilities of D.C., frequency-domain EM, TDEM, CSAMT, and IP. For
cases whore conductive contamination lies very near the surface (<30 meters
depth), continuous, frequency-domain electromagnetic induction is likely to
be useful in mapping the areal extent of the brines. D.C. resistivity may
be successful at slightly greater depths although cultural interferences
(power lines, pipelines, fences) and geologic noise are likely to cause
problems. Neither of these methods are likely to produce useful information
at depths approaching a typical injection horizon (thousands of meters).
However, because the earliest measurements for time-domain
electromagnetic (TDEM) systems are on the order of 50 to 100 microseconds,
near-surface layers can not often be determined with this method (Fitterman
and Steward, 1986). Cultural interferences (power!ines, pipelines, fences)
in the vicinity of the loop arrays may also cause problems. However,
because of its ability to achieve substantial depths of exploration with
relatively small loop arrays, time-domain electromagnetic induction (TDEM)
appears ta offer the greatest chance of success at depths approaching
injection horizons. The case studies reviewed above tend to support this
assertion. However, even TDEM will need a relatively large conductivity
contrast (factor of two or greater) between the target and surrounding
background to do successful mapping. The target must have lateral
dimensions and a thickness which are a substantial fraction of the depth of
exploration to be detectable; mapping requires targets which have
dimensions greater than the depth of exploration.
121
-------
Table 5.1. Summary Table f Surface-Based Electrical Surveys
Maximum
Effective Depth
Method -'f Investigation
D.C. 1000 meters (with
large arrays) .
Frequency- 30 meters (for Geonics
Domain EM EM34) [400 Hz]
IDEM 3000 meters
CSAHT 100 meters
IP 1000 meters
(with large arrays)
Sensitivity to
Geologic Noise
High
High
Low
High
High
Vertical
Resolution
Limited by
Geologic
Noise
Fair
Good
Poor
Limited by
Geologic
Noise
Lateral Survey
Resolution Productivity
Poor Low
Good High
Good Fair
High (If Good
anomaly is
manifested near
surface)
Low
Cc. .;nent
Requires large
arrays for deep
exploration
•Requires good
conductivity
contrast between
soil and target
Cannot be appl ied
to uppermost layer
Often difficult to
use for determin-
ing resistivity
layering
Brines unl ikely to
cause detectable
IP effects;
requires large
arrays
-------
REFERENCES
1.1 U.S. Environmental Protection Agency, 1984; National Primary Drinking
Water Standards, Title 40, Code of Federal Regulations.
1.2 M(Kee and Wolfe, Water Quality Criteria, 2nd Edition, California
State Water Quality Control Board Publication No. 3-A (1963).
•
1.3 Fairchild, D.D., "Review of Data Bases On Injection Wells,"
Environmental and Ground Water Institute, University of Oklahoma,
Norman, OK; June 1984.
1.4 U.S. Environmental Protection Agency, Office of Drinking Water,
Report to Congress on Injection of Hazardous Waste, Washington, D.C.,
May 1985.
1.5 Scalf, M.R., Keely, J.W., and Lafevers, C.J., Ground Water Pollution
in the South Central States," EPA-R2-73-268, Washington, D.C., 1973.
1.6 Miller, D.W., Waste Disposal Effects on Ground Water, Berkeley, CA,
1980
1.7 Collins, A.G., "Saline Groundwater Produced with Oil and Gas,"
EPA-66012-74-010, Washington, D.C., 1974.
1.8 U.S. Environmental Protection Agency, "The Report to Congress: Waste
Disposal Practices and Their Effects On Ground Water," EPA
570/9-77-001, Washington, D.C., 1977.
2.1 Greenhouse, J.P., "Surface and Borehole Geophysics in Contaminant
Hydrogeology," Notes from Hydrogeology Field School, CFB Borden,
1982.-
2.2 McNeill, J.D., "Electrical Conductivity of Soils and Rocks,"
Technical Note TN-5, Geonics Limited, Mississauga, Ontario, Canada,
October 1980.
