United Stales Region 4 EPA 904-B-97-003
Environmental Protection (GWDW-15) May 1997
Agency
EPA
GUIDELINES FOR WELLHEAD AND
SPRINGHEAD PROTECTION AREA
DELINEATION IN CARBONATE ROCKS
US Environmental Protection Agency
Water Resource Center RC-4100
1200 Pennsylvania Ave NW
Washington DC 20460
Prepared for:
GROUNDWATER PROTECTION BRANCH
U.S. Environmental Protection Agency
Region 4
Atlanta, Georgia
Prepared by:
ECKENFELDER INC.
227 French Landing Drive
Nashville, Tennessee 37228
(615) 255-2288
October 1996
-------
ACKNOWLEDGMENTS
This document was prepared for the U.S. Environmental Protection Agency,
Region IV, Ground Water Protection Branch (GWPB) under contract number 2R-
1602-NTLX. The document was authored by Geary M. Schindel,
ECKENFELDER INC.; James F. Quinlan, Quinlan & Associates; Gareth Davies,
Cambrian Water Consultants; and Joseph A. Ray, Kentucky Division of Water.
Mr. Robert Olive and Mr. Ronald Mikulak with the GWPB served as Project
Managers for the U.S. EPA
We want to acknowledge and thank the following people for their contribution to
this manual: Sara Evans with the Kentucky Division of Water, Wellhead Protection
Program, Dr. Ronald A Burt with ECKENFELDER INC., and Robert Olive with the
U.S. EPA for review of this manual; Dr. E. Calvin Alexander, Department of
Geology and Geophysics, University of Minnesota, for permission to reprint the short
course manual "Practical Tracing of Groundwater with Emphasis on Karst
Terranes" in Appendix B; Dr. Joseph Saunders for data on Buckner Cave which was
used in preparation of Plate I; Bill Yarnell and Phil O'dell with the Kentucky
Division of Water, Groundwater Branch, for assistance in preparation of Plate I; and
Jacqueline Telford, Lobby Smith, Jackie Thomas, Nerissa Drury, and Pat Utley with
ECKENFELDER INC. for their assistance in preparation of this document.
This manual is dedicated to the efforts of Dr. James F. Quinlan who passed away
during its preparation. His passion for life and search for truth were an inspiration
to all who knew him.
— one well-designed tracer test, properly done, and correctly
interpreted, is worth 1,000 expert opinions ... or 100 computer
simulations of groundwater flow —
James F. Quinlan
1937 to 1995
-------
TABLE OF CONTENTS
Page No.
Acknowledgments
Table of Contents i
List of Tables iii
List of Figures iii
1.0 INTRODUCTION 1-1
1.1 Carbonate Aquifers: Why Be Concerned? 1-1
1.2 Purpose of Document 1-2
2.0 WELLHEAD PROTECTION AREAS FOR CARBONATE AQUIFERS 2-1
2.1 Definition of Wellhead Protection Area (WHPA) 2-1
2.2 Literature Review 2-3
2.3 Defining What is (or is not) a Karst Aquifer 2-4
2.4 Identification of a Karst Terrane 2-6
3.0 HYDROGEOLOGY OF UNCONFINED CARBONATE TERRANES
(KARST TERRANES) 3-1
3.1 Basic Hydrogeologic Characteristics of Karst Terranes 3-1
3.2 Hydrogeologic Characterization of Carbonate Aquifers 3-2
3.2.1 Recharge 3-2
3.3 Groundwater Movement in Carbonate Rocks 3-3
3.3.1 Flow within Residuum (Soil) 3-3
3.3.2 Flow within the Epikarst (Subcutaneous Zone) 3-4
3.3.3 Rapid Flow 3-5
3.3.4 Slow Flow 3-5
3.4 Groundwater Discharge 3-6
3.4.1 Overflow Springs 3-6
3.4.2 Underflow Springs 3-6
3.5 Groundwater Drainage Basins 3-7
3.6 Transmissivity 3-8
3.7 Storativity 3-8
3.8 Identification of Karst Aquifers 3-9
3.9 Karst Terranes in the USA 3-11
-------
TABLE OF CONTENTS (Continued)
Page No.
4.0 DEVELOPING A WELLHEAD PROTECTION AREA FOR CARBONATE
AQUIFERS 4-1
4.1 Definition of Wellhead Protection Area 4-1
4.2 Protection Goals 4-1
4.3 Hydrodynamic Criteria for Delineation of Wellhead Protection
Areas for Carbonate Aquifers 4-2
4.3.1 Distance 4-3
4.3.2 Drawdown 4-4
4.3.3 Time of Travel 4-5
4.3.4 Flow Boundaries 4-7
4.3.5 Assimilative Capacity 4-8
4.4 Review of Standard Methods for WHPA Delineation 4-8
4.5 Recommended Methods for Wellhead Delineation in Carbonate
Terranes 4-11
4.5.1 Mapping of Flow Boundaries 4-11
4.5.2 Balancing of Discharge 4-14
4.5.3 Tracer Testing 4-15
4.6 Summary of Wellhead Delineation Methods for Carbonate Terranes 4-17
4.6.1 Mapping of Flow Boundaries 4-17
4.6.2 Balancing of Discharge 4-17
4.6.3 Tracer Testing 4-18
5.0 BIBLIOGRAPHY 5-1
APPENDICES
Appendix A - Glossary of Terms
Appendix B - Practical Tracing of Groundwater with Emphasis on Karst
Terranes
Appendix C - Determination of the Recharge Area for the Rio Springs
Groundwater Basin, New Munfordville, Kentucky, an Application
of Dye Tracing and Potentiometric Mapping for Delineation of
Springhead and Wellhead Protection Areas in Carbonate Aquifers
and Karst Terranes
-------
LIST OF TABLES
Table No. Title
3-1 Discharges of Selected Karst Springs
3-2 Some Important Carbonate Aquifers in the United States
4-1 Flow Velocities through Karst Conduits for Straight Line Plan
Distances of More than 10 km
4-2 Normalized Base-Flow (NBF) of Groundwater Basins in
Carbonate Terranes, Chiefly in the Mammoth Cave Region,
Kentucky, But Including the Bluegrass Region of Kentucky
4-3 Normalized Base-Flow of Springs in Florida, Missouri, West
Virginia, and Crotia Plus Cave Streams in Missouri and
Indiana
Follows
Page No.
3-5
3-11
4-5
4-14
4-15
LIST OF FIGURES
Follows
Figure No. Title Page No.
2-1 Comparison of Hydraulic Conductivity of Selected Carbonate
Aquifers 2-6
3-1 Schematic of Karst Drainage System 3-2
3-2 Block Diagram of Hypothetical Carbonate Aquifer 3-2
4-1 Map of Rio Springs Showing 1/2 Mile Arbitrary Fixed Radius 4-3
4-2 Map of Rio Springs Kentucky Showing Tracer Tests for
Delineation of the Zone of Contribution 4-6
4-3 Map of Rio Springs Kentucky Showing Hydrologic Boundaries 4-7
4-4 Map of Rio Springs Indicating the Project Base Map Area 4-12
4-5 Map of Rio Springs Kentucky Showing Flow (Hydrologic)
Boundaries and Zone of Contribution 4-13
4-6 Map of Rio Springs Kentucky Showing 5.2 Square Mile Area
Calculated Using Discharge Balancing 4-15
111
-------
1.0 INTRODUCTION
The 1986 Amendments to the Safe Drinking Water Act (SDWA) of 1974 established
a nationwide program to protect groundwater resources used for public drinking
water supplies. A major provision of the Amendments was the establishment of the
Wellhead Protection Program (Section 1428). Through this program, the U.S.
Environmental Protection Agency (EPA) assists States in protecting areas
surrounding public drinking water supply wells against degradation. The emphasis
of the Wellhead Protection Program is on proactive protection of the groundwater
resource rather than focusing on remediation of individual contamination sources.
It is far easier to protect a water supply than to replace or remediate it. This
technical assistance document, "Guidelines for Wellhead and Springhead Protection
Area Delineation in Carbonate Rocks," was developed to provide technical
information to the States and local entities in their implementation of wellhead and
springhead protection programs in carbonate aquifers.
1.1 CARBONATE AQUIFERS: WHY BE CONCERNED?
Most, if not all subaerially-exposed carbonate rock masses probably comprise karst
aquifers. Approximately 20 percent of the U.S. is underlain by near-surface
carbonate rocks, and all of it comprises various types of karst aquifers. It is
estimated that about one-third of U.S. drinking water is derived from carbonate
aquifers. All of them are characterized by the presence of rapid groundwater flow,
thus making them highly susceptible to groundwater contamination (Quinlan et al,
1988).
Springs and wells in carbonate terranes are used to obtain groundwater for various
purposes including public consumption, industrial, agricultural, and aquacultural
use. Springs in such areas may discharge several million gallons per day. In
Tennessee alone, over 100 spring-fed public water supplies serve more than 200,000
people. EPA, by policy decision, decided to consider springs used for public water
supplies eligible for the same wellhead protection safeguards applied to public
water supply wells (Quinlan et al., 1988).
1-1
-------
The consequences of accidental spills or permitted discharges of harmful materials
into carbonate aquifers are in many ways more severe than in other aquifers.
Groundwater velocities in most carbonate terranes are several orders of magnitude
greater than those in granular (sand and gravel) aquifers. In addition, the rapid
movement of groundwater through carbonate systems may minimise attenuation
and result in high concentrations of contaminants affecting a public water-supply
system. Consequently, there are wellhead delineation problems unique to wells and
springs in carbonate terranes (Quinlan et al, 1988). These include the following
items which are especially important in urban and industrial areas and along
transportation corridors:
• What constitutes the wellhead protection area of a well or spring?
• What criteria can be used to delineate it?
• What methods can be used for delineation?
• What is the stage-related (discharge) variability of boundaries of
groundwater basins?
• What is the time-of-travel and why is it important to know for groundwater
in the basin?
• What considerations should be given for emergency response and backup
water-supply systems?
1.2 PURPOSE OF DOCUMENT
The purpose of this technical document is to provide guidance on the various
approaches to delineate wellhead and/or springhead protection areas within
carbonate aquifers. This document is not a treatise on hydrology or hydrologic
investigative techniques. Furthermore, this document is intended to educate the
reader regarding the sensitivity of carbonate aquifers to contamination.
1-2
-------
2.0 WELLHEAD PROTECTION AREAS FOR CARBONATE AQUIFERS
2.1 DEFINITION OF WELLHEAD PROTECTION AREA (WHPA)
A wellhead or springhead protection area refers to "the surface and subsurface area
surrounding a water well or well field, supplying a public water system, through
which contaminants may potentially pass and eventually reach the water well or
well field" (U.S. Environmental Protection Agency, 1987). EPA includes springs
within its definition of a wellhead protection area. In carbonate terranes, springs
are commonly used as water supply sources in place of wells.
The proximity of most public water supply wells and springs to the populations they
serve makes these groundwater sources vulnerable to contamination from human
activities (U.S. Environmental Protection Agency, 1987). Diverse pollution sources
(e.g. gasoline stations, dry cleaners, septic tanks, accidents along transportation
corridors, industrial facilities, and road salting) can adversely impact water
supplies. Contamination source controls and land management programs that
address physical, microbial, and chemical threats to groundwater are important
tools to help prevent well-water and spring-water contamination (U.S.
Environmental Protection Agency, 1987).
Designating protection zones for areas recharging wells and springs is one way to
protect underground water supplies. Wellhead protection areas range anywhere
from a distance of a few hundred feet (meters) to several miles (kilometers) from
wells and springs. The characteristics of the aquifers surrounding the well, well
field, or spring, the extent of pumping, the vulnerability of the aquifer to surface
contamination and the degree of development and activity surrounding the well are
the primary criteria by which most states, counties, and municipalities delineate
protection areas (U.S. Environmental Protection Agency, 1987).
Management activities commonly employed within these protection areas include:
public education, regulation of land use through special ordinances and permits,
prohibition of specified activities, and acquisition of land (U.S. Environmental
Protection Agency, 1987).
2-1
-------
The U.S. EPA identified five standard criteria to delineate wellhead protection
areas (WHPAs). These criteria include:
• distance
• drawdown
• time of travel
flow boundaries
assimilative capacity
However, these criteria have limited application for WHPA development in
carbonate aquifers. Only one of the five standard criteria, identification of flow
boundaries, is valid in unconfined carbonate aquifers.
Groundwater movement in carbonate aquifers differs from that in granular and
fractured aquifers because the flow within carbonates predominately occurs in
dissolutionally-enlarged bedding-plane partings, joints, faults, fissures and
conduits. The general rule of thumb is that an interconnected void larger than
1 centimeter («1/3 inch) is considered a conduit and is capable of rapidly
transmitting large volumes of water. Conduits in carbonates (and other soluble
rocks) can be quite large with some conduits (caves) exceeding 100 feet (30 meters)
in diameter with lengths measured in miles or kilometers. Mammoth Cave,
Kentucky has a mapped length of more than 340 miles (544 kilometers) of
interconnected conduits distributed over five horizontal levels. Groundwater
velocities there have been measured at more than 1,000 feet (300 meters) per hour.
Groundwater in carbonates may have steep or vertical gradients (slope expressed as
the ratio of the vertical drop over a horizontal distance) approaching or equal to one.
The highest waterfall in the eastern United States is literally in carbonate rock. It
is a shaft (pit or vertical cave passage) 510 feet (155 meters) deep (Fantastic Pit) in
a north Georgia cave.
WHPAs in carbonate aquifers can be delineated using three basic methods:
1) hydrogeologic mapping, 2) discharge balancing, and 3) tracer testing. The
optimum method for delineating a carbonate WHPA is A) tracer testing using
fluorescent dyes, combined with hydrogeologic mapping and contouring of
potentiometric data to describe flow boundaries when appropriate; B) Hydrogeologic
2-2
-------
mapping and discharge balancing, the latter using normalized discharge data
(discharge per unit area), can each be used to roughly estimate a WHPA. However,
hydrogeologic mapping can greatly over or underestimate a basin size and discharge
balancing can only estimate basin size, not location. Hydrogeologic mapping and
discharge balancing must be combined to estimate a basins location. Tracer testing
actually proves the relationship between recharge and discharge points as well as
allows an estimate of groundwater time-of-travel.
2.2 LITERATURE REVIEW
Useful guides to WHPA delineation are those published for workshops by EPA (U.S.
EPA, 1987, 1988a, 1989a, 1989b, 1991a, 1993; and Bradbury et al, 1991). A
valuable introduction to some of the technical, analytical subtleties has been written
by Cleary and Cleary (1991). Several training manuals have also been published for
workshops (U. S. EPA, 1988b, 1991b).
Field (1988) reviewed the vulnerability of carbonate aquifers to contamination.
Schindel (1986) and Baker et al. (1992) wrote some of the few published documents
on WHPA planning in American carbonate terranes. The Water Supply Branch of
the Alabama Department of Environmental Management (ADEM, 1991) published
a guide that specifically addresses karst terranes. Karst drainage was considered
by Ray and O'dell (1993) where they utilized tracer flow velocities in conduits as the
primary criterion for assessing susceptibility to groundwater contamination in
Kentucky. A guide to groundwater protection in Florida includes many legal
insights, but largely ignores the non-Darcian nature of flow in karst and its
consequences in most of the aquifers (Wade, 1991).
Useful discussions of WHPA delineation practices in Europe, protection plans for
European carbonate aquifers, and reviews of the literature on same are given in
several proceedings volumes including Societe Nationale des Distribution d'Eaux
and Commission Nationale de Protection des Sites Speleologiques (1982), and the
International Association of Hydrogeologists (1986). Other discussions of European
practice include those by Lallemand-Barres and Roux (1989), Avdagic and Corovic
(1990), and Kresic et al. (1992).
2-3
-------
In the United States, Quinlan et al., (1995) wrote a guide to WHPA delineation,
which includes a discussion on how to perform WHPA delineation in carbonate
terranes.
2.3 DEFINING WHAT IS (or is not) A KARST AQUIFER
There are widely divergent opinions concerning the definition of a karst aquifer and
nature of groundwater flow in carbonate or karstic aquifers (Quinlan et al., 1992).
However, the American Society for Testing and Materials (ASTM) recognizes that
those aquifers (and fractured-rock aquifers) are special cases (ASTM, 1996). If it is
to be defined in words, Quinlan et aZ.,(1996) suggest:
"the definition of a karst aquifer should include criteria consistent both with an
organized, interconnected, and very anisotropic secondary permeability and with
turbulent flow in the more permeable pathways of the aquifer. Using the Reynolds
number as the criterion for the onset of transitional and turbulent flow in fractures
(Huitt, 1956), and typical conduit velocities as determined from more than 2,250
tracer tests in unconfined carbonates (Worthington, 1994), it appears that the
effects of turbulence may occur in fractures with apertures ranging from a few
millimeters to a few centimeters. A similar conclusion is reached using the cubic
law for fracture flow (Domenico and Schwartz, 1990) when the hydraulic gradient in
the smaller-aperture macro-fissures is assumed to be 0.001 or greater. Thus,
turbulent flow probably occurs in most of the preferential flow paths in triple-
porosity aquifers, regardless of their lithology."
Accordingly, a working definition of a karst aquifer is:
A triple-porosity aquifer in carbonate rocks or other readily soluble rocks with
macrofissures or conduits that typically have hydraulic radii at least as large as a
few millimeters.
This definition is based on measurable Emits within which boundaries to
karstification can be defined and avoids descriptive ambiguity. Freeze and Cherry
(1979) also state that in aquifers referred to as karstic: "Flow rates that exceed the
upper limits of Darcy's Law are common." Thus, carbonate aquifers can not be
2-4
-------
assumed to be an isotropic homogeneous porous medium where Darcy's Law is
valid. It should further be emphasized that a carbonate terrane does not need to
look obviously karstic to have some of the same aquifer properties as a karstic
aquifer or triple-porosity aquifer. Also, it should not be assumed that karst rocks
have a high porosity such as suggested by many standard texts (Freeze and Cherry,
1979; Domenico and Schwartz, 1990). The figure of 5 percent (i.e., the lowest the
two texts quote) is not a value that is close to being accurate for any Paleozoic
limestone block that is devoid of fissures or conduits at the scale of a few
millimeters If the porosity values quoted in Freeze and Cherry (1979) or Domenico
and Schwartz (1990) are used, they may result in significant error.
Examples of major carbonate aquifers which do not have well developed karstic
landscapes occur in the Jurassic Great Oolite, Inferior Oolite, and Cretaceous Chalk
in Great Britain. These aquifers have been called predominantly non-karstic
aquifers, but there is also documentation of turbulent (hence rapid) subsurface flow
in fissures in these formations (Atkinson and Smart, 1977) that discharge to
bedrock springs (Atkinson and Smart, 1981). Tracer velocities of up to
540 meters/hour have been measured for the British Chalk (Atkinson and Smith,
1974) and the French Chalk (Crampon et al., 1993). Caves more than 5,000 meters
in length have also been documented in the British Chalk.
Most unconfined carbonate aquifers and other unconfined soluble-rock aquifers
show tendencies of double-porosity or triple-porosity conditions. Using the ASTM
definition, the Edwards Aquifer and the associated carbonates of Texas, which have
very large springs and cave systems (TDWR, 1979), would be included as karst
aquifers. The same applies to carbonates in Florida, south Georgia, and South
Carolina which all have fissures and conduits that consistently supply some wells
and many very large bedrock springs. Tracer velocities of several hundred
meters/hour are reported in macrofissures and conduits that intersected a well in
north Florida (Robinson and Hutchinson, 1991).
Hydraulic conductivity from in situ aquifer tests (slug tests) and from pump tests
(aquifer tests) are commonly used to assess the "characteristics" of a carbonate
aquifer. These data are commonly "interpreted or misinterpreted" to determine if a
carbonate aquifer is "karst." Because of the extreme anisotropy and heterogeneous
2-5
-------
nature of triple porosity aquifers, many boreholes do not intersect conduits or
macrofissures, but still produce water. Many of these boreholes have a hydraulic
conductivity value of less than 10"6 m/s or so (Figure 2-1). Pumping tests commonly
show similar results and an aquifer is commonly misidentified as not being karst.
Figure 2-1 is a logarithmic-probability plot of hydraulic conductivity of three
aquifers, two of which are acknowledged as karstic (Mindip Hills, Great Britain and
Mammoth Cave, Kentucky) and a third that is not (Great Oolite, Great Britain). It
can been seen that all three geometric means are less than 0.001 m/s and thus
would probably be below the upper limit of Darcy flow. Interestingly, the values
and variability of the Jurassic age Great Oolite (Great Britain) aquifer is the
highest mean and it is not referred to as karstic. The Mindip Hills and Mammoth
Cave aquifers are at least two orders of magnitude less than the Great Oolite value
and are considered karstic. Most significantly, the plot includes data from
monitoring wells in the Mammoth Cave aquifer that were drilled within a few tens
of meters from a conduit that is several meters in diameter and is part of the
longest cave in the world (Mammoth Cave). However, the borehole data indicate
that the hydraulic conductivity for the Mammoth Cave aquifer is very low. This
plot illustrates that a low value for hydraulic conductivity is not a reliable
indication that an aquifer is not karstic or does not have conduits or macrofissures.
2.4 IDENTIFICATION OF A KARST TERRANE
A karst terrane is not always as apparent as the classic karst of the Mississippian
Plateau of Kentucky; with its extensive sinkhole development, absence of surface
streams, the presence large caves (Mammoth Cave), underground rivers, and large
springs. Some karst terranes are much more subdued like the Ordovician
limestones in the Inner Bluegrass surrounding Lexington, Kentucky or the Central
Basin surrounding Nashville, Tennessee.
The identification of a karst terrane can be determined by the presence of any one of
the following types of surface and subsurface features (Appendix A is a glossary of
commonly used terms in this manual.). However, these features are diagnostic and
no specific feature is necessary, and a lack (or apparent lack) of nearly all of them
does not mean a terrane is not karst (Quinlan et al, 1992):
2-6
-------
Probability (%)
O.I
1.0E-01 i
^1.0E-02 -
I10E03:
1 1.0E-04 i
T) :
C
0 1.0E-Q5 ;
0 :
1 1 .OE-Q6 -
•o
1.0E-08 :
t
31 2 16 50 84 98 99.
; Slug tests
•
•
I .
,
A AA
.*£
k*
•'
>
^
I
,
>ut**
^^,
1
7
f
A
A i
•
•
i
i -- ' i ; '\ i
3-2-10123
Mammoth Cave (US)*
•
Mendip(UK)*(1)
A
Great Oolite (UK)**(1)
* Paleozoic
** Mesozolc
50% probability =
geometric mean
Standard Deviations
(1) Smart eta!., 1992
FIGURE 2.1 COMPARISON OF HYDRAULIC CONDUCTIVITY OF SELECTED CARBONATE
AQUIFERS
-------
Sinkhole or doline (any closed surface depression, with internal surface
drainage, with or without a discrete opening at the bottom, formed by
dissolution and/or collapse of bedrock, with flushing and/or collapse of soil into
a subjacent cavity).
Dry valley (in humid climates)
Spring (draining carbonate rocks)
Cave
Sinking or losing stream
Dissolutionally-enlarged joint and/or bedding plane (as seen in rock cores or in
outcrops. This includes voids or mud-filled seams encountered during
drilling).
Grikes (soil-filled, dissolutionally-enlarged joints or grooves; also known as
cutters or soil karren; commonly seen in road cuts)
• Karren (dissolutionally, subaerially, water-carved grooves on rock, commonly
subparallel)
A field reconnaissance is the only reliable way to identify most of the above features.
The first five features may be discernible or named on a topographic or geologic
map, but the remaining three can only be identified through diligent field work.
As mentioned previously, the apparent lack or non-recognition of the above features
in a carbonate terrane, does not rule out the existence of a karst terrane or
underlying carbonate or other triple-porosity aquifer. More field work or drilling
may be necessary to locate the features. The existence of one or all of these features
defines the presence of karst, i.e., presence of subaerially exposed carbonate rock.
Similarly, the presence of just one of these features is diagnostic of a karst terrane
2-7
-------
because that feature is usually representative of a much larger population (Quinlan
and Ray; 1991 and Quinlan et al.\ 1995a).
Whether a terrane is or is not karst is a spurious issue that tends to distract from
the fact that all unconfined carbonate aquifers, whatever their surface may look like
or be called, have conduits and dissolutionally-enlarged fissures characterized by
rapid flow in which the velocity is greater than 0.001 m/s and the assumptions of
Darcy's Law are not met (Ford and Williams, 1989). Such aquifers can not be
modeled successfully as porous-medium equivalent aquifers with respect to
groundwater velocity or contaminant transport, largely because of their high
anisotropy turbulent flow, and triple porosity (Huntoon, 1995 and Quinlan et al.,
1996).
Some of the WHPA delineation and groundwater monitoring literature has
classified carbonate aquifers into two distinct types: diffuse flow and conduit flow.
This distinction is based on an early interpretation of variations of water quality
parameters by various investigators. A conceptual model of diffuse-flow and
conduit-flow aquifers by Shuster and White (1971) was based upon hardness
variation at springs, leading to a direct inference of aquifer differences. This
interpretation was in conflict with that of Newson (1971) who suggested recharge
differences rather than aquifer differences, a model suggested by Worthington et al.
(1992). Most of the diffuse-flow (flow associated with primary porosity and closed
fractures) of all unconfined carbonate aquifers is integrated into its conduit flow
system and the use of diffuse flow as a description of karst-aquifer type should be
discontinued (Davies and Quinlan, 1993; Quinlan et al., 1995).
2-8
-------
3.0 HYDROGEOLOGYOF UNCONFINED CARBONATE TERRANES
(KARST TERRANES)
3.1 BASIC HYDROGEOLOGIC CHARACTERISTICS OF KARST
TERRANES
In some areas, landscapes of carbonate aquifers have the appearance of classical
karst landscapes with unique and characteristic landforms and obviously karstic
hydrogeology (Ford and Williams, 1989). However, many if not most triple-porosity
carbonate aquifers have landscapes that do not obviously look karstic, but have
hydrologic characteristics that are obviously karstic when properly investigated. If
hydrology is considered, almost all subaerially exposed carbonate aquifers in humid
climates would arguably be some form of karstic aquifer. Karst can also occur in
gypsum, salt, and carbonate-cemented siliciclastic (sandstone and shale) rocks.
Soluble subaerially exposed non-carbonate rocks are relatively rare in the eastern
United States and are not specifically addressed in this report.
Most carbonate aquifers are composed of either limestone or dolostone, but lithology
can not be used to infer that such aquifers are or are not karstic; the longest
groundwater tracing test done in the U.S (Mountain View, Missouri to Big Spring,
Van Buren Missouri, 43 miles (67 kilometers)) was performed in a dolostone
(Vineyard and Feder, 1982). Unconfined carbonate aquifers in humid climates are
highly anisotropic and heterogeneous and do not approximate the hypothetical
hydrologic conditions of a porous medium equivalent model (PMEM) except possibly
at a regional scale. Meteoric waters recharging these aquifers absorb carbon
dioxide from the atmosphere and vegetation and forming slightly acidic water.
When acidic water interacts with carbonate rock it causes dissolution and
establishes integrated preferential flow paths that lead to development of conduits
that discharge at springs. The development of preferential flow paths can occur
along bedding plane partings, faults, and fractures or at lithologic boundaries. All
unconfined carbonate aquifers in humid climates and some in more arid climates
are karst aquifers and have rapid flow components with velocities >0.001 m/s
(Worthington, 1994).
3-1
-------
Residual soils (unconsolidated regolith) in karst terranes are thought to be
predominately formed from insoluble minerals in the parent rock (both carbonate
and/or non-carbonate) bedrock or from the overlying stratagraphic units. Residual
soils in karst terranes are typically acidic and clay-rich (Jennings, 1985) and have
great variation in thickness and hydraulic conductivity. However, macroporosity of
soils associated with channels formed by decayed root zones, insect and animal
burrows, desiccation cracks, and mass movement from soil piping, facilitates rapid
recharge to the bedrock aquifer below. In addition, groundwater in the soil zone
may be sufficient to constitute a perched aquifer.
3.2 HYDROGEOLOGIC CHARACTERIZATION OF CARBONATE
AQUIFERS
3.2.1 Recharge
There are two principal forms of recharge that occur in carbonate aquifers:
concentrated and dispersed (Figure 3-1 and 3-2). Concentrated recharge occurs at
the swallets of sinking streams (ponors), and dispersed recharge through the
residual soil or numerous exposed fissures (joints, bedding planes) and pores.
Concentrated recharge consists of relatively large volumes of low-hardness water
rapidly recharging the aquifer (Newson, 1971). Dispersed recharge consists of
smaller volumes of percolation water recharging through numerous narrow
fractures and pores that because of lower velocity is constantly tributary to the
conduits and is the high-hardness, base-flow component of recharge (Newson, 1971).
Natural discharge to carbonate aquifers occurs through a distributary network of
springs that drain both overflow and underflow components (Smart, 1983a;
Worthington, 1991).
Recharge through a residual soil occurs through intergranular pores and
macropores (Figure 3-1 and 3-2). Macropores are formed when animal burrows are
preserved or through fractures formed by desiccation or slumping of soil. The most
important hydrologic effect of macropores is the rapid transmission of recharge to
the underlying bedrock, where conduits expedite the transmission of water into the
carbonate aquifer. Macropore flow may occur at rates close to minimum velocities
in conduits (Quinlan and ALey, 1987). Even when recharge is through a thin mantle
3-2
-------
PRECIPITATION
EVAPOTRANS PI RATION
o
LU
ir
UJ
o
LU
O
00
SURFACE DETENTION
CONCENTRATED RECHARGE DISPERSED RECHARGE
ROCK SOIL
SOIL MOISTURE
STORE
CAVE STREAMS
SUBCUTANEOUS ZONE STORE
(EPI KARST)
CONTINUUM OF PERCOLATION
(SEEPAGES-TRICKLES-FLOWS)
SATURATED
ZONE
STORE
WATER FILLED CAVES
WATER FILLED FISSURES
ANDPORES
ri
Vodose
Spring
W
Shallow Deep
Phreatic
Artesian Marine &
Intertidal
Estovelle
Seepage
I
Deep Loss To
Regional Groundwater
System
FIGURE 3-1 SCHEMATIC OF KARST DRAINAGE SYSTEM
MODIFIED FROM FORD & WILLIAMS. 1989
-------
I/,'',.'.';';/.-'] Sandilono (Impeimeable CapiocKI
[ | EpIHantic Zone
Limo»lor>o-Silislon» Contact
- Sill alone (Slighity Ponneabl
E.Doaotl Limestone
Potenhomelric Surlace
v .*," w Wolei -Filled Conduil
Vado»» Conduit
0 Overland
Inleillow
(T) Subculnneoui
0 ShHli
vadoae
f fl j Spung
D'ltuse Discharge
NO CONS IS fCMT SC«U F
FIGURE 3-2 BLOCK DIAGRAM OF HYPOTHETICAL CARBONATE AQUIFER
-------
of glacial till or outwash deposits, recharge velocities may be high (Crowther, 1989).
Macropores are often not observed because they are smeared shut by auger drilling
and even if test holes are drilled in dollies that take storm water unimpeded,
percolation tests in drilled holes often yield misleadingly low hydraulic conductivity
with the bedrock. Therefore, the presence of clay-rich residual soil mantling a karst
aquifer does not necessarily imply low hydraulic conductivity. The soil may be
highly permeable and efficiently transmit water (Jennings, 1985).
Autogenic recharge is recharge derived entirely from the aerially-exposed portion of
a carbonate aquifer (Ford and Williams, 1989). Most autogenic recharge is
dispersed, but it can be concentrated as by the internal runoff within a sink.
Autogenic recharge is characterized by rapid infiltration and minimal overland flow
of precipitation (Jennings, 1985). Autogenic recharge does not equate to diffuse flow
and no conduits as suggested by Shuster and White (1971). All autogenic
catchments have swallets and conduits. Autogenic catchments are dominant in
nine of the ten deepest cave systems in the world and in five of ten of the world's
longest cave systems (Worthington, et al., 1992).
AUogenic recharge is mostly point recharge derived from precipitation on non-
carbonate or less soluble rocks adjacent to, or overlying a carbonate aquifer. The
greatest difference between allogenic recharge and autogenic recharge is the larger
volumes of water input at stream insurgences for allogenic sources. AUogenic water
is also much lower in hardness and the net effect of this water sinking quickly into a
karst aquifer is a marked drop in specific conductivity at the discharge point after
heavy rains (Ford and Williams, 1989). Allogenic streams may flow a considerable
distance over non-carbonate terrane before sinking into carbonate rocks. Newson
(1971) describes this recharge component as a relatively large volume of low-
hardness water that sinks rapidly at swallets.
3.3 GROUNDWATER MOVEMENT IN CARBONATE ROCKS
3.3.1 Flow within Residuum (Soil)
Autogenic recharge to a carbonate aquifer commonly passes through a soil zone.
The texture of this soil is typically a complex interconnected network of pores, voids,
3-3
-------
and macropores, and also is often not a granular aquifer that behaves like a porous
medium, principally because it contains so many micropores. Pankow et al. (1986),
suggest that the assumption that a EPM exists (EPM, i.e., a hypothetical transport
medium that can be assumed to be homogeneous and itotropic), would depend upon
four parameters: fracture width, interfracture spacing, matrix porosity and matrix
diffusion coefficient. Even though matrix porosity could be high in any residual soil,
there is clear evidence that macropore flow is dominant in most carbonate terrane
soils at velocities that might also be turbulent. Generally, in all unconfined
carbonate aquifers, it should be assumed that flow in the residuum is preferential in
macropores and is relatively rapid, and fracture widths, and flow velocity would
prevent the soil zone from behaving as an EMP. In the soil zone, which is known to
exhibit high porosity, invoking the existence of an EPM would fail on fracture width
and interfracture spacing. Matrix diffusion would also be unlikely based on
evidence from traced velocities which are rapid and most likely turbulent.
3.3.2 Flow within the Epikarst (Subcutaneous Zone)
Meteoric water recharging an unconfined carbonate aquifer causes most dissolution
(corrosion) of the bedrock as it flows through the upper-most part of the bedrock,
called the epikarst, subcutaneous zone or epikarstic aquifer (Ford and Williams,
1989) (Figure 3-1 and 3-2). Movement of water through the epikarst is controlled by
vertical drains along intersections of vertical fractures. Horizontal movement of
recharging water between the sometimes infrequent vertical drains tacks more time
to transit the system, so this zone acts as a zone of temporary storage. However,
the vertical drains in the epikarst are hydrologically connected with conduits below
(Williams, 1983; Ford and Williams, 1989). Drilling experience in this soil/bedrock
zone shows that the greatest amount of water is encountered at or near the
soil/bedrock interface or just below it, where conduits occur. Even though, in many
areas, the epikarst appears to be a perched aquifer (especially when inadequately
studied with a few wells), the most important characteristic is that the vertical
drains exist and typically they are open shafts into the conduits below (Figure 3-2).
3-4
-------
3.3.3 Rapid Flow
Rapid-flow components are the most important hydrologic characteristics of
unconfined carbonate aquifers because contaminants can be transported great
distances in short times (mean velocities in the range of kilometers per day are not
unusual). Rapid flow occurs in most vadose conduits and phreatic (submerged)
conduits and is defined as having a velocity of >0.001 m/s (Quinlan et al, 1996;
Worthington et al., 1992)(Figure 3-1 and 3-2). This value is based upon more than
2,250 tracer tests from swallets to cave streams or springs in 25 countries, and is
used because within more than two standard deviations of the geometric mean of
the 2,220 velocities (0.022 m/s), this is the lowest measured conduit velocity and the
upper limit of Darcy flow (Freeze and Cheery, 1979). Rapid flow also occurs in
vertical fissures in the vadose zone, often even where there is a substantial residual
soil cover (Smart and Frederich, 1986). Rapid flow in conduits can be described
using the Darcy-Weisbach equation (equation for fluid flow in a turbulent regime)
where gravitational acceleration, turbulence, and roughness factor must be
considered. Atkinson (1977) determined that the hydraulic conductivity of the non-
conduit portion of the Mendip aquifer was 89 m/day, nearly 0.001 m/s, numerically
close to the figure derived from other data. This non-conduit hydraulic conductivity,
and the known world-wide conduit velocity value, supports the hypothesis that all
carbonate aquifers have rapid-flow components. Large spring discharge, typical of
rapid convergent flow in integrated conduits in karstic aquifers is cited in Table 3-1.
3.3.4 Slow Flow
Slow (<0.001 m/s) sometimes laminar flow can occur in narrow fissures and pores.
Slow flow may also occur in conduits during base-flow conditions, and in near
horizontal conduits containing lakes or pools (Worthington, 1991) (Figure 3-1 and
3-2). Groundwater tracing results suggest that most wells drilled at random
locations intersect only slow-flow portions of the aquifer as the geometric mean of
well-to-well velocities are about an order of magnitude less than conduit velocities.
A few wells may fortuitously intersect conduits but this has a low probability of
occurrence (Benson and la Fountain, 1987) (Khmchouk, 1994 personal
communication) (Figure 3-2). All carbonate aquifers have slow-flow components as
indicated by the maintenance of base flow discharge from underflow springs; but,
3-5
-------
TABLE 3-1
DISCHARGES OF SELECTED KARST SPRINGS
Discharge (m3 s-1)
Spring
Matali, Papua New Guinea
Dumanli, Turkey
Ljubljanica, Yugoslavia
Chingshui, China
Sv. Ivan, Croatia
Vaucluse, France
Frio, Mexico
Coy, Mexico
Lulangdon, China
Ombla, Yugoslavia
Timava, Italy
Waikoropupu, New Zealand
Maligne, Canada
Lawler Blue Hole, Kentucky
Garvin-Beaver, Kentucky
Echo River, Kentucky
Lost River, Kentucky
Pleasant Grove, Kentucky
Shakertown, Kentucky
Gorin Mill, Kentucky
Turnhole, Kentucky
Graham, Kentucky
Rio, Kentucky
McCoy Blue, Kentucky
Roaring, Kentucky
Johnson, Kentucky
Jones School, Kentucky
Royal, Kentucky
Russell Cave, Kentucky
Garretts, Kentucky
Silver, Florida
Rainbow, Florida
Wakulla, Florida
Big, Missouri
Greer, Missouri
Davis, West Virginia
Blue Spring Cave, Indiana
Tumbling Rock Cave, Missouri
Mean
90
50
39
33
--
29
28
24
--
--
17.4
15
13.5
—
--
--
--
--
--
--
--
--
--
--
--
--
--
-•
-
--
--
-•
—
..
-
--
--
-"
Max
>240
--
132
390
--
200
515
200
74.6
165
138
21
45
—
--
--
--
--
--
--
--
--
--
--
--
--
--
--
-
--
--
--
«
--
--
--
--
••
Min
20
25
4.25
4
0.5
4.5
6
13
8.9
4.1
9
5.3
1
0.059
0.048
0.051
0.339
0.071
0.102
0.711
0.405
0.589
0.113
0.348
0.337
0.312
0.065
0.079
0.028
0.014
15.3
13.8
6.5
9.4
6.2
0.8
0.03
0.06
Basin
Area
(k2)
350
2,800
1,100
1,040
65
2,100
>1,000?
>1,000?
1,000
>600
980
450
730
26.48
18.6
22.8
143.0
41.7
49.2
393.7
234.1
315.0
13.5
93.5
28.0
45.3
7.5
64.8
16.6
19.1
1,976
1,670
1,036
1,399
868
186
26
21
Reference
Maire 198 Ib
Karanjac & Gunay 1980
Gospodaric & Habic 1976
Yuan 1981
Qtdnlan & Ray, 1995
Paloc 1970
Fish 1977
Fish 1977
Yuan 1981
Milanovic 1976
Gams 1976
Williams 1977
Kruse 1980
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Quinlan & Ray, 1995
Modified from Ford and Williams (1989) and from Quinlan and Ray (1995).
Page 1 of 1
-------
even though the volume of slowly-flowing groundwater in narrow fissures is high
(Atkinson, 1977), it is all tributary to rapid flow in conduits that all discharge at
springs (Figure 3-1 and 3-2). The term percolation water, used hy Newson (1971),
could be equivalent to slow flow.
3.4 GROUNDWATER DISCHARGE
Most naturally discharging groundwater from karst aquifers occurs from springs.
Groundwater may also be withdrawn from water wells, caves, springs, waterfilled
sinkholes (karst windows, boomshaws, or blueholes), or other points in the aquifer
where conduits are accessible (Figure 3-1 and 3-2). Karst springs can be classified
into two types: underflow springs and overflow springs. Springs may discharge the
full flow of a karst basin or may share the discharge with other springs as part of a
distributary network. The discharge boundaries of a carbonate aquifer consists of a
vertical and horizontal hierarchy of both overflow and underflow springs (Smart,
1983a; Wortbington, 1991).
3.4.1 Overflow Springs
Overflow springs can be recognized by the fact that they may cease flowing during
certain low flow conditions (Smart, 1983a; Worthington, 1991). They may have a
distinct hydrograph form with little recession and consist of mostly bicarbonate
(HCOs-) waters. Overflow springs are part of a subsurface distributary network
with complementary underflow springs. Overflow springs drain the aquifer at flood
stage after the discharge capacity of other parts of the distributary has been
reached. If there are overflow springs, there must be underflow springs that drain
the base-flow component of an aquifer.
3.4.2 Underflow Springs
Underflow springs are perennial springs draining carbonate aquifers at all stages
and preferentially during base flow; therefore, underflow springs are the most
important springs to recognize and monitor in a carbonate aquifer. Smart (1983b)
recognized that underflow springs drain the karst aquifers at Castleguard and
Maligne in the Rocky Mountains of Canada. Underflow springs are characterized
3-6
-------
by low discharge variability, higher sulfate contents, and lower bicarbonate contents
(Worthington, 1991). Commonly, underflow springs are used for public water
supplies because they provide a more dependable supply of water than overflow
springs.
Full-flow springs are underflow springs that drain an entire groundwater basin.
They have a recognizable quickflow and baseflow (slow flow) separation
(Worthington, 1991). It is reasonable to suggest that few, if any, full-flow springs
actually exist. Most discharge occurs through distributaries of overflow and
underflow springs.
3.5 GROUNDWATER DRAINAGE BASINS
Karst groundwater drainage basins are similar in areal form to surface catchments
in non-carbonate terranes. The principal difference is that carbonate aquifers may
have drainage basin boundaries that are variable based on hydrologic conditions,
i.e. the basin may decant water into adjacent basins at different groundwater
elevations and are often not coincident with topographic boundaries (Quinlan, 1989;
Quinlan and Ray, 1989; Quinlan, 1990a). Quinlan and Ray (1989) provides
examples of groundwater basin delineations in the Mammoth Cave area of
Kentucky. Drainage-basin delineation allows predictions to be made about the
probable flow route and discharge points for accidental spills of pollutants. These
drainage basin features must be understood for effective wellhead and springhead
protection. Drainage-basin boundaries in carbonate aquifers typically do not
coincide with topographic divides and in most instances, they are not defined
(bounded) by surface waters (Thrailkill, 1986; Worthington, 1991).
The depth of circulation in carbonate aquifers, according to Worthington (1991), is
related to the length of the drainage basin, mean dip of the bedrock, and relation of
flow direction to the strike. Worthington (1991) shows that an exponential equation
including drainage basin length, bedrock dip, and flow direction explain 92 percent
of the variation for the 40 longest known flow-paths in carbonate aquifers. Previous
depths of circulation were based upon the depths of looping conduits accessed by
cave divers, or conceptual models (Ford and Ewers, 1976. Ford and Williams (1989)
also suggest that depth of circulation is related to the reduction of hydraulic
3-7
-------
conductivity with depth if transmissivity is to be calculated; however, there are
problems with how representative a well-derived value of hydraulic conductivity can
be in any carbonate aquifer (Quinlan et al., 1996)
3.6 TRANSMISSIVITY
The transmissivity of an aquifer is the product of its hydraulic conductivity and
thickness. Both of these parameters tend to be difficult to estimate in unconfined
carbonates or karstified aquifers because the terms are for aquifers that are
assumed to be isotropic, homogenous porous media, of which triple-porosity and
karstic aquifers are not (Quinlan et al, 1996). However, there are several methods
that could estimate transmissivity, e.g., Atkinson (1977) used the depth of looping
conduits and a baseflow recession curve for the spring that is assumed to be
representative of the narrow fissures in the aquifer, the hydraulic gradient, and
cross section of flow within the aquifer. These data enabled reasonable assumptions
to be made because much groundwater tracing had been done and the scale of
investigation was known. Quinlan et al. (1993) used the relationship between
velocity, hydraulic conductivity, effective porosity, and hydraulic gradient to infer,
that, when hydraulic gradient and effective porosity were the same order of
magnitude (where data using the equation V=Ki/ne (Domenico and Schwartz, 1990)
suggest they might often be so), velocity could be equivalent or the same order of
magnitude as hydraulic conductivity. This would enable velocity (a measurement
more comprehensible than hydraulic conductivity) to be inferred from aquifer test
results and showed that traced velocities were always several orders of magnitude
greater than any other results obtained from pumping tests, double-packer tests, or
slug tests in the same aquifer.
