<>EPA
United States
Environmental Protection
Agency
Office Of
Water
(WH-550G)
EPA 570/9-91 -008
June 1991
Wellhead Protection
Strategies For
Confined-Aquifer Settings
-------
WELLHEAD PROTECTION STRATEGIES FOR
CONFINED-AQUIFER SETTINGS
Bureau of Economic Geology
The University of Texas at Austin
Ground-Water Protection Division
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
Washington, DC 20460
1991
Printed on Recycled Paper
-------
ACKNOWLEDGEMENTS
This document was authored, under cooperative agreement # CX-815385-01-0, by Charles W. Kreitler
and Rainer K. Senger of the Bureau of Economic Geology, University of Texas at Austin, Austin, Texas for the
U. S. Environmental Protection Agency, Office of Ground Water and Drinking Water, Ground-Water Protection
Division (GWPD). Marilyn Ginsberg of GWPD served as Project Manager.
-------
Disclaimer
These are the results of investigations supported by the U.S. Environmental Protection Agency's
Ground-Water Protection Division, Office of Ground Water and Drinking Water as part of its efforts to
provide technical assistance to State, tribal, and local governments on the implementation of the
Wellhead Protection Program. The specific methods and approaches contained in this document have
been peer-reviewed but do not constitute official Agency endorsement or policy recommendations. The
Ground-Water Protection Division, Office of Ground Water and Drinking Water provides this
information to help solve complex technical problems related to the delineation of wellhead protection
areas in confined and semiconfined aquifer settings. Further assistance is available from the Ground-
Water Protection Division, Office of Ground Water and Drinking Water in Environmental Protection
Agency headquarters, as well as the ground-water offices in the ten Environmental Protection Agency
Regions.
-------
CONTENTS
Executive summary xiii
Degree of confinement xiii
Delineating wellhead protection areas xv
Examples of wellhead protection areas in confined aquifers xviii
Acknowledgments xx
Chapter 1. Introduction 1
Confined aquifers: why be concerned? 1
Purpose of document 2
Definition of a confined aquifer 3
Distinction between a semiconfined aquifer and a highly confined aquifer 5
Importance of understanding degree of confinement in context of wellhead protection 5
Chapter 2. Characteristics of a confined aquifer 8
Geologic characteristics 8
Confining beds 8
Confined-aquifer lithology 9
Hydrologic characteristics 9
Elevation of potentiometric surface 11
Direction of vertical ground-water flow 12
Flow velocity and age 15
Storativity 15
Cyclic water-level responses resulting from atmospheric pressure changes 17
Cone of depression 19
Hydrochemical characteristics of ground water in confined aquifers 28
-------
Chapter 3. Approaches for determining the presence and/or the degree of confinement 32
Geologic approach 33
Classic geologic maps 33
Environmental geologic and hydrogeologic maps 34
Subsurface geologic maps 37
Hydrologic approach 37
Water-level elevation in a well 38
Potentiometric surface 38
Pump test for storativity 39
Pump test for leakage 39
Continuous water-level responses 41
Hydrologic measurements in confining strata 42
Numerical modeling 42
Hydrochemical approach 43
General water chemistry 43
Tritium and other anthropogenic chemicals 44
Carbon-14 47
Contamination 48
Hydrochemical measurements in confining strata 49
Changes in water chemistry 49
Quantitatively distinguishing semiconfined from highly confined aquifers 50
Recommendations for evaluating confinement 53
An integrated approach 53
Geologic approaches 53
Hydrologic approaches 54
Hydrochemical approaches 54
VI
-------
Chapter 4. Developing a wellhead protection area 55
Definition of wellhead protection area 55
Protection goals 55
Providing time to react to incidents of unexpected contamination 55
Lowering concentrations of a contaminant to target levels before contaminants
reach a well 56
Protecting all or part of the zone of contribution from contamination 56
Hydrodynamic criteria for delineation of wellhead protection areas for confined aquifers 56
Distance 57
Drawdown 57
Time of travel 58
Row boundaries 58
Assimilative capacity 59
Recommended hydrodynamic criterion for confined aquifers 60
Chapter 5. Methods for calculating wellhead protection areas 62
Methods for calculating wellhead protection areas for confined aquifers with negligible-
gradient regional potentiometric surfaces 62
Cone of depression approach 62
Drawdown in monitoring wells at different distances from a producing well method 62
Drawdown versus time in the producing well or in a monitoring well method 62
Drawdown versus distance simulation using analytical solutions and simple
computer models method 63
Time of travel approach 66
Cone of depression/Time of travel method 66
Cylinder method 69
Semianalytical method (WHPA model) 71
Comparison of approaches and methods 74
vn
-------
Calculation of wellhead protection area for wells in confined aquifers with a regional
sloping potentiometric surface 75
Zone of contribution with identification of flow boundaries method 76
Zone of transport with time of travel contours approach 78
Simple analytical solution method 78
Semianalytical method (WHPA model) 80
Reverse-path calculations method 80
Comparison of methods 80
Chapter 6. Wellhead protection areas for semiconfined and highly confined aquifers 82
Permeability pathway criteria for semiconfined aquifers 82
Permeability pathway criteria for highly confined aquifers 82
Chapter 7. Calculation of wellhead protection areas for well fields 86
Chapter 8. Examples of wellhead protection strategies in confined aquifers 90
Bastrop, Texas, example from the updip section of a confined aquifer 90
Wharton, Texas, example from the downdip section of a confined aquifer 93
Comparison of wellhead protection areas for the two examples 95
Chapter 9. Recommended approach for defining wellhead protection areas for confined
aquifers 97
References 101
Appendix 1. Comparison of wellhead protection areas—two examples 109
Bastrop, Texas, example from the updip section of a confined aquifer Ill
Hydrogeologic setting Ill
Determining confinement 112
Geologic approach 112
Hydrologic approach 116
Hydrochemical approach 118
Conclusions on confinement 121
Vlll
-------
Wellhead protection area delineation 122
Cone of depression approach 122
Analytical methods 122
Numerical methods 124
Time of travel approach 125
Cylinder method 125
Cone of depression/ time of travel method 125
Semianalytical method (WHPA model) 127
Recommended wellhead protection area 129
Wharton, Texas, example from the downdip section of a confined aquifer 131
Hydrogeologic setting 131
Determining confinement 132
Geologic approach 132
Hydrologic approach 134
Numerical model 139
Hydrochemical approach 139
Conclusions on confinement 141
Wellhead protection area delineation 141
Cone of depression approach 141
Analytical solutions and simple computer models method 141
Time of travel approach 144
Cylinder method 144
Cone of depression/time of travel method 144
Semianalytical method (WHPA model) 144
Recommended wellhead protection area 144
IX
-------
Appendix 2. Confined Aquifers of the United States, the Commonwealth of Puerto Rico, and the
Pacific and Caribbean Territories, prepared by James Hamilton, Ground-Water Protection
Division, Office of Ground Water and Drinking Water, U.S. Environmental Protection
Agency 147
Introduction 149
Acknowledgments 149
General description of confined aquifers 151
Physiographic region 1 151
Physiographic region 2 151
Physiographic region 3 153
Physiographic region 4 154
Alaska, the Hawaiian Islands, and the Pacific and Caribbean Islands 155
References 156
Appendix 3 159
Glossary 161
Figures
1. Schematic of a confined aquifer (unconfined in the outcrop area) 4
2. Schematic of a semiconfined (leaky) aquifer 6
3. Aquifers may contain low-permeability strata that are interbedded between permeable
strata and may cause confining conditions 10
4. (a) Confined aquifer where the potentiometric surface is higher than the water table of
the overlying unconfined aquifer, (b) Confined aquifer where the potentiometric surface
is lower than the water table aquifer 13
5. Comparison of potentiometric surface of Floridan aquifer to unconfined, semiconfined and
confined sections of the Floridan aquifer 16
6. Weather-related barometric changes and their effect on the water levels in a well
penetrating a confined aquifer 18
7. Example of daily water-level changes in two wells from the Edwards aquifer,
Georgetown, Texas 20
8. Sources of water from a pumping well in a confined aquifer 21
-------
9. Theis curve and leaky aquifer curves 22
10. Example of pump test (drawdown versus time) for nonleaky aquifer, Dakota sandstone 24
11. Example of pump test (drawdown versus time) for moderately leaky confined aquifer,
Indiana 25
12. Example of pump test (drawdown versus time) for very leaky unidentified aquifer 26
13. Geologic setting and pump test data from confined Oxnard aquifer, overlying and
underlying aquitards, and overlying and underlying aquifers 27
14. Evolution of hydrochemical facies from variable composition from Ca-HCOs to Na-
HCO3 to a Na-Cl for ground-water flow in the Atlantic Coastal Plain, New Jersey 29
15. Example of tritium in ground water from Fresno County, California 30
16. Map of Tom Green County, Texas, showing locations of abandoned oil and gas exploration
boreholes 36
17. Tritium in precipitation data from 1950 to 1986, Ottawa, Canada 46
18. Simulation of drawdown versus log distance for hypothetical aquifer for different
values of leakage 61
19. The lateral extent of a cone of depression of a pumping well can be determined with time
versus distance data 64
20. Simulation of time of travel (in years) for hypothetical aquifer for different values of
leakage 68
21. Example of reverse-path calculation 70
22. Cylinder or volumetric-flow equation approach for calculating time of travel for 40 yr 72
23. Example of reverse-path calculation using wellhead protection area (WHPA) computer
program 73
24. Ground-water flow field for cone of depression of a pumping well with a regional ground-
water flow gradient 77
25. Schematic of areally distributed permeability pathways for semiconfined aquifer 83
26. Example of wellhead protection area for highly confined aquifer where penetration of
confinement has only occurred with abandoned boreholes and a fault 84
27. Example of overlapping wellhead protection areas for two wells in different confined
aquifers 87
28. Overlapping wellhead protection areas based on cones of depression for a highly
confined and a semiconfined aquifer 89
29. Geologic map and cross section of the Gulf Coast area, showing locations of Bastrop
(Camp Swift well field) and City of Wharton 91
XI
-------
30. General highway map of Bastrop County showing the location of the Camp Swift well
field and wellhead protection area for wells 515 and 516 92
31. Map of Wharton, Texas, and vicinity, showing wellhead protection area for city of
Wharton well no. 3 (well 406) 94
32. Flow chart for designing wellhead protection areas for confined aquifers 98
33. Geologic map of outcrop of Wilcox Group, Bastrop, Texas 113
34. North-south cross section of driller's logs and geophysical logs at the Camp Swift well
field 115
35. Log-log plot of drawdown versus time for monitoring wells 503,504, and 505 during
pumping test in well 502, Camp Swift well field 117
36. Log-log plot of drawdown versus time for pumping Camp Swift well 516 during 36-hour
pumping test in 1986 119
37. Distribution of hydrochemical facies and total dissolved solids and calculated carbon-14
ages for the Wilcox Group aquifer and Simsboro Formation 120
38. Radial distance for wellhead protection areas for well no. 516, Bastrop, Texas 126
39. Capture zones for well 516 for the 5-, 10-, 20-, 30-, and 40-year time of travel assuming a
regional hydraulic gradient of 0.002 128
40. Subsurface distribution of sand and shales based on driller's logs and geophysical logs for
City of Wharton wells 133
41. Regional hydrologic cross section through Wharton and adjacent counties showing
vertical distribution of hydraulic heads 135
42. Log-log plot of drawdown versus time in the pumping wells 406, indicating the
drawdown stabilized after about 4 min 136
43. Semilog plot of negative drawdown versus time based on a 3-hr pumping test in well 406 138
44. Distribution of hydrochemical facies along a vertical cross section in the downdip
direction 140
45. Radial distance for wellhead protection areas for well no. 406 for Wharton, Texas 143
46. Major and significant minor confined aquifers of the United States 150
47. Physiographic regions of the United States 152
XII
-------
EXECUTIVE SUMMARY
Improper management of contamination sources has resulted in numerous cases of ground-water
contamination of public water supply wells. One approach toward preventing contamination of public
water supplies is to protect the areas that recharge precipitation and surface water to the aquifer near
the wells. This zone of protection is referred to as a wellhead protection area (WHPA). The potential
for contamination is typically less in a confined aquifer than in an unconfined aquifer. Nevertheless,
contamination of confined aquifers has occurred. Wellhead protection areas should be developed for all
aquifer settings.
A confined aquifer is an aquifer overlain by low-permeability strata. The presence of the low
permeability material reduces the risk of a surface contaminant reaching a producing well. The
potential for contamination of a confined aquifer is controlled by two factors: (1) The presence of
permeable pathways (for example, faults, fractures, permeable sands, or unplugged abandoned
boreholes) that permit contaminant migration and (2) the existence of appropriate hydrologic
conditions (for example, downward flow) that cause contaminants to migrate through the low-
permeability strata.
Confined aquifers occur pervasively from coast to coast in the United States. The coastal plain
aquifers along the Atlantic Ocean and Gulf of Mexico represent some of the largest confined aquifer
systems in the United States. There are numerous other smaller aquifers which exhibit confined
conditions.
Degree of Confinement
Before a wellhead protection area can be delineated, the degree of confinement of the aquifer
setting must be determined. Aquifers can be unconfined or confined. Confined aquifers can be subdivided
into semiconfined and highly confined aquifers. A semiconfined aquifer is an aquifer overlain by strata
xiii
-------
that have relatively low permeability compared to the aquifer. However, the permeability of these
overlying strata may be high enough to allow significant leakage through the strata. A fractured till
is a good example of a relatively low-permeability stratum with significant leakage. In such a setting,
it is inferred that the leakage is areally distributed. In a highly confining strata, leakage is
negligible. If leakage does occur, it is probably restricted to localized zones such as discrete faults or
artificial penetrations such as wells, and abandoned or improperly plugged boreholes. A semiconfined
aquifer is more susceptible to contamination than a highly confined aquifer because of the potential for
significant leakage through the overlying confining strata.
There are several approaches for differentiating confined from unconfined aquifers. These
approaches can be considered as (1) geologic, (2) hydrologic, and (3) hydrochemical. Geologic
approaches include (a) classic geologic mapping, (b) environmental geologic and hydrogeologic
mapping, and (c) construction of geologic cross sections. Hydrologic approaches include evaluations of
(a) water-level elevation in wells, (b) potentiometric surface maps, (c) storativity, (d) leakage, (e)
continual water-level responses in wells, and (f) numerical models. Hydrochemical approaches involve
the evaluation of (a) general water chemistry, (b) tritium and (c) carbon-14 data. Tritium is the
radioactive isotope of hydrogen that has been introduced into the atmosphere in the last 40 yr by
atmospheric nuclear testing. It is now in the recently recharged ground •water in measurable but
nonharmful concentrations. Carbon-14 is the radioactive isotope of carbon that can be used to estimate
the age of ground waters that may be hundreds to thousands of years old.
Though several techniques differentiate confined from unconfined aquifers, only a few
approaches can be used to quantitatively differentiate semiconfined from highly confined aquifers. A
40-yr time of travel (TOT) approach is recommended for making this differentiation (that is, 40 yr is
considered to be a reasonable "rule of thumb" to distinguish between semiconfined and highly confined
conditions). This 40-yr time of travel from the recharge area at the ground surface to the well in the
aquifer can be calculated by hydrologic methods or inferred from tritium analyses. Using the time of
travel equation plus leakage values calculated from a pump test, the rate of vertical leakage through a
low-permeability strata can be estimated. If the calculated time of travel is less than 40 yr the aquifer
xiv
-------
is considered semiconfined. If the time of travel is greater than 40 yr then the aquifer is considered
highly confined. Similarly, if the tritium concentrations in the aquifer are less than 1 to 2 tritium units
(TU), the lower level of detection for many tritium analyses, then the water is older than 40 yr. This is
approximately the amount of time since tritium was first introduced in the hydrosphere by
atmospheric nuclear testing. If the water contains tritium concentrations above 1 to 2 tritium units, then
the confined aquifer has been recharged within the last 40 yr, either by horizontal flow or by vertical
leakage. If horizontal flow cannot explain the presence of tritium, then the tritium must result from
vertical leakage and the aquifer should be considered semiconfined.
It is important to differentiate between semiconfined and highly confined aquifers because, as
previously Stated, semiconfined aquifers are subject to pervasive leakage through the overlying low-
permeability strata, whereas potential leakage to a highly confined aquifer is limited to localized
and discrete permeability pathways. Different types of wellhead protection strategies are needed for
the semiconfined and highly confined aquifers.
Delineating Wellhead Protection Areas
Determining a wellhead protection area for a well or well field in a confined aquifer setting
requires delineating a general area for protection based on hydrodynamic approaches. Subsequently,
critical zones within the general area are defined by identifying potential high-permeability
pathways for downward migration of contaminants through the low-permeability strata overlying the
aquifer.
The hydrodynamically delineated wellhead protection area can be based on either a cone of
depression (COD) (as referred to as zone of influence [ZOI]) approach or a zone of transport (ZOT) (also
referred to as the time of travel [TOT]) approach. The time of travel approach is recommended in
preference to the zone of influence approach.
The cone of depression approach uses the lateral pumping extent of a cone of depression as the
wellhead protection area and, in an area where the prepumping gradient of the piezometric surface is
xv
-------
negligible, cone of depression represents the area for which there is a potential for downward vertical
and lateral flow towards a producing well. The zone of influence approach is one method recommended
for defining the wellhead protection area in unconfined aquifers. However, this approach may not be
appropriate for confined aquifers. As the confining strata become more impermeable, the lateral extent
of a cone of depression in a confined aquifer may become unrealistically large. For example, the radius
of a cone of depression for a semiconfined aquifer may be a few hundred feet, but for a highly confined
setting it may extend more than 10,000 ft. The highly confined aquifer, which is less sensitive to
potential contamination, will have a cone of depression area significantly larger than one for
semiconfined and unconfined aquifers. This increase in lateral extent of the cone of depression is due to
the fact that a pumping well in a confined aquifer must draw more of its ground water from lateral
sources because less water is available from vertical leakage. Therefore, for highly confined aquifers
wellhead protection areas based on cones of depressions may be unreasonably large.
A time of travel approach provides a more realistic estimate of a wellhead protection area for a
confined aquifer. The time of travel approach provides a protection area defined by the lateral
distance that ground water flows for a defined period of time and can be defined by an equal-time
contour line. Inside that contour line, ground water will flow to a pumping well in less than the
specified period of time. Outside that contour, it takes water longer than the specified time to flow to
the producing well. There are two basic methods for calculating a time of travel: (1) A volumetric-flow
equation, which is a modification of Carey's law, provides the distance of flow over a given period of
time. The volumetric-flow equation calculates the radius of a cylinder from which all ground water is
pumped. The wellhead protection area calculated using time of travel may be too large, because it
assumes that there is no vertical leakage and, therefore, that all ground water discharged results from
lateral flow. (2) A second method is to use a time of travel calculation based on the hydraulic gradient
of the cone of depression. The second method, the cone of depression/time of travel, is a more realistic
estimate of time of travel, because it incorporates any vertical leakage into the calculation.
The distance of a time of travel contour from the pumping well for a leaky confined aquifer might
be, for example, a few hundred feet, whereas for a highly confined setting the travel time distance for
xvi
-------
the same period of time might extend to thousands of feet. The cone of depression for the leaky system
stabilizes with a much smaller radius than that for the more confined setting, because in the leaky
setting vertical leakage supplies water to the pumping well, which otherwise has to be supplied by
lateral flow. The more confined an aquifer, the more it approaches the condition of receiving no
vertical leakage, and the closer the time of travel calculated with the cone of depression/time of
travel method approaches the time of travel calculated by the volumetric-flow equation. In general,
the wellhead protection area calculated with time of travel will be smaller than a wellhead
protection area calculated with a cone of depression.
A forty-year time of travel threshold is a reasonable "rule of thumb" for distinguishing between
semiconfined and highly confined aquifers. Forty years is the time frame for which tritium has been
introduced into the atmosphere and therefore into ground water. Well water with no tritium indicates
that it took ground water a minimum of 40 yr to flow horizontally and/or vertically from a point of
recharge to the well. Conversely, well water with tritium indicates ground water that has been
recharged within the last 40 yr; thus, the particular well or aquifer is relatively sensitive to aquifer
contamination.
The shape and size of a wellhead protection area can be affected by the gradient of the regional
potentiometric surface. Nonnegligible gradients cause a wellhead protection area to have a noncircular
shape. The exact shape depends on the rate of pumpage, the transmissivity of the aquifer, and the
regional gradient.
After a general wellhead protection area has been determined using hydrologic criteria, the
permeability pathways through the confining strata should be considered. For a semiconfined aquifer,
permeability pathways such as fractures are considered to be common and evenly distributed and,
therefore, the entire wellhead protection area should be considered highly sensitive to potential
contamination, as is the wellhead protection area for an unconfined aquifer. In contrast, for a highly
confined aquifer, the pathways for contaminant migration probably are limited to a few discrete
breaches of the confining strata. These breaches in confinement might be abandoned boreholes or faults
and should be given a higher level of protection from the rest of the area. In a highly confined aquifer
xvii
-------
setting two levels of protection should be developed. The general hydrodynamic area should be given
one level of protection and the immediate vicinity of discrete pathways, where leakage could occur,
should be given a higher level of protection.
Examples of Wellhead Protection Areas in Confined Aquifers
Wellhead protection areas were determined for two confined aquifer settings in Texas. The first
field case setting was in Bastrop, Texas, where the highly productive Wilcox aquifer crops out. Before
the study it was not known whether the well field would be in the confined or unconfined part of the
aquifer because of its location in the outcrop of the aquifer. The second field case setting was in
Wharton, Texas, where the Gulf Coast aquifer was presumed to be highly confined beneath the
Beaumont Clay. Field studies first evaluated the presence and degree of confinement and then
wellhead protection areas were delineated for municipal well fields in both communities.
In Bastrop, Texas, the Wilcox aquifer was found to be highly confined even though it was located
in the outcrop. The degree of confinement was tested with five techniques, which include (1) evaluating
the regional hydrogeologic setting; (2) conducting a pumping test; (3) monitoring of continuous water
levels; (4) assessing the general hydrochemistry; and (5) determining tritium and carbon-14
concentrations in the well water. The results of the investigations indicate a high degree of confinement
and old waters with ages greater than 4,000 yr. The radius of the wellhead protection area ranged from
3,000 to 18,000 ft, based on the different hydrodynamic approaches. The regional gradient affected the
shape of the wellhead protection area. A wellhead protection area of 3,000 to 7,000 ft in the
downstream and upstream direction, respectively, was considered the most realistic. The most critical
pathways for potential contamination of the ground water are artificial penetrations such as wells and
abandoned boreholes.
In Wharton, Texas, the Gulf Coast aquifer was found to be highly confined. The regional
hydrogeology was investigated, in addition to the evaluation of pumping tests, general
hydrochemistry, and tritium and carbon-14 measurements. The results of the investigations indicate a
xviii
-------
high degree of confinement and old waters with ages greater than 15,000 yr. A pump test indicated
extensive leakage, but it appears that this leakage results from ground-water draining from
interbedded sands within the overlying thick aquitard and not from a shallow aquifer. The calculated
radii of the wellhead protection area based on the different hydrodynamic approaches ranged from
300 to 4,000 ft. The negligible regional gradient of the potentiometric surface did not affect the shape of
the wellhead protection area. A wellhead protection area of 1,000 ft is considered the most realistic.
The most critical pathways for potential ground-water contamination are artificial penetrations such
as wells and abandoned boreholes.
xix
-------
ACKNOWLEDGMENTS
Funding for this project was provided by the U.S. Environmental Protection Agency, Ground-
Water Protection Division, Office of Ground Water and Drinking Water, under Cooperative Agreement
ID No. CX-815385-01-0. We thank Marilyn Ginsberg, Project Manager, and Ron Hoffer from the U.S.
Environmental Protection Agency, Ground-Water Protection Division, Office of Ground Water and
Drinking Water, for their technical input and editorial reviews. Keg Alexander helped in literature
review, pump tests and data interpretation. John Burke from Aqua Water Supply Company, Bastrop,
Texas, and Wayne Popp from the City of Wharton, Texas, provided valuable hydrologic data for their
respective well fields. Original manuscript preparation was by Lucille Harrell, and additional word
processing was by Melissa Sriell and Susan Lloyd. Figures were drafted under the supervision of
Richard L. Dillon. Editorial review was by Kitty Challstrom under the supervision of Susann Doenges.
The efforts of all these people are greatly appreciated.
xx
-------
CHAPTER 1. INTRODUCTION
Nearly half the population of the United States uses ground water as its drinking water supply.
Improper management of contamination sources has resulted in numerous cases of ground-water supply
contamination. One approach toward preventing contamination of these water supplies is to protect the
areas that provide recharge to supply wells.
The 1986 Amendments to the Safe Drinking Water Act created the Wellhead Protection Program.
Through this program, the U.S. Environmental Protection Agency (EPA) assists States in protecting
areas surrounding public drinking water supply wells against contamination. The technical assistance
document, "Guidelines for Wellhead Protection Area Delineation for Confined Aquifer Settings," was
developed to provide technical information to the States in their implementation of wellhead
protection programs.
Confined Aquifers: Why Be Concerned?
Confined aquifers, by definition, are overlain by low-permeability strata. Confined aquifers are
typically less sensitive to surface contamination than water-table (unconfined) aquifers. However,
ground-water contamination has occurred in confined aquifers, demonstrating the need to protect these
sources of ground water.
In general, more confined aquifers are less sensitive to contamination than less confined or
unconfined aquifers, and less restrictive wellhead protection strategies may be appropriate. Unless the
degree of confinement of a well field is known, the potential for contamination is unknown. In some
areas an entire region can be generally characterized because hydrogeologic conditions are relatively
uniform. In other areas, however, it may be necessary to characterize the degree of confinement near
each well or well field.
Some confined aquifers have become contaminated. Confining strata are not impervious to ground-
water movement and to contaminant migration. Long-term pump tests have shown vertical flow
1
-------
through confining strata (Neuman and Witherspoon, 1972; Grisak and Cherry, 1975). Much of this
leakage may be attributable to fractures through clay and silt strata (Williams and Farvolden, 1967;
Gera and Chapman, 1988). Different types of contaminants have also been shown to migrate through
confining layers that consist of clays, silts, and glacial till (Schwartz and others, 1982; Dorhofer and
Fritz, 1988; Jackson and Patterson, 1989; Herzog and others, 1989). Downward migration of
contamination through confining layers can also occur along monitoring-well casings (Meiri, 1989) and in
naturally occurring faults (Keller and others, 1987). In Texas, Thompson and Hayes (1979) identified a
fluorocarbon plume in the confined limestone Edwards aquifer.