2.3 Telford, W.M.; Geldart, L.P.; Sheriff, R.E., Keys, D.A., Applied
Geophysics, Chapter 5, Cambridge University Press, 1976.
2.4 Keys, W.S., and L.M. MacCary, "Applications of Borehole Geophysics to
Water Resources Investigations," in Techniques of Water
Investigations, U.S. Geological Survey, Ch. El, Book 2, 1972.
2.4 Guyod, H., "Interpretation of Electric and Gamma Ray Logs in Water
Wells," Log Analyst, v.6, 1966, p 29-44,
2.5 Sawyer, Clair N., and McCarty, P.L., Chemistry for Sanitary
Enjineers, McGraw-Hill Series in Sanitary Science and Water Resources
Engineering, McGrawHill, 1967.
123
-------
3.1 Kaufman, A.A. and G.V. Keller, Frequency and Transient Soundings,
Elsevier, New York, 1983.
3.2 Kaufman A.A., "Tutorial Distribution of Alternating Electrical
Charges in a Conducting Media," Geophysical Prospecting, v. 133, p.
17;, 1985..
3.3 Alpin, L.M., "Source of the Field in the Theory of Electroprospecting
by Direct Current," Applied Geophysics, v. 3, 1947.
3.4 Van Nostrand, R.G., and K.L. Cook, "Interpretation of Resistivity
Data," Geological Survey Professional Paper 499, 1966.
3.5 Newman, G.A., Gerald W. Hohman, and W.L. Anderson, "Transient
Electromagnetic Response of A Three-Dimensional Body in a Layered
Earth," Geophysics, v.51, p. 1608, 1986.
3.6 Strangway, D.W., Swift, C.M., and Holmes, R.C., "The Application of
Audio-Frequency Magnetotellurics (AMT) to Mineral Exploration,"
Geophysics, v.38, p. 1159, 1973.
3.7 Hoekstra, P., "Electromagnetic Methods for Mapping Shallow
Permafrost," Geophysics, v.43, p. 782, 1978.
3.8 Olhoeft, G.R., "Electrical Properties of Rocks," in The Physics and
Chemistry of Rocks and Minerals, p. 261, 1975, John Wiley & Sons,
N.Y.
3.9 Ogilvy, A.A. and E.N. Kuzmino, "Hydrogeological and Engineering
Geologic Possibilities for Exploring the Method of induced
potentials," Geophysics, v.37, p. 839, 1972.
4.1 "Study of Subsurface Contamination with Geophysical Monitoring
Methods at Henderson, Nevada," E.G. Walther, D. LaBrecque, and D.D.
Weber, Lockheed Engineering and Management Services Company, Inc.,
and R.B. Evans and J.J. van Ee, EMSL-LV, USEPA, Proceedings of the
National Conferences on Management of Uncontrolled HGazardous Waste
Sjtes., 1983, Washington, D.C.
4.2 Olhoeft, Gary R., and D.R. Capron, "The Use of Geophysics in
Hazardous Waste Investigations," U.S. Geological Survey, Denver, CO,
1986
4.3 Wightman, W.E., A.A. Kaufman, and P. Hoekstra, "Mapping Gas-Uater
Contacts in Shallow Producing Formations with Transient EM,"
presented at 53rd Annual Meeting, Society of Exploration Geophysics.
Las Vegas, Nevada (1983)
4.4 Raiinovich, B.I., and V.S. Surkov, "Results of the Use of the 2 SB
Method in the Siberian Platform," in Theory and Use of
Electromagnetic Fields in Exploration Geophysics, Akad. Nouk, USSR,
Novosibirsk, p. 3, 1973.
124
-------
4.5
4.6
4.7
4.8
Fitter-man, D.V., P.V. Raab, and F.C. Fri schknecht, "Detection of
Brine Contamination from Injection Wells Using Transient
Electromagnetic Soundings," EPA, pre-issue copy, January 1986.
Sy>d, T., K.L. Zonge, S. Figgins, and A.R. Anzzolin, "Application of
the Controlled Source Audio Magnetotellurics (CSAMT) Survey to
Delineate Zones of Ground-Water Contamination: A Case History,"
presented at the Second National Conference and Exposition on Surface
and Borehole Geophysical Methods in Ground-Water Investigations,
National Water Well Association, Fort Worth, TX, February 12-14,
1985. -
Reed, P.C., K. Cartwright, and
Surveys Near Brine Holding
Geological Survey, 1981.