3.7 STORATIVITY
The storativity of an aquifer is defined as the product of the specific storage and the
aquifer thickness. Atkinson (1977) calculated the storage for the Mendip aquifer
(Great Britain) from integration using an equation fitted from regression of
baseflow and thus inferred from comparative volumes of conduits and fissure, that
storage was mostly in fissures and pores that were narrow because they were not
dissolutionally widened. Using this storage value, calculating the effective porosity,
3-8
-------
and using velocity from tracer tests, a value for hydraulic conductivity close to the
minimum tracer-test velocity was obtained. Atkinson's technique was criticized by
Ford and Williams (1989) because it inadequately considered the epikarstic aquifer,
which operates as a temporary store because vertical drains are not numerous
enough to always vertically drain this zone efficiently.
3.8 IDENTIFICATION OF KARST AQUIFERS
It should be assumed that if there is an unconfined carbonate aquifer then it is
some form of karstic aquifer. A karstic aquifer is also a special form of triple-
porosity aquifer, not necessarily restricted to having carbonate rocks. Any triple-
porosity aquifer is also likely to be a double permeability aquifer with all three
porosity elements being interconnected. Although Quinlan et. al., (1992) defined a
karst aquifer, any triple-porosity aquifer can be misidentified as a porous medium
aquifer when treated as such and data interpreted accordingly. Criteria for
establishing whether an aquifer behaves as a porous medium are suggested by
Quinlan et al. (1996).
A general rule is that the presence of carbonate rocks indicates the existence of a
karst aquifer. The karstic nature of an aquifer may also be inferred from pumping
tests where the drawdown is stepped, even though pumping is continuous.
However, stepped drawdown curves may also be a consequence of flow through
fractures that have not been dissolutionally enlarged. Whether flow is through
fractures or conduits, groundwater does not occur under Darcian conditions.
Conversely, a smooth drawdown curve does not mean an aquifer is not karstic.
White (1988) noted that values obtained from pump tests vary widely over short
distances, depending on exactly where the wells are drilled. A well that taps a
connection with the conduit system can produce very large quantities of water with
negligible drawdown, leading to extremely large calculated transmissivities. A well
drilled a few meters away in an unfractured block of limestone may have negligibly
small yields. In general, pump-test data are of marginal value in evaluating water
resources in karstic aquifers (White, 1988).
3-9
-------
Other indicators of a karst aquifer, or one that is fractured, or both, are:
• Irregular cone of depressions or no cone of depression, as defined by multiple
observation wells around a pumped well
• Non-linearity of a plot of drawdown vs. discharge
• Stair-stepped, irregular configuration of potentiometric surface (where data
is sufficient to depict the surface with reasonable accuracy)
Bimodal distribution of the logarithms of hydraulic conductivity of a suite of
wells completed in the same formation (Smart et al., 1991). Caution:
carbonate aquifers may also have a unimodal log-normal distribution of
hydraulic conductivity
• Non-coincidence of water levels in closely adjacent wells
Bimodal or polymodal distribution of daily or continuous measurements of
specific conductance (Ford and Williams, 1989; Bakalowicz and Mangin,
1980)
• Significant variation in specific conductance and hydraulic conductivity, as
interpreted with wellbore fluid logs (Padilla et al., 1990)
• Significant variations in the distribution of discharge, as measured by a
current (impeller) meter during pumping at a constant rate (Molz et al.,
1989)
• Significant variations in the distribution of flow in a pumped or unpumped
well, as measured by a movable electromagnetic or thermal flow meter in the
borehole (Nyquist et al, 1991; Kerfoot et al, 1991; Hess and Paillet, 1990)
• Significant differences in the breakthrough patterns for different dyes
injected into several wells and recovered in the same well. (Quinlan, 1992)
3-10
-------
The first four of these indicators are illustrated and briefly discussed by Wisconsin
Geological and Natural History Survey (1991). Proof of appreciable flow through
joints, faults, bedding planes, and cavities can be inferred from cores, video logs,
caliper logs, temperature logs, and packer-test data. Lacking proof to the contrary;
however, if bedrock is limestone or dolomite, one would be wise to assume the
presence of a karst aquifer with or without these features identified.
3.9 KARST TERRANES IN THE USA
Approximately 20 percent of the United States is underlain by limestone or dolomite
rocks. East of the Mississippi River, almost forty percent of the country is
underlain by carbonates at the surface or shallow subsurface (Quinlan, 1989). The
state geologic maps are the most useful guides to the distribution of karst terranes.
The distribution of any formation that has significant amounts of unconfined
carbonate rock corresponds to the distribution of karst terranes.
White and White (1989) produced a good overview of carbonate aquifers in the
United States (Table 3-2). The following is a brief overview of the major carbonate
aquifers in the US. In the Atlantic-Gulf Coastal Plain, most of the karst terranes
are formed in Eocene-Oligocene or Pleistocene limestones located in the Southeast
Coastal Plain (Heath, 1982). Florida and south Georgia share a major karst
aquifer, the Floridan Aquifer, which is a major source of drinking water in Florida.
More than 90 percent of Florida's population is entirely dependent on groundwater
from the carbonates. Florida is also quite famous for its very large springs (e.g.,
Wakulla, Silver, etc.) and karst-related sinkhole-collapse problems.
Major karst terranes are also developed in Paleozoic carbonate rocks of the Valley
and Ridge and Appalachian Plateaus provinces of the Appalachian Mountains belt.
This includes the area from Birmingham, Alabama through Chattanooga,
Tennessee and Knoxville, Tennessee to Bristol, Virginia; the Shenandoah Valley of
Virginia, the Hagerstown Valley in Maryland; the Lehigh Valley of Pennsylvania,
the Greenbriar Valley in West Virginia, etc.
East of the Mississippi River, from northern Alabama through central Tennessee,
central Kentucky, and southern Indiana, is a major area of karst terranes with
3-11
-------
TABLE 3-2
SOME IMPORTANT CARBONATE AQUIFERS IN THE UNITED STATES
Name
Location
Rock Formation
Hydrogeology
Reference
Appalachian Valleys
Appalachian Plateaus
Floridan
Niagara
Illinois Basin
Inner Bluegrass
Highland Rim
Nashville Dome
Ozark Dome
Central PA through VA, WV, TN, Cambrian and Ordovician
QA, to AL limestone and dolomite
WV, TN, KY, and northern AL
Mississippian
Qreenbrier/Oangor/Monteagle/St.
Louis Limestone
Coastal Plan of FL, QA, and SO Eocene through Miocene
Limestones
Northern IL, and along Creat Silurian Niagara dolomite
Lakes
Southern IN, central and western Mississippi Oirkin, Ste. Qenevieve,
KY, southern IL and St. Louis Limestones
Many local aquifers in folded and
faulted carbonates; both fracture
and conduit aquifers common
Many local aquifers; highly
karstified; shallow conduit
systems common
Mainly artesian; much primary
porosity; drowned conduits
Low relief fracture aquifer
Many local aquifers; shallow
conduit systems common
North-central ICY
Central TN, and northern AL
Central TN
Southern MO, northern AL
Ordovician Lexington Limestone Shallow mixed fracture aquifers
Mississippian Limestones
Ordovician Limestones
Mainly localized conduit aquifers
Conduits of moderate depth
Parizeketal. (1971)
Piper (1932), Theis (1936)
Stringfield(1966)
Zeizel et al. (1962)
Thrailkill (1982) Palmquist and
Hall (1961)
Smith (1962)
Moore et al. (1969) Burchett and
Moore (1971)
Mississippian limestone and Many and diverse aquifers
Ordovician limestone and dolomite
PH. 10(1
-------
TABLE 3-2
SOME IMPORTANT CARBONATE AQUIFERS IN THE UNITED STATES
Name
Edwards
El Capitan
Roswell
Pahasapa
Madison Aquifer
Nevada Deep Aquifer
Location Rock Formation
Central TX Cretaceous Edwards, Glen Rose
and related Limestones
Southeastern NM and west TX Permian El Capitan Limestone
Eastern NM Permian San Andres Limestone
and Artesia Group
Western SD Mississippian Pahasapa
Limestone
WY, CO, MT, and UT Mississippian Madison Limestone
NV Sequence of Cambrian and
Ordovician carbonates
Hydrogeology
Deep aquifer feeding large springs
plus smaller shallow systems
deep system with some artesian
parts
Artesian system with some
influence of gypsum beds
Deep artesian aquifer
Many local aquifers, many in
alpine settings
Deep groundwater flow system
Reference
Abbott (1975)
Havener (1968)
Swenson(1968)
Miller (1976)
Hess and Mifflin (1978)
From White and White (1989)
-------
associated carbonate aquifers. The area includes the Nashville Basin, Dissected
Plateaus, and much of western Kentucky and the Bluegrass region of Kentucky.
Almost all groundwater resources in this area are from carbonate aquifers. The
State of Ohio also uses carbonate aquifers as a minor water source (van der Leeden,
1990).
The states of Wisconsin, Michigan, Illinois, Missouri, Iowa, Minnesota, and Kansas
all have extensive terranes underlain by limestone or dolomite, many of which are
till-or loess-covered, unconfined carbonate aquifers.
3-12
-------
4.0 DEVELOPING A WELLHEAD PROTECTION AREA FOR CARBONATE
AQUIFERS
4.1 DEFINITION OF WELLHEAD PROTECTION AREA
As defined previously in Chapter 2, a wellhead protection area refers to "the surface
and subsurface area surrounding a water well or well field, supplying a public water
system, through which contaminants are likely to move toward and reach such
water well or well field" (U.S. Environmental Protection Agency, 1987).
Carbonate aquifers are extremely sensitive to contamination from surface sources
because of their rapid flow components and relatively large recharge areas. For this
reason, it is very important to consider wellhead protection areas for carbonate
aquifers.
4.2 PROTECTION GOALS
The goals of a wellhead protection area for a carbonate aquifer are similar to those
for any aquifer and include one or more of the following:
Providing Time To React to Incidents of Unexpected Contamination
The rapid flow component of groundwater associated with carbonate aquifers
can minimize the reaction time to an unexpected release of potential
contaminants in the recharge area. This reaction time can be extremely short
(on the order of hours to days). The rapid groundwater velocities result in a
management zone for a carbonate aquifer being the entire recharge area (zone
of contribution). Tracer testing in a carbonate aquifer can give a reliable
estimate for the time of travel of contaminants and allow for appropriate
monitoring and implementation of emergency procedures. Tracer testing is
usually conducted in the rapid flow (conduit) portion of an aquifer. Commonly,
sinking streams (or cave streams, open sinkholes, or open joints) are traced to
cave streams or springs. The rapid flow component of groundwater does not
account for the sometimes several orders-of-magnitude lower velocities that
may occur through soil and/or matrix blocks of rock before groundwater
4-1
-------
intersects a conduit. Thus, the time required for water (and pollutants) to
reach a withdrawal or discharge point in a carbonate aquifer may range from
hours to days or months to even years.
• Lowering Concentrations of a Contaminant to Target Levels Before
Contaminants Reach a Well or Spring
Dilution is a major mechanism for lowering contaminant concentrations in the
rapid flow component of carbonate aquifers. However, dilution may not be
sufficient to lower contaminant concentrations to public drinking water
standards. Dilution in carbonate aquifers is dependent upon many factors
including the size, shape and gradient of the recharge area; rainfall intensity
and duration before and during the release; contaminant volume and
concentration; release location and duration; aquifer storage; time of travel;
etc. In most cases, traditional groundwater remediation methods, like
groundwater removal or hydraulic barriers, will not be effective in the rapid
flow component of carbonate aquifers. However, the rapid response to spills on
the land surface with the proper removal or remediation equipment may
prevent pollution from entering the aquifer.
Protection of the Zone of Contribution from Contamination
The purpose of delineating a wellhead protection area is to prevent
contamination of all or part of the zone of contribution. The high groundwater
velocities in carbonate aquifers usually require that the entire zone of
contribution be delineated and protected.
4.3 HYDRODYNAMIC CRITERIA FOR DELINEATION OF WELLHEAD
PROTECTION AREAS FOR CARBONATE AQUIFERS
The U.S. Environmental Protection Agency (1987, 1993) recommended five criteria
as the technical basis for delineating wellhead protection areas. These are
hydrodynamic criteria because they define the wellhead protection area by flow
characteristics of the aquifer:
4-2
-------
Distance
Drawdown
• Time of travel
• Flow boundaries
• Assimilative capacity
All five of the standard criteria generally recognized as forming the technical basis
for delineation of WHPA (see above) are, separately or in partial combination,
usually valid in granular aquifers. They are reviewed by Bradbury et al. (1991), and
Kreitler and Senger (1991). Only one, flow boundaries, is valid in unconfined or
locally partially-confined carbonate (or other karst) aquifers. Distance, drawdown,
and time of travel (TOT) are intrinsically Limited for WHPA determination in
unconfined carbonate aquifers because they erroneously assume Darcian flow.
Assimilative capacity may be valid and has sometimes been used for specific
contaminants in granular and carbonate aquifers. (However, it is not known to
have been used for a wide range of chemical parameters in any aquifer supplying
drinking water).
4.3.1 Distance
Using the distance criterion, a wellhead/springhead protection area is delineated
using an arbitrary fixed radius or dimension measured from the point of water
withdrawal to the protection area boundary. The distance criterion represents the
simplest, least expensive, and most arbitrary criterion used for delineating a
wellhead protection area for any aquifer. Considering that tracer tests have been
conducted in the United States that indicate a linear distance of up to 40 miles (
about 65 kilometers) for the zone of contribution to some springs draining carbonate
aquifers, the distance criteria will most likely be under protective.
An example of an arbitrary fixed radius for a carbonate aquifer is indicated in
Figure 4-1. An arbitrary fixed radius of approximately one-half mile (0.8 kilometers
or 2,500 feet) was placed tangent to, and up topographic gradient of Rio Springs, a
source of water for the Green River Valley Water District (GRVWD) in Kentucky
which serves approximately 30,000 people. (The Rio Springs zone of contribution
was delineated as part of a U.S. Environmental Protection Agency Wellhead
4-3
-------
1/2 MILE ARBITRARY
FIXED RADIUS
Y
TRACED aOW ROUTE OF DYE (Interred Po(h)
(FROM OUINLAN & RAY. 1984)
RIVER AND PERENNIAL SURFACE STREAM
DTT. INJECTION PCHNT
• SPRING
____ RIO SPRINGS ZONE OF CONTRIBUTION AS
DETERMINED EIY TRACER TESTING
(fROU WMNDFI, nillNIAN. ft RAY, 1"»14)
1/2 MILE ARBITRARY FIXED RADIUS
2 MILtS
AREA OF MAP CWERACE
FIGURE 4-1 MAP OF RIO SPRINGS SHOWING 1/2 MILE ARBITRARY FIXED RADIUS
-------
Protection Area Demonstration Project for karst terranes; Schindel et al., 1994).
The arbitrary fixed radius of 2,500 feet (800 meters), a commonly used distance
criterion for granular aquifers, only covers approximately 20 percent of the zone of
contribution of the Rio Springs groundwater basin. Centering the fixed radius at
Rio Springs would result in even less coverage.
The use of an arbitrary fixed radius for a pumping well or spring in unconfined
carbonates would also most likely result in under protection of the resource. A
conventional travel-distance of 2,500 feet (800 meters), which may be practical in
most granular aquifers, could be traveled in only a few hours in many carbonate
aquifers — far too short a response time to offer significant protection.
4.3.2 Drawdown
Drawdown is the decline in water-level elevation resulting from the pumping of a
well or spring. The areal extent over which drawdown occurs is referred to as the
zone of influence or the areal extent of the cone of depression of the pumping well.
Drawdown, although measurable, is almost always invalid as a protection criterion
in carbonate aquifers because most moderate- to high-yield water wells intersect
conduits that have turbulent, non-Darcian flow. The conduit networks violate the
assumptions of homogeneity and isotropy that are made when using the standard
analytical equations for prediction of flow to a well (Dawson and Istok, 1991).
Pumping rate and aquifer properties control the drawdown and zone of pumping
influence in karst aquifers as well, but these are so greatly affected by aquifer
anisotropy as to be incalculable in most, but not all, carbonate aquifers. Rarely does
one have enough data to enable quantification of anisotropy.
Because of the rapid non-Darcian flow characteristics of carbonate aquifers,
pumping from wells which intersect large conduits may not actually produce a
significant cone of depression. In addition, many springs in carbonate rocks
discharge water as gravity flow springs and a cone of depression cannot be
produced.
Rio Springs, Kentucky is an example of a gravity flow spring used as a public water
supply. Rio Springs flows into an impoundment where it is pumped across the
4-4
-------
Green River to the Green River Valley Water District (GRVWD) water treatment
plant. The water from Rio Springs has a lower sediment load than the water from
the Green River and is the preferred water source for the GRVWD.
4.3.3 Time of Travel
Time of travel (TOT) is a criterion using the time for groundwater (or a groundwater
contaminant moving at the same rate) to flow from a point of interest to a water
withdrawal source. Delineating a WHPA with a TOT criterion implies that the TOT
is translated into a distance from the well or spring through determination of
groundwater velocity.
Because of the rapid flow component of carbonate aquifers, TOT has limited
application for groundwater protection in carbonate aquifers. The use of TOT as a
criterion for protection of a withdrawal source is based on groundwater movement
in a porous medium equivalent model where groundwater seepage velocities are
low, generally ranging from inches to feet per day (centimeters to meters).
However, the mean value for seepage velocity in the rapid (conduit) flow portion of
carbonate aquifers is 0.022 m/s (6,200 feet per day). This velocity was obtained by
evaluating the results from more than 2,000 tracer tests in 25 countries (Quinlan et
al., 1993). Table 4-1 lists selected examples of flow velocities through karst conduits
for straight line distances of more than 10 kilometers.
The distance represented by, for example, a 10-year TOT, is unrealistic when
invoked for most carbonate aquifers. Using the mean seepage velocity of 0.022 m/s
stated above, the theoretical average distance that could be traveled in 10 years is
4,300 miles (6,900 kilometers) (greater than the distance between New York and
Los Angeles). If one allows for +/- one order of magnitude range in flow velocities,
0.22 to 0.0022 m/s, extreme higher or lower flow rates would increase or decrease
this 10-year distance by a factor of 10, giving a range of 43,000 (69,000 kilometers)
or 430 miles (690 kilometers). In essence, the use of TOT in karst dictates that the
entire groundwater basin upgradient of a well or spring be included in the WHPA.
TOT-based regulations adopted for WHPAs in unconfined or locally partially-
confined carbonate aquifers, if based on aquifer-test data or computer simulations
4-5
-------
TABLE 4-1
FLOW VELOCITIES THROUGH KARST CONDUITS FOR STRAIGHT LINE PLAN
DISTANCES OF MORE THAN 10 km.
Locality
Taurus, Turkey
Vaucluse, France
Jura
Herault
Aude
Lot
Aach, Germany
Timava, Italy
Ombrie
Maligne, Canada
Missouri, USA
Waikoropupu, NZ
Distance
(km)
134
103
81
35
22
46
24
22
40
21
21
16.5
15.4
11.6
16.6
12.75
12.25
11
12.5
41
11.5
11
16
25
20.2
Descent
(m)
1,095
735
985
1,030
920
545
420
500
360
521
112
60
85
600
550
531
429
136
170
322
326
240
380
<200
50
Time
(h)
8,860
5,424
8,740
816
456
600
1,990
840
2,200
180
2,950
405
672
96
408
336
408
2,400
56
210
15
15
11-80
3-7 x 104
Velocity
(m h-1)
15.2
19.4
9.2
42
48
76.6
12
26.2
18
117
7.2
54
23
120
39.6
50.4
28.8
4.5
221
195
766
733
200-1,450
13-213
0.3-0.7
From Ford and Williams, 1989.
Page 1 of 1
-------
(which generally assume Darcian flow in an equivalent-porous-medium) and not
based on tracer-test data, do not offer significant protection of an aquifer (Quinlan
et al, 1995).
If TOT is based on a properly-designed tracer test or series of tests (testing the
rapid flow component of the aquifer) over a moderate distance, say 500 feet
(170 meters) or more, and if dye is recovered, the TOT determination is probably
representative of the system. Wells, as opposed to sinking streams or sinkholes, are
the least desirable dye-injection or dye-recovery points because of the moderate to
high probability that they do not intersect part of the actively-functioning flow
system of an aquifer.
Although a tracer test for determination of TOT should characterize a maximum
natural flow-velocity, it should not be conducted at the peak of an intense storm-
event. The test should be conducted during recession after a storm, the most
optimal tracing condition. If successful, the test should generally be repeated
during base-flow conditions in order to provide an awareness of the natural range in
flow velocity. The reliability of a tracer test TOT is directly, but imperfectly,
proportional to the distance traveled. The TOT for a tracer to flow the full length of
a very large groundwater basin, say 30 miles (48 kilometers) long, from time of
injection to time of maximum concentration, could be expected to range from about
2.5 to 250 days, with a mean of about 25 days.
Figure 4-2 is a map of the recharge area and tracer tests conducted for delineation
of the Rio Springs drainage basin in Hart County, Kentucky (Schindel, et al, 1994).
Six positive dye traces were conducted during the demonstration project which
indicated travel times of 72 hours to 15 days. The linear distance of the tracer tests
ranged from 2 miles (3.2 kilometers) to 6 miles (9.6 kilometers).
Distance, drawdown, and TOT, can be (but only rarely) valid criteria for WHPA
delineation in unconfined carbonates. For example, these methods may be
appropriate if the well to be protected is located within a large matrix-block of an
aquifer and at maximal distance from (and not connected to) all adjacent,
hydrologically-integrated conduits. Such wells are uncommon. Even in carbonate
aquifers having high primary porosity, such as the Floridan aquifer, nearly all the
4-6
-------
/ j/rvv SPRINGS/.
/ r /ZONE, OF//I 4
/* .-x LCONfRIBIJTION I 4 \
^r ,-' ^CONTRIBUTION I 4
j^
LEGEND
TRACED FLOW ROUTE OF DYE (Inferred Palh)
(FROM QUINLAN it RAY. 1984)
RrVER AND PERENNIAL SURFACE STREAM
DYE INJECTION POINT
SPRING
RK> SPRINGS ZONE OF CONTRIBUTION AS
DETERMINED BY TRACER TESTING
(FROM SCHINOEl, OUINLAN. it RAY. 1994)
TKACEO FLOW ROLTTE OF DYE (Inferred Path)
(FROM SCHINOEL, OUINLAN. It RAY. 1994)
2 MILES
FIGURE 4-2 MAP OF RIO SPRINGS KENTUCKY SHOWING TRACER TESTS
FOR DEUNEATION OF THE ZONE OF CONTRIBUTION
AREA OF UAP COVERAGE
-------
high-capacity wells produce from conduits (caves). Wells in most Paleozoic
carbonates in the midwestern and eastern U.S., including the Appalachians, are not
productive unless they intercept cave streams or cavernous porosity in the phreatic
zone. Without tracing to demonstrate that the well is isolated from the
hydrologically-active conduits, such isolation is unpredictable. A well should not be
assumed to be in an isolated matrix-block until proven by tracer testing.
4.3.4 Flow Boundaries
The flow-boundary criterion for delineating a wellhead protection area uses the
concept of locating groundwater divides or other physical hydrologic features that
control groundwater flow and define geographic areas that contribute groundwater
to a producing well or spring. This area is defined as the zone of contribution.
Depending upon the method and level of effort in determining flow boundaries, the
estimated zone of contribution may be much larger than the actual zone of
contribution.
Flow boundaries can be geologic, caused by changes in lithology (rock) or in bedrock
structure (e.g., faults) across which no flow occurs, or hydrologic, such as
groundwater divides. Groundwater divides can be natural, such as those that
reflect topography, or be human induced, such as those created by a lake or
pumping well.
Figure 4-3 is a map of the Rio Springs area in Hart County, Kentucky indicating the
hydrologic boundaries for the recharge area. The hydrologic boundary for Rio
Springs and the immediately adjacent groundwater basins were determined from
review of the topographic and geologic maps and from interpretation of previous
water-tracing studies of the region conducted by Quinlan and Ray (Schindel et al.,
1994). The hydrologic boundary was identified as: Green River, to the south; Lynn
Camp Creek, to the east; Laurel Branch, Brushy Fork, and Bacon Creek, to the
north; and the inferred southward fiowline of a dye trace to the west. Stream
incision along the Green River and lower part of Lynn Camp Creek has exposed
clayey, silty beds that perch springs at and near the contact between the St. Louis
and Salem-Warsaw Limestone. These beds at the top of the Salem-Warsaw were
4-7
-------
f/RIO SPRINGS/i
I /ZONE OF/ J
V
IKACtU fLOW ROUTE OF DYE (ln(eir«d Poth)
(FROM QUINLAN & RAY. 1984)
DMT!) AND PERENNIAL SURFACE STREAM
DYE INJECTION POINT
SPRING
— - RIO SPRINGS 7ONE OF CONTRIBUTION
(FROM SCHINOEL, OUINLAN, it RAY. 1994)
"•« HYDROLOKIC BOUNDARY
2 MILES
AREA OF MAP COVERAGE
RGURE 4-3 MAP OF RIO SPRINGS KENTUCKY SHOWING HYDROLOGIC BOUNDARIES
-------
considered to be the basal hydrologic boundary of the near-surface aquifer (Schindel
et al, 1994).
As indicated in Figure 4-3, the area delineated by hydrologic boundary mapping
(dotted line) has significantly overestimated the area of the Rio Springs basin by
approximately 10 times as defined by tracer testing (hashed area).
4.3.5 Assimilative Capacity
The assimilative capacity criterion uses the concept that the saturated and/or
unsaturated section of an aquifer can attenuate contaminants to acceptable levels
before the contaminant reaches a well screen or spring. This attenuation process
results from dilution, dispersion, adsorption, and chemical precipitation or
biological degradation. These processes have all been documented to occur and play
important roles in the remediation of contaminated groundwater. However,
consideration of these processes involves sophisticated treatment of contaminant
transport phenomena, which requires detailed information on the hydrology,
geology, and geochemistry of the area of investigation and is typically unavailable.
The inclusion of these processes into wellhead protection strategies is, therefore,
generally not performed.
4.4 REVIEW OF STANDARD METHODS FOR WHPA DELINEATION
All six of the standard methods for delineation of a WHPA -- arbitrary fixed radii,
calculated fixed radii, simplified variable shapes, analytical methods, numerical
flow models, and hydrogeologic mapping (U.S. EPA, 1987, 1993) -- can separately or
in partial combination be valid in granular aquifers. The method to be selected for
such aquifers is determined by evaluating the probable hydrogeology of the area,
the accuracy required, and the budget available.
Only one method, hydrogeologic mapping, is likely to be valid when applied to
unconfined or locally partially-confined carbonate aquifers. The other five standard
methods are intrinsically incapable of WHPA delineation in carbonate aquifers
because of the assumptions of Darcian flow conditions. Notably missing from the
4-8
-------
above listing of methods is groundwater tracing, but it is considered to be an
essential tool in hydrogeologic mapping (U.S. EPA, 1987, 1993).
Arbitrary and fixed radii and simplified variable shapes are too simplistic to be
applied by themselves to the complexities and anisotropy of carbonate aquifers.
Nevertheless, fixed radii have been seriously advocated for carbonate terranes. For
example, compilation of all 17 available tracer-test velocities from one state yields a
mean velocity of 0.019 m/s (not significantly different from the mean of 0.022 m/s for
the aforementioned 2,000 + traces). Assuming a 0.022 m/s velocity of flow in
unconfined carbonates, the resulting mean time of travel of 1.92 hours (or
approximately 2 hours of protection) for a 170 meter (500-foot) radius. Assuming a
+/- one-order-of-magnitude variation, and neglecting the time of travel for
groundwater from the surface to the subjacent conduits, the range in response time
(protection time) ranges from 0.2 to 20 hours — 12 minutes to less than a day! An
arbitrary fixed radius, without a sound technical rationale, is insufficiently
protective for karst aquifers.
Analytical delineation methods also are invalid because of anisotropy, non-Darcian
flow, the scale-effect of measurement of hydraulic conductivity (Quinlan et al., 1993
and Smart et al., 1992) and, among other things, the difficulty of obtaining a
meaningful value for saturated thickness in many unconfined carbonate aquifers.
Numerical flow-models can be useful and can approach validity in unconfined
carbonate aquifers at the regional scale. But they are not valid until or unless they
have been history-matched with tracer-test data (Teutsch and Sauter, 1992; Sauter,
1992a, b, c; Bair, 1994; Oreskes et al., 1994). By the time enough tracer-test and
anisotropy data are available to validate the model, one probably no longer needs
the model because the aquifer is most likely adequately understood (Huntoon, 1995;
Huntoon et al., 1995). The anisotropic, dual- or triple-porosity nature of a carbonate
aquifer is either ignored or rationalized out of existence by modelers (e.g., Blanford
and Birdie, 1993).
A brief review of the relations between the scale of a hydrologic problem and the
applicability of the porous-medium equivalency is given by Bradbury et al. (1991).
More recently, Bradbury and Muldoon (1994) have compared the results of using
4-9
-------
the MODFLOW model with those of a stochastic discrete-flow model using field-
measured data on fractures in dolomite. They found MODFLOW to underestimate
the width of a WHPA by a factor of about nine when compared to the discrete-flow
model. This is not a criticism of MODFLOW. The true width is not known. The
MODFLOW model assumes a porous medium and, unless modified, is incapable of
accurately predicting flow in fractured rocks. But their modeling results emphasize
the importance of tracer data to calibrate the numerical model. Teutsch and Sauter
(1992) successfully modified MODFLOW for use in a karst aquifer as a double-
continuum porous equivalent model, but they did so with massive amounts of data
(18,000 data points over a 10-year period) and with the results of many tracer tests.
To adequately model carbonate aquifers for WHPAs would require similar amounts
of data. Quinlan et al., (1996) point out that even this approach only allowed a
simulation of head variation and discharge and not velocity or contaminant
transport characteristic. Current models can not simulate time critical parameters
such as turbulence and multiple continuum concepts.
Isotope data can yield useful information relevant to WHPAs in confined carbonate
aquifers, but not in the unconfined carbonate aquifers with which this document is
primarily concerned. For example, if tritium is present in a water sample, the time
of travel for some of the water recharged from the surface is less than 50 years.
Interpretation of tritium analyses is less ambiguous if none is present, but such a
concentration is likely only in a confined aquifer (Walsh, 1992).
Interpretation of fracture traces and lineaments on aerial photographs is a
powerful, reliable method for siting water wells with a yield far higher than those
drilled at random locations in the same terrane (Parizek, 1976 and Mabee et al.,
1994). While lineament analysis may improve site specific well yields, they are
irrelevant for delineating a WHPA in unconfined carbonate rocks. Fracture trends,
however many are measured and however convincing their statistical trend, will
most likely result in gross errors in determining the direction of groundwater flow
(for example, see the comparison of fracture trends and tracing results in Blavoux et
al., (1992)). Lineament analysis is no substitute for a well-designed, carefully-
executed series of simultaneously-conducted tracer tests. Fracture-trace analysis is
fraught with methodological problems (Wise, 1982); however, Mabee et al. (1994)
proposed some methods to better filter and manage lineament data. Most EPA
4-10
-------
publications do not mention fracture-trace analysis for WHPA delineation (an
exception is Bradbury et al., 1991) in hydrogeologic investigations of WHPA.
However, lineament analysis is all too often conducted as a misguided end in itself.
Fracture traces and lineaments may indicate some flow routing, but they do not
indicate flow source or destination in a carbonate aquifer; important components in
denning a WHPA. Most discharge in carbonate aquifers is through conduits, and
most conduits in the eastern and midwestern U. S. are preferentially developed
along bedding planes. In essence, a fracture trace and lineament study reveals
nothing significant about catchment area for carbonate aquifers.
4.5 RECOMMENDED METHODS FOR WELLHEAD DELINEATION IN
CARBONATE TERRANES
Three methods of WHPA delineation have been identified for use in unconfined
carbonate aquifers by hydrogeologic mapping. They are: 1) Mapping of Flow
Boundaries, 2) Balancing of Discharge, and 3) Tracer Testing.
4.5.1 Mapping of Flow Boundaries
The effort required to map flow boundaries (hydrologic boundaries) for a carbonate
aquifer can range from a desk-top study to a detailed field study as discussed by
Schindel et al. (1995). The level of effort is dependent upon the budget and level of
accuracy desired. Flow boundaries can be determined from the review of
topographic and geologic maps and from the interpretation of previous water-
tracing studies and carefully assessed potentiometric data, where available.
However, it is best performed by someone experienced in tracer testing in your area.
Mapping of flow boundaries is one of the first tasks performed for tracer testing and
the experience and insight of a professional experienced in your area will help
assure that flow boundary mapping includes the zone of contribution but does not
include an excessive amount of area outside the Zone of Contribution (ZOC).
Flow boundaries are defined as the geologic controls and the major surface and
known subsurface streams or flow routes adjacent to the spring or well of interest
which act as boundaries for near-surface groundwater flow in the region. The
4-11
-------
mapping of flow boundaries should, if possible, include the interpretation of existing
potentiometric data. Caution must be exercised when incorporating potentiometric
data from monitoring wells or water wells as the depth of the well may intersect
water which is not part of the near-surface groundwater flow. Likewise, perched
water levels may lead to underestimate of the true depth of groundwater
circulation. In addition, some geologic features such as lithology may control the
boundary of a basin where structural controls, such as faulting, may either act as a
boundary, or may enhance or concentrate dissolution processes in carbonates.
In addition, two sources of error may be added to produce a variation of ± 7 feet
from the true potentiometric surface when constructed on a 10 foot contour interval
map (Schindel et al., 1994). Contour elevations errors on a traditional 7.5-minute
USGS topographic map may be up to one half the contour interval (Southard, 1983).
Also, depending on the experience of a field worker, estimation of the wellhead
elevation on a map with a contour interval of 10-feet, may be off by 2 feet or more.
Precision surveying of wellhead elevations may not be economically practical.
Consequently, a potential 14 foot error range, especially when applied to a flat-lying
terrain, may discourage the labor-intensive production of a potentiometric map (Ray
and Stapleton, 1996).
The flow boundary of Rio Springs was determined as part of an EPA funded
Wellhead Protection Demonstration Project for karst terranes (Schindel et al.,
1994). The boundary was determined from review of the topographic and geologic
maps and from interpretation of previous water-tracing studies of the region
conducted by Quinlan and Ray (1995). It was defined as the major surface and
subsurface streams adjacent to Rio Springs which most probably act as a boundary
for near-surface groundwater flow in the region. The Rio Springs area was
surveyed to ground truth the topographic and geologic maps; in particular, the blue-
line streams identified on the topographic maps were inspected during low flow
conditions to determine the presence or absence of flowing water.
A project base map of the Rio Springs area (1:24,000 scale) was prepared from four
topographic maps and incorporated the hydrologic flow boundary data (Plate 1).
Figure 4-4 indicates the location of the base map and surface streams at a regional
scale. The major surface and subsurface streams adjacent to Rio Springs were
4-12
-------
N
LEGEND
TRACED FLOW ROUTE OF DYE (Interred Polh)
(FROM OUINLAN & RAT. 1984)
RIVER AND PERENNIAL SURFACE STREAM
DYF INJECTION POINT
SPRING
-- AREA OF PROJECT BASE MAP
(PLATE 1)
2 MILES
AREA OF MAP COVERAGE
FIGURE 4-4 MAP OF RIO SPRINGS INDICATING THE PROJECT BASE MAP AREA
-------
outlined on the project base map. This boundary was identified as: Green River, to
the south; Lynn Camp Creek, to the east; Laurel Branch, Brushy Fork, and Bacon
Creek, to the north; and the inferred southward flowline of the dye trace from near
Bolton Church to Johnson Spring, to the west (Schindel et al, 1994). Figure 4-5
indicates the Rio Springs flow boundary and the Rio Springs zone of contribution as
determined through tracer testing.
As indicated in Figure 4-5, mapping of flow boundaries is likely to overestimate the
size of the area requiring protection, quite possibly by a factor ranging from 2 to 20.
The flow boundary mapping of the Rio Springs area overestimated the zone of
contribution by a factor of 10. It is also important that the study area be surveyed
to ground truth the topographic and geologic maps during low flow conditions.
Blue-line streams indicated on a topographic map often act as a hydrologic
boundary during high flow conditions, a groundwater recharge source during
moderate flow conditions, or to be entirely absent with all flow occurring
underground during low flow conditions. The following are some selected examples
of rivers which are indicated as perennial blue-line streams but which contain dry
reaches during much of the year: Sinking Creek in Breckenridge County, Kentucky;
Roundstone Creek in Rockcastle, Kentucky; the East fork of the Obey River in
Fentress County, Tennessee; Caney Fork in White County, Tennessee; and Fall
Creek in Rutherford County, Tennessee. Many other examples exist, especially
with small rivers and creeks.
It is preferable to have flow boundary mapping conducted by a karst hydrologist or
other professional with tracer testing experience in your area. Mapping of flow
boundaries is one of the first tasks performed for tracer testing and the experience
and insight of a professional will help assure that flow boundary mapping includes
the Zone of Contribution (ZOC) but does not include an excessive amount of area
outside the ZOC.
The mapping of flow boundaries is generally inexpensive, and depending upon
effort, ranges between $1,000 and $10,000 per basin. It is both the least accurate
and least defensible method of WHPA delineation in carbonate terranes when used
by itself.
4-13
-------
LEGEND
V
TRACED FLOW ROUTE OF DYE (Inferred Poth)
(FROM OUINIAN & RAY. 1984)
RIVER AND PERENNIAL SURFACE STREAM
DYE INJECTION POINT
' I Mil'.
- —— - RIO SPRINGS ZONE OF CONTRIBUTION
(FROM SCHINDEl. OUINLAN. & RAY. 1994)
HYOROLOCIC BOUNDARY
.___ PROJECT BASE MAP AREA
2 MILES
AREA OF MAP COVERAGE
FIGURE 4-5 MAP OF RIO SPRINGS KENTUCKY SHOWING FLOW (HYDROLOGIC) BOUNDARIES AND ZONE OF CONTRIBUTION
-------
4.5.2 Balancing of Discharge
Balancing of discharge is a useful method of estimating the recharge area of springs
in a carhonate terrane and is best utilized in conjunction with tracer testing for
WHPA delineation. Discharge balancing employs a comparison of the known
normalized discharge (discharge per unit area) value for a spring or springs in the
same geologic/geographic setting as the spring of interest. By comparing a ratio of
the discharge and basin area of the known (reference) spring to the discharge of the
spring of interest, an estimate can be made of the size of the recharge area. The
limitations of the method are that the base flow and recharge area for a number of
basins in the same geologic/geographic setting should be known for best results.
The method is less valid where a study area encompasses a change in geologic
setting such as a transition from residuum-mantled limestone plain to dissected
upland with sandstone-capped mesas.
The practical use of discharge balancing for WHPA requires that the reference basin
be located in a setting stratigraphically, structurally, and meteorologically similar
to the basin of interest. The reference basin boundaries should have been
delineated using hydrogeologic mapping, potentiometric mapping (if practical), and
tracer testing (Quinlan et al., 1995). A range of base-flow discharge measurements
should be collected from the reference basin and from the basin of interest.
The wellhead protection area of the basin of interest can then be calculated using
the following formula:
Qi/QrxAr= WHPA
Where Qi equals the summer base-flow discharge of the basin of interest, Qr equals
the summer base-flow discharge of the reference basin, Ar equals the area of the
reference basin, and WHPA equals the estimated wellhead protection area of the
basin of interest.
Quinlan and Ray (1995) gave a detailed explanation of discharge balancing
(normalized base flow (NBF)) and presented two tables which are reproduced below.
Table 4-2 presents NBF of groundwater basins in carbonate terranes, in the
4-14
-------
TABLE 4-2
NORMALIZED BASE-FLOW (NBF) OF GROUNDWATER BASINS IN CARBONATE TERRANES, CHIEFLY IN
THE MAMMOTH CAVE REGION, KENTUCKY, BUT INCLUDING THE BLUEGRASS REGION OF KENTUCKY
Autogenic Recharge
Group
1. With up to 25%
Allogenic
Recharge from
Sandstone-
Capped Ridgetops
or Near-Surface,
Leaky, Chert
Aquitard
Spring
Name
Lawler Blue Hole
Garvin-Beaver
Echo River
Lost River
Pleasant Grove
Shakertown
Basin
Area
(miles2)
10.22
7.22
8.82
55.27
16.18
19.09
Summer
Base Flow3
(feet3/sec, cfe)
2.16
1.76
1.86
12.0
2.5
3.6
Normalized
Base Flow, NBF
(cfs/mile2, cfsm)
0.21
0.24
0.21
0.22
0.16
0.19
Geometric Mean
NBF
(cfsm)
0.20
(0.21)
2. With Significant Gorin Mill
Allogenic Turnhole
Recharge from Graham
Carbonate
Terrane4
1522
90.42
1222
25.11°
14.3H
20.8
0.17
0.16
0.17
0.17
3. With Locally-
Thick, Areally-
Significant Sand
and Gravel
Cover13
4. With much
interbedded
shale, Bluegrass
Region
Rio
Rio"
McCoy Blue
Roaring16
Roaringl7
Johnson
Jones Schooll8
Jones School19
Royal
Russell Cave
Garretts
5.2
6.5 "
se.iis
10.8
17.5
3.9i8
2.9i9
25.0
6.4
7.420
4.7W
4.712
12.3
119 ±1
11.0
2.3
2.3
2.8
1.0
0.520
0.91
0.72
0.34
1.19!6
1.0117
0.63
0.59
0.79
0.11
0.15
0.07
0.70
(0.75)
0.11
The geometric mean for NBF, rather than its arithmetic mean, has been calculated because each of the larger data sets
appears to be log-normally distributed. The man of a log-normal distribution is, by definition, the geometric mean.
arithmetic mean, where different, are shown within parentheses.
Page I at 2
-------
TABLE 4-2 (Continued)
EXPLANATORY NOTES
1. The area of basins was measured with a planimeter.
2. Determined by planimeter from Quinlan and Ray (1989) but modified to delete estimated local recharge areas of
overflow springs.
3. All discharge measurements are by Joe Ray, unless otherwise indicated.
4. From Glasgow Upland, underlain by lower third of St. Louis Limestone, which supports surface streams (Quinlan,
1970).
5. Identification number used by Quinlan and Ray (1989).
6. Discharge by Hess and White (1989). We added 0.2 cfs to theirs for Garvin for the Beaver-spring distributary which
they ignored.
7. From Crawford et al. (1987). At Bowling Green, KY.
8. Calculated by Currens (1994). More than 50 miles (80 km) southwest of Mammoth Cave, in limestones with
interbedded shale.
9. Calculated by Howcroft (1992). More than 40 miles (64 km) southwest of Mammoth Cave, in limestones with
interbedded shale.
10. Assumes all base-flow drainage from Three Springs area (Basin 14B of Quinlan and Ray, 1989) flows to Gorin Mill, (Ray,
1994). Average of base-flow measurements by Hess and White (1989) and by Joe Ray. Gorin Mill spring is the major
underflow spring of the Hidden River Complex that is part of the Bear Wallow Complex, and is the major sub-basin
immediately east of Garvin-Beaver Springs. Gorin Mill Spring is part of a distributary in which water from the town of
Horse Cave, 4.5 miles (6.2 km) to the south, flows to 46 springs at 16 locations along a 5-mile (8-km) reach of Green
River (Quinlan and Ewers, 1986).
11. Based on Hess and White (1989) at 14 cfe, plus our estimate of 0.3 cfs at East Window, in Cedar Sink (Quinlan and
Ewers, 1986).
12. Based on 3 unpublished measurements by the U.S. Geological Survey (D. S. Mull, oral communications, June and
September, 1993).
13. The area! percentage and variations in thickness of sand cover are different in various basins and have not been
determined.
14. Inclusive of hypothesized, undetermined north extension (Schindel et al., 1995; Quinlan et al., 1995).
15. The western boundary of this basin is based on unpublished post-1989 tracer tests by Joe Meiman and Marty Ryan,
National Park Service, Mammoth Cave, Kentucky, in files of Kentucky Division of Water, Frankford, Ky.
16. Based on high value for discharge, 12.9 cfs.
17. Based on low value for discharge, 10,9 cfe.
18. Basin area was deliberately over-estimated. The actual area is probably 25% less than this maximum.
19. A more conservative estimate of recharge area of Jones Spring. See note #18.
20. Measurement by Felton and Currens (1994).
From Quinlan and Ray, 1995.
Page 2 of2
-------
Mammoth Cave Region and the Bluegrass Region of Kentucky indicating differences
in NBF. Recharge to these karst groundwater basins is usually autogenic but also
includes some allogenic sources. Table 4-3 present NBF data of springs in Florida,
Missouri, West Virginia, and Croatia plus cave streams in Missouri and Indiana.
Data on karst groundwater basin delineation areas and discharge data may be
available from the USGS and respective state geological surveys.
The area of the ZOC for the Rio Spring groundwater basin was calculated using
discharge balancing methods and tracer test data from Rio Springs Wellhead
Demonstration project (Schindel et al., 1994). The area of the ZOC for Rio Springs
was determined to range from 5.2 to 6.5 square miles (Table 4-2) and is depicted in
Figure 4-6 where it can be compared to the ZOC as determined by flow boundary
mapping and tracer testing.