Purpose of Document
The purpose of this technical document is two-fold. (1) To provide a methodology to define the
sensitivity of an aquifer to contamination. This is accomplished first by determining the degree of
confinement of an aquifer, that is, whether an aquifer is unconfined, semiconfined, or highly confined,
because the more confined the aquifer, the lower the probability for its contamination. (2) To provide
approaches for delineating wellhead protection areas (WHPA's) for highly confined and semiconfined
aquifers.
Chapter 1 defines confinement. Chapter 2 explains the basic mechanics of ground-water flow in a
confined aquifer. Chapter 3 provides methods for characterizing confined aquifers. Chapter 4 describes
general wellhead protection strategies. Chapter 5 describes hydrodynamic approaches for delineating
wellhead protection areas, and Chapter 6 describes the different approaches for developing wellhead
protection areas for semiconfined and highly confined aquifer settings. Chapter 7 provides methods for
determining wellhead protection areas for well fields. Chapter 8 describes two case studies, and
Chapter 9 provides recommended approaches. A detailed description of the two case studies is included
in the appendices, as well as a short discussion on the national distribution of confined aquifers and a
glossary of important terms used in the document.
-------
Definition of a Confined Aquifer
Before wellhead protection areas are delineated for wells, the aquifer setting has to be defined
as to whether it is highly confined, semiconfined, or unconfined. Before addressing the question of
degree of confinement, more basic issues need to be addressed. What is the importance of confinement to
wellhead protection? Are general hydrogeologic definitions of confinement acceptable for wellhead
protection? The following definition is recommended in the context of wellhead protection strategies
and is referred to as the wellhead protection area definition:
"A confined aquifer is a section of an aquifer overlain by low-permeability strata that lower the
probability of ground-water contamination from surface sources" (fig. 1).
The critical elements of this definition are (1) there is a low probability of contamination from
the ground surface and (2) this low probability results from the presence of overlying low-permeability
strata. By this definition, ground water in a confined aquifer need not exist under greater-than-
atmospheric pressure, and/or rise above the top of the aquifer in wells. This definition differs from
classical definitions because its primary focus is the potential for contamination from the surface. The
wellhead protection area definition is an expansion of the definition used in the American Geologic
Institute (AGI) Glossary of Geologic Terms: "An aquifer bounded above and below by impermeable bed
or beds of distinctly lower permeability than the aquifer itself" (Bates and Jackson, 1987).
The wellhead protection area definition is preferred to classical definitions of confined aquifers
because it addresses the hydrogeologic setting that causes confinement rather than the hydrologic
phenomena resulting from confinement. It has implications about the age of the water within the
confined aquifer. If the confining unit prevents contaminants from reaching the confined aquifer, the
unit will also prevent easy movement of water to the aquifer. Geochemical indicators of absolute or
relative age, and numerical or analytical calculations of vertical leakage, provide a very important
approach for identifying confinement.
-------
Recharge nrea
TTT ,Water labio
Confined well
uncofined well
QA14883
Figure 1. Schematic of a confined aquifer (unconfined in outcrop area).
-------
The wellhead protection area definition addresses the presence of confining beds above the
aquifer only, and not "above and below" as Stated in the American Geological Institute definition,
because the dominant source (and therefore the higher probability) of contamination from a wellhead
protection perspective are from disposal practices on or near land surface.
Distinction between a Semiconfined Aquifer and a Highly Confined Aquifer
A confined aquifer can be semiconfined or highly confined. A semiconfined (leaky) aquifer (fig. 2),
as defined by the American Geological Institute glossary, is "A confined aquifer whose confining beds
will conduct significant quantities of water into or out of an aquifer" (Bates and Jackson, 1987). The
sensitivity to contamination of the semiconfined aquifer should be considered higher than that of a
highly confined aquifer because the semiconfined aquifer can receive significant quantities of water
through the confining strata.
A highly confined aquifer, in contrast, receives only minor leakage through confining strata. The
sensitivity to contamination of a highly confined aquifer is low. However, artificial penetrations such
as abandoned boreholes are potentially important pathways that may permit contaminants to pass
through the confining strata and migrate into a producing well.
Importance of Understanding Degree of Confinement in Context of Wellhead Protection
Different wellhead protection strategies are recommended for unconfined (water table),
semiconfined, and highly confined aquifers. These strategies are based on (1) the sensitivities of the
aquifers to contamination, (2) the differences in well hydraulics, and (3) the differences in the
distributions of vertical recharge.
(1) Unconfined, semiconfined, and highly confined aquifers have different sensitivities to
contamination, the water-table aquifer being the most sensitive and the highly confined aquifer being
the least sensitive. The unconfined aquifer is not overlain by confining strata to retard contaminant
-------
A xGround surface
'•.M
• '^"
•"* * ... * ' *'
""N^
r- .'-.- .v
i
-
=
1 x Water table
xX"f •"potent
Uncc
;.;V ;.-.; '•:••':. -V Lea
"ornetric
infined 2
ky aqu
"surface
iquifer
'•:'•'•:'•' :• '•
fer '/:.:
QAU884C
Figure 2. Schematic of a semiconfined (leaky) aquifer.
-------
migration; the semiconfined aquifer has an overlying confining unit, but it is leaky; and the highly
confined aquifer has an overlying confining layer that is essentially impervious to areally distributed
leakage.
(2) For the unconfined aquifer, the size of the cone of depression (COD) from a pumping well is
controlled by the recharge rate and the specific yield (storage) of the aquifer. For the semiconfined
aquifer the amount of leakage from shallower unconfined aquifers affects the size of the cone of
depression. For the highly confined aquifer the cone of depression can become very large because of the
lack of leakage.
(3) The pathways of vertical fluid movement for unconfined, semiconfined, and highly confined
aquifers also differ. In unconfined aquifers, vertical fluid movement to the water table is typically
unimpeded and areally distributed through the unsaturated zone above the water table. A
semiconfined aquifer allows leakage of significant quantities of water through the confining bed;
consequently, flow paths through the semiconfining bed are presumed to be areally distributed and may
include artificial penetrations as well as natural geologic pathways. This is in contrast to a highly
confined aquifer, in which the probability of leakage through the confining unit is very low (but not
necessarily zero). The overlying confining bed of a highly confined aquifer may contain a very small
number of discrete pathways which can include natural penetrations, such as faults and fractures, or
artificial penetrations, such as wells and abandoned bore holes.
The wellhead protection strategies for unconfined, semiconfined, and highly confined aquifers
differ for hydrogeologic reasons. The U.S. Environmental Protection Agency (1987) recommends a
variety of approaches for unconfined aquifers which include (1) time of travel (TOT), (2) zone of
influence (ZOI), that is, extent of cone of depression, and (3) zone of contribution (ZOC) approaches. For
semiconfined and confined aquifers, this document recommends either a time of travel or an integrated
cone of depression/time of travel approach.
-------
CHAPTER 2. CHARACTERISTICS OF A CONFINED AQUIFER
In this section, the typical geologic, hydrologic, and hydrochemical phenomena that are
characteristic of confined aquifers are investigated, and some of the exceptions and complexities are
discussed. Figure 1 is a schematic diagram of a confined aquifer.
Geologic Characteristics
Confining Beds
Confining beds are typically composed of low-permeability materials, composed typically of
shale, silt, or clay. Most low-permeability strata overlying large coastal plain aquifers are composed
of clay and silt. However, any low permeability bed can function as a confining stratum. Dense
limestones and dolomites, chalks and marls, volcanic lava flows, evaporite deposits (for example,
halite and gypsum beds), as well as unconsolidated sediments, may serve as confining units.
There is no established permeability range for confining strata (the term permeability is used
interchangeably with hydraulic: conductivity in this text). Permeability (hydraulic conductivity) for
sand/sandstone aquifers can range from 10~* to 102 cm/sec (10"6 to 1 ft/sec). Low-permeability rocks
typically have permeability values below 10~3 cm/sec (10~5 ft/sec). Permeability of a confining unit
typically is three orders of magnitude lower than the permeability of the producing aquifer.
Confining beds can be extremely heterogeneous, that is, permeability varies significantly in the
horizontal and vertical directions. Variability is in large part a function of the geologic setting and
geologic history of the strata. Marine shales (shales originally deposited under marine conditions)
will be relatively homogeneous, whereas continental shales may be composed of a wide range of
sediment types and, therefore, have a wide range of permeabilities. This is particularly true for
deltaic sediments, continental redbeds, and glacial deposits that may all function as confining strata.
-------
Fractures and faults may cut confining beds and greatly increase their permeability. These
structural features may be areally distributed, for example, in glacial drift in the North-Central
United States, or may only occur in discrete zones, such as a single fault zone. The density and
distribution of these features will have an important impact on degree of confinement and on the type of
wellhead protection strategy employed.
Confined-Aquifer Lithology
A confined aquifer may be composed of a variety of different lithologies. In addition, in a
confined aquifer, permeability may be heterogeneously distributed as it may be in any aquifer. For
example, a sand aquifer is not composed solely of sand; frequently, shales may be interbedded with
permeable sands or sandstones (fig. 3). This presence of low-permeability units within a permeable
aquifer may create confinement even though there is no laterally extensive overlying aquitard (fig. 3).
Furthermore, the contact between the top of an aquifer and the base of an overlying aquitard may be
transitional. Defining the top of an aquifer and the base of an aquitard may be difficult.
The geology (mineralogy, degree of lithification, type of porosity, and so forth) of the confined
aquifer may dictate some of the hydrologic and hydrochemical characteristics often associated with
confined aquifers, such as low storativity and the type of water chemistry that is associated with long
residence times or long flow paths.
Hydrologic Characteristics
Confined aquifers are hydrologically different from unconfined aquifers, as evidenced by the
nature of various hydrologic phenomena, such as elevation of the potentiometric surface, cyclic water-
level response to barometric or tidal phenomena, cone of depression, storage coefficients, and leakage
values.
-------
Land surface
Potentiometric surface
:/.v.-.;::.vA-..
EXPLANATION
High-permeability aquifer
Low-permeability confining zone
Very low-permeability bedrock
Ground-water flow
QAI4885
Figure 3. Aquifers may contain low-permeability strata that are interbedded between permeable strata
and may cause confining conditions. Ground-water production from beneath a low-permeability strata
would be from a confined aquifer even though a geologic map would show the permeable formation
cropping out, a hydrogeologic setting which traditionally would be defined as unconfined.
10
-------
Elevation of Potentiometric Surface
In an unconfined aquifer, there is direct contact between the atmosphere and the ground water
along the entire upper surface (water table) of the saturated section; in comparison, the potentiometric
surface of a confined aquifer (the surface defined by the elevation to which water rises in wells that
are open to the atmosphere) is often above the top of the aquifer. The potentiometric (piezometric)
surface of a confined aquifer may rise above the land surface resulting in flowing (artesian) wells (fig.
1). The reason for this is described next.
Ground water flows in an aquifer from zones of recharge to zones of discharge. The elevation of a
water level in a well represents the potential energy of the ground-water system at that well. Water
flows from higher potential energy to lower potential energy; the highest potential occurs in the
recharge zone and the lowest potential occurs in the discharge zone. The system loses its potential
energy by frictional loss (resistance) as it flows through the aquifer, as expressed by Darcy's law:
q = Ki (1)
where q = the ground-water flow rate,
K = the hydraulic conductivity, and
i = the hydraulic gradient.
In the simplest situation, where aquifer permeability is uniform and flow rate is constant, the potential
energy (head) loss is constant and the potentiometric surface has a constant gradient (fig. 1). A more
complex scenario results when the permeability of the aquifer varies. In coastal plain aquifers,
continental sands/sandstones are interbedded with marine or deltaic shales. Relatively permeable
fluvial sandstones at the outcrop become interbedded with deltaic or marine shales downdip, resulting
in overall average lower down-gradient permeability. According to equation (1), the hydraulic
gradient is inversely proportional to hydraulic conductivity; that is, for a given flow rate, steeper
11
-------
head gradients are required for ground water to flow through a low-permeability zone compared to
ground-water flow through high-permeability zones.
In early days of ground-water exploitation of confined aquifers such as the Dakota sandstone
aquifer, South Dakota, or the Gulf Coast aquifer in Texas, many wells flowed at land surface because
the aquifer was under artesian conditions. Artesian conditions often indicate a confined aquifer setting.
Water elevations below the top of an aquifer do not mean that the aquifer is unconfined. Water
elevations below the top of a confined aquifer may occur naturally or artificially. A potentiometric
surface below the top of a confined aquifer can occur if an aquifer is more easily discharged than
recharged. This phenomenon is being recognized in some of the aquifers in the western United States.
Potentiometric surfaces below the top of confined aquifers may occur locally and regionally
because of ground-water production. Cones of depression from individual pumping wells may result in a
potentiometric surface being beneath the top of an aquifer. Similarly large-scale, regional, long-term,
ground-water production for agricultural and municipal use, such as the San Joaquin Valley, California,
or the greater Houston, Texas, region may result in the regional lowering of a potentiometric surface
that, through time, drops below the top of an aquifer. In the context of the WHPA definition of
confined aquifers, such aquifers are considered to be confined.
Direction of Vertical Ground-Water Flow
The relative elevations of the potentiometric surfaces of a confined aquifer and an overlying
water-table aquifer define the direction of vertical ground-water flow, indicating whether potential
contaminants can migrate from the water-table aquifer to deeper confined aquifers. The direction of
vertical leakage between an unconfined and a lower aquifer is dependent upon whether the
potentiometric surface for the deeper confined aquifer is above or below the upper aquifer's water table.
If the potentiometric surface for the confined aquifer is above the water table, then there is a potential
for upward flow from the deeper aquifer (fig. 4a). Upward flow implies that contaminants cannot move
12
-------
(a)
Inadequately cemented well
Inadequately cased well V
Potentiometric surface in
production/injection zone Production/injection
/ / well
Water table in
unconfined aquifer
y
Unconfined
aquifer
Upward
leakage
^Confining unit;
I
:Fractures =
Inadequately cemented well
Production/injection
well
Unconfined
aquifer
Water table in
unconfined aquifer
Potentiometric surface in
production/injection zone
~* -*. JL
QA14886C
Figure 4. (a) Confined aquifer where the potentiometric surface is higher than the water table of the
overlying unconfined aquifer. The potential for ground- water flow is upward, (b) Confined aquifer
where the potentiometric surface is lower than the water table aquifer. The potential for ground-water
flow is downward. Downward flow is needed for contaminants to migrate from a shallower unconfined
aquifer to a deeper confined aquifer (from U.S. Environmental Protection Agency, 1987).
13
-------
from the shallow to the deep. If the potentiometric surface for the confined aquifer is below the water
table, there is potential for downward flow and, thus, a potential for contamination (fig. 4b).
Downward flow can occur around a well when a cone of depression from a pumping well is lower
than the water table of an upper aquifer. Downward flow can also occur regionally as a result of a
naturally lower potentiometric surface or because of long-term regional ground-water production.
Vertical leakage may contribute a significant percentage of the overall flow of water to an
aquifer on a regional and well field basis. Even though the vertical permeability per unit area of an
aquitard may be low in comparison to the permeability of an aquifer, there may be significant vertical
leakage to the aquifer because of the extensive lateral area of the aquitard in comparison to the
thickness of the aquifer.
Rates of leakage can be calculated by using an equation similar to that for calculating horizontal
flow, that is, by using Darcy's law. Leakage can be defined as
qv = K'(h0-h)/b/ (2)
where qy = rate of vertical leakage per unit area
h0 = water level for the confined aquifer
h = water level at the water table
K' = vertical hydraulic conductivity
b' = thickness of aquitard.
The rate of vertical leakage per unit area is controlled by the vertical hydraulic conductivity of the
aquitard and the hydraulic gradient across the aquitard. K' values are often given as gpd/ft2 or cm/sec.
No one has compiled a range of leakage values, but K' values greater than 10~2 gpd/ft2 (5 x 10~7 cm/sec)
generally will permit significant leakage across the aquitard. The rate of vertical leakage is an
important consideration in differentiating highly confined from semiconfined aquifers.
14
-------
Flow Velocity and Age
Ground water in confined aquifers commonly has a low hydraulic gradient, a low ground-water
flow velocity, and contains relatively old water. Figure 5 compares the potentiometric surface of the
Floridan aquifer with the degree of confinement. Where the low-permeability confining unit is present,
the potentiometric surface of the Floridan is relatively flat compared with the gradient of the
Floridan in northern Florida where the aquitard has been eroded. In confined aquifers of the coastal
plain, hydraulic gradients are very low (<0.0001) and flow velocities may be in the range of 1 to 50 ft
per year. Flow velocities in the Carrizo aquifer, a typical sandstone aquifer dipping toward the Gulf of
Mexico in the Texas Coastal Plain, range from 5 to 30 ft per year with the higher rates in the outcrop
area (Pearson and White, 1967).
Ground water in confined aquifers may be very old because of low velocities. Kreitler and Pass
(1980) identified, with UC, waters that were 5,000 to 15,000 yr old in the updip section of the Wilcox
aquifer, a large Tertiary-aged sandstone formation in East Texas. Pearson and White (1967) measured
water ages of 25,000 yr 20 mi downdip in the Carrizo aquifer in South Texas. Ages of waters in the
confined section of the Chalk aquifer, where it underlies the London Clay of the London Basin
(England) exceed 25,000 yr (Smith and others, 1976). Ground waters from a confined aquifer in
Hermosillo, Mexico were estimated to be 30,000 yr old (Payne and others, 1978).
Storativity
The storativity of an aquifer is defined as the unit volume of water that a unit volume of aquifer
releases "from storage" under a unit decline in hydraulic head (Freeze and Cherry, 1979). For a confined
aquifer with the potentiometric surface above the top of the aquifer, this release of water results from
the compressibility of the aquifer material and a slight expansion of water. In response to a decline in
head, compressible aquifers (unconsolidated sands with interbedded clays) release significantly more
15
-------
85"
Aquifer tyttem unconflned or
nearly w
Aauiter tystem lemi-conflnedi
upper confining unit leu than
JO m thick or breached
Aquifer »y»tem confined i
upper confining unit greater than
Approximate updip llmil of Floridan
Qquiftr lyitem
100 200km
Figure 5. Comparison of potentiometric surface of Floridan aquifer to unconfined, semiconfined, and
confined sections of the Floridan aquifer. The potentiometric surface becomes flatter where the
Floridan becomes highly confined (modified from Johnson and Miller, 1988).
16
-------
water than noncompressible aquifers (limestones and sandstones). In contrast to confined aquifers,
water-level declines in unconfined aquifers cause drainage of water from the pore spaces, that is, the
saturated section becomes thinner. The storage term for unconfined aquifers is referred to as specific
yield.
The release of water from storage for either confined or unconfined aquifers results from a decrease
in head values, for example, as a result of the pumping of a well. The water released by drainage of
pores spaces in an unconfined aquifer is significantly greater than the water released by compressing
the pore spaces in a confined aquifer. Specific yield for unconfined aquifers ranges from 0.30 to 0.01
(Freeze and Cherry, 1979). Confined aquifers commonly have low-storativity values compared to
unconfined aquifers. Storativities for confined aquifers commonly range from 0.005 to 0.00005. However,
the storativity values for very compressible aquifers, characterized by clay compaction, approach
specific yield values for unconfined aquifers. Storativity often is used as a method to differentiate
confined from unconfined aquifers.
Cyclic Water-Level Responses Resulting from Atmospheric Pressure Changes
Water levels in wells of confined aquifers typically exhibit small cyclic changes in elevation,
which may occur with a frequency of once or twice a day. Water levels in wells of unconfined aquifers
typically do not show such a daily cyclic change in elevation. Cyclic responses of the water levels in
wells result from changes in overburden pressures (ocean tides), dilation of the aquifer (earth tides) or
changes in atmospheric pressure at the well bore. Atmospheric pressure changes probably have the
greatest impact on water levels because of the magnitude of the changes and their widespread
occurrence. The water elevation in a well is the elevation to which the water will rise to equilibrate
with atmospheric pressure. Changing weather systems (high pressure and low pressure cells) can cause
atmospheric pressure changes (fig. 6). In addition, atmospheric pressures change continually
throughout the day as a result of heating and cooling of the atmosphere (fig. 6). In a confined aquifer,
the only point where the potentiometric surface is in direct contact with the atmosphere is in the well
17
-------
18 20 22 24 26 28 30 2 4 6 8 10
March, 1939 April, 1939
Days of the month
Figure 6. Weather-related barometric changes and their effect on the water levels in a well
penetrating a confined aquifer (modified from Todd, 1980). Reprinted by permission of John Wiley and
Sons, Inc., New York, New York.
18
-------
bore. Increases in atmospheric pressure will force the water elevation in the well down. Decreases in
atmospheric pressure will permit the water elevation in the well to rise. The rest of the aquifer will
not respond to this change in atmospheric pressure because the overlying aquitard acts as a rigid cover.
Only water levels in water wells open to the atmosphere respond to atmospheric pressure changes. In
contrast, the water table in an unconfined aquifer is in contact with the atmosphere everywhere;
therefore, atmospheric pressure changes are transmitted equally to the water table and not just to the
well; therefore, water elevation in a well does not show daily water-level fluctuations with daily
pressure changes (fig. 7). The presence of these small cyclic water-level changes can be used to
differentiate confined from unconfined aquifer settings.
Cone of Depression
During the pumping of a water well, water levels drop and a cone of depression of the
potentiometric surface develops around a well. The water produced from a well in a confined aquifer
comes from three sources: (1) water flowing laterally from the aquifer into the well; (2) water flowing
vertically from aquitards above or below a producing aquifer. This water either originates from within
the aquitard (aquitard storage) or from leakage through an aquitard; and (3) from storage in the
producing aquifer (fig. 8). In an aquifer with a negligible regional hydraulic gradient, the perimeter of
the cone of depression defines the boundary, at a given time, of the areal extent of the lateral flow in
the aquifer and of vertical flow from adjacent confining units.
A graph of water-level decline, resulting from ground-water pumpage from a highly confined
aquifer, follows a characteristic curve known as the Theis curve and has a generally asymptotic shape
(fig. 9). The only source of water from a highly confined aquifer is the water flowing laterally to the
well. Because there is no vertical leakage, the cone of depression must continue to enlarge over time, and
water levels will continue to decline even after long periods of time. For semiconfined aquifers, the
drawdown of water levels and the lateral extent of the cone of depression stops when the amount of
vertical leakage equals the well discharge. A series of leaky aquifer curves can be used to calculate the
19
-------
21
23 24 ' 23
May 1987
26
Figure 7. Example of daily water-level changes in two wells from the Edwards aquifer, Georgetown,
Texas. The cyclic water-level curve for well 58-27-305 shows two maximum values per day that are
related to barometric changes and exhibit confined aquifer response. The flat water-level response for
well 58- 27-210 exhibits an unconfined aquifer response and shows longer term water- level declines
from local pumpage (modified from Senger and others, 1990).
20
-------
Water table
£ -50
Q.
-------
102
•a
£
10
-2.
0.001
„-. x -0.005
0.05 °-03 \N0.010
X0.015
r/8 = 2.5
10'
1.0
101
102
103
1/H
i I i r ir i ( i
10*
105
10e
107
QAU906C
Figure 9. Theis curve and leaky aquifer curves (from Todd, 1980). Reprinted by permission of John Wiley
and Sons, Inc., New York, New York.
22
-------
amount of leakage (fig. 9); the greater the leakage, the greater the r/B value for the different curves.
For very leaky systems, drawdown can be minimal, and water levels will stabilize rapidly. Figures 10,
11, and 12 show pump-test data for a highly confined aquifer, a moderately leaky semiconfined
aquifer, and a very leaky semiconfined aquifer.
The flow of ground water toward a well is related to changes in head caused by pumping the well.
Horizontal head gradients toward the well permit lateral flow to the well; in the case of confined
aquifers, vertical head gradients across aquitards permit vertical leakage and water from aquitard
storage; and head changes permit compression of the aquifer and the "squeezing" out of water from
aquifer storage. There is no flow to the well from areas where there is no vertical or horizontal head
gradient toward the well. This simple statement offers an important insight toward understanding the
area that contributes water to a producing well. The aquifer external to the cone does not contribute to
the water produced at the well assuming there is no, or a negligible, regional gradient. Once the cone
has stabilized, theoretically, there is no contribution of water from storage. There is no longer any
change in water levels with time and, therefore, no additional compressing and squeezing of water out
of the aquifer. All of the contribution of water comes from vertical leakage.
Leakage through an aquitard has been observed. Neuman and Witherspoon (1972) conducted a 31-
day aquifer test in the Oxnard aquifer, Oxnard, California. The confined Oxnard sand and gravel
aquifer is overlain and underlain by aquitard/aquifer pairs (fig. 13). Monitoring wells were installed
and monitored in the three aquifers and two aquitards. By the end of the 31-day aquifer test, water
levels had dropped in the producing aquifer as well as the aquitards and in the overlying and
underlying aquifers. Vertical permeability was estimated at 2.9 x 10~2 gpd/ft2. This example
graphically demonstrates that leakage from overlying or underlying aquifers does occur and that
contamination through an aquitard can occur.
23
-------
100=1
S. 10:
C '
z :
o
•o
9
ie
•o
2
o
2 1.0:
o :
o
0.1
Q = 120 gpm
r = 2,450 ft
1/u =10
s = 7.25 ft wall, S.D.
pump-test data
10
_ 114.6 Q .... . 114.6(120) .... .„_ ....
T = • W(u) = —-^—i"(10) = 1895 gpd/ft
S /.
-------
10.
c
o
T3
1 -
Q = 25 gpm
r = 96 ft
m = 8 ft
m1 = 14 ft
Nonleaky artesian-type curve trace.. _
Data point.
' Leaky artesian-type
curve trace
f~ Match point
A
-------
E .1
'.01 -
.001
10
10Z
103
104
t (s)
10s
106
M Q - 4.0 x 10'3 m3/s (63 U.S. gal/min)
~ r = 55m (180 ft)
~ b1 -30.5 m (100ft)
K, -7.4 x 10-5 m/s (157 gpd/ft2)
Ss,. 9.0x10-6
K' -2.4x10-6 m/s (5.0 gpd/ft2)
b'-3.05 m (10ft)
QA14909C
-.01
Figure 12. Example of pump test (drawdown versus time) for very leaky unidentified aquifer (from
Freeze and Cherry, 1979). Reprinted by permission of Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
26
-------
\"/ Spontaneous
potential
Resistivity
100-
200-
Q.
0>
Q
300-
400 J
Semiperched
aquifer
Upper aquitard
Oxnard aquifer
Lower aquitard
Mugu aquifer
(b)
10*
10'
c
I
I
10'
,0-
10"
10"
10'
No well storage capacity,
nonleaky artesian-type
curve trace
*• t&
$• 8
^:^
'.$
102
10*
Time after pumping started (mm)
10s
QA14910C
Figure 13. Geologic setting and pump test data from confined Oxnard aquifer, overlying and underlying
aquitards, and overlying and underlying aquifers (from Neuman and Witherspoon, 1972). Example
shows that there is leakage through an aquitard.
