United States
Environmental Protection
Agency
Technology Transfer
EPA/625/4-85/016
Seminar Publication
Protection of Public
Water Supplies from
Ground-Water Contamination
-------
Technology Transfer EPA/625/4-85/016
Seminar Publication
Protection of Public Water
Supplies from Ground-Water
Contamination
September 1985
Center for Environmental Research
Information
Cincinnati, OH 45268
U.S. Environmental Protection
/Region V, Library
230 South Dearborn Street
Chicago, Illinois 60604
-------
NOTICE
The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under Contract 68-03-3130 to
Dynamac Corporation. It has been subject to the Agency's peer and administrative
review, and it has been approved for publication as an EPA document. Mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
Contents
Page
Chapter 1. Introduction 1
Chapter 2. Basic Ground-Water Hydrology 3
Chapter 3. Classification of Ground-Water Regions 53
Chapter 4. Ground Water-Surf ace Water Relationship 83
Chapter 5. Ground-Water Pollution 107
Chapter 6. Management Alternatives 141
Chapter 7. Controlling Volatile Organic Compounds in Ground
Water Used for Drinking 157
Chapter 8. Inground Treatment, Restoration, and Reclamation 175
-------
Acknowledgments
This seminar publication contains material from technology transfer seminars for
the water supply community. The U.S. Environmental Protection Agency's (EPA)
Center for Environmental Research Information (CERI) and Office of Drinking
Water sponsored 14 seminars, held in the 10 EPA Regions and in 4 states, to pro-
vide utility managers and operators, regulators, and technical specialists with pro-
cedures for preventing contamination and for treating public ground-water supplies
that are contaminated. The seminars were designed and delivered in cooperation
with the Water Engineering Research Laboratory in Cincinnati, Ohio, and the
Robert S. Kerr Environmental Research Laboratory in Ada, Oklahoma; Regional
and State coordinators arranged for presentations on local issues, case histories, and
hydrogeologic considerations in dealing with ground-water contamination.
Though many individuals contributed to presenting the 14 seminars and to
preparing and reviewing this publication, several deserve special recognition. Wayne
A. Pettyjohn, Oklahoma State University, was principal speaker and author/editor
for this project.
Major Authors and Speakers
L. Lawrence Graham, EPA, Office of Drinking Water, Washington, DC
Ralph C. Heath, U.S. Geological Survey, Raleigh, NC
With the author's permission, Chapters 2 and 3 are adapted from U.S. Geological
Survey Water-Supply Papers 2220 and 2242, respectively.
Lloyd Hinkle, Dynamac Corporation, Rockville, MD
O. Thomas Love, EPA, Office of Research and Development, Cincinnati, OH
James F. McNabb, EPA, Office of Research and Development, Ada, OK
Richard J. Miltner, EPA, Office of Research and Development, Cincinnati, OH
Wayne A. Pettyjohn, Oklahoma State University, Stillwater, OK
Seminar Coordinators
John H. Mann, EPA, Region IV, Atlanta, GA
Koge Suto, EPA, Region III, Philadelphia, PA
Terry Deen, EPA, Region VII, Kansas City, KS
Harry Von Huben, EPA, Region V, Chicago, IL
Warren Norris, EPA, Region VI, Dallas, TX
Dean Chaussee, EPA, Region VIII, Denver, CO
Bill Thurston, EPA, Region IX, San Francisco, CA
John H. Mann, EPA, Region IV, Atlanta, GA
Bill Mullen, EPA, Region X, Seattle, WA
Thomas Merski, EPA, Region HI, Philadelphia, PA
T. Jay Ray, EPA, Region VI, Dallas, TX
Bobby Savoie, State of Louisiana, Baton Rouge, LA
Stoyell Robbins, EPA, Region II, New York, NY
Mike Burke, State of New York, Albany, NY
Steve Lathrop, EPA, Region I, Boston, MA
Bill Thurston, EPA, Region IX, San Francisco, CA
Locations
Tallahassee, FL
Philadelphia, PA
Kansas City, MO
Chicago, IL
Dallas, TX
Denver, CO
Oakland, CA
Atlanta, GA
Seattle, WA
Pittsburgh, PA
Baton Rouge, LA
Plainville, NY
Boston, MA
Phoenix, AZ
Contract Supervisors
Lloyd Hinkle, Dynamac Corporation, Rockville, MD
Carol Grove and James E. Smith, Jr., EPA, CERI, Cincinnati, OH
IV
-------
Chapter 1
Introduction
Ground water is a valuable drinking water resource
because of both availability and quality. In most
locations of the country, it is available in quantities
large enough to supply 50 percent of the nation's
population. It also is a reliable resource because it is
not subject to extremes in temperature or quality
fluctuations and because it is buffered from the floods
and droughts that affect surface supplies.
The development of subsurface water supplies has
little effect on land use and can usually be accomplished
at relatively low cost compared to the development of
surface supplies. However, the subsurface environment
is a complex system subject to contamination from a
host of sources. Furthermore, the extremely slow
movement of pollutants through this environment
results in a longer residence time and little diffusion of
many pollutants.
Because of the protracted effects of contamination
and problems of accessibility, the restoration of
ground-water quality is difficult and expensive.
Restoration costs generally exceed the short-term value
of the resource when compared to the costs of
alternatives. For this reason, it is widely agreed that the
most viable approach to ground-water quality
protection is one of prevention rather than cure.
There will always be alternative management
practices for obtaining drinking water from subsurface
supplies. One may treat water withdrawn from the
aquifer before use, protect the aquifer totally from all
possible contaminant sources, or reclaim an aquifer
after it has been comtaminated.
Existing Federal authority to address ground-water
quality problems is embodied in at least eight statutes:
National Environmental Policy Act of 1970
Federal Water Pollution Control Act of 1972
Toxic Substances Control Act of 1976
Resource, Conservation and Recovery Act of 1976
Clean Water Act of 1977
Surface Mining Control and Reclamation Act of
1977
Safe Drinking Water Act of 1979
Comprehensive Environmental Response,
Compensation and Liability Act of 1980
At the Federal level, the Safe Drinking Water Act, in
particular, adds to the protection of ground water. In
addition to establishing minimum drinking water
standards, regardless of the source, it addresses the
protection of ground-water quality and provides for
research, technical assistance, and personnel training.
Protection of ground-water quality at the state level
is gained from source control and from statutes dealing
with surface water, which generally are based on public
health concepts. Land-use regulations also are used to
protect ground-water quality by locating waste sources,
such as lagoons and solid waste facilities, in selected
areas in order to minimize the pollution potential.
Often, construction regulations are used in a secondary
manner to protect ground water.
The most promising management option now
available is to protect the ground-water resource from
contamination. This will require many different "best
management practices," including the development of
protection plans at the local level to control activities
that threaten the resource.
An understanding of the processes that affect the
movement and degradation of contaminants in the
subsurface is essential for effective ground-water quality
management. The state of knowledge concerning these
processes is, in many ways, insufficient to ensure
protection of ground-water quality without excessive
restrictions on other surface and subsurface activities.
It is evident from numerous case studies of
contaminated water systems that local communities
have been challenged by contamination of their public
water supplies, and they have been able to find practical
solutions. However, often the cost has been high in
public anxiety and financial burden to the community.
Adequate prevention planning and emergency
response may be beyond the resources of water utilities.
Therefore, it becomes the responsibility of the local
community to carry out planning and prevention
programs that will assess drinking water needs and
protect present and future water supplies.
While statutory authority exists to regulate most
contamination sources there is a need for additional
controls at the local level. Suitable aquifer protection
controls can be adopted under local planning and
zoning laws which take into account land use, industrial
development, health, housing and agriculture. Different
towns have different water supply needs and, thus,
different ground-water protection needs. No one
-------
approach would successfully protect all aquifers. The
entire community, including local government, is
responsible for balancing the risks, costs and benefits
involved in protectng the ground-water supply.
Community planning is particularly evident in some
states, e.g., Connecticut, which has developed a
nationally recognized ground-water program. The
publication, "Protecting Connecticut's Ground Water,"
(available from Natural Resources Center, Connecticut
Department of Environmental Protection, Hartford,
Connecticut 06106) provides a guide for local officials
on the issue of ground-water quality and possible
mechanisms for local protection.
This publication provides an organized approach to
acquiring the knowledge necessary for effective and
efficient management of ground-water supplies. The
information provided is applicable to all regions of the
United States, taking into account differences in
geographic location, from the humid to the arid, and in
geologic composition, from the porous to the
impermeable.
-------
Chapter 2
Basic Ground-Water Hydrology
introduction
The protection of ground water used for public sup-
ply is dependent, to a large extent, on an adequate
understanding of the fundamentals of ground-water
hydrology. Ground-water movement is neither myste-
rious nor occult but rather follows clearly defined prin-
ciples of physics. The purpose of this chapter is to
describe, in an accurate but simplified manner, (1) the
physical characteristics of aquifers and how they func-
tion as storage media and as conduits, (2) techniques
that are commonly applied in order to better under-
stand fluid flow, and (3) methods that are used to
determine or estimate aquifer coefficients.
Rocks and Water
Most of the rocks near the Earth's surface are com-
posed of both solids and voids, as Figure 1 shows. The
solid part is, of course, much more obvious than the
voids, but without the voids, there would be no water
to supply wells and springs.
Water-bearing rocks consist either of unconsolidated
(soil-like) deposits or consolidated rocks. The Earth's
surface in most places is formed by soil and by uncon-
solidated deposits that range in thickness from a few
centimeters near outcrops of consolidated rocks to more
than 12,000 m beneath the delta of the Mississippi
River. The unconsolidated deposits are underlain every-
Primary Openings
Porous Material
Well-Sorted Sand
Poorly-Sorted Sand
Fractured Rock
Secondary Openings
Fractures in
Granite
Caverns in
Limestone
Figure 1. Composition of Rocks Near the Earth's Surface
Figure 2. Different Kinds of Voids in Rocks
-------
where by consolidated rocks.
Most unconsolidated deposits consist of material
derived from the disintegration of consolidated rocks.
The material consists, in different types of uncon-
solidated deposits, of particles of rocks or minerals
ranging in size from fractions of a millimeter (clay size)
to several meters (boulders). Unconsolidated deposits
important in ground-water hydrology include, in order
of increasing grain size, clay, silt, sand, and gravel. An
important group of unconsolidated deposits also in-
cludes fragments of shells of marine organisms.
Consolidated rocks consist of mineral particles of dif-
ferent sizes and shapes that have been welded by heat
and pressure or by chemical reactions into a solid mass.
Such rocks are commonly referred to in ground-water
reports as bedrock. They include sedimentary rocks that
were originally unconsolidated and igneous rocks
formed from a molten state. Consolidated sedimentary
rocks important in ground-water hydrology include
limestone, dolomite, shale, siltstone, sandstone, and
conglomerate. Igneous rocks include granite and basalt.
Figure 2 shows that there are different kinds of voids
in rocks, and it is sometimes useful to be aware of
them. If the voids were formed at the same time as the
rock, they are referred to as, primary openings. The
pores in sand and gravel and in other unconsolidated
deposits are primary openings. The lava tubes and other
openings in basalt are also primary openings.