As Figure 4-6 indicates, the area defined by discharge balancing may be partially
erroneous; however, its reliability is demonstrable (Quinlan and Ray, 1995).
Balancing of discharge estimates the size of the recharge area, but not the
boundaries of the recharge area, i.e., the size of the area needing protection may be
correct, but the location or shape of the inferred zone of protection may not
correspond with the actual zone of contribution (Figure 4-6) However, the
estimated ZOC can be modified to fit an irregular shape using potentiometric data
which would somewhat improve its reliability. An objective estimate of the WHPA
by discharge balancing is second only to that of a tracer-based study and is most
valid when verified by tracing.
Balancing of discharge, combined with flow boundary mapping in the office and in
the field, will cost in the range of $2,000 to $10,000 (1996 dollars), but most likely
will result in a higher degree of accuracy then does flow boundary mapping alone
4.5.3 Tracer Testing
The most reliable method to delineate a WHPA in an unconfined or locally partially-
confined carbonate aquifer is by combining hydrogeologic mapping with tracer-test
data and, if enough relevant and unambiguous water-level data (about 2 wells per
square mile) are available, by mapping the potentiometric surface. Basin
4-15
-------
TABLE 4-3
NORMALIZED BASE-FLOW OF SPRINGS IN FLORIDA, MISSOURI, WEST VIRGINIA, AND CROTIA PLUS CAVE STREAMS IN MISSOURI AND INDIANA
Spring or Cave Name and
Location
Silver (Florida)
Rainbow (Florida)
Wakulla (Florida)
Big (Missouri)
Greer (Missouri)
Davis (West Virginia)
Sv. Ivan (Croatia)
Blue Springs Cave (Indiana)
Tumbling Rock Cave (Mo.)
Catchment
Area
(miles2)
763
645
400
450
335
72
25.1
10.0
8.0
Base Flow
Discharge
(cfs)
539
487
230
332
220
30
20.6
1.0
0.22
Normalized
Base-Flow
(cfsm)
0.71
0.76
0.68
0.74
0.66
0.42
0.82
0.1
0.03
Geometric
Mean NBF
(cfsm)
0.66
(0.68)
0.70
0.42
0.82
0.1
0.03
Source of Discharge
and Catchment
Data
Faulkner (1973)
Faulkner (1973)
Rupert & Wilson (1989)
Vineyard & Feder (1982) plus
Tom Aley (oral communication,
January 1995)
Bill Jones (oral communication,
January 1995) plus Jones (1973)
Bonacci & Magdalenic (1993)
Palmer (1969)
Tom Aley (January 1995)
The florida springs are unconfined portions of the floridan aquifer, delineation of the florida spring-catchments was from potentiometric data, the miasouri springs and cave
streams were delineated by tracing, the west Virginia spring was delineated by tracing and hydrogeologic mapping, the Croatian spring and the blue spring cave stream
were delineated by tracing, hydrogeologic mapping, and analysis of discharge and water-budget data, the geometric mean for nbf, rather than its arithmetic mean, has been
calculated because the bulk of the data appears to be log-normally distributed, the mean of a log-normal distribution is, by definition, the geometric mean. The arithmetic
mean, where different from the geometric mean, is shown within parentheses
From Quinlan and Ray, 1995
Page 1 of 1
-------
TRACED FLOW ROUTE OF DIE (Inferred Poll.)
(FROM OUINLAN if RAY. I9fi4)
I. f
f RIVi
ER AND PERENNIAL SURFACE STREAM
DVE INJECTION POINT
• SPRING
RIO SPRINGS ZONE OF CONTRIBUTION
(FROM SCHINDEL. CoiNLAN. & Rll. liS-l)
o
HYDROLOGIC BOUNUARY
5.2 SO. MILE AREA C4LULATEO USING
DISCHARGE BALANCING
2 MILES
AREA OF MAP COVERAGE
FIGURE 4-6 MAP OF RIO SPRINGS KENTUCKY SHOWING 5.2 SQ. MILE AREA CALULATED
USING DISCHARGE BALANCING
-------
boundaries, inferred from potentiometric mapping, must be tested by tracing on
each side; otherwise, the boundary remains only partially denned. Contouring of
potentiometric data should be performed only if there is sufficient relevant, reliable
potentiometric data-i.e., that data obtained by extensive field work.
Tracer testing builds on and refines what has already been accomplished by flow
boundary mapping and discharge balancing. If properly done, it gives a high degree
of accuracy and is eminently defensible. It alone provides a range in travel times
that are facts rather than inferences. Unlike flow boundary mapping and discharge
balancing, WHPA delineation by tracing, when done with the necessary amount of
tests (see Appendices B and C), minimizes the probability of misidentification of the
zone of contribution to a particular spring or well. The cost can be high, relative to
the other approaches and generally ranges from $5,000 to $75,000.
A discussion of water tracing techniques for carbonate aquifers is beyond the scope
of this manual. Appendix B includes a copy of a paper by Alexander and Quinlan
(1992) entitled "Practical Tracing of Ground Water with Emphasis on Karst
Terranes." This paper has been included to inform the reader on the methods of
tracer techniques. Because of the possibility of discoloration of a public water
supply, this paper should not be considered a how-to manual and is not a
substitute for tracing being performed by a karst hydrogeologist
experienced in tracing techniques.
Appendix C of this document includes the project completion report for the U.S.
EPA Wellhead Demonstration Project entitled: Determination of the Recharge Area
for the Rio Springs Groundwater Basin, Near Munfordville, Kentucky: An
Application of Dye Tracing and Potentiometric Mapping for Delineation of
Springhead and Wellhead Protection Areas in Carbonate Aquifers and Karst
Terranes. This document describes in detail the methodology for tracer testing for
wellhead and spring delineation in carbonate terranes.
Publications and decisions of the U.S. EPA have repeatedly endorsed and/or
required the use of groundwater tracing in carbonate aquifers. The U.S. EPA and
several state agencies employ personnel who routinely conduct tracer tests. Tracing
is an important part of Kentucky's regulatory policy for groundwater management
4-16
-------
(Harker and Ray, 1995). U.S. EPA has also sponsored a tracer-based demonstration
project on WHPA delineation in a karst terrane (Schindel et al., 1995), a how-to
manual on groundwater tracing (Aley et al., in preparation), and several karst-
related investigations of WHPAs by others that are as yet unpublished.
Nevertheless, tracing as a tool for WHPA delineation in carbonate terranes has
been insufficiently stressed in some wellhead protection publications.
Most tracing in carbonate aquifers is conducted with fluorescent dyes because they
are non-toxic in the concentrations employed (Field et al., 1996) and because they
have an ideal combination of desirable properties, including high detectability, low
sorption, and minimal analytical costs. A discussion of their properties is given by
Smart and Laidlaw (1977) and Kass (1992) and of their use is given by Quinlan
(1986) and Kass (1992). No other tracing substance approaches the optimal
combinations of properties and minimal analytical costs of fluorescent dyes for use
as groundwater tracers in karst and other fractured rocks (Alexander and Quinlan,
1994).
4.6 SUMMARY OF WELLHEAD DELINEATION METHODS FOR
CARBONATE TERRANES
4.6.1 Mapping of Flow Boundaries
Mapping of flow boundaries is likely to overestimate the size of the area requiring
protection, possibly by a factor ranging from 2 to 20. It is both the least accurate
and least defensible method of WHPA delineation in carbonate terranes when used
by itself. The mapping of flow boundaries is generally inexpensive, and depending
upon effort, ranges between $1,000 and $10,000 (1996 dollars) per basin.
4.6.2 Balancing of Discharge
Balancing of discharge estimates the size of the recharge area, but not the
boundaries, i.e., the size of the area needing protection may be correct, but the
location or shape of the inferred Zone Of Contribution (ZOC) may not correspond
with the actual ZOC. An objective estimate of the WHPA by discharge balancing is
4-17
-------
second only to that of a tracer-based study and is most valid when verified by
tracing. Balancing of discharge, combined with flow boundary mapping in the office
and in the field, will cost in the range of $2,000 to $10,000 (1996 dollars).
4.6.3 Tracer Testing
Tracer testing is the most reliable method to delineate a WHPA in an unconfined
carbonate aquifer, especially when combined with flow boundary mapping and
discharge balancing. Tracer testing, when conducted with the necessary amount of
tests, minimizes the probability of misidentification of the ZOC for a particular
spring or well. It is also the most expensive approach. The cost of basin delineation
relative to the other approaches, can be high and generally ranges from $5,000 to
$75,000 (1996 dollars).
4-18
-------
5.0 BIBLIOGRAPHY
Abbott, P. L. 1975. On the hydrology of the Edwards limestone, south-central
Texas. Journal of Hydrology, vol. 24, p. 251-269.
Actcs Du ler Colloque National Sur la Protection des Eaux Souterraines Karstiques,
(Organise par la Commission Permanente d'Etude de la Protection des Eaux
Souterraines et des Cavernes). Besancon, Commission Permanente d'Btude de
la Protection des Eaux Souterraines et des Cavernes, 1980, 460 p.
ADEM (Alabama Department of Environmental Management). 1991. Alabama
Wellhead Protection Program. Montgomery, Al., 77 p.
Alexander, E. C., Jr. and Quinlan, J. F. 1992. Practical tracing of ground water,
with emphasis on karst terranes. Geological Society of America, Boulder,
Colorado. 2 vol., 195 + 133 p., 2nd ed. [Short Course Manual for Annual
Meeting].
Aley, T., Quinlan, J. F., Alexander, E. C., Jr. and Behrens, H. (In preparation). The
joy of dyeing: A compendium of practical techniques for tracing groundwater,
especially in karst terranes. Environmental Protection Agency, Washington,
D. C. [in prep].
Atkinson, T. C. 1977. Diffuse flow and conduit flow in limestone terrain in the
Mendip Hills, Somerset (Great Britain). Journal of Hydrology, vol. 35:93-110.
Atkinson, T. C. and Smart, P. L. 1981. Artificial tracers in karst hydrogeology, in A
Survey of British Hydrogeology. The Royal Society, London, p. 173-190.
Atkinson, T. C. and Smart, P. L. (eds.). 1977. Caves and Karst of Southern
England and South Wales, A Guidebook for the International Congress of
Speleology. British Cave Research Association. 83 p.
Avdagic, I. and Corovic, A 1990. Defining of protective zones of karst mountainous
springs, in Water Resources in Mountainous Areas. International Association
of Hydrogeologists, Congress (12th, Lausanne, Switzerland) Proceedings,
International Association of Hydrogeologists. Memoires, vol. 22, pt. 2, p. 1122-
1129.
Bakalowicz, M. and Mangin, A. 1980. L'aquifere karstique. Sa definition, ses
characteristiques et son identification, in hydrogeologie: Interactions entre
1'eau souterrain et son milieu. Societe Geologique de France, Memoire hors-
serie no. 11, p. 71-79.
Baker, R. M., Jackie, J. L. and Raymond, D. E. 1992. Wellhead Protection Program
Development, City of Madison, Alabama. Geological Survey of Alabama,
Information Series 71. 45 p.
5-1
-------
Baker, R. M. and Osborne, W. E. 1994. Geologic and Hydrologic Investigation in
the Valley Head Area, DeKalb County, Alabama in Support of a Wellhead
Protection Program. Geological Survey of Alabama. Circular 180. 32 p.
Baker, R. M. and Raymond, D. E. 1994. Geologic and Ground-Water Assessment of
the Owens Cross Roads Area for the Development of a Wellhead Protection
Program. Geological Survey of Alabama, Information Series 75. 31 p.
Bates, R. L. and Jackson, J. A., eds. 1980. Glossary of Geology. American
Geological Institute, Falls Church, Va. 2nd ed. 749 p.
Benson, R. C. and La Fountain, L. J. 1984. Evaluation of subsidence or collapse
potential due to subsurface cavities, in Beck, B. F., ed. Sinkholes, Their
Geology, Engineering and Environmental Impact, Proceedings. Balkema,
Rotterdam, p. 201-215.
Blanford, T. N. and Birdie, T. R. 1993. Development of wellhead protection areas
for the major public water supply wells in Hernando County, Florida: Final
technical completion report. [consultant report prepared for Southwest
Florida Water Management District, Brooksville, Fla., by HydroGeoLogic, Inc.,
Herndon, Va.. 109 p.].
Blavoux, B., Mudry, J. and Puig, J-M. 1992. The karst system of the Fountaine de
Vaucluse (south-eastern France). Environmental Geology and Water Sciences,
vol. 19, p. 215-225.
Bonacci, O. 1987. Karst Hydrology. Springer Verlag, Berlin. 184 p.
Bonacci, O. and Magdalenic, A. 1993. The catchment area of the Sv. Ivan karst
spring in Istria (Croatia). Ground Water, vol. 31, p. 767-773.
Bradbury, K. R. and Muldoon, M. A. 1994. Effects of fracture density and
anisotropy on delineation of wellhead-protection areas in fractured rock
aquifers. Applied Hydrogeology, vol. 2, no. 3, p. 17-23.
Bradbury, K. R., Muldoon, M. A., Zaporozec, A., and Levy, J. 1991. Delineation of
Wellhead Protection Areas in Fractured Rocks. U. S. Environmental
Protection Agency, Office of Water, Washington, D. C. EPA 570/9-91-009.
144 p.
Burchett, C. R. and Moore, G. K. 1971. Water Resources in the Upper Stones River
Basin, Central Tennessee. Tennessee Division of Water Resources. Water
Resources Series no. 8. 62 p.
Cleary, C. B. F. and Cleary, R. W. 1991. Delineation of wellhead protection areas:
Theory and practice: Water Science and Technology, vol. 24, p. 239-250.
Crampon, N., Roux, and Bracq, J. C. 1993. in collaboration with Delay, F.,
Lepiller, M., Mary, G., Rasplus, L., and Alcayd, G., in Downing, R. A.,
5-2
-------
Price, M., and Jones, G. P. eds., in The Hydrogeology of the Chalk of
Northwest Europe, Oxford Science Publications. 300 p.
Crawford, N. C., Groves, C. G., Feeney, T. P. and Keller, B. J. 1987. Agriculture
and Urban Nonpoint Source Pollution Impacts on Karst Aquifers in the
Pennyroyal Karst Region of Kentucky: Part 1, Hydrogeology of the Lost River
Groundwater Basin, Warren County, Kentucky. Western Kentucky University,
Center for Cave and Karst Studies, Bowling Green, Ky. [Contract research
report prepared for Division of Water, Kentucky Natural Resources and
Environmental Protection Cabinet, Frankfort, Ky., and Barren River Area
Development District, Bowling Green, Ky.] 229 p.
Crowther, J. 1989. Karst geomorphology of South Wales, in Ford, T. D., ed.,
Limestones and Caves of Wales. Cambridge University Press, p. 20-39.
Currens, J. C. 1994. Characterization and quantification of nonpoint-source
pollutant loads in a conduit-flow-dominated karst aquifer underlying an
intensive-use agricultural region, Kentucky [abs.]. Kentucky Water Resources
Symposium, Lexington, KY, p. 3.
Davies, G. J. 1992. Water temperature variation at springs in the Knox Group
near Oak Ridge, Tennessee, in Quinlan, J. F., ed., Hydrogeology, Ecology,
Monitoring, and Management of Groundwater in Karst Terranes Conference
(3rd, Nashville, Tenn.) Proceedings. National Ground Water Association,
Dublin, Ohio, p. 197-212.
Davies, G. J. and Quinlan, J. F. 1993. There is no such thing as a diffuse-flow
carbonate aquifer if that aquifer is unconfined and subaerially exposed [abs.].
Geological Society of America, Abstracts with Programs, vol. 25, no. 6, p. 211.
Davies, G. J., Quinlan, J. F. and Worthington, S. R. H. 1992. Carbonate aquifers:
Distinctions between conduit flow or diffuse flow should be abandoned in their
classification and in the design of monitoring systems. Geological Society of
America, Abstracts with Programs, vol. 24, no. 7.
Dawson, K. J. and Istok, J. D. 1991. Aquifer Testing: Design and Analysis of
Pumping and Slug Tests. Lewis, Chelsea, Mich. 344 p.
Domenico, P. A. and Schwartz, F. 1990. Physical and Chemical Hydrogeology.
Wiley, New York. 824 p.
Downey, J. S. 1984. Geohydrology of the Madison and associated aquifers in parts
of Montana, North Dakota, South Dakota, and Wyoming. U.S. Geological
Survey, Professional Paper 1273-G. 47 p.
Downing, R. A, Ashford, P. L., Headworth, H. G., Owen, M. and Skinner, A. C.
1981. The use of groundwater for river augmentation, in, A Survey of British
Hydrology, The Royal Society, p. 153-172.
5-3
-------
Downing, R. A., Price, M. and Jones, G. P. 1993. The Hydrogeology of the Chalk of
Northwest Europe," Oxford Science Publications, 300 p.
Edwards, A. J. and Smart, P. L. 1989. Waste disposal on karstified carboniferous
limestone aquifers of England and Wales in Beck, B.F. ed., Engineering and
Environmental Aspects of Sinkholes and Karst Conference (3rd, St. Petersburg,
Florida), Proceedings. Balkema, Rotterdam, The Netherlands, p. 165-182.
Faulkner, G. L. 1973. Geohydrology of the Cross-Florida Barge Canal area with
special reference to the Ocala vicinity. U. S. Geological Survey, Water-
Resources Investigations 73-1. 257 p. [Available only through NTIS,
Springfield, Va., as PB-244 669/AS].
Felton, G. K and Currens, J. C. 1994. Peak flow rate and recession-curve
characteristics of a karst spring in the Inner Bluegrass, Central Kentucky.
Journal of Hydrology, vol. 162, p. 99-118.
Field, M. S. 1988. The vulnerability of karst aquifers to chemical contamination.
Hydrological Science and Technology, vol 4, no. 1-4, p. 130-142.
Field, M. S., Wilhelm, R. G., Quinlan, J. F, and Aley, T. 1995. Use and toxicity of
fluorescent dyes. Environmental Monitoring and Assessment. Vol. 38, p. 75-
96.
Fish, J. E. 1977. Karst Hydrogeology and Geomorphology of the Sierra de El Abra
and the Valles-San Luis Potosi Region. Ph.D. thesis. McMaster University.
Ford, D. C. and Ewers, R. O. 1978. The development of limestone caves in
dimensions of length and depth. Canadian Journal of Earth Sciences, vol. 15,
p. 1783-1798.
Ford, D. C. and Williams, P. 1989. Karst Geomorphology and Hydrology. Unwin
Hyman, London. 601 p. [Revised printing published 1992 by Chapman and
Hall, London].
Freeze, R. A. and Cherry, J. A. 1979. Groundwater. Prentice-Hall, Englewood
Cliffs, N.J. 604 p.
Gams, I. 1976. Hydrogeographic review of the Dinaric and alpine karst in Slovenia
with special regard to corrosion, in Problems of Karst Hydrology in Yugoslavia,
Mems. Serb. Geographic Society, vol. 13, p. 41-52.
Gospodaric, R. and Habic, P. 1976. Underground water tracing. Ljubljana:
Institute of Karst Research, Postojna.
Harker, K and Ray, J. A. 1995. Status of Kentucky's regulatory programs that
address karst, in Beck, B.F., ed., Multidisciplinary Conference on Sinkholes
5-4
-------
and Environmental Impacts of Karst (5th, Gatlinburg, Term.), Proceedings.
Balkema, Rotterdam.
Havener, K. C. 1968. Structure, Stratigraphy, and Hydrology of the Northern
Roswell Artesian Basin, Chaves County, New Mexico. New Mexico Bur. of
Mines and Mineral Resources Circ. 93, 30 p.
Heath, R. C. 1982. Classification of ground-water systems of the United States.
Ground Water, vol. 20, no. 4, p. 393-401.
Hess, A. E. and Paillet, F.L. 1990. Application of the thermal-probe flowmeter in
the hydraulic characterization of fractured rocks, in Paillet, F. L. and
Saunders, W. R., eds., Geophysical Investigations for Geotechnical
Investigations. ASTM Special Technical Publication 1101. American Society
for Testing and Materials (ASTM), Philadelphia, p. 99-112.
Hess, J. W. and Mifflin, M. D. 1978. A Feasibility Study of Water Production From
Deep Carbonate Aquifers in Nevada. Desert Research Institute, Water
Resources Center Publication 41054. 125 p.
Hess, J. W. and White, W. B. 1989. Water budget and physical hydrology, in
White, W. B., and White, E. L., eds. Karst Hydrology: Concepts from the
Mammoth Cave Area. Van Nostrand Reinhold, New York, p. 105-126.
Howcroft, W. D. 1992. Ground Water Flow and Water Resources Investigation of
the Auburn, Summers, and Shakertown Springs Karst Groundwater Basins,
Logan and Simpson Counties, Kentucky. M.S. thesis (Geography), Western
Kentucky University, Bowling Green, Ky. 128 p.
Huitt, J. L. 1956. Fluid flow in simulated fractures. American Institute of
Chemical Engineering Journal, vol. 2, no. 2, p. 259-264.
Huntoon, P. W. 1995. Is it appropriate to apply porous media groundwater
circulation models to karstic aquifers?, in El-Kadi, M., ed., Assessment of
Models for Groundwater Resource Analysis and Management. Lewis
Publishers, Boca Raton, Fla. [in press].
Huntoon, P. W., Quinlan, J. F. and Davies, G. J. 1995. Use and abuse of computer
models to characterize flow in unconfined carbonate aquifers, in
Ritchey, J. D. and Rumbaugh, J., eds., Subsurface Fluid-Flow (Groundwater)
Modeling, {Paper accepted for presentation at ASTM (American Society for
Testing and Materials) symposium, Denver, Colo., June 1995, and to be
published by ASTM in a Special Technical Publication].
International Association of Hydrogeologists. 1986. Groundwater protection in
porous, fractured, and karstic aquifer systems, in Symposium on Ground-
Water Protection Areas (Karlovy Vary, Czechoslovakia), Proceedings.
International Association of Hydrogeologists, Memoires, vol. 19, pt. 2,
p. 89-374.
5-5
-------
Jennings, J. N. 1985. Karst Geomorphology, Blackwell, Oxford. 293 p.
Karanjac, J. and Gunay, G. 1980. Dumanli Spring, Turkey - the largest karstic
spring in the world? Journal of Hydrology, vol. 45, p. 219-231.
Kiiss, W. 1992. Geohydrologische Markierungstechnik: Gebriider Borntraeger,
Berlin. 519 p. [Lehrbuch der Hydrogeologie, vol. 9].
Kerfoot, W. B., Beaulieu, G. and Kieley, L. 1991. Direct-reading borehole
flowmeter results in field applications. Ground Water Management.
(Proceedings of the Fifth National Outdoor Action Conference on Aquifer
Restoration, Ground Water Monitoring, and Geophysical Methods), vol. 5,
p. 1073-1084.
4
Kreitler, C. W. and Senger, R. K. 1991. Wellhead Protection Strategies for
Confined-Aquifer Setting. U.S. Environmental Protection Agency, Office of
Water (WH-550G), EPA 570/0-91-008, June. 168 p.
Kresic, N., Papic, P. and Golubovic, R. 1992. Elements of groundwater protection
in a karst environment. Environmental Geology and Water Sciences, vol. 20,
p. 157-164.
Kruse, P. B. 1980. Karst Investigations of Malign Basin, Jasper National Park,
Alberta (Canada). MSc. Thesis. University of Alberta, Canada.
Lallemand-Barres, A. and Roux, J. C. 1989. Guide Methodologique d'Establissment
de Perimetres de Protection des Captages d'Eau souterraine Destinee a
Consommation humaine: Manuels and Methodes, no. 19, Bureau de
Recherches Geologiques et Minieres, Orleans, France. 221 p.
Mabee, S. B., Hardcastle, K C. and Wise, D. U. 1994. A method of collecting and
analyzing lineaments for regional-scale fracture-bedrock aquifer studies:
Ground Water, vol. 32, p. 884-894.
Maire, R. 1981. Karst and hydrogeology synthesis. Spelunca, Supp. no. 3, p. 23-30.
Milanovic, P. T. 1976. Water regime in deep karst. Case study of the Ombla spring
drainage area, in Yevjevich, ed., Karst Hydrology and Water Resources, vol. 1.
Karst hydrology. Colorado Water Resources Publication, p. 165-191.
Miller, W. R. 1976. Water in Carbonate Rocks of the Madison Group in
Southeastern Montana - A Preliminary Evaluation. U.S. Geological Survey,
Water Supply Paper 2043. 51 p.
Molz, F. J., Morin, R. H., Hess, A. E., Melville, J. G. and Guven, 0. 1989. The
impeller meter for measuring aquifer permeability variations: Evaluation and
comparison with other tests. Water Resources Research, vol. 25, no. 7,
p. 1677-1683.
5-6
-------
Moore, G. K, Burchett, C. R. and Bingham, R. H. 1969. Limestone Hydrology of
the Upper Stones River Basin, Central Tennessee. Tennessee Division of
Water Resources. 51 p.
Newson, M. D. 1971. A model of subterranean limestone erosion in the British
Isles based on hydrology. Institute of British Geographers, Transactions.
Vol. 54, p. 51-70.
Nyquist, J. E., Moore, G. K., Young, S. C. and Clapp, R. B. 1991. Use of
electromagnetic borehole flowmeter to delineate groundwater producing
fractures. Tennessee Water Resources Symposium (4th, Knoxville), Extended
Abstracts, p. 26-28.
Padilla, A., Pulido-Bosch, A. and Mangin, A. 1994. Relative importance of baseflow
and quick flow from hydrographs of karst spring[s]. Ground Water., vol. 32,
p. 257-277.
Palmer, A. N. 1969. A Hydrologic Study of the Indiana Karst. Ph.D. dissertation
(Geology), Indiana University. 181 p.
Paloc, H. 1970. La Fontaine de Vacluse et son bassin d'alimentation. Anns. Soc.
Horticulure etHistoire NatureUe d'Herault, vol. 110(3), p. 130-134.
Pankow, J. F., Johnson, R. L., Hewstson, J. P. and Cherry, J. A. 1986. An
evaluation of contaminant migration patterns at two waste disposal sites on
fractured porous media in terms of the equivalent porous medium (EPM)
mode. Journal of Contaminant Hydrology, vol. 1, p. 6-76.
Parizek, R. R. 1976. On the nature and significance of fracture traces and
lineaments in carbonate and other terranes, in Yevjevich, V., ed., Karst
Hydrology and Water Resources. Water Resources Publications, Fort Collins,
Colo., vol. 1, p. 47-108.
Parizek, R. R., White, W. B. and Langmuir, D. 1971. Hydrogeology and
Geochemistry of Folded and Faulted Rocks of the Central Appalachian Type
and Related Land Use Problems. The Pennsylvania State University Earth
and Mineral Sciences Experiment Station, Circ. 82.
Piper, A. M. 1932. Ground Water in North-Central Tennessee. U.S. Geological
Survey Water Supply Paper 640. 238 p.
Price, M., Downing, R. A. and Edmunds, W. M. 1993. The chalk as an aquifer, in
Downing, R. A., Price, M., and Jones, G. P., eds. The Hydrogeology of the
Chalk of North-West Europe. Oxford University Press, Oxford, p. 25-58.
Quinlan, J. F. 1985. Discussion of Ground Water Tracers, by Davis et al., (1985),
With Emphasis on Dye-Tracing, Especially in Karst Terranes: Ground Water,
vol. 24, p. 253-259 and 396-397. [Reply: p. 398-399].
5-7
-------
Quinlan, J. F. 1986. Recommended procedure for responding to spills of hazardous
materials in karst terranes, in Quinlan, J. F., ed., Environmental Problems in
Karst Terranes and Their Solutions Conference (1st, Bowling Green,
Kentucky), Proceedings. National Water Well Association, Dublin, Ohio,
p. 183-186.
Quinlan, J. F. 1989. Ground-water Monitoring in Karst Terranes: Recommended
Protocol and Implicit Assumptions. U. S. Environmental Protection Agency,
Environmental Monitoring Systems Laboratory, Las Vegas, Nev.
EPA/600/X-89/050, 88 p. Draft].
Quinlan, J. F. 1990a. Special problems of ground-water monitoring in karst
terranes, in Nielsen, D. M. and Johnson, A. I., eds., Ground Water and Vadose
Zone Monitoring, ASTM Special Technical Publication (STP) 1053. American
Society for Testing and Materials (ASTM), Philadelphia, p. 275-304.
Quinlan, J. F. 1990b. Types of karst, in Quinlan, J. F., ed., Practical Karst
Hydrogeology, with Emphasis on Groundwater Monitoring. [Course Manual]
5th ed. Association of Groundwater Scientists and Engineers, Dublin, Ohio.
9:H-ltoH-9.
Quinlan, J. F. and Alexander, E. C., Jr. 1987. How often should samples be taken
at relevant locations for reliable monitoring of pollutants from an agricultural,
waste disposal, or spill site, in Beck, B. F., ed., Multidisciplinary Conference
on Sinkholes and the Environmental Impacts of Karst (2nd, Orlando, Florida),
Proceedings. Balkema, Rotterdam, The Netherlands, p. 277-286.
Quinlan, J. F. and Aley, T. 1987. Discussion of "A new approach to the disposal of
solid waste on land", by R. C. Heath and J. H. Lehr. Groundwater vol. 25,
no. 5, p. 615-616.
Quinlan, J. F., Aley, T. and Schindel G. M. 1988. Wellhead protection in karst
terranes: Problems, examples of problems, recommended solutions, and self-
delusory solutions. Wellhead Protection Conference, New Orleans, Louisiana,
EPA meeting, abstract.
Quinlan, J. F., Davies, G. J. and Worthington, S. R. H. 1993. Review of Ground-
water quality monitoring network designs by Loaiciga, H. A., et al., 1992:
Journal of Hydraulic Engineering, vol. 19, p. 1136-1141. [Discussion with
reply, p. 1141-1142].
Quinlan, J. F., Davies, G. J., Jones, S. W. and Huntoon, P. W. 1996. The
applicability of numerical models to adequately characterize ground-water flow
in karstic and other triple-porosity aquifers, in Subsurface Fluid-Flow (Ground
Water) Modeling, ASTM STP 1288. J. D. Ritchey and J. 0. Rumbaugh,
American Society for Testing and Materials.
Quinlan, J. F. and Ewers, R. O. 1985. Groundwater flow in limestone terranes:
Strategy, rationale and procedure for reliable, efficient monitoring of
5-8
-------
groundwater quality in karst areas, in National Symposium and Exposition on
Aquifer Restoration and Groundwater Monitoring (5th, Columbus, Ohio),
Proceedings. National Water Well Association, Worthington, Ohio, p. 197-234.
Quinlan, J. F. and Ewers. R. 0. 1986. Ground water flow in the Mammoth Cave
Area, Kentucky, with emphasis on Principles, contaminant dispersal,
instrumentation for monitoring water quality, and other methods of study, in
Environmental Problems in Karst Terranes and their Solutions Conference
(1st, Bowling Green, Ky.), Guidebook for Field Trip A. National Water WeU
Association, Dublin, Ohio. 129 p. [revised, corrected reprint of Quinlan et al,
1983, with additions].
Quinlan, J. F. and Ray, J. A. 1989. Groundwater basins in the Mammoth Cave
Region, Kentucky, showing springs, major caves, flow routes, and
potentiometric surface. Friends of the Karst, Occasional Publication No. 2,
Friends of the Karst, Mammoth Cave, Ky. [Map with 2-page explanatory text;
revision of map first published in 1981].
Quinlan, J. F. and Ray, J. A. 1991. Groundwater remediation may be achievable in
some karst aquifers that are contaminated, but it ranges from unlikely to
impossible in most: implications of long-term tracer tests for universal failure
in goal attainment by scientists, consultants, and regulators, in Quinlan, J. F.,
ed., Hydrogeology, Ecology, Monitoring, and Management of Ground Water in
Karst Terranes Conference (3rd, Nashville, Tenn.) Proceedings. National
Ground Water Association, Dublin, Ohio, p. 573-635.
Quinlan, J. F. and Ray, J. A. 1995. Normalized base-flow discharge of groundwater
basins: a useful parameter for estimating recharge area of springs and for
recognizing drainage anomalies in karst terranes, in Beck, B. F., ed.,
Multidisciplinary Conference on Sinkholes and Environmental Impacts of
Karst (5th, Gatlinburg, Tenn.), Proceedings. Balkema, Rotterdam.
Quinlan, J. F., Ray, J. A. and Schindel, G. M. 1995. Intrinsic limitations of
standard criteria and methods for delineation of groundwater-source
protection areas (Springhead and Wellhead Protection Areas) in carbonate
terranes: Critical review, technically-sound resolution of limitations, and case
study in a Kentucky karst, in Beck, B. F, ed., Multidisciplinary Conference on
Sinkholes and Environmental Impacts of Karst (5th, Gatlinburg, Tenn.)
Proceedings. Balkema, Rotterdam.
Quinlan, J. F., Smart, P. L., Schindel, G. M., Alexander, E. C., Jr., Edwards A. J.,
and Smith, A. R. 1992. Recommended administrative/regulatory definition of
karst aquifer, principles for classification of carbonate aquifers, practical
evaluation of vulnerability of karst aquifers, and determination of optimum
sampling frequency at springs, in Quinlan, J. F., ed. Hydrogeology, Ecology,
Monitoring, and Management of Ground Water in Karst Terranes Conference
(3rd, Nashville, Tenn.) Proceedings. National Ground Water Association,
Dublin, Ohio, p. 573-635.
5-9
-------
Ray, J. A. 1994. Surface and subsurface trunk flow: Mammoth Cave, Kentucky.
Mammoth Cave National Park Science Conference (3rd, Mammoth Cave, Ky.),
Proceedings. National Park Service, Mammoth Cave, Ky, p. 175-187.
Robinson, J. L. and Hutchinson, C. B. 1991. Ground-Water Tracer Tests in West-
central Florida [abs.]. American Institute of Hydrology, Annual Meeting
(Orlando, Fla.), Abstracts, p. 39.
Rubin, P.A. 1991. Land use planning and watershed protection in karst terranes,
in Quinlan, J. F., ed., Hydrogeology, Ecology, Monitoring, and Management of
Groundwater in Karst Terranes Conference (3rd, Nashville, Tennessee),
Proceedings. National Groundwater Association, Dublin, Ohio.
Rupert, F., and Wilson, W. L. 1989. The Geology and Hydrology of Wakulla
Springs, in Stone, W. C., ed., The Wakulla Springs Project. U. S. Deep Caving
Team, Derwood, Md, p. 163-174.
Schindel, G. M. 1986. Aquifer protection planning for critical karst groundwater
supplies: A coordinated local and state approach for Kentucky in Quinlan, J.
F. ed., Environmental Problems in Karst Terranes and Their Solutions
Conference (1st, Bowling Green, Kentucky) Proceedings. National
Groundwater Association, Dublin, Ohio.
Schindel, G. M., Quinlan, J. F. and Davies, G. J. 1995. Delineation of groundwater-
source protection areas in carbonate (karst) terranes. [consultant report
prepared for Ground-Water Branch, U. S. Environmental Protection Agency,
Atlanta, Georgia]. Eckenfelder Inc., Nashville, Tennessee. 60 p.
Schindel, G. M., Quinlan, J. F. and Ray, J. A. 1994. Determination of the recharge
area for the Rio Springs Groundwater Basin, near Munfbrdville, Kentucky: An
application of dye tracing and potentiometric mapping for delineation of
springhead and wellhead protection areas in carbonate aquifers in karst
terranes, [Consultant Report Prepared for Ground-Water Branch, U. S.
Environmental Protection Agency, Atlanta, Georgia]. Eckenfelder Inc.,
Nashville, Tennessee. 25 p.
Schindel, G. M., Ray, J. A and Quinlan, J. F. 1995. Delineation of the recharge
area for Rio Springs, Kentucky: An EPA demonstration project in wellhead
(springhead) protection for karst terranes in Beck, B. F., ed.,
Multidisciplinary Conference on Sinkholes and Environmental Impacts of
Karst (5th, Gatlinburg, Term.) Proceedings. Balkema, Rotterdam.
Shuster, E. T. and White, W. B. 1971. Seasonal fluctuations in the chemistry of
limestone springs: A possible means of characterizing carbonate aquifers.
Journal of Hydrology, vol. 14, p. 93-128.
Smart, C. C. 1983b. The Hydrology of the Castleguard Karst, Columbia Icefields,
Alberta, Canada. Arctic and Alpine Research, vol. 15, p. 471-486.
5-10
-------
Smart, P. L., Edwards, A. J. and Hobbs, S. L. 1992. Heterogeneity and carbonate
aquifers: Effects of scale, fissuration, lithology, and karstification, in Quinlan,
J. F., ed., Hydrogeology, Ecology, Monitoring, and Management of Ground
Water in Karst Terranes Conference (3rd, Nashville, Tennessee), Proceedings.
National Ground Water Association, Dublin, Ohio, p. 373-388.
Smart, P. L. and Friedrich, H. 1986. Water movement in the unsaturated zone of a
maturely karstified carbonate aquifer, Mendip Hills, England in Quinlan, J. F.
ed., Environmental Problems in Karst Terranes and Their Solutions
Conference (1st, Bowling Green, Kentucky), Proceedings. National Water Well
Association, Dublin, Ohio, p. 59-87.
Smart, P. L. and Hobbs, S. L. 1986. Characteristics of carbonate aquifers: A
conceptual base, in. Quinlan, J. F., ed., Environmental Problems in Karst
Terranes and Their Solutions Conference (1st, Bowling Green, Kentucky)
Proceedings. National Water Well Association, Dublin, Ohio, p. 1-14.
Smart, P. L. and Laidlaw, I. M. S. 1977. An evaluation of some fluorescent dyes for
water tracing. Water Resources Research, vol. 13, p. 15-33.
Smith, P. M. 1964. The Flint Ridge Cave System: 1957-1962. National
Speleological Society Bulletin, vol. 26, p. 17-27.
Stringfield, V. T. 1966. Artesian Water in Tertiary Limestone in the Southeastern
United States. U.S. Geological Survey Professional Paper 517. 226 p.
Swenson, F. A. 1968. New theory of recharge to the artesian basin of the Dakotas.
Geological Society of America Bulletin no. 79, p. 163-182.
Tennessee Department of Environment and Conservation, Division of Water
Supply. 1994. Guidance Document for Wellhead Protection 1200-5-1-.34.
Division of Water Supply, Nashville. 101 p.
Teutsch, G. and Sauter, M. 1992. Groundwater modeling in karst terrains: Scale
effects, data acquisition, and field validation, in Quinlan, J. F., ed.,
Hydrogeology, Ecology, Monitoring, and Management of Ground Water in
Karst Terranes Conference (3rd, Nashville, Tenn.) Proceedings. National
Ground Water Association, Dublin, Ohio, p. 39-57.
Texas Department of Water Resources. 1979. Geohydrology of Comal, San Marcos,
and Hueco Springs. Report 234. 85 p.
Theis, C. V. 1936. Ground Water in South-Central Tennessee. U.S. Geological
Survey, Water Supply Paper 677. 182 p.
Thrailkill, J. 1972. Carbonate chemistry of aquifer and stream water in Kentucky.
Journal of Hydrology, vol. 16, p. 93-104.
5-11
-------
ilTrin, j. 1985. Flow in a limestone aquifer as determined from water tracing
and water levels in wells. Journal of Hydrology, vol. 78, pp. 123-136.
U. S. EPA (Environmental Protection Agency). 1987. Guidelines for Delineation of
Wellhead Protection Areas. EPA 44016-87-010 (reprint EPA 4401-5-93-001)
Office of Ground-Water Protection, Washington, DC. 209 p.
U. S. EPA 1988a. Wellhead protection area delineation: A "hands-on" training
course. Office of Ground-Water Protection, Washington, DC. ca. 500 p.
U. S. EPA. 1988b. Developing a State Wellhead Protection Program: A User's
Guide to Assist State Agencies under the Safe Water Drinking Act. Office of
Ground-Water Protection, Washington, DC. EPA 440/6-88-003. 44 p.
U. S. EPA. 1989a. Local Financing for Wellhead Protection. Office of Water,
Washington, DC. EPA 44016-89-001, 57 p.
U. S. EPA 1989b. Wellhead Protection Programs: Tools for Local Governments.
Office of Water, Washington, DC. EPA 440/6-89-002. 50 p.
U. S. EPA. I991a. Wellhead protection strategies for confined-aquifer settings. U.
S. Environmental Protection Agency, Office of Groundwater and Drinking
Water, Washington, D. C. EPA 570/9-91-008. 168 p.
U. S. EPA. 1991b. Wellhead protection area delineation methods: A training
workshop. U. S. EPA, Kegion H, Drinking/Ground-Water Protection Branch,
New York. ca. 300 p.
U. S. EPA- 1993. Wellhead Protection: A Guide for Small Communities. Office of
Research and Development, EPA/625/R-93/002. 144 p.
van der Leeden, F., Troise, F. L. and Todd, D. K 1990. The Water Encyclopedia.
Lewis, Chelsea, Mich. 808 p.
Vineyard, J. D. and Feder, G. L. 1982. Springs of Missouri. Missouri Division of
Geology and Land Survey, Water Resources Report no. 29. 212 p.
Wade, J. 1991- Guide to local groundwater protection in Florida. University of
Florida, Center for Governmental Responsibility, Gainesville. 3 vols., 38 +
263 + 129 p.
Walsh, J. F. 1992. Tritium in groundwater as a tool to estimate well vulnerability.
Minnesota Department of Health, Minneapolis, Minn. [contract report
prepared for U. S. Environmental Protection Agency. Office of Ground Water,
Region V, Chicago]. 128 p. *
White, W. B. 1988. Geomorphology and Hydrology of Karst Terrains. Oxford
University Press, New York. 464 p.
5-12
-------
White, W. B. and White, E. L., eds. 1989. Karst Hydrology: Concepts from the
Mammoth Cave Area. Van Nostrand Reinhold, New York. 346 p.
Williams, P. W. 1977. Hydrology of the Waikoropupu Springs: A major tidal karst
resurgence in northwest Nelson (New Zealand). Journal of Hydrology vol 35
p. 73-92.
Williams, P. W. 1983. The role of the subcutaneous zone in karst hydrology.
Journal of Hydrology, vol. 61, p. 45-67.
Wisconsin Geological and Natural History Survey. 1991. Delineation of Wellhead
Protection Areas in Fractured Rocks. U.S. Environmental Protection Agency,
Office of Groundwater and Drinking Water, Washington D.C. EPA 570/9-91-
009. 144 p.
Wise, D. U. 1982. Linesmanship and the practice of linear geo-art. Geological
Society of America, Bulletin, vol. 93, p. 886-888. [Discussion and Reply: ibid.,
vol. 94, p. 1377-1379].
Worthington, S. R. H. 1991. Karst Hydrogeology of the Canadian Rocky
Mountains. Ph.D. Dissertation (Geography), McMaster University, Hamilton,
Ontario. 380 p.
Worthington, S. R. H. 1994. Flow velocities in unconfined Paleozoic carbonate
aquifers [Abs.]. Cave and Karst Science, vol. 21, p. 21-22.
Worthington, S. R. H., Davies, G. J. and Quinlan, J. F. 1992. Geochemistry of
springs in temperate carbonate aquifers: Recharge type explains most of the
variation. Colloque d'Hydrologie en Pays Calcaire et en Milieu Fissure (5th,
Neuchatel, Switzerland), Proceedings, Annales Scientifiques de 1'Universite de
Bescancon, Geologie-Memoires Hors Serie, no. 11, p. 341-347.
Yaun Daoxian. 1981. A brief introduction to China's research in karst. Institute of
Karst Geology., Guilin, Guangxi, China
Zeizel, A. J., Walton, W. C., Sasman, R. T. and Prickett, T. A. 1962. Ground-Water
Resources of DuPage County, Illinois. Illinois Water Survey Coop. Ground-
Water Report 2. 103 p.
5-13
-------
APPENDIX A
GLOSSARY OF TERMS
-------
GLOSSARY OF TERMS
The purpose of this Glossary is to provide a list of terms used in this document and
commonly used by hydrogeologists, as well as some specific terms used in
groundwater contamination assessments, wellhead protection, and carbonate
hydrology. The definitions provided in this glossary are not necessarily endorsed by
the Environmental Protection Agency, nor are they to be viewed as suggested
language for regulatory purposes. Many of these definitions are from the U.S.
Environmental Protection Agency (1987).
Advection: The process by which solutes are transported by the bulk motion of the
flowing groundwater.
Allogenic recharge: Recharge derived from an adjacent or overlying non-
carbonate rock (i.e., concentrated recharge that becomes swallet water).
Analytical model: A model that provides approximate or exact solutions to
simplified mathematical forms of the differential equations for water movement and
solute transport. Analytical models can generally be solved using calculators or
computers.
Anisotropy: The condition of having different properties in different directions.
The condition under which one or more of the hydraulic properties of an aquifer
vary according to the direction of flow.
Anthropogenic: Involving the impact of man on nature; induced or altered by the
presence and activities of man.
Aquifer. A formation, group of formations, or part of a formation that contains
sufficient saturated permeable material to yield sufficient, economical quantities of
water to wells and springs.
Aquifer test: A test to determine hydrologic properties of an aquifer, involving the
withdrawal of measured quantities of water from, or addition of water to, a well and
the measurement of resulting changes in head in the aquifer both during and after
the period of discharge or addition. Same as pump test.
