\
GROUND WATER TRACERS
\
7
S.N. DAVIS, D.J. CAMPBELL, H.W. BENTLEY & TJ. FLYNN
NATIONAL GROUND WATER ASSOCIATION
-------
GROUND WATER TRACERS
by
Stanley N. Davis
Darcy J. Campbell
Harold W. Bentley
Timothy J. Flynn
Department of Hydrology and Water Resources
University of Arizona
Tucson, Arizona 85721
Cooperative Agreement CR-810036
Project Officer
Jerry Thornhill
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
Published by
National Ground Water Association
6375 Riverside Dr.
Dublin, OH 43017
-------
FOREWORD
The Environmental Protection Agency was established to coordinate
administration of the major federal programs designed to protect the quality
of our environment.
An important part of the Agency's effort involves the search for infor-
mation about environmental problems, management techniques, and new technol-
ogies through which optimum use of the nation's land and water resources can
be assured. 'S'nd the threat which pollution poses to the welfare of the Ameri-
can peopie can be minimized.
UPA's Offd.ce of Research and Development conducts this search through a
nationwide network of research facilities.
As one of these facilities, the Robert S. Kerr Environmental Research
Laboratory is the Agency's center of expertise for investigation of the soil
and subsurface environment. Personnel at the Laboratory are responsible for
management of research programs to: (a) determine the fate, transport and
transformation rates of pollutants in the soil, the unsaturated zone and the
saturated zones of the subsurface environment; (b) define the processes to
be used in characterizing the soil and subsurface environment as a receptor
of pollutants; (c) develop techniques for predicting the effect of pollu-
tants on ground water, soil, and indigenous organisms; and (d) define and
demonstrate the applicability and limitations of using natural processes,
indigenous to the soil and subsurface environment, for the protection of
this resource.
This report contributes to that knowledge which is essential in order
for EPA to establish and enforce pollution control standards which are
reasonable, cost effective, and provide adequate environmental protection
for the American public.
Clinton W. Hall
Director
Robert S. Kerr Environmental
Research Laboratory
11
-------
ORIGIN OF FUNDING
The information in this document has been funded wholly or in part by
the United States Environmental Protection Agency under cooperative agree-
ment CR-81003601-0 to the University of Arizona. It has been subject to the
Agency's peer and administrative review and it has been approved for publi-
cation as an EPA document.
iii
-------
PREFACE
An Introduction to Ground-Water Tracers has been developed in conjunc-
tion with the U.S. Environmental Protection Agency for use by persons in-
volved in efforts to determine the direction and velocity of ground-water
flow. Techniques described are those which are currently in use and methods
which may be of future significance.
For those concerned with protecting ground water, this document may be
helpful as a ready summary of methods to determine the movement of ground
water and contaminants in an aquifer.
Library of Congress Cataloging-in-Publication Data
Main entry under title:
Ground-water tracers.
"Cooperative agreement CR-810036."
Project coordinated by the Robert S. Kerr Environ-
mental Research Laboratory.
Bibliography: p.
1. Groundwater tracers. 2. Water, Underground—
Pollution. I. Davis, Stanley N. (Stanley Nelson),
1924- . II. Thornhill, Jerry. III. National
Water Well Association. IV. Robert S. Kerr Environ-
mental Research Laboratory.
GB1197.6.G76 1985 628.1'68'0287 85-21785
IV
-------
CONTENTS
Foreword
Origin of Funding
Preface iv
Figures vii*
Tables x
Acknowledgments xi
Chapter 1 1
Introduction 1
General Characteristics of Tracers 1
History of Ground-water Tracing 2
Purpose and Scope 3
Public Health Considerations 5
Chapter 2 7
Hydrogeologic Principles 7
Darcy's Law. 7
Direction of Water Movement 11
Travel Time 14
Sorption of Tracers and Related Phenomena. ..... 14
Hydrodynamic Dispersion and Molecular Diffusion 18
Chapter 3 21
Practical Aspects 21
Planning a Test 21
Types of Tracer Tests 24
Single-Well Techniques 26
Injection/Withdrawal 26
Borehole Dilution 35
Two-Well Techniques 37
Uniform Flow 37
Radial Flow 38
Design and Construction of Test Wells 39
Injection and Sample Collection 47
Interpretation of Results 52
Chapter 4 61
Types of Tracers. ........ 61
Temperature 61
Field Methods 65
Detection and Analysis 65
Additional Information 66
Solid Particles • 66
Paper and Simple Floats 67
Field Methods 67
Detection 67
Additional Information 67
Signal-Emitting Floats 68
v
-------
Yeast 68
Field Methods 69
Detect ion/Sample Analysis 69
Additional Information 69
Bacteria 70
Field Methods 72
Detection 72
Additional Information 72
Viruses 73
Field Methods 76
Additional Information 77
Spores 78
Field Methods 79
Detection and Analysis 81
Additional Information 83
Ions 86
Field Methods 88
Detection and Analysis 91
Discussion of Specific Ion Tracers 93
Chloride 93
Bromide 94
Lithium 96
Ammonium 96
Magnesium 96
Potassium 96
Iodide 96
Organic anions 96
Dyes 97
Field Methods 100
Detection and Analysis ..... 102
Additional Information 103
Discussion of Specific Dye Tracers 110
Green Dyes HO
Fluorescein HO
Pyranine 113
Lissamine FF 113
Orange Dyes 113
Rhodamine B 113
Rhodamine WT 114
Sulfo rhodamine B 118
Blue Dyes 118
Some Common Nonionized and Poorly Ionized Compounds 119
Detection 121
Gases 121
Introduction 121
Inert Radioactive Gases 122
Inert Natural Gases 122
Fluorocarbons 124
Field Methods 1128
Analysis 129
VI
-------
Page
Stable Isotopes 129
Introduction 129
Hydrogen and Oxygen 130
Nitrogen 131
Sulfur
Carbon
Isotopes of Other Elements 135
Field Methods 135
Analyses 13'
Radionuclides 13?
Introduction 137
Injected Tracers 138
Atmospherically Distributed Radionuclides 143
Field Methods 147
Analysis • 148
Appendixes
Appendix A 149
Additional Uses of Water Tracers 149
Appendix B i54
A Discussion of Dispersion and Diffusion 154
Appendix C 263
Factors to Consider in Tracer Selection 263
Appendix D 269
Chemical Supply Companies 269
Appendix E 273
Analytical Methods for the Detection of Tracers 273
References 286
VII
-------
FIGURES
Page
2.1 Darcy's law 8
2.2 Direction of water movement 12
2.3 Divergence from regional direction of water
movement 13
2.4 Average travel time of ground water 15
2.5 Hydrodynamic dispersion 19
2.6 Molecular diffusion 20
3.1 Slope of the water table 23
3.2 Tracer tests at Sand Ridge State Forest, Illinois 25
3.3 Different arrangements for ground-water tracing 27
3.4 Tracer test in alluvium 41
3.5 Two-well tracer test in fractured rock 44
3.6 Tracer test using water temperature 48
3.7 Variation of chemical quality with time 51
3.8 Arrival of tracer front 53
3.9 Dispersion in breakthrough curves 55
3.10 Incomplete saturation of aquifer 56
3.11 Conservative vs. nonconservative tracers 58
3.12 Computer-generated type curves 59
4.1 Results of tracer test using hot water 64
4.2 Results of two-well tracer tests using bromide
and yeast 71
4.3 Two-well tracer test using rhodamine WT and E. Coli 74
4.4 Use of plankton net to catch spores 82
viii
-------
Page
4.5 Comparison of tracer pulse from lycopodium spores
and a dye in a karst system 84
4.6 Comparison of several tracers in a laboratory test 95
4.7 Excitation and emission characteristics of
rhodamine WT 99
4.8 Automatic monitoring system for a stream 104
4.9 Effect of pH on fluorescence 106
4.10 Adsorption of dyes on kaolinite 107
4.11 Comparison of travel time for lycopodium spores, hot
water, and fluorescein 109
4.12 Arrival times of tritium and rhodamine WT in a
field test 116
4.13 Laboratory experiments with fluorocarbon tracers
and tracer elution curves for NaCl and CC1 F 126
4.14 Relation between oxygen-18 and deuterium for
natural waters 132
4.15 Oxygen-18 variations in ground water of the
Tucson basin 133
4.16 Carbon isotopes in methane 136
4.17 Local direction of ground-water movement using
radioactive tracers 141
4.18 Average annual tritium concentration of rainfall
and snow 145
IX
-------
TABLES
2.1 Representative Values of Porosity, Hydraulic
Conductivity, and Permeability ................. 10
4.1 Comparison of Microbial Tracers ................ 75
4.2 Comparison of Lycopodium and Fluorescent Dye
Properties ........................... 85
4.3 Analytical Methods for Ionic Tracers .............. 92
4.4 Description of Dye Tracers ................... 98
4.5 Sensitivity and Minimum Detectable Concentrations
of Dye Tracers ......................... 105
4.6 Relative Costs of Dyes
4.7 Sorption of Dyes on Bentonite
4.8 Compounds Soluble in Water ................... 120
4.9 Gases of Potential Interest as Tracers ............. 123
4.10 Properties of Fluorocarbon Compounds .............. 125
4.11 Commonly Used Radioactive Tracers ............... 139
4.12 Environmental Radionuclides ........ •
B.I Values of Dispersivities Measured by Various
Methods
160
x
-------
Acknowledgments; Mr. Jack Keeley of the Robert S. Kerr Environmental
Research Laboratory encouraged us in launching the initial project. Drs.
Glenn M. Thompson and Emanuel Mazor gave valuable direction in the initial
stages of the work. Much of the library research was done by Michael G.
Wallace. Drafting was completed by Ms. Ann Cotgageorge, and manuscript typ-
ing was done by Ms. Corla Thies. Field assistance in conducting tracer tests
was provided by Jesus Carrera and Morley Weitzman. To these individuals and
to many others who have helped us, we are grateful.
xi
-------
CHAPTER ONE
INTRODUCTION
General Characteristics of Tracers
As used in hydrogeology, a tracer is matter or energy carried by ground
water which will give information concerning the direction of movement
and/or velocity of the water and potential contaminants which might be
transported by the water. If enough information is collected, the study of
tracers can also help with the determination of hydraulic conductivity, por-
osity, dispersivity, chemical distribution coefficients, and other hydro-
geologic parameters. A tracer can be entirely natural, such as the heat
carried by hot-spring waters; it can be accidentally introduced, such as
fuel oil from a ruptured storage tank; or it can be introduced intention-
ally, such as dyes placed in water flowing within limestone caverns (Davis
et al., 1980).
A tracer should have a number of properties in order to be generally
useful. The most important criterion is that the potential chemical and
physical behavior of the tracer in ground water must be understood. The ob-
jective is commonly to use a tracer which travels with the same velocity and
direction as the water and does not interact with solid material. For most
uses, a tracer should be nontoxic. It should be relatively inexpensive to
use and should be, for most practical problems, easily detected with widely
available and simple technology. The tracer should be present in concentra-
tions well above background concentrations of the same constituent in the
natural system which is being studied. Finally, the tracer itself should
not modify the hydraulic conductivity or other properties of the medium
being studied.
1
-------
Obviously, an ideal tracer does not exist. Because of the complexities
of the natural systems which are studied, together with the large number of
requirements for the tracers themselves, the selection and use of tracers is
almost as much of an art as it is a science. This manual will describe some
of this art and also explain some of the important scientific principles
needed to apply the art effectively.
History of Ground-Water Tracing
One of the first tracing experiments was performed almost 2,000 years
ago when Philip, the tetrarch of Trachonitis, threw chaff into a crater
lake. He reported that the chaff appeared downgradient in one of the
springs at the headwaters of the Jordan River. Although Josephus reported
that the experiment was a success, Mazor (1976) demonstrated by chemical and
isotopic measurements that the supposed underground connection would be
highly unlikely. Around the same period of time, Strabo described karst
tracing experiments (Burden, 1963). The karst areas of Europe abound with
folk legends of cavern connections demonstrated by straying ducks and dogs
(Brown and Ford, 1971).
Dyes and salts have been used in Europe since 1869 to find hydraulic
connections in karst areas (Kass, 1964). Among the first dye experiments
was an effort made to establish the water origin of typhoid fever in France
in 1882. Dole (1906) mentioned the work of Dr. Carrieres during this severe
epidemic near Paris. The fluoroscope was invented in France in 1901 by
M. Trillat and perfected by M. Marboutin. This instrument greatly increased
the precision of fluorescent dye measurements. Dole described work in
France with karst and soil tracing and pioneered the use of fluorescein in
English-speaking countries.
-------
During the same time, Thiem used sodium chloride in investigations in
Leipzig to determine the flow velocity of water (Slichter, 1902). Thiem
sampled for chloride, which he analyzed in the laboratory. Slichter modi-
fied Thiem's method by obtaining continuous recordings of electrical con-
ductivity in the field. Ammonium chloride was used in Schlicter's experi-
ments. Slichter (1905) also determined time of travel and direction of flow
in perhaps the first field tracer tests in porous media. His use of shallow
drive point wells and resistivity measurement was modified by the authors of
this manual, and was used in small-scale field tests described in a subse-
quent chapter.
In the 1950's, radioactive tracers were developed (Fox, 1952), allowing
very precise and selective tracer measurement. They were quite popular,
although their use has been curtailed in many countries for public health
reasons. In the 1960's, naturally-occurring radioisotopes and stable iso-
topes became an invaluable tracing tool. In the last two decades, re-
searchers have developed extremely sensitive tracers, including fluorinated
organic acids and halocarbons.
Purpose and Scope
The purpose of this manual is to provide a guide to the use of ground-
water tracers for practicing engineers, hydrologists, and ground-water geol-
ogists. Some parts of the manual may prove to be useful to research scien-
tists; however, emphasis has been placed on the practical rather than on the
theoretical aspects of tracers. Specifically, the manual is concerned with
the selection of tracers, their field application, collection of samples
containing tracers, sample analysis, and interpretation of the results.
-------
Only a general introduction will be provided, however, to laboratory analy-
ses and quantitative interpretation of the results.
The number of possible tracers which can be used in ground water must
number in the thousands, if all trace constituents together with stable iso-
topes and radionuclides are considered. In this manual, emphasis has been
placed on the more practical tracers, while several other tracers have been
mentioned which may be used for special applications. References given in
the bibliography will cover a large number of additional tracers which are
not discussed in detail in the manual.
Except for volatile tracers and tracers which break down in sunlight,
ground-water tracers can be used for surface-water work. The reverse is not
always true. One of the most common errors in ground-water tracer work is
to use dyes which have been applied successfully in surface-water studies.
Many excellent surface-water tracers are available, but cannot be used
directly for ground-water work. This manual is intended for people inter-
ested in ground-water tracing; thus, many of the limitations of surface-
water tracers as applied to subsurface problems have been pointed out.
While some space has been given to natural tracers and tracers intro-
duced accidentally through pollution, most of the manual is focused on mate-
rial injected intentionally into ground water for the purpose of tracing the
movement of fluids in active ground-water systems. Tracer applications in
the petroleum industry are mentioned in Appendix A.
The purpose and importance of tracer tests was eloquently described by
Dole (1906), and his sentiments are consistent with the philosophy of this
manual•
-------
"Consequently, every means for determining the flow and
pollution foci of underground waters should be used. In
studying the potability of a well or spring water, it is
important to know not only its chemical composition, but
also its source, its rate of flow, the area tributary to
it, the nature of the material through which it passes,
and the contaminations to which it may be subjected
before or during its underground journey. It is often
a matter of much importance to know whether the flow is
from a cesspool toward a neighboring well or in the oppo-
site direction; it may be necessary to determine whether
or not water seeps from a contaminated brook into wells
of a neighboring region; whether collecting galleries
for public water supplies receive seepage from well-
established sources of contamination; whether, in general,
known foci of pollution are in immediate, though obscured,
connection with sources of drinking water. Knowledge of
this nature is especially important in the study of waters
passing through formations full of seams or crevices, where
there is opportunity for rapid circulation without much
purification. The determination of the area draining to
the underground supply affords data in regard to the quan-
tity of available water as well as its quality."
Public Health Considerations
Tracers discussed in detail in this manual are mostly harmless and
should pose no public health problems. One cannot emphasize too strongly,
however, that each artificial introduction of tracers must be done with a
careful consideration of possible health implications. Commonly, investiga-
tions using artificially-introduced tracers must have the approval of local
or state health authorities, local citizens must be informed of the tracer
injections, and the results should be made available to the public. In
addition, under some circumstances, analytical work associated with tracer
studies must be done in appropriately certified laboratories. Because of
the extreme variability of local and state regulations and because of the
rapid changes in these regulations, it is impractical to include an exten-
sive discussion of public health aspects of tracers. Therefore, the authors
-------
disclaim any responsibility for judging the health effects of the tracers
covered in this manual.
-------
CHAPTER TWO
HYDROGEOLOGICAL PRINCIPLES
The following discussion is intended only as a brief introduction to
some of the hydrogeological principles necessary for the application of
tracer technology. More complete information can be found in standard text-
books on the topic (Bouwer, 1978; Davis and DeWiest, 1966; Fetter, 1980;
Freeze and Cherry, 1979; Heath and Trainer, 1968; and U.S. Bureau of Recla-
mation, 1977).
Darcy's Law
Most ground-water flow is governed by Darcy's law, which must be under-
stood in order to design successful tracer tests. For a simple flow system,
Darcy's law states that the volume Of water flowing per unit of time, Q,
through a given cross section, A, is directly proportional to the hydraulic
gradient, —, and the hydraulic conductivity, K. Stated as an equation,
AL
this is:
Q = ^ Ah (1)
AL
The meaning of this equation is illustrated in Figure 2.1.
The hydraulic conductivity, K, is in itself a complex measure of a
number of physical factors. One useful equation relating these factors is:
K = d2c g (2)
-------
oo
Figure 2.1. An illustration of Darcy's law using a tube filled with sand.
The energy loss in the flow system is proportional to the change in hydraulic
head, Ah, over an incremental length, AL, and inversely proportional to the
hydraulic conductivity, K, which is a constant only if the fluid properties
and the gravitational field are constant. The discharge, Q, flowing through
the tube is measured in any consistent units of volume per unit time (length3/
time).
-------
where d is some average aperture width, such as the diameter of pores
between sand grains or the width of cracks in rocks; c is a unitless measure
of the geometry of the pores; g is the acceleration of gravity; p is the
density of the fluid; and y is the dynamic viscosity of the fluid. The pro-
duct d2c in Equation (2) is commonly designated as the permeability, k, of
the solid material. The older hydrologic and engineering literature com-
monly uses the term "permeability" to designate hydraulic conductivity. The
permeability, however, is a property of the solid material through which the
water (or other fluid) is moving. In contrast, the hydraulic conductivity
includes the properties of the fluid and the field of gravity as well as the
permeability. Typical values for the hydraulic conductivity, K, and the
permeability, k, of natural materials are given in Table 2.1.
Another equation expressing the conservation of mass of water, assuming
that water is incompressible, is useful in the consideration of tracer move-
ment. This equation in simple form states that:
Q = ^ ngA (3)
in which Q and A are identical to these terms found in Equation (1), and v
is the average velocity of the ground water. The term rig is the effective
porosity, or the pore volume which transmits ground water.
In most sections of this manual, porosity, permeability, and hydraulic
conductivity are assumed to be constants in a given field situation. Under
most conditions, these values can vary widely in space and will even vary
with time. Spatial variations are evident in all geologic materials and
need no explanation at this point. Temporal variations are not as
-------
TABLE 2.1
Representative Values of Porosity,
Hydraulic Conductivity, and Permeability
(Data abstracted from several sources, including Davis, 1969)
Material
Granite, dense
Granite, fractured
Quartz! te, dense
Schist, highly-weathered, clay-rich
Schist, fractured and partly weathered
Basalt, dense
Tuff, friable
Conglomerate, hlghly-lithlf ied
Sandstone, aedlum-grained
Shale, compacted
Line stone, dense
Clay, narine
Sand, •edlum-grained
Sand, aediun to coarse-grained
Sand, fine-grained
Silt, sandy
Silt, loess, fine-grained
Gravel, fine-grained, some sand
Porosity
(Z)
0.3
1.2
0.6
48
5
7.7
36
17.3
15.6
21
10.1
46.5
42.9
37.4
40.1
39.4
50.0
32.1
Hydraulic
Conductivity
(neters/day)
1.5 x 10~6
2 x 10-2
1.4 x 10~6
2.3 x 10-2
1.04
1.04 x 10~S
1.04 x 10-3
3.6 x 10-"
5.6 x 10-2
3 x 10~6
5.7 x 10-3
1.2 x 10~5
13.5
20.4
1.1
2.8 x 10-2
0.24
66
Permeability
(darcys)
2.0
2.7
1.9
3.1
1.4
1.4
1.4
4.9
7.6
4 x
7.7
1.6
18.2
27.5
1.5
3.8
0.33
89
x 10~6
x 10-2
x 10~6
X ID'2
x 10"5
x 10-3
x ID""
x 10~2
10"6
x 10~3
x 10" 5
x 10-2
Mote: With water at 20"C, aaterial having one darcy permeability will have a hydraulic
conductivity of 0.74 neters/day which is equivalent to 2.43 feet/day.
10
-------
self-evident; however, they can also be very large. Examples would be the
variations caused by the effects of natural or artificial compaction of
sediments, the dissolution of minerals making up the solid matrix under con-
sideration, the deposition of mineral matter, and, very importantly, the
expansion or contraction of clays and other fine-grained material in
response to changes in water chemistry. The last example becomes very
important when considering the use of artificially-introduced tracers.
Direction of Water Movement
In order to complete a tracer test using more than one well, the gen-
eral direction of ground-water movement should commonly be known. This is
particularly true if the travel of tracers is to be studied using two wells
with ground water flowing under a natural gradient. As a first approxima-
tion, ground water will flow in the direction of the steepest hydraulic gra-
dient. This direction is generally perpendicular to lines of equal ground-
water levels as defined by water levels in wells penetrating the water-
bearing zone of interest (Figure 2.2).
Unfortunately, local differences in hydraulic conductivity may amount
to several orders of magnitude. Local flow directions may then be highly
distorted, and actual directions will diverge widely from directions pre-
dicted on the basis of widely-spaced water wells (Figure 2.3). It is not at
all uncommon to inject a tracer in a well and not be able to intercept that
tracer in another well just a few meters away, particularly if the tracer
flows under the natural hydraulic gradient which is not disturbed by pump-
ing. This problem will be discussed in more detail later in this manual.
11
-------
land
surface
lines of equal water elevation
water
table
water movement perpendicular to
lines of equal water elevation
Figure 2.2. Contours of the water table are established by measuring the
elevation of water levels in wells. As shown in this figure, ground water
will flow in the general direction in which the water level slopes. Unless
geologic or hydrologic evidence indicates otherwise, ground water is assumed
to flow exactly perpendicular to the lines of equal water elevation.
-------
land surface
contoured on
regional data
water table
less permeable
material
buried
channel
•actual movement almost at right
angles to direction predicted by
regional water levels
Figure 2.3. Although regional data from widely separated wells may
suggest a certain direction of ground-water flow, local zones of high
permeability caused by fractures in rock, solution openings, or local
zones of coarse sediments like that shown in this figure may divert
the flow in an entirely different direction. This effect is one of
the most common causes of failures in ground-water tracing attempts.
Tracers injected in one well simply do not travel to the sampling
point because of heterogeneities in the system.
13
-------
Travel Time
_ AL
The term v In Equation (3) can be replaced by -^ , where At is the
length of time taken by the average water particle to move through a dis-
tance of AL. Then Equations (1) and (3) can be combined and the identical
term A (area) cancelled, resulting in the equation:
n (AL)2
t = (4)
* KAh v '
This equation can be used to estimate the time which would be taken by
water to travel from one point to another. If a tracer which travels with
the water is injected, t is also the travel time of the tracer. The use of
Equation (4) is illustrated in Figure 2.4.
One of the common errors in tracer tests is to conduct tests between
points which are separated by too great a distance. As can be seen in Equa-
tion (4), the expected travel time for a given head drop is a function of
the distance squared (AL2) and therefore increases very rapidly with the
distance, AL. Thus, a tracer test in one region using a specific hydraulic
head drop of Ah over a distance of 1,000 meters would take 10,000 times as
long as a test in another region over a distance of 10 meters which has the
same head drop, provided the effective porosities and hydraulic conductivi-
ties are identical.
Sorption of Tracers and Related Phenomena
Sorption occurs when a dissolved ion or molecule becomes attached to
the surface of a solid or dissolves in the solid. Electrostatic, hydropho-
bic, and chemical forces are involved in sorption. Various types of
14
-------
land surface
Aquifer
If AL = 1000 meters
then t=
(.3)0000)'
300 days
Figure 2.4. The average travel time of ground water between two
points A and B can be estimated by means of Equation (4) where the
gradient, Ah/AL, the hydraulic conductivity, K, and the porosity,
n , are known or can be estimated closely. The value of Ah in the
illustration is the difference in the hydraulic head between points
A and B (490-480 meters).
15
-------
sorption are due to Ion exchange, induced dipole moments, hydrogen bonding,
ligand exchange, and chemical bonding. The term "sorption," as used in this
manual, includes the sum of these physical-chemical phenomena. Commonly,
two different words are used to describe the broad process of sorption.
These are adsorption, a strictly surficial phenomenon, and absorption, a
phenomenon which involves movement of material from solution to sites within
the structure of the solid phase. Most sorption processes which we will
consider are relatively fast, reversible reactions; that is, the dissolved
constituent which is sorbed from the water can be released to the water
again under favorable circumstances. Cation exchange is probably the most
familiar type of adsorption, and is a good example of reversible sorption.
Molecules of some tracers have a tendency to be sorbed on the surfaces
of solids for brief periods, after which they move off the solid and into
the water again. If the water is moving, the tracer molecules move at a
slower rate than the water molecules, because tracer molecules spend part of
their time sorbed on solids. Thus, the sorptive characteristics of a tracer
must be known in order to design meaningful tracer experiments. One equa-
tion for the relative average velocities of water, vw> and of the sorbed
species, vs, is:
(5)
in which Kj is a distribution coefficient, pb is the bulk dry density, and
n is the porosity of the material in question. Values of K^ can range from
almost zero cm3/gram to more than 1,000 cm3/gram. The higher values of K
-------
would mean that the dissolved species is going to be almost stationary in
comparison with the water.
The distribution coefficient of a tracer, Kd, is a complex function of
a number of variables including the temperature of the system, the chemical
nature of the tracer, the concentration of the tracer, and the concentra-
tions and chemical characteristics of other dissolved species in the water
within which the tracer moves. The Kd value also depends upon the total
surface area and the surface chemistry of the solids in contact with the
tracer. It may also be dependent upon the velocity of the water moving past
the solid surfaces. Generally speaking, the value of Kd is lowered by
increasing the concentrations of dissolved species in the water. Solid
materials which tend to sorb material from water will tend to increase the
KJ values of aquifer material. Some natural solids with high sorptive
capacities are clay minerals, metal oxides, organic particles, certain
micas, and natural zeolites.
Certain tracers discussed later in this manual will be virtually unaf-
fected by sorptive processes. Those tracers are commonly called conserva-
tive tracers because their concentrations, and hence their direct relation
to the moving ground water, will be conserved if hydrodynamic dispersion is
not considered.
Although unlikely in most artificially-introduced tracer experiments,
the possibility of mineral dissolution or precipitation should always be
kept in mind. As a simple example, if the sulfate ion is used as a tracer
in water which moves through a natural bed of gypsum, dissolution of the
gypsum will undoubtedly add sulfate to the ground water and may confuse the
interpretation of the experiment.
17
-------
Hydrodynamic Dispersion and Molecular Diffusion
Two natural phenomena, hydrodynamic dispersion and molecular diffusion,
always work together to dilute the concentrations of artificially-injected
tracers. These phenomena are complex and their effects are difficult to
separate in field experiments. The two phenomena are, however, theoreti-
cally quite distinct. Hydrodynamic dispersion is produced by natural dif-
ferences in the local ground-water velocities related to the local differ-
ences in permeabilities (Figure 2.5). Molecular diffusion is produced by
differences in chemical concentrations which tend to be erased in time by
the random motion of molecules (Figure 2.6). Generally, short-term tracer
experiments in permeable material will be affected almost exclusively by
hydrodynamic dispersion. In contrast, the concentrations of natural tracers
moving very slowly in highly heterogeneous materials will be affected pro-
foundly by molecular diffusion.
The phenomena of dispersion and diffusion are discussed in greater
detail in Appendix B. A qualitative picture of the expected effects of dis-
persion, diffusion, and sorption in a simple one-dimensional flow system is
offered at the end of Chapter 3.
18
-------
General direction of water motion
A) Initial position
B)) Position after one hour
CM) Position after two hours
INITIAL
--DISTRIBUTION
OF PARTICLES
DISTRIBUTION OF
.PARTICLES AFTER
/ONE HOUR
DISTRIBUTION OF
PARTICLES AFTER
TWO HOURS
DISTANCE
Figure 2.5. Hydrodynamic dispersion is caused by unequal velocities of the
ground water. In this figure, a few molecules of a tracer are assumed to
have been released at the same time and subsequently carried by the ground
water towards the right side of the diagram. The bottom graph shows the
general distribution of molecules after one hour and after two hours. Only
longitudinal dispersion is shown on this two-dimensional diagram.
19
-------
-spot of dye
-soaked blotter-
(no water movement)
initial conditions
one hour
three hours
one day
Movement by molecular diffusion
Figure 2.6. The movement of dissolved material by molecular diffu-
sion can be seen in a blotter saturated with water. A small dot of
dye will move radially outward in the saturated blotter. If the
blotter is horizontal, the radial movement of the dye is by molecu-
lar diffusion.
20
-------
CHAPTER THREE
PRACTICAL ASPECTS
Planning a Test
The purpose and practical constraints of a potential tracer test must
be understood clearly prior to actual planning of tracer tests (see Appendix
C). Is only the direction of water flow to be determined, or are other
parameters such as travel time, porosity, and hydraulic conductivity of
interest? How much time is available for the test? If answers must be
obtained within a few weeks, then tracer tests using only the natural
hydraulic gradient between two wells which are more than about 20 meters
apart would normally be out of the question because of the long time period
needed for the tracer to flow between the wells. Another primary considera-
tion is the budget. If several deep holes are to be drilled, if packers are
to be set to control sampling or injection, and if hundreds of samples must
be analyzed in an EPA certified laboratory, then total costs could easily
exceed a million dollars. In contrast, some short-term tracer tests may be
possible at costs of less than a thousand dollars.