If the voids were formed after the rock was formed,
they are referred to as secondary openings. The frac-
tures in granite and in consolidated sedimentary rocks
are secondary openings. Voids in limestone, which are
formed as ground water slowly dissolves the rock, are
an especially important type of secondary opening.
It is useful to introduce the topic of rocks and water
by dealing with unconsolidated deposits on one hand
and with consolidated rocks on the other. It is impor-
tant to note, however, that many sedimentary rocks
that serve as sources of ground water fall between these
extremes in a group of semi-consolidated rocks. These
are rocks in which openings include both pores and
fractures—in other words, both primary and secondary
openings. Many limestones and sandstones that are im-
portant sources of ground water are semiconsolidated.
Underground Water
All water beneath the land surface is referred to as
underground water (or subsurface water). The equiva-
lent term for water on the land surface is surface water.
As Figure 3 shows, underground water occurs in two
different zones. One zone, which occurs immediately
below the land surface in most areas, contains both
water and air and is referred to as the unsaturated zone.
The unsaturated zone is almost invariably underlain by
a zone in which all interconnected openings are full of
water. This zone is referred to as the saturated zone.
Water in the saturated zone is the only underground
water that is available to supply wells and springs and is
the only water to which the name ground water is cor-
rectly applied. Recharge of the saturated zone occurs by
percolation of water from the land surface through the
unsaturated zone. The unsaturated zone is, therefore, of
great importance to ground-water hydrology. This zone
may be divided usefully into three parts: the soil zone,
the intermediate zone, and the upper part of the
capillary fringe.
Surface
Water
'/; Well
Rivers and Lakes
Water
Level
Figure 3. Underground Water Zones
4
-------
The soil zone extends from the land surface to a
maximum depth of a meter or two and is the zone that
supports plant growth. It is crisscrossed by living roots,
by voids left by decayed roots of earlier vegetation, and
by animal and worm burrows. The porosity and perme-
ability of this zone tend to be higher than those of the
underlying material. The soil zone is underlain by the
intermediate zone, which differs in thickness from place
to place depending on the thickness of the soil zone and
the depth to the capillary fringe.
The lowest part of the unsaturated zone is occupied
by the capillary fringe, the subzone between the un-
saturated and saturated zones. The capillary fringe
results from the attraction between water and rocks. As
a result of this attraction, water clings as a film on the
surface of rock particles and rises in small-diameter
pores against the pull of gravity. Water in the capillary
fringe and in the overlying part of the unsaturated zone
is under a negative hydraulic pressure—that is, it is
under a pressure less than the atmospheric (barometric)
pressure. The water table is the level in the saturated
zone at which the hydraulic pressure is equal to atmos-
pheric pressure and is represented by the water level in
unused wells. Below the water table, the hydraulic pres-
sure increases with increasing depth.
Hydrologic Cycle
The term hydrologic cycle refers to the constant
movement of water above, on, and below the Earth's
surface. Figure 4 illustrates this cycle. The concept of
the hydrologic cycle is central to an understanding of
the occurrence of water and the development and man-
agement of water supplies.
Although the hydrologic cycle has neither a beginning
nor an end, it is convenient to discuss its principal
features by starting with evaporation from vegetation,
from exposed moist surfaces including the land surface,
and from the ocean. This moisture forms clouds, which
return the water to the land surface or oceans in the
form of precipitation.
Precipitation occurs in several forms, including rain,
snow, and hail, but only rain is considered in this dis-
cussion. The first rain wets vegetation and other sur-
faces and then begins to infiltrate into the ground. Infil-
tration rates vary widely, depending on land use, the
character and moisture content of the soil, and the
intensity and duration of precipitation, from possibly as
much as 25 mm/hr in mature forests on sandy soils to a
few millimeters per hour in clayey and silty soils to zero
in paved areas. When and if the rate of precipitation
exceeds the rate of infiltration, overland flow occurs.
The first infiltration replaces soil moisture, and
thereafter, the excess percolates slowly across the inter-
mediate zone to the zone of saturation. Water in the
zone of saturation moves downward and laterally to
sites of ground-water discharge such as springs on hill-
sides or seeps in the bottoms of streams and lakes or
beneath the ocean.
Water reaching streams, both by overland flow and
from ground-water discharge, moves to the sea, where
it is again evaporated to perpetuate the cycle.
r > ~^_^
Clouds forming
/v;
i*l<$£ SQJtwettr
Figure 4. The Hydrologic Cycle
-------
Movement is, of course, the key element in the con-
cept of the hydrologic cycle. Table 1 shows some
"typical" rates of movement along with the distribution
of the Earth's water supply.
Table 1. Rate of Movement and Distribution of Water
[Adapted from L'vovich (1979), table 1]
Distribution of
Rate of Earth's Water
Location Movement Supply (percent)
Atmosphere
Water on land
surface
Water below the
land surface .
Ice caps and
glaciers
Oceans
100's of kilometers per day
10's of kilometers per day
Meters per year
Meters per day
0001
019
4.12
1 65
9396
Aquifers and Confining Beds
From the standpoint of ground-water occurrence, all
rocks that underlie the Earth's surface can be classified
either as aquifers or as confining beds. Figure 5 illus-
trates these formations. An aquifer is a rock unit that
will yield water in a usable quantity to a well or spring.
(In geologic usage, "rock" includes unconsolidated
sediments.) A confining bed is a rock unit having very
low hydraulic conductivity that restricts the movement
of ground water either into or out of adjacent aquifers.
Ground water occurs in aquifers under two different
conditions. Where water only partly fills an aquifer, the
upper surface of the saturated zone is free to rise and
decline. The water in such aquifers is said to be uncon-
fined, and the aquifers are referred to as unconfined
aquifers. Unconfined aquifers are also widely referred
to as water-table aquifers.
Where water completely fills an aquifer that is over-
lain by a confining bed, the water in the aquifer is said
to be confined. Such aquifers are referred to as con-
fined aquifers or as artesian aquifers.
Wells open to unconfined aquifers are referred to as
water-table wells. The water level in these wells indicates
the position of the water table in the surrounding
aquifer.
Wells drilled into confined aquifers are referred to as
artesian wells. The water level in artesian wells stands at
some height above the top of the aquifer but not neces-
sarily above the land surface. If the water level in an
artesian well stands above the land surface, the well is a
flowing artesian well. The water level in tightly cased
wells open to a confined aquifer stands at the level of
the potentiometric surface of the aquifer.
Porosity
The ratio of openings (voids) to the total volume of a
soil or rock is referred to as its porosity. Porosity is ex-
Water-table
Well
Artesian
Well
Land
Surface
Unsaturated
Zone
Saturated Zone
Unconfined
Aquifer
Confining
Bed
Confined
Aquifer
. .
• •
Sand ••'•'.
\\i\\\* \UHUUMU<
**$t*r'$«MMi-.-V
•-•cW V:-d
~_~_~_~_~_~ _~ Clay ~_~__~_~~_~_~
_ ___ ____ __ _
I I
I II,
I I I I I I I I
I I
II!'
I I I L
III,
I I I
II I I -,
!_ Open „ j
imestone I Hole 1
1 . 1
I I J I I I I I
I I
Figure 5. Aquifers and Confining Bed
6
-------
=0.3 m3
Vt -- l.O
m
— Water — -
h
o o
OOOOoOOOO
0 o o _° e o o o o
oo Dry oo
Oo° , ooo
o o sona oo
OoooooOooo
oooooooo o
0 OOOOoOoOo
OOOpOooOQ
°-« o
n ° °
Ooo
, ° o
• • a
<>'•'.$
O O O O o o <3 OO4
o o b'O o o O'<3 o
o • •* a
o Saturated «
^~sand
oo o'o o
ooo "tf o d o
o o o o o o o o o
o o <> "o o o o
Porosity (/7) =
Figure 6. Illustration of Porosity
Volume of voids ( Vv) 0.3 m3
Total volume (Vt) 1.0 m3
= 0.30
pressed either as a decimal fraction or as a percentage.
Thus,
n -
(i)
where n is porosity as a decimal fraction, Vt is the total
volume of a soil or rock sample, Vs is the volume of
solids in the sample, and Vv is the volume of openings
(voids). Figure 6 illustrates the calculation of porosity.
If we multiply the porosity determined with the equa-
tion by 100, the result is porosity expressed as a per-
centage.
Table 2 lists porosity values for some selected
materials. Soils are among the most porous of natural
materials because soil particles tend to form loose
clumps and because of the presence of root holes and
animal burrows. Porosity of unconsolidated deposits
depends on the range in grain size (sorting) and on the
shape of the rock particles but not on their size. Fine-
grained materials tend to be better sorted and, thus,
tend to have the largest porosities.
Table 2. Selected Values of Porosity
[Values in percent by volume]
Primary Secondary
Material Openings Openings
Equal-size spheres (marbles):
Loosest packing
Tightest packing
Soil
Clay
Sand
Gravel
Limestone
Sandstone (semiconsolidated)
Granite
Basalt (young)
48
26
55
50
25
20
10
10
10
10
1
1
1
Specific Yield and Specific Retention
Porosity is important in ground-water hydrology
because it tells us the maximum amount of water that a
rock can contain when it is saturated. However, it is
equally important to know that only a part of this
water is available to supply a well or a spring. Figure 1
gives some graphical examples of porosity.
Water
Granular Material
Water retained a
o film on rock
surfaces and in
capillary- size
openings offer
gravity drainage
Water
Fractured Rock
Figure 7. Graphical Example of Porosity
-------
Hydrologists divide water in storage in the ground in-
to the part that will drain under the influence of gravity
(called specific yield) and the part that is retained as a
film on rock surfaces and in very small openings (called
specific retention). The physical forces that control
specific retention are the same forces involved in the
thickness and moisture content of the capillary fringe.
Specific yield tells how much water is available for
man's use, and specific retention tells how much water
remains in the rock after it is drained by gravity. As
Sr-Q.\
rr
3 "c
Sy -0.2 m3
n -
,, 0.2 m5 O.I m3 n n
Sr r- = 0.30
I
I m:
Figure 8. Porosity is the Sum of Specific Yield and
Specific Retention
Figure 8 shows, porosity is the sum of specific yield and
specific retention. Thus,
n =
Sr
(2)
(3), (4)
where n is porosity, Sv is specific yield, S, is specific
retention, Vd is the volume of water that drains from a
total volume of Vt, Vr is the volume of water retained in
a total volume of Vt, and Vt is total volume of a soil or
rock sample. Table 3 lists values of porosity, specific
yield, and specific retention for selected materials.
Table 3. Selected Values of Porosity, Specific Yield,
and Specific Retention
[Values in percent by volume]
Specific Specific
Material Porosity Yield Retention
Soil
Clay
Sand ...
Gravel
Limestone .
Sandstone (semiconsolidated)
Basalt (young) ...