27
-------
Hydrochemical Characteristics of Ground Water in Confined Aquifers
Hydrochemical characteristics of ground water typically reflect aquifer lithology and residence
time of ground water. Because of the large geographic extent of many confined aquifers, ground water
within such an aquifer may be relatively old and may have traveled over relatively long distances.
Both age and distance of travel control the chemical and isotopic composition of the waters. The
chemical composition of ground water typically changes as it flows from zones of recharge to zones of
discharge. Recharge zones for confined aquifers are typically oxidizing, have low pH levels, and
relatively high concentrations of nitrate, sulfate, and calcium. As ground water flows downdip, it
becomes more reducing, typically shows an increase in pH, and its total dissolved solids (TDS)
concentrations increase. Nitrate (NOs) and sulfate (804) concentrations decrease significantly, calcium
(Ca) decreases, and sodium (Na) and bicarbonate (HCOa) concentrations increase (Back, 1966; Kreitler
and others, 1977; and Fogg and Kreitler, 1982). Figure 14, a cross section through the Atlantic Coastal
Plain, New Jersey, shows the evolution from a low-total-dissolved-solids mixed-composition water in
the recharge zone to a Na-HCC>3 to Na-Cl water downdip. If the general chemical evolutionary
pathway is known the chemical composition of an individual sample can be used to determine whether
the water came from the recharge zone or from the downdip confined section.
As the water flows down gradient from the recharge zone it also becomes progressively older.
Tritium (3H) concentrations will decrease to zero as the tritium (short-lived radioisotope of hydrogen
in water with a half-life of 12.3 yr) disappears by radioactive decay. Presence or absence of tritium can
be used to indicate whether a water was recharged more or less than approximately 40 yr ago (fig. 15).
Anthropogenic chemicals in the ground water also provide an assessment of the age of the water. The
occurrence of contaminants in a ground water, such as fluorocarbons, nitrates at high levels, and
synthetic organic compounds, also indicates the addition of relatively young waters. Carbon-14
concentrations decrease as ground water flows downdip and becomes older. The age of ground water that
is in the range of thousands of years can be estimated with 14C analyses.
28
-------
A'
Sea level -
-2,000 -
£• -4,000 -
"CD
nj
«
CO
0)
.0
Q.
«
Q
-4,000
QA14911C
Figure 14. Evolution of hydrochemical facies from variable composition from Ca-HCO,. to Na-HCO, to
ij •Cf
a Na-Cl for ground-water flow in the Atlantic Coastal Plain, New Jersey (from Meisler and others,
1988).
29
-------
400-
Fresno
~ -400 H
.O
ro
LU
-800-
-1200-
-1600-
- 100
Figure 15. Example of tritium in ground water, Fresno County, California (Poland and Stewart, 1975).
30
-------
Large-scale pumpage may alter the hydrochemistry of the ground water in a confined aquifer.
Extensive and long-term pumpage may result in increased leakage through confining aquitards and
subsequently alter the chemical composition of the ground water. A water sample collected from a
natural system typically represents ground water that flowed from the outcrop to the point of
collection. In contrast, a water sample collected from a well field that has been pumped at high
volumes continually for 40 yr (as an example), may in fact result from leakage through overlying
aquitards. This sample may have a different chemical composition and may be significantly different
in age from the water sample collected from the natural system.
31
-------
CHAPTER 3. APPROACHES FOR DETERMINING THE PRESENCE AND/OR
THE DEGREE OF CONFINEMENT
Confined aquifers are less sensitive to ground-water contamination from overlying contaminants
than are unconfined aquifers. It is not simple, however, to determine whether a well or well field under
investigation is producing from an unconfined, semiconfined, or confined aquifer. As discussed in the
previous section on the characteristics of confined aquifers, there are several characteristics that can be
used to test for the presence and/or degree of confinement. The prime concerns in determining the
presence and/or degree of confinement are to evaluate the sensitivity of the aquifer to potential
contamination and to identify the potential pathway for contaminants migrating to a producing well.
The methods listed below can be used to describe (1) the presence or absence of confinement, (2) the
presence and degree of confinement (semiconfined versus highly confined), or (3) the degree of
confinement after the presence of confinement has already been identified. Many of the methods,
however, only identify the presence of confinement and not the degree of confinement because we often
measure only the hydrologic, geologic, or hydrochemical phenomena that are caused by confinement
and not the amount of leakage or the zones of leakage. We are limited in our techniques for delineating
highly confined from semiconfined settings and particularly in quantitatively determining the degree
of confinement.
The techniques described below can be used for assessing the presence and/or degree of
confinement. There are three basic approaches for identifying the presence and/or degree of
confinement: geologic, hydrologic, and hydrochemical. Each basic approach can be divided into
different techniques. Geologic techniques identify the presence of confining strata, their spatial
distribution, and their physical characteristics. Because some geologic techniques identify breaches in
confining strata, the degree of confinement can be inferred. Hydrologic techniques identify whether the
aquifer is confined and, for some techniques, the degree of confinement. Hydrochemical techniques
indicate absolute or relative ages of waters, which can in turn be used to infer presence and/or degree of
confinement.
32
-------
The presence and/or degree of confinement should be considered in planning future areas for
ground-water production as well as for safeguarding present water supplies. Highly confined aquifers
will be inherently less susceptible to future contamination than will be unconfined or semiconfined
aquifers. Mapping techniques are of particular benefit to planning and protecting future water supplies
because of the inherent capability of maps to project and infer hydrogeologic properties into areas for
which there are no data.
Geologic Approach
The geologic approach includes several techniques that identify the presence of a confining bed
overlying an aquifer and define the physical characteristics of the bed. These techniques identify the
thickness and areal extent of an aquitard and indicate potential permeability pathways which may
permit contaminants to leak through a confining unit.
Classic Geologic Maps
Geologic maps have been used to determine confinement by depicting geologic formations. A
formation is commonly composed of one predominant lithology, such as shale, limestone or sandstone,
but often other rock types are included. Formations on geologic maps can be interpreted by
hydrogeologists as being aquifers or aquitards, based on the formations' dominant lithologies and on
the estimated ability to produce ground water. Aquifers are often considered to be unconfined because
they crop out, or to be confined because they dip beneath a formation of lower permeability.
Outcrops, soil maps, aerial photographs, and borehole information (electric logs and driller's
logs, for example) are the general types of data that are used for constructing geologic maps delineating
confined aquifers. Many areas have been geologically mapped so published information may be
available. Surface geologic mapping is routinely based on mapping of geologic formations in outcrops.
Outcrop mapping should be supplemented with an aerial photograph interpretation to assist in the
33
-------
mapping of areal distribution of geologic formations. Fractures and faults in confining strata are
important potential pathways for vertical flow and may be identified through aerial-photographic
mapping that is verified in the field. Observation of fracture openings, and mineralization or oxidation
along fractures, indicate that the fractures are a pathway for flow (Grisak and Cherry, 1975). All
mapping approaches provide a two-dimensional, surface picture of the confining unit. They do not
provide any subsurface information.
Environmental Geologic and Hydrogeologic Maps
Environmental geologic and hydrogeologic maps are a subset of classic geologic maps. Instead of
depicting geologic formations, environmental geologic maps typically address a broad range of
environmental issues. For example, in areas where floods are a primary concern flood-prone areas could
be mapped. Hydrogeologic maps typically address only important aspects related to the underlying
ground water. For confined aquifer settings, hydrologic criteria related to confined settings, such as
lithology, faults and fractures, boreholes and wells, and so forth, should be depicted on hydrogeologic
maps. These types of data are available from geologic maps, soil maps, topographic maps, aerial
photographs, borehole information (electric logs, driller's logs), and water-level records, and are
available from organizations, such as the U.S. Geologic Survey (USGS), State geological surveys, State
water and environmental agencies, State public health departments, university geology and civil
engineering departments, regional planning entities and councils of governments, and private
consultants. The technique indicates the presence or absence of confinement to provide information on
the degree of confinement. Geologic data need to be integrated with hydrologic and hydrochemical
data.
Artificial penetration maps are a subset of hydrogeologic maps. A critical pathway for
contaminants to migrate through normally impenetrable confining strata, may be through artificial
penetrations such as abandoned or producing oil and gas wells, abandoned or producing water wells,
seismic shot holes, injection wells, or any other excavations that might breach a confining stratum.
34
-------
Examples of contamination via abandoned wells have been documented by Gass and others, 1977;
Fairchild and others, 1981; Wait and McCollum, 1963. Figure 16 shows the density of abandoned oil and
gas wells in one oil-producing county in West Texas. Anzzolin and Graham (1984) estimated that
approximately 1.65 million abandoned wells exist in the United States. Penetrations may be cased
(abandoned water wells, for example) or uncased (abandoned mineral or oil exploration holes that were
never plugged). Uncased holes in lithified bedrock generally do not collapse and, therefore, remain
open long after abandonment. Uncased boreholes in unconsolidated sediments may collapse from earth
pressures and may be less of a problem. Cased boreholes generally remain open for a long time. Many
uncased and abandoned boreholes may still contain drilling mud which may limit the amount of fluid
flow within the borehole. The amount of leakage down artificial penetrations is difficult if not
impossible to calculate. For this document it is assumed that leakage can occur through an artificial
penetration such as an unplugged borehole. Therefore, any artificial penetration represents a point for
potential vertical migration of contamination.
Mapping the location of artificial penetrations may be extremely difficult. Maps of artificial
penetrations can be produced from a variety of data sources. Maps that depict all known artificial
penetrations generally are not available because such maps would require the mapping of penetrations
associated with different uses. Maps depicting water wells may be available from State water
agencies. Locations of oil and gas wells and other wells used in the mineral industry may be available
from other State agencies regulating water, oil and gas, and the mineral industry. This will vary from
State to State. Abstract companies have ownership maps that may show the location of oil and gas
exploration wells. Many abandoned boreholes, however, may predate State regulations requiring
reports on the exact location and the plugging of artificial penetrations. Field mapping may require
surveys with metal detecting equipment (for example, electromagnetic, resistivity, and magnetic
techniques), aerial photographs, and interviews with present and past landowners. Door-to-door
inventories may be the most effective way to locate artificial penetrations. Uncased, abandoned
boreholes have no electrical signature and may be impossible to find. Hydrologic techniques that may
identify boreholes include (1) monitoring ambient water levels to identify potentiometric highs
35
-------
EXPLANATION
Abandoned borehole
0
I—
0
15
20 mi
10
30 km
QA8328
Figure 16. Map of Tom Green County, Texas, showing locations of abandoned oil and gas exploration
boreholes (from Richter and others, 1990).
36
-------
resulting from discrete points of leakage, (2) injecting water into an aquifer and looking for occurrence of
flowing wells, and (3) pump-testing analysis to identify discrete points of leakage (Aller, 1984;
Javandel and others, 1988). These hydrologic techniques have not been tested in the field.
Subsurface Geologic Maps
Construction of subsurface maps from geophysical logs or driller's logs in the vicinity of a water
well or well field provides a "depth" perspective as to the distribution of low-permeability layers,
which may provide confinement, but may not be evident from surface geologic information at the well or
in the outcrop. When subsurface maps are integrated with surface geologic maps, they provide a three-
dimensional picture of the distribution of confining beds. Well logs are routinely used for determining
the best ground-water producing interval, but generally have not been used to define presence or absence
of confining zones for the purpose of aquifer protection. Geophysical logs can be used to map low-
permeability strata above and within aquifer units. A well log at a specific well or well field provides
particularly relevant data. Where more abundant data are available, cross sections and map views of
structures can be constructed, and thickness of an aquitard and presence of structural and lithological
discontinuities can be determined. Integration of surface geologic maps with subsurface geologic and
hydrologic information allows better assessment of confining conditions.
Hydrologic Approach
The hydrologic approach includes several techniques that generally define whether an aquifer is
confined or not. These techniques include water elevation in a well, potentiometric surface maps, pump
tests for storativity, pump tests for leakage response, continuous water-level responses, hydrologic
measurements in confining strata, and numerical models. Most of the approaches measure or
characterize a hydrologic response within the aquifer. Only two approaches, pump test for leakage
37
-------
and hydrologic measurements in confining strata, evaluate the hydrologic characteristics of the
confining strata itself.
Water-Level Elevation in a Well
Determining the presence of confinement by the elevation of a water level in a well represents one
of the simplest methods for determining confinement. If the water level is above the top of the aquifer,
then the aquifer is confined (fig. 1). Appropriate water-level measurement data may exist or may have
to be collected. Methods of measurement are steel tapes, electric lines, and air lines. Confined aquifers
in which water levels are naturally below the top of the aquifer or in which water levels have
declined below the top of an aquifer because of short-term or long-term pumping, are still considered
confined because of the presence of an overlying low-permeability layer. However, this technique will
not identify these aquifers as confined.
Potentiometric Surface
A potentiometric profile is the line or surface defined by the interpolation of water-level
measurements in different wells (fig. 1). This technique is similar to that previously described for
"Water-Level Elevation in a Well," except the potentiometric surface technique requires the use of
several wells over the area of interest. This technique has the additional capability of determining
how water levels in one well interrelate with other well water levels in the area. A single datum point
often provides little insight into a hydrologic phenomenon. As more data are incorporated in a
potentiometric surface, the presence of confinement can be examined in greater detail. This technique
will not identify confined aquifers in which the potentiometric surface is below the top of the aquifer;
nor will this technique determine the degree of confinement.
38
-------
Pump Test for Storativity
Storativity values can be used to determine whether an aquifer is confined or unconfined, but
should not be used to assess the degree of confinement. Storativity values for confined aquifers are
generally 10~3 or less, whereas Storativity values for unconfined aquifers are 10~2 or greater. The
average Storativity for the Ogallala aquifer, a major unconfined aquifer in the High Plains of Texas, is
.08, whereas the average Storativity for the Gulf Coast aquifer, the major confined aquifer along the
Texas Gulf Coast, is .0009 (compiled from Myers, 1969). The low Storativity values for confined aquifers
result from compression of the aquifer matrix and the concomitant decrease in pore space. The higher
Storativity values from unconfined aquifers result from drainage of pore space. In highly compressible
confined aquifers, such as coastal aquifers that contain interbedded clay strata characterized by high
porosity and compressibility, storage coefficients may approach unconfined values and may not be
characteristic of typically confined aquifers.
Storativity values can be calculated from water-level changes in observation wells during
pumping tests using the Theis nonequilibrium equation or other equations that are modifications of the
Theis equation. Monitoring wells for drawdown observations, however, may be difficult to find because
municipalities often will not have closely spaced wells producing from the same water-bearing
horizon.
Pump Test for Leakage
If drawdown data from an aquifer pump test exhibit leakage, leaky-aquifer solutions can be used
to calculate vertical leakage through an aquitard. The likelihood of an aquifer to receive leakage can
be reasonably well assessed when such information is integrated with a detailed geologic description of
the confining strata. Presence of significant leakage can be determined from the general shape of the
drawdown versus time curve. Figure 10 shows an aquifer test for a nonleaky aquifer, figure 11 shows
39
-------
moderate leakage, and figure 12 shows significant leakage. Leakage can be the result of higher
permeability areas in the confining bed and/or natural or human-induced breaches of the confining
strata.
Long-term pumping-test data may be needed to observe when the change in drawdown
approaches zero, which is characteristic for leaky conditions. Data from observation wells are needed
to quantify rates of leakage because the effects of well loss could impact drawdown in the pumping
well. Estimated leakage values for all aquifers range from 102 to 10~5 gal/day/ft2. These lower values
(10~5 gal/day/ft) approach highly confined conditions with no leakage. Vertical leakage values for
semiconfined aquifers are considered to range from 10~2 gal/day/ft2 to 102 gal/day/ft2.
Calculation of vertical leakage through confining strata probably represents the best hydrologic
method for determining potential for contamination and for delineating highly confined from
semiconfined aquifers. All calculations from pumping-test data, however, represent measurements of
averaged hydrologic properties. Unless the permeability contrast between the pathway of leakage
and the rest of the aquitard is significant, discrete points of leakage probably cannot be seen from
aquifer response. Leakage from confining strata may represent a significant part of the ground water
pumped from a well. Leakage does not necessarily originate from a shallow unconfined aquifer which
may be a potential source for contamination, but may come from storage within the aquitard (Hantush,
1960; Neuman and Witherspoon, 1969a, 1969b, 1972). If the aquitard represents a complex interbedding
of sands and shales, then the source of the water may come from the drainage of the interbedded sands.
A more accurate picture of leakage through an aquitard can be made by installing monitoring wells in
the aquitard itself to see how they respond to pumpage from the confined aquifer (Neuman and
Witherspoon, 1972).
There are several papers on theoretical analysis of leaky aquifers (Hantush, 1959, 1960; Walton,
1962, 1979; and Herrera and Figueroa, 1969; Herrera, 1970; Neuman and Witherspoon, 1969a, b, 1972;
Lai and Su, 1974 ). Calculation of leakage values for well fields, however, is not routine. There is
limited information on which hydrogeologists can base their analyses.
40
-------
Continuous Water-Level Responses
Continuous water-level elevation data can provide a simple and cost-effective method for
determining whether an aquifer is unconfined or confined. Continuous water-level data for confined
aquifers show daily fluctuations of water levels in wells because of daily atmospheric pressure
changes. Water levels in wells of an unconfined aquifer will not show these natural, daily fluctuations
(fig. 7). Major, longer term pressure changes, such as atmospheric pressure changes with weather
changes, will also cause similar effects in wells of confined aquifers. Water-level response of confined
aquifers to recharge events may be significantly different from those in unconfined aquifers. Recharge
to confined aquifers through points of discrete leakage may indicate relatively rapid and large water-
level changes, whereas water-level response in an unconfined aquifer is typically of a smaller
magnitude.
Water-level fluctuations associated with barometric or earth-tide variations are relatively
small and must be measured with equipment that is sensitive enough to measure centimeters of change
and record at least every two hours. Drum recorders with floats or pressure transducers have the
sensitivity and short time interval between measurements needed for these types of measurements.
Measurement periods of at least one day are needed to observe daily fluctuations. Longer term
measurements are needed to observe possible effects of recharge associated with precipitation.
Interpretation of continuous water-level recorder data is a sensitive technique for determining the
presence of confinement, but cannot be used for assessing the degree of confinement. The use of continuous
water-level recorder data for defining confinement may be most appropriate as an initial screening tool
to determine whether an aquifer is confined.
41
-------
Hydrologic Measurements in Confining Strata
The hydrologic characteristics of a stratum suspected of being confining can also be determined by
monitoring hydrologic processes within the stratum itself. Other hydrologic approaches assess the
presence and/or degree of confinement by measuring hydrologic processes in the confined aquifer
beneath the confining stratum. Water-level changes in overlying strata during pumping in an aquifer
indicate communication with the producing aquifer (Neuman and Witherspoon, 1972; Grisak and
Cherry, 1975). Diurnal water-level fluctuations in overlying strata indicate confinement. Conversely,
seasonal water-level changes that correlate to seasonal variations in precipitation suggest leakage
(Williams and Farvolden, 1967).
Hydrologic measurements of leakage through an overlying strata are difficult to make because of
the problem of identifying locations where the leakage is occurring. Permeability pathways through a
suspected aquitard typically are vertical, making monitoring wells particularly difficult to place. The
location and number of monitoring wells should be based on geologic mapping so that monitoring wells
can be installed in leakage locations.
Monitoring wells in overlying strata can be used to test the confining nature of the strata, as well
as to monitor for specific contaminants migrating through the strata. Monitoring of suspected aquitards
is expensive compared with the other techniques described.
Numerical Modeling
Numerical modeling is a sophisticated technique that can be used to determine whether an
aquifer is confined and the degree of confinement. The hydrologic characteristics of confining strata are
estimated by altering hydrologic parameters (referred to as parameter estimation) of the confining
strata and then simulating observed potentiometric surfaces. By estimating vertical leakage in the
confining strata the degree of confinement can be estimated. A numerical model is an excellent method
42
-------
for synthesizing all available geologic and hydrologic information into a comprehensive picture.
Creating a numerical model solely for defining confinement is probably more than is needed for
determining whether an aquifer is confined and the degree of confinement. The previously discussed
techniques are more cost effective for defining confinement.
A numerical model can be of great value in delineating a wellhead protection area. If a numerical
model is to be developed for evaluating wellhead protection areas, then it may also be appropriate to
use the model for determining the degree of confinement. Van der Heijde and Beljin (1988) give a
compilation and review of numerical models appropriate for hydrogeologic characterization and
development of wellhead protection areas.
Hydrochemical Approach
Hydrochemical techniques identify the age of ground water or the flow distance of water within
an aquifer. With general water-chemistry data, we can determine if well water is characteristic of the
recharge zone or of the down-gradient confined section of an aquifer. With radioactive isotopes we can
estimate the age of the water and the approximate time when the water was recharged at land
surface. The sensitivity of an aquifer to contamination can be estimated with the following water
chemistry approaches.
General Water Chemistry
For large coastal plain aquifers with both outcrop and downdip sections, it may be difficult to
determine if a well is located in the recharge zone or downdip in a confined section. This is especially
true in the transitional area between outcrop and downdip sections. Ground-water chemistry may help
determine whether a well is located in an unconfined recharge zone or in downdip confined sections. In
coastal plain confined aquifers, waters in recharge zones are characterized by low pH, high eH, low
TDS, high Ca/Na ratios, low HCOa, low Cl, some NC>3, and some 804. As these waters flow downdip
43
-------
they chemically react with the rock matrix and the water chemistry changes, resulting in increases in
pH, IDS, Na, HCOs, and Cl, and decreases in SO4, NQj, and eH (fig. 14).
For settings known to be confined, significant leakage through aquitards may be identifiable.
Chemical composition of a well water can be compared to the general composition of the ground water
in the region to determine whether the water fits into the chemical composition of the regional-flow
system. If not, local leakage may be occurring. Fogg and Kreitler (1982) observed "recharge" type of
waters downdip in the Carrizo aquifer, East Texas, and concluded that uplifted salt domes had
breached the confining layer permitting leakage to occur at that location.
Tritium and other Anthropogenic Chemicals
Large quantities of tritium, the radioactive isotope of hydrogen, and other anthropogenic
chemicals such as Freon, have been added to the atmosphere in approximately the last 40 yr (1954
through 1990). These chemicals have been recharged through precipitation to the ground water at
concentrations above natural levels on a global basis. The presence of these anthropogenic chemicals
provides an estimate of the absolute age of ground water and, therefore, an estimate of the
susceptibility of an aquifer to contamination by either vertical leakage or lateral flow. The lack of
tritium in an aquifer may indicate the presence of confining strata. Conversely, the presence of modern
concentrations of tritium (see detailed discussion on modern concentrations that follows) indicates
either rapid horizontal flow or vertical leakage. With an understanding of the geologic setting, the
relative importance of horizontal flow versus leakage can be determined. The use of tritium
concentrations in ground water provides a powerful hydrochemical technique for determining the
presence and/or degree of confinement of an aquifer.
The natural tritium in precipitation is estimated to be approximately five tritium units (TU's)
(one tritium unit is equivalent to one 3H atom in 10~18 H atoms). Large quantities of tritium, however,
were added to the atmosphere with the first atmospheric nuclear weapons tests in the early 1950's.
Atmospheric-tritium concentrations in the early 1960's were as high as 6,000 tritium units because of
44
-------
atmospheric testing, but have declined since then because of the U.S./U.S.S.R. moratorium on such
testing (fig. 17) (Fritz and Fontes, 1980). Precipitation and, therefore, ground water that recharged
after the early 1950's contained tritium concentrations significantly above natural background
concentrations. Tritium concentrations in ground water that was recharged before the early 1950's have
decreased by radioactive decay to concentrations below detection levels. Tritium has a half-life of
12.3 yr. Thus, ground waters with no measurable tritium today were recharged before the early 1950's,
whereas ground water with tritium concentrations of two or more tritium units indicates the presence of
a component of water that was introduced into the aquifer after 1954 and is, therefore, younger than
approximately 40 yr.
The tritium techniques should be used to determine only whether a water is younger or older than
40 yr. More specific dates are complicated by the possibility of the mixing of older water (no tritium)
with younger water (high tritium), variable tritium concentrations in atmospheric input, and continual,
radioactive decay of tritium. The tritium in the atmosphere was at its maximum level in the 1960's, but
concentrations have been decreasing ever since (fig. 17). Because of the decrease in nuclear testing, the
atmospheric content and the amount of tritium in recharge water has also been decreasing. This makes
it difficult to calculate specific times of recharge within the period from 1954 to the present. However,
the ability to determine only if well water was recharged more than, or less than, 40 yr ago may be
satisfactory for wellhead protection.
Fluorocarbons (Freon and other artificially created fluorinated organic compounds) have only
been added to the atmosphere in the last 40 yr. These stable organic compounds have been recharged to
the ground water in small but measurable quantities. Presence of fluorocarbons in ground water gives us
an age-dating capability similar to that of tritium (Thompson and Hayes, 1979).
Only atmospherically derived anthropogenic chemicals are considered in this section. Other
anthropogenic chemicals such as Trichloroethane (TCE) and other contaminants also enter aquifers and
can be used to date the age of a water and identify the presence of vertical leakage, but are discussed in
a later section because they are introduced to aquifers through local contaminant plumes rather than on
a worldwide basis.
45
-------
1000
500-
r
- 100-
: 50-
: 10
5-
1 -V
1950
1960
1970
Year
1980
QA14914C
Figure 17. Tritium in precipitation data from 1950 to 1986, Ottawa, Canada (Robertson and Cherry,
1988).
46
-------
The presence of tritium or fluorocarbons in a ground water indicates either recent lateral inflow or
recent vertical leakage. The aquifer, therefore, has the potential to be contaminated from the surface.
The hydrogeologic setting should be evaluated to determine the relative importance of lateral or
vertical flow.
Tritium is measured by the liquid scintillation method on normal or concentrated water samples.
Tritium analyses are not routinely performed on ground water; therefore, there is not an extensive data
base of tritium concentrations. Because of the low concentrations (1 TU = 10~18 3H atoms), care needs to
be taken in water sampling to prevent contamination. Laboratories analyzing tritium should have the
ability to measure tritium concentrations as low as one tritium unit. Fluorocarbons, like tritium, are not
analyzed on a routine basis. Fluorocarbon analyses are made with a gas chromatograph with an
electron capture unit. Fluorocarbons are present in ground water at very low concentrations. Good
sampling procedures are needed to prevent contamination.