Area of influence: Area surrounding a well or spring within which the water table
or potentiometric surface has been changed due to the pumping or recharge of the
well or spring.
Attenuation: The process of diminishing contaminant concentrations in
groundwater due to filtration, biodegradation, dilution, sorption, volatilization, and
other processes.
A-l
-------
Autogenic recharge: Recharge from precipitation falling directly onto a carbonate
aquifer or through the soil above it. Autogenic recharge can be either concentrated
or dispersed, or both.
Carbon-14 (1*C): A radioisotope of carbon with a half-life of 5,730 years.
Carbon-14 concentration can be used to estimate the age of a groundwater (that is,
the time since a groundwater was recharged at land surface and flowed to the point
of collection).
Concentrated recharge: Localized, rapid recharge from a sinking stream or
through an open joint or bedding plane; all concentrated recharge water flows into
conduits.
Conduit flow: A term used extensively for flow in subsurface tubes. It has been
used in three different ways for flow in conduits, for a type of spring having
distinctive properties and ranges in them, and for a type of aquifer. (The latter two
uses are redundant because all karst aquifers have conduits, and all karst springs
flow from conduits. Conduits are dissolutionally enlarged fissures >5 mm in
diameter). The term should be abandoned because of its ambiguous use. However,
if it is used, it should only refer to flow within conduits. Compare with diffuse flow.
Cone of depression (COD): A depression in the groundwater table or
potentiometric surface that has the shape of an inverted cone and develops around a
well or spring from which water is being withdrawn. Its trace (perimeter) on the
land surface defines the zone of influence of a well or spring. Also called pumping
cone and cone of drawdown.
Contaminant: An undesirable substance not normally present, or an unusually
high concentration of a naturally occurring substance in water, soil, or other
environmental medium.
Contamination: The degradation of natural water quality as a result of man's
activities.
Darcian (laminar) flow: That type of flow in which the fluid particles follow
paths that are smooth, straight, and parallel to the channel. In laminar flow, the
viscosity of the fluid damps out turbulent motion.
Diffuse flow: A term used extensively for non-conduit flow but is ambiguous. It,
too, has been used three different ways for slow flow within very small openings, for
a type of spring having distinctive properties and ranges in them, and for a type of
aquifer. The latter two terms are oxymora because all springs hi karst flow from
conduits (although some might be aggraded at their orifice), and all karst aquifers
have conduits. Additionally, the term diffuse flow means Darcian flow as in a
porous medium; and it may mean flow through narrow fractures, tubes, and pores.
However, all aquifers have diffuse flow, so all uses are superfluous.
A-2
-------
Davies et al. (1992) recommend that the terms conduit flow and diffuse flow be
abandoned. The existence or non-existence of a flow type can never be tested
remotely, but can be observed in cave passages. If almost daily temperature data
from several springs draining the same aquifer over a water year are modeled, it is
impossible to rationally explain the data in terms of conduit flow and diffuse'flow;
only mixed proportions of rapid flow and slow flow explain them (Davies, 1992).'
The terms rapid flow and slow flow are definable in numbers (velocity, as
determined by tracer testing). All karst aquifers are also triple-porosity aquifers
(with groundwater occurring in the bedrock matrix, narrow fractures, and in
conduits) with both rapid and slow flow components. Also, from more than
2,225 tracer tests in conduits in 25 countries, it is obvious that there is always rapid
flow at >0.001 m/s in any unconfined carbonate aquifer (Worthington, 1994). There
could be confined carbonate aquifers that have deep flow in fissures <5 mm in
diameter and at Darcian velocities this would be similar to flow in granular aquifers
but, only if the carbonate aquifer was behaving as an equivalent porous medium
model, which, if flow was in fractures, would require the representative elementary
volume (REV) to be at the appropriate scale (Domenico and Schwartz, 1989, p. 84),
If it is assumed that 5 mm is a minimum diameter for conduits, and it is known
that 0.001 m/s is a minimum velocity in conduits, then Darcian flow can only occur
below these values. If they are exceeded, the flow would not be laminar (Huitt,
1956; Ford and Ewers, 1978; Quinlan et al., 1993).
Dispersed recharge: Non-point, relatively slow recharge through a soil or
overlying non-carbonate aquifer.
Dispersion: The spreading and mixing of chemical constituents in groundwater
caused by diffusion and mixing due to microscopic variations in velocities within
and between pores.
Doline: See also sinkhole.
Drainage basin: A portion of a karst aquifer that is a surface and subsurface
catchment for a spring or network of springs.
Drawdown: The vertical distance groundwater elevation is lowered, or the amount
head is reduced, due to the removal of groundwater. Also, the decline in
potentiometric surface caused by the withdrawal of water from a hydrogeologic unit.
The distance between the static water level and the surface of the cone of
depression. A lowering of the water table of an unconfined aquifer or the
potentiometric surface of a confined aquifer caused by pumping of groundwater
from wells.
Epikarst: The uppermost part, commonly about 30 feet (10 meters) thick, of any
karst aquifer; consists of a zone of highly corroded bedrock and has a distinctive
hydrology.
A-3
-------
Fissure: A fracture or crack in a rock mass along which there is a distinct
separation.
Flow line: The general path that a particle of water follows under laminar flow
conditions. Line indicating the direction followed by groundwater toward points of
discharge. Flow lines generally are considered perpendicular to equipotential lines.
Flow path: The path a water molecule or solute follows in the subsurface.
Fracture: A general term for any break in a rock, which includes cracks, joints,
and faults.
Full-flow spring: A spring that is the sole discharge point of an entire drainage
basin or local aquifer.
Groundwater barrier. Rock or artificial material with a relatively low
permeability that occurs (or is placed) below ground surface, where it impedes the
movement of groundwater and thus may cause a pronounced difference in the heads
on opposite sides of the barrier.
Groundwater basin: General term used to define a groundwater flow system that
has defined boundaries and may include more than one aquifer. The basin includes
both the surface area and the permeable materials beneath it. A groundwater basin
could be separated from adjacent basin by geologic boundaries or by hydrologic
divides and boundaries.
Groundwater divide: Ridge in the water table, or potentiometric surface, from
which groundwater moves away at right angle in both directions. Line of highest
hydraulic head in the water table or potentiometric surface.
Groundwater mound: Elevated area in a water table or other potentiometric
surface, created by groundwater recharge.
Head, total: Height of the column of water at a given point in a groundwater
system above a datum plane such as mean sea level. The sum of the elevation head
(distance of a point above datum), the pressure head (the height of a column of
liquid that can be supported by static pressure at the point), and the velocity head
(the height to which the liquid can be raised by its kinetic energy).
Heterogeneity: Characteristic of a medium in which material properties vary from
point to point.
Highly confined aquifer. A confined aquifer that receives only minor leakage
through overlying confining strata.
Homogeneity: Characteristic of a medium in which material properties are
identical throughout.
A-4
-------
Hydraulic conductivity (K): A coefficient of proportionality describing the rate at
which water can move through a permeable medium.
Hydraulic gradient (i): Slope of a water table or potentiometric surface. More
specifically, change in head per unit of distance in a given direction, generally the
direction of the maximum rate of decrease in head. The rate of change in total head
per unit of distance of flow in a given direction. The change in total head with a
change in distance in a given direction. The direction is that which yields a
maximum rate of decrease in head. The difference in hydraulic heads (hi - h£),
divided by the distance (L) along the flowpath.
Hydrogeologic unit: Any soil or rock unit or zone that because of its hydraulic
properties has a distinct influence on the storage or movement of groundwater.
Impermeable: Characteristic of geologic materials that limit their ability to
transmit significant quantities of water under the head differences normally found
in the subsurface environment.
Interference: The result of two or more pumping wells, the drawdown cones of
which intercept. At a given location, the total well interference is the sum of the
drawdowns due to each individual well. The condition occurring when the area of
influence of a water well comes into contact with or overlaps that of a neighboring
well, as when two wells are pumping from the same aquifer or are located near each
other.
Isochrone: Plotted line graphically connecting all points having the same time of
travel for water or contaminants to move through the saturated zone and reach a
well.
Isotropy. The condition in which the properties of interest (generally hydraulic
properties of the aquifer) are the same in all directions.
Karst aquifer: Any unconfined carbonate aquifer. All karst aquifers are triple-
porosity aquifers and have both rapid and slow flow in dissolutionally-enlarged
macrofissures (conduits, fractures, and tubes) >5 mm wide, in small openings
<5 mm wide, and intergranularly.
Leakage: The vertical movement of groundwater; commonly used in the context of
vertical groundwater flow through confining strata.
Macropore: A pore in soil of a large enough size so that water is not held in it by
capillary attraction that can sustain turbulent flow.
Maximum contaminant level (MCL): Maximum permissible level of a
contaminant in water that is delivered to the users of a public water system.
A-5
-------
Maximum contaminant level is defined more explicitly in Safe Drinking Water
Act (SDWA) regulations (40 CFR Section 141.2).
Meteoric Water. Groundwater which originates in the atmosphere and reaches the
zone of saturation by infiltration and percolation.
Observation well: A well drilled in a selected location for the purpose of observing
parameters such as water levels or water chemistry changes.
Overflow spring: A spring that drains an aquifer at high stage or flood stage only.
These springs always have underflow springs that form the remainder of the flow
distributary that discharges the aquifer.
Paleozoic: The era of geologic time from the end of the Precambrian (600 million
years before present) until the beginning of the Mesozoic era (225 million years
before present).
Percolation: Slow downward flow in narrow fissures, tubes and pores within soil
and bedrock.
Piezometric surface: See potentiometric surface.
Pleistocene: An epoch of geologic time of the Quaternary period, following the
Tertiary and before the Holocene. Also known as the Ice Age.
Point source: Any discernible, confined, or discrete conveyance from which
pollutants are or may be discharged, including, but not limited to, pipes, ditches,
channels, tunnels, conduits, wells, containers, rolling stock, concentrated animal
feeding operations, or vessels.
Ponor. The point of a sinking (surface) stream. It can be a large or relatively small
opening through soil or bedrock. It can also occur in overlying non-carbonate cover
beds in covered karst aquifers. Also called a stream sinks or swallow holes.
Porosity: The ratio of the volume of void spaces in a rock or sediment to the total
volume of the rock or sediment.
Potable water: Suitable for human consumption as drinking water.
Potentiometric surface: A surface that represents the level to which water will
rise in tightly cased wells. If the head varies significantly with depth in the aquifer,
then there may be more than one potentiometric surface. The water table is a
particular potentiometric surface for an unconfined aquifer.
Radial flow: The flow of water in all directions away from a potentiometric high or
the flow of water in an isotropic aquifer toward a well.
A-6
-------
Rapid flow: Non-Darcian, mostly-turbulent flow in vadose conduits, and flood flow
in submerged (phreatic) conduits, at a velocity of >0.001 m/s. (Conduits are
dissolutionally enlarged macrofissures >5 mm in wide).
Recharge area: Area in which, water reaches the groundwater flow system by
surface infiltration or direct recharge. An area in which there is a downward
component of hydraulic head in the aquifer.
Residuum: The material eventually resulting from the decomposition of rocks in
place after all but the least soluble constituents have been removed.
Semiconfined aquifer. A confined aquifer whose confining bed may vertically
conduct significant quantities of water.
Sinkhole: Any enclosed karst depression. They may vary in diameter from a few
feet (meters) to over one mile (kilometer), have sides that range from gentle slopes
to vertical and vary from a few feet (meters) to many hundreds or even a thousand
feet (kilometer) deep. They may occur as isolated individuals or in densely packed
groups (Ford and Williams, 1989). Also referred to as a doline.
Slow flow: Mostly-laminar, mostly-Darcian flow in small fissures, tubes, and pores
(<5 mm in diameter) or flow in submerged conduits at base flow; both at <0.001 m/s.
Stagnation point: A place in a groundwater flow field at which the groundwater
is not moving.
Swallet: The opening into which a stream sinks. See Ponor.
Time of travel (TOT): The time required for a contaminant or tracer to move in
the saturated zone or flow system from a specific point to a well or spring.
Tritium (3H): The radioactive isotope of hydrogen with a half-life of 12.3 years.
The presence or absence of tritium in groundwater provides a method for estimating
when the water was recharged at land surface.
Turbulent flow: That type of flow in which the fluid particles move along very
irregular paths. Momentum can be exchanged between one portion of the fluid and
another.
Unconfined aquifer: An aquifer over which there is no confining strata.
Underflow spring: A perennial base-level spring that preferentially drains a basin
or aquifer.
Well field: An area containing two or more wells.
A-7
-------
Wellhead or springhead: The physical feature, facility, or device at the land
surface from or through which groundwater flows or is pumped from subsurface,
water-bearing formations.
Wellhead protection area (WHPA)/springhead protection area (SHPA): The
surface and subsurface area surrounding a water well, well field, or spring
supplying a public water system, through which contaminants are reasonably likely
to move toward and reach such water well, well field, or spring.
Zone of contribution (ZOC): The area surrounding a pumping well or spring that
encompasses all areas and features that supply groundwater recharge to the well or
spring.
Zone of influence (ZOI): The area surrounding a pumping well or spring within
which the water table or potentiometric surfaces have been changed due to
groundwater withdrawal.
Zone of transport (ZOT): The area surrounding a pumping well or spring,
bounded by an isochrone and/or isoconcentration contour, through which a
contaminant may travel and reach the well.
A-8
-------
APPENDIX B
PRACTICAL TRACING OF GROUNDWATER WITH
EMPHASIS ON KARST TERRANES
E. C. Alexander and J.F. Quinlan
1992
Geological Society of America
-------
PRACTICAL TRACING OF GROUNDWATER,
WITH EMPHASIS ON KARST TERRANES
2nd edition*
by
E. CALVIN ALEXANDER, JR.
Department of Geology and Geophysics
University of Minnesota
Minneapolis, Minnesota 55455
(612)624-3517
and
JAMES F. QUINLAN
Quinlan & Associates, Inc.
Box 110539
Nashville, Tennessee 37222
(615)833-4324
(The 1 st edition with same title was by James F. Quinlan
and E. Calvin Alexander, Jr., 1990)
A Short Course Manual presented on the occasion of the
Annual Meeting of the Geological Society of America
October 24, 1992
Cincinnati, Ohio
Geological Society of America
3300 Penrose Place
Boulder, Colorado 80301
-------
Table of Contents
Volume I
Introduction 1
Purposes of tracing 1
Desirable properties of tracers 2
Types of tracers 2
Non-toxicity of tracers 3
Types of tracer tests - 4
Practical tracing in karst terranes 5
Tracing with dyes 5
Dye nomenclature 5
Types of flourescent dyes used for groundwater tracing 8
Dye recovery techniques 12
Placement of dye detectors 19
Estimation of dye quantities needed 19
Injection of dye 21
Sampling frequency 21
Recording of test data 22
Simultaneous use of several dyes 22
Sources of dyes and related materials 22
Standards 22
StabmtyofRhodarnineWT 25
Recent advances in tracer test interpretation 25
Principles which maximize the cost-efficiency, success,
and utility of tracing tests 28
Tracing with environmental isotopes 32
14C and tritium 32
518Oand6D 33
Tracing with ions 33
Cations 33
Anions 33
Unintended tracer tests 33
Costs of tracer tests 34
Volunteer/amateur 34
Professional 34
Bureaucratic 34
Interpretation of tracer tests 34
Positive results 35
Negative results 35
References cited 36
Appendices
A. Ground-water monitoring in karst terranes
B. Groundwater remediation may not be possible....
C Discussion of GROUND WATER TRACERS by Davis et al. (with response and reply)
-i-
-------
Volume n
D. Groundwater Basins in the Mammoth Cave Region, Ky. (map, in envelope)
E. Comparison of tracer mobilities under laboratory and field conditions
F. Recommended admiiu^trative/regulatoiy definition of karst aquifier, principles for
classification of carbonate aquifiers, practical evaluation of vulnerability of karst
aquifiers, and determination of optimum sampling frequency at springs
G. The October 1989 dual dye trace through the Mystery Cave System
H. Water tracing with fluorescent dyes, Part 1: field tests in granular aquifers, soils, and
other sediments, an annotated bibliography
I. Ethical aspects of tracing: a guide for consultants, regulators, and students
J. Use of dyes for tracing ground water: aspects of regulation
-11-
-------
INTRODUCTION
This manual is intended to help you learn about groundwater tracing, chiefly with dyes,
and in karst terranes. You can use it to achieve wealth, influence, and professional adulation. If
you adeptly apply its content you will be acclaimed in world capitals and lionized at professional
meetings and the office. Women will throw themselves at your feet if you are a man. Men will
throw themselves at your feet if you're a woman. But let's worry about tracing now and sex later.
More specifically, this manual is a revised, updated, and expanded and interim preprint
version of part of a chapter on qualitative tracing in The Joy of Dyeing (Aley, et al., in prep.) as a
practical, user-friendly, how-to manual on tracing in karst terranes. It is a successor to the out-of-
date Water Tracer's Cookbook (Aley and Fletcher, 1976). Earlier drafts of this chapter have been
published as part of the multi-volume manual used in a karst hydrogeology and grpundwater
monitoring course taught annually for the National Water Well Association. Appendix A is the
draft EPA protocol for monitoring in karst terranes.
We estimate that approximately 1500 professionally-run groundwater traces have been
made so far in the U.S. We believe that more than 90% of them were done with dyes, in karst
terranes, and that dyes are generally the cheapest, most practical tracers. Therefore, the emphasis
here is on dyes rather than other tracers.
A comprehensive general review of groundwater tracers was made by Davis et al. (1985;
refer to the discussion of it, included herein as Appendix C), so no attempt will be made to review
all of them now in use. We should add, however, that there are serious problems with the
discussion of dyes by Davis et al. Its advice and conclusions concerning dye-tracing should be
used with great caution.
We are not saying in these pages that, "This is the way dye-tracing must be done." Rather,
we are summarizing what we (and others) have found to constantly give reliable results.
The definitive reference on the properties of dyes used for tracing groundwater was written
by Smart and Laidlaw (1977). We won't try to summarize it, but we can not recommend it too
highly.
Excellent introductions to karst, the Disneyland of hydrology, are given in the recently
published texts by Ford and Williams (1989) and White (1988). Ford and Williams include a good
introduction to tracing (p. 219-241). Appendix F is a discussion of a regulatory definition of karst
aquifiers.
PURPOSES OF TRACING
Water tracing has been used for over a century and has been an integral part of karst
hydrogeology for almost as long. Tracing has been used for many purposes: to determine the
resurgence of surface and subsurface flows, to delineate subsurface basins, to estimate
groundwater flow velocities, to identify the sources of pollution, etc. In general, however, the
purpose is to determine and quantify the motion of water in the subsurface. The numerous
concerns about the sub-surface water's motion can be condensed into the following three basic
questions.
1. Where does the water go?
2. How long does it take to get there?
3. What happens along the way?
Water tracing is the direct, often quickest, and (in many cases) the only technique to answer these
questions in a karst terrane. There are other hydrogeologic techniques that can (sometimes) be
used to address these questions, but rarely do these other techniques work as well as tracing.
The tracing techniques and approaches that an investigator might require and/or utilize vary
greatly in levels of sophistication. A question of "Is the septic system of this house connected to
-1-
-------
the nearest sewer main?" can be adequately answered with a few pennies worth of dye and few
minutes of someone's time. Similarly, questions about the internal connections in caves can often
be answered with about the same level of resources and time. Such traces are valid and important
In contrast, questions about the regional-scale dispersal of pollutants from a major Superfund site
in a densely populated karst terrane require a considerably greater investment of time, resources,
and effort. The use of simple visual detection of dyes is rarely acceptable in major projects,
although it can still be a very effective demonstration of a connection in some situations.
Conversely, Superfund protocols are never going to be used or needed in the majority of tracing
tests.
DESIRABLE PROPERTIES OF TRACERS
The following is a modification of the discussion in Ford and Williams (1990, p. 229). If
you are going to go to the trouble to introduce a tracer into a groundwater flow system, you might
as well pick one that makes it easier, rather than harder, to accomplish your goal. Ideally, the tracer
you choose should meet the following criteria.
1. Non-toxic to you, to the karst ecosystem and to potential consumers of the labelled
water.
2. Either not naturally present in your system or there at some very low, constant level.
3. In the case of chemical substances, soluble in water with the resulting solution having
approximately the same density as water.
4. Neutral in buoyancy and, in the case of particulate tracers, sufficiently fine to avoid
significant losses by natural filtration.
5. Unambiguously detectable in very small concentrations.
6. Resistant to adsorptive loss and/or to chemical, physical or biologic degradation.
7. Susceptible to quantitative analysis; the quicker and more economical the better.
8. Easy to administer, inexpensive, and readily obtainable.
Non-toxicity is the most important characteristic. Few tracers satisfy all of these criteria, but
several of the fluorescent dyes meet most in many situations.
TYPES OF TRACERS
Two broad classes of tracers have been used at various times in karst work, labels and
pulses. Both can be usefully subdivided into natural and artificial tracers. The point of labels is to
be able to tell "your" water from the rest of the water. The point of pulse tracing is to send an
identifiable signal through the groundwater system. The following is a partial outline of tracers
that have been used.
-2-
-------
I. Labels:
A. Natural
1. Flora & fauna (chiefly but not exclusively microorganisms)
2. Ions in solution
3. Environmental isotopes
4. Temperature
5. Specific conductance
B. Artificial
1. Dyes and dye-intermediates
2. Radiometrically detected substances
3. Salts and other inorganic compounds
4. Spores
5. Ruorocarbpns
6. A wide variety of organic compounds
7. Effluents and spilled substances
8. Biological entities; bacteria, viruses, yeasts
9. Organic particles; microspheres
10. Inorganic particles; including sediment
11. Temperature
12. Exotica (eels, ducks, marked fish, etc.)
n. Pulses significantly above background or base flow levels:
A. Natural
1. Discharge (stage or flow)
2. Temperature
B. Artificial
1. Discharge
2. Temperature
The different types of tracers have differing strengths and weaknesses and yield differing
types of information about the hydrogeologic system. As the level of sophistication of an
investigation increases, comparison of the results obtained with different tracers can often yield
additional information about the system's properties.
NON-TOXICTTY OF TRACERS
Artificial tracers are foreign substances deliberately introduced into hydrogeologic systems.
Where there is any conceivable possibility for those substances to reach anyone's water supply,
society should and does demand a high standard of safety. In our experience, dye-trace
professionals are concerned, dedicated individuals who do everything in their power to protect the
public. The dye trace community has an excellent record of research designed to evaluate the
toxicity of dye tracing agents. The current definitive reviews of the subject are Smart (1984) and
Field et al (1992) and references listed therein. Smart's (1984) review concluded that Tinopal
CBS-X (or almost any other optical brightener), fluorescein, and Rhodamine WT are the three
safest dyes for tracer work and recommend that the use of Rhodamine B be abandoned. Field et
al. concluded that none of the dyes currently used or evaluated by them pose a threat to health or
the environment if the maximum 24-hour concentration does not exceed 1 part per million (ppm).
-3-
-------
Another recent evaluation of Rhodamine WTs toxicity concludes that the use of Rhodamine WT in
dye tracing "presents a de minimus health risk (<1 x 10"6 ^probability])" to potable water supplies.
Most traces that involve public water supplies are prompted by a desire to protect those
water supplies from existing or potential pollution problems. The traces are not designed to be a
source of pollution. Nevertheless, the ancient human tradition of beheading the bearer of bad news
is all too often the reaction of public officials who have responsibilities over pollution cases. A
disturbing double standard sometimes seems to be in effect Pollutants in a system are often
"presumed innocent until proven guilty beyond the shadow of a reasonable doubt" while tracers,
used to help protect society from the pollutants, are presumed "guilty until proven innocent beyond
the shadow of a (un)reasonable doubt". A public official who wishes to prevent a dye trace can
demand impossible levels of "proof of safety" in any given situation.
A litany of horror stories will accomplish nothing. Suffice to say, the horror stories are
real and all too common. If you are planning a dye trace that can, in any way, impact on public
water supplies, expect trouble. In many situations the best advice is do not ask for permission to
use dyes. If the approval of some public agency is required, and in most cases it is not, do your
homework, pick you tracer, propose your protocol, and try to make it incumbent on the public
official to prove that the tracer is dangerous. Tracing is well-established, necessary, and an
accepted professional practice in karst situations. If you can shift the responsibility for risking
public safety by not doing the trace to the public official, your chances for approval go up
tremendously. Do not make it easy for the public official to stop the trace.
The psychology here is very simple. If you ask a public official to approve your trace, you
are asking that individual to assume some responsibility for any conceivable adverse outcome. The
individuals in question usually do not have technical backgrounds in dye tracing and so are very
reluctant to assume that responsibility. That is just basic human nature. Therefore, do not ask the
public official to assume the responsibility. Make a reasonable and professional proposal. Make it
easy for the official to approve your trace and difficult for them to deny it. (You will still run into
problems — but you will have tried to avoid them. It also helps if you are not a graduate of the
Quinlan & Alexander School of Diplomacy.)
The bottom line here is, to our knowledge, all of the available scientific evidence indicates
that the routine use of fluorescent dyes in water tracing represents no significant hazard to public
water supplies in karst terranes. To the contrary, the use of such a technique is the only practical
way to protect many of the water supplies from pollution. The colored dyes, such as fluorescein
and Rhodamine WT, have an additional, built-in safety feature. The dyes are readily visible at the
hundreds of parts per billion level. Such concentrations are orders-of-magnitude below any
conceivable health-risk level. If anyone ever succeeds in drinking dangerous levels of these dyes,
it will be from glasses of neon green or red water.
Appendix J discusses a proposed regulatory approach that might help with this problem.
TYPES OF TRACER TESTS
Tracer tests can be classified in several ways, according to:
I. Degree of quantification
A. Qualitative
B. Semi-quantitative
C. Quantitative
n. Degree of alterations of hydraulic gradient
A. Natural gradient
B. Forced gradient
-4-
-------
1. Injection (input raises potentiometric surface)
2. Discharge (pumping lowers potentiometric surface)
IEL Type of injection site
A. Natural
1. Sinkhole or swallet
2. Cave stream
B. Artificial
1. From a well or other man-made contrivance
IV. Type of recovery site
A. Natural discharge site
1. Spring
2. Cave stream
B. Artificial discharge site
1. Monitoring well
2. One or more domestic wells
Discussion of classification of tracer test according to their degree of quantification is
included on pages 32-34 of Appendix A.
An introduction to natural-gradient and to forced-gradient tracer tests, using single wells or
multiple wells, is given by Dominico and Schwartz (1990, p. 678-681). Natural gradient tests are
usually complex, extremely expensive, and may involve analysis of tens to several hundred
monitoring points. Single-well and multiple-well tests are also complex and expensive. In the
opinion of Bedmar (1990), only a few hydrologists are familiar with these techniques and their
data are of doubtful validity because of the local scale at which the tests are conducted. An analogy
can be made with pumping tests. Few experienced hydrologists would claim that a single pumping
test is representative of the entire aquifer.
As discussed by Quinlan and Ray (1992) in Appendix B of this manual, all dye-tests suffer
from the possibility of being nonrepresentative of an aquifer. We suggest, however, that dye-tests
from a sinking stream and also from one or more randomly located wells in the area between
conduits will be more representative of a karst aquifer than will single-well or multiple-wells tests.
Tests from a sinking stream and from inter-conduit areas involve orders of magnitude more of an
aquifer than tests between just one well or a few wells nearby.
The effects of testing scale on determination of hydraulic conductivity of an aquifier are
shown in Figure 1, which demonstrates that core, packer, slug, and pumping tests do not yield
hydraulic conductivity values representative of the carbonate aquifier.
PRACTICAL TRACING IN KARST TERRANES
TRACING WITH DYES
Dye Nomenclature
A dye is a substance that can be applied in solution to a substrate, or can be added to a
liquid, thus giving the substrate or the liquid a colored appearance. Fluorescent dyes have an
advantage over non-fluorescent dyes as tracers. Fluorescent dyes can be easily detected in
concentrations which arc one to three orders of magnitude less than those at which non-fluorescent
dyes can be measured spectrometrically.
-5-
-------
CO
E
o
Q
O
O
o
DC
Q
X
10°
ID'2
io-4
10'6
io-8
-10
10
io-12
• Predominant Range
in the Same Aquifer
I Range Reported
in the Literature
P5] 1400-1- Dye Tests
10 in Conduits
A Core (Lab) Tests
B Double Packer Tests
C Slug Tests
D Pumping Tests
E Dye Tests
10°
ID'2
io-4
-6
10
10'8
10
-10
10'12
0.01 0.1 1 10 100 1000 10,000 100,000
SCALE OF MEASUREMENT,
LENGTH OF MEASURED AREA (m)
8
e/
CO
UJ
ui
Q
O
3
UJ
Rgure 1. Range in hydraulic conductivity in karst aquifers—in meters per second (m/s) as a
function of scale of measurement (length of measurement area, m). Test methods A through D
measure hydraulic conductivities Qeft axis) while test method E measures velocity of ground-water
flow in conduits (right axis). In conduits, ground-water velocity and hydraulic conductivity are
assumed to be approximately equivalent. The data represented by heavy bars are from a Jurassic
karst aquifer in the Swabian Alb of Germany, as described by Sauter (1992). the hatchured box
represents velocity data from 1405 dye traces from sinking streams to springs (i.e., in conduits)
from 25 countries. The plot is a comparison of hydraulic conductivity between the conduit portion
and the fracture-matrix portion of any karst aquifer because data from a particular measurement
scale tend to plot in a relatively consistent field. Velocities in conduits are much greater than
velocities in the fracture-matrix portion of a karst aquifer. The presence of conduits in a karst
aquifer requires a dual-porosity approach, rather than a porous-medium approximation, because
the hydraulic conductivity of the conduits and, therefore, the aquifer, would be grossly
underestimated with the porous-medium approach, [modified after Quinlan et al., 1992]
-6-
-------
Smart and Laidlaw (1977) logically classified water-tracing dyes by color of their
fluorescence: blue, green and orange. To these three groups one must add a fourth, yellow. But
we prefer to classify dyes according to the detector (bug) generally used to recover them in
qualitative tests: activated charcoal and cotton. Before discussing the dyes, however, it is
appropriate to discuss dye nomenclature.
The standard industrial reference to dyes is the Colour Index (SDC & ATTC, 1971-1982),
an incredible 7-volume compendium of almost 6500 pages which describes an estimated 38,000
dyes and pigments. (Colour Index is abbreviated CI.) Most dyes listed in it are classified both
according to the dyeing method in which they are used to color various types of textiles, leather,
paper, or other goods (the CI Generic Name) and according to their chemical structure (the CI
Constitution Number). For example, fluorescein is CI Acid Yellow 73 and CI 45350; it is sold
under more than 30 different commercial names such as Soap Yellow F. There are 18 CI Generic
Name categories and 29 CI Constitution Number categories for dyes, but they will not be
discussed here — other than to stress that the Generic Name classification is based on industrial use
and that the four major generic types of dyes used in water tracing are: Acid, Basic, Direct, and
Fluorescent Brightening Agent [References to discussion of dye nomenclature are given by
Quinlan (1986).] Dyes are also classified according to their use in foods, drugs, and cosmetics.
For example, water-soluble fluorescein which has certified purity is designated by the U.S. Food
and Drug Administration as D&C Yellow 8. (D&C = Drug and Cosmetic).
There are numerous commercial names for most dyes. The rules and logic of these
commercial names are as chaotic as those for prescription drugs and as logical as those for
conjugating irregular French verbs. In order to avoid confusion in publications concerning dye
properties and trace results, one should always include specific cation of dyes used by: the CI
Generic Name (rather than CI Constitution Number,) Manufacturer, and the manufacturer's
commercial name.
The first part of the commercial name of a dye should not be confused with the dye itself.
Tinopal and Diphenyl, for example, should always be capitalized; they are trade names belonging
to the Ciba-Geigy Corporation and used for whole series of chemically unrelated dyes made by that
particular company. To use just the first pan of a commercial name for a dye, capitalized or
uncapitalized, is ambiguous and erroneous. There are, for example, seven chemically different
Tinopals and 20 different Phorwites currently sold in the U.S. by Ciba-Geigy and Mobay as
optical brighteners.
The need for careful identification of dyes used for water tracing - particularly if one is
discussing their chemical properties, toxicity, or exitation and emission spectra — and the
importance of knowing something about dye nomenclature and classification, are made obvious
from inspection of the Colour Index itself. For example, five structurally different kinds of
Rhodamine were sold in 1992 in the U.S. under 10 different names by six manufacturers. This
multiplicity of names for what may or may not be the same dye is a result of the fact that, until
recently, the manufacterers of a new dye or an allegedly new dye did not want the customers or the
competition to know the nature of the dye. If the manufacturer knew that his product was the same
as that of another, he did not say so because such identification could increase competition and
force prices lower. Even today, much information about dye structure and composition is
proprietary, is is not released, partly because development costs must be amortized.
When one publishes the results of tracing studies and gives the name of a dye, it is essential
that the full Colour Index Generic Name be given at least once. To refer to , for example, Direct
Yellow without also specifying its number, 96 (as some authors have), keeps the reader ignorant
as to which of more than 180 Direct Yellows was used. Omission of this number in the title or text
is like dropping just one shoe. Only part of the story is known; the rest is left to the imagination.
The above discussion of dye nomenclature is academic and useless if one is going to use
nothing but fluorescein and one or two other dyes. But familiarity with dye nomenclature and
classification is essential if one is going to read and understand the dye literature, make justifiable
decisions about dye toxicity and equivalencies, find alternative sources of supply (particularly
when a supplier has ceased to make a particular dye), and make valid price comparisons.
-7-
-------
Types of Fluorescent Dyes Used For Groundwater Tracing
There are three fluorescent dyes and one group of fluorescent dyes used for most
groundwater tracing in the U.S. Although all of them are sorbed onto activated charcoal, they can
be pragmatically classified into the principal groups according to the method conventionally used to
detect and recover them. They are:
A. Dyes recovered on cotton
1. Optical brighteners (organic compounds added to detergents to "make your
whites whiter")
a. Tinopal 5BM GX (powder)
b. Phorwite BBH Pure (powder)
c. Phorwite AR Solution (13% BBH Pure, in water)
2. Direct Yellow 96 (powder)
B. Dyes recovered on activated charcoal
1. Fluorescein (powder)
2. Rhodamine WT (sold as a 20% solution, in water)
3. Eosin (powder)
Sources of supply, costs, concentration, and Colour Index information on most of these
dyes are given in Tables 1 and 2. Toxicity data for all of them but Tinopal 5BM GX are given by
Smart (1984). Toxicity data for Tinopal 5BM GX are given by Lyman et al. (1975) and Ganz et al.
(1975); in the latter paper, it is identified as DASC-4. Toxicity data for 12 dyes and a dye-
intermediates are reviewed by Field, Wilhelm, and Quinlan (in review).
Similar data for other dyes which can be safely used, and a discussion of dyeing kinetics,
will be given in The Joy of Dyeing, but the major emphasis in it will be on those listed above.
Not all optical brighteners are suitable for water-tracing. Some work best in hot water and
are inefficient in cold (53°F), hard water. Many work best on a specific type of fabric. There are
brighteners and there are brighteners.
There are two dyes called fluorescein. Both of them are CI Acid Yellow 73 but only one of
them, sodium fluorescein, is water-soluble and used for tracing. In the U.S. and England this dye
is generally (and erroneously) called fluorescein rather than its European name, Uranine. We will
follow the U.S. custom. But not all sodium fluorescein is the same. The yield of the dye-
manufacturing process varies from lot to lot In order to achieve a uniform product and, in some
cases, in order to enhance the ease of dyeing a product, it is conventional to add diluents (cutting
agents) to technical-grade dyes. This is standardization, not adulteration. For fluorescein the
diluents commonly used are any one of the following: sodium chloride, sodium sulphate, and
Dextrin (a starch product made from com), in concentrations that range from 10% to 60%; 25%
and 50% are the most common. This means that the fluorescein bought from a company today
may have a different diluent composition and a different strength (diluent percentage) from what
was bought from the same company last year. It is likely to be different from the same dye sold by
another company. It also mans that: 1) calculations of dye-concentration (but not dye-recovery) in
quantitative tests will be higher or lower than the actual value if no allowance is made of the purity
of the dye, especially if a calibration curve is not made for each dye lot), and 2) comparison of dye
costs must consider diluent concentration as well as cost per pound. A dye which costs more per
pound could be cheaper (in terms of active ingredients) than one which costs less. A comparison
of fluorescein prices and concentrations listed in Figure 1 shows that one company offers 50%
more dye per pound at about half the cost per pound charged by another. There are fluoresceins
and there are fluoresceins ~ as shown in Table 3 and analyzed by one of us (E.C.A).
Optical brighteners, fluorescein, and Eosin are very susceptible to photochemical decay,
especially when in low concentrations; Rhodamine WT is less so, but this property is a problem
only if bugs are placed in a stream exposed to sunlight. Direct Yellow 96 does not undergo any
significant photochemical decay; it is ideal if you are tracing a stream that repeatedly sinks,
resurges, and flows along the surface, but its detectability in water is about 0.002 that of
-8-
-------
TABLE 1. Comparison of dye costs.
I. Dyes detected on charcoal:
A. Fluorescein, (7 Acid Yellow 73, a green dye, supplied as a powder.
1. Chemcentral Dyes & Pigments: ($100 minimum order, )
1 Ib can - $55.00/lb; 5 Ib can - $20.65/lb; 10 Ib can - $14.65/lb;
251bdrum-$10.15/lb; 50 Ib drum - $9.65; 100 Ib drum - $9.15/lb
250 Ib drum - $8.65/lb.
2. Pylam:
1 Ib can - $40.00/lb; 5 Ib can - $30.00/lb; 10 Ib can - $29.00/lb;
251bdrum-$23.63/lb; 50 Ib drum-$23.63; 100 Ib drum-$21.63/lb.
A. Rhodamine WT, Cl Acid Red 388, an orange dye, supplied as a 20 % (by weight)
solution, about 10 Ibs per gallon.
1. Crompton & Knowles: Sole U.S. manufacturer.
25 Ib drum - $18.00/Ib; 50 Ib drum - $!8.00/lb; 100 Ib drum - $17.00;
250 Ib drum-$16.00/lb.
2. Chemcentral Dyes & Pigments: ($100 minimum order,)
5 Ib can - $29.80/lb; 10 Ib can - $23.80/lb; 25 Ib drum - $19.30/lb;
50 Ib drum-$18.80; 100 Ib drum - $18.30/lb; 300 Ib drum-$17.80/lb.
3. Pylam:
1 Ib can - $50.00/lb; 5 & 10 Ib cans - $40.00/lb; 25 to 100 Ib cans - $35.00/lb.
II. Dyes detected on cotton:
A. Solophenyl formerly Diphenyl Brilliant Flavine 7GFF, Direct Yellow 96, a yellow dye,
supplied as a powder.
1. Melody Chemicals: Manufactured by Giba-Geigy Corp.
1 to 24 Ibs - $28.87/lb; 25 to 99 Ibs - $26.87/lb; 100 to 219 Ibs - $25.37/lb;
220 Ibs drums-$23.87/lb.
B. Blankophor BBH formerly Phorwite BBH Pure, F.B.A. 28, an optical brightener.
supplied as a powder.
1. Burlington Chemical Co: Manufactured by Mobay Chemical Corp.
25 to 1141bs-$13.10/lb; 115 Ib drums - $10.10/lb.
C. Phorwite AR Solution, F.B.A. 28, an optical brightener, supplied as a liquid.
1. Burlington Chemical Co: Manufactured by Mobay Chemical Corp.
25 to 249 Ibs - $5.10/lb; 250 Ib drums - $2.10/lb.
Cost per pound of tracer dyes currently available from various suppliers, 1992, FOB factory.
All prices subject to change without notice, always check with the supplier. Prices include
repackaging charges, where applicable. Supplier names, addresses and phone numbers listed
in Table 2. Abbreviations: CI = Color Index. F.B.A. = CI Fluorescent Brightening Agent.
-9-
-------
TABLE 2.
Suppliers of dyes, chemical, bug materials, and ultraviolet lamps
Chemeentral Dyes and Pigments
P.O. Box 77073
Detroit, MI 48277-0073
(800) 837-1393
(Ciba-Geigy Corporation)
Small quantities of Ciba-Geigy
dyes are available from:
Melody Chemicals
P.O. Box 524
Charlotte, NC 28225
(800) 334-9481
(800) 228-1589
Crompton & Knowles
Industrial Products Division
P.O. Box 33157
Charlotte, NC 28233-3157
(800) 323-4383
Fisher Scientific Co.
1600 W. Glenlake]
Itasca, IL 60143
(800) 766-7000
(Mobay Chemical Corporation)
Small quantities Mobay dyes
are available from:
Burlington Chemical Co.
P.O. Box 111
Burlington, NC 27216
(919)584-0111
(800) 672-5880
Pylam Products Co., Inc.
1001 Stewart Ave
Garden City, NY 11530
(800) 645-6096
Ultra-Violet Products, Inc.
P.O. Box 1501
San Gabriel, CA 91778
(800) 452-6788
-10-
-------
Table 3.
Chemical Analysis of Commercial Samples of Disodium Fluorescein
(C.I. Acid Yellow 73, Fluorescein, Uranine)
Sample I.
II.
III.
IV.
Chemical Analyses (in wt%)
Na
K
SO4
a
15.89
<0.2
13.35
15.04
0.2
12.06
0.165
21.8
024
35.99
0.229
22.9
0.29
36.93
0.350
15.03
0.25
11.85
0.498
Calculated Composition (in Wt%)
75.9 37.4
Disodium
Fluorescein 77.7
Kd
Nad
unknown
Notes:
19.7
2.6
17.8
0.4
5.9
53.2
0.5
8.9
42.3
54.6
0.6
0.1
2.4
75.6
17.5
0.5
0.5
5.9
Samples: I. = Fisher purified grade uranine, purchased in 1989; H = Chemcentral Dyes
and Pigments technical grade uranine, purchased in 1990; ffl. = Pylam Fluorescent Yellow
purchased in 1979; IV. = Pylam Fluorescent Yellow purchased in 1990: V. = Abby Color and
Chemical, technical grade uranine, batch #1320, purchased in 1979.
Chemical Analyses: The samples were dissolved in deionized water, diluted and analyzed
for common cations and anions. The cation analyses were by Direct Current Plasma Atomic
Emission Spectrometry (DCP/AES) and anion analyses were by Ion Chromatography (1C). In
addition to the listed cations and anions, the following species were analyzed but were <0.1 wt%
in the samples above: Ca, Mg, Fe, Mn, Al, Ti, Sr, Ba, P, Si, F, Br, an
Calculated Composition: For each sample, the measured Cl was balanced first by the
measured K and then by Na if necessary, the measured SO4 was then balanced by Na. The
remaining Na was then calculated as disodium fluorescein using the chemical formula for uranine
as Na2C2oHioOs (F.W. = 376.28). [Some sources list the chemical formula for uranine as
Na2C2oHi2Os. The use of this value would not significantly affect the composition calculations].
The "unknown" component is calculated from mass balance and is presumed to be unidentified
organics and/or water.
-11-
-------
fluorescein. (This means that, for a given distance and given flow-conditions, 600 times more of it
may have to be used)
All tracer dyes tend to react with the environment through which they flow. Fluorescein,
Rhodamine WT (to a lesser extent), optical brighteners, and Direct Yellow 96 (the latter two to a
much lesser extent than Rhodamine WT) are all sorbed by clays. Sorption precludes the efficient
use of fluorescein and Rhodamine WT in granular aquifers but it is less important in karst aquifers.
Optical brighteners and Direct Yellow 96 are more reactive with cellulose (wood, leaves, etc.) than
the other two.
Dye Recovery Techniques
Dye is recovered hi water samples (either grab samples or those taken with an automatic
sampler), with detectors known colloquially as bugs, or with collectors attached to pumped wells
and known as a Dye Intensity-Level Device for Observation Wells.
A bug is used for many reasons. It provides continuous sampling (monitoring) of a site for
hours, days, weeks, or months. It yields an integrated sample with a dye-concentration ten to
several hundred times higher than that which was in the water that passed through it or by it.
Important also is the fact that it is cheap relative to the costs of labor or automatic samplers, and it
is dependable. Bugs don't malfunction, sleep, or need time off during the weekend.
The decision whether or not to use grab samples vs. an automatic sampler vs. bugs is
largely one of economics and purpose of a tracer test. For many tests a qualitative answer is all
that is needed; there is no justification for the expense and complexity of a semi-quantitative or
quantitative test Alternatively, it may be expedient to precede either of the latter with a qualitative
test
Water samples give a degree of reliable quantification that this not possible with a bug. The
accuracy of dye quantification with elutant from a bug is significantly less than that with water
samples because of the kinetics of dye-transit, -absorption, and -elution. A Gluteus Maximus
estimate of the reproductibility of dye concentration in elutant is ± 25%. The fluorometric or
spectrofluorometric reproducibility of a given elutant sample, however, is likely to be ± 0.1%.
No attempt to describe the operation of programmable automatic samplers will be made
here. We have had great satisfaction with the flexibility and reliability of ISCO Model 2700. We
recommend it highly but have no data to suggest it is or is not superior to samples made by other
manufacturers. The Model 2700 is no longer made; it has been superseded by the Model 3700.