55
50
25
20
20
11
.1
11
40
2
22
19
18
6
.09
8
15
48
3
1
2
5
.01
3
Heads and Gradients
The depth to the water table has an important effect
on use of the land surface and on the development of
water supplies from unconfined aquifers. Where the
water table is at a shallow depth, the land may become
"waterlogged" during wet weather and unsuitable for
residential and many other uses. Where the water table
is at great depth, the cost of constructing wells and
pumping water for domestic needs may be prohibitively
expensive.
The direction of the slope of the water table is also
important because it indicates the direction of ground-
water movement. The position and the slope of the
water table (or of the potentiometric surface of a con-
fined aquifer) is determined by measuring the position
of the water level in wells from a fixed point (a measur-
ing point). To utilize these measurements to determine
the slope of the water table, the position of the water
table at each well must be determined relative to a
datum plane that is common to all the wells. The datum
plane most widely used is the National Geodetic Vertical
Datum of 1929 (also commonly referred to as "sea
level").
If the depth to water in a nonflowing well is sub-
tracted from the altitude of the measuring point, the
result is the total head at the well. Total head, as de-
fined in fluid mechanics, is composed of elevation head,
pressure head, and velocity head. Because ground water
moves relatively slowly, velocity head can be ignored.
Therefore, the total head at an observation well involves
only two components: elevation head and pressure
head. Ground water moves in the direction of decreas-
ing total head, which may or may not be in the direc-
tion of decreasing pressure head.
The equation for total head (ht) is
= z + h
(5)
where z is elevation head and is the distance from the
datum plane to the point where the pressure head hp is
determined.
All other factors being constant, the rate of ground-
water movement depends on the hydraulic gradient. The
hydraulic gradient is the change in head per unit of
distance in a given direction. If the direction is not
specified, it is understood to be in the direction in
which the maximum rate of decrease in head occurs.
If the movement of ground water is assumed to be in
the plane of Figure 9—in other words, if it moves from
well 1 to well 2—the hydraulic gradient can be calcu-
lated from the information given on the drawing. The
hydraulic gradient is hjL, where hL is the head loss
between wells 1 and 2 and L is the horizontal distance
between them, or
h^ _ (100m-15m)-(98m-18m) _ 85m-80m _ 5m
L 780m 780m 780m
When the hydraulic gradient is expressed in consistent
units, as it is in the above example in which both the
numerator and the denominator are in meters, any
-------
Well I
Measuring point ( top of casing )
.(Alt 100m) (Alt 98m ) Wel| 2
Distance, L, 780m
Land surface
4_i__£
(National Geodetic
Vertical Datum of 1929!
Figure 9. Gradient is Determined By the Difference in Head Between Two Wells.
other consistent units of length can be substituted with-
out changing the value of the gradient. Thus, a gradient
of 5 ft/780 ft is the same as a gradient of 5m/780 m. It
is also relatively common to express hydraulic gradients
in inconsistent units such as meters per kilometer or feet
per mile. A gradient of 5 m/780 m can be converted to
meters per kilometer as follows:
5m
780m
/1,000m\
km
= 6.4 m km
-1
Both the direction of ground-water movement and
the hydraulic gradient can be determined if the follow-
ing data are available for three wells located in any tri-
angular arrangement such as that shown in Figure 10:
1. The relative geographic position of the wells.
2. The distance between the wells.
3. The total head at each well.
Figure 11 illustrates the following steps in the solution.
well 2
(heod, 26.20m)
0 25 50
i i i
W«ll I
( heod, 26 26m)
Meters
well 3
(heod, 26 07 m )
(b) (26.26-26.20) (26.26-26.07)
26.26m
(o) Well 2
W L =26.20 m
(e) 26 2 - 26.07
Direction of
ground-water
movement
26.07ni
Figure 10. Triangular Arrangement of Wells
Figure 11. Determining the Direction of Ground-Water Move-
ment and the Hydraulic Gradient for a Triangular Arrangement
of Wells
-------
a. Identify the well that has the intermediate water
level (that is, neither the highest head nor the
lowest head).
b. Calculate the position between the well having the
highest head and the well having the lowest head
at which the head is the same as that in the inter-
mediate well.
c. Draw a straight line between the intermediate well
and the point identified in step b as being between
the well having the highest head and that having
the lowest head. This line represents a segment of
the water-level contour along which the total head
is the same as that in the intermediate well.
d. Draw a line perpendicular to the water-level con-
tour and through either the well with the highest
head or the well with the lowest head. This line
parallels the direction of ground-water movement.
e. Divide the difference between the head of the well
and that of the contour by the distance between
the well and the contour. The answer is the
hydraulic gradient.
Hydraulic Conductivity
Aquifers transmit water from recharge areas to
discharge areas and thus function as porous conduits
(or pipelines filled with sand or other water-bearing
material). The factors controlling ground-water move-
ment were first expressed in the form of an equation by
Henry Darcy, a French engineer, in 1856. Darcy's law is
where Q is the quantity of water per unit of time; K is
the hydraulic conductivity and depends on the size and
arrangement of the water-transmitting openings (pores
and fractures) and on the dynamic characteristics of the
fluid (water) such as kinematic viscosity, density, and
the strength of the gravitational field; A is the cross-
sectional area, at a right angle to the flow direction,
through which the flow occurs; and dh/dl is the
hydraulic gradient.1
Because the quantity of water (Q) is directly propor-
tional to the hydraulic gradient (dh/dl), we say that
ground-water flow is laminar—that is, water particles
tend to follow discrete streamlines and not to mix with
particles in adjacent streamlines. (See the "Ground-
Water Flow Nets" section of this chapter.)
If we rearrange equation 6 to solve for K, we obtain
Qdl = (m3d-1)(m) _ nn
Adh (m2)(m) ~ d
(7)
Thus, the units of hydraulic conductivity are those of
velocity (or distance divided by time). It is important to
note from equation 7, however, that the factors in-
volved in the definition of hydraulic conductivity in-
clude the volume of water (Q) that will move in a unit
'Where hydraulic gradient is discussed as an independent entity, as
it is in "Heads and Gradients," it is shown symbolically as hL/L and is
referred to as head loss per unit of distance. Where hydraulic gradient
appears as one of the factors in an equation, as it does in equation 6,
it is shown symbolically as dh/dl to be consistent with other ground-
water literature. The gradient dh/dl indicates that the unit distance is
reduced to as small a value as one can imagine, in accordance with the
concepts of differential calculus.
Unit element
of aquifer
Streamlines
representing
laminar flow
_f — Confining Bed ~^
'*-' "V * ,-s Confined Aquifer ' ;*&r*^""'*'***,-
-. .--",<:' iA-—,*.,- """''
Confining Bed
Height = I m
Unit prism of aquifer
Figure 12. Factors Affecting Hydraulic Conductivity
10
-------
of time (commonly, a day) under a unit hydraulic
gradient (such as a meter per meter) through a unit area
(such as a square meter). Figure 12 illutrates these fac-
tors. Expressing hydraulic conductivity in terms of a
unit gradient, rather than of an actual gradient at some
place in an aquifer, permits ready comparison of values
of hydraulic conductivity for different rocks.
Hydraulic conductivity replaces the term "field coef-
ficient of permeability" and should be used in referring
to the water-transmitting characteristic of material in
quantitative terms. It is still common practice to refer in
qualitative terms to "permeable" and "impermeable"
material.
Figure 13 shows that the hydraulic conductivity of
rocks ranges through 12 orders of magnitude. There are
few physical parameters whose values range so widely.
Hydraulic conductivity is not only different in different
types of rocks but may also be different from place to
place in the same rock. If the hydraulic conductivity is
essentially the same in any area, the aquifer in that area
is said to be homogeneous. If, on the other hand, the
hydraulic conductivity differs from one part of the area
to another, the aquifer is said to be heterogeneous.
Hydraulic conductivity may also be different in dif-
ferent directions at any place in an aquifer. If the
hydraulic conductivity is essentially the same in all
Igneous and Metamorphic Rocks
Unfroot u red
Fractured
Basalt
Unfroctured
Froctured
Sandstone
Lovo flow
Fractured
Semiconsolidoted
Shale
Unfract ured
Fractured
Carbonate Rocks
Fractured
Cavernous
Clay
Silt, Loess
Silty Sand
Clean Sand
Fi ne
Coarse
Glacial Till
Gravel
I L
IO"8 ICT7 IO"6 I0~5 10"" I0~3 I0~2 10"' I 10 10 z IO3 I04
m
J L
_L
ICTT IO"6 IO"5 IO"4 IO"3 I0~2 10"'
ftd-'
I i I i
10 10 2 iO 3 10 4 10 5
J L
IO"7 IO"6 IO"5 IO"4 IO"3 IO"2 IO"1
10 10 2 10 3 10 4 10 5
gal d-' ft
Figure 13. Hydraulic Conductivity of Selected Rocks
11
-------
directions, the aquifer is said to be isotropic. If it is dif-
ferent in different directions, the aquifer is said to be
anisotropic.
Although it is convenient in many mathematical
analyses of ground-water flow to assume that aquifers
are both homogeneous and isotropic, such aquifers are
rare, if they exist at all. The condition most commonly
encountered is for hydraulic conductivity in most rocks
and especially in unconsolidated deposits and in flat-
lying consolidated sedimentary rocks to be larger in the
horizontal direction than it is in the vertical direction.
Functions of Ground-Water Systems
As Figure 14 shows, the aquifers and confining beds
that underlie any area comprise the ground-water sys-
tem of the area. Hydraulically, this system serves two
functions: it stores water to the extent of its porosity,
and it transmits water from recharge areas to discharge
areas. Thus, a ground-water system serves as both a
reservoir and a conduit. With the exception of cavern-
ous limestones, lava flows, and coarse gravels, ground-
water systems are more effective as reservoirs than as
conduits.
Water enters ground-water systems in recharge areas
and moves through them, as dictated by hydraulic
gradients and hydraulic conductivities, to discharge
areas.
The identification of recharge areas is becoming in-
creasingly important because of the expanding use of
the land surface for waste disposal. In the humid part
of the country, recharge occurs in all interstream areas
—that is, in all areas except along streams and their ad-
joining flood plains. The streams and flood plains are,
under most conditions, discharge areas.
In the drier part (western half) of the conterminous
United States, recharge conditions are more complex.
Most recharge occurs in the mountain ranges, on
alluvial fans that border the mountain ranges, and
along the channels of major streams where they are
underlain by thick and permeable alluvial deposits.
Recharge rates are generally expressed in terms of
volume (such as cubic meters or gallons) per unit of
time (such as a day or a year) per unit of area (such as
a square kilometer, a square mile, or an acre). When
these units are reduced to their simplest forms, the
result is recharge expressed as a depth of water on the
land surface per unit of time. Recharge varies from year
to year, depending on the amount of precipitation, its
seasonal distribution, air temperature, land use, and
other factors. Relative to land use, recharge rates in
forests are much higher than those in cities.
Annual recharge rates range, in different parts of the
country, from essentially zero in desert areas to about
eOOmmyr-1 (1,600 nAm-M-1 or l.lxK^gal
mi~2d~') in the rural areas on Long Island and in other
rural areas in the East that are underlain by very
permeable soils.