The degree of confinement can be estimated from the age of the water if the presence of
confinement has already been determined. If a well field contains modern ground water that has flowed
through the confining strata then the aquifer is semiconfined. If the ground water is older than 40 yr,
then the aquifer should be considered highly confined. The tritium technique has the greatest
sensitivity of the geochemical approaches for defining confinement. It does not, however, identify
pathways for leakage and therefore should be integrated with geologic and hydrologic investigations.
Carbon-14
The absolute age of ground water can be estimated from the activity of the carbon-14 (14C) of
dissolved bicarbonate. As with tritium, 14C ground-water dates can be used to estimate the
susceptibility of an aquifer to contamination by either vertical leakage or lateral flow. An old 14C age
could identify the presence of confinement, or, if confining strata had been previously identified, the
degree of confinement. The use of 14C for dating ground water is better suited for dating old waters than
for dating modern waters. Because of its long half life, 14C probably can be most effectively used as a
47
-------
dating tool for ground water for wellhead protection by determining if the 14C age of the water is
greater than 500 yr. Waters younger than approximately 500 yr are considered as "modern." Tritium, on
the other hand, can be used to date ground water that is less than 40 yr old. Tritium is thus the
preferred method for age determination to infer whether an aquifer is confined or not. Determining that
ground water is thousands of years old using 14C dating does provide a level of assurance not obtainable
by any other technique and, therefore, has a role in wellhead protection strategies. Conversely, 14C
analyses should not be considered for aquifers where ground waters are expected to have short residence
times.
Carbon-14, the radioactive isotope of carbon, is produced in the atmosphere by cosmogenic
reactions. Atmospheric 14C originates as dissolved CC>2 in rainwater and is recharged to an aquifer
through normal precipitation/recharge processes. Two geochemical processes decrease the 14C
concentration in the aquifer. The 14C concentration decreases because of radioactive decay. The half-
life of 14C is 5,730 yr. Carbon samples as old as 50,000 yr can be theoretically dated, but are complicated
by geochemical reactions in the aquifer. The 14C in dissolved CC>2 in rain is used in plant growth. Plant
processes create high CO2 and 14C concentrations in the soil zone. This CC>2 with 14C is then recharged
to the ground water as carbonic acid. Carbonic acid may dissolve carbonate mineral material in the
aquifer as the ground water flows through the aquifer. The mineral material being dissolved, however,
contains "dead" carbon, that is, carbon with no 14C. This addition of dead carbon dilutes the 14C
concentration of the bicarbonate in the ground water and requires corrections of calculated ages (Pearson
and Hanshaw, 1970; Wigley, 1975).
Contamination
The presence of surface contaminants in a well field indicates a high sensitivity to future aquifer
contamination, which may result either from lateral ground-water flow or vertical leakage. The
location of the contaminant needs to be known to differentiate the two pathways (for example, lateral
and vertical). Regardless of the pathway, however, the well's zone of contribution is sensitive to
48
-------
contamination. It could be argued that the development of a wellhead protection area in an aquifer
containing contaminants is after the fact; contaminants, however, may have reached the well's zone of
contribution at concentrations below the Environmental Protection Agency's maximum concentration
limit (MCL) or State primary standard. The presence of nonpoint source contaminants, such as nitrate
fertilizer, may indicate pervasive leakage to an aquifer, although specific pathways may not be
identifiable.
Hydrochemical measurements in confining strata
Previously discussed hydrochemical techniques have concentrated on making measurements
within an aquifer to determine presence and/or degree of confinement. It is also appropriate to
characterize the hydrochemistry of the overlying strata to determine the presence and/or the degree of
confinement. The hydrochemical techniques of general water chemistry, tritium, and 14C in confining
strata can be used in a manner similar to that suggested for an underlying aquifer. An investigation of
water chemistry in overlying strata could provide very valuable information on the presence and/or
degree of aquifer confinement, but probably would provide more detail than is needed for defining
confinement and developing a wellhead protection strategy.
Changes in Water Chemistry
Large volume ground-water production from a well or well field may significantly alter the
hydrology and hydrochemistry of a confined aquifer. Head declines from pumpage may result in
significant vertical leakage through the overlying confining strata. General water chemistry and
tritium concentrations may change because of vertical leakage. Salt water contamination (Cohen and
Kimmel, 1970), nitrate contamination (Eccles and others, 1976), and changes in general chemistry
(Smith and others, 1976), are examples of changes in general water chemistry that have resulted from
49
-------
long-term ground-water pumpage. Evaluating water chemistry data through time for a well under
consideration for wellhead protection may document leakage through confining strata.
Quantitatively Distinguishing Semiconfined from Highly Confined Aquifers
The previous discussion of geologic, hydrologic, and hydrochemical approaches provided several
methods for distinguishing confined from unconfined aquifers and/or indicating some degree of
confinement, but do not quantitatively differentiate semiconfined from highly confined conditions. In
the next section on wellhead protection, different wellhead protection strategies are used for
semiconfined and highly confined aquifers. An arbitrary but logical and justifiable division is
presented to quantitatively separate highly confined from semiconfined aquifers.
The suggested method for differentiating semiconfined from highly confined aquifers, from the
perspective of wellhead protection, is based on the ability to quantitatively assess whether an
overlying aquitard can leak contaminants to the underlying aquifer in a reasonable period of time. The
criterion to distinguish semiconfined from highly confined, therefore, is based on a vertical time of
travel calculation. The calculation of vertical time of travel is a sensitive method for assessing the
potential leakage through an aquitard.
Estimation of time of travel can be calculated in two ways. Calculations can be made with tritium
data or with vertical leakage values and hydrogeologic data from a well or well field. Specifically, a
40-yr vertical time of travel is considered to be a reasonable "rule of thumb" for differentiating
semiconfined from highly confined aquifers. A 40-yr time of travel means that the water at a well was
recharged in approximately 1950, which coincides with the beginning of major industrial development,
atmospheric atomic-bomb testing, and extensive agricultural fertilizer and pesticide use. Most
contaminants in ground water in the United States today were probably introduced into the ground
water no earlier than 40 yr ago.
The tritium technique determines whether the ground water in a confined aquifer contains tritium
or not. If there is no appreciable tritium, then the time of travel of ground water is greater than 40 yr
50
-------
from its recharge, and the aquifer would be considered highly confined. If ground water in a confined
aquifer contains more than a couple of tritium units, then the combined vertical and horizontal time of
travel is less than 40 yr, and the aquifer would be considered a semiconfined aquifer and more sensitive
to surface contamination. The tritium technique requires a basic hydrogeologic understanding of the
aquifer to insure that the presence or absence of tritium reflects vertical leakage and not horizontal
flow. For example, ground water in a highly confined, transmissive limestone aquifer might contain
tritium because of lateral flow from a distant point of recharge and not from vertical leakage.
The second approach for differentiating semiconfined from highly confined aquifers is by
calculating vertical time of travel from vertical thickness permeability values, porosity of the
confining strata and vertical hydraulic gradient across the confining strata. The equation for
calculating vertical time of travel across the confining layer is
Tv= 6 L X /K' Ah (3)
where Tv = vertical time of travel (years) across the confining layer
6 = porosity of confining strata
L = thickness of confining strata
X = travel distance across confining strata
Ah = hydraulic gradient across confining strata
K' = vertical permeability of the confining strata.
A hydrogeologic investigation and a pumping test of a well or well field provide the needed data.
The above equation can be rearranged to solve for the vertical permeability (K') that would be
needed to separate a semiconfined from a highly confined aquifer:
K' = 0 L X /Tv Ah. (4)
Assigning hypothetical values of:
Tv = 40 years
9 = .20
51
-------
L = 10ft
X = 10 ft (contaminant is assumed to be at base of unconfined aquifer, that is, top of
aquitard)
Ah = 20ft
then
K' = .025ft/yr
or
K' = .005gpd/ft2.
If the confining strata for this example has a K' larger than .005 gpd/ft2 then water can leak through
the aquitard in less than 40 yr, and the aquifer should be considered semiconfined. For leakage values
smaller than .005 gpd/ft2 the time of travel across the aquitard would be greater than 40 yr, and the
aquifer would be considered highly confined.
Vecchioli and others (1989) used a 5-yr vertical time of travel to differentiate highly confined
from semiconfined aquifers in northern Florida and recommended the 5-yr time of travel as being
practical. A 40-yr vertical time of travel is suggested in this document because it can be calculated not
only by using pump-test data, but also by using tritium data. Having alternate approaches is important
because not enough hydrologic data may be available to calculate accurate times of travel. Conversely,
tritium analyses may be inappropriate, as in the case of a confined limestone aquifer, where horizontal
flow may be fast enough that ground water contains tritium from lateral recharge and not vertical
leakage.
In a case in which a pump test indicates leakage, but the tritium analyses show no tritium, the
tritium data should be given priority and the aquifer should be considered highly confined. The leaky-
pump test may be documenting leakage from within the overlying confining strata and not leakage
through an overlying strata from a surface or shallow source. The lack of tritium indicates that the
confining strata has effectively prevented recently recharged ground water from reaching the producing
well.
52
-------
Recommendations for Evaluating Confinement
The previous section catalogued three basic approaches for defining confined aquifers: geologic,
hydrologic, and hydrochemical. Within each basic approach, several specific techniques were
discussed. Some techniques are more appropriate because they better define the degree of confinement or
because they are less expensive.
An Integrated Approach
The most important recommendation for determining the presence and/or degree of confinement is
that the determination be based on an integration of geologic, hydrologic, and hydrochemical
approaches. The geologic approach is necessary to determine whether there is a confining strata and
whether there are pathways through the confining strata. The hydrologic and hydrochemical
approaches document whether there is actually leakage through the confining bed. Collecting both
hydrologic and hydrochemical data provides a method to compare one approach to another.
Geologic Approaches
Geologic maps or cross sections based on surface and subsurface geologic data are needed to
identify the presence of confining layers. Artificial penetrations should be mapped, because they
represent the most likely pathways for contaminants to leak through confining strata. Sources of
contamination should be identified. Hydrogeologic maps specifically constructed for wellhead
protection areas and based on geologic and artificial-penetration data are recommended.
53
-------
Hydrologic Approaches
The most important hydrologic approach for evaluating degree of confinement is the calculation,
from pump-test data, of the rate of vertical leakage through the aquitard. This technique is a direct
determination of the leakiness of the overlying strata. The pump-test data for calculating vertical
leakage will also be of value for calculating wellhead protection areas. Water-level data,
potentiometric surface data, continuous water-level recorder data are easier and less expensive to
obtain than leakage information but provide less information on the degree of confinement. Their
greatest value will be for initial screening to determine the presence of confinement. Storativity data
are less critical than leakage data and may be expensive to obtain. Monitoring wells in aquitards and
numerical models may provide valuable information on the degree of confinement, but will be
expensive.
Hydrochemical Approaches
The most important hydrochemical technique is the estimating of time of travel with tritium
data, because the technique provides an absolute age for the water and gives a direct measure of the
sensitivity of the aquifer to contamination from combined horizontal flow and vertical leakage.
General water chemistry, presence of contaminants, and 14C data are not as valuable as tritium data.
54
-------
CHAPTER 4. DEVELOPING A WELLHEAD PROTECTION AREA
Definition of Wellhead Protection Area
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, p. 1-2).
Confined aquifers are less sensitive to contamination from surface sources than unconfined aquifers
because of the presence of overlying confining layers. As discussed previously such confining strata may
be semiconfining, that is, they have the potential for extensive leakage on an areal basis, or they may
be highly confining but be penetrated by discrete features such as faults or artificial penetrations.
Even though the potential for contamination of confined aquifers is less than for unconfined
aquifers, contamination of confined aquifers occurs. And so, it is appropriate to consider wellhead
protection areas for confined aquifers.
Protection Goals
The goals of a wellhead protection area for a confined 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
This goal is met by delineating a remedial action zone, that is, an area delineated with a time of
travel long enough to allow identification and cleanup of contaminants before they reach a well.
55
-------
Lowering Concentrations of a Contaminant to Target Levels before Contaminants Reach a Well
This goal is reached by delineating a protection area large enough to attenuate potential
containments to target levels. Attenuation may occur within the confining strata or the underlying
aquifer. Confining strata may or may not attenuate contaminants. The clay minerals of many confining
strata have the potential to adsorb contaminants. However, contaminant migration through an
aquitard probably will be focused along openings such as fractures, where there will be less dispersion
and dilution of a contaminant than through the aquitard material itself. Attenuation within the
confined aquifer, from the wellhead protection area boundary to the well, may represent a significant
proportion of the total attenuation from the contaminant source to the well.
Protecting All or Part of the Zone of Contribution from Contamination
The purpose of delineating a wellhead management zone is the prevention of contamination of all
or part of a well's or well field's zone of contribution. A wellhead management zone that includes the
entire zone of contribution of a well field in a confined aquifer may be very large. This factor combined
with the generally lower susceptibility of contamination in such settings may lead to implementation
difficulties. An alternate approach is to define a wellhead protection area based on some setback zone
such as 10-, 20-, or 40-yr time of travel contours.
Hydrodynamic Criteria for Delineation of
Wellhead Protection Areas for Confined Aquifers
The U.S. Environmental Protection Agency (1987) recommended five criteria as the technical
basis for delineating wellhead protection areas. These criteria are hydrodynamic ones because they
define the wellhead protection area by flow characteristics of the aquifer. For confined aquifers, these
56
-------
criteria should be integrated with a permeability pathway approach which is discussed in a later
section of this document. The hydrodynamic criteria are:
1. Distance
2. Drawdown
3. Time of travel
4. Flow boundaries
5. Assimilative capacity
Distance
Using the distance criterion, a wellhead protection area is delineated by a fixed radius or
dimension measured from the well to the wellhead 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. It is only recommended as a first, initial step until a more complete
analysis can be made.
Drawdown
Drawdown is the decline in water-level elevation resulting from the pumping of a well. 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 (fig. 8). For an aquifer with a negligible regional hydraulic
gradient, the extent of the cone of depression is coincident with the area of downward leakage. This
area of lowered head values provides the proper head gradient to permit potential leakage of surface
contaminants down to a producing interval of a confined aquifer. The hydraulic potential for leakage
decreases rapidly away from the well as head gradient across the aquitard decreases. For the confined
setting, this potential for downward leakage does not automatically translate into the occurrence of
vertical leakage. A permeable pathway must be present in the aquitard for leakage to occur.
57
-------
The extent of the cone of depression may be larger than the area of downward leakage if the
original potentiometric surface of the confined aquifer was higher than the water table of the
overlying aquifer. Considering the typical limitation of data availability and the fact that the extent
of a cone of depression is typically determined by a calculation rather than by measurement, an effort to
delineate an area of downward flow as separate from the extent of the cone of depression may not be
reasonably accomplished. The areal extent and depth of a cone of depression continues to increase with
time until steady-State conditions are reached. Therefore, drawdown thresholds should be related to
specified periods of time.
Time of Travel
Time of travel is a criterion using the time for ground water (or a ground-water contaminant
moving at the same rate) to flow from a point of interest to a well. Isochrons (contours of equal time) of
any required value can be depicted on a map (fig. 8). The lateral area contained within an isochron is
referred to as a zone of transport (ZOT). As previously described, a vertical time of travel can be
calculated for vertical leakage across a confining layer. Time of travel allows wellhead protection
area delineation using calculations that consider both vertical- and horizontal-time of travel flow
components.
Time of travel calculations for this manual are assumed to be based on advective ground-water
flow. Advective flow of contamination represents Darcian flow, which is typically a conservative
approximation for contaminant transport.
Flow Boundaries
The flow-boundary criterion for delineating a wellhead protection area uses the concept of
locating ground-water divides or other physical hydrologic features that control ground-water flow
and define the geographic area that contributes ground water to a producing well. This area is defined
58
-------
as the zone of contribution. These physical boundaries can be geologic, such as faults across which no
flow occurs, or hydrologic, such as ground-water divides. Ground-water divides can be natural, such as
those that reflect topography, or be human induced, such as those created by a pumping well. In an
aquifer with an original horizontal potentiometric surface, the zone of influence perimeter (the lateral
extent of the cone of depression) coincides with a well's ground-water divide; only water within the
zone of influence flows to the well, that is, the zone of influence equals the zone of contribution. Likely
settings for an original potentiometric surface to approach being horizontal are deep, confined aquifers.
Where the original potentiometric-surface gradient is not negligible, the zone of influence and zone of
contribution do not coincide. In such a setting, the well's ground-water divide on the downgradient side
occurs inside the zone of influence; on the upgradient side, the well's ground-water divide occurs outside
and extends upgradient until it intersects a hydrogeologic boundary. The steepness of the original
potentiometric-surface gradient needed to initiate flow external to the zone of influence is dependent on
such aquifer parameters as hydraulic conductivity. The difference between the zone of influence and
the zone of contribution in an aquifer with an original nonnegligible potentiometric-surface gradient
may be quite small for small times of travel. However, as times of travel become large, significant
differences may occur. If there is a significant natural hydraulic gradient across a site, then this
component should be taken into consideration in delineating wellhead protection areas, particularly if
larger times of travel are being used.
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. 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 ground water. However, consideration of these
processes involves sophisticated treatment of contaminant transport phenomena, which requires
59
-------
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 realistic.
Recommended Hydrodynamic Criterion for Confined Aquifers
The recommended criterion for defining a wellhead protection area is time of travel. Distance
does not accurately characterize the recharge zone. Use of the flow-boundary criterion is not generally
recommended because ground-water divides in a confined setting may be difficult to identify.
Assimilative capacity requires complex treatment of contaminant-transport phenomena which is
beyond the scope of a practical application. A comparison of a wellhead protection area delineated
using the time of travel criterion, with a wellhead protection area delineated by the cone of depression
leads to the recommendation that time of travel is preferred to the cone of depression, because the
lateral extent of a cone of depression increases as the leakage through the aquitard decreases, leading
to unrealistically large wellhead protection areas (fig. 18).
Both the cone of depression and the time of travel contours become larger for a more confined
aquifer, because less water is contributed from vertical leakage, and, therefore more water must come
from lateral flow. Consequently, though perhaps counterintuitively, the wellhead protection area for
a highly confined aquifer would be larger than for the semiconfined aquifer, even though the highly
confined aquifer will be less sensitive to contamination than the semiconfined aquifer.
60
-------
ow -
.
1
20 T
^~> I
^,
c
o
•o
03
Q <
10-
.
0-
Q = 500 gpm
T = 50,000 gpd/ft
S = .00001
]
D
D
, a
l 9 a
• . ° P1 (gpd/ft2)
• • a a .001
" . * . ° o • 01
• • D * 1°
• a o 10.0
o •
• D
0 •
O * •
o • • a
• a
0 . * n
0 • • a
• • Q
o *
• • a
o * _
• • o
° 0 ' * • • Q
0 * • * 9 • .
o ¥ • • ¥ • n
i .,.,... | . ....... | . ....... | . ... ^-. . .^ . l l 9 i P( . .......
1 10 100 1,000 10,000 100,000 1,000,000
Radius (ft)
QA14915C
Figure 18. Simulation of drawdown versus log distance for hypothetical aquifer for different values of
leakage using computer code PTIC (Walton, 1987). Note curves are linear. At the well maximum depth
of drawdown can be determined. As drawdown approaches zero, the maximum lateral extent of the cone
of depression can be estimated.
61
-------
CHAPTER 5. METHODS FOR CALCULATING WELLHEAD PROTECTION AREAS
Methods for Calculating Wellhead Protection Areas for Confined
Aquifers with Negligible-Gradient Regional Potentiometric Surfaces
Two approaches are considered for calculating wellhead protection areas for confined aquifer
settings where the regional potentiometric surface gradient is negligible, (1) cone of depression and (2)
time of travel.
Cone of Depression Approach
The lateral extent (as defined by a very small [
-------
to estimate drawdown versus distance. The slope of a semilog plot of drawdown versus distance (fig. 19)
(Driscoll, 1986) is twice the slope of the time versus drawdown curve.
Drawdown versus Distance Simulation Using Analytical Solutions and Simple
Computer Models Method
The lateral extent of a cone of depression can be determined with analytical solutions and
hydrologic parameter values derived from pump-test data or previously collected regional data if
pump-test data are not available. Two techniques are available: the equilibrium technique, used when
the cone of depression has reached equilibrium; or the nonequilibrium technique, used when the cone is
still expanding. The radial distance of zero drawdown for a pumping well that has reached
equilibrium (the cone of depression has expanded as far as it can) can be estimated with the Thiem
equation (Thiem, 1906)
8= Q
2rcKb r (5)
where s = drawdown from original potentiometric surface
Q = discharge
K = hydraulic conductivity
b = aquifer thickness
r = radial distance at point of drawdown observation
re = radial distance of zero drawdown of cone of depression.
Davis and DeWiest (1966) and Lohman (1972) provide a detailed discussion of this equation. The
second technique is to use the nonequilibrium Theis equation (Theis, 1935), from which the lateral
extent of the cone of depression at different times can be calculated
_ 114.6QW(u)
T (6)
63
-------
2 -
4 -
c 6
o
10 -
12 -
14
Pumping rale, Q =
264 Q 264
9.960gpd/ft(125m2/day)
1 235 10 20 30 50 100 200300 500 1,000
Time since pump started (min)
Data from observation well A
4 -
8 -
c 12
o
T3
I16
20 -
24 -
28
observation we I
300 minutes
; 200 gpm (1,090
300 minutes
235 10 20 30 50 100 200300 500 1,000
Distance from pumped well (ft)
QAU916C
Figure 19. The lateral extent of a cone of depression of a pumping well can be determined with time
versus distance data. The slope of drawdown versus log distance is twice the slope of drawdown versus
log time. Example from Driscoll (1986). Used with permission from Groundwater and Wells, Edition 2,
1986, Johnson Filtration Systems Inc.
64
-------
W(u) is the well function of "u" where
Tt (7)
s = drawdown
Q = discharge
T = transmissivity
r = radial distance to point of drawdown observation
S = storativity
t = time.
These equations are written for solution with English units (s[ft], Qtgpm], T[gpd/ft], r[ft], t[days]).
Driscoll (1986) provides a detailed discussion of methods of solution for this general equation. An
appropriate pumping period must be chosen that simulates the normal pumping period for the well
under consideration for wellhead protection.
User-friendly computer programs can also be used to estimate the cone of depression for
equilibrium or nonequilibrium conditions. Computer codes such as those described in Walton (1987) are
semianalytical codes with relatively simple boundary conditions and simple designations of hydraulic
conductivity, storativity, and leakage. More complex models can also be used to calculate drawdown
versus distance where boundary conditions, vertical and horizontal hydraulic conductivity, storativity
values, and so forth, can be varied on an element by element basis. Simulation of well-field hydraulics
with interfering cones of depression from multiple-well production are best accomplished with
numerical codes rather than analytical solutions or some of the simpler numerical models (see Van der
Heijde and Beljin, 1988). The complexity of the code, however, should be matched with the
availability of data. Sophisticated codes are often not appropriate when there are only limited data
available.
65
-------
Time of Travel Approach
Time of travel calculation is based on Darcy's law. Either the distance of flow for a given period
of time or the time of travel for a given distance can be calculated from data on hydraulic gradient,
transmissivity, porosity, and pump discharge. Time of travel calculations can be made either by
incorporating the hydraulic gradient from the cone of depression and transmissivities, both obtained
from pump-test data, into time of travel calculations or by using a simpler cylinder method, which does
not require hydraulic gradient or transmissivity data, or by using WHPA model (Blanford and
Huyakorn, 1990), a semianalytical time of travel model.
A 40-yr period is a convenient period to use for a time of travel calculation because 40 yr is an
approximate break point between recently recharged (post-1950) waters containing tritium, and older
(pre-1950) waters with no "bomb" tritium. Water with no measurable tritium should be older than 40
yr. If there is no tritium in ground water, then it will take at least 40 yr for currently recharging water
to flow to a well either horizontally or vertically.
Cone of Depression/Time of Travel Method
The cone of depression/time of travel method calculates time of travel on the basis of the
hydraulic gradient of the cone of depression. Calculations can be made through (a) simple analytical
solutions such as the following equation, or (b) reverse-path calculation computer codes such as used by
Shafer (1987) or Blanford and Huyakorn (1990).
(a) Analytical time of travel can be calculated from the following equation:
TOT = (Al) * 6/K*i (8)
66
-------
where TOT = time of travel threshold
Al = distance of travel for a given time period
K = hydraulic conductivity
0 = porosity
i = Ah/Al is the hydraulic gradient of the cone of depression between two points of
measurement. Ah is the difference in hydraulic head between two points of
measurement on a flow line (Al).
To calculate time of travel contours, this equation can be arranged in the following form:
Al = (TOTK* i)/ 6. (9)
The hydraulic gradient decreases rapidly away from the well (fig. 8) and, therefore, is not constant
and is a function of Al. The time of travel can be calculated by the following procedure. The time of
travel for various incremental distances is estimated from the hydraulic gradient (i) for each increment
(e. g. 0 to 10 ft, 11 to 100 ft, and 101 to 1,000 ft), pump test data and equation (8) (fig. 18). The total time
of travel is the sum of each time of travel for each increment. The total time of travel is then plotted
versus distance (fig. 20). Because the log of time of travel versus the log of distance is approximately
linear, the distances for different times of travel can be estimated. Extrapolation beyond the farthest
data point should be used with care. (This calculation can easily be made with a spreadsheet program
on a microcomputer.) The distance of travel for a given time of travel can then be contoured to delineate
a wellhead protection area.
(b) Time of travel contours can also be calculated from computer models that map the
potentiometric surface and calculate ground-water flow paths in a reverse direction. Flow paths of a
ground-water flow system can be calculated with either forward or reverse particle tracking numerical
ground-water flow models. Forward tracking predicts where ground water or a contaminant in the
ground water will flow in the future. Most ground water flow models that calculate flow paths are
forward tracking. Forward tracking is particularly valuable for predicting where contamination from a
67
-------
iu- -I
103 -;
102 -j
101 i
~ 10° i
>H ;
1 io-1 -
«
° TO'2 i
0)
E
F IO-,
ID'4 n
ID'5 H
1Q-6 -i
•m-7
Q = 500 gpm
T = 50,000 gpd/ft
S = .00001
40 yr
Syr
P' (gpd/ft2)
o
a .001
• 0.1 0
• 1.0 •
o 10.0 o _ a
3.5 days |
pi
r» •
8hr 0
o i
50 min O H
0 1
o 1
o •
,''
1
•
• D
•
• o
1 "
2,500 ft
300ft
X
•
, °
6,000 ft
\
•
3
10
100
Radius (ft)
1,000
10,000
QA 14917C
Figure 20. Simulation of time of travel (in years) for hypothetical aquifer for different values of
leakage using computer code PTIC (Walton, 1987).