D. Osby. "Electrical Earth Resistivity
Ponds in Illinois," Illinois State
Fitterman, D.V., and M.T. Steward, "Transient Electromagnetic
Sounding for Groundwater," Geophysics, V. 51, p. 995-1005, 1986.
125
-------
APPENDIX 1: 1984 EPA INJECTION WELL INVENTORY
EPA
REGION
I
II
III
IV
STATE
CT
MA
ME -
NH
RI
VT
NJ
NY
DE
MO
PA
VA
WV
AL
Fl
GA
KY
MS
NC
SC
CLASS CLASS
I II
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
23 3859
23 3859
0 0
0 0
5 4073
0 1
7 370
12 444
9 207
105 136
0 0
2 4357
8 1223
3 0
0 0
CLASS
II!
0
0
0
0
0
0
0
0
249
249
0
0
0
24
17
41
10
0
0
0
0
0
0
CLASS
IV
0
0
0
0
0
0
0
0
0
0
0
3
11
3
0
17
0
3
0
0
0
2
3
CLASS
V
173
35
18
33
42
1
302
1327
2324
3651
3
965
15942
1748
169
18827
249
4360
4
307
118
45
33
STATE
TOTAL
173
35
18
33
42
1
1327
6455
5
968
20031
1776
563
475
4604
4
4666
1349
45
36
REGIONAL
TOTALS
302
7782
23341
126
-------
APPENDIX 1 (Cont'd)
EPA
REGION
V
VI
VII
VIII
IX
STATE
Tt-
IL
IN
HI
MN
OH
VI
AR
LA
OK
NM
TX
IA
KS
HO
NE
CO
HT
NO
SO
UT
WY
AZ
CA
HI
NV
CLASS
I
0
127
9
15
86
0
18
0
139
8
80
15
1
160
264
0
59
0
0
59
1
0
1
0
0
3
5
0
39
0
0
CLASS
II
13
5935
18493
3798
1193
0
3911
0
27395
1172
4399
14091
4344
47189
71195
0
30081
300
942
31323
998
1447
570
8
504
4416
7943
6
13844
518
6
CLASS
III
0
10
0
0
110
0
2
0
112
0
227
0
114
30739
31080
0
394
0
0
394
59
0
4
0
30
898
991
484
6
0
0
CLASS
IV
12
20
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
2
0
0
0
0
0
2
5
22
0
0
CLASS
V
33
5149
372
28
2768
19
2814
0
6001
55
0
0
104
436
595
705
931
254
88
1978
6
1
0
0
7
10
24
27
2
1
13
STATE
TOTAL
58
18875
3841
4157
19
6745
0
1235
4706
14106
4564
78524
705
31465
554
1030
1066
1448
575
8
541
5327
522
13913
519
19
REGIONAL
TOTALS
11242
33637
103135
33754
8965
39 14374
490
27
43
14973
127
-------
APPENDIX 1 (cont'd)
EPA
REGION
X
STATE
AK
10
OR
UA
CLASS
I
0
0
0
1
1
659
CLASS
II
160
1
' 0
161
166630
CLASS
III
0
0
0
0
0
33367
CLASS
IV
1
0
0
10
11
78
CLASS
V
3
2042
712
6357
9114
45684
STATE
TOTAL
164
2042
713
6368
REGIONAL
TOTALS
9287
246418
-------
APPENDIX 2
CLASS I WELL INJECTION ZONE CHARACTERISTICS
STATE AVERAGE DEPTH
TO TOP OF
INJECTION
ZONE (feet)
ALABAMA
ALASKA
ARKANSAS
HEAKINS
CALIFORNIA
FLORIDA
ILLINOIS
INDIANA
SPRING
KANSAS
KENTUCKY
LOUISIANA
MICHIGAN
TRAVERSE
MISSISSIPPI
OHIO
2
OKLAHOMA
PENNSYLVANIA
TEXAS
GRAN I
NATIONAL AVERAGE
(by well)
4095
2032
2867
6139
2067
2512
2332
3257
3115
3267
3447
4413
3479
1361
1611
5371
4063
AVERAGE
THICKNESS
OF INJECTION
INTERVAL (feet)
72
115
108
751
513
574
1420
559
2590
281
379
1212
177
964
70
702
556
LITHOLCGY FORMATION NAMES
SS. CLAY. MARL NAHECLA NANAFALIA
SH. SILT. SS TERTIARY.