The rate of movement of ground water from recharge
areas to discharge areas depends on the hydraulic con-
ductivities of the aquifers and confining beds, if water
moves downward into other aquifers, and on the
hydraulic gradients. (See the "Ground-Water Velocity"
section of this chapter.) A convenient way of showing
the rate is in terms of the time required for ground-
water to move from different parts of a recharge area
to the nearest discharge area. The time ranges from a
few days in the zone adjacent to the discharge area to
thousands of years (millennia) for water that moves
from the central part of some recharge areas through
the deeper parts of the ground-water system.
Natural discharge from ground-water systems in-
cludes not only the flow of springs and the seepage of
water into stream channels or wetlands but also evap-
oration from the upper part of the capillary fringe,
where it occurs within a meter or so of the land surface.
Large amounts of water are also withdrawn from the
capillary fringe and the zone of saturation by plants
Discharge
a rea
.-^^d"^^
'« "* ."^-.Jr?*^"**. f&'&'ktf VA?* --V V*., .. ' &' K %£ ,5 \.*V. •?•***>»:•. tt .^ *A< 'A~---^'s' -, ** -?-<>^^A.
Figure 14. The Ground-Water System
12
-------
FluctJOtion of the Wafer Table in the Coastal Plain of Norlh Carolina
~i—r
MftR APR MAY JUNE j JULY
Figure 15. Fluctuation of the Water Level Showing Recharge and Discharge in Observation Wells.
during the growing season. Thus, discharge areas in-
clude not only the channels of perennial streams but
also the adjoining flood plains and other low-lying
areas.
One of the most significant differences between re-
charge areas and discharge areas is that the areal extent
of discharge areas is invariably much smaller than that
of recharge areas. This size difference shows, as we
would expect, that discharge areas are more "efficient"
than recharge areas. Recharge involves unsaturated
movement of water in the vertical direction; in other
words, movement is in the direction in which the
hydraulic conductivity is generally the lowest.
Discharge, on the other hand, involves saturated move-
ment, much of it in the horizontal direction—that is, in
the direction of the largest hydraulic conductivity.
Another important aspect of recharge and discharge
involves timing. Recharge occurs during and immedi-
ately following periods of precipitation and thus is inter-
mittent. Discharge, on the other hand, is a continuous
process as long as ground-water heads are above the
level at which discharge occurs. However, between
periods of recharge, ground-water heads decline, and
the rate of discharge also declines. Most recharge of
ground-water systems occurs during late fall, winter,
and early spring, when plants are dormant and evapora-
tion rates are small. These aspects of recharge and
discharge are apparent from graphs showing the fluctua-
tion of the water level in observation wells, such as the
one shown in Figure 15. The occasional lack of correla-
tion, especially in the summer, between the precipitation
and the rise in water level is due partly to the distance
of 20 km between the weather station and the well.
Capillarity and Unsaturated Flow
Most recharge of ground-water systems occurs during
the percolation of water across the unsaturated zone.
The movement of water in the unsaturated zone is con-
trolled by both gravitational and capillary forces.
Capillarity results from two forces: the mutual attrac-
tion (cohesion) between water molecules and the
molecular attraction (adhesion) between water and dif-
ferent solid materials. Figure 16 shows that a conse-
quence of these forces, water will rise in small-diameter
glass tubes to a height hc above the water level in a
large container.
Most pores in granular materials are of capillary size,
and as a result, water is pulled upward into a capillary
fringe above the water table in the same manner that
water would be pulled up into a column of sand whose
lower end is immersed in water, as Figure 17 shows.
Table 4 shows the approximate capillary rise in selected
granular materials.
Table 4. Approximate Height of Capillary Rise (hc)
in Granular Materials
Material Rise (mm)
Sand-
Coarse ..
Medium
Fine ...
Silt
. 125
. 250
. 400
.1,000
Steady-state flow of water in the unsaturated zone
can be determined from a modified form of Darcy's
law. Steady state in this context refers to a condition in
which the moisture content remains constant, as it
would, for example, beneath a waste-disposal pond
whose bottom is separated from the water table by an
unsaturated zone.
Steady-state unsaturated flow (Q) is proportional to
the effective hydraulic conductivity (Ke), the cross-
sectional area (A) through which the flow occurs, and
the gradients due to both capillary forces and gravita-
tional forces. Thus,
Q = KeA
(8)
where Q is the quantity of water, Ke is the hydraulic
conductivity under the degree of saturation existing in
13
-------
Capillary- sire
glass tube
Sand
column
Wetting
front
<
OQ CO
OOOO
OOOO
&°
2 26
o 24
i
uj 22
20
Lond
surface
^
2 = 3
1
\
1 N
2m
m
—
^
sss
5
—
S^XXXN
hp~2 m
?- —
X
s
>
: , oo
V
4 A
0 L-
, = 31 m
ht = 26 m
Capillary
f
I
Water table.
^r
Datum Plane (National Geodetic Vertical Datum 1929)
Figure 18. Tensiometer Measurements for Measuring the
Capillary Gradient
14
-------
The effective hydraulic conductivity (Ke) is the
hydraulic conductivity of material that is not completely
saturated. It is thus less than the (saturated) hydraulic
conductivity (Ks) for the material. Figure 19 shows the
relation between degree of saturation and the ratio of
saturated and unsaturated hydraulic conductivity for
coarse sand. The hydraulic conductivity (Ks) of coarse
sand is about 60 m d~ '•
o
•a
o
O
O
1.0
0.8
0.6
2: 04
o
o
•*—'
co
OC
0.2
Coorse sand
20 40 60 80
Saturation, in Percent
100
Figure 19. Relation Between Degree of Saturation and the
Ratio of Saturated and Unsaturated Hydraulic Conductivity
for Coarse Sand
stratified deposit, and Figure 21 shows another model
consisting of five layers, three of which were finer
grained and more impermeable than the other two. The
dimensions of the models were about 1.5mxl.2mx
76 mm.
In the nonstratified model, water introduced at the
top moved vertically downward through a zone of con-
Nonstratrfied
model
Figure 20. Single-Size Bead Model Illustrating Water Move-
ment Across the Unsaturated Zone.
Stratification and Unsaturated Flow
Most sediments are deposited in layers (beds) that
have a distinct grain size, sorting, or mineral composi-
tion. Where adjacent layers differ in one of these
characteristics or more, the deposit is said to be
stratified, and its layered structure is referred to as
stratification.
The layers comprising a stratified deposit commonly
differ from one another in both grain size and sorting
and, consequently, differ from one another in hydraulic
conductivity. These differences in hydraulic conductivity
significantly affect both the percolation of water across
the unsaturated zone and the movement of ground-
water.
In most areas, the unsaturated zone is composed of
horizontal or nearly horizontal layers. The movement of
water, on the other hand, is predominantly in a vertical
direction. In many ground-water problems, and espe-
cially in those related to the release of pollutants at the
land surface, the effect of stratification on movement of
fluids across the unsaturated zone is of great impor-
tance.
The manner in which water moves across the un-
saturated zone has been studied by using models con-
taining glass beads. Figure 20 illustrates one model con-
taining beads of a single size representing a non-
Stratified Inflow 0 023 m3 d '
model ,| (6gald ')
\^ r^*::
/***> '•:'•.. ' . . .,1
Explanation
Areas remaining dry after
38 hours of inflow
Figure 21. Five-Layer Model Illustrating Water Movement
Across the Unsaturated Zone.
15
-------
slant width to the bottom of the model. In the stratified
model, beds A, C, and £ consisted of silt-sized beads
(diameters of 0.036 mm) having a capillary height (hc)
of about 1,000 mm and a hydraulic conductivity (K) of
0.8 m d~'. Beds 8 and D consisted of medium-sand-
sized beads (diameters of 0.47 mm) having a capillary
height of about 250 mm and a hydraulic conductivity of
82 md-1.
Because of the strong capillary force and the low
hydraulic conductivity in bed A, the water spread
laterally at almost the same rate as it did vertically, and
it did not begin to enter bed 6 until 9 hours after the
start of the experiment. At that time, the capillary
saturation in bed A had reached a level where the un-
satisfied (remaining) capillary pull in bed A was the
same as that in bed 8. In other words, z in bed A at
that time equaled 1,000 mm - 250 mm, or 750 mm.
(For a definition of z, see the "Capillarity and Un-
saturated Flow" section of this Chapter.)
Because the hydraulic conductivity of bed 6 was 100
times that of bed A, water moved across bed 8 through
narrow vertical zones. We can guess that the glass beads
in these zones were packed somewhat more tightly than
those in other parts of the beds.
Saturated Flow and Dispersion
In the saturated zone, all interconnected openings are
full of water, and the water moves through these open-
ings in the direction controlled by the hydraulic gradi-
ent. Movement in the saturated zone may be either
laminar or turbulent. In laminar flow, water particles
move in an orderly manner along streamlines. In tur-
bulent flow, water particles move in a disordered,
highly irregular manner, which results in a complex
mixing of the particles. Under natural hydraulic gradi-
ents, turbulent flow occurs only in large openings such
as those in gravel, lava flows, and limestone caverns.
Flows are laminar in most granular deposits and frac-
tured rocks.
In laminar flow in a granular medium, the different
streamlines converge in the narrow necks between par-
ticles and diverge in the larger interstices, as Figure 22
illustrates. Thus, there is some intermingling of stream-
lines, which results in transverse dispersion—that is,
dispersion at right angles to the direction of ground-
water flow. Also, differences in velocity result from
friction between the water and the rock particles. The
slowest rate of movement occurs adjacent to the par-
ticles, and the fastest rate occurs in the center of pores.
The resulting dispersion is longitudinal—that is, in the
direction of flow.
As Figure 23 shows, Danel (1953) found that dye in-
jected at a point in a homogeneous and isotropic
granular medium dispersed laterally in the shape of a
cone about 6° wide. He also found that the concentra-
tion of dye over a plane at any given distance from the
inlet point is a bell-shaped curve similar to the normal
probability curve. Because of transverse and longi-
tudinal dispersion, the peak concentration decreased in
the direction of flow.
Cone of dispersion
Figure 23. Changes in Concentration in the Dispersion Cone
The effect of longitudinal dispersion can also be ob-
served from the change in concentration of a substance
(C) downstream from a point at which the substance is
being injected constantly at a concentration of C0. As
Figure 24 illustrates, the concentration rises slowly at
first as the "fastest" streamlines arrive and then rises
rapidly until the concentration reaches about 0.7 C0, at
which point the rate of increase in concentration begins
to decrease.
First oppeorance
of substance
Figure 22. Dispersion in a Granular Deposit
16
Time since start
of injection *"
Figure 24. Effect of Longitudinal Dispersion
Dispersion is important in the study of ground-water
pollution. However, it is difficult to measure in the field
because the rate and direction of movement of wastes
are also affected by stratification, ion exchange, filtra-
tion, and other conditions and processes. Stratification
and areal differences in lithology and other character-
istics of aquifers and confining beds actually result in
much greater lateral and longitudinal dispersion than
that measured by Danel for a homogeneous and iso-
tropic medium.