68
-------
pollution site may flow and in what time period. In contrast, reverse-path calculations estimate where
ground water and contaminants have flowed in the past. This approach is valuable for defining
wellhead protection areas because it defines the "recharge area" for a well and the time of travel for
water or a contaminant to get from a point to a well.
Calculation of reverse flow paths and travel times with numerical models is a two-step process.
First, the water level at the well and the potentiometric surface for the surrounding area is calculated
and, if desired, the problem of vertical leakage associated with semiconfined aquifers can be
addressed. Many ground-water computer models can simulate ground-water flow. Second, reverse flow
paths are calculated with codes such as WHPA (GPTRC-numerical option) (Blanford and Huyakorn,
1990) or GWPATH, the reverse-path numerical model of Shafer (1987) (fig. 21).
The use of reverse flow path and time of travel calculations has advantages and disadvantages.
The advantages are that the method is the most sophisticated and provides the most realistic
simulation. The disadvantage is that the method is the most complex.
An alternate approach to using a reverse-path calculation is to use a solute transport (forward
tracking) code, but use the producing well or field as an injection well and calculate the distance to the
edge of the hypothetical plume as it migrates away from the well for specific times. The plume
boundary for a given period of time (time of travel) can be used to delineate a wellhead protection area.
This approach being used by the Texas Water Commission to delineate wellhead protection areas for
well fields may have advantages, since solute transport modeling specifically considers contamination
migration.
Cylinder Method
The cylinder (volumetric) method is used by the Florida Department of Environmental
Regulation, the U.S. Environmental Protection Agency (1987), and Vecchioli and others (1989). The
69
-------
Feet
8000
8000
A Hydraulic source
• Hydraulic sink
~~- Reverse tracked path
QAH918C
Figure 21. Example of reverse-path calculation (from Shafer, 1987).
70
-------
method uses a volumetric-flow equation which calculates the radius (r) of a cylinder from which all
water would be pumped out after a defined period of time (time of travel) (fig. 22). The equation is as
follows:
r^Qt/neH)1/2 (10)
where Q = discharge
6 = porosity
H = length of screened interval
t = travel time to well
The equation is a modification of Darcy's law for radial flow to a well, even though it uses neither
hydraulic conductivity nor hydraulic gradient (Vecchioli and others 1989). The volumetric-flow
equation assumes all flow is horizontal. In the context of confined aquifers, the aquifer is assumed to be
highly confined and, therefore, there is no vertical leakage into the aquifer. This assumption results in
a larger radius for a given time of travel than would be calculated for a leaky confined aquifer.
Semianalytical Method (WHPA Model)
The WHPA model is an integrated semianalytical model for delineation of wellhead protection
areas (Blanford and Huyakorn, 1990) that was developed for the U.S. Environmental Protection Agency
Office of Ground-Water Protection to calculate wellhead protection areas by calculating time of travel
contours for negligible or sloping regional hydraulic gradients (fig. 23). The WHPA (1.0) originally did
not consider vertical leakage and therefore could have caused time of travel contours and overall
wellhead protection areas to be larger than needed; time of travel contours would be similar to those
calculated by the cylinder method, because both neglect leakage. Recent modifications to the computer
program (WHPA 2.0) allow vertical leakage and will permit time of travel calculations to leaky
aquifer settings. (WHPA 2.0 was not available for testing during preparation of this manual.)
71
-------
Pumping well
Q - 500 gpm
6 - 0.25
Volumetric-flow equation
(Cylinder equation)
R - SQRT(Qt/7t 6H)
when t - 40 yr
r - 6000 ft
QA14919C
Figure 22. Cylinder or volumetric-flow equation approach for calculating time of travel for 40 yr. This
approach gives a conservative time of travel because vertical leakage is not considered.
72
-------
10,000
8,000 -
6,000 -
4,000 -
2,000 -
i i i r i i
2,000 4,000 6,000
Meters
8,000
10,000
QA14920C
Figure 23. Example of reverse-path calculation using wellhead protection area (WHPA) computer
program (from Blanford and Huyokorn, 1990).
73
-------
Comparison of Approaches and Methods
Calculating a wellhead protection area from the cone of depression/time of travel method is
recommended in preference to a method associated with the cone of depression approach or the cylinder
method. The cone of depression/time of travel method is the most versatile of the three because it
provides an accurate assessment of the wellhead protection area for both semiconfined and highly
confined aquifers. By calculating the cone of depression the potential for vertical leakage is accounted
for, and, by using a time of travel calculation the lateral extent of the wellhead protection area is
limited to a reasonable size. The methods associated with the cone of depression approach will
approximate the wellhead protection area calculated from the cone of depression/time of travel
method for a semiconfined aquifer but can become very large for a highly confined aquifer. The cylinder
method is a time of travel calculation which does not account for possible leakage and therefore
considers all aquifers as highly confined. This may result in wellhead protection areas that are larger
than needed.
The difference in size of the wellhead protection areas for semiconfined and highly confined
aquifers can be demonstrated by using the three different methods to calculate a wellhead protection
area for a hypothetical aquifer with: T = 50,000 gpd/ft, Q = 500 gpm, S = .0001, and leakage conditions
2
that vary from highly leaky (P' = 10 gpd/ft ) to highly confined (no leakage). By using the cone of
depression/time of travel method with a 40 yr threshold, the radius of the wellhead protection area
ranges from 300 ft for the very leaky aquifer to 6,000 ft for the highly confined aquifer with most of the
radius values from 2,500 to 6,000 ft for the more confining conditions (fig. 20).
The cone of depression methods create a wellhead protection area which may be significantly
larger than one developed with the cone of depression/time of travel method. The radius of the cone of
2
depression for a very leaky aquifer (P' = 10 gpd/ft ) is approximately 250 ft, whereas the radius of a
cone of depression for a confined aquifer (no leakage) is greater than 20,000 ft (fig. 18). Calculated times
74
-------
of travel for the highly confined scenario from the outer edges of the cone of depression to the pumping
well are greater than 10,000 yr, which is not realistic for implementing a wellhead protection area.
The calculated distance for 40-yr time of travel using the cylinder method is 6,000 ft for highly
confined conditions, which is similar to the time of travel distance for the highly confined aquifer
setting for the cone of depression/time of travel approach. The cylinder method, however, does not
accurately calculate time of travel for the semiconfined condition, because the cylinder equation does
not incorporate any leakage. WHPA (1.0) also calculates the 40-yr time of travel as 6,000 ft.
Calculation of Wellhead Protection Area for Wells in Confined
Aquifers with a Regional Sloping Potentiometric Surface
In the previous section, the approaches for calculating wellhead protection areas assume that
ground-water flow toward a well is dominated by well pumpage from an aquifer with a negligible
initial potentiometric-surface gradient. Potentiometric surfaces in confined aquifers are typically
characterized by very low gradients. Nevertheless, it is possible that steeper initial gradients can
occur within confined aquifers and affect the shape of the cone of depression of a pumping well (fig. 24).
The size and shape of the wellhead protection area is controlled by the regional hydraulic gradient,
the aquifer transmissivity, and well discharge. For aquifers with regional potentiometric gradients
between .0005 and .001 or greater wellhead protection area delineation methods that incorporate a
sloping regional potentiometric surface should be considered (Todd, 1980; Bear and Jacob, 1965; Southern
Water Authority, 1985).
There are two general approaches which incorporate an initial sloping potentiometric surface in
estimating a wellhead protection area: (1) zone of contribution with the identification of flow
boundaries and (2) zone of transport with time of travel contours which can be solved through solution
of simple analytical equations or through computer application.
75
-------
Zone of Contribution with Identification of Flow Boundaries Method
In this method, the zone of contribution is defined by flow boundaries within an aquifer. For a
well pumping from an aquifer having a regional sloping potentiometric surface (fig. 24), the edge of the
cone of depression on the down-gradient side will be relatively close to the well. On the up-gradient
side, the transverse extent (in the Y-direction) of the zone of contribution increases asymptotically to a
maximum, but the lateral extent (in the X-direction) extends infinitely, or until a hydrogeologic
boundary is reached, in the up-gradient direction. The down-gradient null point and the maximum
width of the zone of contribution can be solved analytically (Todd, 1980).
_Y = tan[2aKblY\
X I Q / (ID
where X and Y are coordinates
Q = the pumpage rate at the well
K = hydraulic conductivity
b = the saturated thickness of the aquifer
i = the hydraulic gradient of the initial, sloping potentiometric surface.
The down-gradient flow boundary (null point) is given by
2jrKbi . (12)
The transverse boundary limit is given by
YL=±_S_
2nKbi . (13)
76
-------
(a)
Ground surface
Original
potentiometric
surface
Not to scale
Impermeable
Not to scale
Ground water divide
Uniform-flow equation: « tan(———Yj
Distance to down-
gradient null point:
Q
2jiKbi
Boundary limit: YL » ±
2Kbi
Where: Q = Well-pumping rate
K • Hydraulic conductivity
b = Saturated thickness
i - Hydraulic gradient
it - 3.1416
QAU921C
Figure 24. Ground-water flow field for cone of depression of a pumping well with a regional ground-
water flow gradient. Uniform flow equation (Todd, 1980) can be used to calculate down gradient null
point and lateral extent of zone of contribution.
77
-------
The shape of the flow boundary can be calculated using equation (11), which can be solved by selecting
Y values between zero and YL that are calculated from equation (13). However, no up-gradient flow
boundary can be determined from these equations. The up-gradient boundary is generally selected to be
the first hydrogeological boundary intersected by the zone of contribution or defined by a desired time
of travel. The WHPA code, (Blanford and Huyakorn, 1990) described next, can also be used for
calculating the shape of the flow boundary. Vertical leakage is not considered in equation (11), and so
the welhead protection area using this method will be larger than it needs to be if there is significant
vertical leakage.
Zone of Transport with Time of Travel Contours Approach
A zone of transport with time of travel contours can be calculated using three methods (1) the
simple analytical solution method, (2) the semianalytical method, and (3) the time of travel reverse-
path calculation method. All three methods calculate times of travel from which contours of equal
time can be constructed.
Simple Analytical Solution Method
The time of travel for water to move along a line parallel to the hydraulic gradient, from a point
to a pumping well, can be calculated from the following equation (modified from Bear and Jacob, 1965):
2;tKbi
where TX = travel time from point X to a pumping well
6 = porosity
(14)
78
-------
XL = distance from pumping well over which ground water travels in TX (time); XL is either
positive or negative depending on whether point x is up gradient (+) or down gradient
(-) of the pumping well
Q = discharge
K = hydraulic conductivity
b = aquifer thickness
i = hydraulic gradient
This equation is similar to that used by the Southern Water Authority (1985), which is included in the
Environmental Protection Agency's general guidelines for delineating wellhead protection areas (U.S.
Environmental Protection Agency, 1987).
The equation permits the calculation of the travel time from a given point to a pumping well.
Calculation of travel distances for specific travel times have to be solved by trial and error but can be
easily accomplished through the use of a spreadsheet program with a microcomputer. Travel distances
and travel times can only be calculated along a line through the pumping well parallel to the regional
hydraulic gradient. Complete delineation of the wellhead protection area around a well in an aquifer
with a regional sloping potentiometric surface requires computer solution. The simple analytical
solution method for determining a wellhead protection area does not account for any vertical leakage
through an overlying aquitard if the aquifer is semiconfined. Therefore, as with the cylinder approach
for confined aquifers with low regional potentiometric surfaces having negligible gradients, the
calculated extent of the wellhead protection area should be considered larger than needed.
The best use of this equation may be for determining the importance of the regional
potentiometric gradient on the shape of the wellhead protection area and whether the delineation of
wellhead protection areas should be made with techniques that allow for a regional potentiometric
surface with a non-negligible gradient. The ratio of the distance of ground-water travel in the down-
gradient direction to that in the up-gradient direction for the same time of travel indicates how
noncircular the wellhead protection area will be. As the shape of the wellhead protection area
79
-------
approaches a circle, the influence of the regional hydraulic gradient on times of travel becomes
insignificant.
Semianalytical Method (WHPA Model)
WHPA is an integrated semianalytical model for delineation of wellhead protection areas (fig.
23) (Blanford and Huyakorn, 1990). WHPA is appropriate for calculating time of travel contours for
confined aquifers with regionally sloping potentiometric surfaces. It is recommended in preference to
the simple analytical solution described above because among other reasons the complete time of travel
contours can be calculated, and not just at points along a line intersecting the well and parallel to the
regional-flow gradient.
Reverse-Path Calculations Method
The time of travel from reverse-path calculations can be made with a regional potentiometric
gradient or with a negligible hydraulic gradient. A more detailed description of the method is
included on page 67.
Comparison of Methods
The zone of contribution method defines ground-water flow boundaries, but does not provide an up
gradient limit for a wellhead protection area. It provides a relatively simple method for defining a
wellhead protection area and up-gradient boundaries can be determined by other methods.
A wellhead protection area can be calculated from the simple analytical solution method for
travel times. The equation however limits travel time calculations to a down-gradient point and an up-
gradient point along a line through the well and parallel to the regional flow gradient. The complete
wellhead protection area cannot be delineated.
80
-------
The WHPA computer program, a semianalytical solution for travel times, can be used for
calculating wellhead protection areas. It provides a better approximation of the wellhead protection
area than either the zone of contribution or simple analytical approach because it provides a complete
areal delineation of the wellhead protection area.
Only the WHPA (2.0) computer code accounts for potential vertical leakage in semiconfined
aquifers. Significant vertical leakage will cause wellhead protection areas to be smaller; therefore,
any method that does not account for vertical leakage will result in a larger, that is, more conservative,
wellhead protection area. (The WHPA code [2.0] that incorporates leakage was not available in time
to be tested for this document.)
Reverse-path calculations provide the most sophisticated delineation of a wellhead protection
area. The method requires two steps, (1) calculation of the regional potentiometric surface with a
numerical flow model (this step accounts for vertical leakage) and (2) calculation of the reverse paths
with a second code. Reverse-path particle tracking provides a more accurate delineation of the
wellhead protection area than any other method, but may be more complicated than necessary for the
delineation of many wellhead protection areas in confined aquifers.
81
-------
CHAPTER 6. WELLHEAD PROTECTION AREAS FOR SEMICONFINED
AND HIGHLY CONFINED AQUIFERS
Different permeability pathways are anticipated for semiconfined and highly confined aquifer
settings and determining the locations of these pathways is important for both types of aquifers. The
locations of these pathways should be given a higher level of wellhead protection, because they are
the most probable zones where contamination may enter the aquifer.
Permeability Pathway Criteria for Semiconfined Aquifers
In the case of the semiconfined aquifer, there is, by definition, significant leakage through the
aquitard. The potential for leakage is considered to be areally distributed across the wellhead
protection area (fig. 25). The geologic and artificial penetration mapping techniques described in a
previous section on defining confinement (Chapter 4) are recommended for describing the nature of
leakage and mapping of possible leakage zones. If specific zones of leakage cannot be identified, then
the entire wellhead protection area should be considered sensitive to the leakage of contaminants.
Because the presumption of widespread leakage leads to a high level of protection throughout the
wellhead protection area, identification of specific points or zones of leakage may be less critical than
identification of potential contaminant sources.
Permeability Pathway Criteria for Highly Confined Aquifers
In contrast, the highly confined aquifer has essentially no or negligible, leakage through the
aquitard. Nevertheless, minor leakage that cannot be identified from pumping tests may be important
if it occurs through discrete high-permeability pathways (such as faults or wellbores) (fig. 26).
Mapping geologic and artificial penetrations is recommended for describing the nature of leakage and
for identifying possible leakage. For the highly confined setting, the potential for contamination of
82
-------
. • Semiconfined aquifer
: pCone of depression
jjjjniiy General wellhead protection area
QA 14922C
Figure 25. Schematic of areally distributed permeability pathways for semiconfined aquifer. Example
is of a fractured till aquitard, which causes semiconfinement and an areally extensive potential for
surface contamination. A wellhead protection area should include all the area within the circle.
83
-------
: i . Confined aquifer -'i /^i'-
\
\
X
. ,...., ., .. ... \ ..
>r~t Cone of depression \
|jjjjijj:| General wellhead protection area
| | More sensitive wellhead protection areas created by boreholes and a fault
QA 14923C
Figure 26. Example of wellhead protection area for highly confined aquifer where penetration of
confinement has only occurred with abandoned boreholes and a fault.
84
-------
well water is considered to be lower than for leaky aquifers. Potential pathways such as faults,
fractures, and boreholes may have to be treated as highly restricted zones. Abandoned and unplugged
boreholes may have to be sealed.
85
-------
CHAPTER 7. CALCULATION OF WELLHEAD PROTECTION AREAS FOR WELL FIELDS
The previously described methods for calculating a wellhead protection area are based on the
assumption of a single well. More complex configurations of wells occur and should be considered for
wellhead protection. Three scenarios are considered. (1) Well fields where pumping wells have
interfering cones of depression, (2) well fields where individual wells are screened at different
intervals and cones of depression do not interfere, and (3) well fields where individual wells are
screened in different aquifers, the shallower aquifer is semiconfined and the deeper aquifer is confined.
(1) Well fields in which pumping wells have interfering cones of depression. Ground water
pumpage from multiple wells may result in a composite cone of depression that is deeper and wider
than individual cones of depression and noncircular. Calculation of a wellhead protection area for a
well in an aquifer with a negligible regional gradient should still be based on a cone of depression/time
of travel approach. However, this calculation will probably require the use of numerical models that
calculate the cone of depression and then time of travel contours to accurately assess the more complex
area of time of travel. The wellhead protection area semianalytical solution and the reverse-path
codes are appropriate. The WHPA code and other reverse-path codes are also the most appropriate
methods for calculating wellhead protection areas for sloping regional potentiometric surfaces because
they more accurately portray the interaction between well field hydraulics and the sloping regional
potentiometric surface.
(2) Well fields in which individual wells are screened at different depth intervals and cones of
depression do not interfere. The wellhead protection area should be based on the composite areas
calculated for each well, using one of the previously described approaches (fig. 27). The problem is not
so complex that a numerical model has to be used, since the cones of depression do not interfere; they
only overlap.
(3) Well fields in which individual wells are screened in different aquifers, the shallower
aquifer is semiconfined and the deeper aquifer is highly confined. The total wellhead protection area
86
-------
Highly confined aquifer
Highly confined aquifer
General wellhead protection area
OA14924C
Figure 27. Example of overlapping wellhead protection areas for two wells in different confined
aquifers. Total wellhead protection area is the composite area for the two wells. Cones of depression
are overlapping but not interfering. Wellhead protection areas based on cone of depression.
87
-------
should be the combination of the individual protection zones with each separate zone being protected
according to its sensitivity to potential contamination (fig. 28).
88
-------
: Semi-confined aquifer
Highly confined aquifer .
[jjjjjljjj Wellhead protection area for highly confined aquifer
bs^sj Wellhead protection area for semiconfined aquifer
QA 14925C
Figure 28. Overlapping wellhead protection areas based on cones of depression for a highly confined
and a semiconfined aquifer. The protection area for more sensitive semiconfined aquifers is given the
higher priority than the protection area of the highly confined aquifer where they overlap.
89
-------
CHAPTER 8: EXAMPLES OF WELLHEAD PROTECTION STRATEGIES IN CONFINED AQUIFERS
The following examples describe the development of wellhead protection strategies for two
confined aquifer settings. Wellhead protection areas were developed for two sites, one in Bastrop
County and one in Wharton County, Texas (fig. 29), and are examples from the updip and the downdip
sections, respectively, of a regional confined coastal aquifer. The examples are (1) to discuss assessing
confinement and (2) to discuss determining a wellhead protection area. Evaluating the different
criteria permits the decision on the degree and type of confinement. On the basis of this decision, a
wellhead protection area delineation strategy is presented for each of the two examples. The
development of wellhead protection areas for Bastrop and Wharton Counties is presented in detail in
Appendix 1 to show the complexity of the process.
Bastrop, Texas
Example from the Updip Section of a Confined Aquifer
The first wellhead protection example is a well field in Bastrop County, Central Texas, located
in the outcrop of the Wilcox aquifer. The well field is located about 5 mi north of the City of Bastrop
and south of the Camp Swift Military Reservation (fig. 30). The well field consists of two active wells,
516 and 515, as well as eight inactive and abandoned wells. The well field is bounded to the south and
west by a Federal Prison Facility, to the north by the University of Texas Cancer Research Institute,
and to the east by a trailer park and small industrial park. Within one mile to the west of the well
field, the Lower Colorado River Authority operates a medium-sized open-pit lignite mine. The Camp
Swift well field is operated by the Aqua Water Supply Corporation, a local water cooperative, which
supplies water to the town of Bastrop and rural areas in Bastrop, Lee, and Milam Counties for a
population of about 20,000. The well field is located within the outcrop area of the lower Eocene
Wilcox Group, which is comprised of three formations, (1) the Hooper Formation, (2) the Simsboro
Formation, and (3) the Calvert Bluff Formation. The Simsboro Formation consists of relatively sand-
90
-------
N
EXPLANATION
Alluvium ond
Bolson deposits
Edwards (Balcones fault
zone • Son Antonio region)
Edwards (Balcones fault
zone • Austin region)
High plains
(Ogallala)
Edwards -Trinity
(Plateau)
Trinity group
Carrizo-Wilcox
BASTROP CO ! FAYETTE CO j COLORADO CO
WHARTON CO
A'
BRAZORIA CO
QAI48Z6
Figure 29. Geologic map and cross section of the Gulf Coast area, showing locations of Bastrop (Camp
Swift well field) and City of Wharton.
91
-------
EXPLANATION
Camo Swift well field
• Active wells
O Inactive wells
516 Last three digits of state well
number
© Test wells (abandoned)
General wells (no well number)
<|> Oil or gas test wells
0 Industrial wells
O Domestic or stock
s~ Wellhead protection area
r~ Wellhead protection area,
1 assuming anisotropic
transmisswity. See Appendix I.
QAI4894
Figure 30. General highway map of Bastrop County showing the location of the Camp Swift well field
and wellhead protection area for wells 515 and 516. The wellhead protection area defined by the
dashed line is based on the anisotropic conditions observed during modeling. Appendix 1 provides a
detailed description.
92
-------
rich fluvial deposits and is the main waterbearing unit in the area. The recharge area for the Simsboro
is along a 1- to 3-mi-wide outcrop belt which is about 2 mi west of the well field.
The wellhead protection area delineation strategy for this particular setting followed the steps
outlined above and is discussed in detail in Appendix 1. The first step, determining the presence and/or
degree of confinement was based on evaluation of geologic, hydrologic, and hydrochemical criteria. The
Camp Swift well field was considered highly confined and has a low probability of contamination.
The main indications were the presence of overlying shale strata, the absence of any tritium, relatively
14
old C ground-water ages, and the highly confined response from the aquifer tests.
The second step, delineating the combined wellhead protection areas for the two producing wells,
515 and 516 of the Camp Swift well field, follows the different approaches given above and is
described in detail in Appendix 1. The recommended wellhead protection area is an approximate circle
with a radius 6,000 ft, and is based on a 40-yr threshold and the time of travel approach for the two
producing wells as shown in figure 30. Within the 40-yr capture zone, local higher protection zones are
recommended in the vicinity of the main pathways for potential contamination. These pathways are
considered to be localized, such as abandoned boreholes and existing wells.
Wharton, Texas
Example from the Downdip Section of a Confined Aquifer
The second wellhead protection example is a well field in the City of Wharton, Wharton
County, Texas, located in the Gulf Coastal Plain of southeastern Texas (fig. 29). The well field is
located in the downdip section of the Gulf Coast aquifer, a regionally extensive coastal plain aquifer.
The City of Wharton is about 60 mi west of Houston and about 50 mi north of the coast of the Gulf of
Mexico. The city water wells, serving approximately 70,000 people, are located on empty lots
throughout the city (fig. 31). In particular, a wellhead protection area is designed for City of Wharton
well 3, (also referred to as 406).
93
-------
EXPLANATION
•0- Oil and gas wells
© #2 City water wells
0 2000 4000ft
I | i
1200m
Figure 31. Map of Wharton, Texas, and vicinity, showing wellhead protection area for city of Wharton
well no. 3 (well 406).
94
-------
In Wharton County, the main hydrogeologic units consist of Pleistocene and Pliocene sequences of
gravel, sand, silts, and clay. All Gulf Coast formations thicken toward the coast and crop out in belts
that are nearly parallel to the shoreline. The wells produce from the Chicot aquifer from a depth of
about 600 to 900 ft below sea level. The Chicot aquifer is overlain by a thick sequence of mostly clays
(Beaumont Clay) which is considered the confining unit for the underlying Chicot aquifer. The Willis
Sand is the major waterbearing unit of the Chicot aquifer, which crops out about 30 mi northwest of the
City of Wharton. The outcrop area is the main recharge area.
The development of a wellhead protection area delineation strategy followed two steps, (1)
determining the degree of confinement, and (2) delineating the wellhead protection area. Based on
geologic, hydrologic, and geochemical criteria, discussed in detail in Appendix 1, ground water in well
406 is considered highly confined. Although pumping-test data indicate leaky behavior, leakage is
14
interpreted to come from overlying and underlying sands, which were not screened. The old C ground-
water ages and absence of detectable tritium indicate very old ground water. The overall vertical
hydraulic head distribution indicates a downward gradient; however, vertical permeability of the
confining units is very low, preventing significant fluid movement. The recommended wellhead
protection area for well 406 (fig. 31) is a circular area with a radius less than 1,000 ft and is based on
the cone of depression/time of travel approach using a 40-yr threshold. Within this general area, the
main pathways for contamination are abandoned boreholes and existing wells.
Comparison of Wellhead Protection Areas for the Two Examples
The delineated wellhead protection areas for Bastrop and Wharton, Texas show some differences
owing to their different hydrogeologic settings.
In the Bastrop area the wells are within a highly confined aquifer with a measurable regional
hydraulic gradient. This results in a slightly noncircular wellhead protection area with a radius of
about 6,000 ft.
95
-------
In the Wharton area the well is located in a highly confined aquifer setting with a negligible
horizontal hydraulic gradient. Pump-test data indicate significant leakage, but the leakage is from
adjacent overlying or underlying sands and not from shallow ground-water sources. The wellhead
protection area is a circle with a radius of less than 1,000 ft.
The ground water in both locations is old. The highest priority areas for protection within the
general wellhead protection area are those containing artificial penetrations.
96
-------
CHAPTER 9: RECOMMENDED APPROACH FOR DEFINING
WELLHEAD PROTECTION AREAS FOR CONFINED AQUIFERS
The recommended approach for defining wellhead protection areas for confined aquifers is as
follows, and is diagrammed as a flow chart (fig. 32):
(1) The nature of confinement of the aquifer is considered to be either unconfined, or confined,
(a) If the aquifer is unconfined, recharge to the aquifer is considered pervasive. A wellhead
protection area delineation strategy is developed based on the techniques in EPA's general guide:
Guidlines for Delineation of Wellhead Protection Areas (U.S. Environmental Protection Agency,
1987).