For many applications, the Model 2900 may be quite satisfactory.
Bugs are suspended in streams and springs, commonly on a hydrodynamically stable stand
known as a gumdrop which is explained in Figure 2.
Cotton Bugs
Johnson & Johnson's Red Cross brand surgical cotton, purchased in the one-pound size
rather than a smaller size, is highly recommended. Several other brands of surgical cotton are
treated with an optical brightener during their manufacture; they are, therefore, unsuitable.
Recently, however, Johnson & Johnson has started to use brightener on some of its cotton. We
have found 2.5-inch diameter cotton pads that are useful as bugs. They are sold as Swiss Beauty
Pads and may be found in many drug stores. You can order them directly from the manufacturer,
U.S. Cotton, Inc., at the address given in Table 2 ($2.99 for 100 pads). Swiss Beauty Pads can
be suspended in a packet made of aluminum, nylon, or fiberglass window screening, like those
made for charcoal bugs and described in the caption for Figure 2.
Other cotton pads, swabs, and balls suitable for use as detectors include the following:
Cotton Puffs (triple size) @ Wal-Mart
Demak'up
Curity Super Soft Puffs
-12-
-------
Figure 2. Gumdrop used to suspend dye-detectors (bugs) above stream beds. Total height is about
12 to 14 inches.
A— Concrete semi-hemisphere, approximately 6 inches in diameter and 2 to 3 inches
high. (Concrete is poured into a hydrodynamically stable plastic cereal bowl lined
with Saran Wrap.)
B— Galvanized steel wire, #9 gauge. Note loops bent into it
C~ Nylon cord, 3/32 inches in diameter, tied to loop in wire and to tree or large rock.
(Tan, dirt, or black color is recommended because it blends with dirt)
D- Vinyl-clad #10 copper electrical wire. It is twisted through the steel loop and snugly
around the piece of cotton, E.
E- Surgical cotton, 4 inches long x 2 inches wide and 1 inch thick. The cotton swings
freely in any current and stays free of sediment that might bury it Alternatively, use a
Swiss Beauty Pad in a packet like those used for charcoal.
Note: A second detector, for Fluorescein, Rhodamine WT, or another dye adsorbed onto the
charcoal, can be hung with a paper clip onto the same loop to which C is attached. This detector
consists of one to two heaping teaspoons of activated coconut charcoal in a packet made of nylon
or fiberglass screening, as described in the text
-13-
-------
Cosmetic Puffs @ K Mart
Cosmetic Puffs (triple size) @ Kroger
Note: some products, such as Cosmetic Squares (sold by Super-X Drug Stores) are unsuitable
because they included numerous fluorescent bits of cotton.
A cylindrical cotton bug suitable for use in small-diameter wells is Tampax. We don't
know about the fluorescence or non-fluorescence of other brands of tampons. Tampax would also
be a good "emergency bug" that could be bought almost anywhere.
When recovering bugs in the field, we prefer to put them into individual pre-marked 4x4-
inch Ziploc PVC bags. Before going into the field, label and date each bag with a marking pen
which has an aromatic solvent-based ink. Water-based inks will wash off. After you put the bug
in the bag, but before you seal it, squeeze the excess water out of the bug and drain it.
After collecting a cotton bug, wash it thoroughly with a high speed jet of water. (We use a
garden hose with a squeeze-operated nozzle). This removes clay, silt, and trash from the cotton
mass. Press it flat and examine with a long-wave UV lamp. We recommend the Ultra-Violet
Products model UVL-21 ($72.00) with its Model CC-10 viewing cabinet ($105.00; grotesquely
overpriced, but extremely convenient and almost essential).
Cotton which has reacted with optical brightener is characteristically blue-white. Cotton
which has reacted with Direct Yellow 96 is canary yellow. Cotton which has reacted with both of
these dyes is a distinctive white-white. Fluorescence intensity is directly (but not linearly)
proportional to the amount of dye which has reacted with the cotton. Bugs which have not reacted
with optical brightener will not fluoresce. Bugs many have a few blue-white flecks that fluoresce;
these flecks do not make a bug positive. The flecks are merely brightened fibers from elsewhere
which got mechanically trapped in the bug as it sat in the stream of water.
Some practioners prefer to use cotton cloth rather than surgical cotton as a detector because
the cloth can be more easily scanned by a fluorometer or spectrofluorometer, thus partially
quantifying the degree of positiveness of a bug. We don't. Field tests by the junior author, using
both types of detectors mounted on the same gumdrop, repeatedly show that the surgical cotton is
significantly brighter and more sensitive to lower concentrations of dye. This is probably because
there are fewer fibers per unit area in the cloth.
Charcoal Bugs and Elurion
A charcoal bug adsorbs dye from water that flows through it. The adsorption is a
cumulative process, and the charcoal may be treated, as discussed below, to release a dye-
concentration that is 10 to 400 times higher than what was or is in the water.
A charcoal bug consists of one to two teaspoons of activated coconut charcoal in a packet
made of nylon or fiberglass screening. (Consistently use the same amount.) The packet is about
2.5 inches square. Do not use copper, brass, or aluminum screening. If you use nylon or
fiberglass screening, two of the seams can be closed with a sewing machine, preferably with non-
cotton thread such as polyester, and the third can be closed with a regular-sized paper clip.
Alternatively, all the seams can be closed with a heat-sealing machine. Do not use staples; they are
susceptible to rusting away. The bug is hung on a gumdrop, as described in Figure 2, with a
giant-sized paper clip. Replace paper clips after two to three weeks; they rust and they fail by
fatigue where bent
The following detector and some or all of the indicated reagents are needed:
Activated coconut charcoal, 6-14 mesh (Fisher #5-685-A; cost: $21.00 per pound). Unless
you are going to use hundreds of bugs per month, do not buy charcoal in containers
larger than 1-pond. Charcoal loses its sorption ability by reaction with air over a
period of weeks to months, depending upon how tightly the can is sealed. Do not
use the charcoal designed for water-treatment processes and aquariums.
-14-
-------
isopropyl alcohol (2-propanol): This is rubbing alcohol, a 70% solution, available in
discount drug stores for about $0.35 per pint. You can also use the 99% solution
available from chemical supply houses at a price about 100 times higher, or you can
use the inexpensive 99% solution which is sold for cleaning drums on xerox
machines, but each has to be diluted with water to 70%. If you don't dilute them,
the KOH will not dissolve properly. Make a saturated solution by adding about six
to seven grams KOH to 100 ml of the 70% alcohol. Use just the lighter liquid at
the top, which is a saturated solution of about 5% KOH.
ethanol
1-propanol
potassium hydroxide
ammonium hydroxide
distilled water (tap water would probably be satisfactory when nothing else is available —
unless is is highly chlorinated. Chlorine destroys dyes.)
A Fisher address is given in Table 2. Bugs should be changed every couple of days to
weekly. In streams with smelly or colored water, the charcoal will be rapidly exhausted; bugs
should be changed daily or more often. Too often could be counterproductive.
If elution of dye from charcoal is not done within 6 to 12 hours after collection, the bug
should be dried as soon as possible after collection. This will minimize bacterial reactions that can
destroy some of the fluorescein or other dye in the charcoal. Alternatively, store the bugs (and
water sample) in a refrigerator until they are analyzed. The stability of Rhodamine WT in elutant
and spiked water samples is discussed in a subsequent section of this manual.
A few words should be said about nomenclature. Eluent is a liquid used to remove the dye
from the charcoal. Elutant is the solution of eluent plus dye. Elution is the process by which
eluent becomes elutant Elutriarion (and its derivative words) refers to a straining or decanting.
Elutriation is a word mistakenly used in some of the tracing literature but is has nothing to do with
elution or the process by which dye is recovered.
Elution of dye is simple. Interpretation of elutants, however ranges from obvious to subtle
and can be difficult when the amount of dye recovered is minimal. One has to develop an "eye" for
doing so.
We recommend the following eluants for recovering dye from activated charcoal:
A. For fluorescein: 5% KOH in 70% isopropyl alcohol. (This has a limited shelf-life, do
not use this eluant if it is more than a few days old).
B. For Rhodamine WT: a 5 : 2: 3 mixture of 1-propanol, concentrated NH^OH, and
distilled water. (This is known as a Smart solution. It has an estimated shelf-life of
several months).
C. For instrumental analysis (with either a filter fluorometer or a scanning
spectrofluorophotometer), when it is possible that two or more dyes may be recovered
on activated charcoal: 5% NH4OH in 70% isopropyl (rubbing) alcohol. (This is
known as an Aley solution).
The choice of which hydroxide to use in which alcohol depends upon what dye is to be
eluted, whether two dyes are to be eluted, the adequacy of available ventilation, and one's tolerance
for ammonia fumes. As discussed by Atkinson and Smart (1981, p. 177), the best and most
efficient elutant for Rhodamine WT is a 5 : 2 : 3 mixture by volume of 1-propanol, concentrated
NH4OH, and distilled water, preferably warmed to about 140°F (Smart, 1972; Smart and Brown,
1973). Most practitioners elute at room temperature and we have assumed that the 5:2:3 "Smart
Solution" is also optimal for elution of fluorescein. But it might not be optimal. Staff of the
Missouri Geological Survey prefer to use ethanol rather than 1-propanol because the fumes are
significantly less when ethanol is used. Their laboratory tests have shown that 5% KOH in ethanol
releases significantly more fluorescein from charcoal than does 5% NHUOH in the same alcohol.
In contrast, the recovery of Rhodamine WT from charcoal is far greater with 5% NrfyOH in
-15-
-------
ethanol than with 5% KOH in the same alcohol. For semi-quantitative tests using activated
charcoal, a spectrofluorophotometer and both dyes, the Survey has standardized on elution with
5% NH4OH in ethanol. This is because the significant decrease in the recovery of fluorescein
when NH4OH is used is much less than the great decrease in the recovery of Rhodamine WT when
KOH is used (James W. Duley, oral communication, January 1986). Nevertheless, it has not yet
been shown whether 1-propanol is more efficient or less efficient than isopropyl alcohol, ethanol,
or any other alcohol for elution of fluorescein from activated charcoal. There is a great need for a
systematic study of fluorescein elution, a study similar to that by Smart (1972) of Rhodamine WT
elution. Such a study should start with laboratory tests and be extended to include bugs which
have been subjected to the vagaries of exposure in actual springs and streams.
It is practical to elute dye in baby-food jars on which an identifying number has been
scribed. Similarly-sized jars which can be sealed are also suitable, but investigate what, if
anything, is in their cap-liner. Test them. Be paranoid.
Best practice for elution, however, is to use 3.5-ounce plastic cups that can be capped and
which are disposable. Use a water-proof marking pen for identification. Cups and lids can be
ordered by the case of 2500 from local dealers in paper products of restaurant supplies. [Solo
#P35A cups or Fabri-Kal #PC250 cups, either of which can be used with Fabri-Kal #L250PC lids
(do npl get the corresponding Solo lid.) are ideal]. They come in two styles: transparent and
translucent. Transparent is essential for qualitative tests. One of us (J.F.Q.) prefers translucent
because it minimizes the possibility of photodecomposition of minuscule concentrations of dye.
(This phenomena has not yet been quantitatively tested, but is is one of the ways that
Eosin can be separated from small quantities of fluorescein).
Wash bugs with a high-speed jet of water; this will remove clay and silt which interferes
with the analysis. Pour about half of the charcoal into the numbered cup and cover the charcoal
with about 1/8" to 1/4" of elutant Wait. [Keep the other half of the charcoal in case it is needed
for evidence, or in case there is foul-up such as dropping ajar, etc.]
J.F.Q. has standardized on 10 ml of eluant which is easily decanted from a 1000 ml "Tilt
Dispenser". It is available from Markson Scientific [(800) 528-5114; #A-5060-10 Delivery Head,
10 ml ($49.00) plus A-5065-1000 flask ($19.50)].
The Kelly-green color of fluorescein is distinctive; but interpretation of weakly positive
tests is complicated by confusion between this characteristic color and the color of algae and
organic matter (humic acid, agricultural waste, etc.) which can occur as background. All of them
are sorbed by the charcoal and released from it by the elutant. Many a test has been falsely called
positive by well-meaning people who hadn't yet learned to distinguish the green of this
background from that of fluorescein. The human eye is better able to distinguish between them
than is the fluorometer.
The following is a semi-quantitative scheme proposed by Aley and Fletcher (1976, p. 16)
for assessing the degree of positiveness of fluorescein recovery on charcoal, when doing
quali tative traces:
1. Very strongly positive: Dye can be seen distinctly with the naked eye in sunlight or in
an artificially lighted room within 15 minutes of the time that KOH and alcohol are
added to the charcoal.
2. Strongly positive: Same as above, but after 15 minutes before three (3) hours.
3. Moderately positive: Dye can be seen with the naked eye in sunlight or in an artificially
lighted room but not until three to 24 hours after adding KOH and alcohol. The dye is
indistinct, and the observer feels it is necessary to verify the results by beaming a light
into the sample jar.
4. Weakly positive: Dye cannot be detected by the naked eye in sunlight or in an
artificially lighted room until more than 24 hours after adding KOH and alcohol. Dye
can be distinctly seen by the naked eye when a light is beamed through the sample jar.
For qualitative and semi-quantitative analysis, using a scanning spectrofluorophotometer,
we have standardized on a one hour ±10 minute elution in Aley solution. Other than for reasons
-16-
-------
of reproductibility, this is also because some of the Rhodamine WT tends to be partially resorbed
by the charcoal. In contrast, fluorescein is not resorbed; it continues to be released (T.A. Aley,
written communication, May 1989).
The light-beaming technique increases the detectability of fluorescein; it can be seen in
concentrations as low as one part per billion. The technique is simple and can be done in either of
two ways:
1. Use the focussed beam from a microscope lamp.
2. Use sunlight and a 4-inch reading glass.
The "green beam" is more visible if it is viewed against a black background.
Do not shake the cup. If you do, and if there is any suspended clay that was on the charcoal
and is then in the elutant, the beam will be white; you will have to wait, perhaps as long as several
hours, until it settles.
Elution of dye from activated charcoal does not give reproducible, reliable results on a
fluorometer or spectrofluorometer. There can be dye-losses due to microbiological reactions and
diffusion of adsorbed dye onto internal high-energy sites in the charcoal; not all dye is eluted or the
time necessary for attainment of equilibrium between elutant, charcoal, and dye is long, variable,
and temperature-sensitive (Smart and Friedrich, 1982, p. 108-111).
Jars used for elution should be thoroughly washed between uses. Some practioners prefer
to include a bit of Clorox in the rinse water, in order to eliminate the possibility of contamination
from one test to another. The use of Clorox, however, is probably not necessary. Disposable
plastic cups, at cost of less than two cents each, are much more cost-efficient and consistent with
good QA/QC practice.
No attempt will be made here to describe how to operate a fluorometer or scanning
spectrofluorophotometer for dye-analysis — for two reasons. If you are to use them, you will have
to become familiar with them just as you would with a computer. Also, limited time does not
allow for exposition of procedures for operation of instruments. Such exposition would be almost
useless without access to either instrument. The Jov of Dyeing (Aley, et a/.), however, will
discuss in great detail the use and operation of a fluorometer and a scanning spectrofluoro-
photometer.
Sometimes it is desirable to know whether a given site has an effect on wells in its
groundwater basin. There are four ways to do so. In order of increasing efficiency one could, for
each accessible well:
A. Suspend a bug in the well, allowing it to be in contact with the water. Estimated
efficiency = 1.
B. Suspend a bug inthe tank behind toilets in homes. Estimated efficiency = 10.
C. Pump the well continuously at about one gallon per minute, for the duration of the test
(days, weeks or months) and let it drain into either a:
1. One-gallon milk jug in which a bug is suspended. Estimated efficiency = 30, or
2. Dye-Intensity Level Device for Observation Wells. This device, shown in Figure 3,
samples ail the flow from the pumped well, rather than the small fraction that
makes contact with the bug in the milk jug. Estimated efficiency = 100.
For continuous sampling (generally recommended), attach the sampler to a garden hose, set
flow to one gallon per minute by measuring the time to fill a one-gallon milk bottle, and let water
spill onto the ground. When changing dye-detectors in the field, it is much quicker and easier to
replace them with a canister which has been pre-loaded in the office than to replace them
individually in the field. If one wishes to use this device in a household water line, for long-term,
incidental, use-related sampling rather than for continuous sampling, one can modify the design
and connect the entire contraption, in series, to the water line.
-17-
-------
oo
*"* I
•Ssr
GARDEN HOSE
WASHER
RUBBER
STOP, ON INSIDE
OF THE COUPLING,
RETAINS WASHER
AND DYE-DETECTORS
REDUCING
WASHER
POLYETHYLENE
TO
SCREEN
ALUMINUM
SCREEN
WIRE
SWISS
BEAUTY
PAD®
COTTON
CARTRIDGE
FILLED WITH
ACTIVATED
CHARCOAL
SCREEN
ALUMINUM
SCREEN
WIRE
fx^W^
3 ^
i?
n
py^swsys]"""'-""
W^^^^
VW////////4Z2 VS////////7//A^^^^^^ /
WM/A
m
FLOW
DIRECTION
^'//////)}Z7A W////<
k
s////rfk&±
y
\
\
\
\
* ' " N
C^\S\Vfc^W?
\
s
\
\
\
V
v
>
^
PIPE
PVC
1 * \ L
s?
Screwed Male End Ground To Tapered Glued
Fit, As Shown; Glued
ADAPTER, PIPE MALE ADAPTER MALE ADAPTER
TO GARDEN HOSE
FrIcllon-FIt
Glued
BRASS
PVC
PVC
PIPE COUPLING (CUT END CAP
TO INCLUDE STOP)
PVC PVC PVC
-CARTRIDGE WRAPPER
ALUMINUM
SCREEN WIRE
OR
NYLON
HOSIERY
HOSE ADAPTER
DYE-DETECTOR CANISTER
-------
Use of a Dye Intensity Level sampling device is not without problems. The cotton pad
tends to clog with sediment pumped from the well, thus impeding flow. The build-up of water
pressure may blow the dye-detector canisters off the hose adaptor. This can be partially alleviated
by using a paper punch to put a few holes in the pad. Alternatively one could modify the
construction of the device so that the dye-detector canister is screwed to the adaptor rather than held
onto it just by friciton. This, however, does not solve the problem of clogging.
Placement of Dye Detectors
Place the bugs in a stream or spring and at a location which maximizes the amount of water
which passes through them. Anchor a bug to a rock or use a gumdrop (Figure 1) or a Chaney-pin
(described in he subsequent section on principles of dye-tracing). In our opinion, gumdrops offer
the greatest convenience and flexibility where there is much suspended sediment and when bugs
must be recovered at sites where water levels may fluctuate more than a few feet In the Mammoth
Cave area there is commonly a fluctuation of 10 to 20 feet; 50 feet is possible. Accordingly, the
non-gumdrop end of the cord should be tied at an elevation not likely to be flooded during the
duration of the test being run. If flow velocities are very high, put the bugs in a more sheltered
area. Otherwise cotton may be washed away and charcoal may be washed through the screen; the
test can be lost
Pilfering of gumdrops can be minimized by using a plastic-laminated tag similar to the one
shown in Figure 4. It works.
Estimation of Dye Quantities Needed
This is a matter of experience. We present some rules of thumb. For the average
Kentucky spring (whatever that is), under average conditions (whatever they are), for qualitative
tests, we would start with the following and adjust amounts downward or upward, as initial results
suggest:
Fluorescein ~ one pound per mile, up to five pounds
Direct Yellow 96, Tinopal 5BW GX, Phorwite BBH Pure ~ at least three times as much as
for fluorescein
Phorwite AR Solution -- one gallon per mile, up to four gallons
Rhodamine WT - not recommend for most qualitative tests because the tea-like color of
any organic matter commonly present can mask the pink tint of the dye unless
relatively large quantities are used If you must, try 0.5 gallon per mile.
For quantitative and semi-quantitative tests, using a fluorometer or spectrofluoro-
photometer, use one tenth to one hundreth as much dye. J.F.Q., using a Shimadzu RF-540
scanning spectrofluorophotometer, has detected concentrations of fluorescein (in water) of less
than 4 parts per trillion. Activated charcoal can increase this detectability by a factor of 10 to 400.
A review of various published formulas for calculating how much dye to use for tracing
tests has shown that the "correct" amount for given flow conditions ranges over 11 order of
magnitude (Field, Wilhelm, and Quinlan, in review). Accordingly, an attempt to calculate the
amount of dye to use is an exercise in either naivete or sophistry. There is no substitute for
experience.
It is always a goal to limit the amount of dye to a quantity which is insufficient to impart a
visible coloration to water. This goal is inpired by health and aesthetic (public relation)
considerations, but there are times when, for public confidence in results (or for rabble-rousing), it
is advisable to have the resurgence of the dye vividly obvious to all. The limit of visibility of
fluorescein is about 0.03 ppm; 0.3 ppm can be easily seen with the naked eye.
-19-
-------
PLEASE DO NOT DISTURB
THIS IS PART OF A RESEARCH PROJECT CONCERNED WITH
TRACING KHERF. SPRING HATERS COME FROM. FOR
INFORMATION ABOUT THIS PROJECT AKO ITS RESULTS
CALL Jl" QUINLAN AT: 758-2394.
IF TOO HAVE REMOVED THIS DEVICE FROM THE WATER
WOULD YOU PLEASE PUT IT uct THE WAT YOU FOUKO
IT.
THAN* YOU.
Figure 4. Laminated tag sometimes attached to a gumdrop in order to discourage removal of it
from water by people who are curious as to what it is. (Shown actual size.)
-20-
-------
The U.S. Geological Survey has a policy of limiting the maximum concentration of
fluorescent dyes at water-user withdrawal points to 0.01 ppm (Hubbard et al., 1982, p. 15), but
this is an arbitrary, conservative, non-obligatory limit rather than one based on a comprehensive
assessment of toxicity.
We urge caution in the use of Direct Yellow 96. If you use too high a concentration in
potable water supply you might dye someone's laundry pale yellow. People will wonder where
the yellow came from. There will be no injury to their health but they might be irked. And you
might have to buy some clothes.
Injection of Dye
This too is an art. Whenever possible, use dye pre-mixed in water. It can be stored and
carried in gallon milk jugs. I recommend using one pound per gallon of water for fluorescein and
two pounds per gallon for Direct Yellow 96 and optical brightener. Warning; Do not store Direct
Yellow % in a plastic milk jug for more than a month or two. Direct Yellow 96 eats plastic; the
jugs crack and leak. The timing and location of the consequent disaster follows Murphy's Law.
Although you can easily inject fluorescein powder and Tinopal 5 BM GX where there is
little or no wind, never inject Direct Yellow 96 or Phorwite BBH Pure in powdered form — unless
you arc prepared to have contamination of clothes and everything you come in contact with. When
injecting fluorescein powder and there is a wind, always be upwind of the powder, keep your
equipment and all onlookers upwind and well away from it Try to be to the side of or upstream
from dye being injected into a stream. We have found Wellington-type rubber boots to be ideal
and long, electrician-type rubber gloves to be practical. (Playtex gloves aren't long enough). One
of us (E.C.A) has recently had good luck prewetting fluorescein powder with a roughly equal
volume of ethanol. The resulting slurry does not blow around and mixes readily with water.
A tank-truck of water can be used for dye-injection into the subsurface where natural flow,
is not available. We have had success with a "primer" of 1000 gallons before dye is injected. The
injection is followed by a "chaser" of about 2000 gallons. More water or less water can be used,
as circumstances seem to require, but it is important to have a primer and a chaser.
Sampling Frequency
There are few hard and fast rules about sampling frequency. Most questions about it can
be answered by using common sense after asking the folowing questions:
1. Why is this test being run?
2. How accurately do we need to know the results? (Mere yes or no, or the third decimal
place?)
3. How vital is it to have velocity data?
4. What is the estimated flow velocity?
5. What is the estimated flow duration?
Accordingly, sample frequency may range from a few minutes to daily, for water samples,
and a few hours to weekly for bugs and Dye-Intensity Level Devices attached to pumped wells.
An automatic sampler used for collection of dye in water should always by "backed up" by
the use of a bug — so you don't loose all information when (not if) the sampler fails. We know of
tests (by others) that were lost when this rule was not observed.
-21-
-------
Recording Test Data
Daily tasks associated with a dye-trace are recorded on the bug sheet shown in Figure 5.
Note the column for recording the results of the dye tests. The cup number or jar number is
always recorded on the bug sheet, checked, and rechecked before any eluant is poured into the jar.
The results of a dye test should be recorded on a data sheet similar to that shown in Figure
6. J.F.Q. fills them in with an ordinary pencil but records all positive results with a red pen. This
makes positive results less likely to be overlooked. A weakly positive test is recorded with a single
dotted red cross.
Simultaneous Use of Several Dyes
If a project requires tracing results from more than one input site, it is expedient to run
several traces simultaneously. Doing so can cut time and labor costs by one-half to three-fourths
and it allows the conduct of tests under the same flow conditions. The following recommendations
are based on successful results in hundreds of dye-tests:
Qualitative and semi-qualitative tests (using either visual indentification or a fluorometer):
Use 2 dyes, either fluorescein and optical brightener or Direct Yellow 96. This can be
followed by Rhodamine WT and whichever of the cotton-reactive dyes was not used as the
in the first test
Semi-quantitative and qualitative tests (using a scanning spectrofluorophotometer):
Use 3 or 4 dyes ~ Direct Yellow 96, fluorescein, Eosin (Acid Red 87), and Acid Red 52 or
Rhodamine WT—but only if you are extremely confident about site hydrogeology and test
design. Optical brightener can be used only if it is recovered on cotton detectors. Direct
Yellow 96 is more efficiently recovered on cotton than on charcoal. (Brighteners can not
be reliably analyzed in water or elutant because their molecules are metastable; their
analyses are not reproducible).
One should use extreme caution in designing a dye-test to be performed with two or more
dyes. If the dyes are not recovered somewhere, they can not be used again in the same
groundwater basin or in adjacent basins. This is because one would not know whether dye-
recovery from a second injection could be from the first. Therefore, until one has moderate
experience in the design, operation, and interpretation of tests using a single dye, don't try multiple
traces unless you are willing to risk compromising the trace and site.
Sources of Dyes And Related Materials
Dye costs are given in Table 1. Sources of supply are given in Table 2.
Standards
The preparation of standards will be discussed in The Joy of Dyeing. In the meantime, the
interested reader is referred to Wilson et al. (1986, p. 23-28). [Note: This reference is concerned
solely with surface stream studies. Its discussion of fluorometers is quite useful].
-22-
-------
BUG SHEET- Sites Monitored for Dyes
TEST
RUN BY
DATE
J
A
R
#
AY73
Fluor.
Rhod.
WT
Charcoal
OB
DY
96
Cotton
Site
Code
Location
Pull
Gum
Drop
Pull
All
Bugs
Hang
Cotton
Hang
Charcoal
Rig
Gum
Drop
Figure 5. Sheet for recording tasks performed and results.
-23-
-------
Input Point (Name of test)
Tracer
RECORD OF DYE TEST
Date
Time
Precipitation Before or During Test
Amount
Investigator
Detectors placed at
Date, Number of days since dye input, and results
Date
Number
of Days
Legen± - Negative Result B Background
+ Positive B+ Fluorescence Significantly
Very Positive
Spectacularly, Gloriously
Positive
Bugs not changed
Installed
Remarks:
above background
NR Not Recovered (because of
high water or other reason)
L Bug Lost or Stolen
G New or Extra Gumdrop
Interpretation:.
Figure 6. Record of Dye test
-24-
-------
Stability of Rhodamine - WT
Rhodamine WT has a well-deserved reputation of being among the more stable of the
fluorescent dyes. The most complete reference to the stability and other properties of fluorescent
dyes is Smart and Laidlaw (1977). Rhodamine WT is not subject to severe photodecomposition as
is fluorescein. Standards routinely have shelf-lifes of a year or more. If the Rhodamine WT
standards are in daily repetitive use in situations involving very large numbers of analyses, the
standards may begin to degrade after a few months use. A decade ago we found that mixing in a
1% ammonium hydroxide solution improves the stability of Rhodamine WT standards. This may
be simply an inadvertent disinfection of the water, however.
We have conducted tests in which water samples from springs, wells, and monitor wells
near a landfill in Minnesota were spiked with known concentrations of Rhodamine WT and stored
under varying conditions. Three storage conditions were investigated: 1) room temperature in
normal indoor lighting, 2) room temperature in the dark, and 3) refrigerated at about 4*C in the
dark. No significant decrease of the fluorescence of these samples has been observed in over a
year. Part of these results are shown in Figure 7.
Similar tests have been conducted on the stability of Rhodamine WT after it has been eluted
from a charcoal detector. In these tests, charcoal detectors were placed in a spring, collected after
one week , and eluted in the normal fashion. The resulting sets of conditions listed above, and
then analyzed periodically. The results of this test are shown in Figure 8. The apparent
fluorescence of the elutant increased for a day or two and then began to decrease. Storage in the
cold reduced but did not stop the decrease. There is very little difference in the rate of decay of the
samples stored in the light and dark at room temperature. The general conclusion is that the elutant
from charcoal detectors should always be analyzed within a day or so of the elution.
We have not investigated the stability of charcoal detectors during storage between
collection and elution. We believe it to be prudent, however, to store detectors in a refrigerator
until they are eluted. They could even be frozen, but this too has not been evaluated.
The single stability problem we have encountered with Rhodamine WT has been traced to a
biological source. There are a group of bacteria that live in distilled water, Pseudomonas
fluorescens and other Pseudomonas species (Balows et al., 1985; Morton, 1983; and Parker,
1983). These bacteria are a major, almost ubiquitous problem in hospital and research laboratory
distilled water systems and are resistant to (indeed will happily live on) many common antibiotics.
These bacteria are very omnivorous consumers of organic compounds. They think that
Rhodamine WT, fluorescein, and other dyes are equivalent to a trip to the local fast-food outlet
We have seen Rhodamine WT standards prepared with distilled water containing these bacteria
disappear with a half-life of hours. The 1% ammonium hydroxide solution described above
apparently prevents the bacteria from growing. Boiling of your distilled water for 30 minutes,
immediately before use, will also disinfect the water.
Recent Advances In The Interpretation of Dye Traces
Significant advances have been made during the past few years in the interpretation of dye-
trace results in karst terranes. Repeated traces in the same swallet-to-spring system show that the
travel times in several British systems show an inverse, non-linear (hyperbolic) relation with
discharge (Smart, 1981; Stanton and Smart, 1981). They showed that the breakthrough curve
(time vs. dye concentration) indicates increasing dispersion with decreasing discharge, a
consequence on the increasing "dead volume" of the conduit Multiple dye-peaks suggest that
branched flow-routes are present and that the relative significance of each branch changes as spring
discharge changes. Where the travel time shows a 1 : 1 inverse relationship to discharge, a system
is interpreted to be phreatic; this characteristic of simple phreatic streams allows graphic analysis to
distinguish them from vadose and complex phreatic streams. They also showed that the amount of
Rhodamine WT which is sorbed in transit from swallet to resurgence is appreciably greater at low
flows than at high flows. A pulse-test (a powerful technique explained and revised by Smith,
-25-
-------
an
con
K'S^ 31
£3 N"* Q. n
rt
-------
11
M>
g.00
(re ey
ef
5-5-
312.
era c
'8
I
§
f
o
o
i
Rhpdamine WT Stability
in Elutant Solution
0.0
07/11/89 08/30/89 10/19/89
12/08/89 01/27/90 03/18/90 05/07/90 06/26/90 08/15/90
date
-------
1977, p. 99-101) made during low flow in one of the systems indicates that only a small
proportion of the conduit is vadose (Smart and Hodge, 1980). Significant advances in quantitative
dye-tracing have also been made by C.C. Smart (1983a, 1983b) and Smart and Ford (1982) in a
glacienzed Canadian karst, and by Lang et al. (1979), Collings (1982), Burkimsher (1983), and
Brugman (1986) in glaciers -- as well as by Mull et al. (1988) in karst. A useful 2-volume review
of tracing has been published by Caspar (1987a, 1987b).
Interpretation of results of intensive tracing, especially when done with complementary
chemical analyses and discharge measurements, has made it possible to construct mental models
that enable visualization and comprehension of water and pollutant movement through karst
terranes - as shown by Jones (1973) Quinlan and Rowe (1977), Aley (1977,1978), Quinlan and
Ray (1981), Crawford (1984), Friederich and Smart (1982), C.C Smart (1984a, 1984b),
Hallberg et al. (1985), Quinlan and Ewers (1985), Vandike (1985), Gunn (1986), and many
o iers. A recently proposed mathematical technique for analysis of transport of material through
any kind of natural system examines residence-time distributions and can relate tracer studies to
empirical models (Buffham, 1985; discussed by Woods, 1985). Perhaps this technique and one
proposed by Rathor et al. (1985) can be used for karst studies.
Four very significant recent papers relevant to the interpretation of tracer tests have been
published by Dreiss (1989a, 1989b) and by Maloszewski and Zuber (1990, 1992). Dreiss used
continuous monitoring of stage and water chemistry as a tracer and made a sophisticated
mathematical analysis of residence time, travel time, and effective dispersitiy. The behavior of
tracers in fissured double-porosity aquifers has been described and successfully modeled in a
seminal paper by Maloszewski and Zuber (1990,1992).
Everts et al. (1989; included herein as Appendix E) have shown movement of Rhodamine
WT and other tracers through packed soil columns in the laboratory at a velocity of 10"4 cm/sec
(about 0.012 foot/hour). In the field, however, where much flow was in macropores, flow
velocities of about 9 feet per minute in undisturbed soils were observed. Their experiments have
many important implications about the movement of fertilizers and pesticides through the soil
profile and into aquifers.
Excellent guides to who is doing what, where, and how in tracing, especially in non-karstic
settings, are the abstracts for: 1) the 1989 N.W.W.A. symposium, "Tracers in Hydrogeology:
Principles, Problems, and Practical Applications", that were published in Ground Water, v. 27, p.
718-728, and v. 28, p. 154-155 and 2) the 1990 "International Conference and Workshop on
Transport and Mass Exchange Processes in Sand and Gravel Aquifers: Field and Modeling
Studies".
Principles Which Maximize The Cost-Efficiency, Success, Reliability,
and Utility of Dye-Tracing Studies
In aggregate, we have had more than thirty years of experience in running and directing
hundreds of dye-tests in carbonate rocks. We are still learning. This experience has taught us
much about how to and how not to run dye tests. The following are some of the lessons we have
learned and their applications to the study of karst hydrology. The principles are not listed in any
particular order of significance.
PRINCIPLE 1. Dye-tests should be designed so that there is always a positive result ~
somewhere.
DISCUSSION: Negative results to a monitored site are always questionable. But if they
are accompanied by strongly positive results to another site they can be considered
reliable for the flow conditions during the test ~ and they can be used. (The results
might be different when the flow is much lower or higher.) Also, until one knows
where a tracer has gone, that tracer can not be used in the same groundwater basin
for a long time. If one has not accounted for the dye from the first test, and if the
-28-
-------
same dye is used for a second, third, or even fourth or fifth test, one never knows
whether the positive results of a given test are a result of dye injection from the
first, second, third, fourth, or fifth test
PRINCIPLE 2. Always run background tests before tracing with optical brightener.
DICUSSION: This is always necessary with any dye if there is a possibility of litigation
over interprettiton of tracer results, or if it is possible that another investigator may
be tracing in the same area. Litigation aside, many karst areas have a slight to
moderate brightener background because brighteners are in laundry detergents;
laundry wastewater is part of the effluent from septic tanks. Brightener can still be
used, and the background can be overridden, but a standard for comparison is
needed. [One can use anomalously high background for brighteners as a
prospecting guide to locating effluent from sewage treatment plants, sewer lines,
and septic field systems (Quinlan and Rowe, 1977; Aley, 1985).] Direct Yellow 96
can be ideal for use in areas where there is a possibility that someone else might
have injected dye recently, check for it. Cave explorers, for example, sometimes
run tests, generally with fluorescein. Similarly, if you know of someone who may
be doing some tracing in you area of interest, check with him. The courtesy is
appreciated and may be mutually beneficial. Also, he or she might have and be
willing to share data that will be useful to you. Return the favor.
PRINCIPLE 3. Generally speaking, never follow a qualitative test to a given recovery
point with the same dye if the second test might also go the same recovery point —
unless a major storm has occurred after the first test is over and before the second is
started.
DISCUSSION: For traces to a given recovery site one should always alternate from one
dye to another — unless a major rainstorm has occurred. This is because a rain that
occurs after the first test may flush dye-laden water which was locally left stranded
in the system (either in pools or as coatings on passage walls), thus giving a falsely
positive result The rain that occurs after the first dye-test may also flush dye which
had been temporarily sorbed onto clays and left "stranded" by receding
floodwaters, thus also giving a second maximum in a plot of dye-concentration vs.
time. Our preference is to switch from a dye recovered on one type of bug to a dye
recovered on the other type.
PRINCIPLE 4. If possible, always try to run two tests simultaneously ~ to the same
possible recovery site or to adjacent sites.
DISCUSSION: This doubles productivity and cuts labor costs per test to approximately
half. For a given potential recovery site, and depending upon circumstances, one
should use one dye detected on charcoal and another detected on cotton. For
example, use fluorescein and optical brightener. This can be followed by
Rhodamine WT and Direct yellow 96. (Of course one can start with fluorescein
and Direct Yellow 96 and follow it with Rhodamine WT and Tinopal 5BM GX or
Phorwite BBH Pure or AR Solution). In some aquifers, however, where dilution
(and perhaps also sorption) is high, fluorescein is the only dye which can be used
for qualitative tests; Rodamine WT could be used for quantitative tests.
PRINCIPLE 5. Qualitative dye-tests give useful results much more rapidly and cost-
efficiently than quantitative tests; often, they are all that is needed.
DISCUSSION: Qualitative tests, using activated coconut charcoal and/or cotton as
detectors can be run with a detector cost of less than $2.00 each. The detectors
work 24 hours per day and need to be changed only once or twice per week. (Of
course, samples can be taken much more frequently if there is a need for doing so).
Although the materials cost is trivial, the preparation, setting, changing, elution,
-29-
-------
and analysis of bugs is very labor-intensive. Much extremely useful information
can be gained by using grab-samples (or an automatic sampler) and a fluorometer or
a spectrofluorophotometer. Fluorescein is the most practical dye to use with
activated charcoal; Rhodamine WT is not recommended for general qualitative use.
Each of the other three dyes is suitable for quantitative use, but activated charcoal
will not give accurate quantitative results with any of them. Optical brightener and
Direct yellow 96 work equally well with cotton detectors.
PRINCIPLE 6. Set bugs at all the likely places to which dye might flow, many of the
unlikely places, and a few of the stupid, incredibly impossible places.
DISCUSSION: Use enough dye to reach the most distant of the above places. Dye is
expensive, but it is cheap relative to the time and labor costs necessary to re-run a
trace which isn't done properly the first time. The cost of charcoal and cotton
detectors is minuscule.
PRINCIPLE 7. Delineation of a groundwater basin by dye-tracing should generally
involve partial delineation of its neighboring basins.
DISCUSSION: This is a corollary of Principle #1. You should generally set bugs in the
adjacent groundwater basins. By having most dye flow to them, by simultaneously
working opposite sides of the basin to be delineated, and by alternating the dyes
used, one can work more rapidly and efficiently. Both sides of a suspected divide
should be tested by tracing, but the concept of an imaginary line that neatly
separates one basin from another doesn't always apply in karst terranes. In most,
dye which is injected near the center of a basin goes to a spring or to a distributary
group of springs. But in some terranes, dye injected close to the assumed
boundary between two basins flows to each of them. Smart (1977) has suggested
that, when it is desirable to compute a water balance, the basin boundary would be
chosen so as to coincide with dye-inputs in which the dye was divided evenly
between the springs of two adjacent basins. The "neatness" of basin boundaries is
also shattered by results in West Virginia described by Jones (1984c). Dye from a
sinking stream flows about a mile in 24 hours to a particular spring. But a total of
only about 5% of the dye goes there; the remainder of the dye was recovered a
month later at a spring 12 miles away. Although calculation of basin area for a
water balance may seem like an interesting but merely academic exercise,
delineation of the boundary, and determination of whether it is fuzzy or sharp, is
extremely practical and necessary if one must monitor a site or make an emergency
response to a spill of toxic agents. Such a determination could only be done if
springs outside the basin to be delineated were also bugged.
PRINCIPLE 8. Always, always replace one set of bugs with another, even if you are sure
that the test is over.
DISCUSSION: Bugs are cheap relative to labor costs and the risk of losing a test. Also,
sites should be monitored until you know the dye is out of the system. Meanwhile,
the dye may also show up elsewhere in either distributary flow or radial flow.
PRINCIPLE 9. Be paranoid about the possibility of contamination of samples, tampering,
or removal of bugs by people curious as to what they are.
DISCUSSION: Contamination happens, sometimes in he strangest ways. Anticipation of
its possibility is the key to its prevention. Recognition of possible tampering is
made easier if: 1) extra, well-hidden bugs are set, 2) the dye (or dyes) used and
the bugged locations are kept secret until the test is over, 3) control bugs are set
upstream from tributary junction, and 4) optical brightener is secretly used with, for
example, fluorescein, as a check. Also, careful monitoring of the timing, location,
and dye-concentration of the positive results will detect tampering by all but the
-30-
-------
most skillful attempt at manipulation of the test results. Minimize loss of bugs by
using 3/16" goldline or black nylon cord on gumdrops; it "blends with dirt Even
rub dirt onto new cord. Sometimes a monofilament fishing line is best When
goldline was unavailable, J.F.Q. has had black nylon cord woven to his
specifications. Never use white nylon cord. People steal it If there is any
possibility of public access to the bugs (by hikers, fishermen, etc.), it is necessary
to set an extra bug or two at each site. Steel "Chancy pins" can be driven into a
stream bottom and used for holding bugs without the aid of telltale cords. If
necessary, when Chancy pins are buried by alluvium, they can be recovered with
the aid of a metal detector. The use of laminated tags like the one shown in Figure
3 is effective.
PRINCIPLE 10. If you have a choice, inject dye during moderate flow conditions, while
stream flow is in recession.
DISCUSSION: Dilution will probably be minimal and flow times will be average. This is
the optimum time for tracing. See Principle 11. Spring and fall flows are often
higher than summer flows. The evapotranspiration component is larger in the
summer.
PRINCIPLE 11. Although the first dye-test should be run when conditions seem
optimum, they should also be run at both low-flow conditions and high-flow
conditions — in many situations, but not all.
DISCUSSION: Flow-routes that function only during moderate and flood-flow conditions
may divert some of the water to springs in adjacent groundwater basins. During
such conditions dye (or leachate from a site) can go to springs which it does not
reach during low flow. In the Mammoth Cave Region the flow time between two
points about 5 miles apart (Parker Cave and Mill Hole) ranges from 18 days to less
than 24 hours. The average flow time is 3 to 5 days.
PRINCIPLE 12. Do not use a dye unless you know what it is.
DISCUSSION: There are many dyes sold as "leak-tracers" which are carcinogenic,
mutagenic, etc. The companies selling them disclaim responsibility for their use,
but you can't — morally or legally. See the toxicity review by Smart (1984). In
brief, fluorescein, Rhodamine WT, and optical brighteners are the safest dyes to
use. Available data for Direct Yellow 96 does not suggest any problems with its
use but it has not been tested for mutagenicity. Rhodamine B is a known
carcinogen and possible mutagen; it should never be used unless you can give a
rigorous justification for using it rather than a different dye. It should be stressed,
however, that the toxicity of Rhodamine B has been shown to be due to impurities
(diluents) within some batches of technical grades of the dye, not the dye itself
(Smart, 1984). According to Smart, no water-tracing dye is acutely hazardous
because of short exposure to the locally very high concentrations of it which can
briefly occur where dye is injected. A concentration of 1 ppm for 48 hours can be
endured by the more sensitive organisms.
PRINCIPLE 13. Always set bugs before you handle or dump dye.
DISCUSSION: This lessens the probability of contamination of bugs. Also, carry the dye
containers in plastic bags and in the back of a pickup truck. If you are working by
yourself and must load your dye after making up the bugs and loading them, wash
your hands and inspect them under an ultraviolet lamp after you load the dye. [We
often have to decontaminate spots of optical brightener on the steering wheel of a
pickup truck. Keep rubbing it with cotton swabs until no more brightener transfers
to cotton. Ideally, two people and different vehicles could be sued, but this is not
necessary. Be careful, very careful.
-31-
-------
PRINCIPLE 14. If there are enough accessible wells, if there is not perched water above
the main water mass, and if project needs can justify and afford it, make a map of
the potentiometric surface, preferably before dye-tracing is started or completed.
DISCUSSION: The map can pay for itself, many times over. It is a useful supplement
which can greatly aid decision-making in designing dye-tests and it can greatly
decrease the number of test otherwise needed. Such a map can be almost useless,
however, if there are not enough wells to allow reliable contouring, or if there are
siltstones or shales within the carbonate section; they tend to perch the water and
make the data misleading and difficult to interpret
PRINCIPLE 15. Know your dye.