Ground-Water Movement and Topography
It is desirable, wherever possible, to determine the
position of the water table and the direction of ground-
water movement. To do so, it is necessary to determine
-------
the altitude, or the height above a datum plane, of the
water level in wells. However, in most areas, general
but very valuable conclusions about the direction of
ground-water movement can be derived from observa-
tions of land-surface topography, as Figure 25 illus-
trates.
Arrows show direction of
ground-water movement
Figure 25. Shallow Ground-Water Movement Generally
Conforms to the Surface Topography.
Gravity is the dominant driving force in ground-water
movement. Under natural conditions, ground water
moves "downhill" until, in the course of its movement,
it reaches the land surface at a spring or through a seep
along the side or bottom of a stream channel or an
estuary.
Thus, ground water in the shallowest part of the
saturated zone moves from interstream areas toward
streams or the coast. If we ignore minor surface irregu-
larities, we find that the slope of the land surface is also
toward streams or the coast. The depth to the water
table is greater along the divide between streams than it
is beneath the flood plain. In effect, the water table
usually is a subdued replica of the land surface.
In areas where ground water is used for domestic and
other needs requiring good-quality water, septic tanks,
sanitary landfills, waste ponds, and other waste-disposal
sites should not be located uphill from supply wells.
The potentiometric surface of confined aquifers, like
the water table, also slopes from recharge areas to dis-
charge areas. Shallow confined aquifers, which are
relatively common along the Atlantic Coastal Plain,
share both recharge and discharge areas with the sur-
ficial unconfined aquifers. This sharing may not be the
case with the deeper confined aquifers. The principal
recharge areas for these are probably in their outcrop
areas near the western border of the Coastal Plain, and
their discharge areas are probably near the heads of the
estuaries along the major streams. Thus, movement of
water through these aquifers is in a general west to east
direction, where it has not been modified by with-
drawals.
In the western part of the conterminous United
States, and especially in the alluvial basins region, con-
ditions are more variable than those described above. In
this area, streams flowing from mountain ranges onto
alluvial plains lose water to the alluvial deposits; thus,
ground water in the upper part of the saturated zone
flows down the valleys and at an angle away from the
streams.
Ground water is normally hidden from view; as a
consequence, many people have difficulty visualizing its
occurrence and movement. This difficulty adversely af-
fects their ability to understand and to deal effectively
with ground-water-related problems. This problem can
be partly solved through the use of flow nets, which are
one of the most effective means yet devised for illus-
trating conditions in ground-water systems.
Ground-Water Flow Nets
Flow nets consist of two sets of lines. One set, re-
ferred to as equipotential lines, connects points of equal
head and thus represents the height of the water table,
or the potentiometric surface of a confined aquifer,
above a datum plane. The second set, referred to as
flow lines, depicts the idealized paths followed by par-
ticles of water as they move through the aquifer.
Because ground water moves in the direction of the
steepest hydraulic gradient, flow lines in isotropic
aquifers are perpendicular to equipotential lines—that
is, flow lines cross equipotential lines at right angles.
There are an infinite number of equipotential lines
and flow lines in an aquifer. However, for purposes of
flow-net analysis, only a few of each set need be drawn.
Equipotential lines are drawn so that the drop in head
17
-------
is the same between adjacent pairs of lines. Flow lines
are drawn so that the flow is equally divided between
adjacent pairs of lines and so that, together with the
equipotential lines, they form a series of "squares."
Flow nets not only show the direction of ground-
water movement but can also, if they are drawn with
care, be used to estimate the quantity of water in transit
through an aquifer. According to Darcy's law, the flow
through any "square" is
and the total flow through any set or group of
"squares" is
0 = nq
(10)
where K is hydraulic conductivity, b is aquifer thickness
at the midpoint between equipotential lines, w is the
distance between flow lines, dh is the difference in head
between equipotential lines, dl is the distance between
equipotential lines, and n is the number of squares
through which the flow occurs.
Figures 26 and 27 show a flow net in both plan view
and cross section for an area underlain by an uncon-
fmed aquifer composed of sand. The sand overlies a
horizontal confining bed, the top of which occurs at an
elevation 3 m above the datum plane. The fact that
some flow lines originate in the area in which heads ex-
ceed 13m indicates the presence of recharge to the
aquifer in this area. The relative positions of the land
surface and the water table in Figure 27 suggest that
recharge occurs throughout the area, except along the
stream valleys. This suggestion is confirmed by the fact
that flow lines also originate in areas where heads are
less than 13 m.
As Figures 26 and 27 show, flow lines originate in
recharge areas and terminate in discharge areas. Closed
contours (equipotential lines) indicate the central parts
of recharge areas but do not normally indicate the limits
of the areas.
In the cross-sectional view in Figure 27, heads
decrease downward in the recharge area and decrease
upward in the discharge area. Consequently, the deeper
a well is drilled in a recharge area, the lower the water
level in the well stands below land surface. The reverse
is true in discharge areas. Thus, in a discharge area, if a
well is drilled deeply enough in an unconfmed aquifer,
the well may flow above land surface. Consequently, a
flowing well does not necessarily indicate artesian condi-
tions.
Figures 28 and 29 show equipotential lines and flow
lines in the vicinity of a stream that gains water in its
headwaters and loses water as it flows downstream. In
the gaining reaches, the equipotential lines form a V
pointing upstream; in the losing reach, they form a V
pointing downstream.
Figure 26. Plan View of the Flow Net for an Area Underlain by an Unconfined Aquifer Composed of Sand.
18
-------
Land surface
0
II 1 I I I
Horizontal scale
2000
Figure 27. Cross Section of the Flow Net in Figure 26.
Gaming
stream
Plan view
Gaming
/stream
Figure 28. Plan View of the Flow Net for the Vicinity of a
Stream that Gains Water in its Headwaters and Loses Water
as it Flows Downstream.
4000 Meters
I
^
Horizontal scale
1000 2000
__l I
3000 Meters
I
Figure 29. Cross Section of the Flow Net in Figure 28.
Ground-Water Movement and Stratification
Nearly all ground-water systems include both aquifers
and confining beds. Figure 30 shows that ground-water
movement through these systems involves flow not only
through the aquifers but also across the confining beds.
The hydraulic conductivities of aquifers are tens to
thousands of times those of confining beds. Thus,
aquifers offer the least resistance to flow, the result be-
ing that, for a given rate of flow, the head loss per unit
of distance along a flow line is tens to thousands of
times less in aquifers than it is in confining beds. Con-
sequently, as Figure 31 shows, lateral flow in confining
beds usually is negligible, and flow lines tend to "con-
centrate" in aquifers and be parallel to aquifer bound-
aries.
Differences in the hydraulic conductivities of aquifers
and confining beds cause a refraction or bending of
flow lines at their boundaries. As flow lines move from
aquifers into confining beds, they are refracted toward
the direction perpendicular to the boundary. In other
words, they are refracted in the direction that produces
the shortest flow path in the confining bed. As the flow
lines emerge from the confining bed, they are refracted
19
-------
Figure 30. Movement of Water Through Ground-Water
Systems
Figure 31. Concentration of Flow Lines in Ground-Water
Systems
back toward the direction parallel to the boundary.
Figure 32 shows that the angles of refraction (and the
spacing of flow lines in adjacent aquifers and confining
beds) are proportional to the differences in hydraulic
conductivities (K) such that
tan
tan
(11)
In cross section, the water table is a flow line. It
represents a bounding surface for the ground-water
system; thus, in the development of many ground-
water flow equations, it is assumed to be coincident
with a flow line. However, during periods when re-
charge is arriving at the top of the capillary fringe, the
water table is also the point of origin of flow lines, as
Figure 30 shows.
Aquifer
Confining
bed
Aquifer
Figure 32. Angles of Refraction in Ground-Water Systems
The movement of water through ground-water sys-
tems is controlled by the vertical and horizontal
hydraulic conductivities and thicknesses of the aquifers
and confining beds and the hydraulic gradients. The
maximum difference in head exists between the central
parts of recharge areas and discharge areas. Because of
the relatively large head loss that occurs as water
moves across confining beds, the most vigorous cir-
culation of ground water normally occurs through the
shallowest aquifers. Movement becomes more and
more lethargic as depth increases.
The most important exceptions to the general situa-
tion described in the preceding paragraph are those
systems in which one or more of the deeper aquifers
have transmissivities significantly larger than those of
the surficial and other shallower aquifers. Thus, in
eastern North Carolina, the Castle Hayne Limestone,
which occurs at depths ranging from about 10 to
about 75 m below land surface, is the dominant
aquifer because of its very large transmissivity,
although it is overlain in most of the area by one or
more less permeable aquifers.
Ground-Water Velocity
The rate of movement of ground water is important
in many problems, particularly those related to pollu-
tion. For example, if a harmful substance is intro-
duced into an aquifer upgradient from a supply well, it
becomes a matter of great urgency to estimate when
the substance will reach the well.
The rate of movement of ground water is greatly
overestimated by many people, including those who
think in terms of ground water moving through
"veins" and underground rivers at the rates commonly
observed in surface streams. It would by more appro-
priate to compare the rate of movement of ground
water to the movement of water in the middle of a
very large lake being drained by a very small stream.
The ground-water velocity equation can be derived
from a combination of Darcy's law and the velocity
equation of hydraulics.
(Darcy's law) (6)
Q = Av (velocity equation) (12)
where Q is the rate of flow or volume per unit of time,
K is the hydraulic conductivity, A is the cross-sectional
area, at a right angle to the flow direction, through
which the flow Q occurs, dh/dl is the hydraulic gradi-
ent, and v is the Darcian velocity, which is the average
velocity of the entire cross-sectional area. Combining
equations 6 and 12, we obtain
Canceling the area terms, we find that
'-"(3)
Because this equation contains terms for hydraulic
conductivity and gradient only, it is not yet a complete
expression of ground-water velocity. The missing term
is porosity (n) because, as we know, water moves only
20
-------
through the openings in a rock. Adding the porosity
term, we obtain
v =
Kdh
ndl
(14)
In order to demonstrate the relatively slow rate of
ground-water movement, equation 14 is used to deter-
mine the rate of movement through an aquifer and a
confining bed.
1. Aquifer composed of coarse sand
K = 60 m/d
dh/dl = 1 m/1,000 m
n = 0.20
n dl
60 m2
200 md
60rn
d
= 0.3
1
1 m
0.20 1,000m
2. Confining bed composed of clay
K = 0.0001 m/d
dh/dl = 1 m/10 m
n = 0.50
_
v —
0.0001 m 1
d 0.50
0.0001 m2
5 m d
1 m
10m
= 0.00002 m c
Velocities calculated with equation 14 are, at best,
average values. Where ground-water pollution is in-
volved, the fastest rates of movement may be several
times the average rate. Also, the rates of movement in
limestone caverns, lava tubes, and large rock fractures
may approach those observed in surface streams.
Further, movement in unconfined aquifers is not
limited to the zone below the water table or to the
saturated zone. Water in the capillary fringe is sub-
jected to the same hydraulic gradient that exists at the
water table; water in the capillary fringe moves, there-
fore, in the same direction as the ground water.