(b) If the aquifer is confined, one should determine whether it is a semiconfined or highly
confined system through methods that calculate a time of travel. A 40-yr vertical time of travel
is suggested, but other time periods may be more appropriate for specific well settings. If the
aquifer is semiconfined, the aquifer is overlain by a leaky aquitard in which leakage is assumed
to be areally distributed throughout, in addition, there may be localized leakage through fault
zones and boreholes. If the aquifer is highly confined, the aquifer is overlain by a nonleaky
aquitard, and the only potential points of leakage are through discrete permeability pathways
such as faults, fracture zones, and abandoned boreholes.
(2) The prepumping gradient of the regional potentiometric surface is determined. As a rule of
thumb, if the regional gradient is 0.0005-0.001 or greater, it may affect the size and shape of the
wellhead protection area. The impact of the regional gradient on the shape of the wellhead protection
area can be estimated with equation (14). If the gradient is less than 0.0005, the size of the wellhead
protection area will be controlled by the hydraulics of the pumping well.
For either scenario, a time of travel delineation criterion is recommended. For the scenario with a
very low regional hydraulic gradient, assuming some degree of confinement, the time of travel
calculation can be made with either the cone of depression/time of travel or the cylinder methods. If
the necessary data are available, the cone of depression/time of travel method is recommended in
97
-------
Start
Define aquifer setting:
geologic, hydrologic,
hydrochemeical approach
Highly confined
discrete permeability
pathways
End
Confined aquifer
WHPA defined
Hydrodynamic
WHPA defined
Degree of
confinement
Integrate
hydrodynamic and
pathway WHPA
Semiconfined
areally-distributed
permeabilty pathways
QA14893C
Figure 32. Flow chart for designing wellhead protection areas for confined aquifers.
98
-------
preference to the cylinder method. For the scenario with a regional hydraulic gradient that could cause
a noncircular wellhead protection area, one of the methods recommended for a sloping potentiometric
surface should be used.
(3) After the general wellhead protection area is delineated, a permeability pathway map is
made. This map defines the zones of potential, natural and artificial pathways through the aquitard,
and is important to management of activities in the wellhead protection area. High-permeability
pathways are distinguished to allow more protective measures to be taken in the more sensitive areas.
(a) For semiconfined aquifers, where significant leakage through the aquitard occurs, the entire
regional area of the wellhead protection area should be considered as having a potential for
vertical leakage.
(b) For the highly confined aquifer, the location of natural and artificial zones of leakage to the
aquifer are of prime concern, because they represent the only pathways for contaminants to reach
the producing aquifer.
99
-------
REFERENCES
Aller, L., 1984, Methods for determining the location of abandoned wells: U.S. Environmental
Protection Agency publication no. EPA-600/2-83-123,130 p.
Anzzolin, A. R., and Graham, L. L., 1984, A regulatory perspective, in Fairchild, D. M., ed., Proceedings
of the First National Conference on Abandoned Wells: Problems and Solutions, Environmental
and Ground-Water Institute, University of Oklahoma, p. 17-36.
Back, W., 1966, Hydrochemical facies and groundwater flow patterns in northern part of Atlantic
Coastal Plain: U.S. Geological Survey Professional Paper 498-A, 42 p.
Barnes, V. E., Project Director, 1974, Austin Sheet: The University of Texas at Austin, Bureau of
Economic Geology, Geologic Atlas of Texas, scale 1:250,000.
Bates, R. L., and Jackson, J. A., eds., 1987, Glossary of Geology, Third Edition: Alexandria, Virginia,
American Geological Institute, 788 p.
Bear, J., and Jacob, M., 1965, On the movement of water bodies injected into aquifers: Journal of
Hydrology, v. 3, p. 37-57.
Blanford, P., and Huyakorn, P., 1990, WHPA: an integrated semi-analytical model for the delineation
of wellhead protection areas: Report prepared for the U.S. Environmental Protection Agency,
Office of Ground-Water Protection, under subcontract to ICF, Inc., contract no. 68-08-0003.
Cohen, P., and Kimmel, G. E., 1970, Status of salt-water encroachment in 1969 in southern Nassau and
southeastern Queens Counties, Long Island, New York: Geological Survey Research, 1970-D,
p. D281-D286.
Davis, S. N., and DeWiest, R. J. M., 1966, Hydrogeology: New York, New York, John Wiley and Sons,
463 p.
Dorhofer, G., and Fritz, J., 1988, Hydrogeological experiences with hazardous waste disposal sites in
near-surface claystone formations and conclusions for the planning of future sites, in Hydrogeology
and safety of radioactive and industrial hazardous waste disposal: Orleans, France, Proceedings,
International Symposium of the International Association of Hydrogeologists, v. 1, p. 283-296.
101
-------
Driscoll, F. G., 1986, Groundwater and wells, second edition: St. Paul, Minnesota, Johnson Division,
1089 p.
Dutton, A. R., and Richter, B. C, 1990, Regional geohydrology of the Gulf Coast aquifer in Matagorda
and Wharton Counties, Texas: development of a numerical model to estimate the impact of
water-management strategies: Prepared for the Lower Colorado River Authority, The University
of Texas at Austin, Bureau of Economic Geology, under contract no. IAC(88-89)0910,116 p.
Eccles, L. A., Klein, J. M., and Hardt, W. F., 1976, Abatement of nitrate pollution in a public-supply
well by analysis of hydrologic characteristics: Ground Water, v. 14, no. 6, p. 449-454.
Fairchild, D. M., Hall, B. J., Cantu, L. W., 1981, Prioritization of ground water pollution potential of
oil and gas field activities in the Garber-Wellington area: National Center for Ground-Water
Research, University of Oklahoma Report 81-21.
Fogg, G. E., and Kreitler, C. W., 1982, Ground-water hydraulics and hydrochemical facies in Eocene
aquifers of the East Texas Basin: The University of Texas at Austin, Bureau of Economic Geology
Report of Investigations no. 127,75 p.
Follett, 1970, Ground-water resources of Bastrop County, Texas: Texas Water Development Board
Report 109,138 p.
Freeze, R. A., and Cherry, J. A., 1979, Groundwater: Englewood Cliffs, New Jersey, Prentice-Hall Inc.,
604 p.
Fritz, P., and Fontes, J. Ch., eds., 1980, Handbook of environmental isotope geochemistry: Amsterdam,
Elsevier Scientific Publishing Company, 545 p.
Gass, T. E., Lehr, J. H., Heiss, H. W., Jr., 1977, Impact of abandoned wells on ground water: U.S.
Environmental Protection Agency (EPA/600/3-77-095).
Gera, F., and Chapman, N. A., 1988, Large-scale fracture permeability of mudrocks, in Hydrogeology
and safety of radioactive and industrial hazardous waste disposal: Orleans, France,
International Symposium of the International Association of Hydrogeologists, v. 1, p. 357-366.
Greis, J. P., Rahn, P. H., and Baker, R. K., 1976, A pump test in the Dakota Sandstone at Wall, South
Dakota: South Dakota State Geological Survey Circular 43, 9 p.
102
-------
Grisak, G. E., and Cherry, J. A., 1975, Hydrologic characteristics and responses of fractured till and clay
confining a shallow aquifer: Canadian Geotechnical Journal, v. 12, no. 23, p. 23-43.
Hantush, M. S., 1959, Nonsteady flow to flowing wells in leaky aquifers: Journal of Geophysical
Research, v. 64, no. 8, p. 1043-1052.
1960, Modification of the theory of leaky aquifers: Journal of Geophysical Research, v. 65,
no. 11, p. 3713-3725.
Henry, C. D., and Basciano, J. M., 1979, Environmental geology of the Wilcox Group Lignite Belt, East
Texas: The University of Texas at Austin, Bureau of Economic Geology Report of Investigations
No. 98,28 p.
Herrera, I., 1970, Theory of multiple leaky aquifers: Water Resources Research, v. 6, no. 1, p. 185-193.
Herrera, I., and Figueroa, G. E., 1969, A correspondence principle for the theory of leaky aquifers:
Water Resources Research, v. 5, no. 4, p. 900-904.
Herzog, B. C., Griffins, R. A., Stohr, C. J., Follmer, L. R., and Su, U. S., 1989, Investigation of failure
mechanisms and migration of organic chemicals at Wilsonville, Illinois: Ground Water
Monitoring Review, v. 9, no. 2, p. 78-81.
Jackson, R. E., and Patterson, R. J., 1989, A remedial investigation of an organically polluted outwash
aquifer: Ground Water Monitoring Review, v. 9, no. 3, p. 119-125.
Javandel, I., Tsang, C. F., and Witherspoon, P. A., 1988, Hydrologic detection of abandoned wells near
proposed injection wells for hazardous waste disposal: Water Resources Research, v. 24, no. 2,
p. 261-270.
Johnson, R. H., and Miller, J. A., 1988, Region 24, southwestern United States, in Back, W., Rosenshein,
J. S., and Seaber, P. R., eds., Hydrogeology, the geology of North America, v. 0-2, p. 229-236.
Keller, B., Hoylman, E., and Chadbourne, J., 1987, Fault controlled hydrogeology at a waste pile:
Ground Water Monitoring Review, v. 7, no. 2, p. 60-63.
Kreitler, C. W., Guevara, E. H., Granata, G. E., and McKalips, D. G., 1977, Hydrogeology of Gulf Coast
aquifers, Houston-Galveston: Texas Gulf Coast Association of Geological Societies Transactions,
103
-------
v. 25, p. 72-89. Reprinted as The University of Texas at Austin, Bureau of Economic Geology
Geological Circular 77-4,18 p.
Kreitler, C. W., and Pass, D., 1980, Carbon-14 dating of Wilcox aquifer ground water, in Kreitler, C. W.,
and others, Geology and geohydrology of the East Texas Basin, a report on the progress of nuclear
waste isolation feasibility studies (1979): The University of Texas at Austin, Bureau of Economic
Geology Geological Circular 80-12, p. 79-82.
Lai, R. Y. S., and Su, C. W., 1974, Nonsteady flow to a large well in a leaky aquifer: Journal of
Hydrology, v. 22, p. 333-345.
Lohman, S. W., 1972, Ground-water hydraulics: U.S. Geological Survey Professional Paper, 70 p.
Loskot, C. L., Sandeen, W. M., and Follett, C. R., 1982, Ground-water resources of Colorado, Lavaca, and
Wharton Counties, Texas: Texas Department of Water Resources Report no. 270, 240p.
McDonald, M. G., and Harbaugh, A. W., 1980, A modular three-dimensional finite-difference ground-
water flow model: Reston, Virginia, U.S. Department of the Interior, U.S. Geological Survey
National Center, 528 p.
McGowen, J. H., Brown, L. F., Jr., Evans, T. J., Fisher, W. L., and Groat, C. G., 1976, Environmental
geologic atlas of the Texas Coastal Zone: The University of Texas at Austin, Bureau of Economic
Geology, 98 p.
Meiri, D., 1989, A tracer test for detecting cross contamination along a monitoring well column: Ground
Water Monitoring Review, v. 9, no. 2, p. 78-81.
Meisler, H., Miller, J. A., Knobel, L. L., and Wait, R. C., 1988, Region 22, Atlantic and eastern Gulf
Coastal Plain, in Back, W., Rosenshein, J. S., and Seaber, P. R., eds., Hydrogeology, the geology
of North America, v. 0-2, p. 209-218.
Myers, B. N., 1969, Compilation of results of aquifer tests in Texas: Texas Water Development Board
Report 98,532 p.
Neuman, S. P., and Witherspoon, P. A., 1969a, Theory of flow in a confined two-aquifer system: Water
Resources Research, v. 5, no. 4, p. 803-816.
104
-------
1969b, Applicability of current theories of flow in leaky aquifers: Water Resources
Research, v. 5, no. 4, p. 817-829.
1972, Field determination of the hydraulic properties of leaky multiple aquifer systems:
Water Resources Research, v. 8, no. 5, p. 1284-1298.
Payne, B. R., Quijano, L., and Latorre, C, 1978, Study of the leakage between two aquifers in
Hermosillo, Mexico, using environmental isotopes, in Application of isotope techniques to arid-
zone hydrology, p. 113-130.
Pearson, F. J., and Hanshaw, B. B., 1970, Sources of dissolved carbonate species in groundwater and their
effects on carbon-14 dating: Isotope Hydrology, p. 271-285.
Pearson, F. J., and White, D. E., 1967, Carbon-14 ages and flow rates in Carrizo sand, Atascosa County,
Texas: Water Resources Research, v. 3, no. 1, p. 251-261.
Poland, J. F., and Stewart, G. L., 1975, New tritium data on movement of ground water in western Fresno
County, California: Water Resources Research, v. 11, no. 5, p. 716-724.
Richter, B. C, Dutton, A. R., and Kreitler, C. W., 1990, Identification of sources and mechanisms of salt-
water pollution affecting ground-water quality: a case study, West Texas: The University of
Texas at Austin, Bureau of Economic Geology Report of Investigations No. 91,43 p.
Robertson, W. D., and Cherry, J. A., 1988, Tritium as an indicator of recharge and dispersion in a
groundwater system in Central Ontario: Water Resources Research, v. 25, no. 6, p. 1097-1109.
Schwartz, F. W., Cherry, J. A., and Roberts, J. R., 1982, A case study of a chemical spill:
polychlorinated biphenyls (PCB's) 2. Hydrogeological conditions and contaminant migration:
Water Resources Research, v. 18, no. 3, p. 535-545.
Senger, R. S., Collins, E. W., and Kreitler, C. W., 1990, Hydrogeology of the northern segment of the
Edwards aquifer, Austin Region: The University of Texas at Austin, Bureau of Economic Geology
Report of Investigations No. 192,58 p.
Shafer, J. M., 1987, GWPATH: Interactive ground-water flow path analysis: Illinois State Water
Survey, Champaign, Bulletin 69, 42 p.
105
-------
Smith, D. B., Downing, R. A., Monkhouse, R. A., Otlet, R. L., and Pearson, F. J. 1976, The age of
groundwater in the chalk of the London Basin: Water Resources Research, v. 12, no. 3, p. 392-404.
Southern Water Authority, 1985, Aquifer protection policy: Worthing, West Sussex, England,
Guildbourne House, B1V11LD, 47 p.
Theis, C. V., 1935, The relationship between lowering of the piezometric surface and the rate and
duration of discharge of a well using ground-water storage: American Geophysical Union, v. 16,
p. 519-526.
Thiem, A., 1906, Hydrologische Methoden: Leipzig, Gebhardt, 56 p.
Thompson, G. M., and Hayes, J. M., 1979, Trichlorofluoromethane in groundwater—a possible tracer and
indicator of groundwater age: Water Resources Research, v. 15, no. 3, p. 546-554.
Todd, D. K., 1980, Groundwater hydrology, second edition: John Wiley and Sons, 535 p.
U.S. Environmental Protection Agency, 1987, DRASTIC: a standardized system for evaluating ground-
water pollution potential using hydrogeologic settings, EPA/600/2-87/035, Ada, Oklahoma,
455 p.
U.S. Environmental Protection Agency, 1987, Guidelines for delineation of wellhead protection areas:
U.S. Environmental Protection Agency, Office of Ground-Water Protection, 139 p.
U.S. Soil Conservation Service, 1979, Soil survey of Bastrop County, Texas: U.S. Department of
Agriculture, 73 p.
U.S. Soil Conservation Service, 1979, Soil survey of Bastrop County, Texas: U.S. Department of
Agriculture, 43 p.
Van der Heijde, P., and Beljin, M. S., 1988, Model assessment for delineating wellhead protection areas,
U.S. Environmental Protection Agency, Office of Ground-Water Protection.
Vecchioli, J., Hunn, J. D., and Aucott, W. R., 1989, Evaluation of methodology for delineation of
protection zones around public-supply wells in west-central Florida: U.S. Geological Survey
Water Resources Investigations Report 88-4051, 36 p.
Wait, R. L., and McCollum, M. J., 1963, Contamination of fresh-water aquifers through an unplugged oil
test well in Glynn County, Georgia: Georgia Survey Mineral Newsletter, v. 16 (3-4), p. 74-80.
106
-------
Walton, W. C, 1962, Selected analytical methods for well and aquifer evaluation: Illinois State
Water Survey Bulletin 49.
Walton, W. C., 1979, Review of leaky artesian aquifer test evaluation methods: Ground Water, v. 17,
no. 3, p. 270-283.
Walton, W. C., 1987, Groundwater pumping tests: design and analysis: Chelsea, Michigan, Lewis
Publishers, 201 p.
Wigley, T. M. L., 1975, Carbon-14 dating of groundwater from closed and open systems: Water Resources
Research, v. 11, no. 2, p. 324-328.
Williams, R. E., and Farvolden, R. N., 1967, The influence of joints on the movement of ground water
through glacial till: Journal of Hydrology, v. 5, p. 163-170.
107
-------
APPENDIX 1
COMPARISON OF WELLHEAD PROTECTION AREAS
TWO EXAMPLES
-------
Bastrop, Texas
Example from the Updip Section of a Confined Aquifer
The first wellhead protection example for confined aquifers is the Camp Swift well field in
Bastrop County, Central Texas (fig. 29). The well field (fig. 30) is located about 5 mi north of the City
of Bastrop and south of the Camp Swift Military Reservation. The well field consists of two active
wells and eight inactive and abandoned wells. Specifically, wellhead protection areas have been
established for wells 516 and 515 (fig. 30); well 516 is the main water supply well and produces from an
approximate depth of 500-700 ft, and well 515 is used as backup during high demand in the summer
months and produces from a shallower depth of approximately 250-550 ft. The Camp Swift well field
is operated by the Aqua Water Supply Corporation, a local water cooperative which supplies water to
the town of Bastrop and to rural areas in Bastrop, Lee, and Milam Counties, Texas, for a population of
approximately 20,000. The well field is bounded to the south and west by a Federal Prison Facility, to
the north by the University of Texas Cancer Research Institute, and to the east by a trailer park and
small industrial park. Within 1 mi to the west of the well field, the Lower Colorado River Authority
operates a medium-sized open-pit lignite mine.
Hydrogeologic Setting
The area is characterized by a dry, subhumid climate with an annual precipitation of about
36.7 inches, which is less than the average annual potential evaporation (Follett, 1970). The
topography is characterized by gently rolling to undulating hills with generally less than 150 ft of
relief.
The area is in the updip part of the Gulf Coast Sedimentary Basin, a thick wedge of sedimentary
rocks, ranging in age from Cretaceous to Quaternary. Ground water is produced from the Wilcox aquifer,
which is composed of fluvial, deltaic, and marine deposits of Eocene age. The Wilcox strata crop out in
111
-------
broad parallel bands that trend to the northeast and dip gently to the southeast at approximately 2 to
3 degrees (fig. 33).
The well field is located within the outcrop area of the lower Eocene Wilcox Group, which is
comprised of three formations, (1) the Hooper Formation, (2) the Simsboro Formation, and (3) the
Calvert Bluff Formation. The Simsboro Formation consists of relatively sand-rich fluvial deposits and
is the main waterbearing unit in the area. The recharge area for the Simsboro is along a 1- to 3-mi-wide
outcrop belt that is about 2 mi west of the well field. Several faults have been identified in the
vicinity of the Camp Swift area, but are relatively minor and probably have no influence on the
regional ground-water flow regime.
Determining Confinement
The degree of confinement of the Camp Swift well field has been evaluated using geologic,
hydrologic, and hydrochemical criteria described in earlier chapters. Only a limited number of
methods were found appropriate, and they are discussed below.
Geologic Approach
1. Geologic Map and Cross Section
The Camp Swift well field is located on an outcrop of the Calvert Bluff Formation, the uppermost
unit of the Wilcox Group (fig. 33). The Calvert Bluff Formation consists of fine- to coarse-grained sands
and sandstones, interbedded with clays and mudstones and is generally less than 500-ft thick in the
area. In general, this formation produces small amounts of water for domestic and livestock uses. The
underlying Simsboro Formation, the main waterbearing unit of the Wilcox, consists of fine- to coarse-
grained sands with smaller amounts of interbedded clay and mudstones and ranges in thickness from
about 100 to 300 ft. Most of the wells in the area and all of the wells at the Camp Swift well field are
completed in the Simsboro.
112
-------
Outcrop of
Wilcox aquifer
Figure 33. Geologic map of outcrop of Wilcox Group, Bastrop, Texas (Barnes, 1974).
113
-------
The subsurface distribution of sand and shales is depicted along a cross section through the north-
south oriented wells (fig. 34) based on driller's logs. The upper section is dominated by shales, whereas
the deeper section is sand-rich and is the producing zone of the different wells. Although it is difficult
to correlate the sand geometry across the section, the uppermost shale section, as much as 300-ft thick,
appears to be continuous throughout the well field and indicates a relatively thick, confining aquitard
on top of the producing aquifer. The presence of this thick, low-permeability layer shown on the
geophysical and driller's logs is not evident from surface geologic maps.
Although the Camp Swift well field is located on a Wilcox outcrop, the Calvert Bluff Formation,
which is considered regionally a minor aquifer, may act as a confining or semiconfining unit for the
aquifer unit (Simsboro Formation) due to abundant clay and shale layers within the Calvert Bluff.
2. Other Mapping Methods
Henry and Basciano (1979) developed environmental geologic maps for the Wilcox Group of East
Texas that identify areas of critical natural resources, such as aquifer recharge areas and areas of
natural hazards such as flood-plain areas. The Camp Swift well field is located in a moderate-relief,
sandy mud-oak forest, with shallow geology characterized by interbedded sand and mud and muddy
sand. The general area of the well field is considered a recharge area; however, it is not as important a
recharge area as the area to the west, corresponding to the outcrop of the Simsboro Formation.
The general soil map of Bastrop County (U.S. Soil Conservation Service, 1979) classifies the soil
at the Camp Swift well field as Axtell fine sandy loam. This type of soil formed in clayey sediments
interbedded in places with shale and sandstone. The soils have a loamy surface layer and low-
permeability lower layer with high-water capacity. The soil characteristics suggest limited recharge
potential.
Mapping artificial penetrations of the confining unit is crucial for the development of wellhead
protection strategies. Abandoned boreholes are the most likely pathways for contaminants to migrate
into a confined aquifer. Figure 30 denotes those known wells in the vicinity of the Camp Swift well
field, including those which are abandoned and no longer used.
114
-------
Southwest
Northeast
>0
6
10
6
s
.2
3
(A
o
-------
Hydrologic Approach
1. Water-Level Data in Wells
Water-level elevations for the different wells in the well field are shown on figure 34. Water
elevations are generally above the top of shale layers, indicating confined aquifer conditions. The
regional potentiometric surface for Bastrop County, based on water-level measurements primarily of
the Simsboro Formation, indicates a hydraulic gradient of 0.002 to the east-southeast in the general
dip direction of the formations. This hydraulic gradient is typical for the outcrop region of regionally
confined aquifers along the Gulf Coast.
The pattern of daily water-level variations from continuous water-level recorders can
distinguish confined and unconfined aquifers. Continuous water-level records measured from well 505
show semidiurnal variations of approximately 1 inch and thus indicate confined conditions.
2. Pumping-Test Data
(a) Extent of the cone of depression
Water levels in observation wells, as far as 3,200 ft away from the producing well, drop during
pumping. However, no water-level response was observed in well 502 during pumping of 516, which is
located 1,800 ft away; the screened interval in 516 is somewhat deeper than those in the other wells
(fig. 34), suggesting a lack of hydraulic communication between well 502 and well 516.
(b) Storativity
Calculated Storativity values from the pumping test in the well field range between 0.0003 to
0.0005 with an average value of 0.0004 (Myers, 1969). These values are typical for confined aquifers in
the Texas Gulf Coast.
(c) Leakage
In a confined or semiconfined aquifer, effects of leakage may be reflected in the drawdown curve
during a pumping test. Drawdown in wells 503, 504, and 505 (fig. 35) from the pumping test in well 502
follow the typical Theis nonleaky curve (fig. 9), suggesting a highly confined condition.
116
-------
100-
10-
1 -
Swift well field: pumping test (1942)
wells: 503, 504, and 505
well: 502 at 810gpm
Drawdown (503)
Drawdown (504)
O Drawdown (505)
.01
Time (hr)
10
100
QA 14897C
Figure 35. Log-log plot of drawdown versus time for monitoring wells 503,504, and 505 during pumping
test in well 502, Camp Swift well field.
117
-------
In comparison, the pumping test in well 516 (fig. 36) indicates a relatively flat slope, more
characteristic of leakage through an overlying aquitard. Note that the screened interval in well 516 is
somewhat deeper than those of the other wells (fig. 34). Furthermore, relatively thick sands are
shown above the screened intervals in well 516, which are separated from the screened sand interval by
a relatively thin-shale layer. Consequently, leakage inferred from the pumping test in the deeper well
516 (fig. 36) apparently does not represent leakage from a shallow unconfined aquifer, but rather is
leakage from a shallower sand layer of the confined aquifer that is not screened (fig. 34).
Hydrochemical Approach
1. General water chemistry
The chemical composition of ground water in a regionally extensive, confined sandstone aquifer
typically shows a general change from a Ca-HCOs water in the shallow recharge sections to an Na-
HCOs type for deeper ground water as a result of chemical reaction with aquifer rock. Thus, the general
chemical composition of ground water can be used to infer the relative age of the ground water.
Figure 37 shows the distribution of hydrochemical facies in Bastrop County for the Wilcox Group
aquifer. Those wells completed in the Simsboro Formation are marked separately. In the vicinity of the
Camp Swift well field, the Ca-HCOa-type water, a recharge-type water, extends relatively far
downdip in the Simsboro. Toward the south, water in the Simsboro Formation is mostly a Na-HCQj-
type ground water, a water typical of older waters in a confined section. To the north, ground water
shows a more complex facies distribution which is probably related to mixing of different waters and
possibly different water-rock reactions. Although the water from the Camp Swift well field appears
chemically to be recharge-type waters, the regional ground-water chemistry appears to be sufficiently
complex to prevent a conclusion on the presence of confinement. A simple downdip evolution of ground
water is not apparent.
118
-------
100'
S
own
«
Q
Camp Swift well
36-hour pumping text (1986)
.01
Time(hr)
10
100
QAU898C
Figure 36. Log-log plot of drawdown versus time for pumping Camp Swift well 516 during 36-hr pumping
test in 1986.