The initial step in determining the physical feasibility of a tracer
test is to collect as much hydrogeologic information as possible concerning
the field area. The logs of the wells at the site to be tested, or logs of
the wells closest to the proposed site, should be obtained. Logs will give
some idea of the homogeneity of the aquifer, layers present, fracture pat-
terns, porosity, and boundaries of the flow system. Local or regional piez-
ometric maps, or any published reports on the hydrology of the area (includ-
ing results of aquifer tests) are valuable, as they may give an indication
of the hydraulic gradient and hydraulic conductivity.
21
-------
The hydrogeologic information is used to estimate the direction and
magnitude of the ground-water velocity in the vicinity of the study area
(Fetter, 1981). One method to arrive at a local velocity estimate is the use
of water-level maps together with Darcy's Law, if transmissivity, aquifer
thickness, and head values are available (see Chapter 2). The second method
involves using a central well with satellite boreholes, and running a prelim-
inary tracer test. The classical method for determining the regional flow
direction is to drill three boreholes at extremities of a triangle, with the
sides 100-200 meters apart (Figure 3.1). The water levels are measured and
the line of highest slope gives the direction of flow. However, regional
flow is generally not as important as local flow in most tracer tests, and
the importance of having an accurate flow direction cannot be overemphasized.
Caspar and Oncescu (1972) described a method to determine local flow direc-
tion by drilling 5-6 satellite wells in the general direction of flow. They
noted that the satellite boreholes should be at a minimum distance of 8x the
well diameter from the injection well. The boreholes should be screened and
gravel packed to avoid well-bore effects. Commonly, the boreholes are 2-3
meters from the central well. The advantage of knowing the general flow
direction is that fewer observation wells will eventually be drilled. If a
preliminary value of the magnitude of the natural velocity of the aquifer is
available, then the injection or pumping rate necessary to obtain radial flow
can be determined. Also, when a velocity magnitude is obtained from the pre-
liminary test or available data, a decision as to the distance from the
injection well to observation well(s) can be made. This decision depends on
whether the test is a natural flow or induced flow (injection or pumping)
type test. Natural flow tests are less common due to the greater amount of
time involved.
22
-------
Water level elevation • 391 m
Observation Well *M
NJ
OJ
Area of
proposed
tracer test
Water level elevation « 382 m
Observation Well # 3
Water level elevation • 387
Observation Well # 2
Figure 3.1. For tests using artificially-injected tracers which will flow by a natural ground-water
gradient from one well to another, it is essential to know the direction of ground-water flow prior
to the final design of the tests. One method to estimate this direction is to construct a local
water-level map near the site of the test and assume that the flow is going to be perpendicular to
the lines of equal water elevation. The minimum number of wells needed for the water-level map is
three, as shown in this hypothetical example.
-------
A second major consideration when planning a test is which tracers are
the best for the conditions at the site and the objectives of the test. Sam-
ples of well water should be analyzed for background values of relevant
parameters, such as temperature, major ions, natural fluorescence, fluoro-
carbons, etc. Choice of a tracer will depend partially on which analytical
techniques are easily available (see Appendix E) and which background con-
stituents might interfere with these analyses. Various analytical techniques
incorporate different interferences, and consultation with the chemist or
technician who will analyze the samples is necessary.
Determination of the amount of tracer to inject is based on the natural
background concentrations detection limit for the tracer and the dilution
expected (Figure 3.2). If a value for porosity can be estimated, the volume
of voids in the medium can be calculated as the volume of a cylinder with one
well at the center and the other a distance away. Adsorption, ion exchange,
and dispersion will decrease the amount of tracer arriving at the observation
well, but recovery is usually not less than 20% (of the injected mass) for
two-hole tests using a forced recirculation system and conservative tracers.
The concentration should not be increased so much that density effects become
a problem. Lenda and Zuber (1970) gave graphs which can be used to estimate
the approximate quantity of tracer needed. The values are based on estimates
of the porosity and dispersion coefficient of the aquifer.
Types of Tracer Tests
The variety of tracer tests is almost infinite when one considers the
various combinations of tracer types, local hydrologic conditions, injection
methods, sampling methods, and the geological setting of the site (Appendix
24
-------
100
IO
1.0
Mg/1
Injected 4/25
nlected 4/27
AMINO 6 ACID
RHODAMINE WT
Injection well
to r
i.o -
O.OOI
4/25 5/1 5/10 5/20 6/1 6/IO 6/20 7/1 7/10 7/2O
DATE
Figure 3.2. Results of tracer tests at the Sand Ridge State Forest,
Illinois. The aquifer was a fine to medium coarse dune sand in the
upper part and a medium to coarse sand in the lower part. Three in-
jection wells 3 ft apart were used to make 3 separate injections of
Lissamine FF (green dye), Amino G Acid (blue, optical brightener),
and Rhodamine WT (orange dye). A slug having a uniform concentration
of ]00 mg/1 was used. Lissamine was not detected in any of the obser-
vation wells during the duration of the test. Dilution in the injec-
tion wells and movement of the dye was entirely by ground water flowing
under a natural gradient of 1.5 x 10~3. Variations of shapes of break-
through curves are caused by heterogeneities in the aquifer. Note the
ten-thousand fold decrease of concentration of Rhodamine produced by
only 50 ft of flow. (Data from Naymik and Sievers, 1983).
25
-------
C). Some of these varieties are shown in Figure 3.3. The following sections
discuss a few of the more common types of tracer tests.
Differences in the tests are due to the parameters (such as velocity,
dispersion coefficient, and porosity) which are to be determined, the scale
of the test, and the number of wells to be used.
Single-Well Techniques
Two techniques, injection/withdrawal and point dilution, give values of
parameters which are valid at a local scale. Advantages of single-well tech-
niques are: (1) less tracer is required than for two-well tests; (2) the
assumption of radial flow is generally valid so natural aquifer velocity can
be ignored, making solutions easier; and (3) knowledge of the exact direction
of flow is not necessary.
Injection/Withdrawal
The single-well injection/withdrawal (or pulse) technique results in a
value of pore velocity and the longitudinal dispersion coefficient. The
method assumes that porosity is known or can be estimated with reasonable
accuracy. A given quantity of tracer is instantaneously added to the bore-
hole, the tracer is mixed, and then 2-3 borehole volumes of fresh water are
pumped in to force the tracer to penetrate the aquifer. Only a small quan-
tity is injected so that natural flow is not disturbed. After a certain
time, t, the tracer has traveled a distance X, due to uniform flow. Then the
borehole is pumped out at a constant rate which is large enough to overcome
the natural ground-water flow. Tracer concentration is measured with time or
pumped volume. This enables one to find the distance traveled, X, by the
relationship:
26
-------
Figure 3.3. A number of common configurations for ground-water
tracing by the use of artificially-injected tracers are shown in
the following diagrams. Although single tracers are shown in
most of the diagrams, most tests can use more than one tracer.
Also, the purposes are varied and only the most obvious ones are
mentioned. Sampling of the initial mixture of the tracer and
water prior to injection is not shown but is almost always re-
quired if quantitative results are to be obtained.
27
-------
sink
sampling point
spring
Fractured rock
&;.• Tracer
(O)
Determine if trash in sinkhole contributes to contamination of spring.
sampling point
cave stream
71
(b.)
Measure velocity of water in cave stream.
28
-------
Tj sinking stream
sampling point
CcJ
Check source of water at rise in stream bed.
sampling point
(d.)
Determine if tile drain from septic tank contributes to contamination
of well.
29
-------
three
different
tracers
waste-water
lagoon
yy.M .......,«. ^rn n nun!
' *
' toilet
landfill
sampling poinf
Determine source of pollution from three possibilities.
_
_—jwater table
, sampling point
IlllIB
mm
nun
nun
Him
mm
nun
* f / // / // / / / //
(f.)
Determine velocity and direction of ground-water flow under natural
conditions. Injection followed by sampling from same well.
-------
/// //TV// /////'
water table
• zone of injected tracer
(gj
Test precipitation of selected constituents on the aquifer material
by injecting multiple tracers into aquifer then pumping back the
injected water.
^M>*
1
sampling point
77 j in i
_ '»
TTTTTTTTTT
water
" • .".'.'•
^^v>
7 / / / 1 1 1 / /
table
(h.)
Test velocity of movement of dissolved material under natural
ground-water gradients.
-------
multi-level
sampling
• T
1 t
////////// ///
'. •"•••"..
:••:••••::••
• • ••.«.
•• • • . . •
. • • . . • •
.• • . • • *
i
''*//// //
^^^ ^"^^* ^^^M
~*S
• 1
• .... 1
• • ' . .
« * * 1
. « »
^^•^^^^^^
•
777777777777777777"
J^ater^ fable
(i.)
Test hydrodynamic dispersion in aquifer under natural ground-water
gradients.
sampling
point
pumped
well
Itafi
injection
well
/////////////////
////// ////))} f 1
(j.)
Test a number of aquifer parameters using a pair of wells with forced
circulation between wells.
32
-------
sampling
point
packers
| \y I} )n i >) it 7-
\ ^,x- fractured
granite
packers
uncased
holes
(k.)
Determine the interconnect fractures between two uncased holes.
Packers are inflated with air and can be positioned as desired in
the holes.
33
-------
sampling points
water
table
(1.)
Determine the direction and velocity of natural ground-water flow by
drilling an array of sampling wells around a tracer injection well.
4 sampling point
at pumping well
till i 11 j f urrrn
J
If 11111' i ill II'
(m.)
Verify connection between surface water and well
-------
?rbn
where VSQ = volume pumped to recover 50% of the mass injected;
b = aquifer thickness; and
n = porosity.
Average velocity is then —, where t is time from initial injection
to the time when pumping started. If concentration is measured at various
depths with point samplers, relative permeability of layers can be deter-
mined. The dispersion coefficient is obtained by matching experimental
breakthrough curves with theoretical curves based on the general dispersion
equation. A finite difference method is used to simulate the theoretical
curves (Fried, 1975). Some assumptions of the theory are homogeneous, hori-
zontal, and independent strata. Fried concluded that the method is useful
for local information (2-4 meters) and for detecting the most permeable
strata. An advantage of this test is that nearly all of the tracer is
removed from the aquifer at the end of the test.
Borehole Dilution
Borehole dilution techniques are also described in Chapter 4 under radi-
oactive tracers. This technique can be used to measure the magnitude and
direction of horizontal tracer velocity and vertical flow. Also, hydraulic
conductivity values can be obtained by applying Darcy's law.
The procedure is to introduce a known quantity of tracer instantaneously
into the borehole, mix it well, and then to measure the concentration
decrease with time. The equation used to determine velocity is
35
-------
r In (C /c)
V = -
4 tn
where r = borehole radius;
t = time of observation; and
n = effective porosity.
Often a correction term for distortion of flow due to the borehole is
added. The tracer is generally introduced into an isolated volume of the
borehole using packers. Radioactive tracers have been used frequently for
borehole dilution tests, but other tracers can be used.
The lower limit of the aquifer velocity for use of this method is
V = 0.01 m/a, due to diffusion. The upper limit is a few hundred meters per
day because flow is no longer laminar. Other assumptions related to this
technique are:
(1) Borehole dimensions are well known.
(2) Measurements are taken after steady flow has been estab-
lished (well screen does not alter flow).
(3) If possible, borehole construction should be such that
vertical flow is not present.
(4) If the borehole is screened, the gravel pack should be
homogeneous with respect to permeability. Also, the
screen and gravel pack should be arranged concentrically
within the borehole.
Other factors to keep in mind when conducting a point dilution test
are the homogeneity of the aquifer, effects of drilling (mudcake, etc.),
homogeneity of the mixture of the tracer and the well water, degree of
tracer diffusion, and density effects. A number of methods are available to
correct for well construction, vertical currents, and other factors (Caspar
and Oncescu, 1972).
36
-------
The ideal condition for conducting the test is to use a borehole with
no screen or gravel pack. If a screen is used, it should be next to the
borehole as dead space alters the results. Samples should be very small in
volume so that flow is not disturbed by its removal.
The time versus concentration curve will be linear in a middle section
of the plot. Velocity determinations are reasonably accurate if the linear
region is in the area of c/co < 0.50. For more information on this type of
test, see Caspar and Oncescu (1972), Fried (1975), and Klotz et al. (1978).
The direction of ground-water flow can be measured in a single borehole
by a method similar to point dilution. A tracer (often radioactive) is
introduced slowly and without mixing. A section of the borehole is usually
isolated by packers. After some time, a compartmental sampler (4-8 compart-
ments) within the borehole is opened. The direction of minimum concentration
corresponds to the flow direction. Another similar method is introduction of
a radioactive tracer and subsequent measurement of its adsorption on the
borehole or well screen walls by means of a counting device in the hole. The
method is described in more detail in Caspar and Oncescu (1972).
Two-Well Techniques
These methods consist of two types, uniform (natural) flow and radial
flow tests. The parameters measured (dispersion coefficient and porosity)
are assumed to be the same for both types of flow.
Uniform Flow
A tracer is placed in one well without disturbing the flow field and
a signal is measured at observation wells. This test can be used at a local
(2-5 m) or intermediate (5-100 m) scale, but the time involved in the test
is much larger than that related to radial tests. The direction and
37
-------
magnitude of the velocity must be known quite precisely, or a large number of
observation wells are needed. The quantity of tracer needed to cover a large
distance can be expensive. On a regional scale, environmental tracers are
generally used, including seawater intrusion, radionuclides, or stable iso-
topes of hydrogen and oxygen. Man-made pollution has also been used. For
regional problems, a mathematical model is calibrated with concentration ver-
sus time curves from field data, and the same model is used to predict future
concentration distributions.
Analysis of local or intermediate scale uniform flow problems can be
done analytically, semi-analytically, or by curve-matching. Layers of dif-
ferent permeability can cause distorted breakthrough curves, which can usual-
ly be analyzed (Caspar and Oncescu, 1972). One- or two-dimensional models
are available. Analytical solutions can be found in Fried (1975) and Lenda
and Zuber (1970).
Radial Flow
These techniques are based on imposing a velocity on the aquifer, and
generally solutions are easier if radial flow is much greater than uniform
flow. A value for natural ground-water velocity is not obtained, but poros-
ity and the dispersion coefficient are obtained.
A diverging test involves constant injection of water into an aquifer
with a slug or continuous flow of tracer introduced instantaneously into the
injected water. The tracer is detected at an observation well which is not
pumping. Very small samples are taken at the observation well so that flow
is not disturbed. Packers can be used in the injection well to isolate an
interval. Sampling can be done with point samplers or an integrated sample
can be taken.
38
-------
Converging tests involve introduction of the tracer at an observation
well, and another well is pumped. Concentrations are monitored at the pumped
well. The tracer is often injected between two packers or below one packer,
and then 2 to 3 well bore volumes are injected to push the tracer out into
the aquifer. At the pumping well, intervals of interest are isolated (par-
ticularly in fractured rock), or an integrated sample is obtained.
A recirculating test is similar to a converging test, but the pumped
water is injected back into the injection well. This tests a significantly
greater part of the formation because the wells inject to and pump from 360
degrees. The flow lines are longer, partially canceling out the advantage of
a higher gradient. Theoretical curves are available for recirculating tests
(see Sauty, 1980).
Design and Construction of Test Wells
In many tracer tests, the construction of test wells is the single most
expensive part of the work. It also can be the source of major difficulty
if the construction is not done properly. Several texts cover the general
details of drilling technology and well construction (California Department
of Water Resources, 1968; Campbell and Lehr, 1973; Johnson Division, UOP,
Inc., 1972; Todd, 1980) and therefore they are not discussed in this manual.
Five common types of problems are encountered with tracer tests. The
first problem relates to site selection. If heavy equipment is to be moved
into an area, lack of overhead clearance, narrow roads, poor bearing capaci-
ties of bridges, and the lack of flat ground at the site can all be major
problems. Also, overhead electrical power lines at the site should be
avoided. One of the most common hazards is accidental grounding of power
39
-------
lines by drill rigs and auger stems with subsequent electrocution of
workers.
The second problem relates to the improper choice of drilling equipment.
For some purposes, cheap systems using hand augers and drive points are suit-
able to install wells for shallow tracer tests (Figure 3.4). To be sure, a
large drilling rig could be moved into the site to do the same job, but with
at least a ten-fold increase in cost which would be a major misuse of funds.
The error is commonly the other direction, however, with attempts to hold
down the cost resulting in the use of drilling equipment which is unable to
handle the needs of the project. Another general problem relating to drill-
ing is the use of drilling fluids which will affect the tracer tests. Cer-
tain drilling muds and mud additives have a very high capacity for the sorp-
tion of most types of tracers. The muds could also clog small pores and
alter the permeability of the aquifer near the drill hole. The use of com-
pressed air for drilling may avoid some of these problems but it could intro-
duce atmospheric fluorocarbons which could interfere with tracer tests using
fluorocarbons.
A third problem is the choice of casing diameter. Ideally, packers
should be used to isolate the zones being sampled from the rest of the water
in the well (Figure 3.5). For a number of reasons which include economics,
insufficient time, and lack of technical training, packers are often not used
in tracer tests. In this case, the diameter of the sampling well should be
as small as possible in order to minimize the amount of "dead" water in the
well during sampling. The diameter, however, cannot be too small because the
well must be adequately cleaned after installation and the well must accommo-
date bailers, pumps, or other sampling equipment. Common casing diameters
40
-------
'
Figure 3.4a. Well installation for a simple tracer test.
Two home-made drive points of iron pipe crimped at the ends
and perforated by drilled holes are on the left and two
standard wire-wound commercially available drive points are
on the right. Extension pipe is screwed onto the ends of
the drive points in order to reach desired depths. Unless
special jetting equipment is used, drive points can usually
penetrate only 20 to 30 feet of alluvium.
-------
-**» * .
Figure 3.4b. Hammer used to install drive points. The hammer
which is shown in this picture is a hollow weighted tube with
one closed end and side handles. If alluvium has coarse gravel
or cobbles, home-made drive points will collapse easily.
42
-------
: v
•a*.
Figure 3.4c. Tracer water being injected into a shallow
test hole. Instrument is a thermistor thermometer.
43
-------
SOLUTE TRANSPORT TESTING SYSTEM
TRACER INJECTION SYSTEM
TRACER SAMPLING SYSTEM
HIGH FLOW
RATE INJECTION
PANEL
High
Flow Rota
Rftcirciilgf
TRACER SUPPLY
LOW FLOW
RATE INJECTION
PANEL
Low
Flow
Tracer
FRACTURED ROCK AQUIFER
ELECTRONIC DATA
LOGGING
EQUIPMENT
HOSE
REEL
SAMPLING
SYSTEM
CONTROL
UNIT
GROUND LEVEL
PRESSURE
TRANSDUCER
HOUSING
UPPER
PACKER
CONNECTING
PIPE
DEAD VOLUME
SECTION
TRACER INJECTION
NOZZLES
CONNECTING PIPE
LOWER
PACKER
MIXING PUMP
SAMPLE PUMP
MIXING NOZZLES
Culler.,1984
Figure 3.5. (a) Schematic diagram for a two-well tracer test in
fractured rock completed at Oracle, Arizona, by the Department of
Hydrology and Water Resources, University of Arizona. The follow-
ing photograph shows the control panel for the high-flow rate
injection with storage tanks for four different tracers arranged
on top of the panel box. This is an example of a tracer test which
is more complex than commonly attempted for practical applications.
Diagram, courtesy of James Cullen.
-------
Figure 3.5. (b) High flow rate injection panel shown in
diagram in Figure 3.5a.
45
-------
used range from about 1" to 4" for relatively shallow test holes to as much
as 6" to 8" for very deep tests.
The type of casing to be used is a fourth concern primarily if low-level
concentrations of tracers are to be used, and in particular if these tracers
are organic compounds or metallic cations. For plastic casings, TEFLON
absorbs and releases less organics than does PVC. Adhesives used to connect
sections of plastic pipes may be also a troublesome source of interferring
organic compounds. Metal casing could release trace metals but it is gener-
ally superior to plastic casing in terms of strength and sorptive character-
istics. Inexpensive metal casing, however, will have a short life if ground
waters are corrosive.
A fifth problem is that of the choice of filter construction for the
wells which depends on the aquifer and the type of test to be completed. If
the aquifer being tested is a very permeable coarse gravel and if the casing
diameter is small, then numerous holes drilled in the solid casing may.be
adequate. In contrast, for a single-well test with an alternating cycle of
injection and pumping of large volumes of water into and out of loose, fine-
grained sand, an expensive well screen with a carefully placed gravel pack
may be required. Regardless of the type of filter used, it is absolutely
essential that the casing perforations, gravel pack, or screen as well as the
aquifer at the well be cleaned of silt, clay, drilling mud, and other mate-
rial which would prevent the free movement of water in and out of the well.
This process of cleaning or development is so critical that it should be spe-
cified in clear terms in any contract related to well construction.
46
-------
Injection and Sample Collection
Injection equipment depends on the depth of the borehole and the funds
available. In very shallow holes, the tracer can be lowered through a tube,
placed in an ampule, which is lowered into the hole and broken, or just
poured in. Mixing is desirable and important for most types of tests and is
simple for very shallow holes. For example, a plunger can be surged up and
down in the hole or the release of the tracer can be through a pipe with many
perforations. Flanges on the outer part of the pipe will allow the tracer to
be mixed by raising and lowering the pipe. For deeper holes, tracers must be
injected under pressure and equipment can be quite sophisticated. Figure 3.5
is an example of a high-pressure injection system. The interval of interest
in the borehole is usually packed off. This equipment is often custom built
for a specific experiment, as tracer injection systems for water wells are
not yet available commercially. As mentioned before, instantaneous injection
is the ideal condition. For a pulse test, this may mean an injection period
of a minute or an hour, depending on the equipment. The equipment shown in
Figure 3.5 is described in detail in Simpson et al. (1983) with details of
work conducted in fractured rock by the Department of Hydrology at the Uni-
versity of Arizona.
Sample collection can also be simple or sophisticated. For tracing
thermal pulses, only a thermistor needs to be lowered into the ground water
(Figure 3.6). For chemical tracers at shallow depths, a hand pump may be
sufficient. Bailers can also be used, but they mix the tracer in the bore-
hole which, for some purposes, should be avoided. A TEFLON bottom-loading
bailer is described in Buss and Bundt (1981). It may be desirable to clear
the borehole before taking a sample, in which case a gas-drive pump can be
47
-------
Figure 3.6a. Small digital thermometer with thermistor
line in observation well. A thermal pulse produced by
injecting warm water is being measured at this point.
48
-------
Figure 3.6b. Recording data from test. Although thermistor
signals are easily recorded automatically, hand recording is
satisfactory for many low-budget, short-term tests.
49
-------
used to evacuate the well. For a nonpumping system, the decision as to how
much water must be withdrawn from a borehole in order to obtain a sample
which is representative of the water adjacent to the borehole is not a
trivial problem. If not enough water is taken out, the sample composition
will be influenced by semistatic water which will normally fill much of the
well. If too much water is drawn out, a gradient towards the well will be
created and the natural movement of the tracer will be distorted. A common
rule of thumb is to pump out four times the volume of water which is in the
well before the sample is taken (see Figure 3.7).
If existing wells which have been drilled for water-supply purposes are
used for tracer tests, extreme care is required because of the complex rela-
tionship among such variables as pumping rates, patterns of water circulation
within the well, and the yields of different parts of the aquifers which are
penetrated. This complexity is reflected commonly in the variability of
water chemistry as a well is being pumped (Keith et al., 1982; Schmidt,
1977). Stated simply, for wells drawing water from complex aquifers or a
series of aquifers, an analysis of a single water sample taken at a given
point in time cannot yield definitive information about the water chemistry
of any individual zone.
Many systems for sampling in wells have been described in recent years.
Ground Water Monitoring Review is a good source of current techniques.
Multi-level samplers are described in this journal by Cherry and Johnson
(1982) and Pickens et al. (1981). For more information on gas-driven and
positive displacement sampling devices, see Robin et al. (1982), Morrison
and Brewer (1981), and Gillham and Johnson (1981).
50
-------
Zinc (February)
Zinc (August)
2468
WELL VOLUMES PUMPED
Figure 3.7. A difficult problem in field tests is to obtain
a representative sample from an open test hole. Results of
the analyses of successive samples taken from a small test hole
are shown in this diagram which show that useful samples appear
to be obtained after pumping 4 well volumes out of the hole.
However, the number of well volumes needed varies with the hy-
draulic gradient, the well construction, the permeability of the
zone being sampled, the type of tracer used, and the volume of
water initially in the well. Diagram is adapted from Gibb,
Sehuller, and Griffin (1981).
51
-------
The preservation and analysis of samples is covered in Chapter 4 and
Appendix C. Keith et al. (1982) also cover some of the practical problems
involved with sample collection, analyses, and quality control.
Interpretation of Results
The following remarks and figures are intended only as a brief qualita-
tive introduction to the interpretation of the results of tracer tests. More
extensive and quantitative treatments are found in the works of such authors
as Halevy and Nir (1962), Theis (1963), Fried (1975), Custodio (1976), Sauty
(1978), Grisak and Pickens (1981), and Gelhar (1982).
The basic plot of the concentration of a tracer as a function of time
or water volume passed through the system is called a breakthrough curve.
The concentration is either plotted as the actual concentration (Figure 3.2)
or, quite commonly, as the ratio of the measured tracer concentration at the
sampling point, C, to the input tracer concentration, CQ (Figure 3.8).
The measured quantity which is fundamental for most tracer tests is the
first arrival time of the tracer as it goes from an injection point to a
sampling point. The first arrival time conveys at least two bits of informa-
tion. First, it indicates that a connection for ground-water flow actually
exists between the two points. For many tracer tests, particularly in karst
regions, this is all the information which is desired. Second, an approxima-
tion of the maximum velocity of ground-water flow between the two points may
be obtained if the tracer used is conservative.
Interpretations more elaborate than the two simple ones mentioned
depend very much on the type of aquifer being tested, the velocity of ground-
water flow, the configuration of the tracer injection and sampling systems,
and the type of tracer or mixture of tracers used in the test.
52
-------
ditch filled with tracer
having a concentration of
sampling well with water
having a tracer concentration of C
/
tracer front
(a.)
c/c
0
1.0
0.5
0.0
time of first
arrival
^tirne of maximum
rate of change of C
•> time
B
(b.)
Figure 3.8(a). Ditch into which a tracer is injected continuously
and mixed with the water in the ditch to produce water with an ini-
tial fixed tracer concentration of CQ. The arrival of the tracer
front is studied by taking samples from the well that is downgradi-
ent from the ditch. (b) The breakthrough curve obtained from in-
jecting tracers into the ditch.
53
-------
Next after the first arrival time, the most interest is commonly centered on
the arrival time of the peak concentration for a slug injection, or for a
continuous feed of tracers, the time since injection when the concentration
of the tracer changes most rapidly as a function of time (Figure 3.8). In
general, if conservative tracers are used, this time is close to the theoret-
ical transit time of an average molecule of ground water traveling between
the two points. The "spread" of the curve is also of interest, and can be
related to the combined effects of hydrodynamic dispersion and molecular dif-
fusion (Figure 3.9).
If a tracer is being introduced continuously into a ditch penetrating
an aquifer as shown in Figure 3.8, then the ratio C/CQ will approach 1.0
after the tracer starts to pass the sampling point. The ratio of 1.0 is
rarely approached in most tracer tests in the field, however, because waters
are mixed by dispersion and diffusion in the aquifer and because wells used
for sampling will commonly intercept far more ground water than has been
tagged by tracers (Figure 3.10). Ratios of C/CO in the range of between
10~5 and 2 x 10"1 are often reported from field tests.
If a tracer is introduced passively into an aquifer but is recovered by
pumping a separate sampling well, then various mixtures of the tracer and
the native ground water will be recovered depending on the amount of water
pumped, the transmissivity of the aquifer, the slope of the water table,
and the shape of the tracer plume. Keely (1984) has presented this problem
graphically with regards to the removal of contaminated water from an
aquifer.
With an introduction of a mixture of tracers, possible interactions
between the tracers and the solid part of the aquifer may be studied. If
54
-------
.0
0.5
0.0
time
Figure 3.9. Breakthrough curves _a and b_ were obtained from
tests in two different media. Test a_ shows only a moderate
amount of dispersion while test b_ shows a rather high amount
of dispersion. Tests _§_ and c_ were conducted at the same time
in the same material but with different tracers. The displace-
ment of the test curve _c to the right of the diagram is caused
by sorption of the tracer on the solid material in contact with
the water.
55
-------
ditch filled with tracer
which supplies 1/4 of
downgradient ground-
water flow.
sampling well
/ f ////////// //////// ' f s
(a.)
'0
0.50^
0.25
0.00
time
(b.)
Figure 3.10. Most tracer tests do not fully saturate the
aquifer with the tracer being injected. This is shown in
diagram (a). The resulting breakthrough curve, diagram (b),
therefore, will never increase to the C/C value of 1.0.
o
56
-------
interactions take place, they can be detected by comparing breakthrough
curves of a conservative tracer with the curves of the other tracers being
tested (Figure 3.11). A common strategy for these types of tracer tests is
to inject and subsequently remove the water containing mixed tracers from a
single well. If injection is rapid and pumping to remove the tracer follows
immediately, then a recovery of almost all the injected conservative tracer
is possible. If the pumping is delayed, the injected tracer will drift
downgradient with the general flow of the ground water, and the percentage
of the recovery of the conservative tracer will be less as time progresses.
Successive tests using longer delay times between injection and pumping can
then be used to estimate ground-water velocities in permeable aquifers with
moderately large hydraulic gradients.
The methods of quantitative analyses of tracer breakthrough curves are
generally by curve-matching of computer-generated type curves, or by analyt-
ical methods. Analytical methods are covered in Fried (1975). Sauty (1978)
provided solutions for solute transport for different flow fields (linear
and radial) and for diverging and converging conditions. He covered contin-
uous and slug injection. Sauty (1978) presented a finite difference method
to be used with converging and diverging problems; the program is called
RAMSES. Carrera and Walter (1985, manuscript in preparation) developed a
similar, more accurate program called CONFLO for use in converging problem.
An example of a type curve is given in Figure 3.12. The match can be
done by eye or by computer.
57
-------
0.05 -
0.00
time
Figure 3.11. In this hypothetical diagram, four different tracers
are mixed and injected as a single slug into an aquifer. As can be
seen in the resulting breakthrough curves, tracer a_ is conservative,
tracer b_ shows some effect of sorption on the aquifer, tracer c_
shows a large effect of sorption, and tracer d_ is precipitated or
destroyed before a significant amount reaches the sampling point.