Water - table
well
fringe
0 Velocity >•
Figure 33. Rate of Lateral Movement in the Capillary Fringe
As Figure 33 shows, the rate of lateral movement in
the capillary fringe decreases in an upward direction
and becomes zero at the top of the fringe. This con-
sideration is important where unconfined aquifers are
polluted with gasoline and other substances less dense
than water.
Transmissivity
The capacity of an aquifer to transmit water of the
prevailing kinematic viscosity is referred to as its
transmissivity. The transmissivity (T) of an aquifer is
equal to the hydraulic conductivity of the aquifer
multipled by the saturated thickness of the aquifer.
Thus,
T = Kb (15)
where T is transmissivity, K is hydraulic conductivity,
and b is aquifer thickness.
dl = 1000m
^-— _
...
Confinina
f^lavr
Confined
Aquifer * ° '
.- -Sand «K=50
Confining Bed
— Bed ^=^^-'—
-
.°/>=1<
• ' \;
.™ • • •
)0m .*B
. o
V. .*' .
~ Clay - - "]
Figure 34. Water Flow Through an Aquifer
As is the case with hydraulic conductivity, transmis-
sivity is also defined in terms of a unit hydraulic gra-
dient.
If equation 15 is combined with Darcy's law (see the
"Hydraulic Conductivity" section of this chapter), the
result is an equation that can be used to calculate the
quantity of water (q) moving through a unit width (w)
of an aquifer. Darcy's law is
q = KA |^| (6a)
Expressing area (A) as bw, we obtain
q = Kbw^jj\ (9)
Next, expressing transmissivity (T) as Kb, we obtain
(16)
21
-------
Equation 16 modified to determine the quantity of
water (0) moving through a large width (W) of an
aquifer is
Q = TwW
dl
(17)
or, if it is recognized that 7 applies to a unit width (w)
of an aquifer, this equation can be stated more simply
as
(18)
If equation 18 is applied to Figure 34, the quantity of
water flowing out of the right-hand side of the figure
can be calculated by using the values shown on the
figure as follows:
T = Kb =
0= TW
dh\ 5,000m2 1,000m
dll d 1
1
1m
1,000m
= 5,000 m2d-1
= 5,000 m3d-1
Equation 18 is also used to calculate transmissivity,
where the quantity of water (Q) discharging from a
known width of aquifer can be determined as, for ex-
ample, with streamflow measurements. Rearranging
terms, we obtain
W \dh
(19)
The units of transmissivity, as the preceding equation
demonstrates, are
T= (m3d~1)(m) _ m^
(m)(m) d
Figure 35 illustrates the hydrologic situation that per-
mits calculation of transmissivity through the use of
stream discharge. The calculation can be made only
during dry-weather (baseflow) periods, when all water
in the stream is derived from ground-water discharge.
For the purpose of this example, the following values
are assumed:
,3 c-1
Average daily flow at
stream-gaging station A: 2.485 m3 s
Average daily flow at
stream-gaging station B: 2.355 m3 d ~1
Increase in flow due to
ground-water discharge: 0.130 m3 s~1
Total daily ground-water
discharge to stream: 11,232 m3 d~1
Discharge from half of aquifer
(one side of the stream): 5,616 m3 d ~1
Distance (x) between stations
A and B: 5,000 m
Average thickness of aquifer (b): 50 m
Average slope of the water table
(dhldl) determined from
measurements in the observa-
tion wells: 1 m/2,000 m
tream-gaging
station B
Stream-gaging
station A
Confining bed
Figure 35. Hydrologic Situation Allowing Calculation of Transmissivity
22
-------
By equation 19,
7- -
1 ~ W
_dl_ = 5,616 m3
dh ~ d x 5,000 m
2,000m=
1 m
The hydraulic conductivity is determined from equa-
tion 15 as follows:
b d x 50 m
Because transmissivity depends on both K and b, its
value differs in different aquifers and from place to
place in the same aquifer. Estimated values of transmis-
sivity for the principal aquifers in different parts of the
country range from less than 1 m2 d ~ 1 for some frac-
tured sedimentary and igneous rocks to 100,000 m2 d" 1
for cavernous limestones and lava flows.
Finally, transmissivity replaces the term "coefficient
of transmissibility" because, by convention, an aquifer
is transmissive, and the water in it is transmissible.
Storage Coefficient
The abilities (capacities) of water-bearing materials to
store and to transmit water are their most important
hydraulic properties. Depending on the intended use of
the information, these properties are given either in
terms of a unit cube of the material or in terms of a
unit prism of an aquifer.
Property
Unit cube of material Unit prism of aquifer
Transmissive capacity
Available storage
Hydraulic conductivity (K) Transmissivity (T)
Specific yield (Sy) Storage coefficient (S)
The storage coefficient (S) is defined as the volume
of water that an aquifer releases from or takes into
storage per unit surface area of the aquifer per unit
change in head. The storage coefficient is a dimension-
less unit, as equation 20 shows, in which the units in
the numerator and the denominator cancel:
S =
volume of water
(unit area)(unit head change)
V111 I _ '" OQ)
(m^m) m3
The size of the storage coefficient depends on
whether the aquifer is confined or unconfined, as Figure
36 shows. If the aquifer is confined, the water released
from storage when the head declines comes from expan-
sion of the water and from compression of the aquifer.
Relative to a confined aquifer, the expansion of a given
volume of water in response to a decline in pressure is
very small. In a confined aquifer having a porosity of
0.2 and containing water at a temperature of about
15°C, expansion of the water alone releases about
3 x 10~7 m3 of water per cubic meter of aquifer per
meter of decline in head. To determine the storage coef-
ficient of an aquifer due to expansion of the water, it is
necessary to multiply the aquifer thickness by 3 x 10~7.
Thus, if only the expansion of water is considered, the
storage coefficient of an aquifer 100 m thick would be
3 x 10'5. The storage coefficient of most confined
aquifers ranges from about 10-5 to 10-3 (0.00001 to
0.001). The difference between these values and the
value due to expansion of the water is attributed to
compression of the aquifer.
Figure 37 will aid in understanding this phenomenon.
It shows a microscopic view of the contact between an
aquifer and the overlying confining bed. The total load
on the top of the aquifer is supported partly by the
solid skeleton of the aquifer and partly by the hydraulic
pressure exerted by the water in the aquifer. When the
water pressure declines, more of the load must be sup-
Con fin ing be_d — -
Unit declines
m heads
Water released
from storage
Figure 36. Size of the Storage Coefficient in Confined and Unconfined Aquifers
Potentiometnc surfoce
~C orvfj n ing b e d—•'-
23
-------
ported by the solid skeleton. As a result, the rock par-
ticles are distorted, and the pore space is reduced. The
water forced from the pores when their volume is
Confining bed — — ^_~ ~~
_I_" -H Total "load on aquifer .
Suppor
through
rock
ske leton
Support
through
water
Figure 37. Microscopic View of the Contact Between an
Aquifer and the Overlying Confining Bed
Land surface
reduced represents the part of the storage coefficient
due to compression of the aquifer.
If the aquifer is unconfined, the predominant source
of water is from gravity drainage of the sediments
through which the decline in the water table occurs. In
an unconfined aquifer, the volume of water derived
from expansion of the water and compression of the
aquifer is negligible. Thus, in such an aquifer, the
storage coefficient is virtually equal to the specific yield
and ranges from about 0.1 to about 0.3.
Because of the difference in the sources of storage,
the storage coefficient of unconfined aquifers is 100 to
10,000 times the storage coefficient of confined
aquifers, as Figure 36 illustrates. However, if water
levels in an area are reduced to the point where an
aquifer changes from a confined condition to an uncon-
fined condition, the storage coefficient of the aquifer
immediately increases from that of a confined aquifer
to that of an unconfined aquifer.
Long-term withdrawals of water from many confined
aquifers result in drainage of water both from clay
layers within the aquifer and from adjacent confining
beds. This drainage increases the load on the solid
skeleton and results in compression of the aquifer and
subsidence of the land surface. Subsidence of the land
surface caused by drainage of clay layers has occurred
in Arizona, California, Texas, and other areas.
Potentiometric
--=• surface-^
- --- . ^
- •- _
— —
—
•
— .
.
— , ,
• •,•-•.•.•«•••••.•.'•
'.•••.•.•.•.•••••I-
_• • :_i_l_'_l_: l-J—'
Confining bed
Aquifer
Total
storage
Porosity
Available
storage
Artesian
storage
coefficient
Specific yield
Drainage of fine-grained
beds, including confining bed
Sources of
available storage
Expansion of
water and
compression
of the aquifer
Partial drainage
of pores
Reduction of porosity
of fine-grained beds
Bedrock
Figure 38. Potential Sources of Water in a Two-Unit Ground-Water System
24
-------
The potential sources of water in a two-unit ground-
water system consisting of a confining bed and a confin-
ed aquifer are shown in Figure 38. The figure is based
on the assumption that water is removed in two
separate stages—the first while the potentiometric sur-
face is lowered to the top of the aquifer and the second
by dewatering the aquifer.
The differences in the storage coefficients of confined
and unconfined aquifers are of great importance in
determining the response of the aquifers to stresses such
as withdrawals through wells.
Cone of Depression
Both wells and springs serve as sources of ground-
water supply. However, most springs having yields large
enough to meet municipal, industrial, and large com-
mercial and agricultural needs occur only in areas
underlain by cavernous limestones and lava flows.
Therefore, most ground-water needs are met by with-
drawals from wells.
The response of aquifers to withdrawals from wells is
the well, the hydraulic gradient must get steeper toward
the well.
Several important differences exist between the cones
of depression in confined and unconfined aquifers.
Figure 39 shows that withdrawals from an unconfined
aquifer result in drainage of water from the rocks
through which the water table declines as the cone of
depression forms. Because the storage coefficient of an
unconfined aquifer equals the specific yield of the
aquifer material, the cone of depression expands very
slowly. On the other hand, dewatering of the aquifer
results in a decrease in transmissivity, which causes, in
turn, an increase in drawdown both in the well and in
the aquifer.
Figure 40 shows that withdrawals from a cofifined
aquifer cause a drawdown in artesian pressure but do
not (normally) cause a dewatering of the aquifer. The
water withdrawn from a confined aquifer is derived
from expansion of the water and compression of the
rock skeleton of the aquifer. (See the "Storage Coeffi-
cient" section of this chapter.) The very small storage
coefficient of confined aquifers results in a very rapid
Land surface
Limits of cone
depression
Land surface
Figure 39. Withdrawals from an Unconfined Aquifer
an important topic in ground-water hydrology. When
withdrawals start, the water level in the well begins to
decline as water is removed from storage in the well.
The head in the well falls below the level in the sur-
rounding aquifer. As a result, water begins to move
from the aquifer into the well. As pumping continues,
the water level in the well continues to decline, and the
rate of flow into the well from the aquifer continues to
increase until the rate of inflow equals the rate of
withdrawal.