119
-------
EXPLANATION
Woter
Cq
I No-
Wllco» group (no Simsboro)
totol dissolved solids [mg/L]
Simsboro Formation
total dissolved solids
types •
HC03
Mixed-HCO,
ID Ca-»
A Co-<
Co-
Cou
HCOj facies
boundary
•<> Mi«ed-CI
O No-CI
9 No-Mi«,d
• Mi»d-Mi>«d
5040 Corrected 14C age
\
\
QAI4895\
Figure 37. Distribution of hydrochemical facies and total dissolved solids and calculated carbon-14
ages for the Wilcox Group aquifer and Simsboro Formation.
120
-------
Changes in the chemistry of water over time from a well may indicate vertical leakage through
an overlying aquitard. At the Camp Swift well field, which has produced for almost 50 yr, no trends in
variations of the chemical composition of ground water could be identified.
2. Carbon-14 age determination
Selected samples in the Bastrop area were analyzed for 14C. Corrected 14C ground-water ages,
using the 813C approach (Pearson and White, 1967), range from about 4,490 yr to as much as 18,400 yr
(figs. 34 and 37). The generally old age determined of the ground water in the area indicates a
relatively long flow path from the recharge area to the well. Note that the Na-HCOa-type ground
water is much older than the Ca-HCOj-type water.
3. Tritium
Tritium analyses performed on the same water samples (figs. 34 and 37) as those with 14C
analyses showed zero tritium concentration and indicate that the water is older than 40 yr. This is
expected due to the old age determined from the 14C analyses. The absence of tritium also indicates
that no water has recharged relatively quickly by leakage along fractures or artificial penetrations
and mixed with old ground water.
Conclusions on Confinement
The Camp Swift well field is considered highly confined. The main indications are the absence of
any tritium, the old 14C ages, and the highly confined response from aquifer pump testing. Although
the general ground-water chemistry at the well field is characterized by Ca-HCOs-type water,
typical for recharge water, the tritium and 14C data indicate that it is, nevertheless, ground water
that was recharged a long time ago. Pumping-test data from wells 503, 504, and 505, representing the
shallower zone, exhibit highly confined conditions. Pumping-test data from the deeper confined zone in
well 516 indicate some leakage. The observed leakage in the deeper confined zone most likely
originates from the shallower confined strata that were not screened rather than from shallow water-
table aquifers.
121
-------
Wellhead Protection Area Delineation
A wellhead protection area is delineated for the two main wells of the Camp Swift well field,
well 516 in the main, deeper producing zone and well 515 in the shallower zone. Although pumping-test
data from well 515 were not available, this particular well is located between wells 502 and 503 (fig.
34). Pumping tests were performed in well 502 using wells 503,504, and 505 as monitoring wells (fig. 35).
Screens in well 515 are assumed to be at similar intervals as 502; it is therefore reasonable to assume
that hydraulic properties determined from a pump test in 502, and measured monitoring wells 503, 504,
and 505, are representative for well 515.
Cone of Depression Approach
The lateral extent of the cone of depression for the shallower and deeper production zones has
been estimated with two methods: (1) analytical methods that either calculate or measure drawdown
versus distance and (2) numerical modeling to calculate the extent of the cone of depression. The
analytical methods assume that the regional hydraulic gradient is zero. Only for the numerical
modeling method is the regional gradient considered.
Analytical Solutions and Simple Computer Models Method. Well 516—The radius of the cone of
depression is estimated from the 36-hr pumping test (fig. 36) at well 516, using a semilog plot of
drawdown versus time. The corresponding semilog plot of drawdown versus distance can be constructed
by multiplying the slope of the time-drawdown curve by (-2) and plotting the curve on a semilog plot of
distance versus drawdown. The latter curve passes through a point representing measured drawdown at
the pumping well (distance equals zero) or at a monitoring well (at known distance from pumping well);
when the curve is extrapolated to 0 ft drawdown, the lateral extent of the cone of depression was
determined to be approximately 3,500 ft.
122
-------
Some uncertainty exists because the distance-drawdown curve is based on the measured drawdown
at the pumping well and not at an observation well. Drawdown at the well may be affected by well loss
and could be greater than actual water levels in the formation adjacent to the well. The distance-
drawdown curve may therefore overestimate the extent of the cone of depression. Water-level
measurements in observation wells would yield better information on the cone of depression, as they are
not affected by well loss.
Analytical solutions for equilibrium (Thiem equation) and nonequilibrium conditions (Theis
equation) can also be used to estimate the extent of the cone of depression. Calculating the extent of the
cone of depression requires: estimates of transmissivity (a value of 34,500 gal/day/ft is obtained from
the 36-hr pumping test in well 516 [fig. 36]), the pumpage rate (1,200 gpm), the well radius (0.5 ft), and
a drawdown value at the well (84 ft). Assuming equilibrium conditions (Thiem equation), the radius of
influence extends to 18,600 ft. For nonequilibrium, fully confined conditions (Theis equation) the radius
of the 1-ft drawdown contour extends to 8,300 ft after 36-hr pumpage. Although some leakage could be
inferred from the pumping-test data (fig. 36), the leakage rate was small and did not decrease the
extent of the cone of depression when using either the Theis curve or the leaky type curves.
Well 515—The lateral extent of the cone of depression for the shallower aquifer (for example,
well 515) was also calculated. Measured drawdown in monitoring wells 503, 504, and 505 (figs. 30 and
35) during the pumping test in well 502 were used to estimate the extent of the cone of depression. For
each monitoring well, the measured drawdown at a given time after pumping started was plotted
against the distance of the monitoring well from the pumping well. The intercept with the zero
drawdown line gives the extent of the cone of depression. The drawdown measurements from the three
monitoring wells after 8-hr pumpage indicate a similar lateral extent of the cone of depression of about
3300 ft. After 55 hr of pumpage, the cone extends to about 10,000 ft.
Assuming equilibrium conditions, the zone of influence around well 515 ranges between 5,750 and
6,750 ft, based on a pumpage rate of 810 gpm, an average transmissivity of 27,770 gal/day/ft, and a well
123
-------
radius of 0.5 ft. By using the Theis equation, the radius of the 1-ft drawdown contour extends to about
8,790 ft after pumping for 60 hr.
Well 515—The cone of depression was simulated for the production zone of well 515.
Transmissivities calculated from the pumping-test data can be used in a numerical model to check the
analytical approach and to incorporate complexities, such as heterogeneous transmissivity and
regional hydraulic gradients. For the Camp Swift well field, a numerical model was constructed that
incorporates the aquifer as a single layer with initially uniform transmissivity. In addition, a uniform
hydraulic gradient of 0.002 was assumed across the model in a west-east direction representing the
regional hydraulic gradient in the Wilcox aquifer. The lateral dimensions of the model were
10,000 x 10,000 ft; the area was discretized by a 40 x 40 finite-difference grid. The model was
implemented with the program MODFLOW, a USGS finite-difference ground-water flow model
(McDonald and Harbaugh, 1980).
Using a uniform transmissivity value of 27,700 gal/day/ft based on the pumping test results at
well 502 (pumping test in the shallow unit), the model could not reproduce the observed water-level
declines in well 503, 504, and 505 (fig. 35). However, by reducing transmissivity by a factor of 4 in the
west-east direction, perpendicular to the north-south orientation of the wells in the well field (fig. 30),
simulated drawdown compared reasonably well with observed values (fig. 35). The simulated cone of
depression is an ellipse with the long axis in the direction of the well configuration and the short axis
perpendicular to a line through the wells. The short axis is approximately parallel to the dip
direction of the hydrostratigraphic units. The drawdown ellipse along the axis extends approximately
5,000 ft (1-ft drawdown contour) along the short axis for a 60-hr pump test, whereas the ellipse along
the long axis extends as far as 9,000 ft. Due to the reduced transmissivity in the general direction of the
regional hydraulic gradient, the downdip extent of the cone of depression is only 500 ft shorter than the
updip extent of the cone. The regional hydraulic gradient may not significantly alter the shape of the
cone of depression for well 515. In this case geologic variability may be a more important control on the
shape of the cone of depression than the regional potentiometric gradient.
124
-------
Well 516—Water-level declines from the 36-hr pumping test in well 516 (deeper production zone)
were also simulated using the MODFLOW model. A reasonable drawdown in the pumping well could be
simulated assuming either isotropic or anisotropic conditions. For anisotropic conditions, transmissivity
values of 34,500 gal/day/ft in the direction of the Camp Swift wells and 8,625 gal/day/ft
perpendicular to the well alignment were used. The drawdown ellipse for the 36-hr pumping test in
well 516 extends 4,500 ft along the short axis, and 8,000 ft along the long axis (along the line through
the wells in the well field). MODFLOW only calculates the cone of depression and does not calculate
flow paths.
Time of Travel Approach
Time of travel calculations were used to estimate the wellhead protection area for the shallower
and deeper production zones. Calculations of times of travel for the two wells were done independently
because the two main producing wells are not in hydraulic communication.
Cylinder Method. The cylinder method used by the U.S. Environmental Protection Agency (1987),
described in an earlier section, uses a volumetric-flow equation that determines the radius of a cylinder
from which all the water would be pumped out after a defined period of time. Using the 40-year time of
travel, a radius of about 5,000 ft is calculated for well 516 (fig. 38), based on a pumpage rate of
1,200 gpm, a screened interval of 175 ft, and a porosity of 0.25. In comparison, the cylinder radius for
well 515 is only 3,400 ft, based on a pumpage rate of 810 gpm, screened interval of 250 ft, and porosity of
0.25.
Cone of Depression/Time of Travel Method. An analytical estimate (cone of depression/time of
travel method) of the position of the 40-yr time of travel contour can be obtained from the slope of the
drawdown curve (fig. 36) for the 36-hr pumping test for well 516. The calculated radius is 4,000 ft,
which is slightly greater than the inferred radius of the cone of depression using the semilog plot.
125
-------
WHPA APPROACHES
WHPA model
o1 o o o £. Well 51 6
M- CO CM T- LO
I I
— — 1 — — •" •— • ' — t 1 — —
a,b c 12
i i
10,000 5000
I i
0 5000
1
3
4
10,0
Time of travel and
Cone of depression/40 yr
time of travel approaches
a WHPA 40 yr time of travel (5000 ft)
b Cylinder equation (40 yr time of
travel) (5000 ft)
c Cone of depression/40 yr time of
travel (4000 ft)
Cone of depression
approaches
1 Cone of depression based on
Jacob plot (3500 ft)
2 Cone of depression based on
numerical modeling of the
36 hour pump test (5000 ft)
3 Cone of depression based on
36 hour pump test using Theis
equation (8300 ft)
4 Cone of depression based on
Thiem equation (18,000 ft) QA 16404
Figure 38. Radial distance for wellhead protection areas for well no. 516, Bastrop, Texas. Those
distances on right side of figure used cone of depression approaches. Radial distance on left side of
figure used time of travel and cone of depression/time of travel approaches.
126
-------
Drawdown measurements from well 515 were not available, and the cone of depression/time of travel
approach could not be applied to this well.
Semianalytical Method (WHPA Model). Calculation of 40-yr time of travel toward the pumping
well was done using the wellhead protection area software package (Blanford and Huyakorn, 1990),
which was developed for Environmental Protection Agency's wellhead protection program. WHPA is
an integrated semianalytical model for the delineation of wellhead protection areas.
Figure 38 shows capture zones for the 5-, 10-, 20-, 30-, and 40-yr time of travel for well 516. The
configuration is completely symmetric assuming isotropic transmissivity and no regional hydraulic
gradient. Using an isotropic transmissivity of 34,500 gal/day/ft and a pumpage rate of 1,000 gpm, the
40-yr capture zone extends about 5,000 ft from the pumping well. Assuming a regional hydraulic
gradient from left to right (west to east) of 0.002, the capture zones for the different time periods
become asymmetric (fig. 39). The 40-yr capture zone extends 7,000 ft in the upgradient direction,
whereas the downgradient extent is 3,300 ft. The lateral extent perpendicular to the regional gradient
remains constant.
The WHPA program (Blanford and Huyakorn, 1990) does not incorporate the anisotropic
transmissivities which were inferred from the numerical model simulations of water-level declines
associated with the pumping test in well 502. However, when calculating capture zones that correspond
to reduced transmissivity (lower by a factor of 4), the asymmetry of the capture zones is significantly
reduced. With this lower transmissivity in the WHPA program, the resulting difference in distance
between the upgradient and downgradient extent is less than 500 ft.
A similar flow pattern and capture zone was obtained for well 515 in the shallower production
zone, based on a pumpage rate of 810 gpm, an isotropic transmissivity of 27,700 gal/day/ft, and a
regional hydraulic gradient of 0.002. For this well, the upstream extent of the 40-yr capture zone is
4,700 ft, whereas the downstream boundary extends to 2,400 ft.
127
-------
20,000
15,000 -
CD
CD
LL
10,000 -
5000 -
40-yr TOT,
regional gradient
.002
40-yr TOT,
no regional gradient
_ Down _
gradient
5000
I 1
10,000
Feet
15,000
20,000
OA14900C
Figure 39. Capture zones for well 516 for the 5-, 10-, 20-, 30-, and 40-yr time of travel assuming a
regional hydraulic gradient of 0.002. The 40-yr time of travel contour for the no-gradient scenario is
included for comparison.
128
-------
Recommended Wellhead Protection Area
Because of the higher pumpage rate, the 40-yr capture zone of well 516 includes nearly the entire
capture zone of well 515, which is about 1,800 ft from well 516. Drawdown in the two wells is assumed
not to interfere, based on the different hydrologic responses and the fact that screened intervals in well
516 are deeper than those in the other wells (fig. 34). It is therefore assumed that the capture zones for
the two wells overlap, but do not interfere with each other.
Figure 38 shows the cone of depression calculation for well 516, using different methods.
Analytical solutions of the Thiem equation for equilibrium conditions and the Theis equation for
nonequilibrium conditions result in very large wellhead protection areas, which exceed the 40-yr time
of travel contour, as computed by the WHPA program. The cone of depression/time of travel method
using a 40-yr threshold, the cylinder method, and the WHPA program are recommended. Calculated
radii of protection zones range from 4,000 to 5,000 ft, assuming isotropic conditions and no regional
hydraulic gradient. Using the observed regional hydraulic gradient of 0.002, capture zones computed by
the WHPA program become asymmetric, that is, the 40-yr capture zone extends 7,000 ft in the
upgradient direction and 3,300 ft in the downgradient direction (fig. 39). The WHPA program does not
incorporate effects of anisotropy. Anisotropy of transmissivity was inferred from the numerical model
calibration of pumping-test results, with reduced transmissivity in the direction of the regional
hydraulic gradient. Incorporating effects of anisotropy reduces the effect of the regional hydraulic
gradient, resulting in a more circular wellhead protection area with a shorter upstream distance but
increased downstream distance. Therefore circular wellhead protection areas were chosen (fig. 30).
As discussed earlier, the aquifer at the Camp Swift well field is considered to be highly
confined, that is, it has a low probability of contamination. The main pathways for contamination are
localized, such as improperly sealed, abandoned wells and boreholes. Figure 30 shows the
recommended wellhead protection area, which includes an overlay of the 40-yr capture zone for the
two producing wells and local protection zones in the vicinity of any existing well, representing higher-
129
-------
priority protection zones. In case the exact locations of abandoned wells are not known, the local, high-
priority protection zone is enlarged to be certain that reported wells are included (noted by dashed
circles).
130
-------
Wharton, Texas
Example from the Downdip Section of a Confined Aquifer
The second example of delineation of a wellhead protection area in a confined aquifer is for a
well in the well field of the City of Wharton, Wharton County, located in the Gulf Coastal Plain of
southeastern Texas (fig. 29). The City of Wharton is about 60 mi west of Houston and about 50 mi north
of the coast of the Gulf of Mexico. The city water wells are located on empty lots throughout the city
(fig. 31). A wellhead protection area is designed for City of Wharton well 3 (also called 402), which is
screened from 600 to 900 ft in the Willis Sand of the Chicot aquifer.
Hydrogeologic Setting
The area is humid, subtropical, and annual rainfall averages 41 inches per year, which is less
than average annual potential evaporation (Loskot and others, 1982). The topography is relatively
flat, characteristic of coastal plains of low relief. In Wharton County, the main hydrogeologic units
consist of Pleistocene and Pliocene sequences of gravel, sand, silt, and clay. All formations crop out in
belts that are nearly parallel to the shoreline and dip towards the Gulf of Mexico. The stratigraphic
sequence can be divided into three hydrogeologic units (Loskot and others, 1982): (1) the Chicot aquifer,
that includes the Willis Sand, Bentley Formation, Montgomery Formation, the Beaumont Clay of
Pleistocene age, and Holocene Alluvium; (2) the underlying Evangeline aquifer, that includes the
Pliocene Goliad Sand; and (3) the Burkeville confining layer that consists of the Upper Miocene
Fleming Formation and underlies the Evangeline aquifer. The shallow Beaumont Clay consists of a
thick sequence of mostly clays with only local sands and is considered a major confining unit for the
Chicot and underlying Evangeline aquifers. Locally, the Beaumont Clay can produce some ground water
from interbedded sand bodies. The Chicot aquifer reaches a depth of about 600 ft below sea level in the
vicinity of Wharton. The Willis Sand is the major waterbearing unit of the Chicot aquifer. Its updip
131
-------
outcrop, the main area of recharge from Wharton County, is in Colorado County, which is
approximately 30 mi northwest of the City of Wharton.
Determining Confinement
The degree of confinement of the well field has been evaluated using geologic, hydrologic, and
hydrochemical criteria.
Geologic Approach
1. Geologic Map and Cross Section
The wells of the City of Wharton are located on alluvium of the Colorado River. Beneath the
alluvium, the Beaumont Formation consists of thick clay with interbedded sand and acts as a confining
unit for the underlying Willis Sand. The potential for confinement is not apparent from the outcrop map
but from the subsurface data.
The subsurface distribution of sand and shale is depicted in figure 40, showing driller's logs and
geophysical logs of the municipal wells of the City of Wharton. The entire geologic section contains
interlayered sands and shales indicating the presence of confining layers, with thicker sands of the
Willis Sand occurring at greater depth. Except for wells 1 and 3, which are about 70 ft apart, the sands
of the different wells cannot be correlated to assess the lateral continuity of the clay layers (fig. 40).
2. Various Mapping Methods
Environmental geology maps by McGowen and others (1976) show clayey sands and silts as the
dominant surficial deposits. The deposits are characterized by moderate permeability, drainage, and
water-holding capacity in the Wharton area. The general soil map of Wharton County published by
the U.S. Soil Conservation Service (1979) shows the predominant soil in the City of Wharton to be of
the Miller-Norwood association. This soil type is characterized by moderately well-drained
132
-------
No. 1
402
Alabama Street
(1949)
No. 3
406
Alabama Street
(1965)
Wo. 2
404
Cloud Street
(1953)
Wo 4
Walhalla
0 -,
100 -
200 -
300 -
400 -
g 500
_o
0)
-O
600 -
0 0
i 700
«
Q
800 -
900 -
1000 -
1100 -
1200 -I
[
[
[
[
[
E
c
—
c
t
™
L_
I
I::::::::
ill
iiiiii
;•*•••**•
tJS-rr-Hg.
il.l'^'l'jP
-_T_-
— _T"_T"
Driller'
log
S3 l
H c
:!B 5
~Jj !
I] <
r** *— - CO
1960
Sea
level
Trit - 0
C-14 -
24,000 yr
5 Ge
.oam or sc
ilay
Shale
Sand (drille
Band (geoj
i .
>------
. .
ophysical
log
il
2r's log)
physical log)
ig:
JSSSSSS
!!!!!!!!!
•••••••*
••••••i:
-«
i ^f ••» . MM
_"1."V
_-;I-I-
==
0
o
y
1965
E
Trit « 0 P
... I-
C-14 = _
15,000 yr==
E
E
£
:
i— «—••—••
.
«z^zj;
1
-1.-L-
— — —
— — ^
.,•• ^ ••
. — — — .
Drillers Geophysical
log log
Mil
E
i
•.•.
iiiii
•.
:::
Illlliil!
iiiiiiiii
—
------
— •
i •» ^ M
_-^_-
— — i
_
:'-L
r_i~_i~_
=3:
-_-
1=1
L1^-"^"
-^_r.
to ^ 1~.
1975
— — —
_
-T-T-T
—_—_—_
_~^_-
~I_"7_~I.
1
C
E
C
F
t
E
^z
--T-T.
.
_ _ .
r- -r-
------
------
~ '
-
£=
1
H?I-!
—
:- —
T
--_-_-.
— L~I_"
H Screened interval
D
^ Water level
Driller's Geophysica Geophysical
log log log
OA 149B1C
Figure 40. Subsurface distribution of sand and shales based on driller's logs and geophysical logs for
City of Wharton wells.
133
-------
calcareous soils on flood plains that are underlain by Recent loamy and clayey alluvium. The
variability of surficial deposits does not give a clear indication of the potential for confinement.
Potential locations of artificial penetrations of the confining unit were obtained from maps
available from the Texas Railroad Commission, which regulates the oil and gas industry in the State.
The Commission's records indicate several oil wells at the outskirts of the City of Wharton, some of
which are within a mile of the city water wells (fig. 31), but no drilling has been conducted in the city.
If they are abandoned and inappropriately sealed, the oil wells may represent potential pathways for
contamination. The well records, however, may not be complete, and additional information on the
potential locations of artificial penetrations may be obtained from land-use maps and air photos that
indicate industrial developments in the area.
Hydrologic Approach
1. Water-Level Elevations in Wells
Water levels for the different wells in Wharton are shown on figure 40. A potentiometric surface
of the Chicot aquifer in Wharton County, however, was not constructed because of a wide range in
measured water levels vertically, laterally, and over time within the aquifer. Ground water in the
aquifer is used extensively in the county for agricultural and municipal uses which has resulted in
water-level declines as much as 50 ft over the last 20 yr. A regional hydrologic cross section (Dutton and
Richter, 1990) indicates a relatively small lateral gradient but a significant downward hydraulic
gradient (fig. 41). The regional lateral hydraulic gradient is less than 0.0005. The vertical hydraulic
gradient indicates a potential for shallow ground water to leak into the deeper aquifer units.
2. Pumping-Test Data
(a) Extent of cone of depression
A pumping test conducted at well 40 (fig. 42) did not produce drawdown in well 402, which is
located about 70 ft away from the pumping well. Note, however, that the screened intervals are at
134
-------
North
South
MATAGORDA CO
-500-
o
o
>
.2
u
-1000
Beaumont
WATER COLUMN
— Water level
— Well bottom
EXPLANATION
gO'~"~ Hydraulic head (ft)
_.,- Base of fresh waler
20 mi
30km
OAII9I4
Figure 41. Regional hydrologic cross section through Wharton and adjacent counties showing vertical
distribution of hydraulic heads (from Dutton and Richter, 1990).
135
-------
10--
I
ra
Q
1 -•
.001
.01
.1
Time (hr)
10
QA14903C
Figure 42. Log-log plot of drawdown versus time in the pumping well 406, indicating the drawdown
stabilized after about 4 min.
136
-------
different elevations in the two wells (fig. 40) and may be in poor hydraulic connection; thus, the cone of
depression of well 406 may extend more than 70 ft.
(b) Transmissivity
Attempts to calculate transmissivity from a pumping test at well 406 on May 25, 1989, were
limited due to measurement problems. Drawdown in the pumping well was determined from pressure
changes in an air line. Leakage of the air line was noticed, and the drawdown curve may be somewhat
affected by this leakage.
Using the straight-line segment of the second part of the drawdown curve (fig. 43) a
transmissivity value of about 40,000 gal/day/ft is calculated. This value is relatively high for typical
transmissivity for other wells in the Chicot aquifer. Matching the log-log plot of drawdown versus
time (fig. 42) with leaky type curves gives an estimate of about 3,870 gal/day/ft.
A transmissivity of 14,000 gal/day/ft was estimated from model calibration with available
data from the area. The 14,000 value is considered a better estimate of transmissivity.
(c) Storativity
Storativity was not obtained from the pumping test in well 406. Reported storativities for the
Chicot aquifer in the vicinity of the city of Wharton are in the order of 10~2 (Dutton and Richter, 1990).
Although the value is higher than the typical value of 10~* for a confined aquifer, abundant clay
layers within the aquifer (fig. 40) probably account for the higher Storativity in the aquifer.
(d) Leakage
The log-log plot of drawdown versus time for a pump test on well 406 follows a very typical leaky
type curve where the rate of drawdown became significantly reduced about 4 min after pumping started
(fig. 42). However, without water-level measurements in a nearby monitoring well, leakage cannot be
quantitatively estimated. Leakage from sand layers above and below the producing zone may occur
(fig. 40), because not all of the sand intervals are screened, and some of the shale layers adjacent to the
screened sand intervals are relatively thin.
137
-------
Distance (ft)
100
0 -i
1,000
10,000
-100
Well
10
I wen _i
I radius |
100
Time (min)
1,000
10,000
100,000
Calculated extent of
Cone of Depression
City of Wharton well: 406
Affected by
wellbore storage
100,000
QA14904C
Figure 43. Semilog plot of negative drawdown versus time based on a 3-hr pumping test in well 406.
Based on the slope of the straight line section, a distance-drawdown curve can be inferred, but is
considered to be unrealistic.
138
-------
Numerical Model. Due to the uncertainty in estimated transmissivities from the pumping test in
well 406, a numerical model was used to test the sensitivity of drawdown to transmissivity. The results
of the numerical model are discussed later in the wellhead protection area delineation section.
Hydrochemical Approach
\. General Water Chemistry
The distribution of hydrochemical facies along a vertical cross section in the direction of the
regional dip of the hydrostratigraphic units is shown in figure 44. Most of the ground water in Wharton
County in the Chicot and Evangeline aquifers is of a Ca-HCOs type.
The ground waters in the overlying Beaumont Formation are Na-HCOj- and Na-Cl-type waters.
This supports the hydraulic data indicating that the shallow Beaumont Formation is hydraulically
separated from the deeper aquifer units. Although an overall downward hydraulic gradient is
observed (fig. 41), shallow ground water has not reached the deeper aquifers because of the relatively
low vertical permeability of the Beaumont Formation.
Records of water-chemistry data from the Wharton City wells do not show any changes through
time, which indicates that the source of ground water has remained constant and has not been changed
by extensive pumpage during the last several decades.
2. Carbon-14 Age Determination
Absolute ground-water ages based on 14C analyses at two wells 406 and 402, were 15,000 and 24,000
yr (fig. 44), corresponding to the deeper Ca-HCOa-type and shallower Na-HCOa-type ground water,
respectively. Both waters show very great ages. The Ca-HCC»3-type water in the deeper, but more
transmissive, Chicot aquifer is younger than the Na-HCOs-type water from the shallower, less
transmissive Beaumont Formation. Ground water recharged to the Chicot from the west where the
Beaumont is absent appears to have flowed beneath the overlying Beaumont Formation.