The destruction can be by radioactive decay, by chemical decompo-
sition, or by the metabolic action of microorganisms.
58
-------
Figure 3.12. Computer-generated type curves are used for a two-well test in
which one well is used for injection and the other for sampling to find dis-
persivity (a) and porosity (<)>)• The vertical axis is dimensionless concen-
tration, defined as the following:
cD
where irr2b = volume of the cylinder defined by the injection and
withdrawal of wells;
b = thickness of aquifer;
r = distance between wells;
C = measured concentration at time t; and
M = mass of tracer injected during the test.
The horizontal axis is reduced time, defined as:
where t = time of sampling;
Q = pumping rate.
When analyzing a test, the tracer test results are plotted as log C versus
log t on the vertical and horizontal axes, respectively. The experimental
curve is matched with a type curve, keeping axes parallel. From the match
curve, the Peclet number is found. The Peclet number (Pe) is equal to r/a,
so the dispersivity is obtained. Next, a match point is chosen for any
point on both curves. The equation for reduced time is used, and all values
except are known. Then,
Q =
m:2btR
(10)
To verify the validity of the method, the dimensionless concentration equa-
tion is used. From the matchpoint, C and CD are known. If CD = (irr b<|>C)/m,
the method has been verified. These type curves were developed by Hydro-
GeoChem, Inc., of Tucson, Arizona, 1984.
59
-------
I I I I I
3 456789
10
Reduced Time (Log Scale)
60
-------
CHAPTER FOUR
TYPES OF TRACERS
In this chapter, information is presented concerning various types of
tracers, including water temperature, solid particles (yeast, bacteria,
spores, etc.)> ions, organic acids, dyes, and radioactive tracers. The
final section of the chapter deals with environmental tracers, such as
stable isotopes and radionuclides. Each tracer type will be discussed
regarding its applicability in different hydrologic settings, the field
methods used (necessary equipment and sampling techniques), and type of
detection used. Additional information (interpretation of results, cost of
the tracer, and environmental and health concerns) is presented at the con-
clusion of each subsection.
Temperature
The temperature of water changes slowly as it migrates through the sub-
surface, because water has a high specific heat capacity compared to most
natural materials. For example, temperature anomalies associated with the
spreading of warm wastewater in the Hanford Reservation in south-central
Washington have been detected more than 8 km (5 miles) from the source (U.S.
Research and Development Adm., 1975).
Water temperature is a potentially useful tracer, although it has not
been used frequently. The method should be applicable in granular media,
fractured rock, or karst regions. Keys and Brown (1978) traced thermal
pulses resulting from the artificial recharge of playa lake water into the
Ogallala Formation in Texas. They described the use of temperature logs
(temperature measurements at intervals in cased holes) as a means of
61
-------
detecting hydraulic conductivity differences in an aquifer. Temperature
logs have also been used to determine vertical movement of water in a bore-
hole (Keys and MacCary, 1971; Sorey, 1971).
Heat is transmitted by convection (transport of heat by fluid flow) and
conduction (due to temperature gradients within the saturated material).
Assuming that convection dominates, Keys and Brown (1978) demonstrated the
use of a simple temperature model to estimate a ratio of 1:3 for the veloc-
ity of the temperature pulse compared to the water velocity in a granular
material. The actual ratio depends on aquifer porosity, density of the
aquifer material and of water, and the heat capacity of the aquifer material
and of water. They concluded that the actual relationship between the rate
of transmission of a thermal wave in an aquifer and the velocity of water
was unknown. However, water most certainly has a higher velocity than the
temperature pulse.
Laboratory column tests have been performed to compare the travel times
of chloride, yeast, and temperature (Keys and Brown, 1978). The chloride
concentration began to increase at 0.8 pore volumes and reached input con-
centration at 1.2 pore volumes. The yeast began to increase at 0.95 pore
volumes and reached input concentration at 1.25 pore volumes. Temperature
began to rise at 0.7 pore volumes and reached input temperature at 3.25 pore
volumes. The heat traveled faster than the other tracers as far as initial
detection is concerned, but the center of mass of the thermal pulse arrived
later than the chloride or yeast. This illustrates the point that changes
in water temperature are accompanied by changes in density and viscosity of
the water. This, in turn, alters the velocity and direction of flow of the
water. For example, injected ground water with a temperature of 40°C will
62
-------
travel more than twice as fast in the same aquifer under the same hydraulic
gradient as will water at 5°C. Because the warm water has a slightly lower
density than cold water, buoyant forces give rise to flow which "floats" on
top of the cold water. In order to minimize problems of temperature-induced
convection, small temperature differences with very accurate temperature
measurements should be used if hot or cold water is the introduced tracer.
Temperature was used as a tracer for small-scale field tests, using
shallow drive-point wells two feet apart in an alluvial aquifer. The
transit time of the peak temperature was about 107 minutes, while the
resistivity data indicate a travel time of about 120 minutes (see Figure
4.1). The injected water had a temperature of 38°C, while the ground-water
temperature was 20°C. The peak temperature obtained in the observation well
was 27°C.
In these tests, temperature served as an indicator of breakthrough of
the chemical tracers, aiding in the timing of sampling. It was also useful
as a simple, inexpensive tracer for determining the correct placement of
sampling wells.
The use of cold water as an injected tracer was attempted by Simpson
(personal communication, 1984). Icicles of water containing I131 were
deposited In a borehole penetrating an alluvial aquifer. No temperature or
radiation change was detected at sampling points, while breakthrough did
occur when the liquid tracer was used. The higher density of the cooler
tracer is believed to have caused the tracer to sink and miss the sampling
points.
Another application of water-temperature tracing is the detection of
river recharge in an aquifer. Most rivers have large water temperature
63
-------
G
£-
LLi
oe
3
tr
Ul
CL
s
Ul
27.0
26.0
25.0
24.O
23.0
oo Ct
£c,.\J
21.0
i i I i i i 1 i l i l 1 l 1 1 1 — 1 I I I
A A Well 1 f"
o o Well 2 7
x x Well 3 r
e . Well 4 A
.-
Initial temperature fi
of injected fluid = 47.l°C /
/
fj, Tc=r%rf-
^=^^^(X<4^^*««»it^*i
-
l I l i I l 1 1 1 1
4-A
""A
'A
V
\
\
A
a
v
h
2
c-
Ul
a
200
175
I5O
125
100
75
50
25
:
/
i . t i i i i i i i i .1
10 20 30 40 50 60 70 80 90 100 110 I2O 130 140 150
TIME AFTER INJECTION (minutes)
Figure 4.1. Results of a field test in which hot water was injected
into a well in a shallow aquifer of coarse alluvium. The sampling
wells were perforated metal pipes driven two feet from the injection
well and arranged in a semicircle in the downstream direction. Only
Well #1 intercepted the injected water, thus establishing the local
direction of ground-water flow. Resistivity of the aquifer was mea-
sured between the injection well and the sampling wells by passing a
small electrical current under a 6v potential from the injection well
to the various sampling wells. The lowering of resistivity caused by
the hot water verifies the flow direction which was determined by
measuring water temperatures. Because the resistivity of the entire
volume of the aquifer between the wells is measured, the initial drop
in resistivity does not signal the arrival of the hot water in Well
#1; it simply indicates that the hot water was started on its way
towards Well #1.
-
-------
fluctuations in response to seasonal effects. If the river is recharging
an aquifer, the seasonal fluctuations can be detected in the ground water
adjacent to the river (Rorabaugh, 1956).
• Field Methods
One of the attractive aspects of the use of temperature as a tracer is
the relatively simple and inexpensive equipment required. Temperature is
usually measured by means of a temperature probe, which utilizes a ther-
mistor. The instrument measures resistance, which is converted to tempera-
ture electronically or manually by a calibration curve. The probes are
available with meters or with digital readouts. Recording devices can be
attached, and logs may be in analog or digital form.
Any of the tracer test configurations (Figure 3.3) could theoretically
be used. The tests performed by Keys and Brown (1978) were natural gradient
tests with pressure injection. The authors also used a natural gradient
test. A short-term test using natural gradients has not been used success-
fully for a travel distance greater than 46 meters. Plumes of warm water a
number of miles long have been documented at many places where there has
been a constant feed of warm water into an aquifer over periods of many
years.
The logging method requires movement of the probe up and down in the
borehole. An alternative is to leave the probe at a constant depth, which
yields an average travel velocity for a small interval of the aquifer above
and below the sampling point.
• Detection and Analysis
The lack of laboratory analyses and the easy means of obtaining direct
measurements in the field are advantages of using a thermal tracer. Temper-
ature can be detected in sealed pipes, while chemical, bacterial, and
65
-------
particulate tracers are generally sampled and identified after entering a
borehole from a screened segment. This makes multi-level sampling for non-
thermal tracers more difficult, and a vertical distribution of tracers is
seldom obtained.
Temperature measurements can be quite sensitive using modern equipment.
Keys and Brown (1978) used a probe with an accuracy, repeatability, and sen-
sitivity of approximately 0.02°C. With very expensive temperature detection
equipment, this performance can probably be improved by an order of magni-
tude.
• Additional Information
The velocity measurements obtained from temperature tests are generally
not equal to water velocity, as discussed previously. A conservative tracer
such as chloride could be used to determine the temperature lag for site-
specific tests. Temperature is currently most useful in obtaining relative
velocities of various zones within an aquifer.
The expenses involved in this type of test are minimal in comparison to
other tracer tests. A relatively inexpensive probe and a recording device
(if desired) are the only capital expenses. Labor is minimized due to the
lack of laboratory analyses.
Environmental effects should not be a problem in this type of test pro-
vided high quality water is used for injection. For more information on
temperature as a ground-water tracer, see Stallman (1963), Sorey (1971), and
Combarnous and Bories (1975).
Solid Particles
Solid material in suspension can be a useful tracer in areas where
water flows in large conduits, such as some basalt, limestone, or dolomite
66
-------
aquifers. Aley (1976) reported that geese, bales of hay, and wheat chaff
have been used in Missouri in karst regions. In the past decade, small
particulate tracers such as bacteria have been used successfully in porous
media*
This section of the manual will briefly describe the following particu-
late tracers: paper and simple floats, signal-emitting floats, yeast, bac-
teria, viruses, and spores.
Paper and Simple Floats
Some examples of these tracers are small bits of paper (as punched out
from computer cards, for example), or multicolored polypropylene floats.
Due to the large size of these tracers, they are useful only when flow is
through large passages. The particles must be of such a size and density as
to pass through shallow sections of flow without settling out. Because
these particulates generally float on the surface, they travel faster than
the water's mean velocity. These tracers are most useful for approximating
the flow velocity and establishing the flow path.
Dunn (1963) described the use of polypropylene floats of approximately
3/32-inch diameter and one-inch length.
• Field Methods
This type of test requires very little equipment. The tracer is intro-
duced in a sinkhole or other convenient locations and is recovered by siev-
ing water as it emerges from springs or karst openings.
• Detection
The particulates are counted manually.
• Additional Information
This method is very inexpensive. Environmental effects are minimal.
67
-------
Signal-Emitting Floats
A novel tracer is a small delayed time bomb which floats through a cave
system. When the bomb explodes, the location of the explosion is determined
by seismic methods at the surface (Arandjelovic, 1969 and 1977). W.A.
Schnitzer (1972) described karst tracing experiments in which blasting takes
place in dolines. Sound impulses are detected by microphones in adjacent
springs. The impulses can be recorded by oscillographs. Another method of
tracing underground streams is described by Lange (1972). The method utili-
lizes natural noise impulses generated by moving water. The signal is
detected on the surface by seismometers, then amplified and recorded on mag-
netic tape or on a chart recorder. It combines an acoustic tracking method
and a procedure used by seismologists to locate the foci of earthquakes.
Problems with this method include noise interference from wind, traffic, and
surface streams.
Because these methods are relatively expensive and have seldom been
used, they will not be discussed in further detail.
Yeast
The use of baker's yeast (Saccharomyces cerevisiae) as a ground-water
tracer in a sand and gravel aquifer has been reported by Wood and Ehrlich
(1978). Yeast is a single-celled fungus which is ovoid in shape. The
diameter of a yeast cell is 2 to 3 ym, which closely approximates the size
of pathogenic bacterial cells. This tracer is probably most applicable in
providing information concerning the potential movement of bacteria.
68
-------
• Field Methods
In Wood and Ehrlich's experiments, tracer tests were conducted in wells
1.50 m (5 feet) apart. However, the tracer was detected at an observation
well 7.6 m (25 feet) downgradient. The injection concentration was 16 kg
(35 pounds) of baker's yeast to 45 1 (12.8 gallons) of water.
Samples for yeast analysis can be collected in sterile bottles at regu-
lar intervals and prepared for analysis in the field.
• Detection/Sample Analysis
The samples must first be filtered through membrane filters. The fil-
ters are then placed on absorbent pads saturated with M-Green Yeast and Mold
Broth and incubated at 30°C for 36 hours. Colonies can then be counted
under low magnification. This type of analysis is fairly simple, relatively
inexpensive, and requires little specialized equipment, other than a source
of heat for incubation. A wide range of concentrations can be analyzed
because the sample can be diluted if the colonies are too numerous to count.
One advantage of yeast is negligible background levels.
• Additional Information
Wood and Ehrlich (1976) found that the yeast penetrated more than 7
meters into a sand and gravel aquifer in less than 48 hours after injection.
The relative mobilities of yeast and chloride were also compared in this
study. Yeast cells are generally mechanically filtered as they pass through
the intergranular pore space. It appears that microbial cells such as yeast
or bacteria become trapped at the soil-water interface (e.g., of an injec-
tion well), and as the mat of cells increases, it becomes a more effective
filter (Vecchioli et al., 1972). This causes the breakthrough curve to
increase to an abrupt maximum and then decrease sharply as the mat of cells
69
-------
builds up (Figure 4.2). In Wood and Ehrlich's study, the yeast arrived
faster than the conservative ionic tracers, bromide and iodide.
In this case, the time lag in the peak concentrations of bromide and
iodide was due to their flow occurring simultaneously in solution channels
and intergranular pores. The yeast cells are believed to have traveled
through the solution openings. The cells were probably filtered from sus-
pension, and did not flow through the intergranular pores. The two types of
flow would explain the abrupt peak for yeast and gradual rise and long tail
for bromide.
The tracer is very inexpensive, as is analysis. The lack of environ-
mental concerns related to this tracer is one of its advantages.
Bacteria
Bacteria are the most commonly used microbial tracers, due to their
ease of growth and simple detection. Recently, Keswick et al. (1982)
reviewed case studies of bacteria used as tracers. Some of the bacteria
which have been used successfully are Escherichia coliform (E. coli),
Streptococcus faecalis, Bacillus stearothermophilus, Serratia marcescens,
and Seratia indica. They range in size from one to ten microns and have
been used in a variety of applications.
A fecal coliform, E. coli, has been used to indicate fecal pollution at
pit latrines, septic fields, and sewage disposal sites. A "marker" such as
antibiotic resistance or H S production is necessary to distinguish the
tracer from background organisms. Hagedorn et al. (1978) and Rahe et al.
(1978) used antibiotic-resistant strains of E. coli and Streptococcus
faecalis to trace movement through a saturated soil. Bacteria movement
70
-------
24.0
0.30
• Bromide
o Yeast
60 120 180 240 300 360
TIME AFTER INJECTION (min)
Figure 4.2. Results of a two-well tracer test in an alluvial
aquifer using bromide and yeast. Although the bulk of the
yeast was probably filtered out, some particles moved through
the largest openings to produce an early breakthrough peak on
the graph. This apparent anomaly where a nonconservative par-
ticulate tracer arrives ahead of the bulk of the conservative
tracer is caused by the fact that the largest openings which
carry the particles are also the paths of highest velocities.
(Graph is redrawn from Wood and Ehrlich, 1978).
71
-------
through fractured bedrock was studied by Allen and Morrison (1973). Rippon
(1963) used a bacterial tracer for detecting water movement in an estuary,
and Wimpenny et al. (1972) used an antibiotic strain of Serratia marcescens
as a tracer in a polluted river.
• Field Methods
Most bacterial tracer tests reported in the literature are two-well
natural gradient tests. The tracer can be injected by siphoning from the
container, through Tygon tubing, to the desired depth. Injection under
pressure has been used. The wells may be relatively far apart, as Sinton
(1980) reported recovery at a distance of 920 meters.
The samples can be obtained by bailer or hand pump. Most workers place
the samples on ice and transport them to the laboratory. Samples should
then be stored at 4°C or otherwise refrigerated until analyzed.
• Detection
As mentioned previously, some type of "marker" (such as antibiotic
resistance) is necessary to distinguish tracer bacteria from background
bacteria that are almost always present.
The average time for lab analysis is one to two days. Cells are recov-
ered from the water sample by membrane filtration, and the bacteria are then
diluted by serial dilution, if necessary. The filters are incubated on
plates of agar. A normal temperature is 37°C, and the time required ranges
from 24 to 48 hours. Colonies are counted under low magnification.
Methods for growing inoculum (bacteria) for use as a tracer are
described in Ormerod (1964), Rahe et al. (1978), and Sinton (1980).
• Additional Information
In choosing a bacterial tracer, a reasonable survival rate, with no
reproduction, is desired. Some bacteria are capable of growth in aquifers,
72
-------
yielding erroneous tracer results- Another factor causing ambiguous results
is the fact that bacteria, like yeast cells, are large enough to be filtered
by some soils. Also, they may adsorb to a variety of surfaces. However, in
field tests, bacterial tracers show faster transit times than dyes (Pyle,
1981; Rahe, 1978). Figure 4.3 illustrates relative travel times.
The tracer itself is relatively inexpensive, but may be difficult to
obtain. The most obvious remedy is to culture the bacteria personally.
Analysis requires a laboratory, incubator, and microscope. Cost (if samples
are sent to a commercial laboratory) is comparable to a chemical analysis.
The greatest health concern in using these tracers is that the bacteria
must be nonpathogenic to man. Even E. coli has strains which can be patho-
genic. Davis et al. (1970) reported that Serratia marcescens may be patho-
genic.
Another concern is related to the injection of antibiotic-resistant
strains. The antibiotic resistance can be transferred to potential human
pathogens. This can be avoided by using bacteria which cannot transfer this
genetic information. As is true with most other injected tracers, permis-
sion to use bacterial tracers should be obtained from the proper federal,
state, and local health authorities.
For more information concerning bacterial tracers, see Schaub (1977),
Vecchioli (1972), Ormerod (1964), and Romero (1970).
Viruses
Animal, plant, and bacterial viruses have been recently used as ground-
water tracers. Viruses are generally much smaller than bacteria, ranging
from 0.2 to 1.0 microns (see Table 4.1). In general, human enteric viruses
cannot be used, due to disease potential. Certain vaccine strains, such as
73
-------
^^
I
0>
0.
c
3
6.0
5.0
o 4.0
O
~
O
•P
(O
3.0
£ 2.0
—
O
O
ui
1.0
f\
ro
1 I I I
A
• i
X *
* \
/r^ \
/ ' V K
• / ; \iv
• x ^^
1 ' X^**«^^_ -
" 1 ' S^c
1 '
1 i
/ x X X Rhodamine
/ I WT
i'
i i i i
X
_l
100 o»
o.
o>
80 ~
d
z
^ o
60 O
1-
£
40
UJ
«j
20 <
Q
0
X
n cc
0 IO 2O 30 40 50
TIME (hours)
Figure 4.3. A comparison of travel time in
a two—well tracer test using rhodamine WT dye
and E. coli. The E. coli, which is a particu-
late tracer, arrived slight ahead of the dye,
probably for reasons explained in connection
with Figure 4.2. (Figure redrawn from Pyle and
Thorpe, 1981).
74
-------
TABLE 4.1
Comparison of Microbial Tracers
Tracer
Bacteria
Spores
Size
(um)
1-10
25-33
Time
Required for
Assay (days)
1-2
1/2
Essential
Equipment
Required
incubatora
microscope
Yeast
Viruses:
Animal (enteric)
Bacterial
2-3
0.2-0.8
0.2-1.0
1-2
3-5
1/2-1
plankton nets
incubatora
incubator
tissue culture
laboratory
incubator3
a Many may be assayed at room temperature.
75
-------
a type of polio virus, have been used but are considered risky. Most animal
enteric viruses are considered safer, as they are not known to infect nan
(Keswick et al., 1982). However, both human and animal viruses are gener-
ally not considered to be suitable tracers for field work. Bacterial
viruses (bacteriophage) were first used by Wimpenny et al. (1972). Their
properties are similar to those of animal viruses, but the health risk is
lower. Virus tracers have several advantageous properties. The injection
culture is easily grown in the laboratory. The viruses are specific to the
host bacteria so that they may be mixed, injected, and then distinguished on
different host bacteria. Also, viruses have shown good survival in ground-
water studies.
The most useful application of virus tracers is in modeling the move-
ment of microbial pathogens (such as hepatitis) in ground water. The move-
ment of viruses from septic tank drainfields was traced with the use of a
bovine enterovirus by Scandura and Sobsey (1981). In karst terrain in
Missouri, Fletcher and Meyers (1974) used a bacteriophage which traveled a
distance of 1600 meters. In granular media, the aquifer must be very perme-
able to observe long travel distances. Martin and Thomas (1974) used a bac-
teriophage In sandstone strata in South Wales, with a travel distance of 680
meters.
A possible use of virus tracers is in evaluating the suitability of
potential land treatment sites (Keswick et al., 1982). A "standard" virus
with well-defined properties could be used.
• Field Methods
An advantage of bacterial or virus tracers is the small injection
volume needed to label large water volumes. A typical concentration of
76
-------
injected tracer might be 5 x 1010 phage per ml. In determining the injec-
tion quantity, Aley and Fletcher (1976) suggested that 97% of the tracer may
be lost and noted that the minimum detection is 10 plaque-forming units per
ml of water. The methods to prepare the tracer are described by Schaub et
al. (1975) and Sargeant (1969). The stock to be injected can be grown rela-
tively easily in a well-equipped microbiology laboratory in 10 to 24 hours.
The method used to assay viruses is described by Schaub and Sorber
(1977), Schaub et al. (1975), and Aley and Fletcher (1976). In general, a
portion of a sample is put on a plate of jelly-like bacteria. The plate is
incubated for various lengths of time, depending on the bacteria and virus.
The virus feeds off of the bacteria and leaves a clean area (plaque) of dead
bacteria on the milky surface of the plate. The clear patches on the plate
are counted manually, assuming that one phage is associated with one plaque.
It is best to have 30-300 plaques per plate. The sample can be serially
diluted to obtain this concentration.
The procedure is fairly complex and time consuming, and it may be diff-
icult for hydrologists who generally lack knowledge of microbiological
techniques.
An immunochemical type of virus assay (analysis) has the potential to
reduce virus detection time to one to three hours (Keswick et al., 1982).
However, this method is not yet available for water tracers.
• Additional Information
Some considerations in planning and interpreting virus tracer tests are
die-off rates, background levels, and adsorption. The die-off rate should
be investigated before choosing the tracer. Martin and Thomas (1974) found
that the bacteriophage population which they used was reduced to 10% of the
original value in about 28 days. The die-off rate is increased by higher
77
-------
temperature and exposure to ultraviolet light. Background levels of viruses
in ground water might cause tracer test results to be incorrect and should
be determined before the test. Aley and Fletcher (1976) described a method
to determine if interfering bacteriophage are present in the water, which
would infect the tracer phage's host bacteria.
Neglecting the time factor, the cost of using virus tracers is rela-
tively small if access to a microbiology laboratory is available. The pri-
mary cost would be related directly to wages of laboratory personnel.
Spores
Lycopodium spores have been used as a water tracer since the early
1950's, and the techniques are well developed. Spore tracing was initiated
by Mayr (1953) and Maurin and Zotl (1959). Their methods were modified by
Drew (1968a). As is true with all larger particulate tracers, spores can be
used only where significant interconnected large pores exist. Almost all
applications of spore tracers have been in karst regions characterized by
large solution openings in the aquifers.
Lycopodium is a clubmoss which has spores that are nearly spherical in
shape, with a mean diameter of 33 microns. It is composed of cellulose and
is slightly denser than water, requiring some turbulence to keep the mate-
rial in suspension. Some advantages of lycopodium use are: (1) the spores
are relatively small; (2) they are not affected by water chemistry or
adsorbed by clay or silt; (3) they travel at approximately the velocity of
the surrounding water; (4) the injection concentration can be very high
(e.g., 8 x 10 6 spores per cubic centimeter); (5) no health threat is posed;
(6) the spores are easily detectable under the microscope; and (7) at least
five dye colors may be used, allowing five tracings to be conducted
78
-------
simultaneously in a karst system. Some disadvantages associated with its
use include the large amount of time required for preparation and analysis
of the spores, and the problem of filtering of spores by sand or gravel if
flow is not sufficiently turbulent.
The basic procedure involves the addition of a few kilograms of dyed
spores to a cave or sinking stream. The movement of the tracer is monitored
by sampling downstream in the cave or at a spring, with plankton nets
installed in the stream bed. The sediment caught in the net is concen-
trated, and treated to remove organic matter. The spores are then examined
under the microscope.
Tracing by lycopodium spores is most useful in open joints or solution
channels (karst terrain). It is not useful in wells or boreholes unless the
water is pumped continuously to the surface and filtered. A velocity of a
few miles per hour has been found sufficient to keep the spores in suspen-
sion. According to Smart and Smith (1976), lycopodium is preferable to dyes
for use in large-scale water resource reconnaissance studies in karst areas.
This holds if skilled personnel are available to sample and analyze the
spores and a relatively small number of sampling sites are used.
The spores survive well in polluted water, but do not perform well in
slow flow or in water with a high sediment concentration. Lycopodium spores
have been used extensively in the United States, Great Britain, and other
countries to determine flow paths and to estimate time of travel in karst
systems.
• Field Methods
Various pieces of equipment are required for spore preparation, sam-
pling, and analysis. Tracer preparation is described in detail in Drew and
79
-------
Smith (1969), Gardner and Gray (1976), and Aley and Fletcher (1976). A. res-
pirator should be worn during the process, and extreme care must be used to
avoid powder explosions when working with the dry spores. The spores and
dyes can be obtained from a biological supply house (see Appendix D). The
preparation involves heating the wetted spores, adding the dye and boiling
for about an hour, and finally adding chemicals to fix the color in the
spores. Next, the dyed spores are dried in an oven and refrigerated until
used. The dyes found to be most easily distinguished with a regular micro-
scope by Gardner and Gray (1976) were safranine orange, crystal violet, mal-
achite green, sudan black, and crystal blue.
The equipment needed for sampling includes a conical plankton net and a
trap (wood or metal frame) to hold the net- Nylon or silk nets are avail-
able from biological supply houses. Nylon is more expensive, but more tear-
resistant than silk. A 25-micron mesh is generally used. One rule of thumb
in determining the net opening diameter to be used is that the net opening
should be no less than 10% of the cross-sectional area of flow at the loca-
tion of the trap (Gardner and Gray, 1976). The nets are tapered at one end
and fitted with a rubber tube and clip to allow emptying into a bottle
during sampling.
Various suggestions have been made concerning injection quantity. A
large quantity of spores is necessary because probably 99% of the injected
spores are lost in transit, and only a few of those which are transported to
the sampling site are caught in the nets. Drew and Smith (1969) and
Atkinson (1968) recommended using 600 grams (dry weight) of spores (per 0.3
m3/sec discharge) for every estimated kilometer of straight-line travel.
This recommendation is based on the discharge of the largest spring and
80
-------
assumes that approximately 10% of the flow passes through the nets at the
spring. A high silt content reduces the number of spores arriving at the
sampling points. Maurin and Zotl (1959) used 2 to 3 kilograms for a dis-
charge of 500-2,000 cubic meters per hour. Using silk nets, Atkinson, Drew,
and High used one kilogram of spores per 10,000 gallons per minute dis-
charge, per mile of travel. With nylon nets, they successfully used 0.75
kilograms per 50,000 gallons per hour discharge, per mile of travel.
Approximately 8-10% of the outflow was netted in this study.
The sampling nets must be installed before injection. The traps are
placed securely (see Figure 4.4) in the portion of the stream or spring with
the highest discharge. Extra traps should be available in case one is
broken or lost.
Sampling consists of emptying the sediment trapped in the net into a
bottle. The nets are washed in the stream and reused. The frequency of
sampling will depend partly on the amount of sediment in the water, with a
higher sediment load requiring more frequent sampling. If the purpose of
the test is to determine where a sink resurges, sampling once every two days
may be sufficient. If time of travel is desired, the interval should be no
less than one-fifth of the estimated travel time (Gardner and Gray, 1976).
• Detection and Analysis
Laboratory analysis is fairly time consuming for this type of tracer
experiment. The basic equipment includes a good quality microscope and a
centrifuge. The analysis is described in detail in Aley and Fletcher
(1976). Samples are filtered to separate the spores from larger solids,
then concentrated with a centrifuge and analyzed under the microscope.
81
-------
STAKES --^
DRIVEN INTO
STREAM BED
^PLANKTON
NET
-•-SAMPLE
BOTTLE
ROCKS TO PREVENT
FRAME FROM MOVING
Figure 4.4. Tracing with spores is most commonly done
in karst systems where cave streams or streams fed by
large springs are available for sampling. Spores can
be collected from the streams by anchoring a plankton
net in the stream as shown in this sketch (adapted from
Gardner and Gray, 1976).
82
-------
Prevention of contamination is of utmost importance in the analysis lab.
Analysis should not be performed in the same room in which spores were dyed.
• Additional Information
As mentioned above, contamination is the major concern in lycopodium
tracing. Natural systems may have a background level, which should be
tested for. Also, injected spores have been found in streams long after
injection. If the time vs. concentration curve doesn't look reasonable,
contamination is very likely. Gardner and Gray (1976) discussed precautions
which should be taken to prevent contamination in all stages of spore prepa-
ration, injection, sampling, and analysis.
The performance of lycopodium in comparison to other tracers demon-
strates its conservative nature. Buchtela et al. (1968) compared spores,
uranine (a fluorescent dye) and sodium chloride, and found that lycopodium
traveled most rapidly. This may be attributed to the fact that lycopodium
stays in suspension only in fast, turbulent flow, so it probably travels
faster than the average water velocity. Atkinson et al. (1973) compared
lycopodium with pyranine, another fluorescent dye, and found similar peak
arrival times (see Figure 4.5). However, the first arrival of lycopodium
was much earlier than that of pyranine.
The cost of using lycopodium is generally considered greater than that
of using dyes. Smart and Smith (1976) noted that capital costs are similar,
but labor is higher for spore use. See Table 4.2 for a comparison of lyco-
podium and dye tracer properties.