Figures 39 and 40 illustrate that the movement of
water from an aquifer into a well results in the forma-
tion of a cone of depression. Because water must con-
verge on the well from all directions and because the
area through which the flow occurs decreases toward
i Potintiomttrie turfoct
------,
^/'
Drawdown
Confining
////////
0
Confined
\
bed '
////////
0 +~
*-
„_ ^_
aquifer
1
1
^
1
I
1
1
N^\
X" Cone of
/ ^^ depression
/
1
SS//S/// ////// / /
*+ 0
-«— 0
-^ 0
^ B
Conf in ng bed
Figure 40. Withdrawals from a Confined Aquifer
expansion of the cone of depression. Consequently, the
mutual interference of expanding cones around adjacent
wells occurs more rapidly in confined aquifers than it
does in unconfined aquifers.
Cones of depression caused by large withdrawals
from extensive confined aquifers can affect very large
areas. Figure 41 shows the overlapping cones of depres-
sion that existed in 1981 in an extensive confined
aquifer composed of unconsolidated sands and inter-
bedded silt and clay of Cretaceous age in the central
part of the Atlantic Coastal Plain. The cones of depres-
sion are caused by withdrawals of about 277,000 m3
d-1 (73,000,000 gal d-') from well fields in Virginia
and North Carolina. (See the "Source of Water Derived
From Wells" section of this chapter.)
25
-------
VIRGINIA
CAROLINA
36"
10
70
so Kilometers
Explanation
Water levels are in feet
National Geodetic Vertical Datum 1929
Figure 41. Potentiometric Surface of the Lowermost Cretaceous Aquifer in Southeastern Virginia and Northeastern North Carolina
26
-------
Source of Water Derived from Wells
Both the economical development and the effective
management of any ground-water system require an
understanding of the response of the system to with-
drawals from wells. The first concise description of the
hydrologic principles involved in this response was
presented by C. V. Theis in a paper published in 1940.
Theis pointed out that the response of an aquifer to
withdrawals from wells depends on:
1. The rate of expansion of the cone of depression
caused by the withdrawals, which depends on the
transmissivity and the storage coefficient of the
aquifer.
2. The distance to areas in which the rate of water
discharging from the aquifer can be reduced.
3. The distance to recharge areas in which the rate
of recharge can be increased.
Figure 42 shows that over a sufficiently long period
of time under natural conditions—that is, before the
start of withdrawals—the discharge from every
ground-water system equals the recharge to it. In other
words,
natural discharge (D) = natural recharge (R) (21)
In the eastern part of the United States and in the
more humid areas in the West, the amount and dis-
tribution of precipitation are such that the period of
time over which discharge and recharge balance may
be less than a year or, at most, a few years. In the
drier parts of the country—that is, in the areas that
generally receive less than about 500 mm of precipita-
tion annually — the period over which discharge and
recharge balance may be several years or even cen-
turies. Over shorter periods of time, differences be-
tween discharge and recharge involve changes in
ground-water storage. In other words, when discharge
exceeds recharge, ground-water storage (S) is reduced
by an amount AS equal to the difference between dis-
charge and recharge. Thus,
D = R+AS
(22)
Conversely, when recharge exceeds discharge, ground-
water storage is increased. Thus,
D = R - AS (23)
Figure 43 shows that when withdrawal through a
well begins, water is removed from storage in its
vicinity as the cone of depression develops. Thus, the
withdrawal (Q) is balanced by a reduction in ground-
water storage. In other words,
0 = AS
(24)
As the cone of depression expands outward from the
pumping well, it may reach an area where water is
discharging from the aquifer. Figure 44 shows that the
hydraulic gradient will be reduced toward the dis-
charge area, and the rate of natural discharge will
decrease. To the extent that the decrease in natural
discharge compensates for the pumpage, the rate at
which water is being removed from storage will also
decrease, and the rate of expansion of the cone of
depression will decline. If and when the reduction in
natural discharge (AD) equals the rate of withdrawal
(Q), a new balance will be established in the aquifer.
This balance in symbolic form is
(D - AD) + Q = R
(25)
Conversely, if the cone of depression expands into a
recharge area rather than into a natural discharge area,
the hydraulic gradient between the recharge area and
the pumping well will be increased. If, under natural
conditions, more water was available in the recharge
area than the aquifer could accept (the condition that
Theis referred to as one of rejected recharge), the in-
crease in the gradient away from the recharge area will
permit more recharge to occur, and the rate of growth
of the cone of depression will decrease. If and when
the increase in recharge (Afl) equals the rate of with-
drawal (Q), a new balance will be established in the
aquifer, and expansion of the cone of depression will
cease. The new balance in symbolic form is
D + Q = fl +
(26)
In the eastern part of the United States, gaining
streams are relatively closely spaced, and areas in
which rejected recharge occurs are relatively unimpor-
tant. In this region, the growth of cones of depression
first commonly causes a reduction in natural dis-
charge. Figure 45 shows that, if the pumping wells are
near a stream or if the withdrawals are continued long
enough, ground-water discharge to a stream may be
stopped entirely in the vicinity of the wells, and water
may be induced to move from the stream into the
aquifer. In other words, the tendency in this region is
for withdrawals to change discharge areas into
recharge areas. This consideration is important where
the streams contain brackish or polluted water or
where the streamflow is committed or required for
other purposes.
To summarize, the withdrawal of ground water
through a well reduces the water in storage in the
source aquifer during the growth of the cone of
depression. When and if the cone of depression ceases
to expand, the rate of withdrawal is being balanced by
a reduction in the rate of natural discharge and (or) by
an increase in the rate of recharge. Under this condi-
tion,
Q = AD + Afl
(27)
27
-------
Stream -
— /S » '" ™ ^ "-•—••! iWif^ifrTffiJttyrTg^jfg
-_lormning ^ bed^p —^_____^rj__^;
Figure 42. Discharge (D) = Recharge («)
-^^
Figure 43. Withdrawal (Q) = Reduction in Storage (AS)
'•'OV ~v' '•.''-'•'''-' •'''•»V,-\S*L'^-'V','•};-''-,-. - - "si- V.1 JVs '.'
^4: ,v-.;- - -;>;-' v/, ;,• -;^-;-;7^V^--?>- -SL^-?ri*
•-^* "**»'•' •'•• V -1 • - * '* v-'-'-':-.\- --^*- -f*" -JS. ***
..^^^•5^;*^;-^»---iLk,--^£.^
Figure 44. Withdrawal (0) = Reduction in Storage (AS, + Reduction in Discharge (AD) )
Figure 45. Withdrawal (O) = Reduction in Discharge (AD) + Increase in Recharge
(A/?)
28
-------
Aquifer Tests
Determining the yield of ground-water systems and
evaluating the movement and fate of ground-water
pollutants require, among other information, knowl-
edge of:
1. The position and thickness of aquifers and con-
fining beds.
2. The transmissivity and storage coefficient of the
aquifers.
3. The hydraulic characteristics of the confining
beds.
4. The position and nature of the aquifer bound-
aries.
5. The location and amounts of ground-water with-
drawals.
6. The locations, kinds, and amounts of pollutants
and pollutant practices.
3. Accurate water-level measurements made at
precisely known times during both the drawdown
and the recovery periods.
Drawdown is the difference between the water level
at any time during the test and the position at which
the water level would have been if withdrawals had not
started. Drawdown is very rapid at first. As pumping
continues and the cone of depression expands, as
Figure 47 shows, the rate of drawdown decreases.
The recovery of the water level under ideal condi-
tions is a mirror image of the drawdown. The change
in water level during the recovery period is the same as
if withdrawals had continued at the same rate from the
pumped well but, at the moment of pump cutoff, a
recharge well had begun recharging water at the same
point and at the same rate. Therefore, the recovery of
the water level is the difference between the actual
measured level and the projected pumping level.
In addition to the constant-rate aquifer test men-
tioned above, analytical methods have also been devel-
4
5
6
r
8
9
10
1 1
1?
r t i '
xPump on
— ^— — — ^^^. Jr n
f "
t
w>
i i
egiona
i i r r
trend
S'
1 1
-1 - o
•o
\ Z - -^
Vf
^
Prepumpinq
"period
i 1
Pumping
" period '
, „ i „ i
off \
"----. 4l
>•>
•>
>
o
w
(K
_ I r _ _
1 1 1 1 I I . .
10
12 13 14 15 16
Days
Figure 46. Map of Aquifer Test Site
Figure 47. Change of Water Level in Well B
Acquiring knowledge on these factors requires both
geologic and hydrologic investigations. One of the
most important hydrologic studies involves analyzing
the change, with time, in water levels (or total heads)
in an aquifer caused by withdrawals through wells.
This type of study is referred to as an aquifer test and,
in most cases, includes pumping a well at a constant
rate for a period ranging from several hours to several
days and measuring the change in water level in obser-
vation wells located at different distances from the
pumped well. Figure 46 shows the map of an aquifer
test site.
Successful aquifer tests require, among other things:
1. Determination of the prepumping water-level
trend (that is, the regional trend).
2. A carefully controlled constant pumping rate.
oped for several other types of aquifer tests. These
methods include tests in which the rate of withdrawal
is variable and tests that involve leakage of water
across confining beds into confined aquifers. The
analytical methods available also permit analysis of
tests conducted on both vertical wells and horizontal
wells or drains.
The most commonly used method of analysis of
aquifer-test data—that for a vertical well pumped at a
constant rate from an aquifer not affected by vertical
leakage and lateral boundaries—will be covered in the
discussion of "Analysis of Aquifer-Test Data." The
method of analysis requires the use of a type curve
based on the values of W(u) and Mu listed in Table 5.
Preparation and use of the type curve are covered in
the following discussion.
29
-------
Table 5. Selected Values of W(u) for Values of 1/u
1/0
10-1
1
10
102
103
10"
105
106
107
108
10«
1010
10"
1012
1013
1014
10
0.219
1.82
4.04
6.33
8.63
10.94
13.24
15.54
17.84
20.15
22.45
24.75
27.05
29.36
31.66
33.96
7.69
0.135
1.59
3.78
6.07
8.37
10.67
12.98
15.28
17.58
19.88
22.19
24.49
26.79
20.09
31.40
33.70
5.88
0.075
1.36
3.51
5.80
8.10
10.41
12.71
15.01
17.31
19.62
21.92
24.22
26.52
28.83
31.13
33.43
5.00
0.049
1.22
3.35
5.64
7.94
10.24
12.55
14.85
17.15
19.45
21.76
24.06
26.36
2866
30.97
33.27
4.00
0.025
104
3.14
5.42
7.72
10.02
12.32
14.62
1693
19.23
21.53
23.83
26.14
28.44
30.74
3305
3.33
0.013
.91
2.96
5.23
7.53
984
12.14
14.44
16.74
19.05
21.35
23.65
25.96
28.26
30.56
32.86
2.86
0.007
.79
2.81
5.08
7.38
9.68
11.99
14.29
16.59
18.89
21.20
23.50
25.80
28.10
30.41
32.71
2.5
0.004
.70
2.68
4.95
7.25
9.55
11.85
1415
16.46
18.76
21.06
23.36
25.67
2797
30.27
32.58
2.22
0.002
63
2.57
4.83
7.13
9.43
11.73
14.04
16.34
18.64
20.94
23.25
25.55
27.85
30.15
32.46
2.00
0.001
.56
2.47
4.73
7.02
9.33
11.63
13.93
16.23
18.54
20.84
23.14
25.44
2775
30.05
32.35
1.67
0.000
.45
2.30
4.54
6.84
9.14
11 45
13.75
16.05
18.35
20.66
22.96
25.26
27.56
29.87
32.17
1.43
0.000
37
2.15
4.39
6.69
8.99
11.29
13.60
1590
18.20
20.50
22.81
25.11
27.41
29.71
32.02
1.25
0.000
.31
2.03
4.26
6.55
8.86
11.16
13.46
15.76
18.07
2037
22.67
24.97
27.28
29.58
31.88
1.11
0.000
.26
1.92
4.14
6.44
8.74
11.04
13.34
15.65
17.95
20.25
22.55
24.86
27.16
29.46
31.76
Examples: when 1/u = 10X1Q-1, W(u) = 0.219, when 1/u = 3.33x102, W(u) = 5.23.