DISCUSSION: Play with it in your office or laboratory, before you start running tests for
the first time. Make up a fluorescein solution that is approximately 1 ppm. Pour
approximately 500 ml of the 1 ppm solution through a bug at such a rate that it takes
about 15 to 20 seconds to pass through the bug. (Alternatively, and with a pair of
tongs, swish a bug through about 500 ml of the 1 ppm solution for several
minutes). Rinse the bug in clean tap water for about a minute and elute. You
should get a very strongly positive result. Repeat, using 500 to 1000 ml of 0.1,
0.01, and 0.001 ppm solutions. Results from these elutions are only semi-
quantitative, partly because not all of the dye will be released from the charcoal.
But it will give you a feeling for how to elute and how to recognize fluorescein.
Having run this series of test, leave a "background" bug in a stream for a week and
run the same series of tests on the elutant from it when you repeat the above series
on a 1 ppm solution of fluorescein. Compare the results.
As with love-making, there is only so much you can learn from the printed word. There is
no substitute for experience or experimentation.
TRACING WITH ENVIRONMENTAL ISOTOPES
Isotopic techniques have been recognized as powerful tools in hydrogeologic investigations
for half a century. Environmental isotopes are among the most ideal of tracers. Tritium and
oxygen and hydrogen are part of the water molecule itself and are the only tracers that actually trace
the water.
If you are not familiar with isotope geochemistry, we are not going to teach it to you in this
short course. If you are interested, start with a text on isotope geochemistry such as Faure (1986)
and then go to a specialized text such as Fritz and Fontes (1980) and Fontes (1983) for the
hydrologic applications. If you are already familiar with isotopic studies, you know that they can
yield very useful and in some cases unique information on hydrogeologic systems but are neither
cheap nor simple to interpret Single measurements do you little good. You need typically at least
tens of measurements before the patterns begin to emerge and most workers expect mat reasonably
complete chemical analyses should accompany each isotopic analysis.
14C and Tritium
Radiocarbon and tritium can be used as straight tracers but they are usually used in attempts
to define the "age", or "residence time" of the water. Conceptually, one is trying to quantify the
length of time the water sample has been isolated from the surface. These concepts are deceptively
simple and are very effective public education tools. Many professional hydrogeologists are
-32-
-------
justifiably skeptical of such ages. Mixing is always present in ground water situations and at best
the ages are averages.
Those averages can, however, fundamentally change how you and your client view your
system. If the water from a sampling point does not contain tritum from the atmospheric testing of
nuclear weapons in the last 35 years, it will be surprising if the tracer shows up at that point
(unless you have injected it in a long residence-time flow path). Such information can save you
and your client lots of time, money, and frustration.
Tritium determinations range in cost from about $50 to more than $300 per sample,
depending on the detection limit, accuracy, and turn around time desired. Radiocarbon
determinations are currently in the $250 per sample range.
5^0 and 5D
Stable isotope measurements of water currently cost about $75 from a commercial
laboratory.
TRACING WITH IONS
Ionic species have long been used in tracing studies. They work very well as tracers. In
general the analytical costs of the necessary chemical analysis are greater than for dyes and the
detection limits are often 100 to 1000 times less for ionic species. Natural background levels of
many of the best ionic tracers can be a limiting factor. Nevertheless, ions will continue to be used
extensively as tracers. Davis et al. (1985) reviews much of the relevant literature. Appendix E
reports on direct comparisons of Rhodamine WT with lithium, bromide and nitrate ions in a tracer
study.
Cations
Any soluble cation is a potential tracer. Good results have been obtained using ammonia,
sodium, lithium, and potassium ions as tracers. These are among the most conservative of the
cationic species. Other metal ions are often inadvertent tracers in pollution situations.
Anions
The simple anions such as chloride, bromide, iodide, nitrate, and sulfate are generally taken
to be the most conservative of the tracers. They are extensively used in tracing. The limiting factor
is usually the natural background.
UNINTENDED TRACER TESTS
Almost every groundwater pollution problem can be viewed as an uncontrolled, unintended
tracer test. These situations are messy, usually potically sensitive, and usually difficult to work
with. Nevertheless, most of us will be working on such situations. Society is committing a lot of
financial resources to these situations. We should be able to extract some useful information from
all of the effort being spent
-33-
-------
COSTS OF TRACER TESTS
The costs of tracer tests range from essentially free to millions of dollars. In some cases,
someone has already done the tracer test for you. If you are working in the Mammoth Cave area,
the map in Appendix D already exists. You may need to refine some of theat information but you
do not need to start over. Tracer tests can be divided roughly into three levels of economic
investment but a continuum of investment, effort, and sophistication exists.
VOLUNTEER/AMATEUR
A number of very sophisticated and useful traces are done with direct costs of a few
dollars. These traces are done by volunteers. If managed correctly, such traces are as useful and
valid as much more expensive traces. Indeed, it is often possible to do traces with motivated
volunteers that would be physically and financially impossible otherwise. The volunteers
(students, cave-explorers, local residents, public interest groups, etc.) are most often used to
collect the samples for analysis but surprisingly talented analysits and very sophisticated analytical
equipment are sometimes available free of charge.
PROFESSIONAL
Traces are often run entirely with professional field, laboratory, and interpretation
personnel. The costs of such traces start in the range of thousands of dollars for the simplest traces
and range up into the hundreds of thousands of dollars. This type of trace is increasingly
common. These tracer studies sometimes attract individuals and companies who have very little
experiece in tracing although they may have impeccable credentials as consultants. If you are the
client, you may not want to pay for the learning curve of that individual or organization. In other
cases you may be willing to pay for that learning curve in order to establish in-house or local
expertise.
BUREAUCRATIC
If Superfund or analogous public programs are involved, the costs of the simplest traces
will be in the tens of thousands of dollars range and tracer studies costing millions of dollars are
known. In most cases, it is very difficult to convince oneself that the resulting product is worth the
extra cost The traces are rarely ten to a hundred times "better" than a professional trace or
thousands to millions of times better than volunteer traces. Indeed, the complexity of the
bureaucratic QA/CA often degrades the test
INTERPRETATION OF TRACER TESTS
Now that you have done all of the field work, the real fun begins. You must interpret your
results and present the results in a convincing fashion. Jones (1984c) is a good place t start
learning how to interpret your results. The interpretation of dye tracing results is explicitly
illustrated in the several appendices of this manual.
-34-
-------
posrnvE RESULTS
Positive results are the simplest kind to interpret You have found your dye and
demonstrated a connection between the input point(s) and recovery point(s). Depending on the
frequency with which the recovery point(s) have been sampled, travel time can be determined to
varying accuracies. If the trace has been quantitative, details of the dilution, despersion, and
underground plumbing may be evident in your results.
The major caution is to beware of false positives. The smaller the positive result, the
greater the danger of false positives. The best protection against false positives involve repeating
the test, perhaps with a different dye. As more workers begin to do dye traces, the odds of
inadvertently detecting someone else's dye becomes progressively greater. Knowing who else is
working in a given area is desirable but is not always possible. We have encountered cases of
deliberate sabotage in which someone introduced dye into a spring to confuse and confound a
trace. Such cases are rare but have happened.
Not only need you worry about another dye trace being conducted in you area, the chances
of detecting one or more dyes from the disposal of commercial products containing dye is an
increaseing possibility. Rhodamine B is used in a surprising number of commercial applications.
Commercially treated seed is colored with Rhodamine B, for example. Ideally, background testing
should be conducted for a length of time comparable to the projected trace. Unfortunately, we
have rarely encountered a client willing to patiently wait and pay for months of background testing
before dye is injected. Repeated tests and varying dyes can establish the validity of a positive
result.
NEGATIVE RESULTS
Negative results in dye tracing are much more difficult to interpret and are therefore
frustrating bom to the tracer and to the client The fundamental principle is that you can not prove a
negative. The absence of proof is not proof of absence of a connection. There are three broad
categories of reasons for negative results.
1. The trace really is negative. The water flow went to some other location which you
did not sample. This is always possible. The best defense is to sample
everywhere. Sample every accessible location that the dye can conceivably reach —
and several totally inconceivable locations. Karst hydrogeologic systems continue
to surprise even experienced workers.
2. The dye simply had not yet reached the sampling locations when the trace was
terminated. Every trace is conducted under some type of economic and time
constraints. If the transit time is longer than you can afford to wait you will not
see the dye. Some other hydrologjc technique should be used.
3. The dye was diluted below analytical detection limits and/or adsorbed before it
reached the sampling locations. The use of conservative tracers and the use of
adequate amount of tracers are the best defense to avoid this case. The cost of
tracer dye is usually one of the least expensive components of large traces. Trying
to economize here is usually very foolish.
The pressure to "interpret" negative traces as proof of an absence of connections is often
enormous. Resist that pressure. Appendix I discusses some of the ethical principles that are
relevant to this problem.
-35-
-------
REFERENCES CITED
Aley, T. 1985. Optical brightener sampling; a reconnaissance tool for detecting sewage in karst
groundwater. Hydrological Science and Technology: Short Papers, v. 1, p. 45-48.
Aley, T., and Fletcher, M.W. 1976. The Water Tracer's Cookbook. Missouri Speleology, v. 16,
no. 3, p. 1-32.
Aley, T. Quinlan, J.F., Alexander, E.G., Jr., and Behrens, H. 1991. The Joy of Dyeing: A
Compendium of Practical Techniques for Tracing Groundwater, Especially in Karst
Terranes. National Water Well Association, Dublin, Ohio, (in prep)
Balows, A., Hausler, W.J., Jr., and Shadomy, HJ. (eds.). 1985. Pseudomonas Species. la:
Manual of Clinical Microbiology (4th ed.), Amer. Soc. Microbio., Washington, p. 353-
356.
Bedmar, A.P. 1990. Report to the Director General [on the] Advisory Group Meeting on Use of
Artificial Tracer in Hydrology. International Atomic Energy Agency, Vienna. 61p.
Brugman, M. 1986. Dye-tracing events at Variegated Glacier, Alaska. Ph.D. dissertation
(Geology), California Institute of Technology.
Buffman, B.A. 1985. Residence-time distributions in regions of steady-flow systems. Nature, v.
314, p. 606-608.
Buridmsher, M. 1983. Investigations of glacier hydrological systems using dye tracer techniques:
Observations at Pasterzengletscher, Austria. Journal of Glaciology. v. 29, p. 403-416.
Collins, D.N. 1982. Row-routing of meltwater in an alpine glacier as indicated by dye tracer test
Beitrage zur Geologic der Schweiz - Hydrologie. v. 28, pL 2, p. 523-534.
Crawford, N.C. 1984. Karst landform development along the Cumberland Plateau Escarpment of
Tennessee, in LaFluer, R.G., ed. Groundwater as a Geomorphic Agent Allen & Unwin,
Boston, p. 294-339.
Davis, S.N., Campbell, D.W., Bentley, H.W., and Flynn, T.J. 1985. Ground Water Tracers.
National Water Well Association, Worthington, Ohio. 200 p.
Domenico, P.A., and Schwartz, F.A. 1990. Physical and Chemical Hydrogeology. Wiley, N.Y.
824 p.
Dreiss, S.J. 1989a. Regional scale transport in a karst aquifer. 1. Component separation of
spring flow hydrography. Water Resources Research, v. 25, p. 117-125.
Dreiss, S.J. 1989b. Regional scale transport in a karst aquifer. 2. Linear systems and time
moment analysis. Water Resources Research, v. 25, p. 126-134.
Everts, C.J., Kanwar, R.S., Alexander, E.G., Jr. and Alexander, S.C. 1989. Comparison of
tracer mobilities under laboratory and field conditions. Journal of Environmental Quality,
v. 18, p. 491-498.
Faure, G. 1986. Principles of Isotope Geology (2nd ed.). Wiley, New York, Chapters 22, 24,
and 26.
Field, M.S., Wilhelm, R.G., and Quinlan, J.F., in review. Use and toxicity of fluorescent dyes
for tracing groundwater. Submitted to Ground Water.
Ford, D.C., and Williams, P.W. 1989. Karst Geomorphology and Hydrology. Unwin Hyman,
N.Y. 601 p.
Friederich, H., and Smart, P.L. 1982. The classification of autogenic percolation waters in karst
aquifers: A study in G.B. Cave, Mendip Hills, England. University of Bristol
Spelaeological Society, Proceedings, v. 16, p. 143-159.
Ganz, C.R., Schulze, J., Stensby, P.S., Lyman, F.L., and Macek, K. 1975. Accumulation and
elimination studies of four detergent fluorescent whitening agents in Bluegill (Lepomis
Macrochirusl Environmental Science and Technology, v. 9, p. 738-744.
Caspar, E. 1987. Modern Trends i nTracer Hydrology. CRC Press, Boca Raton, Fla. 2 v. 145
and 137 p.
Gunn, J. 1986. A conceptual model for conduit-flow dominated karst aquifers. International
Symposium on Karst Water Resources (Ankara, 1985), Proceedings, p. 587-596.
-36-
-------
Hallberg, G., Libra, R.D., and Hoyer, B.E. 1985. Nonpoint source contamination of ground
water in karst-carbonate aquifers in Iowa, in Perspectives in Nonpoint Source Pollution.
W.S. Environmental Protection Agency, EPA 440/5-85-001, p. 109-114.
Hubbard, E.F., Kilpatrick, F.A., Martens, L.A., and Wilson, J.F., Jr. 1982. Measurement of
travel time and dispersion in streams by dye tracing. U.S. Geological Survey, Techniques
of Water Resources Investigations. Book 3, Chapter A9. 44 p.
Jones, W.K. 1973. Hydrology of limestone karst in Greenbriar County, West Virginia. West
Virginia Geological and Economic Survey, Bulletin 36. 49 p.
Jones, W.K. 1984b. [1986]. Analysis and interpretation of data from tracer tests in karst areas.
NSS Bulletin, v. 46, no. 2, p. 41-47.
Lang, H., Liebundgut, C, and Festel, E. 1979. Results form tracer experiments on the water
flow through the Aletschgletscher. Zeitschrift fur Gletscherkunde und Glazieal-geologie.
v. 15, p. 209-218.
Lyman, F.L., Schultze, J., Ganz, C.R., Stensby, P.S., Keplinger, M.L. and Calendra, J.C.
1975. Long-term toxicity of four fluorescent brightening agents. Food and Cosmetic
Toxicology, v. 13, p. 521-527.
Maloszewski, P., and Zuber, A. 1990. Mathematical modeling of tracer behavior in short-term
experiments in fissured rocks. Water Resources Research, v. 26, p. 1517-1528.
Maloszewski, P., and Zuber, A. 1992. On the calibration and validation of mathematical models
for the interpretation of tracer experiments in groundwater. Advances in Water Resources.
v. 15, no. 1, p. 47-62.
Morton, H.E. 1983. Pseudomonas. la: Block, S.S., ed. Disinfection, Sterilization, and
Preservation (3rd ed.), Lea & Febiger, Philadelphia, p. 401-411.
Mull, D.S., Lieberman, T.D., Smoot, J.L, and Woosley, L.H., Jr. 1988. Application of dye-
tracing techniques for determining solute-transport characteristics of groundwater in karst
terranes. U.S. Environmental Protection Agency, Atlanta. EPA 904-/6-88-001. 103 p.
Parker, M.T. 1983. Pseudomonas. la: Principles of Bacteriology, Virology, and Immunity (7th
ed), Williams & Wilkins, Baltimore, v. 2, p. 246-252.
Quinlan, J.F. 1989. Ground-Water Monitoring in Karst Terranes: Recommended Protocols and
Implicit Assumptions. U.S. Environmental Protection Agency, Environmental Monitoring
Systems Laboratory, Las Vegas, EPA/600/X-89/050. 87 p.
Quinlan, J.F., Davies, G J., and Worthington, S.R.H. 1992. Rationale for the design of cost-
effective groundwater monitoring systems in limestone and dolomite terranes: Cost-
effective as conceived is not cost-effective as built if the system design and sampling
frequency inadequately considers site hydrogeology. Waste Testing and Quality Assurance
Symposium (8th, Washington, D.C.), Proceedings. U.S. Environmental Protection
Agency, Washington, D.C., p. 552-570.
Quinlan, J.F., and Ewers, R.O. 1985. Groundwater flow in limestone terranes: Strategy
rationale and procedure for reliable, efficient monitoring of ground water quality in karst
area, national Symposium and Exposition on Aquifer Restoration and Ground Water
Monitoring (5th, Columbus), Proceedings, p. 197-234.
Quinlan, J.F., Ray, J.A., and Fehrman, J.G. 1990. Groundwater Remediation may be achievable
in some karst aquifers that are contaminated, but it ranges form unlikely to impossible in
most n. Implications for the Mammoth Cave area of long-term tracer tests and universal,
nationwide failure in goal attainment by scientists, consultants, and regulators (abs.).
Mammoth Cave Karst Research Conference, Proceedings, (in press)
Quinlan, J.F., and Ray, J.A. 1992. Ground-water remediation may be achievable in some karst
aquifers that are contaminated, but it ranges from unlikely to impossible in most: I. Impli-
cations of long-term tracer tests for universal failure in goal attainment by scientists,
consultants, and regulators. In: (Quinlan, J. and Stanley, A., eds.) Conference on
Hydrogeology, Ecology, Monitoring, and Management of Ground Water in Karst
Terranes, (3rd, Nashville, Tenn.), Proceedings, National Ground Water Association,
Dublin, Ohio, p. 553-558.
-37-
-------
Quinlan, J.F., and Rowe, D.R. 1977. Hydrology and water quality in the Central Kentucky
Karst: Phase I. University of Kentucky, Water Resources Research Institute, Research
Report no. 101. 93 p.
Rathor, M.N., Giblilaro, L.G., and B.A. Huffman. 1985. The hopping model for residence time
distributions of systems with splitting and merging streams. American Institute of
Chemical Engineers, Journal, v. 31, p. 327-329.
SDC & AATCC [Society of Dyers & Colourists and American Association of Textile Chemists],
1971-1982. Colour Index, 3rd ed. SDC & AATCC, Bradford. 7 v. 6460 pp.
Smart, C.C. 1983a. Hydrology of a glacierized alpine karst Castlefuard Mountains, Alberta.
Ph.D. dissertation (Geography), McMaster University. 343 p.
Smart, C.C. 1983b. Hyrology of the Castleguard karst, Columbia Icefields, Alberta, Canada.
Arctic and Alpine Research, v. 15, p. 471-486.
Smart, C.C. 1984 a. Glacier hydrology and the potential for subglacial karstification. Nork
geografiske Tidskrift v. 38, p. 157-161.
Smart, C.C. 1984b. Overflow sedimentation in an alpine cave system. Norsk Geografiske
Tidskrift v. 38, p. 171-175.
Smart, C.C., and D.C. Ford. 1982. Quantitative dye tracing in a glacierized alpine karst.
Beitraege zur Geologic der Schweiz - Hydrologie. v. 28, pt 1, p. 191-200.
Smart, P.L. 1972. A laboratory evaluation of the use of activated carbon for the detection of the
tracer dye Rhodamine WT. M.S. thesis (Geography), University of Alberta. 188 pp.
Smart, P.L. 1977. Catchment delineation in karst areas by the use of quantitative tracer methods.
International Symposium on Underground Water Tracing (3rd, Ljubljana-Bled, 1976),
Papers, v. 2, p. 198-291.
Smart, P.L. 1984 [1986]. A review of the toxicity of twelve fluorescent dyes used for water
tracing. NSS Bulletin, v. 46, no. 2, p. 21-33.
Smart, P.L., Atkinson, T.C., Laidlaw, I.M.S., Newsom, M.D., and Trudgill, ST. 1986.
Comparison of the results of quantitative and non-quantitative tracer tests for determination
of karst conduit networks: An example form the Traligill basin, Scotland: Earth Surface
Processes and Landforms. v. 11, p. 249-261.
Smart, P.L., and Brown, M.C. 1973. The use of activated carbon for the detection of the tracer
dye Rhodamine WT. International Speleological Congress (6th, Olomouc,
Czechoslovakia), Proceedings, v. 4, p. 285-292.
Smart, P.L., and Friederich, H. 1982. An assessment of the methods and results of water-tracing
experiments in the Gunung Mulu National Park, Sarawak. British Cave Research
Association, Transactions, v. 9, p. 100-112.
Smart, P.L. and Hodge, P. 1980. Determination of the character of the longwood sinks to
Cheddar resurgence conduit using an artificial pulse wave. British Cave Research
Association, Transactions, v. 7, p. 208-211.
Smart, P.L. and Laidlaw, I.M.S. 1977. An evaluation of some fluorescent dyes for water tracers.
Water Resources Research, v. 13, p. 15-33.
Smith, D.I. 1977. Applied geomorphology and hydrology of karst regions, in Hails, J.R., ed.
Applied Geomorphology. Elsevier, Amsterdam/New York. p. 85-118.
Vandike, I.E. 1985. Movement of shallow groundwater in the Perryville area, southeastern
Missouri. Missouri Geological Survey, Miscellaneous Publication no. 4. 53 p.
White WB. 1988. Geomorphology and Hydrology of Karst Terranes. Oxford, N.Y. 464 p.
Wilson, J.F., Jr., Cobb, E.D., and Kilpatrick, F.A. 1986. Fluorometric Procedures for Dye
Tracing. U.S. Geological Survey, Techniques for Water-Resource Investigations, Book
3, Chapter A12 (revised). 34 p. _
Woods, J. 1985. Residence times of water masses in regions of the ocean. Nature v. 314, p.
578-579
Worthington, S^R.H. 1991. Karst Hydrogeology of the Canadian Rocky Mountains. Ph.D.
thesis, Geography, McMaster University, 370 p.
-38-
-------
APPENDIX C
DETERMINATION OF THE RECHARGE AREA FOR THE RIO SPRINGS
GROUNDWATER BASIN, NEW MUNFORDVILLE, KENTUCKY, AN
APPLICATION OF DYE TRACING AND POTENTIOMETRIC MAPPING
FOR DELINEATION OF SPRINGHEAD AND WELLHEAD PROTECTION
AREAS IN CARBONATE AQUIFERS AND KARST TERRANES
-------
DETERMINATION OF THE RECHARGE AREA FOR THE RIO SPRINGS
GROUNDWATER BASIN, NEAR MUNFORDVILLE, KENTUCKY: AN
APPLICATION OF DYE TRACING AND POTENTIOMETRIC MAPPING
FOR DELINEATION OF SPRINGHEAD AND WELLHEAD PROTECTION
AREAS IN CARBONATE AQUIFERS AND KARST TERRANES
PROJECT COMPLETION REPORT
Prepared for
Ground-Water Branch
U.S. Environmental Protection Agency, Region IV
Atlanta, Georgia
by
Geary M. Schindel
ECKENFELDER INC.
227 French Landing Drive
Nashville, Tennessee 37228
(615) 255-2288
James F. Quinlan, Ph.D.
Quinlan and Associates
Box 110539
Nashville, Tennessee 37222
(615) 833-4324
and
Joseph A. Ray
Groundwater Branch
Kentucky Division of Water
18 Reilly Road
Frankfort, Kentucky 40601
(502) 564-3410
August 1994
-------
TABLE OF CONTENTS
Page No.
SUMMARY 1
INTRODUCTION 2
PROJECT PARTICIPANTS 3
BACKGROUND INFORMATION 3
PREVIOUS STUDIES 5
GEOLOGIC SETTING 6
Project Hydrologic Boundary and Estimated Boundary of Basin 8
SEARCH OF PUBLIC GROUND WATER DATA-BASES 8
Dye Trace Information g
Water Well Survey 9
Public Water-Supply Resources 10
Compilation of Data 11
FIELD INVESTIGATIONS H
General n
Spring Survey ii
Survey for Dye-Injection Locations 12
Sinkholes 12
Sinking Streams 12
Water Wells 13
Potentiometric Mapping 13
DESIGN OF TRACER TESTS 13
Selection of Monitoring Points 14
Selection of Dye-Injection Points 14
Selection of Dyes 14
Recovery of Dyes in the Field 15
Laboratory Analysis for Dyes 16
Documentation of Results 17
Glen Lilly Road Spring to Buckner Spring Cave 17
Charles Ash Sinkhole to Jones School Spring 18
Glen Lilly Sinkhole to Rio Springs 18
Bail Road Ditch Sinkhole to Bailey Falls Spring 18
Knox Creek Sinkhole 18
Christene Dye Well to Johnson Spring 18
Walter Well 18
Route 357 Sinkhole 19
Unrecovered Dyes 19
-------
TABLE OF CONTENTS (Continued)
Page No.
CONCLUSIONS AND DISCUSSIONS 20
REFERENCES 23
APPENDICES
Appendix A - Kentucky Ground-water Tracing Forms
11
-------
DETERMINATION OF THE RECHARGE AREA FOR THE RIO SPRINGS
GROUNDWATER BASIN, NEAR MUNFORDVILLE, KENTUCKY: AN
APPLICATION OF DYE TRACING AND POTENTIOMETRIC MAPPING
FOR DELINEATION OF SPRINGHEAD AND WELLHEAD PROTECTION
AREAS IN CARBONATE AQUIFERS AND KARST TERRANES
SUMMARY
Dye traces and a potentiometric map based on water-wells, spring, and stream
elevations were used to delineate the Rio Springs groundwater basin located east of
Munfordville, Kentucky. This investigation was performed as a Springhead
(Wellhead) Delineation Demonstration Project supported by the U.S. Environmental
Protection Agency, Kentucky Division of Water, and the Green River Valley Water
District. The results of the series of dye traces were used to iteratively revise the
potentiometric maps that guided the design of successive trace tests. The rationale
for various investigative techniques used and decisions made is included in this
report.
The recharge area for the Eio Springs groundwater basin is approximately
4.9 +0.5 square miles and is shown in Plate I. The area includes groundwater
drainage from an adjacent surface-water basin, Bacon Creek. Such inclusion is
inferred because the boundary of the Rio Springs groundwater basin is beyond (and
outside) the boundary of its surface water basin (the topographic divide) where this
latter boundary can be drawn.
The long-term, sustained flow of the Rio Springs groundwater basin (its normalized
base flow), as measured by its base flow discharge per square mile, is five to six
times greater than that measured in any other groundwater basin in the Mammoth
Cave area. This significantly greater sustained flow is a response to attenuation
(damping) of aquifer response to storm-induced recharge—attenuation caused by
thick masses of slumped sand and gravel that overlie most of the Rio Springs basin.
The hydrogeologic properties of the sand and gravel increase the time it takes for
the aquifer to respond to storms. They impart a storage that is significantly higher
than that of nearby karst aquifers which lack a thick granular, non-clay mantle
above the carbonate bedrock.
-------
Normalized base flow can be used to reliably estimate the recharge area of a spring
but only if its probable geology is already known. This principle, quantified during
this investigation, should be widely applicable elsewhere.
The results of this investigation may be used for response to environmental
emergencies, local and regional planning, resource protection through a Springhead
Protection Program for the Rio Springs area, and public education.
Many results of this delineation project are relevant to the study, interpretation,
and protection of water supplies in other karst terranes. These results, plus
conclusions applicable to maximizing the efficiency and reliability of similar
investigations elsewhere, are discussed.
INTRODUCTION
This report describes the findings of a hydrogeologic study of the Rio Springs
groundwater basin, east of Munfordville, Kentucky. Rio Springs is a raw water
source for the Green River Valley Water District (GRVWD). This study was
conducted to define the area of recharge as part of a Karst Springhead (Wellhead)
Protection Demonstration Project for public water supply springs. This
hydrogeologic study was designed to delineate the recharge area of the Rio Springs
groundwater basin. Tasks included the collection and review of background
information, determination of physical setting, field reconnaissance, tracer-test
design, tracer testing, and report generation.
A spring and the conduit network draining to it can be considered as a near-
horizontal well. It follows, then, that springhead is the spring-equivalent of a
wellhead.
This project completion report is written as a tutorial for technical personnel and
others who may be considering establishment of a wellhead protection program in a
carbonate rock terrane elsewhere. Accordingly, we have included background
information to give perspective, and discussion of the rationale for why many
decisions were made. However, this is not a "how-to" manual for tracing delineation
-------
of wellhead protection in non-carbonate rocks, or for organization of a wellhead
protection program. The latter two topics are well described by U.S. EPA (1987,
1989).
Although many wellhead protection studies routinely include analysis of fracture-
traces and lineaments as a guide to flow direction and routing of groundwater, such
a study would have been irrelevant in the Rio Springs area and was not performed.
In spite of the excesses described by Wise (1982, 1983), these features have been
repeatedly shown to be a guide to the siting of highest-yield wells, and they do
indicate the most easily recognized possible flow routes, but they are not a predictor
of flow destination (Blavoux et al., 1992), or of major flow routes in carbonate
aquifers. Most flow in carbonate aquifers of the Mammoth Cave/Rio Springs area is
in conduits developed near-parallel to bedding planes rather than along joints.
Accordingly, although much orientation data could have been acquired, analyzed,
and presented, we did not consider fracture-trace and lineament analysis to be
judicious, cost-efficient, or relevant to springhead delineation in the study area.
PROJECT PARTICIPANTS
Project participants included ECKENFELDER INC., Nashville, Tennessee;
Quinlan & Associates, Nashville, Tennessee; Groundwater Branch, Kentucky
Division of Water, Frankfort, Kentucky; the Green River Valley Water District,
Cave City, Kentucky; and ATEC Associates Inc., Nashville, Tennessee. Funding for
this project was supplied by the Ground Water Branch, U.S. EPA, Region IV,
Atlanta, Georgia. The Kentucky Division of Water and the Green River Valley
Water District also contributed personnel to work on this project. Most of the field
work for this project was carried out by Joseph Ray, Kentucky Division of Water;
Geary Schindel, ECKENFELDER INC.; and Tray Lyons, Green River Valley Water
District. Robert Olive, Environmental Scientist, USEPA, Ground Water Branch
was Project Coordinator for the EPA.
BACKGROUND INFORMATION
The GRVWD supplies more than 25,000 people with rural service connections in
Hart County and portions of Green, LaRue, Barren, Metcalfe, and Edmonson
-------
Counties. It supplies water directly to Mammoth Cave National Park and to the
cities of Horse Cave and Cave City, to the LaRue County Water District
(2,627 people), the Green-Taylor Water District (8,751 people), Bonnieville Water
District (752 people), and the Munfordville Water District (2,627 people). Rio
Springs is also the water source for the Glenbrook Trout Farm, located below the
Rio Springs reservoir.
Rio Springs consists of several contiguous springs on the north side of the Green
River, near the former community of Rio in Hart County, Kentucky. The springs
are approximately one-half mile west of U.S. Highway 3IE and one-quarter mile
north of the Green River. All perennial flows are on the west bank of a small south-
flowing intermittent tributary of Rocky Hollow. None of the springs are shown on
the U.S. Geological Survey Canmer, Kentucky, 7.5-minute topographic map, but
they are shown on Plate I of this report as Sites 6 and 7.
The stream which flows from the springs has been dammed with a concrete
structure to form a small reservoir covering less than one acre. The natural spring
orifices were backflooded by the reservoir and aggraded by sand. This reservoir is
presently fed by seven perennial springs located on the west side of the ravine.
Three of these flows comprise the major portion of Rio Springs. During high-flow
conditions, a 0.6-mile long intermittent stream, draining south from the community
of Linwood, conveys surface water to the reservoir. The total discharge of Rio
Springs is approximately 4.8 cubic feet per second, as averaged from two
measurements reported by the U.S. Geological Survey (Donald S. Mull, oral
communication, June 1993). A subsequent gauging on September 3, 1993
determined a discharge of 4.4 cubic feet per second.
The Green River Valley Water District reports an average use of 500,000 gallons of
water a day from Rio Springs and approximately 1,500,000 gallons a day from the
Green River. However, during some months of the year, no water from Rio Springs
is used by the GRVWD. The District would prefer to use additional water from Rio
Springs because it requires less treatment than water from the Green River.
However, there is a conflict regarding allocation of water between the GRVWD and
the trout farm, which also requires a high-quality water supply.
-------
Map coverage of the study area is available on four 7.5-xninute, 1:24,000 scale
topographic and geologic maps published by the U.S. Geological Survey: the
Canmer, Hammonville, Hudgins, and Magnolia quadrangles. Each of the
topographic and geologic maps was reviewed for the presence of surface streams,
sinking streams, springs, caves, and other karst features. Most of the Canmer
quadrangle and a small part of each of the other three topographic maps were
assembled into a working project map which was photographically reproduced and
used to plot all points possibly suitable for tracer injection and monitoring.
PREVIOUS STUDIES
Although there are no previous hydrogeologic studies of the Rio Springs basin,
extensive investigations were conducted in the area adjacent to Rio Springs by
James F. Quinlan and Joseph A. Ray when each was employed by the National Park
Service at Mammoth Cave National Park. Those investigations north of the Green
River were not completed and have not been published. Their work on groundwater
basins south of the Green River has been published (Quinlan and Ray, 1989).
Available information from the following organizations was also reviewed: Kentucky
Division of Water, Groundwater Branch; Green River Valley Water District; U.S.
Geological Survey, Kentucky District; and Kentucky Geological Survey. A map of
Buckner Spring Cave was graciously provided by Dr. Joseph Saunders.
Data on dye traces in the Johnson Spring and Lanes Mill Spring groundwater
basins, adjacent to the Rio Springs groundwater basin, were obtained from James F.
Quinlan. Discharge and water-quality data for Rio Springs, plus well location and
water-level data on 14 of the 61 wells shown on Plate I, were obtained from the U.S.
Geological Survey. The Kentucky Division of Water supplied available records of
water wells drilled after 1985. The National Park Service supplied water-well
locations and water-level data for wells drilled before 1986. James F. Quinlan
supplied information on the location of numerous springs found by him and
Joseph A. Ray during pre-1986 studies.
The most recent syntheses of regional hydrogeology of the karst have been
published by White and White (1989) and Quinlan et al. (1983) but neither of these
original works specifically address the Rio Springs area.
-------
A published potentiometric map at a scale of 1:250,000 (about 1 inch = 4 miles)
includes the area of Plate I (Plebuch et al., 1985), but it was contoured at a 50-foot
interval. For the Mammoth Cave area south of the Green River, it reproduced the
1981 version of Quinlan and Ray (1989), but partially recontoured at a 50-foot
interval. [The Quinlan and Ray map averaged about 100 wells per quadrangle
(approximately two per square mile) and had been published with a 20-foot contour
interval.] For most of the remaining coverage of the Plebuch et al. map, there are
significantly fewer wells measured per quadrangle, commonly less than 15, than
were used in this report. Also, no springs are shown on it. Accordingly, the Plebuch
et al. map can be used for only very general predictions. It does not include, nor can
it be used to determine, boundaries of groundwater basins, or for response to
environmental incidents.
GEOLOGIC SETTING
The Rio Springs groundwater basin, located in west-central Kentucky near the
southeastern edge of the Illinois Basin, is in Mississippian-age limestones overlain
by Mississippian and Pennsylvanian sandstones. The rocks throughout most of the
map area shown in Plate I dip gently to the west at about 20 to 50 feet per mile.
The north edge of Plate I coincides approximately with the axis of an anticline
extending to the east; the anticline is used as the Magnolia Gas Storage Field
(Moore, 1975). The stratigraphic units in the study area, from oldest to youngest,
include the Salem-Warsaw, St. Louis, Site. Genevieve, and Girkin Limestones, the
Big Clifty Sandstone, and the Caseyville and Tradewater Formations (mostly
sandstone and conglomerate highly weathered to sand and gravel).
The Rio Springs area is a karst terrane. It is characterized by sinkholes, sinking
streams (most of which are ephemeral), caves, springs, and a well-integrated
subsurface drainage network. Much of the study area is a highly dissected part of
the Mammoth Cave Plateau (Dicken, 1935), which is also known as the Chester
Cuesta (Quinlan, 1970). The northern half of Plate I includes a sandy terrane that
is known as the Brush Creek Hills (Sauer, 1927). The sand and gravel is part of the
west-southwest-trending Brownsville Channel, which occupies a paleo-valley up to
several hundred feet deep that is filled with Pennsylvanian sandstone, shale, and
-------
conglomerate that unconformably overlie several of the Mississippian lime stone and
sandstone formations. Much of this has been intensely weathered, disaggregated,
and lowered during dissolutionally-induced subsidence.
All land south of the Green River is part of the Sinkhole Plain that, at Sims Bend
and Davis Bend, extends up to 3 miles north of the river. All of Plate I is underlain
by the same relatively pure limestones that crop out in the Mammoth Cave area
and which are locally capped by the Big Clifty Sandstone.
A geologic map has been published for each of the four topographic map
quadrangles listed above (Miller, 1969; Miller and Moore, 1969; Moore, 1972, 1975).
These maps were spliced together and interpreted in order to determine what
relationships may exist between stratigraphy, structure, and the distribution of
springs.
Extensive field observations in this area, coupled with interpretation of published
geologic maps, have shown that there are three lithologic controls on groundwater
movement in the karst of the Rio Springs area. They are:
1. Impermeability of the Big Clifty Sandstone and associated shale. This
locally preserves the caves below from erosion and dissolutional
destruction, but favors the development of vertical shafts that help
accomplish such destruction at the edge of ridges. The impermeability of
the Big Clifty is much less important in the study area where the ridges
are narrower and more highly dissected than to the west, where the beds
dip slightly more steeply, dissection is less, and the ridges are wider.
2. Impermeability of the clayey, silty limestones at the top of the Salem-
Warsaw Limestone and lowermost part of the overlying St. Louis
Limestone, about 40 feet above their contact. Rio Springs and several
other springs appear to be perched on these upper beds. Many additional
springs are perched on the top of the Salem-Warsaw.
3. Recharge attenuation and storage capacity of up to several hundred feet
of slumped sand and gravel (disaggregated sandstone and conglomerate)
-------
that overlie the limestones in the northern half of Plate I. This
widespread sand and gravel impede rapid or direct recharge into the
aquifer at sinkholes; limestone outcrops are uncommon. As a result,
many of the springs have a more subdued response to storms and lower
turbidity than those appreciably fed by sinking streams and sinkholes
draining into open holes in limestone.
Project Hydrologic Boundary and Estimated Boundary of Basin
The preliminary project hydrologic boundary for Rio Springs and immediately
adjacent groundwater basins was determined from review of the topographic and
geologic maps and from interpretation of previous water-tracing studies of the
region conducted by Quinlan and Ray. It was defined as the major surface and
subsurface streams adjacent to Rio Springs which most probably act as a boundary
for near-surface groundwater flow in the region. This boundary was identified as:
Green River, to the south; Lynn Camp Creek, to the east; Laurel Branch, Brushy
Fork, and Bacon Creek, to the north; and the inferred southward flowline of the dye
trace from near Bolton Church to Johnson Spring, to the west. Some of these
streams were known to be beyond the estimated boundary of the actual
groundwater basin.
Stream incision along the Green River and the lower part of Lynn Camp Creek has
exposed clayey, silty beds that perch springs at and near the contact between the
St. Louis, and Salem-Warsaw. These beds and the top of the Salem-Warsaw were
considered to be the basal hydrologic boundary of the near-surface aquifer.
Interpretation of the map of Buckner Spring Cave (the spring location shown on the
published topographic map is about 120 feet above its actual elevation), plus the
pre-project trace to Lanes Mill Spring, suggested that the actual boundary of the
Buckner Spring Cave basin would be closer to Rio Springs. The orientation of
Buckner Spring Cave led its mapper, Dr. Joseph Saunders, to hypothesize that the
cave functioned as a high-level overflow for Rio Springs. Such a distributary could
exist, but it was not demonstrated during this study.
8
-------
SEARCH OF PUBLIC GROUNDWATER DATA-BASES
Dye Trace Information
The Kentucky Division of Water and the Kentucky Geological Survey had no records
of dye traces in the study area. Karst researchers known to be actively working in
this region of Kentucky were contacted for background information and to
determine if any dye tracing was currently being conducted in the area; none was.
The unpublished tracing studies by Quinlan and Ray are the only previous work
known in the area.
Water Well Survey
The Kentucky Division of Water's Water-Well Drillers Program, supplied a copy of
logs from two wells drilled within the area of Plate I since 1985. One well was
inaccessible for measuring of water level; the other could not be found. A copy of
unpublished data on well locations and water levels, based on pre-1986 field work
by Quinlan and Ray and inclusive of data from the U.S. Geological Survey, was
obtained from the files of the National Park Service. These data had been used by
them to construct a draft potentiometric map of the region north of Green River, but
more well data were needed for the Rio Springs area. That draft map guided much
of the field reconnaissance for this project and was basically correct, but its contours
were repeatedly modified after additional water-level data and tracing results were
acquired and interpreted.
Experience has repeatedly shown that, unless a locally intensive well survey has
already been made, state and federal records in many states rarely include more
than about 10 percent of the wells that exist in an area. Accordingly, most of the
well survey was performed by conducting a house-to-house quest for wells. The
pre-1986 survey north of the Green River was part of a research program. Field
work in that study yielded 0 to 12 measurable wells per day and averaged about 5,
but this was a function of local stratigraphy and population density. Daily
productivity of water-level data was higher in the Sinkhole Plain to the south.
-------
When a measurable well was found, project-relevant data was recorded. Location
and ground-surf ace elevation was estimated from the topographic map, and water
level was measured with an electric tape. Data from all measurable wells was
collected during low-flow conditions. Some of the wells first measured by the U.S.
Geological Survey were remeasured in order to confirm the static conditions.
U.S. Geological Survey national mapping standards require that 90 percent of the
elevations on a topographic map have a error of no more than half a contour
interval. The practical application of this standard is ambiguous, however, unless,
one has information on the statistical distribution of the errors contributory to
meeting or failing the standard. But if it is assumed that: 1) the maximum possible
error is half a contour interval, and 2) the study area is in a low relief, non-forested
terrane such as the Sinkhole Plain where most homes and barns are shown on the
topographic map, the elevation of a well can be estimated to about one-fifth of a
contour interval. The two sources of error can be added to calculate the maximum
error. Therefore, where the contour interval is 10 feet, the elevation of most wells
can be estimated to within + 7 feet; where the contour interval is 20 feet, twice as
much.
For most of the area of Plate I, the slope of the potentiometric surface ranges from
about 40 to 100 feet per mile. Therefore, the elevation "noise" on the potentiometric
surface in the study area (± 7 or 14 feet, depending upon the contour interval) does
not greatly affect the accuracy of the surface being contoured.
The effects of possible error induced by some of the "topography" having been locally
contoured on vegetation canopy rather than on the ground surface has not been
evaluated by us. It could be a problem in some of the more densely wooded parts of
the study area.
The potentiometric surface in Plate I was contoured manuaDy rather than with a
computer program. The surface is subjective and was revised as tracing data
became available. The working maps with the revised potentiometric contours and
tracer-test results were used to guide planning and interpretation of additional
tracer tests.
10
-------
Public Water-Supply Resources
The Kentucky Division of Water's Water Withdrawal Program, furnished
information on three water-withdrawal permits for the area: The Green River
Valley Water District is permitted to remove approximately 2.6 minion gallons per
day (mgd) from Rio Springs and the Green River; the Glenbrook Trout Farm is
permitted to use 1.446 mgd from Rio Springs; and the Powder Mill Trout Ranch,
Inc., adjacent to Lynn Camp Creek and near Sites 14 and 15, near the eastern edge
of Plate I, used an average of 0.735 mgd in 1992.
Compilation of Data
Data collected as part of the background study for this project were compiled onto a
working map produced from the four topographic maps. These data included the
injection and recovery points for the previous dye traces performed by Quinlan and
Ray, locations and depths to water in domestic water wells, and the locations of
springs near and within the estimated hydrologic boundary of the Rio Springs
basin. These data were used to construct a potentiometric surface map and to
identify areas where additional field work was required.
FIELD INVESTIGATIONS
General
Field reconnaissance was made of the area near the estimated hydrologic boundary
not checked during the earlier studies by Quinlan and Ray. Most of this field work
centered on the northern part of the map area and included Martis Branch, Tampa
Branch, Laurel Branch, and Brushy Fork. Twenty-eight locations (19 springs and
9 streams) were initially identified for the placement of dye receptors (detector).
One spring outside of the estimated hydrologic boundary, Handy Culvert Spring
(Site No. 13 on Plate I) was monitored as a quality control procedure. It drained
from the east.
11
-------
Spring Survey
A spring survey was conducted in order to find springs at or within the preliminary
hydrologic boundary of the study area. Springs identified during the uncompleted
pre-1986 field work by Quinlan and Ray were incorporated into this survey.
Additional springs were also located during field work for this project.
Only 4 of the 19 springs monitored during this study are shown on the U.S.
Geological Survey topographic maps. An additional spring (James School Spring,
Site No. 21) is shown on a geologic map but not on the corresponding topographic
map. All other springs were found as a result of field work. Only 1 of the
19 springs found in the area may have been detectable on aerial photos.
[Throughout the Mammoth Cave area and most of the U.S., generally less than
5 percent of springs relevant to regional hydrology of karst terranes are shown on
topographic maps.]
Survey for Dye-Injection Locations
Field reconnaissance for dye-injection locations, conducted near the estimated
hydrologic boundary of the Rio Springs basin, was done concurrently with the
spring survey. An attempt was made to identify sinkholes, sink-points (swallets) of
sinking streams, and water wells that might be usable for injecting dye into the
aquifer. Suitable dye-injection sites are rare in the study area; locating them
required a significant field effort.