Lycopodium is not associated with any known health effects, and it is
considered one of the most harmless tracers.
83
-------
O.O7
^ O.06
Q_
z O.05
d °-04
O O.03
O
UJ 0.02
Q 0.01
o
1 1 r~
-
LYCOPODIUM
-
-
_
i i r~^
•m*
—
f
1
K
/DY
1 '
-
-
\
\
•^
'E
,—.
,
\
-
-
^.
^A \ *«tj— >
468
TIME (hours)
140 uj
I2O ^
IOO S$
8O
UJ
60 ^ x
-------
TABLE 4.2
Comparison of Lycopodium and Fluorescent Dye Properties
Lycopodium spores
Fluorescent dyes
Require only periodic sampling
Sampling requires the use of
special plankton nets
Cost of capital equipment
(microscope, centrifuge)
moderate
Cost of non-capital equipment
(nets, glassware, etc.) high
Pre-treatment to color spores
time consuming and moder-
ately expensive
Post-collection treatment
time consuming
Analysis time consuming and
requires skilled personnel
Immediate field analysis not
possible
Cost of tracers moderate
Unaffected by water chemistry
and pollutants
Affected by high sediment
concentrat ions
Require frequent sampling
Sampling requires no special equip-
ment and is possible using an
automatic water sampler
Cost of capital equipment (fluorom-
eter) high
Cost of non-capital equipment
(glassware) low
No pre-treatment required
No post-collection treatment
Analysis straightforward and fast,
requires no skilled personnel
Immediate field analysis possible
Cost of tracers moderate
May be detrimentally affected by
water chemistry and pollutants
Affected only at extremely high
sediment concentrations
85
-------
Ions
Ionic compounds such as common salts have been used extensively as
ground-water tracers. This category of tracers includes those compounds
which undergo ionization in water, resulting in separation into charged
elements possessing a positive charge (cations) or a negative charge
(anions). The charge on an ion affects its movement through aquifers by
numerous mechanisms, which will be discussed for each specific tracer. The
ionic tracers which will be mentioned include chloride (Cl~), bromide (Br ),
_.!_ _L- i i _L-
lithium (Li), ammonium (NH ), magnesium (Mg ), potassium (K ), iodide
(I~), sulfate (SO =), organic anions (such as benzoate), and fluorinated
organic anions (such as M-TFMBA). Ions listed are those which have been
found to be successful as tracers under various field and laboratory
conditions.
Ionic tracers have been employed as tools for a wide range of hydro-
logic problems dealing with the determination of flow paths and residence
time and the measurement of aquifer properties. Selection of the approp-
riate ionic tracer should be based on the purpose of the study; the type of
aquifer system, such as karst, granular media, and fractured rock; the site
specific aquifer characteristics, including the degree of heterogeneity and
extent of clay lenses; the natural background concentration of specific ions
in the ground water; and the analytical techniques available. The general
characteristics of an ideal tracer have been outlined previously in the
introduction. Specific characteristics of individual ions or ionic groups
may approach those of an ideal tracer, particularly in the case of dilute
concentrations of certain anions.
-------
In most situations, anions (negatively-charged ions) are not affected
by the aquifer medium. Mattson (1928), however, has shown that the capacity
of clay minerals for holding anions increases with decreasing pH. Under
such conditions of low pH, anions in the presence of clay, other minerals,
or organic detrites may undergo anion exchange. Other effects which may
occur include anion exclusion and precipitation/dissolution reactions.
Cations (positively-charged ions), however, will react much more frequently
with clay minerals through the process of cation exchange which in turn dis-
places other cations such as sodium and calcium into solution. For this
reason, little work has been done with cations due to the interaction with
the aquifer media. Versene (a tetra-sodium salt) has been used in the labo-
ratory to prevent ion exchange (Haas, 1959). Kaufman (1956) has shown that,
when permeabilities and flow rates are low, often indicative of a large clay
fraction, the solid phase may have a considerable capacity for adsorption of
an ionic component. This is significant for cationic tracers and may have
some significance for certain anionic tracers.
One advantage of the simple ionic tracers is that they do not decompose
and thus are not lost from the system. One consideration in the application
of specific ions is the background concentration of the tracer in the natu-
ral ground water or receiving waters. A large number of ions (including Cl
and NO ~) have high natural background concentrations. Use of these ions
under such situations would require the injection of a tracer of high con-
centration which may result in density separation and gravity segregation
during the tracer test (Grisak, 1979). Density differences will alter flow
patterns, the degree of ion exchange, and secondary chemical precipitation,
which may change the aquifer permeability.
87
-------
Various applications of ionic tracers have been described in the liter-
ature. Common salt was used by Adolph Thiem and other German hydrologists
as early as 1889 to determine the flow rate of ground water in sandstone and
other media. Similar methods used for Cl~ were also postulated for ions
such as nitrate (NO ~), dichromate (Cr20?), and ammonium (NH4+) (Haas,
1959). Murray (1981) used lithium bromide (LiBr) in carbonate terrain to
establish hydraulic connection between a landfill and a fresh-water spring
where use of rhodamine WT dye tracer proved inappropriate. Sodium chloride
(NaCl) was used by Mather (1969) to investigate the influence of mining sub-
sidence on the pattern of ground-water flow. Tennyson (1980) used bromide
(Br~) to evaluate pathways and transit time of recharge through soil at a
proposed sewage effluent irrigation site. Chloride (Cl~) and calcium (Ca+)
were used by Grisak (1979) to study solute transport mechanisms in frac-
tures. Potassium (K*~) was used to determine leachate migration and extent
of dilution by receiving waters located by a waste disposal site (Ellis,
1979).
• Field Methods
The field techniques and required equipment for use of ion tracers are
fairly simple and standard for all of the ionic elements in this group. It
is primarily in the detection and analysis phases of a field study that
techniques and required analytical equipment vary substantially. The basic
equipment necessary to conduct a multiple-well fluid tracer test would
include an injection well, observation wells or piezometric nest, auger
(manual or power), well-casing driver (manual or power), steel measuring
tape, tracer mixing and injection container, hand pump or automatic sampler,
sample bottles, and break-through detection equipment (i.e., electrical
-------
conductivity/resistivity meter). Tests may be run utilizing a single bore-
hole for injection and observation (Saleem, 1971) as discussed below, or by
utilizing only one observation well given that the flow direction is estab-
lished. Variations from the standard multiple-well test require modifica-
tions in equipment and techniques- The monitoring network (well configura-
tion), sampling instrumentation, sampling frequency, and detection methods
are dependent on the flow velocity and direction of the measured system.
This information can be obtained using a conservative ionic tracer with a
multi-observation well configuration. Knowledge of the ground-water direc-
tion and flow velocity is critical when conducting a single-well or two-well
tracer study.
Several types of tracer tests have been performed successfully with
ionic tracers, including two-well recirculating tests; radial flow tests;
convergent flow slug tests; point dilution tests; and packer tests.
The concentration of ion to be injected should be such that it can be
detected well above the natural background concentration level that exists
in the receiving water. Density effects should be considered when determin-
ing injection concentration. One method to offset density effects is to
raise the temperature of the injection mixture above that of the receiving
ground water. Tracer dilutions of as much as six or seven orders of magni-
tude or greater may be unavoidable in field tracer tests (Thompson, 1980).
The ion injection concentration should thus be high enough to ensure detect-
able levels (based on analytical techniques) in the observation well(s).
The ion tracer may be introduced as a powdered salt and allowed to dis-
solve in solution in the injection borehole. This passive injection tech-
nique results in negligible disturbance of the normal ground-water flow
89
-------
velocity and direction. This would be employed when the flow velocity is
large or the distance between the injection well and the observation well is
short. The ion tracer may also be introduced at a known flow rate and con-
stant concentration. This technique would be a forced injection with a con-
stant injection flow rate. This is useful in situations where the ground-
water flow velocity (average pore velocity) is small and/or the distance
between the injection and observation wells is large. The tracer may either
be injected as a slug or as a continuous source input.
The simplest ion tracer tests do not require the collection of samples
from the observation well. The technique developed by Slichter (1902) mea-
sures tracer recovery by changes in electrical conductivity of the ground
water, and thus does not require further laboratory analysis. Sampling is
required for other detection techniques (outlined in the following section),
which are employed when density effects are significant and the injected ion
concentrations are very low. There are two types of sampling: constant
depth sampling and multi-level sampling. Sampling may be conducted manually
using a hand pump or automatically using an electric or battery-operated
sampler.
Lee (1980) employed a multi-level sampler to obtain pore water from
various depths in a flow field. The sampler is a vertical bundle of poly-
propylene tubes which terminate in a small patch of nylon screen and are set
at selected depths- The multi-level sampler is described as an effective
and relatively inexpensive means of defining the spatial distribution and
temporal variations of the tracer zone.
The number of samples kept for laboratory analysis can be minimized by
making field measurements of electrical conductance within, ahead of, and
90
-------
behind the tracer slug (Lee, 1980). This field measurement provides an idea
of relative breakthrough, and thus indicates when sampling frequency should
be increased or decreased. When using samplers, each tube should be flushed
before taking a sample by withdrawing and discarding one pump and tube
volume.
• Detection and Analysis
As mentioned previously, the simplest and most inexpensive detection
and analysis technique for ionic tracers is the measurement of electrical
conductance as described by Slichter (1902). Electrical conductance can be
used (as a break-through indicator) in two-well or multi-well tracer tests.
The movement of an ionic tracer from the injection well towards the obser-
vation well is observed by means of an electric circuit that utilizes the
conductivity of the ground water. As the tracer moves towards the observa-
tion well, the conductivity increases. An electric circuit within each
observation well is used for detecting the time of arrival (break-through
time) of the tracer (Haas, 1959).
Additional detection and analysis techniques are used if ionic tracers
are injected at low concentrations, when greater accuracy is required, or in
aquifer systems where electrical conductance is difficult to measure. The
numerous detection and analytical techniques require that samples be col-
lected in the field and that analyses be performed in the laboratory. The
techniques applicable to specific ions are presented in Table 4.3. Several
common techniques include specific ion electrodes, "Hach kit" analysis,
liquid chromatography, gas chromatography, and mass spectrometry. Appendix
E provides a further discussion of analytical methods for detection of
tracers.
91
-------
TABLE 4.3
Analytical Methods for Ionic Tracers
Ion
Li+ K+
i-l J. y IX
Mn^"1", Mg^,
K1", Na+
Cl~
S0,=
NH ,-N
1+
Br'
Method
Atomic absorption
Spectrophotome try
Flame emission
Spectrophotome try
Coulometric filtration
Mercuric thiosulfate method
Turbidimetric
U.V.-Visible spectrophotometry
coupled with chemical procedures
Brucine-sulfanilic acid
Sulfanilamide-napthylenediamine
Phenolhypochlorite-nitroprusside
Specific ion electrode
Spectrophotometry
Neutron activation
Reference
Brown et al. (1970)
Pickett (1969)
Cotlove (1964)
Lee (1980)
Hach (1969)
USDI (1969a)
Tennyson (1980)
92
-------
The overall costs of ionic tracer tests include both the cost of the
injected ion (salt) and the costs of analysis. Common salts such as NaCl are
relatively inexpensive, but some of the organic anions and fluorinated
organic anions are expensive. With the exception of the electrical conduc-
tance, specific ion electrode, and Hach methods of analysis, the detection
and analysis costs are significantly greater than the cost of the ionic com-
pound. Cost is a limiting factor in the use of several detection methods.
For example, Schmotzer et al. (1973) applied post-sampling neutron activation
in a Br~ tracer test and pointed out that a major disadvantage with this
technique is the significant cost of analysis. Each sample requires irradia-
tion and generally chemical separation, counting, and quantitative analysis-
The concentrations of ionic solutions typically used in field tracer
tests generally pose no measurable environmental or health effects. Water
containing such concentrations of ions is much less palatable, but in most
cases is potable. Schmotzer et al. (1973) reported that the only toxicity
data on bromide (Br~) was a result of medical research by Dreisback (1955)
and Von Oettingon (1958). The report concluded that 50-100 mg of Br"/100 ml
of blood is the lower toxic limit for humans. For an adult, this limit is
2.4 grams of Br~ in the blood. A person drinking water with a bromide con-
centration of 200 ppm would have to drink 12 liters in order to ingest 2.4
grams of bromide. Natural background levels of bromide in ground water are
usually low, and therefore low concentrations are typically employed in
tests.
• Discussion of Specific Ion Tracers
Chloride (Cl~): Background levels in ground water are typically moder-
ate to high. Chloride can be used satisfactorily where density effects can
93
-------
be avoided and dispersion of clays is not likely. A chloride front proceeds
at a high velocity and exhibits little distortion, resulting in sharp elution
curves (Kaufman, 1956). See Figure 4.6 for a comparison of chloride, dex-
trose, fluorescein, and 131I. The problem with chloride is the necessity of
using high doses of NaCl to provide detectable concentrations at distant
wells, and the danger of altering the permeabilities of high-clay soils by
ion exchange. Kurty (1972) found that Cl~ and nitrate (NO ~) move at equal
rates. Davis et al. (1980) reported that the injection concentration of NaCl
should not exceed 3,000 mg/1 (ppm) because of the increased density effects.
Cl~ is a fairly conservative tracer which may be weakly adsorbed by some
soils.
Bromide (Br~); Bromide has low background levels in ground water, thus
allowing low injection concentrations relative to chloride. Br~ is perhaps
the most commonly used ion tracer. Jester and Uhler (1974) concluded that
bromide was superior to chloride, iodide, fluoride, and vanadium when used as
a tracer in soil-water systems with post-sampling neutron activation analy-
sis. Schmotzer et al. (1973) reported that Br~ is biologically stable, and
appears not to be lost by precipitation, absorption, or adsorption. Smith
and Davies (1974) found that NO ~ lags behind Br~ as a tracer. Expected
background concentration of bromide will be <1 mg/1 in most aquifers contain-
ing potable water (Davis et al., 1980). There are numerous techniques for
detection and analysis of bromide, ranging from inexpensive methods (electri-
cal conductance or specific ion electrode) to more expensive methods (neutron
activation analysis or liquid chromatography).
-------
0 100 200 300 400 500 600 700 800
VOLUME ELUTED (liters)
Figure 4.6. A comparison of several tracers in a laboratory
test. Note that chloride shows less of an effect of sorption
than the other tracers (figure adapted from Kaufman, 1956).
95
-------
Lithium (Li+); Lithium has a low (0.05 to 0.3 mg/1) background concen-
tration in potable ground water, but has a high loss to ion exchange (Haas,
1959).
Ammonium (NH"t'); This ion exhibits relatively high loss to ion
exchange and the analysis of ammonia is more difficult than most other com-
mon ions (Haas, 1959). Natural background values in most potable water are
below 5 mg/1.
• i i i
Magnesium (Mg ): As is true with other positive ions, Mg is subject
to sorption and ion exchange. However, analyses are simple and inexpensive.
Natural background values are commonly between 2 and 40 mg/1 in potable
ground water.
Potassium (K^): A simple potassium ion will be sorbed and concentra-
tions in water will be modified by ion exchange. Analyses are rapid and
simple with atomic absorption or emission techniques. Expected background
values in potable ground water are relatively low, ranging from about 0.2 to
10 mg/1.
Iodide (I~): Iodide has very low background concentrations (generally
<0.01 mg/1). Methods for sensitive analysis of I~ are relatively simple.
However, iodide tends to be sorbed to a greater extent than either Br~ or
Cl~ (Davis et al., 1980) and it is affected by microbiological activity.
Organic anions: These compounds have very low background concentra-
tions, are nonsorbed, nonvolatile, and are highly to moderately stable.
High precision measurement techniques are available with a detection sensi-
tivity of 50 ppt. One disadvantage is the high cost of these compounds,
which include benzoate and m-TFMBA. Malcolm et al. (1980) found that these
compounds are highly mobile and have a good sensitivity of detection and a
high precision of measurement using liquid chromatography.
96
-------
Dyes
Various organic dyes have been used for surface-water and ground-water
tracing since the late 1800's. Dyes are relatively inexpensive, simple to
use, and effective. The extensive use of fluorescent dyes for water tracing
began around 1960. Fluorescent dyes are preferable to non-fluorescent
varieties due to much better detectability. Some non-fluorescent dyes in-
clude Congo Red and Malachite Green, which have been used in conjunction
with cotton strip detectors (Drew, 1968) or with visual detection, often in
soil studies. This discussion will concentrate on fluorescent dyes, which
are more suitable for ground-water studies.
The most commonly used tracer dyes to be discussed include fluorescein,
pyranine, lissamine FF, rhodamine B, rhodamine UT, and sulfo rhodamine B.
Photine CU and amino G acid, two optical brighteners, will also be men-
tioned.
Table 4.4 gives the dyes by color, lists alternative names, and pro-
vides spectra wavelengths and filter combinations for their analysis. Sev-
eral dyes may be used in a single tracer test if the absorption and emission
spectra do not overlap. For example, Smart and Laidlaw (1977) recommended a
combination of lissamine FF, amino G acid, and rhodamine WT. In general,
the spectra do overlap, particularly for dyes of the same color. Figure 4.7
illustrates the excitation and emission spectra of rhodamine WT.
Although fluorescent dyes exhibit many of the properties of an ideal
tracer, a number of factors interfere with concentration measurement.
Fluorescence is used to measure dye concentration, but it may vary with sus-
pended sediment load, temperature, pH, CaC03 content, salinity, etc. Other
variables which affect tracer test results are "quenching" (some emitted
97
-------
TABLE 4.4
Generic and Alternative Names
And Chemical Structure of the Tracer Dyes
Blue Fluorescent Dyes
Amlno C acid
Photlne CD
Alternative
Names
Maximum
Excitation,
7-anlno 1,3 napthalene
dlsulphonlc. acid
355 (310)
345
Ma xImun
Emission,
nm
445
435(455)
Primary
Filter
7-37*
Mercury
Line,
nm
36 S
Secondary
Filter
98**
Green Fluorescent Dyes
FluoreaceIn
Llssamlne FF
Pyranlne
Fluoresceln LT 490
Uranlne
Sodium fluoresceln
Ussamlne yellow FF 420
Brilliant sulpho flavlne FF
Brilliant acid yellow 8G
Pyranlne Cone. 455(405)
D & C green 8
520
5V 5
515
98**
436
55**
Orange Fluorescent Dyes
Rhodamlne B
Rhodamine UT
Sulfo rhodanlne B
Pontacyl brilliant pink B
Llssamlne red 4B
Klton rhodatnine B
Acid rhodamlne B
555
555
565
580
580
590
2xl-60*+61**
546
4-97+3-66*
Figured In parentheses re!er to secondary maxima- For all spectra, pH la 7.0.
* Corning filter.
** Hodak Wratten filter.
Source: Smart and Laldlaw (1977)
-------
1.0
2 _J 0.8
UJ
f-
~ S 0.6
(- ^ 0.4
< UJ
UJ Lu
a o 0.2
r i i—r
0
400
Excitation spectrum
(peak 558
Rhoclamine WT
Emission spectrum
(peck 582 m/z)
Secondary filter peak
(590
\
500 600
WAVELENGTH (m//)
700
800
Figure 4.7. Excitation and emission characteristics of rhodamine WT,
a fluorescent dye commonly used as a tracer (source, Wilson, 1968).
-------
fluorescent light is reabsorbed by other molecules), adsorption, and photo-
chemical and biological decay. These effects will be discussed in more
detail in reference to specific dyes. Another disadvantage of fluorescent
dyes is their poor performance in tropical climates, due to chemical reac-
tions with dissolved carbon dioxide.
The advantages of using these dyes include their very high detecta-
bility, rapid field analysis, and relatively low cost and low toxicity. The
theory of fluorescence is described by McLaughlin (1982) and Skoog and West
(1980). As described by McLaughlin, the process of fluorescence involves
the following steps: (1) energy is absorbed by the molecule from sunlight
or an ultraviolet lamp and a transition to a higher, excited electron state
takes place; (2) the molecule relaxes from the highest to the lowest
vibrational energy of that state, losing energy in the process; and (3) if
the excess energy is not further dissipated by collisions with other mole-
cules (quenching), the electron returns to the lower ground electron
state. This emission of energy due to the transition from the higher to the
lower state is fluorescence.
• Field Methods
The basic equipment necessary for dye tracing is a manual or automatic
sampler and a field or laboratory detection device. Sampling is performed
by adsorption of dye onto packets of activated charcoal suspended in the
water (in karst topography), or by taking grab samples at a discharge point
(for karst, porous media or fractured rock studies). A filter fluorometer
or a spectrofluorometer is generally used for analysis, although visual
detection is sometimes used for qualitative results.
The tracer is introduced at a sink hole or well. The detection limit
for fluorescent dyes is very low, so the quantity of tracer used is
100
-------
relatively small. The amount of tracer needed has been approximated for
karst systems by Drew and Smith (1969). They recommended using 60 grams of
dye per kilometer of underground travel, per 0.15 cubic meters per second of
discharge, at the largest likely rising. Atkinson et al. (1973) also
described a method to calculate dye injection quantities for karst tests.
One sampling technique used in karst tracing is the Dunn method,
developed in 1957. Small packets of fine mesh nylon or window screen are
filled with activated charcoal and suspended in the watercourse at the
sampling point. The dye adsorbs very strongly onto the charcoal, and is
later eluted by placing the bag in a solution of 5% NH OH and ethyl alcohol.
After soaking for one hour, the dye can be analyzed. The charcoal packets
must be changed periodically, depending on flow rates and dilution of the
tracer. Some examples of the use of this method are given in Gann and
Harvey (1975) and Drew and Smith (1969). Cotton strip detectors have been
used in a similar manner. Marston and Schofield (1962) described a tracer
test using rhodamine B and cotton detectors.
Flow-through fluorimeters are sometimes used, which eliminate the need
for sample collection. However, the most common method is collection of
samples in sample bottles. Automatic samplers have been discussed in
Chapter 3. Glass bottles should be used rather than polyethylene to avoid
adsorption (Hubbard et al., 1982). Reznek et al. (1979) described pro-
cedures for sampling and analyzing fluorescein, and many of the procedures
apply to other dyes. The samples and standards should be buffered to within
a 5 to 11 pH range before analysis. If the samples are turbid, it is pref-
erable to centrifuge the samples rather than filter them, as dye adsorbs
101
-------
onto the filter. The dye to be injected and the samples should both be
stored out of sunlight and preferably in light-proof containers. Feuerstein
and Selleck (1963) found that some fluorescent dyes exhibit a 50% photochem-
ical chemical decay in two days, even when stored in light-proof flasks.
Obviously, it is advisable to analyze samples as soon as possible after
sampling.
• Detection and Analysis
It is possible to visually detect some dyes in water at a concentration
of about 40 ppm (Corey, 1968). This concentration is much higher than 10
ppb, which is the maximum permissible concentration allowed at drinking
water intakes (Wilson, 1968). The visual detection method is qualitative
and rarely used.
Other detectors are the filter fluorometer and the spectrofluorometer.
The filter fluorometer (or fluorimeter) consists basically of an ultraviolet
light source, glass curvets (sample holders), and sets of primary and
secondary filters which correspond to the absorption and emission wave-
lengths of the dyes used. The filter fluorometer must be calibrated with
standard solutions at the same temperature as the samples to be analyzed.
As mentioned before, the fluorescence of a sample is affected not only by
concentration of the dye, but also by background fluorescence, temperature,
pH, turbidity, photochemical decay, and adsorption. Temperature control
apparatus and correction charts are available, and methods to avoid the
other interferences have been briefly discussed. Two U.S. Geological Survey
publications are very useful for planning a fluorescent dye test and avoid-
ing these interferences. "Fluorometric Procedures for Dye Tracing," by
Wil son (1968) is a classic report, and in 1982, Hubbard et al. published a
very useful updated report, "Measurement of Time of Travel and Dispersion in
102
-------
Streams by Dye Tracing." These two are excellent references, as is the
article by Smart and Laidlaw (1977).
Fluorometers are available with individual sample analysis capability,
or with flow-through sampling (see Figure 4.8). They can be equipped with
strip-chart recorders, and can be powered in the field with a portable
generator (with a transformer) or a car battery (Hubbard et al., 1982).
Some of the most well-known fluorometers are made by American Instrument
Company and by G. K. Turner Associates.
Spectrofluorometers are more expensive and more complex to operate than
filter fluorometers. They are generally not taken into the field. An
example of this type of instrument is the Aminco-Bowman ultraviolet spectro-
photofluorometer made by American Instrument Company. Table 4.5 shows sen-
sitivity and minimum detection for certain dyes.
• Additional Information
The effects of temperature, pH, and suspended solids concentration on
fluorescence have been mentioned. Fluorescence intensity is inversely pro-
portional to temperature. Smart and Laidlaw (1977) described the numerical
relationship and provide temperature correction curves. The effect of pH on
rhodamine WT fluorescence is shown in Figure 4.9. An increase in the sus-
pended sediment concentration generally causes a decrease in fluorescence.
Adsorption on kaolinite caused a decrease in the measured fluorescence of
several dyes, as measured by Smart and Laidlaw (see Figure 4.10).
The detected fluorescence may decrease, as in this example, or actually
increase due to adsorption. If dye is adsorbed onto suspended solids, and
the fluorescence measurements are taken without separating the water samples
from the sediment, the dye concentration is a measurement of sediment
103
-------
o
Constant-voltage
transformer
Voltmeter
Standard
solution
(optional)
generator
Figure 4.8. Cave streams or large springs can be monitored for dye tracers by using continuously
recording fluorometers as shown schematically in this diagram by Wilson (1968) . Less expensive
but also less precise methods of monitoring for dye can use packets of activated charcoal which
are placed in the stream. The dye, if present, is strongly sorbed on the charcoal which is then
taken into the laboratory for analysis. The charcoal packets are replaced periodically in the
stream until the test is finished.
-------
TABLE 4.5
Sensitivity and Minimum Detectable
Concentrations for the Tracer Dyes
Dye
Amino G Acid
Photine CU
Fluorescein
Lissamine FF
Pyranine
Rhodamine B
Rhodamine WT
Sulfo rhodamine B
Sensitivity*
yg/1 Per Scale
Unit
0.27
0.19
0.11
0.11
0.033
0.010
0.013
0.061
Background
Reading**
Scale Units
0-100
19.0
19.0
26.5
26.5
26.5
1.5
1.5
1.5
Minimum
Detectability***
yg/1
0.51
0.36
0.29
0.29
0.087
0.010
0.013
0.061
For a Turner 111 filter fluorometer with high-sensitivity door and
recommended filters and lamp at 21°C.
* At a pH of 7.5.
** For distilled water.
*** For a 10% increase over background reading or 1 scale unit,
whichever is larger.
Adapted from Smart and Laidlaw (1977).
105
-------
pH
Figure 4.9. The fluorescence of most dyes is
dependent on pH and the types of dominant ions
present. Results of some experiments on the
fluorescence of rhodamine WT are shown in this
figure adapted from Smart and Laidlaw (1977).
106
-------
Sulfo
Rhodamine B
5 10 15 20
KAOLiNITE CONCENTRATION (g/l)
Figure 4.10. Most dyes will be adsorbed on fine par-
ticulate material, particularly on organic fragments
and clays. Results of experiments with the adsorption
of dyes on kaolinite (a type of clay) as reported by
Smart and Laidlaw (1977) are shown in this illustration.
107
-------
content and not of water flow. As mentioned before, the ideal separation is
with a centrifuge, as the dye can adsorb onto filter paper. Adsorption can
can occur on organic matter, clays (bentonite, kaolinite, etc.), sandstone,
limestone, plants, plankton, and even glass sample bottles. These adsorp-
tion effects are a strong incentive to choose a non-sorptive dye for the
type of medium tested.
Dyes travel slower than water due to adsorption, and are generally not
as conservative as the ionic or radioactive tracers. See Figure 4.6 in the
ion section for a comparison of chloride, dextrose, fluorescein, and 1*31.
Drew (1968) compared lycopodium, temperature, and fluorescein as karst
tracers and found fluorescein breakthrough to be slowest (Figure 4.11). He
questioned the ability of fluorescein to give accurate data on flow rates.
Field data comparing the more recently developed dyes are not yet available.
Atkinson et al. (1973) stated that an advantage of fluorescent dye measure-
ment over lycopodium analysis is the ability to make deductions about dis-
charges, changes in storage, and the geometry of the system. They suggest
that dyes are more useful than spores for obtaining the maximum amount of
quantitative information about a small karst system.
A final point concerning the interpretation of tracer tests is empha-
sized by Brown and Ford (1971). They obtained some very interesting results
by running three identical dye tracer tests in the same karst system. These
yielded three different flow-through times. One of the values differed by
50% from the original test value. Although only one test is generally run
due to economic considerations, it may be advisable to run several tests to
check reproducibility if accuracy is important.
108
-------
— — —— Lycopodium
Temperature
Fluorescein
2.5 5.0
7.5 10.0 12.5 15.0
TIME (minutes)
17.5 20.0 22.5 25.0
Figure 4.11. A comparison of the results of three simul-
taneous tracer tests in a karst system (data from Drew, 1968).
-------
A comparison of the cost of various fluorescent dyes is given in Table
4.6. Prices are given in British pounds per kilogram for bulk dye. Volume
labeled per unit cost is also listed, and rhodamine B appears to be the most
cost effective. However, problems with its use will be discussed in a sub-
sequent section.
Some of the available toxicity data will be mentioned in regard to
specific dyes in the following section. Smart and Laidlaw (1977) discussed
the toxicity of dye tracers, but regulations may change rapidly and should
be researched before conducting a test. Current World Health Organization,
Environmental Protection Agency, and state health standards should be con-
sulted.
• Discussion of Specific Dye Tracers
Green Dyes
Fluorescein
Fluorescein, also known as uranin, sodium fluorescein, and pthalien,
has been one of the most widely used dyes. Like all green dyes, its use is
commonly complicated by high natural background fluorescence, which lowers
sensitivity of analyses and makes interpretation of results more difficult.
It has a very high photochemical decay rate compared to other dyes
(Feuerstein and Selleck, 1963), but this is generally of little concern in
ground-water tracing.
Feuerstein and Selleck (1963) recommended that fluorescein be restricted
to short-term studies of only the highest quality water. Because this dye is
affected strongly by pH (it becomes colorless in acidic conditions), they
suggested that the sample pH be adjusted to greater than 6 before analysis.