SAGAVANIRKTOK
SS. SH. CLAY GRAVES. TOKIO
BLOS, SOM.
SS. SILT RIO BRAVO
LS CEDAR KEYS. LAUSON.
LOWER FLORIDIAN
OOL. SS. LS POTOSI. EMINENCE.
MT. SIMON, SALEM
SS MT. SIMON, BETHEL,
CYPRESS. TAR
EAU CLAIRE
DOL. LS. CHERT ARBUCKLE LIR
OOL KNOX
SS. CLAY. SILT HOSSTON. FLEMING.
SPARTA
SS. LS. DOL MT. SIMON. EAU CLAIRE
GALESVILLE. FRANCON
IA. DUDEE
SS HOSSTON
SS. DOL HT. SIMON. MAYNARD-
V1LLE. ROME
OOL. LS. SS. ARBUCKLE
CHERT
LS BOSS ISLAND
SS. CLAY. SHALE CATAHOULA. OAKVILLE
FRIO. SAN ANDRES,
BLOSSOM. JACKSON,
ANHUAC, LOWER
WASH. GLORIETTA
LEGEND: SS - SANDSTONE SH - SHALE OOL - DOLOMITE LS- LIMESTONE SILT - SILTSTONE
129
-------
AVERAGE
CONFINING ZONE
STATE THICKNESS (ft)
ALABAMA
ALASKA
ARKANSAS
CALIFORNIA
FLORIDA
ILLINOIS
INDIANA
KANSAS
KENTUCKY
LOUISIANA
MICHIGAN
OHIO
OKLAHOMA
PENNSYLVANIA
TEXAS
150
15000
521
700
311
319
256
3089
700
442
. 538
1254
83
395
1442
LITHOLOGY
CLAY
SS
SH.
SS.
CLAY
SH.
SS.
LS.
ML.
SH.
SH.
DOL.
SH.
SH.
SH .
MARLS. CHALK
SH. SILT
, OOL. ANHY
LS. DOL. SILT
SH. SILT
SH. SS
LS
CLAY, SS. SILT
DOL. LS
SH. LS
LS
LS. CHERT
CLAY. SS
NATIONAL AVERAGE 928
(by w«ll)
SS - SANOr'ON£
SH - SHALE
DOL - DOLOMITE
LS - LIMESTONE
SILT- SILTSTONE
APPENDIX 3
CONFINING ZONE CHARACTERISTICS
OF
CLASS I INJECTION WELLS
FORMATION NAHES
PERHAFRONT
SARATOGA. ANNONA, BROWNSTOWN.OZAN
FREEHAN-JEWETT. VALLEY SPRING-IONE
CEDAR KEYS. BUCATUNNA
PRAIRIE DU CHIEN. EAU CLAIR.
MAQUOKETA. NEW LABANY. ST. GENEVIEVE
EAU CLAIRE. TAT SPRINGS
WELLINGTON TO SIMPSON
TRENTON. BLACK R. CHAZY
SLIGO. BURKEVILLE. FLEMING
ANTRIM. PRAIRIE OU CHIEN. ELLSWORTH,
BELL. BAYPORT-HICHIGAN
EAU CLAIRE. ROCHESTER. ROME. TOMSTOVN
WOOOFORO. CHATTANOOGA
ANAHAU. JASPER. BEAUMONT. OAKVILLE,
LAGARTO. LISSIE. MONTGOMERY. BETTY.
GRAY8URG, YATES, FR10. FLEMING.
BURKEVILLE. ANAHUAC. GLENROSE
130
------- |