Sinkholes. Sinkholes with an opening through which water might drain readily,
especially after storms, were sought during field reconnaissance and plotted on the
working map. Many sinkholes drain runoff only after heavy storms. Therefore,
each potential dye-injection site had to be evaluated for its accessibility by tank-
trucks or other sources of water. All sinks judged to have potential as dye-injection
points were also evaluated for their proximity to the estimated boundary between
the Rio Springs basin and adjacent groundwater basins. A site for potential
injection of dye could be technically exceUent, but if it were near the probable
middle of the Rio Springs basin rather than near its boundary, tracing from it was
considered unnecessary.
12
-------
Sinking Streams. Although there are many sinking streams in the area shown on
Plate I, fewer than five of those shown as such on the 7.5-minute topographic maps
occur within or adjacent to the estimated boundaries of the Rio Springs basin.
These few are all ephemeral, flowing only after major storms, and they are all in
locations either near the probable middle of the basin or obviously, on the basis of
pre-project tracer tests, draining to other basins. Other ephemeral sinking streams
exist within or adjacent to the Rio Springs basin but are not shown on the
topographic maps.
One perennial sinking stream was identified for which the sink-point shifts,
depending on stage height. Extensive field work was done in order to locate
ephemeral streams that convey stormwater runoff to discrete sink-points. The few
perennial and ephemeral sink-points found were plotted on the working map and
evaluated for location relative to the estimated groundwater boundary of the Rio
Springs Basin. The sink-points were further evaluated for their ease of access by a
tank-truck or other source of water.
Water Wells. Several unused water wells were found during pre-project
investigations by Quinlan and Ray. High priority was given during this study to
finding additional unused wells suitable for injection of dye because of their
accessibility and their location relative to the tentatively inferred boundary of the
Rio Springs basin. Landowners were extremely cooperative. The presence of
unused water wells was generally a consequence of the availability in recent years
of public water from the Green River Valley Water District.
Potentiometric Mapping
There are fewer domestic water wells per square mile in the Rio Springs area than
in most of the area south of Green River previously studied by Quinlan and Ray
(1989). In part, this is because of the lower population density near Rio Springs,
and because the thickness and loose nature of slumped sands reportedly makes it
difficult for local drillers to complete a well successfully without it collapsing or
producing excessive sand. An intensive house-to-house search for additional wells
was made, but only five more were found.
13
-------
DESIGN OF TRACER TESTS
Information obtained from the background study and field reconnaissance was used
to design the tracer tests. Evaluation of these data indicated that multiple-dye
traces, using up to four dyes for each series of tests, could be conducted
simultaneously. The use of multiple dyes allowed for greater cost efficiency in
collection, analysis, and evaluation of dye receptors (detector). Data from each
series of dye tests were evaluated and additional dye-injection sites were selected.
After each series of tracer tests, the location and number of monitoring sites were
evaluated for their relevance to the study objectives.
Selection of Monitoring Points
Springs and streams were evaluated as potential monitoring sites for the tracer
tests. Major springs were individually monitored with dye detectors. In areas
where no major springs could be found, streams were monitored instead. After
careful evaluation and in an effort to reduce costs, some springs were monitored at
their confluence or in streams. If dyes were to be recovered, there would be time to
place dye detectors in individual springs before dye cleared from the system.
Twenty-eight locations, including 19 springs, were initially identified for the
placement of dye detectors. Additional stream sites were added after the tests
began. Plate I shows the name and location of all monitoring points used.
Selection of Dye-Injection Points
Approximately 15 potentially usable dye-injection points were identified in the
study area. These included sinkholes, sinking streams, and water wells. Tracer-
test injection points selected for use were those considered likely to yield access to
the aquifer and to be near the suspected boundary of the Rio Springs basin, as
inferred from the working draft of the potentiometric map.
14
-------
Selection of Dyes
Four types of fluorescent dye were selected for use in this study. They were picked
on the basis of non-toxicity, availability, analytical detectability, low cost, and ease
of use. All of these dyes have been previously used as tracers and their properties
have been documented in the karst-related literature. The following dyes were
used:
Dye Colour Index Generic Name
Rhodamine WT Acid Red 388
Fluorescein (Uranine) Acid Yellow 73
Solophenyl Direct Yellow 96
Optical Brightener
Burcofluor AF Solution Fluorescent Brightening Agent 28
Tinopal CBS-X Fluorescent Brightening Agent 351
The quantity of dyes used in this study was based on the experience of the authors
in similar terranes. Factors evaluated in determining those quantities include:
detection limit of the analytical method to be used for dye analysis, a desire not to
induce visible coloration to spring waters or streams, and a desire not to "overload"
the aquifer with dye that would persist for a much longer time than if a minimal
quantity was employed (thus delaying completion of the project). The desire not to
induce visible coloration to waters was a matter of aesthetics and public relations,
not possible toxicity. All of the dyes used are non-toxic (non-carcinogenic, non-
mutagenic, non-tumorogenic, non-teratogenic, non-poisonous, etc.), especially in the
concentrations to which they were diluted and discharged at springs (Smart, 1984;
Field et al., in review), and posed no threat to the quality of private or public water
supplies. Non-toxic, fluorescent dyes rather than other tracing agents are used
because they are safe, practical, most cost-efficient, and most easily detected tracers
available.
Recovery Of Dyes in the Field
The rationale and the techniques for conducting tracer tests and methods for the
analysis of dyes are discussed by Alexander and Quinlan (1992). They are
summarized briefly in this report.
15
-------
Dye detectors, consisting of either granular activated charcoal or non-fluorescent
cotton, both of which sorb dye, are used in lieu of water samples for two reasons:
economy and enhancement of dye concentration. More specifically, detectors yield
an integrated sample, that barring interference from other organic compounds, is a
product of continual sorption of dye, whenever dye is present in water. Therefore
sampling can be weekly or biweekly rather than hourly or daily. Some tracer tests
require, for various reasons, quantification of frequently collected water samples.
However, most projects, including this one, are only interested in determining if dye
was "present or absent" from a monitoring site. The consequent cost-savings in
time, labor, materials, and analysis can be considerable. Further, the amount of
dye accumulated on a detector increase with time. Charcoal, for example, when left
for a week in a spring or stream that has had a constant concentration of dye, will
commonly yield an elutant that has a dye-concentration of 100 to 400 times greater
than ever present in the stream. Therefore, use of detectors rather than water
samples enables one to sense the presence of dye that might not be detectable in a
water sample.
Optical brighteners were detected with the use of non-fluorescent cotton which had
been checked with an ultraviolet light before use. Both the cotton and charcoal
were placed in nylon-screen bags and suspended in water from a wire attached to a
concrete stand (gum drop) or wired brick. Detectors were generally changed once a
week. However, the collection of detectors was dependent upon weather and access.
A longer period of time between detector collections was used during the last part of
the investigation.
Detectors were set, collected, analyzed, and evaluated for background
concentrations of dye or dye-like substances over a period of several weeks before
any dyes were put into the ground. Once it was established that there was no
background, or that the background present was manageable, the final decisions
could be made about what dye would be most suitable for tracing to a given site.
All sites at which dye was detected were monitored for its continued presence until
the dye was either no longer detectable or was present at a concentration so low as
to not interfere with the interpretation of any subsequent tracer tests. This allowed
16
-------
time for the dye to be possibly detected at additional sites if connections existed.
Data on the frequency and duration of monitoring for each site, along with what
dyes it was tested for, and whether they were present, are summarized in
Appendix A.
Laboratory Analysis for Dyes
Each detector was placed in an individually marked bag in the field and shipped to
the laboratory for analysis. All detectors were thoroughly washed with a high-
intensity jet of tap water before being analyzed. Cotton detectors were evaluated by
using an ultraviolet light over a dark, light-proof box. Cotton that was positive for
brightener fluoresced a brilliant blue-white. Cotton that was positive for Solophenyl
fluoresced canary yellow. The results of detector evaluation were recorded on a
Tracer Test Form by date (see Appendix A). Charcoal detectors were evaluated by
eluting them for one hour in a solution containing 95 percent of a 70 percent
solution of isopropyl alcohol in water and 5 percent of ammonium hydroxide. The
elutant was then decanted for storage in a closed, labeled glass vial until analyzed.
Laboratory analyses for fluorescein were conducted with a Turner Designs Model 10
filter fluorometer by ECKENFELDER INC. Analysis for Rhodamine WT was
performed with a Turner Associates Model 111 filter fluorometer by
ECKENFELDER INC. Analyses for both dyes were performed with a Shimadzu
RF-540 scanning spectrofluorophotometer by Quinlan & Associates. Although it
was not critical for this study, because the dye concentrations in elutant were not
minimal and were well above detection limits, a scanning spectrofluorophotometer
is the optimal instrument for dye analysis because it can unambiguously detect
smaller concentrations of dye than a filter fluorometer and can readily and
unambiguously separate three or more dyes used simultaneously. More
importantly, a scanning spectrofluorophotometer can unambiguously distinguish
between dyes and non-dyes that may have fluorescence which overlaps that of dyes.
A filter fluorometer is an optimal instrument where one or more dyes are to be
detected in a setting not likely to have industrial contaminants that may fluoresce,
when fluorescein and Rhodamine WT are recovered in sub-equal quantities, or when
the concentration of one dye does not exceed that of the other by more than a factor
of about 20.
17
-------
Documentation of Results
Results of the tracer tests are shown in Plate I. The data sheets supporting it are
included in this report as Appendix A, which is comprised of Spring Survey Forms
and Tracer Test Forms.
The following summaries state where and when dyes were injected and recovered,
and for how long they were recovered. For details of what other sites were
monitored, for when and for how long all sites were monitored, and for how long
dyes were subsequently detectable, see Appendix A.
Glen Lilly Road Spring to Buckner Spring Cave (Dye Trace A). Two pounds of
Rhodamine WT (20 percent solution) were injected into the Glen Lilly Road Spring
on March 3, 1993. This dye was first recovered on a detector collected at Buckner
Spring Cave (Site 5) on March 12, 1993. It was present for five weeks.
Charles Ash Sinkhole to Jones School Spring (Dye Trace B). Seven pounds of
fluorescein were injected into the Charles Ash Sinkhole on March 3, 1993. This dye
was first recovered on a detector collected at Jones School Spring (Site 21) on
March 12, 1993. It was present for four weeks.
Glen Lilly Sinkhole to Rio Springs (Dye Trace C). Six and one-half pounds of
Solophenyl were injected into the Glen Lilly Sinkhole on March 25, 1993. The
presence of dye was first indicated on a detector collected at Rio Springs and Rio
Springs East (Sites 6 and 7) on April 2, 1993. It was present for three weeks.
Bail Road Ditch Sinkhole to Bailey Falls Spring (Dye Trace D). Forty and
one-half pounds (4.5 gallons) of optical brightener were injected into the Bail Road
Ditch Sinkhole on March 25, 1993. This dye was first recovered on a detector
collected at Bailey Falls Spring on April 2, 1993. It was present for two weeks.
Knox Creek Sinkhole (Dye Trace E). Fifteen pounds (1.75 gallons) of optical
brightener were injected into the Knox Creek Sinkhole on April 24, 1993 and was
not detected. The trace was repeated on June 16, 1993 using 6 pounds of another
18
-------
optical brightener (as a powder). Again, the dye was not detected in samples
collected over a one-month period. The trace was repeated on July 16, 1993 with
3 pounds of fluorescein. This trace was repeated on March 19, 1994 with 5 pounds
of fluorescein. This trace was positive at Lanes Mill Spring on March 21, 1994.
Christene Dye Well to Johnson Spring (Dye Trace F). Eleven pounds of
fluorescein were injected into the Christene Dye [sic, owner's name] Well on
April 24, 1993. This dye was first recovered on a detector collected from Johnson
Spring on May 9, 1993. Prior to this test, a small quantity of fluorescein was
present as background at this spring, but the fluorescein recovered had
concentrations several orders of magnitude higher than the background, and the
progressive decrease in its concentration was characteristic of tracer-test results. It
was present for five weeks.
Walter Well (Dye Trace G). Twelve pounds of Rhodamine WT (20 percent solution)
were injected into the Walter Well on April 24, 1993. As of August 15, 1993, this
dye was not found by analysis of detectors collected on a weekly to bi-weekly basis
(see section on unrecovered dyes, below). It is assumed to have flowed to Johnson
Spring, as shown on Plate I.
Route 357 Sinkhole (Dye Trace H). Three pounds of Solophenyl were injected into
the Route 357 Sinkhole on April 24, 1993. This dye was never recovered (see section
on unrecovered dyes, below).
Unrecovered Dyes
Eight tracer tests were conducted, as shown on Plate I. Two of them, Walter Well
(Site G), and Route 357 Sinkhole (Site H), were unsuccessful.
The Walter well test (Site G) was considered during its design to be a difficult one.
It was thought likely that the dye might be injected into the slumped sand and
gravel above the limestone rather than into it directly. Flow through slumped sand
could result in a very long time of travel and sorption onto the formation matrix.
Dye could be diluted to below the detection limits of the analytical instruments
used. Alternatively, the dye might have been (or will be) discharged at one of the
19
-------
14 small springs shown down-gradient along Bacon Creek [their discharge ranges
from 0.05 to 0.2 cubic feet per second] or at other springs beyond the west edge of
Plate I. The fact that dye was not detected at the Wabash Bridge (Site 28) could be
a consequence of its dilution by Bacon Creek or non-discharge into it. Flow could
also be partially through sand to Jones School Spring (Site 21) or even to Rio
Springs (Sites 6 and 7), but, after six months, it was not yet detected in any spring.
Each, some, and all of these explanations could be correct. If the test were to be
repeated, a greater quantity of a different dye should be used.
The Route 357 Sinkhole test (Site H) was conducted simultaneously with that from
the Christene Dye well (Site F). The rationale for this test was that the project was
drawing to a close and that if Site F were to drain to Rio Springs, it would be
desirable to test the other side of the inferred boundary between sites F and H. But
there would not be time to do so. More dye should have been used but the results of
the test from Site F makes repetition of the Route 357 Sinkhole test unnecessary.
Site H most certainly flows to Johnson Spring (Site 2); the injection point at Site H
is bracketed on each side by flow to this spring.
In summary of this section, and in retrospect, Site G was a known gamble and was
conducted prudently. Site H was a technical and strategic decision and is now both
academic, obvious, and unnecessary. These tests were not failures; there just was
not sufficient time (or funding) in the project schedule to complete them by changing
methodology or reinjection. On a more positive note, all other tracer tests, both
those done as part of this project and those done before it, were conducted at
optimal locations that allowed accurate approximation of the actual boundary of the
Rio Springs groundwater basin.
CONCLUSIONS AND DISCUSSIONS
There are at least 16 important general conclusions that can be made relevant to
goal attainment, predictability and accuracy of results, logistics, and interpretation
of results of this project. Although some of these conclusions have also been made
concerning other projects in karst terranes of Kentucky and elsewhere, they have
been reinforced by the Rio Springs project and they are listed and discussed here.
These conclusions are:
20
-------
1. The position and configuration of the boundary of the Eio Springs
groundwater basin, shown in Plate I, is based primarily on inferences
from six successful tracer tests, interpretation of the configuration of the
potentiometric surface, and deductions from field observations of the
highest perennially flowing reaches of seepage-fed surface streams. To a
far lesser extent, in part because we are aware of the fallacy of using
negative evidence to prove anything, the boundary position is also
influenced by the non-recovery of dye in one unsuccessful tracer test that
is tentatively interpreted to have gone to somewhere other than to Rio
Springs. [The rationale for this interpretation is: If some of the more
than adequate amount of dye injected had reached Rio Springs, it would
have been easily detected there.] We would prefer to have conducted
additional tracer tests, especially within the basin, but budget and time
constraints limited the number of tests that could be performed.
2. Delineation of groundwater basins in karst terranes can be done. It
requires extensive field work, sometimes under difficult conditions. But
groundwater tracing yields a greater quantity of highly reliable data
concerning the flow direction, velocity, destination, and sometimes
routing of groundwater and pollutants in karst terranes than any other
technique for investigation.
3. Delineation of the Rio Springs groundwater basin was done empirically,
by tracing. It could only have been done by tracing . No analytical model
or computer model now available would have been capable of doing such
delineation with similar accuracy.
4. The Rio Springs groundwater basin has an area of 4.9 ±0.5 square miles.
This area was determined by using a compensating polar planimeter to
trace its boundary. This boundary is approximate and subjective, but
usable for planning, protection, and emergency response. The
+0.5 square mile error is estimated.
21
-------
5. Although many groundwater basins in carbonate terranes have been
properly delineated -- by tracer tests performed on each side of their
inferred boundary (see, for example, Quinlan and Ray, 1989) - and tracer
tests have been routinely employed by state agencies and various
consultants in the delineation of numerous wellhead and springhead
protection areas, we believe the Rio Springs Basin is the first springhead
protection area in the U.S. to have been delineated as such and to have a
published map showing the inferred relations between the boundary, the
path between dye-injection and dye-recovery points, and the
potentiometric surface.
6. The general dip within the Rio Spring basin is west (at about 20 to
30 feet per mile); flow within it is generally south, along the strike, as
shown in Plate I. Basins to the east of it flow up-dip or along the strike.
Those to the west generally flow down-dip and then along the strike.
Those to the north generally flow northwest to west, along the strike.
7. The distribution of data points makes it likely that the boundary of much
of the Rio Springs groundwater basin north of the 600-foot potentiometric
contour could be shifted east or west by up to 1,000 feet. Nevertheless,
we have attained a reasonably accurate representation of the probable
boundary and general flow within the Rio Springs basin. If this map
were to be used for evaluating potential threats to water quality in the
basin by a facility proposed within a mile beyond the boundary north of
the 600-foot potentiometric contour, site-specific additional tracing would
have to be done.
8. The inferred divide between groundwater draining south to the Green
River and groundwater draining north and west to Bacon Creek, a
tributary of the Nolin River, is locally more than two miles north of the
surface water (topographic) divide between the two streams. This
subsurface piracy of surface drainage is dramatically shown in the
western area of Plate I where three traces that went to the south were
injected significantly north of the surface-water divide, well within the
surface-watershed of Bacon Creek. Although the topographic map shows
22
-------
blueline streams draining northwest from the topographic divide, these
are actually intermittent waterways utilized only during storm events.
The Green River, because of the relatively steeper gradient to it, is
capturing groundwater from the flanks of Bacon Creek. [Unpublished
tracing results by Quinlan and Ray show this relationship even more
dramatically in the contiguous extensive area west of Plate I.] Rio
Springs includes drainage from beyond the northern surface-water divide
of its basin. This fact is extremely relevant to spill-response and
protection of water quality.
9. The basin boundary is shown by smooth curvilinear lines. In actuality,
the boundary could be irregular, even interdigitate, and it could
temporarily shift in response to storms and seasonal changes in static
water levels. The northern boundary of the Rio Springs groundwater
basin was inferred to be nominally midway between the topographic
divide (shown as a line of dots in Plate I) and the imaginary line formed
by linking the highest perennially flowing segment of the easternmost
five small steams (shown as arrow heads) draining to Bacon Creek.
These flowing segments are fed by groundwater. Therefore, barring
perching, the groundwater divide must be southeast of them. No attempt
was made to locate additional highest perennially flowing stream
segments west of State Highway 357.
10. Previous studies have shown that topographic maps, both in Kentucky
and generally, show less than five percent of relevant springs and actual
sink-points of sinking streams. Similar observations were made during
this study. The topographic maps and geological maps are an essential
guide for planning field work, but interpretation of them is not a
substitute for field work. Field work is required for locating karst
features and must be done. The field work necessary for design of a
rigorous tracing plan can require use of 20 to 50 percent of a project
budget. Aerial photographs are rarely useful in field work for the design
of a tracer test. Most of the project-relevant karst features are too small
to be recognized in photos. Alternatively, in humid climates, they are
obscured by vegetation.
23
-------
11. There are not enough accessible wells in the area of Plate I to accurately
map the potentiometric surface. The surface shown is subjective but
consistent with a prudent, skeptical interpretation of all available
groundwater elevations (at springs, perennial streams, and wells) and
tracing data.
12. In a karst terrane similar to the one studied, it is necessary to extend the
field work at least three miles beyond the estimated boundary of the
basin being studied. This is because delineation of a groundwater basin
must entail partial delineation of each of the basins adjacent to it -- in
fulfillment of the maxim that "A boundary is not a boundary until and
unless it has been tested by traces on each side of its alleged position."
Data for interpretation of the potentiometric surface of an area similar to
that shown in Plate I, if it is to be reliably contoured, must also be
acquired from the area at least a mile (and preferably at least two miles)
beyond the map edges.
13. Dry weather significantly slowed completion of the project because of the
rarity of wet weather flow; many potential sites for dye injection had no
water draining into them — except after major storms. These problems
were solved by scheduling tracing during the rainy season, when flow
velocities are faster, test duration is shorter and, consequently, analytical
and labor costs are less. Where water was not naturally entering the
proposed injection point, alternative sources were used: siphoning by
hose from a pond, injecting tap water via a garden hose from two homes
on Green River Valley Water District line, and employing a 1,500-gallon
tank-truck.
14. A potentiometric map, if based on an adequate amount of data
(preferably about 2 wells per square mile in areas similar to the one
described herein; only 0.73 wells per square mile could be measured in
the area of Plate I north of the Green River), and where there are not
extensive perching beds within an aquifer, is an extremely useful guide to
the design of a tracer test. Nevertheless, as tracing results are acquired,
24
-------
the draft potentiometric map needs to be repeatedly revised in order to be
consistent with tracing data. Tracing data are real; all potentiometric
maps are inferential and subjective.
15. The approximate normalized base flow (NBF) of the 4.9-square-mile Rio
Springs basin is 0.90 cubic feet per second per square mile (cfsm). In
contrast, the approximate NBF of six other groundwater basins in the
Mammoth Cave area south of Green River ranges from 0.15 to 0.20 cfsm,
with a mean of 0.17. These six basins range in size from 8.8 to
190 square miles but there is no significant correlation between the area
of these basins and their NBF. The NBF of what has been called conduit-
flow (low-storage) karst aquifers ranges from about 0.01 to 0.2; the NBF
of diffuse-flow (high-storage) karst aquifers ranges from about 0.2 to 0.4
(White, 1975). [The continuum between what have been called conduit-
flow and diffuse-flow aquifers is reflected in the continuity of NBF values
from 0.01 to 0.4.]* For granular aquifers, the NBF is commonly 0.5 or
more. Why, then, is the NBF of the Rio Springs basin almost six times
*The terms conduit flow and diffuse flow have been used in at least four different senses
since 1971, to refer to idealized end-members of continua describing types of discharge,
recharge, flow, and storage. The consequent ambiguity of these concepts, and the ambiguity
of their implied properties for the carbonate aquifers they purportedly describe, have caused
much confusion, both to investigators and to regulators. Worthington et al. (1992) analyzed
data in the world .literature and concluded that the traditional criterion for distinguishing
between types of recharge and types of flow within aquifers in temperate, climate, hardness
(or its directly related surrogate, specific conductivity), was invalid. Hardness and
conductivity of aquifer discharge are, instead, a measure of percentage of recharge from
sinking streams. Worthington et al. (1992), followed by Quinlan et al. (1993) and Davies
and Quinlan (1993), interpreted the velocities of 1,800+ tracer tests in carbonate aquifers of
25 countries and concluded that flow should be divided, on the basis of velocity, into two
types, rapid-flow and slow-flow, with 0.001 meters/second being the separation value
between them. This holds no matter whether flow is through conduits or through small,
dissolutionally enlarged pores and joints. [This 0.001 m/s value is empirical, being based on
the tracer-test velocities, rather than arbitrary.] For all these reasons, Daves and
Quinlan (1993) recommended that the terms conduit flow and diffuse flow be abandoned,
except to non-generically distinguish between flow within conduits and flow within pores
and minimally enlarged joints. In support of the rapid-flow/slow-flow paradigm
Davies (1992), Davies and Quinlan (1993), show that neither long-term, almost daily
measurements of temperature variation of springs, nor similar measurements of their
conductivity, can be explained by invoking conduit flow or diffuse flow (in any of the
aforementioned sense of these terms). The variations can only be explained by invoking
mixed proportions of rapid flow and slow flow. The bottom line: One cannot use the
concepts of conduit flow or diffuse flow to validly justify decisions about springhead or
wellhead protection area boundaries or groundwater monitoring strategy.
25
-------
higher than the regional average? We believe this higher ratio is a
consequence of differences in storage and yield of the aquifers. The five
basins south of the Green River (Echo River, Pike Spring, Turnhole
Spring, Lower Blue Hole, Graham Springs, and Bear Wallow) have low
storage, respond rapidly to storms, and drain rapidly. The NBF of the
Rio Springs basin is significantly higher for four reasons:
A) Approximately 75 percent of its recharge area is sand-mantled
(Miller, 1969), more than any other basin yet delineated in the Mammoth
cave area; there is no sand and gravel mantle in the five basins that are
compared to the Rio Springs basin. B) Open sinkholes are rare in the
Rio Springs basin; they are common in the other five. C) There is
relatively higher storage of available water in the thick sand and gravel
that overlie the limestone. And D) the relatively non-flashy response of
Rio Springs to storms occurs because there is attenuation of its rate of
recharge and discharge by this sand and gravel above the limestone.
They enhance the quality and yield of waters from Rio Springs, making
them unique and perhaps the best in the Mammoth Cave area for use as
a water supply.
16. We estimate that the inferred 4.9-square-mile area of the Rio Springs
basin is accurate to within + 10 percent. Even if its actual size were
30 percent larger than is shown (6.4 square miles), the normalized base
flow would be 0.69 - still significantly higher than the mean NBF of
other basins in the Mammoth Cave area and supportive of the hypothesis
that the hydrologic responses of the Rio Springs basin are damped by the
effects of the thick sands that blanket most of its recharge area.
REFERENCES
Alexander, E. C., Jr. and Quinlan, J. F. 1992. Practical Tracing of Groundwater,
with Emphasis on Karst Terranes. Short-course manual for Geological Society
of American Annual Meeting. Geological Society of America, Boulder,
Colorado. 2 vols. 161 + 130 p.
Davies, G. J. 1992. Water Temperature Variation at Springs in the Knox Group
near Oak Ridge, Tennessee, in Quinlan, J. F., and Stanley, A., eds. Conference
on Hydrogeology, Ecology, Monitoring, and Management of Ground Water in
26
-------
Karst Terranes (3rd, Nashville, Tennessee, 1991), Proceedings. National
Ground Water Association, Dublin, Ohio.
Davies, G. J., and Quinlan, J. F. 1993. There is no such thing as a diffuse-flow
carbonate aquifer if that aquifer is unconfined and subaerially exposed [abs.]
Geological Society of America, Abstracts with Programs, Vol. 25, No. 6, p. 211.
Dicken, S. N. 1935. Kentucky Karst Landscapes. Journal of Geology, v. 43,
p. 708-728.
Field, M. S., Wilhelm, R. G., and Quinlan, J. F. In review. Use and toxicity of
fluorescent dyes for tracing groundwaters. Submitted to Ground Water and in
revision.]
Hess, J. W., and White, W. B. 1988. Water Budget and Physical Hydrology, in,
White, W. B., and White, E. L., eds. Karst Hydrology: Concepts from the
Mammoth Cave Area. Van Nostrand Reinhold, New York. p. 105-126.
Miller, R. C. 1969. Geologic map of the Cannier quadrangle, Kentucky. U.S.
Geological Survey, Geologic Map GQ-816.
Miller, R. C., and Moore, S. L. 1969. Geologic map of the Hudgins quadrangle,
Kentucky. U.S. Geological Survey, Geologic Map GQ-834.
Miller, R. C., and Moore, F. B. 1972. Geologic map of the Hammonville quadrangle,
Kentucky. U.S. Geological Survey, Geologic Map GQ-1051.
Miller, R. C., and Moore, F. B. 1975. Geologic map of the Magnolia quadrangle,
Kentucky. U.S. Geological Survey, Geologic Map GQ-1280.
Plebuch, R. O, Faust, R. O., and Townsend, M. A. 1985. Potentiometric surface and
water quality in the principal aquifer, Mississippian Plateaus region. U.S.
Geological Survey, Water Resources Investigation Report 84-4102. 45 p.
Quinlan, J. F. 1970. Central Kentucky Karst, Mediterranee, Etudes et Travaux,
v. 7, p. 235-253.
Quinlan, J. F., Davis, G. J., and Worthington, S.R.H. 1993. Review of "Ground-
Water Quality Monitoring Network Design" by Loaiciga, H. A., et al., 1992.
Journal of Hydraulic Engineering, v. 19, no. 12, p. 1136-1141. [Discussion,
with reply, p. 1141-1142.]
Quinlan, J. F., Ewers, R. 0., Ray, J. A., Powell, R. L., and Krothe, N. C. 1983.
Groundwater Hydrology and Geomorphology of the Mammoth Cave Region,
Kentucky, and of the Mitchell Plain, Indiana, in Shaver, R. H., and
Sunderman, V. A., eds. Field Trips in Midwestern Geology. Geological Society
of America and Indiana Geological Survey, v. 2, p. 1-85
27
-------
Quinlan, J. F. and Ray, J. A. 1989. Groundwater Basins in the Mammoth Cave
Region, Kentucky, showing Springs, Major Caves, Flow Routes, and
Potentiometric Surface. Friends of the Karst. Occasional Publication No. 2
[revised edition of map first published 1981].
Sauer, C. O. 1927. Geography of the Pennyroyal. Kentucky Geological Survey,
Series 6, v. 25. 303 p.
Smart, P. L. 1984 [1986]. A Review of the Toxicity of Twelve Fluorescent Dyes Used
for Water Tracing. National Speleological Society [NSS] Bulletin, v. 46, no. 2,
p. 21-33.
U.S. EPA (U.S. Environmental Protection Agency). 1987. Guidelines for
Delineation of Wellhead Protection Areas. Office of Ground-Water Protection,
Washington, DC. 209 p.
U.S. EPA (U.S. Environmental Protection Agency). 1989. Wellhead Protection
Programs: Tools for Local Governments. Offices of Water, Washington, DC.
EPA 440/6-89-002. 50 p.
White, E. L. 1977. Sustained Flow in Small Appalachian Watersheds Underlain by
Carbonate Rocks. Journal of Hydrology, v. 32, P. 71-86.
White W. B. 1988. Geomorphology and Hydrology of Karst Terranes. Oxford
University Press, New York. 464 p.
White, W. B., and White, E. L., eds. 1989. Karst Hydrology: Concepts from the
Mammoth Cave Area. Van Nostrand Reinhold, New York. 346 p.
Worthington, S.R.H. Davies, G. J., and Quinlan, J. F. 1992. Geochemistry of
Springs in Temperate Carbonate Aquifers: Recharge Type Explains Most of
the Variation. Colloque d'Hydrologie in Pays Calcaire et en Milieu Fissure
(5th, Neuchatel, Switzerland), Proceedings. Annales Scientifiques de
1'Universite de Besancon, Geologic - Memoires Hors Serie, no. 11, v. 2,
p. 341-347.
28
-------
APPENDIX A
KENTUCKY GROUNDWATER TRACING FORMS
-------
TRACER INJECTION SITE
#J.
1. Name of Dye Trace (Site Location): Ash Farm Trace - Crhaf-frcS A*yk ^
2. Date of Injection: March / 3 / 1993 Time: 10:45
Month Day Year
3. Owner of Injection Site: Charles Ash Farm Phone: (
4. Quadrangle/County: Hammonville Quad, KY / Hart County
5. Elevation: 780 (*}map ( ) measured
6. Latitude: 37° 22' °5" N Longitude: 85° 45' 27" W
7. Description of Injection Site:
( ) sinking stream ( ) losing stream ( ) karst window
( ) cave ( ) water well ( ) injection well
( ) lagoon ( ) septic system other
Remarks
8. Formation Receiving Tracer Injection: Stc. Genevieve Limestone
9. Flow Conditions: ( ) low (x) moderate ( ) hiqh
10. Field Conditions (precipitation, runoff, etc): runoff occurring into sinkhole
used water from nearby pond to Flush dye into sinkhole
~fs?/<:kc>/
-------
RECORD -OF DYE TRACE
#R-
Project Rio Springs Wellhead Delineation Project Injec
Name of
Princlpa
Preclpfc
Dye Trace (injection s
I Investigator Geary
fte) CAar/eS Aslj Smk'Uo/'C.
tionDa
Trac
te ^ ds'/Zr^ZJaS-
^
H R Not Recovered (high water, stolen receptor, etc)
L Receotor tost
G New or Extra Receptor Installed
Z-*^tA«^c^ £>-£_ \Js?-L«Ji^4 ^>chff-0*L l)/i^4^t-d
("fetlf,
-------
RECORD -OF DYE TRACE
#R-
Project Rio Springs Wellhead Delineation Proiect Infection Date @ 3 / & 3 / ^5
Name of Dye Trace (injections
Principal Investigator Geary
Precipitation before & during ti
ID
23
I^J^^^fe^r
-S^-^Si?^-^ ' ^'- ^'iS???''
Month Day YetJ
Schindel Field Personnel Joe Ray - Tray Lyons
•ace
Date
Duration
Location of Dye Detectors
Tampa Branch
24 iMartis Branch
25
26
27
Caddie Cemetery
Creek
Honey Run Creek
Warren East Creek
28 IBacon Ck. /Wabash Bde.
29
30
31
32
Honey Run at Bridge
Martis Branch @
Beaver
Tampa Branch East
Tampa Branch South
33 I Tampa Branch at
Bridge
-
Back-
ground
i
Dam
\
3~l |3-/z|3^7 \3-26 \*/-z.
v-ft>
Results
—
-
->
—
—
—
—
—
—
~
—
—
NM l/VM
NM
//M
N^
^. —
—
—
—
— .
—
Wf
KM
_ —
_ —
__ ~
— _
— —
/(/Af /Y«/
AfM\A/M
HWI i MM \ /vw MM
NM l/u/if IA/W
-
—
—
—
—
~
frttf
J(/M
JV(M
A/W WM \j\iM
;V^ \&M iA'W lA"^i Wm
i
\
!
(
1
i
MM
i
i
J
!
\
i
— Negative Results 8 Perceptible Background (slight)
+ Positive B + Significant Background (problematic)
Legend; +•*• Verv Positive NR Not Recovered (high water, stolen receptor, etc)
+ •+• •+• Extremely Positive L Recepior lost
/ Receptor Not Changed G New or Extra Receptor Installed
Remarks /V^\ /VW*' /W
-------
TRACER RECOVERY-SITE(S)
AKGWA#
1. Kame of Recovery She:
2. Owner: Uft
3. Quadrangle/County^
4.Etevatlon: ' '
I
AKGWA*
1. Name of Recovery She:_
2. Owner
( J measured
I. Latitude: 37 2 3' "S/'IV Longitude:
6. She Description:
(Xlspring { ) cave ( ) stream
( } water well ( ) monitoring well other_
7. Discharge al Baceflow: £?•
8. Background Status.: Fluor _
9. Dye Detected: (
other
Unit
Rhod.
( ) Rhod
( ) karst window
( •"Jest. ( ) measured
OB DY other
( )OB { )DY
3. Quadrangle/CounTy:_
4. Elevation:
5. Latitude:
) map ( ) measured
Longitude:
6. She Description:
( )spring ( )
( ) water well ( ) monitoring well other
7. Discharge a3 Baceflow^
cave ( ) stream
( ) karst window
10. Method of Detection:
(>/) charcoal/cotton
( ) on-sile fluorometer
11. Method of Analysis:
( ) visible in elutanl
( ) grab sample
( j visual other_
( ) auto sampler
Unt
8. Background SUtuc: Fluor Rhod
9. Dye Detected: ( ) Fluor ( ) Rhod
other
( ) est ( ) measured
OB DY other
( )OB ( )DY
( ) auto sampler
( C ~ ^>£jv) / HCilf!,/
'
AKGWA* #C-
1 . Name of Recovery Site:
olher
1Z Dale of Detection: / /
Monm Day Yu/
13. Initial Dye Breakthrouoh: f ) a.m. ( ) o.m.
14. Duration of Dye Curve:
1 5. Principal Investigator:
1 6. Field Personnel:
AKGWA# #C-
1. Name of Recovery Sile:
2. Owner:
3. Quadrangle/Coumy7_
<. Bevation:
5. Latitude:
( ) map ( } measured
Longitude:
2. Owner:
3. Quadrangle/County:_
4. Elevation:
5. Latriude:
( ) map ( ) measured
Longitude:
6. Site Description:
( ) spring ( ) cave ( ) stream
( ) water well ( ) monitoring well other_
7. Discharge at Baseflow:
8. Background Status: Fluor
9. Dye Detected: ( ) Fluor
other
{ ) karst window
( ) est. ( ) measured
OB DY other
( ) Rhod ( ) OB ( ) DY
Unit
Rhod
1 0. Method of Detection:
( ) charcoal/cor.on ( ) grab sample ( ) auto sampler
( ) on-sile fluoromeler ( ) visual other_
11. Method of Analysis:
( ) visible in elulan; ( ) spectrophoiomeier ( ) Jluorometer
other
12, Date of Detection:
Month
3. Initial Dye Breakthrough:
4. Duration of Dye Curve:
5. Principal Investigator
6. Field Personnel:
Day Year
( ) a.m. ( ) p.m.
6. Site Description:
( ) spring ( ) cave ( ) stream
( ) water well ( ) monitoring well o(her_
7. Discharge at Baseflow:
8. Background Status: Fluor
9. Dye Detected: { ) Fluor
other
( ) karst window
( ) esl. ( ) measured
OB DY other
( ) Rhod { ) OB ( ) DY
UnC
Rhod
10. Method of Detection:
( ) charcoal/cor.on ( ) grab sample
( j on-she fiuorometer ( j visual other
11. Method of Analysis:
( ) visible in elulant
( ) aulo sampler
( ) spectrophotomeler ( ) ftuoromeier
other
12. Date of Detection:
13. Initial Dye Breakthrough:_
14. Duration of Dye Curve:
15. Principal Investigator:
16. Field Personnel:
Dey Yea;
( ) a.m. ( } p.m.
IDENTIFY RECOVERY S!TE(S) ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
' \*XN 8 V>,
ASH FARM TRACE-CHARLES ASH SINKHOLE
HAMMOKVILLE, KY QUAD
ROAD CLASSIFICATION
Heavy-duty.. ._== - — — — — Light-duty
Mediurr-d'jty
U. S. Route
Unimproved dirt
State Route
KENTUCKY
HAMMONVILLE. KY.
NE.'« MUNFGRDV1LLE ! 5' CU'AOR ANCLE
N 3722.5— W 854 5/7.5
1954
PHOTOREVISED 1982
DMA 3B58 IV NE- SERIES V853
-------
TRACER INJECTION SITE
1. Name of Dye Trace (Site Location):
2. Date of Injection: March
Glen Lilly Road Spring
3 / 1993
Time: 1:10
Month
Day
)a,m. (x)P-m.
Year
3. Owner of Injection Site: Royce Noe
Phone:( 502 ) 528-5730
4. Quadrangle/County: CANMER
/ Hart County
5. Elevation: 750 feet
37° 20' 12" N
(X)map ( ) measured _
( ) unknown
6. Latitude:
Longltude:_
85 46' 39" W
7. Description of Injection Site:
(X) sinking stream
( ) cave
( ) lagoon
( ) losing stream
( ) water well
( ) septic system
( ) karst window
( ) injection well
other
' ( ) sinkhole
( ) monitoring well
Remarks
8. Formation Receiving Tracer Inlectlon: Ste. Genevieve Limestone
9. Flow Conditions: ( ) low (x) moderate ( ) high.
10. Field Conditions (precipitation,runoff,etc): water flows from small spring beneath shelf of
rock and sinks at base of small sinkhole.
11. Rate of Flow: 5-7
)cfs
)cms
( ) measured ( ) estimated
( } permanent injection site ( ) intermittent
( ) multiple sites possible
12. Induced Flow? $% no ( )yes
13. Tracing Agent:
( ) Sodium Fluorescein
(X)RhodamineWT
( ) Optical Brightener
( ) Direct Yellow 96
Other '
minutes
Pre-injection Post-injection
Color
Index
Acid Red 388
Elapsed Time
% Active
ingredient
20 percent
Amount
14. Reason for Investigation: Rio Springs Wellhead Delineation Project
15. Principal Investigator: Geary Schindel
16. Agency or Organization: ECKENFELDER INC.
227 French Landing Drive
Nashville
TN
37228
AJOress
( 615 ) 255-2370
City
State
Phone
FAX #
17. Field Personnel: Joe Ray - Geary Schindel
IDENTIFY INJECTION SITE ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
RECORD -OF DYE TRACE
Project Rio Springs Wellhead Delineation Project injection Date O 3 /_ <£?
Monm Day
Name of Dye Trace (injection stte) frfen *
Principal investigator Geary Schindel
Precipitation before & during trace
3 /
Year
Field Personnel Joe Ray - Trav Lvnn«^
Date
Duration
3-21?
ID
Location of Dye Detectors
Back-
grount
Results
Boiling- Springs Conf lui
i Johnson Spring
Cottrell Spring
Log: Spring
Buckner Spring
Rio Springs Conflu.
Rio Springs East
8 I Rio Springs Creek
!Scottv Spring
10
Lanes Mill Sjpring
11 i Bridge Spring
-12 ! Knox Creek Spring
JL3 ' Handy Culvert Spring.
14 Powder Mill Trout Conf
Powder Mill Sn-.--Spri-n
ing
16 I Ba i 1 PV Fa 11 g Spr i -n£
17
Mystery Springs Cpnf.
18
Rumble Spring
12 I Aetna Furnace at Bdg.
_ i
20 -i Branch "Fork at Bridge
21 i Jones School Spring
22 ' Jones School Creek
— Negative Results
+ Positive
Legend: ++ very Positive
+ •+••*• Extremely Positive
/ Receptor Not Changed
B Perceptible Background (sligh:)
B •+• Signifiean: Background (pro&lemaDc)
HR Not Recovered (nign waier, stolen receptor, etc)
L Receptor lost
G New or Extra Beosptor Installed
Remarks
Interpretation
-------
BSE
RECORD -OF DYE TRACE
#R-
Project Rio Springs Wellhead Delineation Proiect Infection Date O 3 / 63 f •^'3
Name of Dye Trace (Injection s
Principal Investigator Geary
Month Day Year
j^ / /It ^^ •
•He) /yy&M L-lU~V K6aJl *?0VM<>i Tracer /?Ap<^#»w/*f.£ UJJ
/ ' W
Schindel Field Personnel Joe Ray - Tray Lvons
Precipitation before & duringtrace
.•:•"-'£..-';•''. .->v
• ''^7;V« ^; / • - -[';", ' " ,~* /, - Iv^'-.l'T
* - *"-"T7' ^- '_-.£ ! ',ri~ -: -«~t£r't rr~~Tl?r.' ^ „* •"-- £'*H -.5?'
ID
23
24
25
26
27
28
29
30
31
32
Date
Duration
Location of Dye Detectors
1 Tampa Branch
iMartis Branch
Caddie Cemetery
Creek
Honey Run Creek
Warren East Creek
Ifiacon Ck./Wahash Bde.
Honey Run at Bridge
Martis Branch @
Beaver
Tampa Branch East
Tampa Branch South
33 i Tampa Branch at
Bridge
i
1
1
I
I
1
BttCK-
grounc
Dam
! i
Legend:
Remarks
3-t2\ J~1J
3*zdv-2 \y-tc
&cu.'&rov^
Results
—
—
—
—
—
-
Nrt
HW
NW
A/W
y /»•
—
—
—
—
-
—
_ — —
—
—
—
—
-
— —
- —
— —
- -
~ —
^.
! ^S
\
\ \ T^»
1
1
' *s
1
1
I
~ \ \
—
1
—
—
—
i
i
1 j
1
1
j
1
— Negative Results B Percepu&le Background (slight)
+ Positive B + Significani Background (pio&iemauc)
-t- + Very Positive H R Not Recovered (nigft wa:er. stolen receplor. etc)
+ + + Ext/emely Positive L Recepior tosi
/ Receptor No! Changed G New or Extra Receptor Installed
f\IW A/V%" fl40*i /~C&Y"
-------
TRACER RECOVERY-SITE(S)
AKGWAlF
1. Name of Recovery She:
2. Owner: UM K\AOUSV\,
4. Etevatlon:
(/-'frnap ( } measured
5. Latitude: 37° / & ¥7 "tf Longrtude:
6. Site Description:
'spring ( t^cave ( ) stream
water well ( ) monrtoring well other_
(
( ) karst window
7. Discharge at Bas-eflow7_,
E. Background Statue: Fluor _
9. Dye Detected: ( ) Fluor
other
Unit
Rhod
(X) Rhod
( ) measured
_OB DY olhe
( )OB ( )DY
10. Method of Detection:
(XT charcoal/cotton
( ) on-she fluorometer
11. Method of Analytic:
( ) visible in elulant
other
( ) grab sample
( ) visual other_
( ) auto sampler
>*
( ,/}, s
( ,/}, spectrophotomeler (
12. Date of Detection:
Montn
13. Initial Dye Breakthrough:
Af A-
Day
( }
( )
14. Duration of Dye Curve:_
15. Principal Investigator: 5P,^IM rf-C f
16. Field Personnel:
U
AKGWA*
1. Name of Fiecovery Sfte:_
2. Owner:
3. Quadrangle/CounTy:_
4. Elevation:
5. Latitude:
( ) map ( ) measured _
Loogrtude:
6. She Description:
spring { ) cave ( )
water well ( ) monitoring well
II:
7. Discharge al &ateflow:_
stream
other
( ) karst window
a. Background Statue: Fluor
9. Dye Detected: ( ) Fluor
other
( ) est ( ) measured
OB DY other
( ) Rhod ( ) OB ( ) DY
Unfc
Rhod
10. Method of Detectior:
( ) charcoal/cotton ( ) grab sample ( ) auto sampler
( ) on-srte fluorometer ( ) visual other
11. Method of Analytic:
( ) visible in elulant ( ) spectropholometer ( } fluorometer
other
12. Date of Detection:
Monm
13. Initial Dye Breakthrough:
14. Duration of Dye Curve:_
15. Principal Investigator:
15. Field Personnel:
Day YCJU
( ) a.m. ( ) p.m.