Fluorescein also exhibits an appreciable decrease in fluorescence with
110
-------
TABLE 4.6
Relative Costs of Dyes
Dye
Ami no G Acid
Photine CU
Fluoresceln
Lissamlne FF
Pyranine
Rhodamlne B
Rhodanlne WT
Sulfo rhodamine B
State
powder
20% solution
powder
powder
powder
powder
201 solution
powder
Relative Cost
Per Kilogram*
4
1
4
14
13
5
7
9
Volume Labeled
Per Kilogram**
(lOV/kg)
2
3
4
4
12
100
77
16
Volume Labeled
Per Unit Cost
(105m3/cost)
5.7
6.0
10.0
2.8
9.2
200.0
22.0
17.8
After Smart and Laidlaw (1977).
* Costs are approximate and based on 1975 prices; the higher the number, the higher the price of the dye.
** Based on minimum detectabllltles In Table 4.5.
-------
increasing salinity, and is similarly affected by oxidizing agents and sus-
pended solids (Reznek et al., 1979).
Some examples of flucreseein use include a fractured rock study by
Lewis (1966). Borehole dilution tests resulted in hydraulic conductivity
values similar to pump test values. Another example is a mining subsidence
investigation in South Wales, where more than one ton of fluorescein was used
in a sandstone tracer test (Mather et al., 1969). A distance of 1,100 feet
was traversed. Tester et al. (1982) used fluorescein to determine fracture
volumes and diagnose flow behavior in a fractured granitic geothermal reser-
voir. He found no measurable adsorption or decomposition of the dye during
the 24-hour exposures to rocks at 392°F. Omoti and Wild (1979) stated that
fluorescein is one of the best tracers for soil studies, but Rahe et al.
(1978) did not recover any injected dye in their hillslope studies, even at a
distance of 2.5 meters downslope from the injection point. The same experi-
ment used bacterial tracers successfully. Figure 4.11 compares fluorescein,
lycopodium, and temperature as karst tracers.
An advantage of using fluorescein (or any of the green dyes) is its
emission in the green band of the visible spectrum. Fluorescein can be
visually detected at a concentration of about 40 ppm, but other means of
detection are preferred since this is a relatively high concentration. The
approximate sensitivity and minimum detection limit for fluorescein are
given in Table 4.5.
Fluorescein is less costly in bulk than many of the dyes (see Table
4.6), but due to its high photochemical decay rate and high amount of adsorp-
tion, it increases in relative cost as the length of the test increases (more
dye must be added to compensate for loss).
112
-------
Pyranine
Another green fluorescent dye, pyranine, has a stronger fluorescent
signal than does fluorescein, but Is much more expensive. It has been used
in several soil studies, and Reynolds (1966) found it to be the most stable
dye used in an acidic, sandy soil. Omoti and Wild (1979) recommended pyra-
nine and fluorescein as the best tracers for soil tests, although pyranine
is relatively unstable if the organic matter content of the soil is high.
Drew and Smith (1969) stated that pyranine is not as easily detectable as
fluorescein, but is more resistant to decolorization and adsorption. Pyra-
nine has a very high photochemical decay rate, and is strongly affected by
pH in the range found in most natural waters (McLaughlin, 1982).
Lissamine FF
This green dye has been used primarily for aerosol tracing (Yates and
Akisson, 1963), and has not been used extensively in ground-water tests.
Little information is available on the performance of lissamine FF; however,
Smart and Laidlaw (1977) recommended it as the best quantitative tracer of
the three green dyes discussed. The dye is extremely stable and resistant
to adsorption losses, but is much more expensive than most dyes.
Orange Dyes
Rhodamine B
Rhodamine B was the first of the three orange (or red) dyes to be used in
water tracing. Its high adsorption losses make it a less suitable tracer
for ground-water work than rhodamine WT or sulfo rhodamine B, although it
has been used more frequently. Aulenbach et al. (1978) concluded that rho-
damine B should not be used as a ground-water tracer due to sorption losses,
and Feuerstein and Selleck (1963) reported significant adsorption. They
also found that the fluorescence of rhodamine B is affected by large
113
-------
salinity changes. Knuttson (1968) reported that the dye is relatively unaf-
fected by pH in the range found in most natural waters (5-10). The dye how-
ever, is sensitive to temperature (Omoti, 1977) and exhibits optical quench-
ing by suspended solids. Like fluorescein, rhodamine B has a large inter-
ference from high background fluorescence in tropical areas. It is less
affected than the other rhodamine dyes by bacteria and light, but it adsorbs
readily on bentonite, sand and gravel, till, and karst channels, pure quartz
sand, and even plastic and glass laboratory columns (Table 4.7). Hubbard et
al. (1982) compared rhodamine B and rhodamine WT and found high adsorption
of rhodamine B on aquatic plants, suspended clays, and glass and plastic
sample bottles. They found rhodamine WT easier to handle and more economi-
cal than rhodamine B. Although the unit cost of rhodamine B is lower, its
loss rate is much higher than that of rhodamine WT.
Rhodamine B was decertified for use in cosmetics by the U.S. Food and
Drug Administration in the 1960's. In 1968, it was illegal for use as a
water tracer in the U.S. (Drew, 1968). Both rhodamine B and fluorescein
were placed on toxicological classification Gill by the Food and Agriculture
Organization/World Health Organization. Of the dyes discussed in this
article, rhodamine B is generally recognized as the most toxic to man, as it
is readily adsorbed onto body tissue. Currently, the U.S. Geological Survey
recommends that tracer tests should result in a final concentration less
than 10yg/l. Numerous studies related to toxicity tests for various aquatic
organisms are reported by Smart and Laidlaw (1977), and they recommend that
the dye not be used as a water tracer.
Rhodamine WT
This dye has been considered one of the most useful tracers for quanti-
tative studies, based on minimum detectability, photochemical and biological
114
-------
decay rates, and adsorption (Smart and Laidlaw, 1977; Wilson, 1968; and
Knuttson, 1968). Hubbard et al. (1982) stated that it is the most conserva-
tive of dyes available for stream or karst tracing.
Some recent uses of rhodamine WT include projects by Burden (1981),
Aulenbach et al. (1978), Brown and Ford (1971), Gann (1975), and Aulenbach
and Clesceri (1980). Burden successfully used the dye in a water contami-
nation study in New Zealand in an alluvial aquifer. Aulenbach and Clesceri
also found rhodamine WT very successful in a sandy medium.
Gann (1975) used rhodamine WT for karst tracing in a limestone and
dolomite system in Missouri. He used grab samples and activated charcoal
packets, and traced a 14 km (8.7 mile) path. Three fluorescent dyes (rhoda-
mine B, rhodamine WT, and fluorescein) were used by Brown and Ford (1971) in
a karst test in the Maligne Basin in Canada. The highest recovery of dye
(98%) was obtained for rhodamine WT. The fluorescein was not recovered at
all. The horizontal flow path was 1.3 miles, and a Turner III fluorometer
was used for analysis.
Aulenbach et al. (1978) compared rhodamine B, rhodamine WT, and tritium
as tracers in a delta sand. The project involved tracing effluent from a
sewage treatment plant. Sampling was performed with drive points, pumped
wells, and lysimeters. The rhodamine B was highly adsorbed, while the rho-
damine WT and tritium yielded similar break-through curves (Figure 4.12).
Rhodamine WT seems to be adsorbed less than rhodamine B or sulfo rhodamine B
(Table 4.7). Wilson (1971) found that in column and field studies, rhoda-
mine WT did show sorptive tendencies.
Rhodamine WT is thought to be slightly less toxic than rhodamine B and
sulfo rhodamine B (Smart and Laidlaw, 1977). This source notes that rhoda-
mine WT and fluorescein are of comparable toxicity, but Aley and Fletcher
115
-------
30O
c
"E
20O -
O
u
cc
Rhodamine WT
E
0.
a.
IOO -
Z
llJ
O
z
O
O
UJ
>
Q
TIME (days)
Figure 4.12. Although many researchers have found that
rhodamine WT is sorbed on aquifer material, data presented
by Aulenbach et al. (1978) suggest that this dye can be
used in coarse, permeable sand. Comparative data from the
study by Aulenbach et al. (1978) using tritium and rhoda-
mine WT indicate little difference between the two tracers
as shown in this figure adapted from their study.
116
-------
TABLE 4.7
Measured Sorption of Dyes on Bentonite Clay
Losses Due to
Dye Adsorption on Clay
Rhodamlne WT 28%
Rhodamine B 96%
Sulfo Rhodamine B 65%
Source: Repogle et al. (1966)
117
-------
(1976) stated that rhodamine WT is not as "biologically safe" as
£lucreseein.
Sulfo rhodamine B
Sulfo rhodamine B, also known as pontacyl brilliant pink, has not been
used extensively as a ground-water tracer. Its fluorescence is affected
slightly by high salinity, and it exhibits low adsorption on suspended
sediment (Feuerstein and Selleck, 1963). Table 4.7 compares the adsorption
of the rhodamine dyes onto bentonite. This dye is more expensive than the
other rhodamine dyes, and its toxicity appears to be slightly higher than
that of rhodamine WT.
Blue Dyes
The optical brighteners are blue fluorescent dyes which have been used
in increasing amounts in the past decade in textiles, paper, and other
materials to enhance their white appearance. Water which has been contam-
inated by domestic waste can be used as a "natural" tracer, if it contains
detectable amounts of the brighteners. Glover (1972) described the use of
optical brighteners in karst environments. Examples of the brighteners are
amino G acid and photine CU. These two are the least sensitive of the dyes
reviewed (Table 4.5), but the blue dyes have much lower background levels in
uncontaminated water than do the green or orange dyes.
Photine CU is significantly affected by temperature variations, and
both dyes are affected by pH below a pH of 6.0. The dyes have high photo-
chemical decay rates, similar to those of pyranine and fluorescein. Amino G
acid is fairly resistant to adsorption.
Toxicity studies on optical brighteners were reviewed by Akamatsu and
Matsuo (1973). They concluded that the brighteners do not present any toxic
hazard to man, even at excessive dosage levels.
118
-------
Some Common Nonionized and Poorly Ionized Compounds
A number of chemical compounds will dissolve in water but will not
ionize or will ionize only slightly under normal conditions of pH and Eh
found in ground waters. Some of these compounds are relatively difficult to
detect in small concentrations, others present a health hazard, and still
others are present in moderate to large concentrations in natural waters,
thus making the background effects difficult to deal with in most settings.
A list of a few of these compounds is given in Table 4.8.
The use of these and similar compounds as injected tracers in ground
water is limited to rather special cases. Of those listed, boric acid would
probably act most conservatively over long distances of ground-water flow.
Boric acid has been used successfully as a tracer in a geothermal system
(Downs et al., 1983). Large concentrations, 1,000 mg/1 or more, would need
to be used for injected tracers which, unfortunately, would pose difficult
environmental questions if tracing were attempted in aquifers with potable
water. From the standpoint of health concerns, sugars would be the most
acceptable; however, they decompose rapidly in the subsurface and also tend
to be sorbed on some materials. Results of an experiment using dextrose are
shown in Figure 4.6. Alcohols such as ethanol would also tend to be sorbed
on any solid organic matter which might be present. Another problem with
the use of most of these compounds as tracers is that they would need to be
introduced in moderately large concentrations which in turn would materially
change the density and viscosity (particularly for glycerin) of the injected
tracer mixture.
119
-------
TABLE 4.8
A List of Some Simple Compounds Which are Soluble in Water
Name
Formula
Remarks
Silicic Acid
Boric Acid
H.S-iO,
it i ^
(After combination
with water)
H3B°3
Phosphoric Acid
Acetic Acid
H3P<\
Present in normal ground
water in non-ionized form in
concentrations of between 4
and 100 mg/&. Low toxicity.
Present in normal ground
water in nonionized form in
concentrations of 0.05 to 2.0
mg/fc. Toxic to plants above 1
to 5 mg/£. Toxic to humans in
higher concentrations.
ionizes above pH of 6.0. Will
form complexes with other dis-
solved constituents. Sorbs on
or reacts with most aquifer
materials. Natural concentra-
tions mostly between 0.05 mg/£
and 0.5 mg/JU
Moderately toxic in high con-
centrations. Water soluble.
Natural concentrations are less
than 0.1 mg/Jl in ground water.
Ethyl Alcohol
(Ethanol)
Sugars
Sucrose
Maltose
Lactose
Clucose
Glycerol
(Glycerin)
C2H6°
C12H22°11
C12H22°11
C12H22°11
C6 H12°6
C3H6<\
Major component of alcoholic
drinks. Water soluble.
Natural concentrations are less
than 0.05 mg/£ in ground water.
Major components of human and
animal foods. Water soluble.
Probably less than 0.2 mg/£ in
most ground water.
Water soluble. Low toxicity.
Probably absent in natural
ground water.
120
-------
Some of these compounds such as sugars, nevertheless, may be useful for
simulating the movement of other compounds which are also subject to rapid
decomposition but which are too hazardous to inject directly into aquifers.
Detection; Silica and phosphates can be determined by rather simple
colorometric methods using standard solutions and photometric detectors.
Boron is also detected by colorometric methods but the chemical procedure is
more complicated than for silica and phosphate. The organic compounds
listed in Table 4.8 are probably best detected by chromatographic methods.
Also, high concentrations of glycerin and sugars are detected easily by
optical refraction techniques.
Gases
Introduction; Numerous natural as well as artificially produced gases
have been found in ground water. Some of these gases can serve as tracers
which are already introduced, generally by natural processes, into the
ground-water system. In addition, gas can be injected into ground water
and the gas which is consequently dissolved can then serve as an injected
tracer. Only a few examples of injected gases used for ground-water tracers
are found in the literature.
The amount of gas which is dissolved in water increases with the gas
pressure, decreases with an increase of temperature, and decreases with an
increase of the salinity of the water. In most situations, once gas is dis-
solved in ground water at near-atmospheric pressures, the gas will tend to
stay in solution as the water enters the ground-water system. This is due
to the fact that fluid pressure increases rapidly as water moves downward
into an aquifer and the gas will effectively be under a pressure far above
121
-------
the original pressure. If gas such as methane (CH ) is being generated in
the subsurface in large quantities, however, this gas may work its way as
undissolved bubbles of gas through the aquifer and will remove much of the
preexisting dissolved gases from the ground water.
Gases of potential use in hydrogeologic studies are listed in Table
4.9.
Inert Radioactive Gases; Chemically inert but radioactive 133Xe and
85Kr appear to be suitable for many injected tracer applications (Robertson,
1969; and Wagner, 1977), provided legal restrictions can be overcome. Of
the natural inert radioactive gases, 222Rn is the most abundant. It is one
of the daughter products from the spontaneous fission of 238U. Radon is
present in the subsurface, but owing to the short half-life (3.8 days) of
its principal isotope, 222Rn, and the absence of parent uranium nuclides in
the atmosphere, radon is virtually absent in surface water which has reached
equilibrium with the atmosphere. Surveys of radon in surface streams and
lakes have, therefore, been useful in detecting the locations of places
where diffuse ground water enters surface waters (Rogers, 1958).
Inert Natural Gases; Because of their nonreactive and nontoxic nature,
noble gases are potentially useful tracers. Helium is used widely as a tra-
cer in industrial processes. It also has been used to a limited extent as a
ground-water tracer (Carter et al., 1959). Neon, krypton, and xenon are
other possible candidates for injected tracers because their natural concen-
trations are very low (Table 4.9). Although the gases do not undergo chem-
ical reactions and do not participate in ion exchange, the heavier noble
gases (krypton and xenon) do sorb to some extent on clay and organic
material.
122
-------
TABLE 4.9
Gases of Potential Interest as Tracers
Name of Gas
Argon
Neon
Helium
Krypton
Xenon
Carbon Monoxide
Nitrous Oxide
Approximate Natural Background Assuming
Equilibrium with Atmosphere at 20°C
(mg gas/liter water)
0.57
1.7 x 10"14
8.2 x 10~6
2.7 x ID""
5.7 x 10~5
6.0 x 10"6
3.3 x lO""
Maximum Amount In Solution Assuming 100Z
Gas at Pressure of i atra at 20°C
(mg gas/liter water)
60.6
9.5
1.5
234
658
28
1100
-------
The very low natural concentrations of noble gases in ground water make
them useful as tracers, particularly in determining ground-water velocities
in regional aquifers. The solubility of the noble gases decreases with an
increase in temperature. The natural concentrations of these gases in
ground water are, therefore, an indication of surface temperatures at the
time of infiltration of the water. This fact has been used to reconstruct
the past movement of water in several aquifers (Sugisaki, 1969; Mazor, 1972;
Andrews and Lee, 1979).
Fluorocarbons; Numerous artificial gases have been manufactured during
the past decade and several of these gases have been released in sufficient
volumes to produce measurable concentrations in the atmosphere on a world-
wide scale. One of the most interesting groups of these gases are the
fluorocarbons (Table 4.10, Figure 4.13a). The gases generally pose a very
low biological hazard, they are generally stable for periods measured in
years, they do not react chemically with other materials, they can be
detected in very low con-centrations, and they sorb only slightly on most
minerals. They do sorb strongly, however, on organic matter (Figure 4.13b).
Fluorocarbons have two primary applications. First, as an environmen-
tal tracer, they can be used in the same way tritium is used. Because large
amounts of fluorocarbons were not released into the atmosphere until the
late 1940's and early 1950's, the presence of fluorocarbons in ground water
indicates that the water was in contact with the atmosphere within the past
30 to 40 years and that the ground water is very young (Thompson and Hayes,
1978). The second application of fluorocarbon compounds is for injected
tracers (Thompson, Hayes, and Davis, 1974). Because detection limits are
so low, large volumes of water can be labeled with the tracers at a rather
124
-------
TABLE 4.10
Properties of Fluorocarbon Compounds
(from Thompson)
Common Name
Chemical
Formula
Boiling Point
at 1 atm (°C)
Solubility in Water
at 25°C (weight %)
Freon-11
Freon-12
Freon-113
CC13F
CC12F2
CC12F-CC1F2
CBrClF.
23.8
-29.8
47.6
-4.0
0.11
0.028
0.017
unknown
CBr2F2
C BrI-CBrF.
24.5
47.3
unknown
unknown
125
-------
Br-
BROMIDE
BCF CCIBrF2
F-ll CC13F
F-II3 C2CI3F3
.6 .8 1.0 1.2 1.4
1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
PORE VOLUMES
Figure 4.13(a). Laboratory experiments with fluorocarbon tracers and bromide flowing through
a column of quartz sand. Note the reduced concentration peaks, the "tailing" of the curves,
and the delay of the arrival of the peak concentrations relative to bromide caused primarily
by sorption of the fluorocarbon tracers on the quartz (data from Thompson and Hayes, 1978) .
-------
.05
.04
.03
.02
.Ol
c/cc
.07
.06
.05
.04
.03
.02
.01
.NaCI
.8 1.6 2.4 3.2 4.0
PORE VOLUME
NaCI
O .4 .8 1.2 1.6 2.O 2.4 2.8 3.2 3.6
PORE VOLUME
NaCI
CCI3F
234
PORE VOLUME
Figure 4.13(b). Tracer elution curves for laboratory experi-
ments with NaCI (common salt) and CCl-jF (Freon-11 of trichloro-
fluoromethane) using (A) Ottowa sand (no fine material), (B)
Yolo sandy loam (small amount of clay and some silt), and (C)
crushed coal. Note that NaCI curves are similar for all experi-
ments but that fine inorganic material reduces the peak concen-
tration and delays the breakthrough curve for CCloF. Crushed
coal, like most solid natural organic material, will adsorb most
of the CCl-jF and will release it very slowly to the water as it
passes through the test column. Data from Brown (1980).
127
-------
modest cost. Despite the problem of sorption on natural material and espe-
cially on organics (Figure 4.13), initial tests have been quite encouraging.
• Field Methods
Because the tracers are gases, it is most convenient to transport them
to the field in pressurized containers for tracer injection. The gas is
then bubbled into the water which is used for the tracer. For qualitative
work, this is a simple task. If the initial tracer concentrations are to be
established quantitatively, the gas injection should be made first into a
container where the gas and injected water are turbulently mixed and brought
into equilibrium at a known temperature and pressure. Provision should be
made to sample this labeled water just before it is injected into the aqui-
fer to ensure that the initial concentrations are constant during the test.
For most fluorocarbons, the tracers are dissolved first in the labora-
tory in methanol or some other solvent. The solvent is then injected as a
liquid into the water which is used for the tracer test.
The most critical aspect of the field work is the sample collection and
preservation. All gas tracers will be lost rapidly to the atmosphere unless
samples are sealed in metal or glass containers. Most plastic containers
are somewhat permeable to gas. Even certain types of glass are slightly
permeable to light gases. Furthermore, all seals and caps should be metal
or glass if fluorocarbons are being used because these compounds are sorbed
strongly on many greases and plastic sealers.
The problem of the storage and shipping of water with fluorocarbon tra-
cers is one of the major limitations of this method. Glenn Thompson, who
has worked extensively with these tracers, has developed an analytical sys-
tem for field use which largely eliminates the problem of sample integrity
(Thompson and Hayes, 1978).
128
-------
• Analysis
Although a number of inexpensive gas-analysis kits are available, these
are generally unsatisfactory for tracer studies. Quantitative analyses of
gas should be done either with a gas chromatograph (GC) or a mass spectrom-
eter (MS). Commonly a combined instrument, the GCMS, is used. The use of
these analytical instruments is standard and within the training of all good
analytical chemists. The difficult or nonstandard part of the analyses for
most chemists, however, is in the method by which the tracer gases are
removed quantitatively from the sample and fed into the analytical system.
For most laboratories, the development of a gas stripping system for the
samples is not a trivial task unless the chemists have had previous ex-
perience with the analysis of gas from water samples.
The measurement of fluorocarbon compounds is generally accomplished
with an electron-capture detector used in conjunction with a gas chromato-
graph. Special care should be taken that no plastic connectors and valves
are in contact with the sample being analyzed.
Stable Isotopes
Introduction: In this short section, we will look briefly at the use
of natural stable isotopes for water tracers. A detailed treatment of the
topic, however, is beyond the scope of this manual. The reader is referred
to several excellent summaries of the topic (Gat, 1971; Fritz and Fontes,
1980; and Ferronsky and Palyakov, 1982).
An isotope is a variation of an element produced by differences in the
number of neutrons in the nucleus of that element. Thus, hydrogen has two
stable isotopes. One isotope (*H) has only a proton and no neutron in the
-------
nucleus; the other (2H) has a proton plus a neutron in the nucleus. In
addition, hydrogen has an unstable, or radioactive, isotope (3H) which has
two neutrons in addition to the proton in the nucleus. An important char-
acteristic of isotopes is that isotopes of an element, for all practical
purposes, will react chemically in an identical way. For example, varia-
tions of sulfur isotopes (as 32S, 3l+S, and 36S) in the sulfate ion will not
affect the way in which the ion moves with the water. Thus, the water can
be labeled with the isotope without affecting significantly the movement of
the constituent.
In general, the uncertain ability to detect small artificial variations
of most isotopes against the natural background, the high cost of their
analysis, and the expense of preparing isotopically enriched tracers, means
that stable isotopes are rarely used for artificially injected tracer stud-
ies in the field. They are, however, quite widely used to detect sources of
pollution and to help determine areas of natural recharge.
Research into the topic of stable isotopes of various elements in natu-
ral waters is progressing rapidly, and the potential usefulness of these
isotopes to ground—water tracing will undoubtedly increase markedly in the
near future.
Hydrogen and Oxygen: The two stable isotopes of hydrogen (-H and 2R)
and the three stable isotopes of oxygen (1&0, 170, and 180) form part of the
water molecule, and analyses of their natural concentrations have been used
widely to help understand the movement of ground waters. Natural variations
in shallow ground water are generally related to variations within the orig-
inal recharge water coming from the surface. Because of the large differ-
ences in mass between the two hydrogen isotopes, they tend to fractionate
130
-------
whenever evaporation or condensation of water takes place. Other factors
being equal, waters with a higher 2H (commonly called deuterium) content
will be found near the coastlines, at low elevations, in warm rains, and in
water which has undergone partial evaporation such as in soil moisture dur-
ing periods of little rain or in saline lakes. Although mass differences
among oxygen isotopes are not as large as those of hydrogen, natural frac-
tionation of those isotopes also takes place. The variations in i80 and 170
contents of shallow ground water generally follow those of deuterium. That
is, if the water has a larger than normal 2H/-H ratio, it will generally
have also a larger than normal i80/i60 ratio (because 'l '0 is much less
abundant than either *80 or i60, it is rarely reported in routine isotopic
studies). This general relationship is defined by Craig's line and is shown
in Figure 4.14. Possible departures from this line can be caused by exces-
sive evaporation, by reactions between minerals and hot water, and by other
less important effects.
The most common use of studies of 2H and *80 has been to trace the
large-scale movement of groundwater and to locate areas of recharge
(Figure 4.15).
Nitrogen: The two abundant isotopes of nitrogen ( 14N and 15N) can vary
significantly in nature. Ammonia escaping as vapor from decomposing animal
wastes, for example, will tend to remove the lighter (14N) nitrogen and will
leave behind a residue rich in heavy nitrogen. In contrast, many fertili-
zers with an ammonia base will be isotopically light. Natural soil nitrate
will be somewhat between these two extremes. As a consequence, nitrogen
isotopes have been useful in helping to determine the origin of unusually
high amounts of nitrate in ground water.
131
-------
8 D %o
-I-100
o
-100
-200
-300
Water
from
coastal
areas
-Ocean water
Water
in mountains
and inland
Snow frorn
high mountains
in Arctic
and Anarctic
Direction of
shift due to
intense evaporation
18
Direction of 5 0 shift
due to high-temperature
interaction with minerals.
Snow at
South
r~ / Pole
Figure 4.14. Relationship between deuterium and oxygen-18 for
natural waters. Large arrow shows the direction of compositional
change found in geothermal waters where heavy oxygen found in
rock-forming minerals will exchange with the lighter oxygen in
normal ground water (data from Ferronsky and Polyakov, 1982).
132
-------
Santa Cotalina Mountains
N
Alluvial basin
LEGEND
8130%o
Summer runoff from
large bosin to south. -S to -7
Ground water recharged
from Santa Cruz River. -8
Ground water recharged
from summer runoff
from small arid basins
in low mountains. -7
Winter runoff from
high mountains. -10 to-12
Ground water recharged
from winter runoff
mixed with some
summer recharge. -10
Figure 4.15. Differences in the stable isotope of oxygen (o in
ground water of the Tucson basin in Arizona reflect different sour
of water. Because all values are negative, the larger number repre-
sents isotopically lighter water. Although the chemical character-
istics of the ground x\Tater are quite similar throughout most of the
basin, distinctive isotopic differences help to determine the origin
of recharge for the basin. Data are from several unpublished M.S.
theses at the University of Arizona. Diagram is not to scale.
133
-------
Most nitrogen in ground water will be in the form of the nitrate anion
(NO" ) or dissolved nitrogen gas (N ) from the atmosphere. Locally in zones
devoid of dissolved oxygen, the chemically reduced form (NHj1") may predomi-
nate. In general, nitrate will move as a conservative tracer and is an
important indicator of pollution. If nitrate concentrations exceed about
10 mg/1, the health of infant mammals including humans may be adversely
affected. Also, the presence of more than about 5 mg/1 of nitrate commonly
is an indirect indication of other forms of contamination including those
from chemical fertilizers and sewage.
Sulfur: Most dissolved sulfur within shallow ground water is bound
within the sulfate ion (SCT^). The stable sulfur isotopes (32S, 34S, and
35S) found in the sulfate ion will vary quite widely and, under certain cir-
cumstances, can be useful indicators of the origin of the sulfate. This is
particularly true if, for example, one wishes to distinguish sulfate orig-
inating from natural dissolution of gypsum (CaSO '2H 0) from sulfate orig-
inating from an industrial spill of sulfuric acid (f^SO^).
Carbon: Two stable isotopes of carbon (12C and 13C) and one unstable
isotope ( 11+C) are used in hydrogeologic studies. Most of the carbon dis-
solved in normal potable ground water is within the bicarbonate ion (HCO~ ) .
Contaminated water may also have large amounts of organic materials which
contain carbon. Other forms of carbon dissolved in natural water are car-
bonate (C0= ) and carbonic acid (H CO ) , the concentrations of which are
O £. 0
pH-dependent, and the gases carbon dioxide (C02) and methane
Most isotopic studies of carbon in water have been centered on 1UC
which will be discussed in a later portion of this chapter. Although not
as commonly studied as lifC, the ratio of the stable isotopes, 13C/12C, are
134
-------
potentially useful in sorting out the origins of certain contaminants found
in water. For example, methane (CH.) originating from some deep geologic
deposits is isotopically heavier than methane originating from near surface
sources (Figure 4.16). This contrast forms the basis for identifying aqui-
fers contaminated with methane from pipelines and from subsurface storage
t ank s.
Isotopes of Other Elements: The potential exists for the use of stable
isotopes of a number of other elements as natural tracers of water. Some of
these are chlorine (37C1 and 35C1), strontium (85Sr and 87Sr), boron (i°B
and ^3), and the isotopes of the noble gases. In general, studies of these
isotopes are related more to the determination of regional directions of
ground-water flow than to problems of the identification of sources of con-
tamination.
• Field Methods
Collection of field samples must take into consideration problems of
obtaining a representative sample as discussed in Chapter 3. Also, the
sample must be preserved so that isotopic fractionation does not take place
prior to analysis. For oxygen-deuterium samples, small glass bottles with
vapor-proof caps which hold about 20 to 50 nil are sufficient for most
purposes. For boron, nitrogen, carbon, and sulfur, a larger sample should
be taken. The size of the sample will depend on the water chemistry and the
analytical methods used. Generally, sample sizes are from 1 to 10 liters
for normal potable water. Samples should be stored in the dark and a growth
inhibitor should be added to water samples taken for boron, nitrogen, car-
bon, and sulfur analyses, because biological activity within the sample can
cause significant isotopic fractionation. Analyses of stable chlorine will
generally require samples of 1 to 2 liters of potable water and much less
135
-------
LJ
_J
CL
15 r
IO -
u_
o
UMBER
Ol
z
n
— i~-
i
i
!""' Bedrock """j
, r" , i n
i
J-]
Lr
Glacial
i drift ill]
-40
-50
-60
-70
-80
-90 %o
£C13 FROM CH4
Figure 4.16. Histogram showing composition of carbon
isotopes from methane from bedrock and from glacial
drift. The contrast in isotopic composition allows
the identification of methane from storage and from
pipelines which may leak out and contaminate ground
water. Natural methane generated in shallow aquifers
is much different isotopically than bedrock methane
that is distributed commercially. (Redrawn from Cole-
man et al., 1977).