139
-------
North
South
MATAGORDA CO
Beaumont
-5OO
-KXX>
0 Well bottom
—• Base of fresh water
-soo— Total dissolved
solids (mg/L)
EXPLANATION
Ca-HCOj
Mixed-cation-
Na-HCO3
Chicot
2O mi
30 km
QAI4905
Figure 44. Distribution of hydrochemical fades along a vertical cross section in the downdip direction
(from Dutton and Richter, 1990).
140
-------
3. Tritium
Tritium analyses of water samples collected in wells 406 and 402 both indicated tritium
concentrations below detection limit. This is consistent with the relatively great age based on 14C
analysis. The absence of any tritium also indicates that no rapid recharge occurs through localized
features such as faults and fractures allowing mixing with younger ground water.
Conclusions on Confinement
The 14C and tritium concentrations in well 406 indicate very old ground water. The producing zone
of well 406, therefore, is considered highly confined. The hydraulic head distribution indicates an
overall downward gradient; however, the difference in water chemistry and ground-water ages
between the shallow and deep sections indicates a lack of significant downward ground-water
movement. Although pumping-test data indicate a leaky behavior, leakage is interpreted to come from
vertically adjacent sand units, which are not screened.
Wellhead Protection Area Delineation
A wellhead protection area was delineated for well 406 (Wharton city well 3; fig. 31), using the
cone of depression and time of travel approaches.
Cone of Depression Approach
Analytical Solutions and Simple Computer Models Method. The semilog plot of negative
drawdown versus time (fig. 43) shows two straight-line sections, (1) from 0.2 to 3 min and (2) from 3 to
200 min. The first section is affected by well-bore storage and does not represent aquifer conditions. The
second section is affected by leakage. Using the relationship between the slope of the time-drawdown
curve and the distance-drawdown curve, the extent of the cone of depression is estimated at about
141
-------
100,000 ft (fig. 43). This value, however, is considered a gross overestimation due to the observed
leakage into the aquifer. Leakage reduces the rate of drawdown, and thereby decreases the slope of the
distance-drawdown curve. Calculating the lateral extent of a cone of depression with drawdown versus
time data may not be appropriate if the aquifer is semiconfined and characterized by significant
leakage.
Analytical solutions describing well discharge for equilibrium conditions (Thiem equation) and
nonequilibrium conditions (Theis equation) are also used for estimating the radius of the cone of
depression. Using a transmissivity of 14,000 gal/day/ft, based on the model calibration (discussed
below), a pumping rate of 940 gpm, and a storativity of 0.01, the calculated radius of the 1-ft drawdown
contour extends to 342 ft after 3 hr of pumpage.
For equilibrium conditions (Thiem equation), the extent of the cone of depression is calculated at
243 ft based on T = 14,000 gal/day/ft (fig. 45).
Because of the uncertainty in transmissivities estimated from the pumping test in well 406, a
numerical model was used to test the sensitivity of transmissivity and storage on drawdown. Although,
the hydrostratigraphy shows a highly heterogeneous aquifer (fig. 40), the simulations were performed
using a one-layer representation of the aquifer. The regional hydraulic gradient in the vicinity of
Wharton is very small (less than 0.001) and was assumed to be negligible. A semianalytical software
package (Walton, 1987) was used to simulate drawdown in a single well under a variety of conditions.
In this case Walton's program is well suited for wellhead protection delineation, as it computes not
only drawdown in a pumping well but also calculates the distance-drawdown relationship.
In a series of simulations where transmissivity, storativity, and leakage were varied, the best fit
with the observed data was obtained when using a transmissivity of 14,000 gal/day/ft, a storativity of
0.01, and an aquitard permeability that is only one order of magnitude lower than aquifer
permeability. Both high storativity in and high leakage to the aquifer can be expected, considering
the overall hydrostratigraphy (fig. 40). Based on these calibrated hydrologic properties, the
calculated 1-ft drawdown contour extends to about 350 ft from the pumping well after 3 hr of pumpage.
142
-------
WHPA APPROACHES
WHPA
model 4°yr 3°yr20yr 10yr Syr
well 406
a,b
1 2,3
5000 4000 3000 2000 1000
Time of travel and
Cone of depression/40 yr
time of travel approaches
a WHPA 40 yr time of travel (4100 ft)
b Cylinder equation (40 yr time of
travel) (4100ft)
c Cone of depression/40 yr time
of travel (1000 ft)
1000 2000 3000 4000 50C ft
Cone of depression
approaches
1 Cone of depression based on
Thiem equation (243 ft)
2 Cone of depression stabilized
by leakage based on Theis
equation (342 ft)
3 Cone of depression based on
semianalytical simulations using
the Walton method (350 ft)
QA 16405
Figure 45. Radial distance for wellhead protection areas for well no. 406 for Wharton, Texas. The
distances on right side of figure used cone of depression approach. Radial distance on left side of figure
used time of travel approach and cone of depression/time of travel approaches.
143
-------
Time of Travel Approach
Cylinder Method. Using the cylinder method, a radius of about 4,100 ft (fig. 45) was calculated
for a 40-yr time period for well 406, based on a pumpage rate of 940 gpm, a screened interval of 200 ft,
and a porosity of 0.25. This approach assumes no vertical leakage.
Cone of Depression/Time of Travel Method. Using the cone of depression/time of travel method
with a 40-yr threshold, and a hydraulic gradient based on the lateral extent of the cone of depression
of 350 ft and 90 ft drawdown at the pumping well, the wellhead protection radius is approximately
1,000 ft (fig. 45). In this case the 40-yr time of travel contour is larger than the lateral extent of the cone
of depression.
Semianalytical Method (WHPA Model). The 40-yr time of travel for the pumping well was
calculated using the WHPA computer program (Blanford and Huyakorn, 1990). As mentioned before,
the version of the WHPA program that was used did not include the effects of leakage and thereby
assumed a distance-drawdown curve typical for highly confined aquifers. Using a transmissivity value
of 14,000 gal/day/ft, the 40-yr time of travel contour computed by the WHPA program extends to about
4,000 ft from the pumping well (fig. 45). With the regional hydraulic gradient assumed to be less than
0.0005 (fig. 41), all wellhead protection areas were circular in shape.
Recommended Wellhead Protection Area
A comparison of the results of the different methods is shown in figure 45. The time of travel
calculations generally yield greater capture zones than the cone of depression calculations. The 350-ft
radius for the cone of depression was based on a simulated 3-hr pump test. The actual size of the cone of
depression could not be determined. A wellhead protection radius of 1,000 ft is considered a reasonable
144
-------
approximation for this leaky aquifer based on the cone of depression/time of travel method with a 40-
yr threshold. The 4,000-ft radius from the 40-yr time of travel method using the volumetric-flow
equation (cylinder method) or the WHPA computer program was calculated without consideration of
the effects of leakage; thus, 4,000 ft overestimates the lateral extent of the cone of depression where
the cone of depression is based on a 3-hr pump test. The WHPA program and the cylinder method give a
conservative estimate of the wellhead protection area and are appropriate when information about
aquifer properties, for example, transmissivity, leakage, and storativity, is not available. Local
protection zones in the vicinity of existing wells should be established to provide higher priority
protection zones.
145
-------
APPENDIX 2
CONFINED AQUIFERS OF THE UNITED STATES, THE COMMONWEALTH OF PUERTO RICO,
AND THE PACIFIC AND CARIBBEAN TERRITORIES
by
James Hamilton
Ground-Water Protection Division
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
Washington, D.C.
-------
Introduction
Major and significant minor confined aquifers (hereafter referred to only as "confined aquifers")
occur throughout the United States, the Commonwealth of Puerto Rico, and the Pacific and Caribbean
Territories (Back and others, 1988).
The map of confined aquifers of the United States (fig. 46) primarily is based on U.S. Geological
Survey (USGS) information contained in Moody and Chase (1985). Other significant information comes
from Heath (1984), Sun (1987, 1988), Weeks and Sun (1987), and Moody and others (1988). Figure 46 also
incorporates information from Gerlach, 1970, Davies and others, 1984, and from telephone interviews
with scientists at USGS district offices. Fenneman's 1946 map was used as a guide for confined-aquifer
boundaries.
Only aquifers of drinking-, irrigation-, or stock-water quality are depicted on figure 46. Where
researchers in adjoining States do not believe that strata serving as a significant aquifer in one State
constitute a significant aquifer in the adjoining State, it was necessary to approximate the position of
the boundary separating the presence and absence of a confined aquifer, to be near the States' common
border. Dashed lines are used in figure 46 to represent such a boundary.
Acknowledgments
Numerous scientists from the U.S. Geological Survey (USGS) and the U.S. Environmental
Protection Agency (EPA) provided information on the geographic distribution of confined aquifers. The
time and efforts of these scientists are greatly appreciated.
The USGS scientists are: Gary Balding, Rick Benson, David Brown, William Carswell, David
Click, Timmy Cummings, Dan Davis, Robert Faust, Herbert Freberger, Ector Gann, Joseph Gates, Roy
Glass, Robert Graves, Steven Hindall, William Horak, Thomas Huntzinger, Jeffery Imes, Ivan James,
Richard Karsten, James Kircher, John Kline, Alfred Knight, Richard Krause, James Kroheski, Larry
149
-------
EXPLANATION
Confined aquifer
Approximate boundary
Other (does not contain recognizable/
delineatable confined aquifers)
0 600km
Scale I : I4,000,OOO
QAI527I
Figure 46. Major and significant minor confined aquifers of the United States.
150
-------
Land, Gerald Lindholm, Robert MacNish, Joe Moreland, William Oakley, Glenn Patterson, Kathy
Peter, Michael Planert, Stanley Robson, Michael Shulters, Dennis Stewart, Arturo Torres, Donald
Vaupel, John Vecchioli, John Williams, and Thomas Winterstein.
The EPA scientist is John Malleck.
General Description of Confined Aquifers
For ease of discussion, aquifers of the continental United States are grouped into four general
physiographic regions (fig. 47) (after Fenneman, 1946): (1) the Atlantic Coastal and Gulf of Mexico
Coastal Plains from New York to Mexico; (2) the Appalachian Highlands from Maine to central
Alabama, and the geologically similar Laurentian Uplands of Minnesota and Wisconsin; (3) the
Midcontinent section, consisting of the Interior Plains and Interior Highlands, with mature basins and
dissected plains; and (4) the western portion of the United States, consisting of the Rocky and Pacific
Mountain Systems and the Intermontane Plateaus and Basins.
Physiographic Region 1
The Atlantic Plain and the Gulf of Mexico Plain contain confined aquifers. The unconfined areas
within the general Physiographic Region include such aquifers as the Floridan, which is unconfined in
the outcrop area but confined where buried deeply (Sinnott and Gushing, 1978; Burchett, 1986; and
Moody and Chase, 1985).
Physiographic Region 2
All the New England States contained fractured, crystalline bedrock aquifers overlain by glacial
deposits. In New Hampshire, Vermont, Massachusetts, and Connecticut, the bedrock aquifers are
confined by overlying glacial till, and in some places by glacial-lake sediments. Rhode Island is not
151
-------
LAURENTIAN UPLANDS
(included in region 2)
GULF OF MEXICO
' COASTAL PLAIN
EXPLANATION
Physiographic region
boundary
600km
QAI5270
Figure 47. Physiographic regions of the United States.
152
-------
depicted as containing confined aquifers (fig. 46) because the till that overlies bedrock is not considered
to be confining in this State. Maine's aquifers of till, of glaciofluvial outwash, and of ice-contact
deposits are generally unconfined and are found in most of the State. Carbonate aquifers in extreme
northeastern Maine are confined (Sinnott and Gushing, 1978). Primarily unconfined crystalline aquifers
that are pervasive throughout most of Maine may be locally confined in areas too small to depict in
figure 46.
New York contains significant minor aquifers. These are: primarily unconfined carbonates;
primarily unconfined stratified drift; and in small areas, confined sandstone aquifers and confined
valley-fill deposits (Waller and Finch, 1982). South of New York, most of the northwestern half of the
Appalachian Highlands essentially is an area of confined aquifers. The southeastern half consists of
the Blue Ridge Mountains and Piedmont which contain crystalline aquifers that are primarily
unconfined in Virginia, South Carolina, and Georgia. In eastern Tennessee, northern Alabama, and
northern Georgia, the crystalline rocks are primarily unconfined (Zurawski, 1978). In North Carolina,
similar crystalline rocks have been defined as confined, low-yield aquifers by the USGS. For purposes
of this report, however, these aquifers are not considered significant because the sustained water yields
come from the overlying, saturated regolith.
The Laurentian Upland is a recently glaciated surface on unconfined crystalline rocks (Weist,
1978). In Wisconsin the aquifers are unconfined, and in the northeastern part of Minnesota, the aquifers
are a generally confined combination of crystalline, sandstone, and volcanic rocks. The southern extent
of the Laurentian Upland in Wisconsin approximates the boundary between the northern unconfined
aquifers and the more southern (Physiographic Region 3) Sandstone aquifer, that is confined in the east
by the Maquoketa Shale and is locally confined elsewhere (Moody and Chase, 1985).
Physiographic Region 3
The Midcontinent portion of the United States consists of the Interior Plains and the Interior
Highlands (Bloyd, 1974). A very large confined-aquifer area containing several extensive aquifers
153
-------
extends from Wisconsin to western Montana. One of the confined aquifers is the Fort Union Coal, which
covers large parts of Montana, Wyoming, and the Dakotas. The coal is confined except for a narrow area
around its perimeter where it either crops out or is shallow. In Wyoming, the Fort Union Coal is
underlain by carbonate and sandstone aquifers that are also confined (Reeder, 1978).
The remainder of the Region, including the extensive High Plains aquifer area, is predominantly
unconfined. Glacial-drift aquifers of northern Missouri are confined in buried valleys where overlain by
relatively thick deposits of low-permeability outwash (Taylor, 1978; Moody and Chase, 1985). These
minor aquifers are indiscernible at the scale of figure 46.
Physiographic Region 4
The Western United States is the Region of Intermontane Plateaus and Basins and the Rocky and
Pacific Mountain Systems. This Region includes the extensive Columbia River Plateau, which
encompasses southeastern Washington, eastern and central Oregon, the Snake River Plain of southern
Idaho, and the northern portions of California and Nevada (Foxworthy, 1979; Whitehead, 1986). The
confined aquifers within the area of the plateau are the Columbia River Basalt aquifers of
Washington and Oregon and the western Snake River aquifer. The volcanic and sedimentary aquifers of
the rest of the Columbia River Plateau are unconfined but may be locally confined in areas too small to
be shown in figure 46.
The aquifers in most of the rest of the Region are unconfined except for the carbonate aquifers of
the Great Basin's eastern half, located mostly in eastern Nevada and western Utah (Dettinger, 1989).
Sediments of the Central Valley in California constitute one of the Region's most extensive aquifer
systems; the southern half of the valley is confined (fig. 46) (Thomas and Phoenix, 1976; Moody and
Chase, 1985).
154
-------
Alaska, the Hawaiian Islands, and the Pacific and Caribbean Islands
Alaska has a varied and relatively complex geology. To date only one area has been determined
to contain confined aquifers. This area is on the south coast near Cook Inlet, where basins and valleys
surrounding a south-central embayment are filled with glacial till and fine-grained, glaciolacustrine
materials that are interbedded with more permeable water-worked deposits of sand and gravel. The
glacial outwash alluvium is confined by glacial, lacustrine, and estuarine deposits (Zenone and
Anderson, 1978).
The Hawaiian Islands are composed of complex volcanics that are, for the most part, unconfined.
Some basal ground-water (that is, water that floats on, or is in hydrodynamic equilibrium with, salt
water) areas of the Island of Oahu have been described as being locally confined where cap rock is
present.
In the Virgin Islands, ground water is primarily under water-table conditions except on the Island
of St. Thomas. On that island, sand and gravel beds are locally confined by overlying alluvium.
Information is not available to delineate these areas (Cosner and Bogart, 1972; Jordan and Cosner, 1973;
and Jordan, 1975).
Most of the water in Guam is produced from limestone aquifers that are primarily unconfined
(Ward and others, 1965).
Puerto Rico has a confined-aquifer area along the western and central portions of the north coast
of the main island. In this area, the Cibao Formation and the Lares Limestone are unconfined at
outcrops but are confined at depth (Torres, 1985; 1986).
155
-------
References
Back, W., Seaber, P. R., and Rosenshein, J. S., eds., 1988, Hydrogeology, Geology of North America:
Series of Geological Society of America's Decade of North American Geology project: Geological
Society of America, v. O-2, 534 p. 3 pocket plates.
Bloyd, R. M., Jr., 1974, Summary Appraisals of the Nation's Ground-Water Resources—Ohio Region:
U.S. Geological Survey Professional Paper 813-A, 41 p.
Burchett, C. R., 1986, Edwards Aquifer. U.S. Geological Survey, in cooperation with the Edwards
Underground Water District: U.S. Geological Survey Monograph, 38 p.
Cosner, O. J., and Bogart, D. B., 1972, Water in St. John, U.S. Virgin Islands, U.S. Geological Survey, in
cooperation with the National Park Service and the Government of the U.S. Virgin Islands,
Caribbean District, Open-File Report, 46 p.
Davies, W. E., Simpson, J. H., Ohlmacher, G. C., Kirk, W. S., and Newton, E. G., 1984, Engineering
Aspects of Karst: U.S. Geological Survey National Atlas of the United States of America, two-
page map.
Dettinger, M. D., 1989, Distribution of Carbonate-Rock Aquifers in Southern Nevada and the Potential
for Their Development: Summary of Findings, 1985-1988, U.S. Geological Survey Summary
Report No. 1, in cooperation with the State of Nevada, 37 p.
Fenneman, N. M., 1946, Physical Divisions of the United States: prepared in cooperation with the
Physiographic Committee of the U.S. Geological Survey, one-page map.
Foxworthy, B. L., 1979, Summary Appraisals of the Nation's Ground-Water Resources—Pacific
Northwest Region: U.S. Geological Survey Professional Paper 813-S, 39 p.
Gerlach, A. C., Editor, 1970, Productive aquifers and withdrawals from wells: U.S. Geological Survey
National Atlas Map, p. 122-123, 128.
Heath, R. C., 1984, Ground-Water Regions of the United States: U.S. Geological Survey Water-Supply
Paper 2242, 78 p.
156
-------
Jordan, D. G., 1975, A Survey of the Water Resources of St. Croix, Virgin Islands: U.S. Geological Survey
in cooperation with the National Park Service and the Government of the U.S. Virgin Islands,
Caribbean District, Open-File Report, 51 p.
Jordan, D. G., and Cosner, O. J., 1973, A Survey of the Water resources of St. Thomas, Virgin Islands:
U.S. Geological Survey in cooperation with the National Park Service and the Government of the
U.S. Virgin Islands, Caribbean District, Open-File Report, 55 p.
Moody, D. W., and Chase, E. B., 1985, National Water Summary 1984-Hydrological Events, Selected
Water-Quality Trends, and Ground-Water Resources: U.S. Geological Survey Water-Supply
Paper 2275,467 p.
Moody, D. W., Carr, Jerry, Chase, E. B., and Paulson, R. W., compilers, 1988, National Water Summary
1986-Hydrological Events and Ground-Water Quality: U.S. Geological Survey Water-Supply
Paper 2325, 560 p.
Reeder, H. O., 1978, Summary Appraisals of the Nation's Ground-Water Resources—Souris-Red-Rainy
Region: U.S. Geological Survey Professional Paper 813-K, 25 p.
Sinnott, Allen, and Cushing, E. M., 1978, Summary Appraisals of the Nation's Ground-Water
Resources—Mid-Atlantic Region: U.S. Geological Survey Professional Paper 813-1, 32 p.
Sun, R. J., 1987, Regional Aquifer-System Analysis Program of the U.S. Geological Survey—Summary of
projects 1978-1984: U.S. Geological Survey Circular 1002, 264 p.
1988, Regional Aquifer-System Analysis (RASA) Program: U.S. Geological Survey, Open-
File Report 88-118, water fact sheet, 2 p.
Taylor, O. J., 1978, Summary Appraisals of the Nation's Ground-Water Resources—Missouri Basin
Region: U.S. Geological Survey Professional Paper 813-Q, 41 p.
Thomas, H. E., and Phoenix, D. A., 1976, Summary Appraisals of the Nation's Ground-Water
Resources—California Region: U.S. Geological Survey Professional Paper 813-E, 51 p.
Torres, A., 1985, North Coast Limestone Area of Puerto Rico: U.S. Geological Survey North Coast
Limestone Progress Report 5,16 p.
157
-------
Torres, A., 1986, North Coast Limestone Area of Puerto Rico: U.S. Geological Survey North Coast
Limestone Progress Report 7,26 p.
Waller, R. M., and Finch, A. J., compilers, 1982, Atlas of Eleven Selected Aquifers in New York:
U.S. Geological Survey Water Resources Investigations Open-File Report 82-553: prepared in
cooperation with the New York State Department of Health, Albany, New York, 255 p.
Ward, P. E., Hoffard, S. H., and Davis, D. A., 1965, Geology and Hydrology of Guam, Mariana Islands:
U.S. Geological Survey Professional Paper 403-H, Chapters A-I, 28 p.
Weeks, J. B., and Sun, R. J., 1987, Regional Aquifer-System Analysis Program of the U.S. Geological
Survey—Bibliography 1978-1986: U.S. Geological Survey Water Resources Investigations Report
87-4138,81 p.
Weist, W. G., Jr., 1978, Summary Appraisals of the Nation's Ground-Water Resources—Great Lakes
Region: U.S. Geological Survey Professional Paper 813-J, 30 p.
Whitehead, R. L., 1986, Geohydrologic Framework of the Snake River Plain, Idaho and Eastern
Oregon: U.S. Geological Survey Hydrological Investigations Atlas HA-681, three sheets.
Zenone, C., and Anderson, G. S., 1978, Summary Appraisals of the Nation's Ground-Water Resources—
Alaska Region: U.S. Geological Survey Professional Paper 813-P, 28 p.
Zurawski, Ann, 1978, Summary Appraisals of the Nation's Ground-Water Resources—Tennessee Region:
U.S. Geological Survey Professional Paper 813-L, 35 p.
158
-------
APPENDIX 3
-------
Glossary
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 ground-water contamination assessments and
wellhead protection. 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 ground
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
161
-------
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 pumping or recharging well within which the water table or
potenn'ometric surface has been changed due to the well's pumping or recharge.
Attenuation. The process of diminishing contaminant concentrations in ground water, due to filtration,
biodegradation, dilution, sorption, volatilization, and other processes.
Carbon-14 (14C). A radioisotope of carbon with a half life of 5,730 years. Carbon-14 concentration can be
used to estimate the age of a ground water (that is, the time since a ground water was recharged at land
surface and flowed to the point of collection).
Cone of depression (COD). A depression in the ground-water table or potentiometric surface that has
the shape of an inverted cone and develops around a well from which water is being withdrawn. Its
trace (perimeter) on the land surface defines the zone of influence of a well. 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.
Dispersion. The spreading and mixing of chemical constituents in ground water caused by diffusion and
mixing due to microscopic variations in velocities within and between pores.
162
-------
Drawdown. The vertical distance ground-water elevation is lowered, or the amount head is reduced,
due to the removal of ground water. 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 ground water from wells.
Fissure. A fracture or crack in a rock 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 ground water toward points of discharge. Flow lines generally are
considered perpendicular to equipotential lines.
Flow model. A computer model that calculates a hydraulic head field for the study area using
numerical methods to arrive at an approximate solution to the differential equation of ground-water
flow.
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.
Ground-water barrier. Rock or artificial material with a relatively low permeability that occurs (or is
placed) below ground surface, where it impedes the movement of ground water and thus may cause a
pronounced difference in the heads on opposite sides of the barrier.
Ground-water basin. General term used to define a ground-water flow system that has defined
boundaries and may include more than one aquifer. The basin includes both the surface area and the
163
-------
permeable materials beneath it. A rather vague designation pertaining to a ground-water reservoir
that is more or less separate from neighboring ground-water reservoirs. A ground-water basin could be
separated from adjacent basins by geologic boundaries or by hydrologic boundaries.
Ground-water divide. Ridge in the water table, or potentiometric surface, from which ground water
moves away at right angles in both directions. Line of highest hydraulic head in the water table or
potentiometric surface.
Ground-water mound. Raised area in a water table or other potentiometric surface, created by ground-
water recharge.
Head, total. Height of the column of water at a given point in a ground-water 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.
Hydraulic conductivity (K). A coefficient of proportionality describing the rate at which water can
move through a permeable medium.
164
-------
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 - h2), 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 ground water.
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.
165
-------
Leakage. The vertical flow of ground water; commonly used in the context of vertical ground-water flow
through confining strata.
Maximum contaminant level (MCL). Maximum permissible level of a contaminant in water that is
delivered to the users of a public water system. Maximum containment level is defined more explicitly
in Safe Drinking Water Act (SDWA) regulations (40 CFR Section 141.2).
Observation well. A well drilled in a selected location for the purpose of observing parameters such as
water levels or water chemistry changes.
Piezometric surface. See potentiometric surface.
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.
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 an aquifer toward a well.
166
-------
Recharge area. Area in which water reaches the ground-water reservoir by surface infiltration. An
area in which there is a downward component of hydraulic head in the aquifer.
Semiconfined aquifer. A confined aquifer whose confining bed may vertically conduct significant
quantities of water.
Stagnation point A place in a ground-water flow field at which the ground water is not moving.
Time of travel (TOT). The time required for a contaminant to move in the saturated zone from a specific
point to a well.
Tritium (3H). The radioactive isotope of hydrogen with a half-life of 12.3 years. The presence or
absence of tritium in ground water provides a method for estimating when the water was recharged at
land surface.
Unconfined aquifer. An aquifer over which there is no confining strata.
Well field. An area containing two or more wells supplying a public water supply system.
Wellhead. The physical structure, facility, or device at the land surface from or through which ground
water flows or is pumped from subsurface, water-bearing formations.
Wellhead protection area (WHPA). The surface and subsurface area surrounding a water well or well
field, supplying a public water system, through which contaminants are reasonably likely to move
toward and reach such water well or well field.
167
-------
Zone of contribution (ZOO. The area surrounding a pumping well that encompasses all areas and
features that supply ground-water recharge to the well.
Zone of influence (ZOI), The area surrounding a pumping well within which the water table or
potentiometric surfaces have been changed due to ground-water withdrawal.
Zone of transport (ZOT). The area surrounding a pumping well, bounded by an isochrone and/or
isoconcentration contour, through which a contaminant may travel and reach the well.
•£ US GOVERNMENT PRINTING OFFICE !<»•— 5 17- 003' 47013
168
-------
------- |