AKGWA#
AKG\VA#
1. Name ot Recovery Site:_
2. O>vner:
1. Name of Recovery Si1e:_
2. Owner: .
3. Quadrang)e/Coumy:_
<. Elevation:
5. Latitude:
( ) map ( ) measured _
Longitude:
3. Quadrangle/County:_
4. Elevation:
5. Latitude:
( ) map ( ) measured
Longitude:
6. Site Description:
( ) spring ( ) cave ( ) stream ( ) karst window
( ) water well ( ) monrtoring well other
7. Discharge at Baseflow:
( ) est ( ) measured
6. Site Description:
( ) spring ( ) cave ( ) stream
{ ) water~well ( ) monitoring well other_
7. Discharge at Baseflow:
( ) karst window
£. Background Status: Fluor Rhod OB DY other
£. Dye Detected: ( ) Fluor ( } Rhod ( ) OB ( ) DY
other
10. Method of Detection:
{ ) charcoal/cotton ( ) grab sample
( } on-siic fluorometer ( ) visual other
11. Method of Analysis:
( ) visible in eluler,: ( ) specuophotometer ( ) fluorometcr
E. Background Status: Fluor
9. Dye Detected: ( ) Fluor
other
( ) est. ( ) measured
O3 DY other
( ) Rhod ( ) O3 ( ) DY
Unr
Rhod
auto sampler
10. Method of Detection:
( ) charcoal/cotton ( ) grab sarrole ( ) auto sampler
( ) on-srte fiuoromcler ( ) visual o'.her
11. Method of Analysis:
( } visible in clutant ( ) spsctropho'.omeler ( ) lluorome'.cr
other
12. Date of Detection:
Month
13. Initial Dye Breakthrough:
14. Duration of Dye Curve:
1 5. Principal Investigator:
15. Field Personnel:
/
Day
( } a-rr- ( ) p.m.
12. Date of Detection:
Montn
13. Initial Dye Breakthrough:
14. Duration of Dye Curvc:_
1 5. Principal Investigator:
1 G. Field Personnel:
Day Yea;
( ) a.m. ( ) p.m.
IDENTIFY RECOVERY S!TE(S) ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
.-.RD
1 990000 FETT
KENTUCKY—HART CO.
7.5 MINUTE SERIES (TOPOGRAPHIC)
SEM KUNFORDVILLE 15' QUADRANGLE
HODKCWILLF. 17 Ml , RCCM
egg*-* CN'OLI.* 5.3 /.'/ ' . 6JQ OS-fl
f-37c22'30"
--
C-i
-=J T~ '\
Ar&L
^
'-A *"v"/~fc
3jfc-«ff
(•^V ./•-;,
- -5P/2
r^\ \^y
N\"
'3£
^ATV^
w
^ I/A
^% -SP
;^J^
;-^_>j-
7;TT!
o
•' • ~
;r
ifiirv' . X.-V
«;
\v?
Cinwoo
'
>" v vrr3?oooo
No
,
. t— . -.Ir -^? -*•»>•.
:
i! ;= «••— -
-^ r?~
^•*i'Z£l2fiS&S&&
m., \^sg$m&^-}g3&3
'} *J/-z-—&^ *~ ' -x\ ">^ °jLx/^
GLEN LILLY ROAD SPRING -4' --;./-'" \ {j-^ £
CAM-dZR, KY QUAD
^ ^.^/.-r
.1?'.
~y
.AHHI-.K, KX QUAD V-«. 4^/'-4 O P\ ^
:,; a^^LCH^^;^%f
&V^^^^^T i^^fe^j
•9=^
"32
;ev>
\Three__K
/'Knob---,./
.-TV-'
-------
TRACER INJECTION SITE
#J-
1. Name of Dve Trace (Site Location): Glen Lilly Sink
2. Date of Injection: March / 25 / 1993 Time: 12:
Month Day Year
3. Owner of Injection Site: N/A Phone: (
4. Quadrangle/County: Cannier Quad.-, TCY / Hart Co.
5. Elevation: 890 K^rnap ( ) measured
6. Latitude: 37° 21' 06" N Longitude: 85° 47' 12"
7. Description of Injection Site:
$$ sinking stream ( ) losing stream ( )karst window
( )cave ( ) water well ( ) injection well
( ) lagnnn ( ) septic system other
Remarks
8. Formation Receivina Tracer Injection: Girkin Formation
P. Flow Conditions: ( ) low (yXfnoderate ( )hiqh
40 ( ) a.m. fa) p.m.
)
( ) unknown
W
( ) sinkhole
( ) monitoring well
10. Field nondillons (precipitation. runoff. etd: Dye injected after rain on previous day -
hard rain on evening after trace.
11. Rate of Flow: 10-] 5 ( ) cfs fcx) gpm ( ) l/s ( ) cms ( )
( ) permanent injection site (X^ intermittent
( ) multiple sites possible
12. Induced Flow? (x)no ( ) ves /
Pre-injection Posl-injection hlapsed I ime
13. Tracing Agent: Color % Active
Index Ingredient
( ) Sodium Fluorescein
( ) Rhodamine WT
( ) Optical Brightener
(XjJ Direct Yellow 96 unknown
measured £x) estimated
minutes
Amount
6.5 Ibs
O^er • Solophenyl (formerly marketed as Diphenyl Brilliant Flavine
14 Reason for Investiaation: Rio Springs Wellhead delineation Project
15. Principal Investigator: Geary Schindel —
16. Aoencv or Oraanization: ECKENFELDER INC.
227 French Landing Drive Nashville TN
Address City Stale
< 615 ) 255-2288 ( )
Phone FAX*
17. Field Personnel: Joe Ray - Geary Schindel
37228
±\P
TGF
-------
RECORD -OF DYE TRACE
Wellhead Belipearion
NameofDyeTrace (injection site)
Principal im/estlaator Cearv Schindel
Precipitation before & during trace
Tracer
Field Personnel, .Toe
- Trav
ID I Location of Dye Detectors |0*0cUn«j
1 I Boiling- Springs Conf lui.
[ 1
14 ! -povAe-r Mill Trour Confl.
Results
BY ft
Rio Springs Creek
16
Conf.
Legend:
Remarks y€/
— Negative fieiuto
+ Posrjve
•<• + Very Positive
+ + -(- Sxuemely Positive
/ JVscepior Nol Changed
B Percepu'We EacKcrouni (shorn)
B •»• Sionifican: Backarounc (proaiemauc)
N R No: Recovered (r»o^ wa:er. stolen receptor, etc)
L Receplor tos;
G Newor Extra BeccDio^ Installed
y<-
lrrJerpretation_
X2.
-77-
-------
BSD
RECORD -OF DYE TRACE
#R-
Project Rio Springs Wellhead Delineation Project Injection Date ^> 3 /
Monm Day
Name of Dye Trace (injection stte) G"/t.n £.1 LLy ~>/*lkl>)0'l<^ Tracer 'So/ophttyf
/ t- f
Principal Investigator Geary Schindel
Precipitation bef ore & during trace
Yoar
Field Personnel Joe Ray - Trav Lyons
Date
Duration
Y-/7\
ID
Location of Dye Detectors
Back-
pround
Results
23
Tampa Branch
In.-
rt±s Branch
25
Caddie Cemetery Creek
26
Honey Run Creek
27
Warren East Creek
28 (Bacon Ck. /Wabash
29
Honey Run at Bridge
30
Kartis Branch @ Beaver Dam
31
Tampa Branch East
32 I Tampa Branch South
33 I Tampa Branch at Bridge
Legend:
Remarks
— Negative Results
+ Positive
+ •*- Very Positive
+ + +• Exuemely Positive
/ Fvscepior No! Chanoeti
B PerceouNe Sackcround (slight)
B •*• Significant BacKcrounfi (problemauc)
N R No! Recovered (n<~n water, stolen receptor, etc)
L fvwcstor los;
G New or £x.*a R«ce3:cx Installed
lnterpretation_
-------
TRACER RECOVERY-SITE(S)
AKGWA*
1. Name of Recovery She: Afc?
2.
/
*.. Elevation:
5.
( X map ( ) measured
( ) karst window
6. SHe Description:
f>-f spring ( ) cave ( ) stream
{ ) water well ( ) monitoring well other
7. Discharge «t Raseflow: 5"- £ OP? (
Unit
8. Background Status: Fluor Rhod OB DY
9. Dye Detected: ( ) Fluor ( ) Rhod
other
( ) measured
OB DY other
JOB
10 Method of Detection:
(>
-------
KENTUCKY—HART CO.
7.5 MINUTE SERIES (TOPOGRAPHIC)
SE/4 MUNFORDVILLE 1 5' QUADRANGLE
;Nol'"V/«/. \ .ijo B5'45'
X'r-Vv-'" v "?'-' -.; "-
)^. •) N C^ _';.'-r i- «"'T-
f
' *• • ~~
GLEN LILLY SINKHOLE -; f
CANMER, KY QUAD
-------
TRACER INJECTION SITE
1. Name of Dye Trace (Site Location): Bail Road Ditch
2. Date of Injection: March / 25 / A99!
Time: 2:2°
( ) a,m. p.) p.m.
Month Day Yew
3. Owner of Infection Site: County Highway Dept. Right of Way Phone: ( )_
4. Quadrangle/County: Hudgins Quad., KY / Hart County
5. Elevation: 73° ft (x)map ( ) measured .
( )unknown
6. Latitude: 37° 21' 44" N
Lonoltude: 85° A4' 27" W
7. Description of Injection Site: .
(x) sinking stream ( ) losing stream ( ) karst window ) sinkhole
( )cave ( ) water well ( ) injection well ( ) monrtonng well
( ) lagoon ( ) septic system other _—
Remarks Placed dye in water sinking in ditch along N side of road
8. Formation Receiving Tracer Injection: ste. Genevieve Limestone.
9. Flow Conditions: ( ) low (x) moderate ( ) high
10. Field Conditions (proHpitatinn rnnnff PAr.V Dve inlected after rain on previous day -
hard rain after trace .—
11. Rate of Flow: 1 to 3 ( ) cfs (x) gpm ( ) l/s ( ) cms
( ) permanent injection site (x) intermittent
( ) multiple sites possible . .
( ) measured (x) estimated
12. Induced Flow? (x)no ( ) yes
13. Tracing Agent:
( ) Sodium Fluorescein
( )RhodamineWT
(x) Optical Brightener
( ) Direct Yellow 96
Other '
minutes
Pre-injection Post-injection
Color
Index
Elapsed Time
% Active
Ingredient
Amount
Fluorescent Brightening Agent 28
40.5 pounds
14. R»a«»n for Invgstioation: Rio Springs Wellhead Delineation Project
15. Principal Investigator: Geary Schindel . _
16. Agency or Organization:_
227 French Landing Drive
ECKENFELDER INC.
Nashville
TN
37228
Address
City
( 615 ) 255-228
( 615 1 256-8332
17. Field Personnel:
Phone
Joe Ray - Geary Schindel
FAX
IDENTIFY INJECTION SITE ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
RECORD -OF DYE TRACE
ProlectRlo Springs Wellhead Delineation Project Injection Date_
Name of Dye Trace (injection s\He}J^LS<2C^L
Principal Investigator Geary Schindel
Precipitation before &during trace
Field Personnel Joe Ray — Tray Lyons
Date
Duration
ID
Back-
| Location of Dye Detectors |proun"d
Results
1 I Boiling- Springs Conflul
Johnson Spring
Cottrell Spring
L
pring
Buck
?ring
s Coiiflu.
Ri<
East
Rio Springs Creek
Spring
10
Mill
ipring
11 iBridi
17 I Knnx
JL
O:
>der Mil] TT
JL5.
mn
V-f-
JL7_
18
.Rwnbl'
JL2.
_Bjij
2Q--vBrar.r-h-FnT-V "at Bridge
21 I Jones School Spring,
:hool Ci
lk
— Nc-garive Bcsula
t- Positive
Legend: + + ver/positive
•»• + + Ejcuemely Positive
/ ReccplorNotChangf-d
B PereepUbie BackorounC (sliohl)
B+ Signilican: Backa round (problematic)
NR Not Recovered (liionwaier.siolen receptor, etc
L Receptor lost
G Newer cxira Receptcx Installed
Remarks
Irrterpreiatic
-------
RECORD -OF DYE TRACED
. Wellhead Delineation Prelect Injection Date
„ ....... T..Trnrr(.nJ— -1^
Principal lnveStigator_Gear2_ScMSdel
Precipitation before & during trace_
Field Personnel TOP **y - Tray
Date
Duration
ID
Location of Dye Detectors |B*0cot,*d
Tampa Branch
iMartis Branch
25
26
27
Caddie Cemetery Creek
Hooev Run Creek
Warren East Creek
28
Jacon Ck./Wabash Bdg.
29
30
Honev Run at Bridge^
Martis Branch P Beavei
Dam
31
aa Branch South ;
na KT-anrh at Bridge
Results
— \—
— Ncgaiwe fiesote
-*• Positive
Legend: ++ very Positive
+ + + Ext/emety Positive
/ Receptor Not Changed
-
—
—
—
—
—
B Perceptible Background (slicrtl)
B •+ Significan: EackcrounC (problemauc)
N R Not Recovere<: (nrcn water, stolen receptor, etc)
L Receptor lost
G New or Extra Receptor Installed
Remarks
Irrterpretation_
-------
TRACER RECOVERY SITE(S)
AKGWA# *K>
1. Name of Recovery She: &&(./'&<••}' /^/^ ( West
Unh
8. Background Statue: Fluor Rhod OB
9. Dye Detected: ( ) Fluor ( ) Rhod ( ^f6i
other
1 0. Methodj>f Detection:
(*xfcharcoa!/cotlon ( ) grab sample (
( ) on-s'rte lluorometer ( ) visual other
11. Method of Analysis:
( ) visible in elutant ( ) spectrophotometer (
other & I/ ^t^^'C'
* Day
13. Initial Dye Breakthrough: ( )
14. Duration of Dye Curve:
a.m. ( ) p.m.
15. Principal Investigator:
16. Field Personnel:
AKGV/A* #C-
1. Name of Recovery Site:
2. Owner:
3. Quadrangle/County: /
4. Elevation: ( ) map ( ) measured
5. Latitude: Longitude:
6. Sile Description:
( ) spring ( ) cave ( ) stream ( )
( ) water well ( ) monitoring well other
7. Discharge at Baseflow: ( ) est, (
Unr.
8. Background Status: Fluor Rhod O3
9. Dye Delected: ( ) Fluor ( ) Rhod ( ) OB
other
10. Method of Detection:
( ) charcoal/cotton ( ) grab sample ( )
( ) on-site fiuorometer ( ) visual other
1 1 . Method of Analysis:
( ) visible in elutant ( } spect/opholometer (
other
karst window
) measured
DY other
( )DY
aLrto sampler
) fluorometer
12. Date of Detection: / /
Monm Dey Yea;
13. Initial Dye Breakthrough: ( ) a.m. ( ) p.m.
14. Duration of Dye Curve:
15. Principal Investigator:
16. Field Personnel:
IDENTIFY RECOVERY S!TE(S) ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
UNITED STATES
DEPARTMENT OF THE INTERIOR
GEOLOGICAL SURVEY
U^DC^y^':@
}, \J/—ixi If'djtN^'A
^5V*^r^\cS
\T-SV^-. ^UL^/7 ;W=^J
5^0;? 9^^
^ »v •:; :\; -^ ?r v, ?>. c <~T •, / L-TOO-^^CX^ v^/
} |1 SljAXr^7 :'^^-dirf^^y^;
——--'ra u/v - /'Vy' f^* IS
m^C$^r
^;(w6i
-------
TRACER INJECTION SITE
y>ii
1. Name of Dye Trace (Site Location^: Knox Creek Sink
2. Date of Injection: APril_ / 2* / 1993
Time:
)a,m. (x)p.m.
Month Day Yea/
3. Owner of Injection S HP -County Highway Right of Way -
4.
5. Elevation:
Phone: (_
Hart County
690 feet
(x) map ( ) measured
( )unknown
6. latitude: 37° 21' 36" H
7. Description of Injection Site:
( ) sinking stream ( ) losing stream
( )cave ( ) water well
( ) lagoon ( ) septic system
Longitude: 85° A5T 11" W
( ) karst window
( ) injection well
other
(x) sinkhole
( ) monitoring well
Remarks Water injected into sinkhole in bed of .Knox Creek
8. Formation Receiving Tracer Injection: Sre. Genevieve - St. Louis Limestone
9. Flow Conditions: (x) low ( ) moderate ( ) high .
10. Field Conditions (precipitation, runoff, etc): Tool, clear day - T?o precipitation
11. Rale of Flow: 100 ( ) cfs (x)gpm ( ) l/s ( ) cms
( ) permanent injection site ( ) intermittent
( j multiple sites possible_
12. Induced Flow? ( ) no (x)yes Iff^a! /-T_-—r
( ) measured Gc ) estimated
13. Tracing Agent:
Color
Index
15
Elapsed Time
% Active
Ingredient
minutes
Amount
( ) Sodium Fluorescein
( ) Rhodamine WT
(x) Optical Brightener
( ) Direct Yellow 96
Other
Fluorescent Brightening, Agent 28
15 pounds
14. Reason for Investigation^
Wellhead Delineation Project
15. Prirrir1 '™»g*in*t"r- Geary Schindel
16. Agpnry nr Organization: ECKENFELDER INC
227 French Landing Drive
Address
_615 } 255-2288
17. Field Personnel:
37228
Z'P
r_615__\ 256-8332
IDENTIFY INJECTION SITE ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
ESC
RECORD -OF DYE TRACE
Project "Rio Springs Wellhead Delineation. Project Injed
Name of I
Principal
Preclptta
8iiS
3ye Trace (injection £
Investigator Geary
rtei fafi't C>reje,l<: S/H&iak
Schindel
JonDat
£. Trac
e fi?V / 2V / £ ^^72^x^^/5/l^-^^L£--t. */?
'
&f^ "
-------
RECORD -OF DYE TRACE
#R-
Project Rio Springs Wellhead Delineation Project Injecr
Name of Dye Trace (injection E
Principal Investigator Geary
He) At^X f^xT^e-TSA*
UonDa
Trac
te O^f I "2-^ 1 ^3
Wonth Day Year
^OfltlCMf f^lSl^hfeL-Mtr-
i m ' ' \j
Schindel Field Personnel Joe Ray - Trav Lvons
Precipitation before & during trace
"' '' - ~ ^ :'.; '. ••:. i". : ' '•'• .' "•'^"'•cJJ.-' -V'-'^.' ^ ~'?.^-" "^^
T ".^'--^l-J^'TPSiH^** * '. ~ ~ •"-"• :"-i is'iVK^
ID
23
24
25
26
27
Date
Duration
Location of Dye Detectors
Tampa Branch
Back-
ground
Martis Branch
Caddie Cemetery
Boney Run Creek
Creek
i
Warren East Creek
28 Isacon Ck. /Wabash Bde.
29
30
31
32
Boney Run at Bridge
Martis Branch @
Beaver
Tampa Branch East
Tampa Branch South
33 Tampa Branch at
Bridge
Dam
i
-f-2?lr-/ l?-r b-/^
*?-?£
£-/?l6-26|7-Z5 }7-V
1
Results
— "
—
—
—
• —
—
—
/y**/
yV^tf
A"v/
yV^j
—
—
NM
/vW
—
—
—
—
—
— •
-'
I
i i
M*f 1/VtM
/V/M
/V^M I/VIM A/ At
/^ I/VIM
Ani
—
—
—
-
—
—
—
/NM
—
—
—
—
—
- —
A^/«
/v« MIH\NH
kllM \ A/W. \/Vj*t \ /]/£*!
Nfa. \NiM lA/^f \Ntt4
j\m \NM
—
—
—
—
—
—
— ~
—
—
—
—
• —
• —
. —
/IrVH
JVfr \Nh>( |/yAf
- h
^- —
— —
— __
• — —
— . —
— —
i
|
i
i
i j j
—
—
—
• —
-
—
—
— Ne-garive Results B Perceptible EoCKC'Gund (slight)
•+ Positive B •*• Sionidcan; Eacxcroune (problematic)
Legend! •++ Verv Positive NR Not Re-covered (ni;n waief. stolen receptor, eic)
+ + + Exi/emely Positive L Receptor los;
/ fveocpior No! Changed G New of txrz Beo^atof Installed
Remarks /Vn/1 NoT Pb&Wt 'C0V{_cb
Irrterpreta
Jon
-------
TRACER INJECTION SITE
1. Name of Dye Trace (She Location):
Creek Sink
19
94
2. Date of Inlectlon: M""^ _. —
MomS Day Year
County Highway Right of Way
Time:
1:00
( ) a.m. * ) p.m.
3.
phope: (
/
4. Quadrangle/County:_ .
5. Elevation: 690 (x) map ( ) measured 6. Latitude: 37 21' 36" K Longitude:^
17*'
7. Description of Injection Site:
fe) sinking stream () sinkhole () water well
( ) losing stream ( ) karst window ( ) monitoring well
( ) lagoon ( ) cave stream ( ) other
Sinkhole in dry stream bed of Knox Creek
( } injection well
( ) septic system
Remarks
8. Formation Receiving Tracer Injection:
9. Flow Conditions: (*) low ( ) moderate
10. Induced Flow? ( ) no (x)yes 1,500-:-
Ste. Genevieve - St. Louis Limestone
( )high
/ 1,50.0-
39
minutes
11. Tracing Agent: Am! 5 Ibs
Fre-injeclion Posl-injection Elapsed Time
Fluor. ( )Rhod.WT ( ) OB ( ) DY96 ( ) olher_
RECORD OF DYE TRACE
PrlndpannvestlQator G- Sc
Field Personnel
G. Schindel, Joe Ray
—_^^_^_ , p—. ,
Heavy rain night after trace. Appro. 1.5 xnches_
Date [3/12
-v:JDuration
3/1J8 |3/2jl
i
ID
(Location of Dye
Results
Ri.
rings
10
Lanes Mill
Bridge Spring
12
Knoz Creek Spring |
13
Handy Culvert Spring |
Powder Mill Trout SpJ
16
Baily Falls
i
21
Jones School Sp
Brushy Fork
33
Tampa Branch at Bdg |
Mouth ot Knox ureeic
I !
35
Green River at Rio Sp
I I I
I I
, ,
Legend:
PosiUve
t--n- Extremely Positive
— Negatrve Fvssute
B Feroep3't>l« Backpround (slioht)
B+ Sip--.:ftcam Background (protdemalic)
NR No: fv&coveted (hiph waler. siolen receptor, eic)
R fveoepior removed
/ Receptor
L Reoe&lOf tos!
N New Recepior Installed
Remarks Dye vas injected during low flow condition, recovered during high flow
Interpret stion___fjj
iditions.
Please identity injection and recovery si;es on photocopy ol topographic map. Kentucky o^on 0; water ion993
-------
i 990000 F?
r — t-iAKJ cu.
7.5 MINUTE SERIES (TOPOGRAPHIC)
SE/< t/.UNFORDVILLE 15' QUADRANGLE
WvtCKDLIt- 53 '-'I \ -MO
£|ip° '• ^ ''-^ 7^-:^'" fc"^ " '^ ^^S^^^l^
-r-37°22'3CP
"37
CANMER,
mm
%iy.i^- ^
\£ y^ij.*?. •-.:!
< f'V !^f^H
Y?,fc^q-
'-5i i !:* : A\ M "35
• J ii"^1- ' '.SX"0 ,^'/ 1
^^ru^
•.^—-^~ -H^CJ / _ ^}
'.]
-760-
o. c*^--:-:
NS:
i=x -• ^-fr^Cl^
^vfc,-^
©*":^-l \-
§11^: |f^r v/p:
^fe.^.||g^
^vtf^s>;;f$
^ / - - ~~~ • - T*~^~ . *v • • • rt- 1 /
SsNTwi
:;;
•J^^Nv
'^•.
-, AW
'fe&^O^T*'
>i_ f.
••-'-*•
:-
"-^:^^J "• <-
\\9SSA.
-~^- >• - . ^N'-
' - - S> - •• •*- j; ~ \~l /
- ,- *•*»£"'^JU
.--
<<•
00
.,,
"34
~--'
.^•^
^C'Vr
tncxx --~ (~-?2
"N
*?TV
••*a
-------
TRACER INJECTION SITE
1. Name of Dye Trace (Sile Location!: Christene Dye Well
2. Date of Infection: April
/ 24 / 1993
Time: 3:45
Month Day
3. Owner of Injection site! Christene Dye
( ) a.m. Qty p.m.
Year
Phone: (_
4. Quadrangle/County: ranmer Quad .. KY
5. Elevation: 830' _p$map ( ) measured.
_/. Hart
( )unknown
6. Latitude:
37° 21' 59" N
7. Description of Injection Site:
( ) sinking stream ( ) losing stream
( )cave (X) water well
( ) lagoon ( ) septic system
Remarks Abandoned Private Water Well
_LongItude:_
85° 47' 43" W
( ) karst window
( ) injection well
other
( ) sinkhole
( ) monitoring well
8. Formation Receiving Tracer Injection: Girken Format ion/ Ste. Genevieve Limestone
9. Flow Conditions: (x)low ( ) moderate ( ) high.
10. Field Conditions (proHpitatinn mnoff. etrt: Cool clear day - no precipitation
11. Rate of Flow: 3
)cfs
)cms
measured ( ) estimated
(30$ permanent injection site ( ) intermittent
( ) multiple sites possible
12 Induced Flow? ( ) no (X)yes 30 gallons/300 gallons
. _ Pre-injection Post-injection
100
minutes
13. Tracing Agent:
(X) Sodium Fluorescein
( )RhodamineWT
( ) Optical Brightener
( } Direct Yellow 96
Other
Color
Index
Acid Yellow 73
Elapsed Time
% Active
Ingredient
80 percent
Amount
11 Ibs
14. Reason for Investigation: Rio Springs Wellhead Delineation Project
15. Principal Investigator: Geary Schindel
16. Agpncy or Organization: ECKENFELDER INC.
227 French Landing Drivt
Nashville
TN
37228
Address
C.ty
615 ) 255-2288
J_
17. Field Personnel:
Phone
Joe Ray - Geary Schindel
FAX*
IDENTIFY INJECTION SITE ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
RECORD -OF DYE TRACE
5SJ
SI
t
Project Rio Springs Wellhead Delineation Prolect Injection Date &*i 1 %-*/ 1 93
Nameofl
Principal
Preclptta-
L^vr^^v;.?-
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
1 s
16
17
18
19
20"-
>ye Trace (injection s
Invest) aator Geary
te\f (kf !<$£&* r. ftst, Ut&lf
Schindel
Trac
Month Day Vow
Field Personnel Joe Ray - Trav Lvons
Lion before &. during trace
:-£^::-.-:-.M&~^r
^j^^^fi^r^^
Date
Duration
•
Location of Dye Detectors {f*0cun"d
Boiling- Springs Conflul
Johnson Spring
Cottrell Snring
Log Spring
Buckner Spring
Rio Springs Conflu.
Rio Springs East
Rio Springs Creek
Sf-otty Spring
Lanes Mill Spring
Bridge Spring
y-nny CrppV Sp r i p £
JHandy Culvert Spring
Powripr Mill Trout Conf
•Povdpr Mil 1 So/.
Mr q -f
Rumble Spring
Aetna_ "Furnace ^
-K-r'a-nr'h--'Fork at
-^-J
iri-np-
Lj>nf -
it Bdg.
Bridce
21 '• Jones Srhnol Sorinc
22
Legend:
Remarks
Interpret.
J on es School Qreek
&-Z3
*r-f
^_"T* ^f
£YA
S--&\6-lti \6-26 7-2$
Results
• —
—
—
- —
—
. —
—
—
-
—
—
—
. —
—
_____
I
— 1
—
i-hi-
—
—
—
—
—
—
-
—
i
— i — ~-
—
—
_^_
—
—
—
- —
_
_
—
—
—
• —
—
— •
—
—
—
—
—
-
A/w-
A/w
A/w
/V*vi
—
—
—
J-+
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
-hi-
_
"
—
—
—
—
—
—
1
—
+ -h
—
• —
—
—
—
— .
• —
—
—
—
—
__ . —
•M —
— —
— —
— • —
— —
— — •
—
- —
— —
_ —
- — - —
— | —
— —
__ - —
—
— ~-
i x
1 1 ' -^
i ^j
7
I ^
i ! *"
— •
— -
—
— — — — i
— No-arive Rcsuiti B Pcrc«puble Sackoround (slion;)
•* Po'ilive B •<• Significan: Eackprou.id (problematic)
•* + Verv Positrve H R Not Recovered (liigri waier. stolen receptor, etc)
+ -n- Exu'emcly Positive 1- Reoeatorlos;
/ Receptor Not Changed G New or txuz Recep: ex Installed
flirt c /l/&&fl/ftf/'EZ:>ir>g12{
•tion
-------
RECORD OF DYE TRACE
Project Rio Springs Wellhead Delineation Project Injection Date &£f / 2^ /
Month Day
Nameof Dye Trace (Injection srte)^/)KVS££»tg Dy£, &S&U Tracer
Principal Investigator Geary Schindel
Precipitation before & during trace
Year
Field Personnel Joe Rav - Tray Lyons
Date
Duration
10
Location of Dye Detectors
fiack-
Results
23
Tampa Branch
2k ttartis Branch
25
Caddie Cemetery Creek
26
Honey Run Creek
27
Warren East Creek
28 IBacon Ck./Wabasb Bde.
29
Honey Run at Bridee
30
Martis Branch (? Beaver iDam
31
Branch East
32 ITamoa Branch South
33 [Tampa Branch
Legend:
Remarks
Negative Besuiis
Positive
Very Positive
Extremely Positive
Receptor Not Chanced
B PercepiiSle Backgtound
8 + Sionifican; Backp'OimC (pro&lemauc)
KB NOI Recovered (rnjn water, stolen receptor, eic)
L Receptor tos:
G New Of cx.ra Receptor Installed
Interpretation
^
-------
TRACER RECOVERY SITE(S)
as
AKGWA#
1. Name of Recovery She:
2, Owner:
/
3, Quadrangle/County:
4. Elevation: <5"27£? (*-fma.p ( } measured
5. Latitude: ~ ~~ ~" "
Longrtude:
52?
6. SHe DeaCriplion:
( •-'I spring ( ) cave ( ) stream ( ) karsl window
( } water well ( ) monitoring well other
7. Discharge «1 Bascflow: ^ "t
B. Background Statue: Fluor
9. Dye Detected: (
otner
Unit
Rhod
( <^fest. ( ) measured
OB
DY
othei
( ) Rhod ( ) OB ( ) DY
10. Method- of Detection:
( "1 charcoal/cotton ( ) grab sample ( ) auto sampler
( ) on-she fluoromeler ( ) visual other
11. Methodic! Analysis: -^
( /Divisible in elutant ( •^jspectrophotometer (ixjiluorometer
other
12. Date of Detection:
13. Initial Dye Breakthrough:
14. Duration of Dye Curve:
1S. Principal Investigator:
f
16. Field Personnel: .
i x ii
A//J.
Doy
( ) a.m. ( ) p.m.
(M
^ I
AKG\VA#
1. Name of Recovery SHe:_
2. Owner:
3. Quadranple/County:_
4. Elevation:
5. Lalttude:
{ ) map ( ) measured
Longitude:
6. SMe Description:
( 1 spring ( ) cave ( ) stream
( ) water well ( ) monitoring well other_
7. Discharge al Eoceflow:
8. Background Statue: Fluor
9. Dye Detected: { ) Fluor
other
( } karst window
( ) est. ( ) measured
OB DY other
( ) Rhod ( } OB ( ) DY
Unf.
Rhod
10. Method of Detection:
( ) charcoal/coaon ( ) grab sample ( ) auto sampler
( ) on-sHe fluoromeler ( ) visual other
11. Method of Analytic:
( ) visible in elutant ( ) spectrophoiomeler ( } fluoromeicr
other
12. Date of Detection:
Month
13. Initial Dye Breakthrough:
14. Duration of Dye Curve:_
15. Principal Investigator:
16. Field Personnel:
Day Ye*;
( ) a.m. ( ) p.m.
AKGWA#
1. Name of Recovery Site:
2. Owner:
1. Name of Recovery Site:
2. Owner:
3. Quadrangle/County:
4. Elevation:
5. Latitude:
( ) mep ( ) measured
Longitude:
3. Quadrangle/County:
4. Elevation:
5. Latitude:
( ) map ( ) measured _
Longitude:
6. Site Description:
( ) spring ( ) cave ( ) stream
( ) water well ( ) monitoring well other_
7. Discharge at Baseflow:
E. Background Status: Fluor
9. Dye Detected: ( ) Fluor ( ) Rhod
Unit
Rhod
{ ) karst window
( ) est ( ) measured
O5 DY other
( )OB ( )DY
6. Site Description:
( ) spring ( ) cave ( ) stream
( ) water well ( ) monitoring well other_
7. Discharge at Easeflow:
8. Background Status: Fluor
9. Dye Delected: ( ) Fluor
other
UnS
_Rhod
( ) Rhod
( ) karst window
( ) est. ( ) measured
OB DY other
( ) 03 ( ) DY
10. Method of Detection:
( ) charcoal/cor.on ( ) crab sample ( ) auto sampler
( ) on-sile lluorometer ( ) visual other
11. Method of Analysis:
( ) visible in elutan; ( ) spectrophotomcter { ) fiuorometer
' ( ) a.m. ( ) p.m.
1 0. Method of Detection:
( j charcoal/cotton ( ) crab sample ( ) aulo sampler
( ) on-siie fluoromeler ( ) visual other
11. Method of Analysis:
( ) visible in eiulant ( ) spectrophoiomeler
( } iluoromc'.er
other
12. Date of Detection: I
Wontn Day
13. In.'tial Dye Breakthrough:
14. Duration of Dye Curve:
1 5. Principal Investigator:
-i 6. Fic'c7 Personnel:
12. Dale of Detection:
Monin
13. Initial Dye Breakthrough:
14. Duration of Dye Curve:
1 5. Principal Investigator:
1 6. Field Personnel:
Dav
( } a.m. ( ) p.m.
IDENTIFY RECOVERY SITE(S) ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
KENTUCKY—HART CO.
7.5 MINUTE SERIES (TOPOGRAPHIC)
SE/< MUNFORDVILLE 15' QUADRANGLE
ff; "of o
Si&A>&
'Knob
-------
TRACER INJECTION SITE
1. Name of Dye Trace (Site Location): Walter's Well
2. Date of Injection: April
/ 24 / 1993
Time: 12:12
Month
)a.m.
Dey
Year
3. Owner of Injection Site: Ms. Walter
Phone: (
4. Quadrangle/County: Hamnonville, KY _ __
5. Elevation: 845 feet _ (x)map ( ) measured,
Hart County
( ) unknown
6. Latitude:
37° 23' 18" N
Longltude:_
85° 45' 25" W
7. Description of Injection Site:
( ) sinking stream ( ) losing stream ( ) karst window ( ) sinkhole
j )Cave (x) water well ( ) injection well ( ) monitoring well
( ) lagoon ( ) septic system other
Remarks
Ab
idoned private water supply well - casing
ground
8. Formation Receiving Tracer Injection: Beaver Bend and Paoli Limestone - Slnnmed Sand.srnrie
9. Flow Conditions: (*)low ( ) moderate ( )high
10. Field Conditions (precipitation.runoff.etck Cool, clear dav - nn p-rt»n'piran'nTi - vpll
behind house beneath concrete slab, approximately 2 to 3 feet below ground
11. Rate of Flow:
)cfs ( )gpm
)cms
( ) measured ( ) estimated
( ) permanent injection site ( ) intermittent
( ) multiple sites possible
12. Induced Flow? ( ) no (x)yes
13. Tracing Agent:
( ) Sodium Fluorescein
(x) Rhodamine WT
( ) Optical Brightener
( ) Direct Yellow 96
Other '
75 gal / 440 gal
Pre-injection Post-injection
Color
Index
14(L
minutes
Elapsed Time
% Active
Ingredient
Amount
Acid Red 388
20 percent
14. Reason for Investigation^
15. Principallnvestigator:
DP Tin pa firm
16. Anencv or Organization: ECKENFELDER INC.
227 French Landing Drive
Nashville
TN
37228
Address
City
615 ) 255-2288
( 615 ) 256-8332
17. Field Personnel:
Phone
Joe Ray - Geary Schindel
FAX#
IDENTIFY INJECTION SITE ON PHOTOCOPY OF TOPOGRAPHIC MAP
-------
RECORD -OF DYE TRACE
35
31
Projeci Rio Springs Wellhead Pelinearion Project Injection Date <^V /
Montn
/ *f 3
Name of Dye Trace (injection stte)
Principal Investigator Geary Scaindel
Preclphation before 4 during trace
Day
Tracer
UJ7~~
Field Personnel Joe Ray - Tray Lyons
Datel
Duration
y-z.31 y
"-23 I 7/3 f
JD
Location of Dye Detectors
|&»CK-
jpround
Results
Boiling- Springs Conflul
I Johnson Spring
Cottrell Spring
Log Spring
Buckner Spring
Rio Springs Conflu.
Rio Springs East
Rio Springs Creek
Scotcy Spring
10 I Lanes Mill Spring
11 Bridge Spring
1? I Knoy
r Spring
13 I Bandy Ctil vert Spring
14 IPovderM-m Trout Conf
POV^P-T Hi 11 So-.--
ley Tall fi Spring I
JLL
Kysccry Spricgs Conf.
Spring
19 ' Aerns FiiT-nar,e^at Bdg.
20'—-Branch-Forlc-ar Bridge
21 ! Jones School Spring
22 ' Jones Schoo] Creek
Leend:
•* -•••»- Exuemcly Positive
/ fveoepior No: C/vancetf
E PerceoUbl* &acKorour«; (slioht)
B t- Sipnilican: tackpround (piobiemauc)
N R No: Recovered (ragn waier. stolen receptor,
L Reoeoior tos;
G Ne^v or cxira r*cc3:o? ln:ialk-d
Remarks
Irrterpretation
-------
RECORD -OF DYE TRACE
Project RjLo Springs Wellhead Delineation Project Injection Date & V / "2-V
Moom Day
Nameof Dye Trace (injection stte) USc>-t&e,{r USCsll _ Tracer
Yon/
Principal Investigator Geary Schindel
Precipitation before & duringtrace
Field Personnel Joe
- Tra-v- LTOPS
Date
- Duration
*-z.?l 7-3/
ID
Location of Dye Detectors
Buck-
ground
Results
23 IjTagrpa JBranch
— I - I AM
iMartis Branch
25
Gaddle Cemetery Creek
26
Hooey Run Creek
27
Warren East Creek
28
(Bacon Ck./Wabash Bdg.
29
Boney Run at Bridge
30
Martis Branch @ Beaver
Dam
31
Tasrpa Branch East
Tampa Branch South
33 lTampa Branch at Bridee
i
Leaend:
- Positive
•*• Very Positive
••»••*• £2rjemely Positive
r No;
E Percepiiolc Bac'-proune (slicr-.t)
B -f Sionidc^n: &ackcrounc ^pro&Jemauc)
H K No: Rccove'eC (nipn water, slolen receptor, eic)
L Reoeotof los;
G Nov or txira fvjccpiof Insalljrd
Remarks
nterprelation
-------
-25'
II
Heavy-duty =-
Wec'iun-cuty,
ROAD CLASSIFICATION
Light-duty _
Unimproved dirt
O State Route
QUADRANGLE LOCATION
HAMMONVILLE. KY.
NE/-i MUNFGP.3VILLZ IS' QUADRANGLE
N 3722.5-W 85^5/7.5
1954
PHOTOREVISED 1932
DMA 1858 IV NE-SERIES V853
-------
TRACER INJECTION SITE
33
ST
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
#J-
Name of Dye Trace (Site Location): Route 357 Sinkhole
April 24 1993 11:
Date of Injection: / / Time:
Month Day Year
Owner of Injection Site: Unknown Phone: (
Quadrangle/County: Cs&Ylfw-e-Vi /CV {^tJa-tf, $&£& 1 H^L^d (
r /
Elevation: ~/^^O SuCf pOmap ( ) measured
Latitude: Longitude:
Description of Injection Site:
( ) sinking stream ( ) losing stream ( ) karst window
( ) cave ( ) water well ( ) injection well
( ) lagoon ( ) septic system other
Remarks
Formation Receiving Tracer Injection: (~FtV*rZ\V\. h-t?*riMti.'&)/9i/\^
Flow Conditions: pflow ( ) moderate ( ) hiqh
Field Conditions (precipitation, runoff, etc):
Rate of Flow: /- 3 ( ) cfs (X) gpm ( ) 1/s ( ) cms (
( ) permanent injection site (A3 intermittent
( ) multiple sites possible
Induced Flow? f"^h° ( ) ves /
00 X
( )am. ( ) p.m.
}
'^oe/is'Cv
( ) unknown
(^ sinkhole
( ) monitoring well
) measured (^estimated
minutes
Pre-injection Post-injectton Elapsed Time
Tracing Agent: Color % Active
Index Ingredient Amount
( ) Sodium Fluorescein
( ) Rhodamine WT
( ) Optical Briahtener
f*) Direct Yellow 96 5o/£>flh-e M *//
Other
Reason for Invesiiqation: Rio Springs Wellhead Delineation Project
Geary Schindel
Principal Investiqator:
ECKENFELDER INC.
Aqencv or Orqanization:
227 French Landing Drive Nashville TN
Adcress Ciry • Stale
615 255-2288
Phone FAX
Field Personnel" Joe Ray-Geary Schindel
3" /h
-------
RECORD -OF DYE TRACE
S3
ProleclHao Springs Wellhead Delineation Project Injection Date
Yea.'
NameofDveTraeefinlectlonsHel
Principal investigator Geary Behind el
Precipitation bef ore & during trace
35"7
Tracer
Field Personnel Joe Ray - Trav Lyons
Date
Duration
-/y!
ID
j Location of Dye Detectors ]{:"„"„" f\-
33 I Ha-ntlv Culvert Spring
JLL
Mil_1_Trout Conf
J_5_
HPT- Hi 11 Sn.-J
1T1«
JL£_
JLZ.
Mystery Springs Conf.
JLS.
19 ' Aetna, Pur-nan* at Bdg. J
20
-'-
Branrh-ForV--ar Bridge
21
Jones School Spring
22 I Jones School Creek
Results
Positive
r No^ Cftino ef Inslalied
feo«?ior, e:c)
Remarks
Interpretation
(7
-------
B32
RECORD OF DYE TRACE
#R-
Project Pd.0 Springs Wellhead Delineation Project Inject
Nameof
Princlpa
Prec-lpte
[_•— "j£o«-~.i-.:
Dye Trace (injection sfte)
UonDa
Trac
te / /
Wonm Day Yea;
:er
1 Investigator Gearv Scbindel Field Personnel Joe Hav - Tray Lyons
rtionbef ore tdurino trace
^^:'Qii~r^-r-^^i Date
~nvr '^^SsT^'^z -^£| Duration
ID j Location of Dye Detectors
23 iTamDa Branch
24 iMartis Branch
25
26
27
2B
23
30
31
Caddie Cemetery Creek
Honey Run Creek
Warren East Creek
Bacon Ck./Wahash BdE.
Honey Run at Bridee
Kartis Branch @ Beaver
Tanroa Branch East
32 JTaxana Branch South
33 Tampa Branch at Bridee
i
I
1
Back-
ground
Dam
*/-?& tr-/
5-?
*T'k
er-^-t
£-/^
Resufts
—
• —
—
/vn
' 1/VtM
*- l/tym
—
—
—
—
—
f^
^_
H^l
//XU I//M
. —
^
, —
^-~
—
. — .
. — ,
•
1
1
1
V-s
Srj
I --..
1 ^
— ~ _
—
• —
—
—
—
—
— '
—
—
— —
I
1
— .
• —
•
-
!
1
i
!
!
i !
Legend:
Remarks
Interpretation
- Posiuve
- -*• Very Posiliv-
r No: Cf-.a-ioed
B PtrccD'Jblr &ackp'Oor,<; (slich;)
B -t- Sipnllican: EacKciounc (.D'OOIemaiic)
N R Not Beoovere-o (mon warer. stolen recopior. cic)
L F^ceoiot los;
G New or Exva
------- |