136
-------
for saline water. Changes in the isotopic ratios of chlorine will not take
place under normal conditions of storage.
Analyses: Analyses of stable isotopes are made with expensive mass
spectrometers which require highly-trained technicians to run. Further
details are given in Appendix E.
Radionuclides
Introduction: This section includes a description of some of the
hydrogeologic applications of radioactive isotopes of various elements,
which are called collectively radionuclides.
In the early 1950's, great enthusiasm was evident for the use of radio-
nuclides both as natural, "environmental" tracers and as injected artificial
tracers. The environmental use has been expanded greatly until it is a
major component of many hydrochemical studies of today. In contrast, the
use of artificially injected radionuclides has all but ceased today in many
countries including the United States. Most use of artificially introduced
radioactive tracers in these countries is confined to carefully controlled
laboratory experiments or to deep petroleum production zones which are de-
void of potable water.
A brief explanation of some aspects of radioactivity is necessary be-
fore discussing isotopes of specific elements. For any radioactive element,
the radiation given off is in short, almost instantaneous, pulses which are
randomly distributed in time. If enough individual nuclei are considered,
however, the process of radioactive decay can be expressed as
137
-------
in which x is the number of nuclei present, t is time, and X is the decay
constant which is unique to each radionuclide. If xo is the number of
nuclei at zero time and xt is the number of nuclei at time t, then
xt
(12)
The half-life of a particular radionuclide is the time which is taken for
one-half the original number of nuclei to decay, or
h/2
The foregoing equations apply to all types of radioactive reactions even
though some reactions produce alpha particles (^He ions), others produce
beta particles (electrons, both negatrons and positrons), and still others
produce gamma rays (an electromagnetic radiation similar to X-rays). A
number of other types of radiation may also be produced but they will not
be discussed in this brief summary.
Injected Tracers: For a number of reasons, the detection and counting
of y-radiation is much easier than either B or a radiation. Radionuclides
which have a strong gamma emission are, therefore, commonly chosen for
tracers. A number of these radionuclides as well as others are listed in
Table 4.11. In addition, tracers are selected which can be injected into
ground water in a form which is highly mobile in the water phase. This
usually is either in a neutral or anionic form.
Most radioactive tracers are superior to other tracers because they can
be detected easily by field equipment in very small concentrations which are
far below levels that would alter the flow characteristics of the ground
water. Also, tracers can be selected which have half-lives so short that
they are essentially decayed after a few hours to a few days. Despite the
138
-------
TABLE 4.11
Commonly Used Radioactive
Tracers for Ground-Water Studies
Half-Life
y=year ,
d=day ,
Radionuclide Radiation h=hour)
2H f 12. 3y
32P 6~ 14.3d
51Cr -y 27. 8d
60Co jT,Y 5.25y
82Br g~,Y 35. 4h
85Kr g~,Y 10. 7y
i31l iT,-f 8. Id
i98Au g-,y 2.7d
Chemical Compound
H 0
Na2HPO,.
EDTA-Cr and CrCl,
EDTA-Co and K^Co (CN J
J b
NH^Br, NaBr, LiBr
Kr (gas)
I and KI
AuCl0
139
-------
demonstrated safety of many of the techniques and tracers, the complexity of
local and federal regulations makes their field use impractical in many
countries, including the United States.
Radioactive tracers, besides being used for tracers which move from one
well to another, have been used for studies of the local hydraulics near and
within a well. Radioactive gold (198Au), when mixed with water in a well,
will plate out on the downstream side of the well as the water moves through
the well. A directional counter will detect this concentrated radioactivity
and thus indicate the direction of water movement in the vicinity of the
well (Figure 4.17). Also, the rate of removal of the radioactivity from the
well water will be a function of the volume of water moving through the well
per unit time. Although giving only conditions near the well, this dilution
technique is useful in obtaining estimates of hydraulic conductivity accord-
ing to the following equations:
Co
in which
Ct = concentration of tracer in the well of time t;
Co = original concentration of tracer in the well at
time = 0; and
B = a factor which is constant for simple, steady-state
conditions.
If B is constant, then
Q
v
140
-------
DIRECTION OF FLOW
60
60
860 900 940
concentrotion in
cpm
60
Figure 4.17. The local direction of ground-water movement
as determined by the movement of a radioactive tracer within
a borehole. The hole was not pumped during the test. The
ground water is flowing under natural conditions and enters
the well from the west and leaves the well towards the east.
After release of the radioactive tracer, the gamma radiation
is measured in different directions by rotating a shielded
counter within the well. Although the surveys may be highly
useful, it must be remembered that flow directions within
the well are influenced by well-construction methods and by
local heterogeneities in the aquifer. The measured direc-
tions, therefore, may not give a reliable indication of
regional directions of ground-water flow. Diagram is re-
drawn from Rodriguez (1977) .
141
-------
in which
Q = the volume of water per unit of time flowing through
the well and V is the volume of water in the well.
For fully penetrating wells in isotropic and homogeneous aquifers,
Q = 2dmne:v (15)
d = the effective diameter of the well;
m = the saturated thickness of the aquifer;
ne = the effective porosity of the aquifer; and
v = the average velocity of the ground water outside
of the well (in the aquifer).
If the hydraulic gradient, i, of the ground water is known, then the
hydraulic conductivity, K, is given by
(16)
If the experiment has a duration which is 5% or more of the length of
the half-life of the radioactive tracer, then Equation (12) should be used to
correct for radioactive decay during the experiment. Thus in Equation (14),
Ct is the calculated concentration at time t assuming no radioactive decay
has taken place. It is not the actual observed concentration of radio-
activity.
In summary, the progressive dilution of a tracer in a well can be used
to obtain the hydraulic conductivity of an aquifer near the well provided
dimensions of the well are known and estimates can be made of the effective
porosity of the aquifer and the hydraulic gradient near the well.
142
-------
Atmospherically Distributed Radionuclides: A number of radionuclides
are present in the atmosphere from natural and artificial sources. Many of
these radionuclides will be carried into the subsurface by rain water. The
radionuclides of greatest interest are listed in Table 4.12. The most common
hydrogeologic use of these radionuclides is to obtain some estimate of the
average length of time ground water has been isolated from the atmosphere.
Because of dispersion in the aquifer and mixing in wells that sample several
hydrologic zones, a unique age of the ground water does not exist. Neverthe-
less, it can be commonly established that most or virtually all of the ground
water is older than some given limiting value. In many situations we can
say, based on atmospheric radionuclides, that the ground water was recharged
more than 1,000 years ago or that, in another region, all the ground water in
a given shallow aquifer is younger than 30 years.
Tritium, the radioactive isotope of hydrogen (3H) with a 12.4-year half-
life, was produced at low levels by natural processes prior to the detonation
of thermonuclear devices in the early 1950's. Since that time, atmospheric
tritium has been dominated by tritium from man-made sources. Most commonly,
tritium concentrations are measured in tritium units (TU) which is the number
of tritium nuclei per 1018 stable hydrogen nuclei. Prior to the 1950's,
natural levels in rain ranged from 5 to 15 TU, the exact number depending on
several local and regional factors. Owing to the decay of the tritium, water
recharged during the early 1950's will only have 0.8 to 2.5 TU today if the
water has been isolated from the atmosphere since that time. Thermonuclear
explosions increased local rainfall to more than 1,000 TU in the Northern
Hemisphere by the early 1960's (Figure 4.18). Tritium analyses of ground
water are used widely to determine the "age" of young ground water. In
general, ground water in the Northern Hemisphere which has more than about 5
143
-------
TABLE 4.12
Environmental Radionuclides
Radtonuclide Half-life Useful "age" range Comments
(years) (years)
3H 12.4 5-50 Analyses are done routinely at several laboratories'
Useful for Identifying young ground water In the
aubaurface•
1>4C 5730 500-30,000 Analyses are done routinely In several laboratories.
Sample collection and interpretation of results re-
quire experienced Isotope hydrologist.
JZSi 103 50-100 Analyses difficult and done by only a few labora-
tories in the world. Interpretation of results Is
difficult.
36C1 3 x 10s 5 x lQk-2 x 106 Analyses done in only two or three laboratories In the
world. Potentially an excellent radlonuclide with
which to study very old water. Also, sharp anthropo-
genic source In I960'a has produced a useful recent
hydrologlc tracer for ground water of recent origin.
39Ar 269 100-1,000 Sample collection and analyses are extremely diffi-
cult. Has been utilized in Europe but new techniques
are needed before the method can be applied widely.
8SKr 10.7 3-30 Almost all 85Kr Is from anthropogenic sources. Sample
collection and analyses are very difficult and have
been done successfully In only a few studies. Poten-
tially more useful than 3H because Increases in con-
centrations with time have been less erratic than
Increases of ^H.
-------
4000 r
350O
3OOO
~ 2500
2000
1500
IOOO
50O -
' -* 1 L
_J L
56 58 60 62 64 66 68 70 72 74 76 78 80 82
YEAR
Figure 4.18. Average annual tritium concentration of rain-
fall and snow for the states of Arizona, Colorado, New
Mexico, and Utah. During any single year, however, tritium
concentrations may vary by more than 300% with the maximum
concentrations in rainfall during the summer. In northcen-
tral United States and central Canada, concentrations have
been higher than those shown for the western states. Con-
centrations in precipitation along coastlines, in the trop-
ics, and in the Southern Hemisphere are generally much lower
than those shown here. (Diagram redrawn from Vuataz et al.,
1984).
145
-------
TU is less than 30 years. Very small amounts, 0.05 to 0.5 TU, can be
produced by natural subsurface processes, so the presence of these low levels
does not necessarily indicate water 40 to 60 years old or small amounts of
more recent water mixed with very old water.
The radioactive isotope of carbon, 11+C, is also widely studied in ground
water. Most 11+C in potable ground water is contained in the HCO ~ ion in the
water. Other carbon-bearing material dissolved in water such as CO , C0,=,
CH , H.CO , and organic acids may also contain variable amounts of 4C. As
a first approximation, the initial number of ^C nuclei per total carbon
nuclei, or Xo in Equation (2), in a water sample is considered to have been
constant prior to 1950 due to the almost constant natural production of ^C
in the atmosphere by cosmic radiation interacting with the atmosphere. If
the only source of 14C in the water is originally from the active biosphere,
then the ^**C which is measured in carbon from the water sample can be con-
sidered to be Xt in Equation (2). Because X is known from experimental work,
the "age" of the sample, or t, in Equation (2) can be determined directly.
In practice, however, the use of 14C is rarely as simple as just
described. Sources of old carbon, primarily from limestone and dolomite,
will dilute the sample. A number of processes, such as the formation of CH
gas or the precipitation of carbonate minerals, will fractionate the isotopes
and alter the apparent age. The complexity of the interpretation of 11+C
"ages" of water is so great that it should be attempted only by hydrochemists
specializing in isotope hydrology.
Despite the complicated nature of 14C studies, they are highly useful in
determining the approximate residence time of old water (500 to 30,000 years)
in aquifers. For certain practical problems, this information is essential
and cannot be obtained in any other way.
146
-------
Other radionuclides listed in Table 4.12 are not used routinely in
hydrogeologic work owing either to problems of sampling or to problems of
analyses. Of those listed, 36C1 will probably be used routinely in another
decade after the present analytical bottleneck is solved. The major advan-
tages of 36C1 are the ease of sampling, the stability of the sample in stor-
age, and the fact that 36C1 can give information concerning extremely old
water.
• Field Methods
Injected radioactive tracers are handled with great care to avoid radia-
tion exposure and to avoid sample contamination. Otherwise, they are gen-
erally treated as normal chemical tracers. Special down-hole devices to mea-
sure in-place tracer dilution for the application of Equations (14), (15),
and (16) are fabricated by the Institut fur Radiohydrometrie, Gesellschaft
fur Strahlenund Umweltforschung MBH, Neuherberg, Ingolstadter Landstrasse 1,
D-8042 Oberschleissheim, West Germany.
Field collection of samples for the determination of environmental
levels of tritium must be done with great care to avoid contamination from
the atmosphere from local sources of tritium such as watch dials, and from
high levels of tritium commonly present in laboratories. From two to four
liters of water are needed if anticipated tritium levels are below 15 TU.
Sample containers should be metal or high-quality glass. Some plastic con-
tainers are permeable to gases, so plastic containers are to be avoided
unless the properties of the plastic are known.
Field collection of samples for 14C is highly specialized and should be
done by individuals experienced with this type of sampling. For routine 14C
samples, large volumes of water (from 10 to 1,000 liters) are required and
147
-------
the carbon is extracted either by large batch or by flow-through systems
The use of the tandem accelerator mass spectrometric (TAMS) method for
analysis has greatly reduced the amount of carbon required so that one liter
of water or less can be used. Access to the TAMS system, however, is not
routine.
Samples for 36C1 analyses are relatively simple to obtain. About 30 mg
of chlorine should be available for the analysis. Most potable water con-
tains between 10 and 100 mg/1 of chloride, so a sample of a few liters of
water generally is enough. Silver nitrate, AgNO.., is mixed with the water
sample, and AgCl is formed. The AgCl precipitate is placed in an amber bot-
tle and stored out of sunlight and excessive heat until analyses can be com-
pleted.
• Analysis^
The analysis of radioactive materials is a highly specialized branch of
chemistry and is not easy to accomplish except where the field determination
of gamma radiation can be related directly to the concentration of injected
tracers. Scintillation counting using special liquid scintillation fluids is
normally required for beta emitters.
Environmental radionuclides such as tritium, lLkC, and 36C1 require very
special equipment for their determination. Low-level tritium is concentrated
by electrolysis and counted by liquid scintillation. A number of methods are
used to determine il+C. All processes are complicated. Many end with the
carbon in a gaseous form which is placed into counters designed to receive
gas. The TAMS method can be used for both llfC and 36C1 analyses. The accel-
erator used is a multi-million dollar instrument and only a few of these are
presently in operation.
148
-------
APPENDIX A
ADDITIONAL USES OF WATER TRACERS
The purpose of this manual is to describe ground-water tracing
techniques. However, tracers are widely used in other areas of hydrologic
study, such as surface water, the unsaturated (vadose) zone, and the atmos-
phere. Numerous engineering applications also involve tracer use, including
petroleum exploration, leak detection, flow of sewage, and biological and
medical research. A brief description of these uses is given with reference
articles.
Ground Water
Tracers have been used to determine the flow path, velocity, and resi-
dence time of solutes, and aquifer characteristics such as hydraulic con-
ductivity, dispersivity, and effective porosity. Ground-water velocity and
aquifer characterization studies have been described in the text.
Examples of flow path measurements are most numerous in karst studies.
The Water Tracer's Cookbook (Aley and Fletcher, 1976), published by the
Missouri Speleological Survey, is an excellent introduction to karst mapping
and characterization through use of a wide variety of tracers. Another
application of karst flow tracing is described by Caspar and Oncescu (1972),
and deals with water exchange between karst mines, depressed regions, and
ground water. Karst tracing has also been used to delineate catchment
boundaries (Smart, 1975). Flow path studies in non-karst regions include
evaluation of the movement of sewage in ground water (Sinton, 1980), and the
determination of the potential for chemical or bacterial pollution of a New
149
-------
Zealand aquifer (Thorpe, 1979). Vecchioli et al. (1972) studied the travel
of indicator bacteria through the Magothy aquifer in New York.
Residence time studies include the determination of ground-water
recharge using environmental isotopes (Vogel et al., 1974; Fontes and Fritz,
1975). Ground-water dating, involving the use of cosmic-ray and bomb-
induced radioisotopes, is a growing field of study (Davis and Bentley,
1982). Environmental isotopes have recently been used to demonstrate the
effect of ground water on storm runoff hydrographs (Sklash and Farvolden,
1979).
Surface Water
Tracers have been widely used in surface-water studies to determine
flow patterns (dispersion), flow volume, and time-of-travel (velocity).
Kilpatrick et al. (1967) described flow measurements with fluorescent
tracers. A more recent, general work on the subject is "Measurement of Time
of Travel and Dispersion in Streams by Dye Tracing" (Hubbard et al., 1982),
a handbook published by the U.S. Geological Survey.
Determination of flow patterns yields information concerning movement
of contaminants (such as factory effluents, radioactive waste, and sewage)
in streams (Caspar and Oncescu, 1972). Study of dispersion under turbulent
flow results in determination of eddy-diffusion coefficients. White (1981)
discussed estuary mixing through the use of environmental radionuclides.
Caspar and Oncescu (1972) reviewed the use of tracers in measuring flow
rates in natural streams, closed conduits, and reservoirs. Dilution studies
are used to find the time required for inflowing contaminants to be reduced
to acceptable levels.
150
-------
Storm runoff studies employ tracers to obtain travel time measurements
in order to help establish flood hydrographs. Smith (1973) noted the use of
environmental tritium in river recharge investigations.
Sediment transport is another aspect of surface-water systems which has
been studied with tracers. Elrick and Lawson (1969) looked at sediment
movement in rivers, irrigation canals, estuaries, harbors, and the open
ocean. River bank and bed erosion have also been investigated (Caspar and
Oncescu, 1972). White (1981) discussed the dating of sediments and surface
water with environmental radionuclides.
Soil
In the unsaturated zone, soils have been investigated extensively
through the use of various tracers. Infiltration, drainage, and evapo-
transpiration are fields of interest. Recent research includes: the use of
bromide as a tracer in the root zone of soils (Tennyson and Settergren,
1980); the use of radioactive tracers to determine the impact of deforesta-
tion on the soil profile (Ryckborst, 1981); and a general study of water
distribution and movement in the unsaturated soil profile (Ligon, 1980).
Atmosphere
Environmental and injected tracers are utilized in estimating the
travel of pollutants, studying precipitation and evaporation, and tracking
air motion on a global scale through the use of nuclear debris (Elrick and
Lawson, 1969).
151
-------
Petroleum Industry
The oil and gas industry has developed tracers for a number of oilfield
applications. Wagner (1977) described the use of chemical and radioactive
tracers for waterfloods and gas drives in the tertiary oil recovery process.
Some of the information to be obtained from diagnosis of interwell hetero-
geneities includes: identification of problem injection wells; directional
flow trends and fluid velocity; and delineation of flow barriers. Preferred
water and gas tracers are listed by Wagner (1977). Greenkorn (1962) also
compared waterflood tracers.
Additional Engineering Applications
Leak detection in water and sewer pipes, embankments, and dams is
another branch of tracer use (Caspar and Oncescu, 1972). Zuber et al.
(1979) discussed tracing of water leakage into salt mines, and Alburger
(1977) described leak testing with dyes as a non-destructive technique for
soils, sewers, electronics components, boilers, tanks, pipelines, etc.
Koerner et al. (1979) reported non-destructive tracer testing methods for
detecting dam seepage.
Sea-water intrusion around the foundation of a nuclear power plant was
modeled by Myer (1981), using I131 as a tracer. Sewage system tracing has
been performed by Renard (1982), and Aulenbach and Clesceri (1980) used
tracers in monitoring the land application of waste water. Finally, sani-
tary landfill leachate has been traced by Ellis (1980) and Murray et al.
(1981), using potassium (from the leachate) and injected lithium bromide,
respectively.
152
-------
Potential Uses
Radioactive, hazardous waste, and sanitary landfill disposal site
evaluations are likely to employ tracer test results. In addition, soluble
tracers can be mixed in dry form with wastes which are buried so that any
water percolating later through the waste will carry the tracer which in
turn could provide an early warning for the arrival of the bulk of the
slower moving and hazardous leachate from the waste.
153
-------
APPENDIX B
A DISCUSSION OF DISPERSION AND DIFFUSION
One of the purposes of many tracer tests is to obtain a value of the
aquifer parameter, dispersivity (a). The intent of this Appendix is to dis-
cuss briefly the theoretical background of the parameter, and to present
some current attitudes concerning dispersion.
The transport of a tracer or contaminant in a porous medium is analyzed
by some form of the convection-dispersion equation, introduced by Ogata and
Banks (1961), and discussed by Bear (1961a, 1969). Convection is the bulk
movement of water at the mean velocity of the flow system, u (where u equals
specific discharge divided by porosity, as defined in Chapter 2). Convec-
tion may be caused by differences in density of the water (natural convec-
tion), regional movement in the aquifer (advection), and the pumping of
wells (forced convection) (Sauty, 1980).
Dispersion is the mechanism which causes a solute to mix and spread to
positions which would not be expected by convection alone. Dispersion in
ground water is a combination of mechanical dispersion (mixing) and molecu-
lar diffusion, and it causes a dilution of the solute. Mechanical disper-
sion is due to variations in fluid velocity and to the tortuous flow paths
in the voids of the porous medium at the microscopic scale (Sudicky and
Cherry, 1979). On a larger scale, mixing is due to the presence of zones of
different permeabilities.
Molecular diffusion is caused by the kinetics of the molecules which
give rise to phenomena such as Brownian motion. Molecular diffusion is
often considered insignificant in magnitude in comparison with mechanical
dispersion for rapidly flowing ground water. In most tracer tests in porous
154
-------
media, diffusion is neglected because the rate of ground-water flow is too
high for pore-to-pore equalization of concentration (Perkins and Johnson,
1963). A reasonable value for the diffusion coefficient for non-adsorbed
species in porous media is 1 x 10" 1° m2/s (Freeze and Cherry, 1979), while
the dispersion coefficient is generally orders of magnitude larger.
Derivation of the Convection-Dispersion Equation
The convection-dispersion equation used in contaminant transport model-
ing is based on Fick's first and second laws. Formulated by analogy to heat
conduction, the first law states that the flux of a diffusion or dispersing
substance in a given direction is directly proportional to the concentration
gradient in that direction. The negative sign indicates that flux is posi-
tive in the direction of decreasing concentration. In the following text,
dimensions are given in brackets.
F - - D -
x ~ x 3x'
M
where FX = mass flux [——] in the x direction;
D = coefficient of proportionality [•=— ]; and
c = concentration [—1.
L3
Fick's second law is derived from the law of conservation of mass, as
applied to the first law. It states that:
9c
It
155
-------
~ 2 ^ 2 "
where V2 = -?— i + -?— j +
8x2 9y2 3z2
k
This assumes that D Is constant, while it is actually a function of tempera-
ture, concentration, and other factors.
The convection-dispersion equation for a non-reactive solute is stated,
in one-dimensional form, as:
— = -u — + D
3t 9x
where
L2
D = coefficient of dispersion [=— ] ;
u = average linear flow velocity.
This assumes that flow is parallel to the x direction, with steady-state
velocity, u. It also assumes that the fluid is incompressible.
The coefficient of dispersion, D, may be thought of as a correction
factor which describes the variation of solute distribution about the mean.
The coefficient, D, is a combination of the effects of hydrodynamic disper-
sion and molecular diffusion.
D = oLu = %
Here, otL = longitudinal dispersivity (in the x direction) [L] ; and
L2
D^j = molecular diffusion coefficient [— ] .
The term "dispersivity" was introduced by Scheidegger (1954). This
parameter has components in three orthogonal directions. The longitudinal
dispersivity is in the direction of flux. Horizontal transverse dispersivity
156
-------
may be called "lateral dispersivity", and vertical transverse dispersivity
may be referred to as "vertical dispersivity." In laboratory experiments,
the transverse dispersivities are generally 5 to 20 times smaller in magni-
tude than the longitudinal dispersivity (Freeze and Cherry, 1979).
Solution of the Convection-Dispersion Equation
The one-dimensional solution of the convection-dispersion equation for
a step-function input of tracer into a semi-infinite aquifer with natural
flow velocity (Ogata and Banks, 1961) is:
c/c0(x,t) -1 [erfc (x "J^ + exp (ux/D) erfc (2L±« £)] (21)
2 2 -y Dt 2 "Y Dt
where
c/c0 = normalized concentration (relative to source):
x = distance from the measuring point to the source;
u = average linear velocity;
t = time;
D = dispersion coefficient; and
erfc = the complimentary error function.
The boundary conditions are:
c(x<0, t) = 0, for all t
c(0, t) = C0, for all t > 0
c(», t) = 0, for all t
The solution above can be approximated, after a short period of time,
by:
157
-------
c/c0 = - erfc [X -^5] (22)
0 2 2-yDt
This equation can be solved for various boundary conditions, flow
regimes, and types of injection (e.g., uniform flow, radial flow, continuous
injection, slug injection). Fried (1975) provided a number of solutions, and
Sauty (1977) developed type curves for uniform or radial flow to characterize
response to continuous or instantaneous pulse input at a point. Lenda and
Zuber (1970) developed analytical solutions in normalized form for different
measurement geometries. They presented type curves for point injection and
line injection in an infinite aquifer. Sudicky and Cherry (1979) developed
type curves for a finite—width pulse injection.
Hoopes and Hareleman (1967a) presented a general equation describing
the nonsteady-state concentration of a tracer during plane radial flow.
Analytical solutions to this equation for a constant input concentration have
been given in that paper and by Gelhar and Collins (1971). These solutions
can be used for single-well injection/withdrawal tests.
For a two-well tracer test, Webster et al. (1970) and Grove and Beetem
(1971) provided solutions. The tracer addition can be continuous or a pulse,
and recirculation can be accounted for.
Measuring Dispersivity
The error function is related to the normal distribution * ($), as:
* (z) - i [1 + erfc (-^r)] (23)
This holds for tables of the normal distribution with negative infinity
as the lower limit.
158
-------
Then,
c/c0 = 1 -
-------
TABLE B.I
Values of Dispersivities
Measured by Various Methods
Type of
Aquifer
Alluvial
Single-Well Injection Withdrawal Test
Location
oL
(meters)
Lyons, France
0.1-0.5
Reference
Fried, 1975
Multiple-Well Tracer Test (including two-veil tracer tests)
Distance Between
Injection and
Type of
Aquifer
Chalk
Alluvial
Alluvial
Fractured
dolomite
Fractured
carbonate
Fractured
crystalline
Location
Dorset, England
Lyons, France
Eastern France
Carlsbad, NM
So. Nevada
Savannah River
Plant, S.C.
Observation Hells
(meters)
8
6 & 12
6 & 12
55
121
538
cL
(meters)
3.1
4.3
11.0
38.0
15.0
134.0
Reference
Ivanovich and
Smith, 1978
Fried, 1975
Fried, 1975
Grove and
Beeten, 1971
Classen and
Cordes, 1975
Webster et al. ,
1970
Type of
Aquifer
Single-Well Tracer Test with Surface Geophysics
Location
Distance Traveled aL aT
by Tracer (meters) (meters) (meters)
Reference
Alluvial Lyons, France
- 80 m
5-12 0.009-14.5 Fried, 1975
160
-------
TABLE B.I (continued)
Dlspersivities Measured on a Regional Scale by Model Calibration
Type of
Aquifer
Alluvial
Limestone
Alluvial
Alluvial
Glacial
deposit
Basalt
Location
Lyons, France
Brunswick, GA
Rocky Mtn.
Arsenal, CO
Arkansas River
Valley, CO
Long Island, NY
Snake River
Approximate
Distance Traveled al«
by Solute (meters) (meters)
1,000 12
1,500 61
4,000 30
5,000 30
1,000 21.3
4,000 91
Reference
4 Fried, 1975
18 Bredehoeft &
Finder, 1973
30 Konikow, 1977
9 Konikow 4
Bredehoeft,
1974
4.3 Finder, 1973
137
Robertson,
1974
161
-------
Pickens et al. (1976) suggested that large dispersivities obtained from
analysis of two-well tracer tests are a result of mixing of water from dif-
ferent levels, which occurs at the well bore.
Most researchers feel that the primary cause of the scale effect is the
heterogeneity of an aquifer (Warren and Skiba, 1964; Matheson and de Marsily,
1980; and Gelhar et al., 1979). Recent research indicates that, for certain
hydraulic conductivity distributions, the longitudinal dispersivity approaches
a constant at large time or large mean travel distance. Gelhar et al. (1979)
suggested an improved form of the convective-dispersive transport equation
which incorporates the statistical properties of the hydraulic conductivity
distribution. However, the traditional convection-dispersion equation and its
solutions continue to be used to obtain values of dispersivity until a better
alternative is found.
162
-------
APPENDIX C
FACTORS TO CONSIDER IN TRACER SELECTION
Determination of:
PURPOSE OF STUDY
flow path
velocity (solute)
velocity (water)
porosity
dispersion coefficient
distribution coefficient
Delineation of contaminant plume
Recharge
Dating
Tracer Type to be Used
Nonconservative
Conservative
Conservative
Conservative
Nonconservative
Constituent of plume
Environmental isotope
or anthropogenic
compound
Radionuclides
AVAILABLE FUNDS
Manpower and equipment to run tests to completion (e.g., drilling, tracer
cost, sampling, analysis).
TYPE OF MEDIUM
Karst
Porous media (alluvium, sandstone,
soil)
Fractured rock
Tracer Type
Fluorescent dyes, spores,
tritium, as well as
other tracers
Wide range of choices.
Dyes and particulate
material are rarely
useful.
Wide range of choices.
Dyes and particulate
material only
occasionally useful.
163
-------
STABILITY OF TRACER
Distance from injection to sampling Must be stable for length
point of test and analysis
Approximate velocity of water and
approximate estimate of time re-
quired for test, given: distance
from injection to sampling point,
porosity, thickness of aquifer
DETECTAB1LITY OF TRACER
Background level
Dilution expected in test (function of type
of injection, distance, dispersion, porosity,
and hydraulic conductivity)
Detection limit of tracer (ppm, ppb,
PPt)
Interference due to other tracers and
natural water chemistry
DIFFICULTY OF SAMPLING AND ANALYSIS
Factors to Consider Example of Difficult Tracer
Availability of tracer Radioactive (must have special
permits)
Ease of sampling Gases (will escape easily from
poorly sealed container)
Availability of technology for Cl-36 (only one or two labora-
and ease of analysis tories in the world can do
analyses)
164
-------
PHYSICAL/CHEMICAL/BIOLOGICAL PROPERTIES OF TRACER
Density, viscosity May affect flow (e.g., high
concentrations of Cl~)
Solubility in water Affects mobility
Sorptive properties Affects mobility
Stability in water Affects mobility
Physical Chemical Biological
radioactive decomposition degradation
decay and precipi-
tation
PUBLIC HEALTH CONSIDERATIONS
Toxicity
Dilution expected
Maximum permissible level—determined by federal, state, provincial,
and county agencies.
Proximity to drinking water
165
-------
Summary of Some Important Tracers
Tracer
Characteristics
A. Particulates
Spores
Bacteria
Viruses
Used in karst tracing; inexpensive
Detection: high, multiple tests possible
by dying spores different colors
Low background
Moderately difficult sampling and analysis
(trapping on plankton, then microscopic
identification and counting)
No chemical sorption
May float on water, travels faster than
mean flow rate
Most useful for studying transport of
microorganisms
Detection: highly sensitive
Sampling: filtration, then incubation and
colony counting
No diffusion, slight sorption
Detection: highly sensitive
Sampling: culturing, colony counting
Some sorption
Smallest particulate
B. Ions (Non-radioactive,
excludes dyes)
Chloride
Bromide
Conservative
Inexpensive
Stable
Detection: 1 ppm by titration, electrical
conductivity, or selective ion electrode
High background may be problematic
In large quantities, affects density which
distorts flow
No sorption
Inexpensive
Stable
Detection: 0.5 ppm by selective ion
electrode
Low background
No sorption
166
-------
Tracer
Characteristics
C. Dyes
Rhodamine WT
Fluorescein
Used in karst and highly permeable sands
and gravels
Inexpensive
Moderate stability
Detection: 0.1 ppb by fluorimetry
Low background fluorescence
Moderate sorption
Properties similar to Rhodamine WT, except:
Degraded by sun
"Chlorella" bacteria interferes
High sorption
D. Radioactive Tracers
Tritium
EDTA-
High stability
Detection: > 1 ppt by weak 3 radiation
Varying background
Complex analysis (expensive field and
lab equipment)
Half-life = 12.3 years
Radiation hazard
Handling and administrative problems
No sorption
High stability
Detection: high sensitivity by measuring
3 and a emission
Background negligible
Complex analysis
Half-life = 8.2 days
Radiation hazard
Sorption on organic material
Moderately stable (affected by cations)
Detection: highly sensitive, by radiation
or post-sampling neutron activation
analysis
No background
Half-life = 28 days
Radiation hazard
Little sorption
167
-------
Tracer
Characteristics
82Br
High stability
Detection: high sensitivity by measuring
B emission
No background
Half-life = 35 hours
Radiation hazard
No sorption
E. Other Tracers
Fluorocarbons
Organic anions
High stability
Detection: 1 ppt by gas chromatography
with electron capture detection
Low background
Difficult to maintain integrity of
samples
Non-degradable, volatile, low solubility,
strong sorption by organic materials
Low toxicity
Detection: few ppb by HPLC
Low background
Expensive analysis
Very low sorption
Low toxicity
168
-------
APPENDIX D
CHEMICAL SUPPLY COMPANIES
A list of general chemical suppliers is provided, followed by a more
specific list according to type of tracer. It is recommended that several
companies be contacted, as prices can be quite variable. Prices are not
quoted here because they are subject to change. Current prices can be ob-
tained from the supplier by requesting a catalogue and price list, or by
telephone inquiry.
169
-------
General Chemical Supplies
Company
J.T. Baker Chemical Company*
222 Red School Lane
Phillipsburg, New Jersey 08865
Eastman Kodak Company*
343 State Street
Rochester, New York 14650
Fisher Scientific Company*
711 Forbes Avenue
Pittsburgh, Pennsylvania 15219
Hach Company*
P.O. Box 389
Love land, Colorado 80537
LaMotte Chemical Products Company*
P.O. Box 329
Chestertown, Missouri 21620
Union Carbide Corporation*
270 Park Avenue
New York, New York 10017
Telephone
(201) 859-2121
(716) 722-2915
(412) 562-8300
(303) 669-3050
(301) 778-3100
(212) 551-3763
Bacter iophage
American Type Culture Collection**
12301 Parklawn Drive
Rockville, Maryland 20852
Dyes and Biological Stains
Eastman Kodak Company
343 State Street
Rochester, New York 14650
Hach Company
P.O. Box 389
Loveland, Colorado
80537
E.I. du Pont de Nemours and Company, Inc.*
1007 Market Street
Wilmington, Delaware 19898
(716) 722-2915
(303) 669-3050
(302) 774-2421
Sources: * Analytical Chemistry Lab Guide, 1982
** Water Tracer's Cookbook (Aley, 1976)
*** Personal Communication (Thompson and Bentley, 1983)
170
-------
Fluorescent Dyes
Company Telephone
Aldrich Chemical Company, Inc.* (414) 273-3850
940 W. St. Paul Avenue
Milwaukee, Wisconsin 53233
Pylam Products Company, Inc.**
95-10 218th Street
Queens Village, New York 11429
E.I. du Pont de Nemours and Company, Inc. (302) 774-2421
1007 Market Street
Wilmington, Delaware 19898
Gases
Allied Chemical Corporation* (201) 455-4400
Specialty Chemicals Division
P.O. Box 2064 R
Morristown, New Jersey 07960
Union Carbide Corporation (212) 551-3763
270 Park Avenue
New York, New York 10017
AIRCO Industrial Gases* (201) 464-8100
575 Mountain Avenue
Murray Hill, New Jersey 07974
Matheson (201) 933-2400
P.O. Box 85
932 Paterson Plank Road
East Rutherford, New Jersey 07073
Halogens
Alfa Products* (617) 777-1970
Thiokol/Ventron Division
152 Andover Street
Danvers, Mississippi 01923
Edmund Scientific Company* (609> 547-3488
7082 Edscorp Building
Barrington, New Jersey 08007
171
-------
Isotopes (Stable and Radioactive)
Company Telephone
Monsanto Company (314) 694-1000
800 N. Lindbergh Boulevard
St. Louis, Missouri 63166
Alfa Products (617) 777-1970
Thiokol/Ventron Division
152 Andover Street
Danvers, Mississippi 01923
Edmund Scientific Company (609) 547-3488
7082 Edscorp Building
Harrington, New Jersey 08007
Lycopodium Spores
Carolina Biological Supply Company**
Burlington, North Carolina 27215
Lithium
Foote Mineral Company*** (215) 363-6500
Rt. 100
Exton, Pennsylvania 19341
Lithium Corporation*** (213) 728-6658
Fluorinated Benzoic Acids
Saber Laboratory, Inc.*** (312) 998-5950
Box 232
Morton Grove, Illinois 80039
Aldrich Chemical Company, Inc. (414) 273-3850
940 W. St. Paul Avenue
Milwaukee, Wisconsin 53233
172
-------
APPENDIX E
ANALYTICAL METHODS FOR THE DETECTION OF TRACERS
Electrical Conductance
An indication of the total dissolved ionic constituents can be obtained
by determining the capability of the water to conduct an applied electrical
current. The relative change in the ability of the ground water to conduct
an electrical current (above the background resistivity prior to injection
during an ion tracer test) will allow the determination of breakthrough time
(travel time) of the tracer in the flow field. The ability of a solution to
conduct an electrical current is a function of the concentration and charge
of the ions in solution and of the rate at which the ions can move under the
influence of an electrical potential. Conductivity or velocity of the ions
is also a function of temperature; thus, it is important to adjust the con-
ductivity readings for any change in temperature.
The device most commonly used for measuring electrical conductivity is
a conductivity meter, read in micromhos. An alternating current is estab-
lished between two points in the flow field and the conductivity (inverse of
resistivity) is measured. A plot of the time versus resistivity or conduc-
tivity readings will indicate the breakthrough time of the tracer. This
technique is very inexpensive and simple to use with various ionic species.
The concentration of the tracer passing through a system at the breakthrough
point cannot be determined by this method. It will, however, provide a
quick method to determine when to sample so that concentration of tracer at
the inflection point (peak conductivity) can be determined analytically.
173
-------
Specific Ion Electrode
Specific ion electrode analysis (see additional discussion under
"Coulometric Techniques") is similar to pH measurement with a pH meter.
Like the pH meter which measures the H*" ion, this technique is ion-specific
and thus, given data from an ionic tracer test, the concentration of the
tracer can be determined using a calibration curve (millivolts versus mg/1).
The reading is a function of temperature, type of ions present, and concen-
tration of various ions, particularly the ion being measured. Specific ion
electrodes can be used in the field or samples can be taken and analyzed by
this method in the laboratory.
Many pH meters used in the field can also read millivolts from specific
ion electrodes. The electrode should be checked using a standard before
initial use and should be checked daily during regular use. This method is
a fast and inexpensive technique for ionic tracers which have concentrations
of about 0.05 mg/liter or greater. Commonly, ions different than those
being measured will produce part of the measured voltage, so the electrodes
should be used with standard solutions having a composition similar to the
water sample being measured.
Titration
Titration is the procedure by which a solution of known concentration
(standard solution) is added to a water sample of unknown tracer concentra-
tion until the chemical reaction between the two solutes is complete. The
point at which stoichiometrically equivalent quantities of substance have
been brought together is known as the equivalence point of the titration,
which is usually indicated by a change in color produced by an added dye.
174
-------
In acid-base titrations, organic dyes known as acid-base indicators are used
for this purpose. A pH meter can be used instead of a colormetric pH indi-
cator if greater precision is needed. The titration method of analysis
varies in complexity based on the type of chemical tracer involved, and is
very time-consuming if a large number of samples require analysis. Examples
of tracers which can be analyzed by titrimetric techniques include Cl~, I~,
SCN~, N03~, and SO^".
Laboratory Culturing
The analysis of various bacteria, bacteriophage, and yeast as ground-
water tracers requires sample collection in sterile containers (in order to
minimize the potential of sample contamination by normal soil and water
microorganisms) and the preparation of specific media on which to assay or
culture the desired species. These microbial tracers are usually selected
because of their ease of identification by a microscope on prepared media,
or because they are "marked" by such characteristics as antibiotic resis-
tance.
Once samples are collected, known volumes obtained from serial dilu-
tions of the samples are filtered through membrane filters. These filters
are then placed on prepared nutrient media plates (i.e., agar-agar or mold
broth for yeast) and maintained at the optimum growth temperatures either in
an incubator or at room temperature for the appropriate species-specific
time period. The plates are then analyzed under a microscope for the char-
acteristic markers such as pigmented colonies or other traits. In the case
of bacteriophage, samples can be frozen at the study site and analyzed at a
later date.
175
-------
Microscopic Inspection of Spores
Various species of spores (i.e., Lycopodium) used as ground-water
tracers are injected into the flow system at locations such as sink holes
and are trapped with plankton nets at potential resurgencies. The spores
(typically marked by dyes) are then examined and counted under a microscope.
Colorimetric Techniques
Analysis by colorimetric methods consists of comparing the extent of
absorption of radiant energy at a particular wavelength by a solution of the
test material with a series of standard solutions. Work with visual compar-
ators requires simple equipment, but is subject to the vagaries of the human
eye; in particular, fatigue and unavoidable low sensitivity under 450 nm and
above 675 nm. The precision of measurement by unaided visual observation is
always less than that attainable with photoelectric instruments. Such
instruments, including filter photometers, are suitable for many routine
methods that do not involve complex spectra. Precise work is done with a
spectrophotometer which is able to employ narrow band-widths of radiant
energy and which can handle absorption spectra in the ultraviolet region if
equipped with fused silica optics.
The limitations of many colorimetric procedures lie in the chemical
reactions upon which these procedures are based. Although very few reac-
tions are specific for a particular substance, many reactions are quite
selective, or can be rendered selective through the introduction of masking
agents, control by pH, use of solvent extraction techniques, adjustment of
oxidation state, or by prior removal of interferents (Dean, 1969). Both the
176
-------
color-developing reagent and the absorbing product must be stable for a
reasonable period of time.
Numerous ground-water tracers can be analyzed by colorimetric tech-
niques, specifically, the large class of organic dyes (see Chapter 4).
Fluorometry
Fluorometric analysis is a photoluminescent method in which the elec-
tronic state of a molecule is elevated by absorption of electromagnetic
radiation, and as a consequence, the molecule emits light in order to reduce
its energy and return to the ground electronic state. With the exception of
X-ray fluorescence, most of the work lies in the wavelength region between
2000 and 8000 angstroms. Fluorescence provides two kinds of spectra for
identification, the excitation and emission spectra.
Instruments used for fluorometric analysis range from simple filter
fluorometers to highly sophisticated spectrophotofluorometers. Each will
contain four principal components: (1) a source of excitation energy; (2) a
sample cuvette; (3) a detector to measure the photoluminescence; and (4) a
pair of filters or monochromators for selecting the excitation and emission
wavelengths (Willard, 1965).
Fluorescence measurements usually are made by reference to some arbi-
trary chosen standard. The standard is placed in the instrument and the
circuit is balanced with the reading scale at any chosen setting. Without
readjusting any circuit components, the standard is replaced by standard
solutions of the test material and the fluorescence of each recorded.
Finally, the fluorescence of the solvent and cuvette alone is measured to
establish the true zero concentration.
177
-------
Measurement of fluorescent intensity permits the quantitative determi-
nation of inorganic and organic species in trace amounts. Such ground-water
tracers as dyes can be analyzed by this method. The technique is also very
sensitive; the lower limits for the method frequently are less than those
for the absorption method by a factor of ten or better, and are in the range
of a few thousandths to one-tenth of a part per million.
Coulometric Techniques
Coulometric methods of analysis measure the quantity of electricity
(in coulombs) required to carry out a chemical reaction. The coulomb is
that amount of electricity which flows during the passage of a constant
current of one ampere for one second. Reactions may be carried out either
directly by oxidation or by reduction at the proper electrode (primary
coulometric analysis), or indirectly by quantitative reaction in the solu-
tion with a primary reactant produced at one of the electrodes (secondary
coulometric analysis). In either case, the fundamental requirement of
coulometric analysis is that only one overall reaction must occur, and that
the electrode reaction used for the determination proceeds with 100% current
efficiency.
There are two general techniques used in coulometry. One method, the
controlled-potential method, maintains a constant electrode potential by
continuously monitoring the potential of the working electrode as compared
to a reference electrode. The current is adjusted continuously to maintain
the desired potential. The second method, known as constant-current coulom-
etry, maintains a constant current throughout the reaction period. In this
method, an excess of a redox buffer substance must be added so that the
178
-------
potential does not rise to a value which will cause some unwanted reaction
to occur. The product of the electrolysis of the redox buffer serves as an
intermediate in the reaction, and must react quantitatively with the
substance to be determined.
Coulometric techniques are particularly useful in trace analyses, being
accurate in the range from milligram down to microgram quantities. This
technique can be used for various ionic tracers such as Cl , Br , I , or
SCN~.
Liquid Chromatography
Chromatography encompasses a diverse group of separation methods used
to separate, isolate, and identify components of mixtures which might other-
wise be resolved with great difficulty. In its broadest sense, chromatog-
raphy refers to processes that are based on differences in rates at which
individual components of a mixture migrate through a stationary medium under
the influence of a moving phase. This rate of movement of a specific com-
ponent is referred to as its retention time. Liquid chromatography is a
specific class of chromatography where the mobile phase (injected sample) is
a liquid and, depending on the specific method, the stationary phase is
either liquid or solid.
In order to employ chromatographic techniques, the components to be
separated must be soluble in the mobile phase. They must also be capable
of interacting with the stationary phase either by dissolving in it, by
being absorbed by it, or by reacting with it chemically. Thus, during the
separations, the components become distributed between the two phases.
The most widely used chromatographic method is elution analysis. In
the elution method, a small portion of sample is injected and introduced at
179
-------
the head of the separation column. A differential migration process occurs
in which each component of the sample interacts with the stationary phase,
retarding its flow at a rate characteristic of that specific component down
the length of the column. The time required for a specific component to
reach the end of the column, which is referred to as the retention time, is
a function of the distribution coefficient of the component. The concentra-
tion of each component present is then determined based on the comparison of
its retention time to that of a known concentration standard.
There are numerous chromatographic methods employing a liquid mobile
phase. These include partition, adsorption, ion exchange, paper, and thin-
layer chromatography. All are based on the same chromatographic principles
of separation and isolation as described previously, with variation in the
constituents of the mobile and stationary phases.
Liquid chromatography can be used for the analysis of a wide range of
tracers at very low detection levels. Fluorinated organic acids can be
detected down to concentrations from 1 ppm to 0.01 ppb using reverse phase
and ion exchange high-pressure liquid chromatography (Stetzenbach, 1982).
Halide tracers including Cl~, Br~, and I" can be analyzed using liquid (ion
exchange) chromatography.
Gas Chromatography
In gas chroraatography, the components of a vaporized sample are frac-
tionated as a consequence of partition between a mobile gaseous phase and a
stationary phase which is either a liquid held on a solid support (gas-
liquid chromatography) or a solid (gas-solid chromatography). In principle,
gas and liquid chromatography techniques differ only in that the mobile phase
in the former is a carrier gas rather than a liquid.
180
-------
In gas chromatography, the sample containing the solutes is injected
into a heating block where it is immediately vaporized and swept as a plug of
vapor by the carrier gas stream into the column inlet. The solute components
having a finite solubility in the stationary phase distribute themselves
between that phase and the gas according to the equilibrium law. This par-
titioning process occurs repeatedly as the sample is moved toward the outlet
by the carrier gas- Each component (solute) will travel at its own rate
through the column, and consequently, a band corresponding to each solute
will form. The bands will separate to a degree which is determined by the
partition ratios of the solutes and the extent of band spreading. The sol-
utes are eluted, one after another, in the increasing order of their parti-
tion ratios and enter a detector attached to the column exit. If a recorder
is used, the signals appear on the chart as a plot of time versus the compo-
sition of the carrier gas stream. The retention time or time of emergence of
a peak identifies the component, and the peak area reveals the concentration.
of the component in the sample. Although the gas chromatographic method is
limited to volatile materials (about 15% of all organic compounds), the
availability of gas chromatographs working at temperatures up to 450°C, pyro-
lytic techniques, and the possibility of converting many materials into a
volatile derivative extend the applicability of the methods (Willard, 1965).
Gaseous tracers such as f luorocarbons (i.e., Cd-^F ar>d CC1 F ) are
easily detectable in low concentration of between 1 and 100 parts per tril
lion by gas chromatographic methods.
Mass Spectrometry
Mass Spectrometry techniques involve converting the compounds of a
sample into charged ionic particles consisting of the parent ion and ionic
181
-------
fragments of the original molecule, and resolving them according to their
mass/charge ratio. A mass spectrometer consists generally of four units:
(1) the inlet system; (2) the ion source; (3) the electrostatic accelerating
system; and (4) the detector and readout system. This ionization process
results in a mass spectrum which is a record of the numbers of different
kinds of ions. The relative numbers of each type of ion are characteristic
for every compound, including isomers.
Sample size requirements for solids and liquids range from a few milli-
grams to submicrogram quantities as long as the material can exist in the
gaseous state at the temperature and pressure existing in the ion source.
The average sample size for routine gas analysis is about 0.1 ml at standard
conditions, but with special instrumentation, samples of 10 8 ml can be
analyzed (Skoog, 1980). Information useful for elucidating chemical struc-
tures and for accurate determination of molecular weight can be obtained from
the mass spectra literature. Mass spectra can also be employed for the quan-
titative analysis of complex mixtures. In such cases, the magnitude of ion
currents at various mass settings is related to concentration.
Mass spectrometry is often used in conjunction with gas chromatography
techniques. Such is the case for the analysis of fluorinated organic acids
used as ground-water tracers. Lithium salts used for tracing are often
analyzed by mass spectrometry. Stable isotopes (deuterium, tritium, •1-L+C,
sulfur, etc.) are also analyzed using mass spectrometry.
Gamma-Ray Emission
Gamma emission is one type of radiation encountered in radiochemical
analysis of both natural and artificial radioactive isotopes which have been
182
-------
used as tracers in hydrologic systems. There are three general types of
radiochemical methods: (1) activation analysis; (2) isotope dilution; and
(3) radiometric analysis. In activation analysis, activity is induced in
one or more elements of the sample by irradiation with suitable particles
and the resulting radioactivity is measured. In isotope dilution, a pure but
radioactive form of the substance to be determined is mixed with the sample
in a known amount. After equilibrium, a fraction of the component is iso-
lated and its activity analyzed. In a radiometric analysis, a radioactive
reagent is employed to separate completely the component from the bulk of the
sample. The activity of the isolated portion is then measured.
Gamma rays (high-energy photons) are monoenergetic and have a penetrat-
ing power which is much greater than that of either alpha or beta particles,
but a lower ionizing power. The gamma-ray emission spectrum, in contrast to
the alpha and beta emission spectra, is characteristic for each nucleus and
is thus useful for identifying radioisotopes (Skoog, 1980).
One type of detection method for gamma-ray emission is photon counting.
This is a signal processing method where the individual pulse of electricity
produced as a quantum of radiation is absorbed by the transducer and counted.
The power of the beam is then recorded digitally in terms of counts per unit
of time. This operation requires rapid response times for the detector and
signal processor with respect to the rate at which quanta are absorbed by the
transducer. Thus, photon counting is only applicable to beams of relatively
low intensity.
Other types of detectors include gas-filled detectors, the geiger tube,
proportional counters, ionization chambers, and semiconductor detectors. In
most techniques, interference from alpha and beta radiation is readily
183
-------
avoided by filtering the beam with a thin window of aluminum or mylar.
Radioactive ground-water tracers such as *3*I can be analyzed by gamma-ray
emission.
Beta Particle Emission
Beta particle emission is another type of radiochemical analysis. Beta
particles interact primarily with the electrons in the material penetrated by
the particle. The molecules may be dissociated, excited, or ionized. Beta
particles are produced within a nucleus by the spontaneous transformation of
a neutron to a proton or a proton to a neutron.
Beta particle decay is characterized by production of particles with a
continuous spectrum of energies which is characteristic of each decay pro-
cess. Beta-energy ranges in air are difficult to evaluate. Thus, they are
based upon the thickness of an absorber, such as aluminum, required to stop
the particle. Thin-windowed geiger or proportional tube counters are used to
count a uniform layer of the sample for beta sources having energies greater
than 0.2 MeV. For low-energy beta emitters, such as carbon-14, sulfur-35,
and tritium, a liquid scintillation counter is used. For the liquid scin-
tillation counter method, the sample is dissolved in a solution of the
scintillation compound. A vial containing the solution is then placed be-
tween two photomultiplier tubes housed in a light-tight container. The
output from the two tubes is fed into a counter which records a count only
when pulses from the two detectors arrive at the same time.
Beta particle emission techniques are used for analysis of radioactive
tracers.
184
-------
Neutron Activation Analysis
Neutron activation analysis involves the production of a radioactive
isotope by the capture of neutrons by the nuclei of the substance to be
analyzed. Irradiation is accomplished by placing the sample to be analyzed
in an intense flux of either thermal or fast neutrons for a length of time
sufficient to produce a measurable amount of the desired radioisotope. Radi-
ation detectors are used to analyze the radiation emitted by each sample and
the unique radiation characteristics of the sample are sought.
The method known as post-sampling activation analysis has been described
by Schmotzer (1973) as a tracer technique using low concentrations of Br~.
Although this method of tracer analysis reduces the amount and subsequently
the cost of the chemical tracer, it is an expensive technique.
185
-------
REFERENCES
Akamatsu, K., and Matsuo, M., 1973, Safety of optical whitening agents (in
Japanese): Senryo to Yakuhin, Vol. 18, No. 2, pp. 2-11.
Alburger, J. R. , 1977, Leak testing with dyed liquid tracers: Materials
Evaluation, Dec., pp. 60-64.
Aley, T., and Fletcher, M. W. , 1976, Water tracer's cookbook: Journal of
the Missouri Speleological Survey, Vol. 16, No. 3.
Allen, M. J., and Morrison, S. M. , 1973, Bacterial movement through frac-
tured bedrock: Ground Water, Vol. 11, No. 2, pp. 6-10.
Andre, J. C., and Molinari, J., 1976, Mises au point sur les differents
facteurs physiochimique influant sur la mesure de concentration de
traceurs fluorescents et leurs consequences practiques en hydrologie:
Journal of Hydrology, Vol. 30, pp. 257-286.
Andrews, J. H., Lee, D. J., 1979, Inert gases in groundwater from the ^
Bunter Sandstone of England as indicators of age and paleoclimatic
trends: Journal of Hydrology, Vol. 41, pp. 233-252.
Arandjelovic, D., 1969, A possible way of tracing groundwater flows in
karst: Geophysical Prospecting, Vol. 17, No. 4, pp. 404-418.
Arandjelovic, D., 1977, Determining groundwater flow in karst using "Geo-
bomb." In Karst Hydrology, edited by J. S. Tolson and F. L. Doyle:
Memoirs of the 12th Congress of the Int. Assoc. Hydrogeologists,
University of Alabama Press in Huntsville, pp. 399-400.
Atkinson, T. C., 1968, Tracing swallet waters using locopodium spores:
Transactions Cave Research Group of Great Britain, Vol. 10, No. 2,
pp. 99-106.
Atkinson, T. C., Drew, D. P., and High, C., 1967, Mendip karst hydrology
research project, phases one and two. Occasional Publ. Wessex Cave
Club, Vol. 2, No. 1, 38 p.
Atkinson, T. C., Smith, D. I., Lavis, J. J., and Whitaker, R. J., 1973,
Experiments in tracing underground waters in limestones. Journal
of Hydrology, Vol. 19, pp. 323-349.
Aulenbach, D. B., Bull, J. H., and Middlesworth, B. C., 1978, Use of tracers
to confirm ground-water flow: Ground Water, Vol. 16, No. 3, pp.
149-157.
Aulenbach, D. B., and Clesceri, N. L., 1980, Monitoring for land application
of wastewater: Water, Air, and Soil Pollution, Vol. 14, pp. 81-94.
186
-------
Bear, J. , 1961, On the tensor form of dispersion in porous media, J- Geo-
phys. Res., 66(4), pp. 1185-1197.
Bear, J. , 1969, Hydrodynamic dispersion, jLn Flow Through Porous Media,
DeWiest (Ed.), Academic Press.
Bentley, H. W-, and Alweis, S., 1982, Use of sulfur isotopes to evaluate
groundwater contamination by mining and industrial processes, ACS
Symposium on the Hydrology of Mining and Industrial Wastes, El Paso,
December 2 (abstracts with programs).
Borg, I. Y., and others, 1976, Information pertinent to the migration of
radionuclides in ground water at the Nevada Test Site: Lawrence
Livermore Laboratory, University of California, Publication UCRL-52078
Part I, 216 p.
Bouwer, H., 1978, Groundwater Hydrology: New York, McGraw-Hill Book Co.,
480 p.
Brown, E., Skougstad, M. W., and Fishman, M. J., 1970, Methods for collec-
tion and analysis of water samples for dissolved minerals and gases:
U. S. Geological Survey Tech. Water Resources Investigation, Vol. 5,
No. Al, 160 p.
Brown, J. D., 1980, Evaluation of fluorocarbon compounds as ground-water
tracers: Soil column studies: Unpublished M.S. thesis, University
of Arizona, Department of Hydrology and Water Resources, 97 p.
Brown, M. C. , and Ford, D. C., 1971, Quantitative tracer methods for inves-
tigation of karst hydrology systems, with reference to the Maligne
Basin area: Transactions of the Cave Research Group of Great Britain,
Vol. 13, No. 1, pp. 37-51.
Brown, M. C., Wigley, T. L., and Ford, D. C., 1969, Water budget studies in
karst aquifers: Journal of Hydrology, 9(1), pp. 113-116.
Buchtela, K., Mairhofer, J., Maurin, V., Papadimitropoulos, T., and Zotl,
J., 1968, Comparative investigations into recent methods of tracing
subterranean water. National Speleological Society Bulletin, Vol. 30,
13, pp. 179-195.
Burdon, D. J., et al., 1963, The use of tritium in tracing karst ground
water in Greece, in Radioisotopes in Hydrology IAEA Symposium, Tokyo,
pp. 309-320.
Buss, D. F., and Bandt, K. E. , 1981, An all-teflon bailer and an air-driven
pump for evacuating small-diameter ground-water wells: Ground Water,
Vol. 19, No. 4, pp. 100-102.
California Department of Water Resources, 1968, Water well standards, State
of California: Calif. Dept. Water Res. Bulletin, Vol. 74, 205 p.
187
-------
Campbell, M. D., and Lehr, J. H., 1973, Water well technology: McGraw-Hill,
New York, 681 p.
Carrera, J., and Walter, G. R., 1985 CONFLO, a new numerical model for
analyzing convergent flow tracer tests. Sandia Contractors Report
(in preparation).
Carter, R. C., and others, 1959, Helium as a ground-water tracer. Journal
of Geophysical Research, Vol. 64, pp. 2433-2439.
Cherry, J. A., and Johnson, P. E., 1982, A multilevel device for monitoring
in fractured rock: Ground Water Monitoring Review, Summer 1982,
pp. 41-44.
Claassen, H. C., and Cordes, E. H. , 1975, Two-well recirculating tracer test
in fractured carbonate rock, Nevada: Hydrological Sciences Bulletin
20(3), pp. 367-382.
Coleman, D. D. , Meents, W. F., Liu, C. L., and Keogh, R. A., 1977, Isotopic
identification of leakage gas from underground storage reservoirsa
progress report: Illinois State Geological Survey, Illinois Petroleum
No. Ill, 10 p.
Combarnous, M. A., and Bories, S. A., 1975, Hydrothermal convection in
saturated porous media, rn Advances in Hydroscience, V. T. Chow (ed.):
pp. 232-307.
Corey, J. C., 1968, Evaluation of dyes for tracing water movement in acid
soils: Soil Science, Vol. 106, No. 3, pp. 182-187.
Cory, C. C., and Horton, J. H., 1968, Movement of water tagged with 2H, 3H,
and 180 through acidic kaolinitic soil: Soil Sci. Soc. America Proc.,
Vol. 32, pp. 471-475.
Cotlove, E., 1964, Determination of chloride in biological materials: Meth.
Biochem. Anal., Vol. 12, 287 p.
Custodio, E., 1976, Trazadores y tecnicas radioisotopicas en hidrologia
subterranea, Section 12 in Hidrologia Suberrranea, Vol. 2, edited by
E. Custodio and M. R. Llamas; Ediciones Omega, Barcelona, Spain,
pp. 1165-1312.
Dansgaard, W., 1964, Stable isotopes in precipitations: Tellus, Vol. 16,
pp. 436-468.
Davis, J. T., Flotz, E., and Balkeraore, W. S., 1970, Serratia marcescens, a
pathogen of increasing clinical importance: J. Am. Med. Assoc., Vol.
214, pp. 145-150.
Davis, S. N., 1969, Porosity and permeability of natural materials, in Flow
through porous media, R. M. DeWiest, ed., Academic Press, New York,
pp. 54-89.
188
-------
Davis, S. N. , and Bentley, H. W. , 1982, Dating groundwater, a short review,
in Nuclear and chemical dating techniques, Lloyd Curie, ed., Am.
Chemical Society Symposium Series No. 176, Chapter 11, pp. 187-222.
Davis, S. N., and DeWiest, R. J. M. , 1966, Hydrogeology: John Wiley and
Sons, New York, 463 p.
Davis, S. N., Thompson, G. M., Bentley, H. W., and Stiles, G., 1980, Ground-
water tracers—a short review: Ground Water, Vol. 18, pp. 14-23.
Dean, J. A., 1968, Chemical Separation Methods: D. Van Nostrand, New York.
Dole, R. B., 1906, Use of fluorescein in the study of underground water:
U. S. Geological Survey Water Supply and Irrigation Paper, No. 160,
pp. 73-83.
Downs, W. F., McAtee, R. E., and Capuano, R. M., 1983, Tracer injection
tests in a fracture dominated geothermal system (Abstract): Am.
Geophysical Union, EOS, Vol. 64, No. 18, p. 229.
Drew, D. P., 1968, A study of the limestone hydrology of St. Dunstans Well
and Ashwick drainage basins, Eastern Mendip: Proc. Univ. Bristol
Speleol. Soc., Vol. 11, No. 3, pp. 257-276.
Drew, D. P., and Smith, D. I., 1969, Techniques for the tracing of subter-
ranean drainage: Br. Geomorphol. Res. Group Tech. Bulletin, Vol. 2,
36 p.
Drewry, W. A., and Eliassen, R. , 1967, Virus movement in ground water:
Jour. Water Pollution Control Federation, Vol. 40, No. 4, pp. 257-271.
Drost, W., Klotz, D., Koch, A., Moser, H., Neumaier, F., and Ravert, W.,
1968, Point dilution methods of investigating ground-water flow by
means of radioisotopec: Water Resources Research, Vol. 4, pp. 125-
146.
Dunn, J. A., 1963, New method of water tracing: Journal of Eldon Pothole
Club, Vol. 5, p. 5.
Ellis, J., 1980, A convenient parameter for tracing leachate from sanitary
landfills: Water Research 14(9): pp. 1283-1287.
Elrick, D. E., and Lawson, D. W. , 1969, Tracer techniques in hydrology,
Proceedings of the Canadian Hydrology Symposium No. 7: National Re-
search Council of Canada, Subcommittee on Hydrology, (1), pp. 155-187.
Ferronsky, V. I., and Polyakov, V. A., 1982, Environmental isotopes in the
hydrosphere: John Wiley and Sons, Interscience Publications, New
York, 466 p.
Fetter, C. W., Jr., 1980, Applied hydrogeology: Columbus, Ohio, Charles E.
Merrill Publishing Co., 488 p.
189
-------
Fetter, C. W., Jr., 1981, Determination of the direction of ground water
flow: Ground Water Monitoring Review: Vol. 1, No. 3, pp. 28-31.
Feuerstein, D. L., and Selleck, R. E. , 1963, Fluorescent tracers for dis-
persion measurements: Journal of the American Society of Civil
Engineers, August, pp. 1-21.
Fletcher, M. W., and Myers, R. L., 1974, Ground-water tracing in karst
terrain using phage T-4: Amer. Soc. Microbiol. Abstr. Ann. Mtg.,
p. 52.
Fontes, J. C., and Fritz, P., 1975, Isotope hydrology 1974—a review of the
IAEA Symposium on isotope techniques in groundwater hydrology: Earth
and Planetary Science Letters, Vol. 4, pp. 321-324.
Fontes, J. C., and Gamier, J. M. , 1979, Determination of the initial lkC
activity of the total dissolved carbon, A review of the existing
models and a new approach: Water Resources Research, Vol. 15, No. 2,
pp. 399-413.
Fox, G. S. , 1952, Radioactive isotopes trace underground waters: Public
Works, Vol. 83, pp. 57-58.
Freeze, R. A., and Cherry, J. A. , 1979, Groundwater: Prentice-Hall, Inc.,
New Jersey, 604 p.
Fried, J. J., 1975, Groundwater pollution; theory, methodology modeling,
and practical rules. Elsevier Scientific Publishing Co., pp. 59-113.
Gann, E. E., and Harvey, E. J., 1975, Norman Creek: A source of recharge
to Maramec Spring, Phelps County, Missouri: Journal of Research of
the U. S. Geological Survey, Vol. 3, No. 1, pp. 99-102.
Gardner, G. D., and Gray, R. E., 1976, Tracing subsurface flow in karst
regions using artificially colored spores: Association of Engineering
Geologists Bulletin, Vol. 13, pp. 177-197.
Caspar, E., and Oncescu, M., 1972, Radioactive tracers in hydrology:
Elsevier Publishing Co., pp. 77-154.
Gat, J. R. , 1971, Comments on the stable isotope method in regional ground-
water investigations: Water Resources Research, Vol. 7, pp. 980-993.
Gelhar, L. W., 1982, Analysis of two-well tracer tests with a pulse input:
Rockwell International (Hanford, Washington) Report RHO-BW-CR-131P,
96 p.
Gelhar, L. W., and Collins, M. A., 1971, General analysis of longitudinal
dispersion in nonuniform flow: Water Resource Research 7(6), pp.
1511-1521.
190
-------
Gelhar, L. W., Gutjahr, A. L., and Naff, R. L., Stochastic analysis of
macrodispersion in a stratified aquifer, Water Resources Research,
Vol. 15(6), pp. 1387-1397.
Gillhara, R. W. , and Johnson, P. E., 1981, A positive displacement ground-
water sampling device: Ground Water Monitoring Review, Summer 1981,
pp. 33-35.
Glover, R. R., 1972, Optical brighteners—a new water tracing reagent.
Transactions Cave Research Group, Great Britain, Vol. 14, No. 2,
pp. 84-88.
Goyal, S. M., Zerda, K. Z., and Gerba, C. P., 1980, Concentration of coli-
phages from large volumes of water and waste water: Appl. Environ.
Microbiol., Vol. 39, pp. 85-91.
Greenkorn, Robert A., 1962, Experimental study of waterflood tracers:
Journal of Petroleum Technology, January, pp. 87-92.
Grisak, G. E., and Pickens, J. F. , 1980, Solute transport through fractured
media 1. The effect of matrix diffusion: Water Resources Research,
Vol. 16, No. 4, pp. 719-730.
Grisak, G. E., Pickens, J. F. , and Cherry, J. A., 1979, Solute transport
through fractured media 2. Column study of fractured till: manu-
script submitted to Water Resources Research, July 1979.
Grove, D. B. , and Beetem, W. A., 1971, Porosity and dispersion constant
calculations for a fractured carbonate aquifer using the two-well
tracer method: Water Resources Research, Vol. 7, No. 1, pp. 128-134.
Haas, J. L. , 1959, Evaluation of groundwater tracing methods used in spele-
'ology: Bulletin Natl. Spel. Soc. (U. S. A.), 21(2), pp. 67-76.
Hach Chemical Company, 1969, Colorimetric procedures and chemicals for
water and wastewater analysis: Hach Chem. Co., Ames, Iowa, 91 p.
Hagedorn, C. , Hansen, D. T., and Simonson, G. H., 1978, Survival and move-
ment of fecal indicator bacteria in soil under conditions of saturated
flow: J. Environ. Quality, Vol. 17, pp. 55-59.
Halevy, E., and Nir, A., 1962, The determination of aquifer parameters with
the aid of radioactive tracers: Jour. Geophysical Research, Vol. 61,
pp. 2403-2409.
Heath and Trainer, 1968, Introduction to Ground-Water Hydrology: John Wiley
and Sons, New York, 283 p.
Higgins, C. H., 1969, Evaluation of the ground-water contamination hazard
from underground nuclear explosions: Jour. Geophys. Research, Vol. 64,
p. 1509.
191
-------
Hubbard, E. F., Kilpatrick, F. A., Martens, L. A., and Wilson, J. F., Jr.,
1982, Measurement of time of travel and dispersion in streams by dye
tracing, in: Techniques of Water Resources Investigations of the
U. S. G. S., Chapter A9, Book 3, Applications of Hydraulics.
HydroGeoChem, 1984, Hydrologic investigation of existing pond seepage at
Jim Budger power plant, Report to Pacific Power and Light, Portland,
Oregon.
Isotope Hydrology Section, International Atomic Energy Agency, 1973, Nuclear
techniques in ground-water hydrology, In: Ground-water studies:
UNESCO, Paris, Sections 10.1-10.4, 38 p.
Ivanovich, M., and Smith, D. B., 1978, Determination of aquifer parameters
by a two-well pulse method using radioactive tracers, Journal of
Hydrology, Vol. 36, No. 1/2, pp. 35-45.
Jennings, A. R., and Schroeder, M. C., 1968, Laboratory evaluation of selec-
ted radioisotopes as ground-water tracers: Water Resources Research,
Vol. 4, pp. 829-838.
Johnson Division, UOP, Inc., 1972, Ground water and wells, 2nd ed.: Edward
E. Johnson, Co., St. Paul, Minnesota, 440 p.
Kaas, W., 1964, Die unmittelbare Bestimmung von Uranin-Spuren bei Farbver-
suchen: Steirische Beitrage zur Hydrogeologie. Jahrgang 1963/64,
pp. 37-66.
Kaufman, W. J., 1961, Tritium as a ground-water tracer: Am. Soc. Civil
Eng. Trans. Paper 3203, pp. 436-446.
Kaufman, W. J., and Orlob, G. T., 1956, Measuring ground-water movements
with radioactive and chemical tracers: Amer. Waterworks Association
Journal, 48, pp. 559-572.
Keely, J. F. , 1984, Optimizing pumping strategies for contaminant studies
and remedial actions: Ground Water Monitoring Review, Vol. 4, No. 3,
pp. 63-74.
Keith, S. J., Wilson, L. G., Fitch, H. R., Esposito, D. M., 1982, Sources of
spatial-temporal variability in ground-water quality data and methods
of control: Case study of the Cortaro Monitoring Program, Arizona:
Second National Symposium on Aquifer Restoration and Ground-Water Moni-
toring, National Water Well Assoc., Worthington, Ohio, pp. 217-227.
Keswick, B. H. , Wang, D., and Gerba, C. P., 1982, The use of Microorganisms
as Ground-Water Tracers; A Review: Ground Water, Vol. 20, No. 2,
pp. 142-149.
Keys, W. S., and Brown, R. F., 1978, The use of temperature logs to trace
the movement of injected water: Ground Water, Vol. 16, No. 1, pp.
32-48.
192
-------
Keys, W. S., and MacCary, L. M. , 1971, Application of borehole geophysics
to water-resources investigations: U. S. Geological Survey Techniques
in Water Resources Inv., Book 2, Chapter 1.
Kilpatrick, F. A., Sayre, W. W., and Richardson, E. V., 1967, (Discussion
of Repogle et al.)> Flow measurements with fluorescent tracers (loc.
cit): Proceedings of the ASCE, Journal of the Hydraulics Division,
93, pp. 298-308.
Klotz, D., Moser, H., and Trimborn, P., 1978, Single-borehole techniques;
present status and examples of recent applications: Isotope Hydrology,
IAEA, Vienna, Part 1, pp. 159-179.
Knuttson, G., 1968, Tracers for ground-water investigations, in: Ground
Water Problems. Eriksson, E. , Gustafsson, Y., and Nilsson, K., (eds):
Pergamon Press, London, pp. 123-152.
Koerner, R. M., Reif, J. S., and Burlingarae, M. J., 1979, Detection methods
for location of subsurface water and seepage: Journal of the Geotech-
nical Engineering Division, ASCE, Vol. 105, pp. 1301-1316.
Konikow, L. F., and Bredehoeft, J. D. , 1974, Modeling flow and chemical
quality changes in an irrigated stream-aquifer system, Water Resources
Research, Vol. 10(3), pp. 562-596.
Krothe, N. C., 1982, Sulfur isotopes and hydrochemical variations as an
indicator of flow in groundwater: jLn Isotope studies of hydrologic
processes, E. C. Perry, Jr. and C. W. Montgomery, editors: Northern
Illinois University Press, DeKalb, Illinois, pp. 75-82.
Lange, A. L., 1972, Mapping underground streams using discrete natural noise
signals: A proposed method: Caves and Karst, Vol. 14, pp. 41-44.
Lee, R., Cherry, J. A., and Pickens, J. F., 1980, Groundwater transport of
a salt tracer through a sandy lakebed: Limnol. Oceanogr. 25(1),
pp. 45-61.
Lenda, A., and Zuber, A., 1970, Tracer dispersion in groundwater experi-
ments: Isotope Hydrology (Proc. Symp. Vienna, 1970), IAEA, pp. 619-
641.
Lewis, D. C., Kriz, G. J., and Burgy, R. H., 1966, Tracer dilution sampling
technique to determine hydraulic conductivity of fractured rock: Water
Resources Research, Vol. 2, pp. 533-542.
Libby, W. F., 1961, Tritium Geophysics: Jour. Geophys. Research, Vol. 66,
pp. 3767-3782.
Loosli, H. H. , and Oeschger, H. , 1978, Argon-39, carbon-14, and krypton-85
measurements in groundwater samples, In: Isotope Hydrology 1978:
Internat. Atomic Energy Agency, Vienna, Vol. 2, pp. 931-945.
193
-------
Malcolm, R. L., Aiken, G. R., Thurman, E. M., and Avery, P. A., 1980, Hydro-
phllic organic solutes as tracers in groundwater recharge studies:
Contaminants and Sediments, Vol. 1, pp. 71-87.
Marston, T. K. , and Schofield, J. , 1962, An improved method of tracing
underground waters using Rhodamine B: Cave Research Group Newsletter,
No. 84, pp. 4-13.
Martin, R., and Thomas A., 1974, An example of the use of bacteriophage as
a ground-water tracer: Journal of Hydrology, Vol. 23, pp. 73-78.
Mather, J. D. , Gray, D. A., and Jenkins, D. G., 1969, The use of tracers to
investigate the relationship between mining subsidence and groundwater
occurrence of Aberdare, South Wales: Journal of Hydrology, Vol. 9,
pp. 136-154.
Mattson, S., 1929, The laws of soil colloidal behavior I: Soil Science,
Vol. 27-28, pp. 179-220.
Maurin, V., and Zotl, J., 1959, Die Untersuchung der Zusammenhange
unteirir-discher Wasser nit besonderer Berucksichtigung der Karstver
haltnisse: Steierische Beitrage zur Hydrologie, Jahrgang, 1959, Graz,
Austria.
Mayr, A., 1953, Bluten pollen und Pflanzl Sporen als Mittel zur Untersuchung
von Quellen und Karstwassen: Anz. Math-Natw. Kl. Ost. Ak. Wiss.
Mazor, E., 1972, Paleotemperatures and other hydrological parameters deduced
from noble gases dissolved in ground waters, Jordan Rift Valley, Israel:
Geochimica et Cosmochimica Acta: Vol. 36, pp. 1321-1336.
Mazor, E., 1976, The Ram Crater Lake, a note on the revival of a 2,000-year-
old ground-water tracing experiment: in Interpretation of Environmental
Isotope and Hydrochemical Data, jin Groundwater Hydrology. IAEA,
Vienna, pp. 179-181.
Mclaughlin, M. J., 1982, A review of the use of dyes as soil water tracers:
Water S. A., Water Research Commission, Pretoria, South Africa, Vol.
8, No. 4, pp. 196-201.
Morrison, R. D., and Brewer, P. E. , 1981, Air-lift samplers for zone-of-
saturation monitoring: Ground Water Monitoring Review, Spring 1981,
pp. 52-55.
Murray, J. P., Rouse, J. V., and Carpenter, A. B., 1981, Groundwater contami-
nation by sanitary landfill leachate and domestic wastewater in carbon-
ate terrain: principle source diagnosis, chemical transport character-
istics and design implications: Water Research, 15(6), pp. 745-757.
Naymik, T. G., and Sievers, M. E., 1983, Ground-water tracer experiment (II)
at Sand Ridge State Forest, Illinois: Illinois State Water Survey
Division, SWS Contract Report 334, 105 p.
194
-------
Ogato, A., and Banks, R. B. , 1961, A solution of the differential equation of
longitudinal dispersion in porous media, U. S. Geological Survey, Pro-
fessional Paper 411A.
Omoti, 0., 1977, Laboratory and field studies of pathways of solute move-
ment in soils: unpublished Ph.D. dissertation, University of Reading,
England.
Ormerod, J. G., 1964, Serratia indica as a bacterial tracer for water
movement: Appl. Bact., Vol. 27, pp. 342-349.
Payne, B. R. , 1972, Isotope hydrology, in V. T. Chow, editor, Advances
in Hydroscience, Vol. 8, pp. 95-138.
Pearson, F. J., Jr., White, D. E., Carbon-14 ages and flow rates of water
in Carrizo Sand, Atascosa County, Texas: Water Resources Research,
Vol. 3, No. 1, pp. 251-261.
Perkins, T. K., and Johnston, 0. C., 1963, A review of diffusion and dis-
persion in porous media, Soc. Pet. Eng. Jour., Vol. 3(1), pp. 70-84.
Pickens, J. F., Cherry, J. A., Coupland, R. M., Grisak, G. E., Merrit, W. F.,
and Risto, B.A., 1981, A multilevel device for ground-water sampling:
Ground Water Monitoring Review, Spring 1981, pp. 48-51.
Pickens, J. F., and Grisak, G. E. , 1981, Scale-dependent dispersion in a
stratified granular aquifer. Submitted to Water Resources Research,
January, 1981, pp. 1-83.
Pickens, J. F., Merritt, W. F., and Cherry, J. A., 1976, Field determination
of the physical contaminant transport parameters in a sandy aquifer.
Paper presented at the Advisory Group Meeting on "The Use of Nuclear
Techniques in Water Pollution Studies."
Pickett, E. E., and Koirtyohann, S. R. , 1969, Emission flame photometry—a
new look at an old method: Analyt. Chem., Vol. 45, pp. 28a-42a.
Finder G. F. 1973, A Galerkin-finite element simulation of ground-water
contamination on Long Island, New York, Water Resources Research, Vol.
10(3), pp. 546-562.
Plata Bedmar, A., 1972, Isotopos en Hidrologia: Editorial Alhambra, S. A.,
Madrid, 328 p.
Pyle, B. H., and Thorpe, H. R., 1981, Evaluation of the potential for
microbiological contamination of an aquifer using a bacterial tracer:
Proceedings of Ground-Water Pollution Conference, 1979. Australian
Water Resources Council Conference Series, No. 1, pp. 213 224.
195
-------
Rahe, T. M., Hagedorn, C., McCoy, E. L., and Kling, G. F., 1978, Transport
of antibiotic-resistant Echerichia coli through western Oregon hill
slope soils under conditions of saturated flow: J. Environ. Qual.,
Vol. 7, pp. 487-494.
Repogle, J. A., Myers, L. E., and Brust, K. J., 1966, Flow measurements with
fluorescent tracers: Journal of the Hydraulics Division ASCE, Vol. 92,
PP. 1-15.
Reynolds, E. R. C., 1966, The percolation of rainwater through soil demon-
strated by fluorescent dyes: Journal of Soil Science, Vol. 17, No. 1,
pp. 127-132.
Reznek, S. , Hayden, W., and Lee, M., 1979, Analytical note—fluoresceln
tracer technique for detection of ground-water contamination: Journal
of the American Water Works Association, Vol. 71, No. 10, pp. 586-587.
Rippon, J. E., 1963, The use of a colored bacterium as an indicator of local
water movement: Chem. Ind., Vol. 11, pp. 445-446.
Robertson, J. B., 1969, Behavior of xenon-133 gas after injection under-
ground: U. S. Geol. Survey Open File Report TD022051, 37 p.
Robin, M. J. L., Dytynyshyn, D. J., and Sweeney, S. J., 1982, Two gas-drive
sampling devices: Ground Water Monitoring Review, Winter 1982, pp.
63-66.
Rodrigeuz, C. 0., 1977, Hidrologia isotopica en Colombia: Instituto de
Asuntos Nucleares, Bogota, 81 p.
Rogers, A. S., 1958, Physical behavior and geologic control of radon in
mountain streams: U. S. Geological Survey Bulletin 1052E, pp. 187-211.
Romero, J. C., 1970, The movement of bacteria and viruses through porous
media: Ground Water, Vol. 8, No. 2, pp. 37-48.
Rorabaugh, M. I., 1956, Ground water in northeastern Louisville, Kentucky:
U. S. Geol. Survey Water-Supply Paper 1360-B, pp. 101-169.
Saleem, M., 1971, A simple method of ground water direction measurement in a
single borehole: Journal of Hydrology, 12, pp. 387-410.
Sanitary Engineering Research Laboratory, University of California, Berkeley,
1954, Report on the investigation of travel of pollution: California
State Water Pollution Control Board Pub. 11, 218 p.
Sargeant, K., 1969, The deep culture of bacteriophage, _rn Methods in Micro-
biology, Norris, J. R., and Ribbons, D. W., (eds.), Vol. 1: Academic
Press, New York, pp. 505-520.
196
-------
Sauty, J. P., 1978, Identification of hydrodispersive mass transfer param-
eters in aquifers by interpretation of tracer experiments in radial
converging or diverging flow (in French): Journal of Hydrology, Vol.
39, pp. 69-103.
Sauty, J. P., 1980, An analysis of hydrodispersive transfer in aquifers:
Water Resources Research, Vol. 16, No. 1, pp. 145-158.
Scandura, J. E., and Sobsey, M. D. , 1981, Survival and fate of enteric
viruses in on-site waste-water disposal systems in coastal plains
soils: Abs. Ann. Mtg. Am. Soc. Microbiol., p. 175.
Schaub, S. A., Meier, E. P., Kolmer, J. R., and Sorber, C. A., 1975, Land
application of wastewater: the fate of viruses, bacteria, and heavy
metals at a rapid infiltration site: Report No. TR 7504, AD A011263,
U. S. Army Medical Bioengineering Research and Development Laboratory,
Ft. Detrick, Frederick, Maryland.
Schaub, S. A., and Sorber, C. A., 1977, Virus and bacteria removed from
waste water by rapid infiltration through soil: Applied Environmental
Microbiology, Vol. 33, pp. 609-619.
Scheidegger, A. E., 1954, Statistical hydrodynamics in porous media, J.
Applied Physics, Vol. 25(8), pp. 994-1001.
Schmidt, K. D., 1977, Water quality variations for pumping wells: Ground
Water, Vol. 15, No. 2, pp. 130-137.
Schmotzer, J. K., Jester, W. A., and Parizek, R. R. , 1973, Groundwater
tracing with post sampling activation analysis: Journal of Hydrology,
Vol. 20, pp. 217-236.
Simpson, E. S., 1984, Personal communication, Department of Hydrology, Uni-
versity of Arizona, June, 1984.
Simpson, E. S., Neuman, S. P., and Thompson, G. M., 1983, Field and theo-
retical investigations of mass and energy transport in subsurface
materials: Progress Report for the Nuclear Regulatory Commission by
the Department of Hydrology and Water Resources, University of Arizona,
Tucson, Arizona.
Sinton, L. W., 1980, Investigations into the use of the bacterial species
Bacillus stearotheraiophilus and Echerichia coli as tracers of
ground-water movement: Water and Soil Technical Publication No. 17,
MWD, Wellington, 24 p.
Sklash, M. G., and Farvolden, R. N. , 1979, The role of groundwater in storm
runoff: Journal of Hydrology, Vol. 43, pp. 45-65.
Skoog, D. A., and West, D. M., 1980, Principles of instrumental analysis:
Holt, Rinehart, and Winston, Philadelphia, Pennsylvania, 2nd edition,
p. 760.
197
-------
Slichter, C. S., 1902, The motions of underground waters: U. S. Geological
Survey Water-Supply Paper No. 67, p. 106.
Slichter, C. S., 1905, Field measurements of the rate of movement of under-
ground waters: U. S. Geological Survey Water Supply and Irrigation
Paper No. 140, pp. 9-34.
Smart, P. L., 1976, Catchment delimitation in karst areas by the use of qual-
itative tracer methods: Proc. 3rd Internatl. Symp. of Underground Water
Tracing, Bled, Yugoslavia, 1976, pp. 291-298.
Smart, P. L., and Laidlaw, I. M. S., 1977, An evaluation of some fluorescent
dyes for water tracing: Water Resources Research, Vol. 13, No. 1, pp.
15-33.
Smart, P. L., and Smith, D. I., 1976, Water tracing in tropical regions;
the use of fluorometric techniques in Jamaica: Journal of Hydrology,
Vol. 30, pp. 179-195.
Smith, D. B., 1973, Flow tracing using isotopes: Groundwater Pollution in
Europe, Proc. Reading Conf., Water Research Association, pp. 241-250.
Smith, S. J., and Davies, R. J. , 1974, Relative movement of bromide and
nitrate through soils: Jour. Environmental Quality, Vol. 3, pp.
152-155.
Sorey, M. L., 1971, Measurement of vertical ground-water velocity from
temperature profiles in wells: Water Resources Research, Vol. 7,
No. 4, pp. 963-970.
Stallman, R. W., 1963, Computation of ground-water velocity from tempera-
ture data: in. u* s> Geological Survey Water Supply Paper 1544-H.,
R. Bental (ed.), pp. 36-46.
Stetzenbach, K. J., Jensen, S. L., and Thompson, G. M., 1982, Trace enrich-
ment of fluorinated organic acids used as ground-water tracers by
liquid chromatography: Environmental Science and Technology, Vol. 16,
p. 250.
Sudicky, E. A., and Cherry, J. A., 1979, Field observations of tracer dis-
persion under natural flow conditions in an unconfined sandy aquifer,
Water Pollution Research (Canada), Vol. 14.
Sugisaki, R., 1969, Measurement of effective flow velocity of groundwater
by means of dissolved gases: American Jour. Science, Vol. 259, pp.
144-153.
Tennyson, L. C., and Settergren, C. D., 1980, Percolate water and bromide
movement in the root zone of effluent irrigation sites: Water Re-
sources Bulletin, Vol. 16, No. 3, pp. 433-437.
198
-------
Tester, J. W., Bivens, R. L., and Potter, R. M., 1982, Interwell tracer
analysis of a hydraulically fractured grantitic geothermal reservoir:
Society of Petroleum Engineers Journal, August, 1982, pp. 537-554.
Theis, C. V., 1963, Hydrologic phenomena affecting the use of tracers in
timing ground-water flow: Radioisotopes in Hydrology. Internatio.nal
Atomic Energy Agency (Tokyo Symposium) Vienna, Austria, pp. 193-206.
Thompson, G. M., and Hayes, J. M. , 1978, Trichlorofluoromethane in ground
water. A possible tracer and indicator of ground-water age: Water
Resources Research, Vol. 15, No. 3, pp. 546-554.
Thompson, G. M., Hayes, J. M., and Davis, S. N., 1974, Fluorocarbon tracers
in hydrology: Geophysical Research Letters, Vol. 1, pp. 177-180.
Thompson, G. M., and Jensen, S. L. , 1980, (unpublished manuscript), New
organic tracers for waste monitoring: Department of Hydrology and
Water Resources, University of Arizona, Tucson, Arizona.
Thorp, J., and Gamble, E. E., 1972, Annual fluctuation of water levels in
the soils of the Miami catena, Wayne County, Indiana: Earlham College,
Science Bulletin No. 5, 26 p.
Thorpe, H. R., 1979, Movement of contaminants into and through the Here-
taunga Plains Aquifers, Hawkes Bay, New Zealand: Paper presented at
AWRC Conference on Groundwater Pollution, Perth, Australia.
Todd, D. K., 1980, Groundwater hydrology, 2nd ed.: John Wiley and Sons, New
York, 535 p.
U. S. Department of the Interior, 1969, Methods for chemical analysis of
water and wastes: Federal Water Pollution Control Admin., Cincinnati,
Ohio, 280 p.
Vecchioli, J., Ehrlich, G. G., and Ehlke, T. A., 1972, Travel of pollution
indicator bacteria through the Magothy aquifer, Long Island, New York:
U. S. Geological Survey Prof. Paper 800-B., pp. B237-B239.
Vogel, J. C., Thilo, L., Van Dijken, M. , 1974, Determination of groundwater
recharge with tritium, Journal of Hydrology, Vol. 23, pp. 131-140.
Vuataz, F. D., Stix, J., Goff, F., and Pearson, C. F., 1984, Low-temperature
geothermal potential of the Ojo Caliente warm springs area in northern
New Mexico: Los Alamos National Laboratory Publication LA-10105-OBES,
VC-666, 56 p.
Wagner, 0. R. , 1977, The use of tracers in diagnosing interwell reservoir
heterogeneities: Jour. Petroleum Technology, November, 1977, pp.
1410-1416.
Warren, J. E., and Skiba, F. F., 1964, Macroscopic dispersion, Soc. Pet.
Eng. Jour., Vol. 4(3), pp. 215-230.
199
-------
Water and Power Resources Service, 1981, Ground water manual: U. S. Depart-
ment of the Interior, Denver, Colorado, 480 p.
Webster, D. S., Proctor, J. F., and Marine, I. W., 1970, Two-well tracer
test in fractured crystalline rock: U. S. Geological Survey Water
Supply Paper 1544-1, pp. 1-22.
White, K. E., 1981, Hydrological studies possible with radionuclides of
bomb-test. Primordial and natural origin to complement investigations
using manufactured radiotracers: Water Pollution Control, 80(4),
pp. 498-512.
Willard, H. H. , Merritt, L. L. Jr., and Dean, John A., 1965, Instrumental
methods of analysis, D. Van Nostrand, New York, 5th ed., 859 p.
Wilson, J. F., 1968, Fluorometric procedures for dye tracing: Chapter A12
in Techniques of Water-Resources Investigations of the U.S.G.S., U. S.
Geological Survey, pp. 1-31.
Wilson, L. G., 1971, Investigations on the subsurface disposal of waste ef-
fluent at inland sites: Water Resources Research Center, Tucson, Ari-
zona, Final report to Office of Saline Water, Grant #14-01-0001-1805,
pp. 1-82.
Wimpenny, J. W. T., Cotton, N., and Strathem, M., 1972, Microbes as tracers
of water movement: Water Research, Vol. 6, pp. 731-739.
Wood, W. W., and Ehrlich, G. G., 1978, Use of baker's yeast to trace micro-
bial movement in ground water: Ground Water, Vol. 16, No. 6, pp.
398-403.
Yates, W. E., and Akesson, N. B. , 1963, Fluorescent tracers for quantitative
microresidue analysis: Trans. ASAE, Vol. 16, pp. 104-114.
Zuber, A., Grabczak, J., and Kolonko, M., 1979, Environmental and artificial
tracers for investigating leakages into salt mines: Isotope Hydrology,
I.A.E.A., Vienna, 1979, Part 1, pp. 45-63.
200
------- |