Analysis of Aquifer-Test Data
In 1935, C. V. Theis of the New Mexico Water
Resources District of the U.S. Geological Survey devel-
oped the first equation to include time of pumping as a
factor that could be used to analyze the effect of with-
drawals from a well. Thus, the Theis equation per-
mitted, for the first time, determination of the hydraulic
characteristics of an aquifer before the development of
new steady-state conditions resulting from pumping.
The importance of this capability may be realized from
the fact that, under most conditions, a new steady state
cannot be developed or that, if it can, many months or
years may be required.
Theis assumed in the development of the equation
that:
1. The transmissivity of the aquifer tapped by the
pumping well is constant during the test to the
limits of the cone of depression.
2. The water withdrawn from the aquifer is derived
entirely from storage and is discharged instan-
taneously with the decline in head.
3. The discharging well penetrates the entire thickness
of the aquifer, and its diameter is small in com-
parison with the pumping rate, so that storage in
the well is negligible.
These assumptions are most nearly met by confined
aquifers at sites remote from their boundaries. How-
ever, if certain precautions are observed, the equation
can also be used to analyze tests of unconfined
aquifers.
The forms of the Theis equation used to determine
the transmissivity and storage coefficient are
^ Q W(u)
4nS
S =
4Ttu
(28)
(29)
where T is transmissivity, S is the storage coefficient,
Q is the pumping rate, s is drawdown, t is time, r is
the distance from the pumping well to the observation
well, W(u) is the well function of u, which equals
//2 i/3 i|4
:3! 4x4!
and
u = (r2S)/(4Tt).
The form of the Theis equation is such that it can-
not be solved directly. To overcome this problem,
Theis devised a convenient graphic method of solution
that involves the use of a type curve, shown in Figure
48. To apply this method, a data plot of drawdown
versus time (or drawdown versus r/r2) is matched to
the type curve of W(u) versus 1/u as shown in Figure
49. At some convenient point on the overlapping part
of the sheets containing the data plot and type curve,
values of s, t (or t/r2), W(u), and 1/u are noted. These
values are then substituted in equations 28 and 29,
which are solved for T and S, respectively.
A Theis type curve of W(u) versus 1/u can be
prepared from the values given in the table contained
in the preceding section, "Aquifer Tests." The data
points are plotted on logarithmic graph paper—that is,
graph paper having logarithmic divisions in both the x
and y directions.
The dimensional units of transmissivity (T) are
L2/~1, where L is length and t is time in days. Thus, if
Q in equation 28 is in cubic meters per day and s is in
meters, T will be in square meters per day. Similarly,
if, in equation 29, T is in square meters per day, t is in
days, and r is in meters, S will be dimensionless.
Traditionally, in the United States, T has been ex-
pressed in units of gallons per day per foot. The com-
mon practice now is to report transmissivity in units of
square meters per day or square feet per day. If Q is
measured in gallons per minute, as is still normally the
case, and drawdown is measured in feet, as is also nor-
30
-------
Figure 48. Theis Type Curve
10
t, in minutes
10*
I0
I04
io-
10
0 1
0.01
/
Match
Point
+ x
/
..'•'
W(u) --
__.,.
Match-Point
1, s-. 2.2
1 , r = 1 8
Data Plot
Q- 1.9 m3 min"1
r- 187 m
-•-•-•-
Coordinates
0 m
min
Type Curve
U)
10 •-
0>
e
c
O.I
0 I
10
I0
I0
I0
Figure 49. Data Plot of Drawdown Versus Time Matched to Theis Type Curve
mally the case, equation 28 is modified to obtain T in
square feet per day as follows:
ft3
^ Q W(u) _ gal „ l,440minx
4ns min d 7.48 gal ft 4n
„
W(u)
or
(when Q is in gallons per minute and s is in feet). To
convert square feet per day to square meters per day,
divide by 10.76.
The storage coefficient is dimensionless. Therefore,
if T is in square feet per day, t is in minutes, and r is
in feet, then, by equation 29,
4 ^ ft2 r min x d
1 d ft2 1,440 min
s _
or
o _
Ttu
360 r2
(When T is in square feet per day, / is in minutes, and
r is in feet).
Analysis of aquifer-test data using the Theis equa-
tion involves plotting both the type curve and the test
data on logarithmic graph paper. If the aquifer and
31
-------
the conditions of the test satisfy Theis's assumptions,
the type curve has the same shape as the cone of
depression along any line radiating away from the
pumping well and the drawdown graph at any point in
the cone of depression.
Use of the Theis equation for unconfined aquifers
involves two considerations. First, if the aquifer is
relatively fine grained, water is released slowly over a
period of hours or days, not instantaneously with the
decline in head. Therefore, the value of S determined
from a short-period test may be too small.
Second, if the pumping rate is large and the observa-
tion well is near the pumping well, dewatering of the
aquifer may be significant, and the assumption that
the transmissivity of the aquifer is constant is not
satisfied. The effect of dewatering of the aquifer can
be eliminated with the following equation:
(30)
where s is the observed drawdown in the unconfined
aquifer, b is the aquifer thickness, and s' is the draw-
down that would have occurred if the aquifer had been
confined (that is, if no dewatering had occurred).
To determine the transmissivity and storage coeffi-
cient of an unconfined aquifer, a data plot consisting
of s' versus t (or t/r2) is mathed with the Theis type
curve of W(u) versus 1/u. Both s and b in equation 30
must be in the same units, either feet or meters.
As noted above, Theis assumed in the development
of his equation that the discharging well penetrates the
entire thickness of the aquifer. However, because it is
not always possible, or necessarily desirable, to design
a well that fully penetrates the aquifer under develop-
ment, most discharging wells are open to only a part
of the aquifer that they draw from. Such partial pene-
tration creates vertical flow in the vicinity of the dis-
charging well that may affect drawdowns in observa-
tion wells located relatively close to the discharging
well. Drawdowns in observation wells that are open to
the same zone as the discharging well will be larger
than the drawdowns in wells at the same distance from
the discharging well but open to other zones. The
possible effect of partial penetration on drawdowns
must be considered in the analysis of aquifer-test data.
If aquifer-boundary and other conditions permit, the
problem can be avoided by locating observation wells
beyond the zone in which vertical flow exists.
Time-Drawdown Analysis
The Theis equation is only one of several methods
that have been developed for the analysis of aquifer-
test data. (See the "Analysis of Aquifer-Test Data"
section of this chapter.) Another method, and one that
is somewhat more convenient to use, was developed by
C. E. Jacob from the Theis equation. The greater con-
venience of the Jacob method derives partly from its
use of semilogarithmic graph paper instead of the
logarithmic paper used in the Theis method and from
the fact that, under ideal conditions, the data plot
along a straight line rather than along a curve.
However, it is essential to note that, whereas the
Theis equation applies at all times and places (if the
assumptions are met), Jacob's method applies only
under certain additional conditions. These conditions
must also be satisfied in order to obtain reliable
answers.
To understand the limitations of Jacob's method,
we must consider the changes that occur in the cone of
depression during an aquifer test. The changes that are
of concern involve both the shape of the cone and the
rate of drawdown. As the cone of depression migrates
outward from a pumping well, its shape (and, there-
fore, the hydraulic gradient at different points in the
cone) changes. We can refer to this condition as
unsteady shape. At the start of withdrawals Figure 50
shows that the entire cone of depression has an
unsteady shape. After a test has been underway for
some time, Figure 51 shows that the cone of depres-
sion begins to assume a relatively steady shape, first at
the pumping well and then gradually to greater and
Land surface
Cone of depression---**1'^!
(unsteady shape) ^
/xxyxw/xxxx/x/xxxxxxxxxxxxxxxxxxxxxx.
f
jT
Confining bed
'//////Ssssssssssssssss/s/f/
jJLi/
1
5 .*_ \^f
'• Confined aquifer
^^yxxxxxxxx^ixxi^5?^xxxx^^
Figure 50. Cone of Depression at Start of Withdrawal
Unsteady shape
Steady shape
Figure 51. Cone of Depression During Test
River
Figure 52. Steady-State Cone of Depression
32
-------
A
?
CD
4-«
o> 4
2 H
"5?
i fi
s c
1
CO Q
Q
ID
1?
/' -r-^ijii I I
T X
lo
As* - 1
S - 1.
-
—
r- 75 m
Q = 9.3m 3
f0 = 2.5x
t I i i i i i I
•^ncrtc
K __
21-., -b
III **1
mirT1 ( 24!
I0~5d
1 1 1 1 1 M 1
^*~""~~"^»t-^
~~*fc.
L 0 Q
cycle
55 gal mm"1
t i i p 1 1 1 1
1 1 1 1 M 1 1
~~~*"^--)l^
^-*~~-
)
1 i i 1 1 1 i I
1 1 i 1 1 1 1 '
_ .
U ro w a
/^measu
>-cL
^*~ >L
^~— *.
i 1 1 1 1 1 11
own _
re ments
*--x
-
i i i 1 1 1 1 1
10-5
10-4
10-3
10-2
Time, in Days
0.
10
Figure 53. Time-Drawdown Graph
greater distances. If withdrawals continue long enough
for increases in recharge and (or) reductions in dis-
charge to balance the rate of withdrawal, drawdowns
cease, and the cone of depression is said to be in a
steady state, as Figure 52 shows.
The Jacob method is applicable only to the zone in
which steady-shape conditions prevail or to the entire
cone only after steady-state conditions have developed.
For practical purposes, this condition is met when
u - (r2S)/)4ff) is equal to or less than about 0.05.
Substituting this value in the equation for u and solv-
ing for t, we can determine the time at which steady-
shape conditions develop at the outermost observation
well. Thus,
_ 7,200r2S ,,n
C£ — _. \J I)
where tc is the time, in minutes, at which steady-shape
conditions develop, r is the distance from the pumping
well, in feet (or meters), S is the estimated storage
coefficient (dimensionless), and T is the estimated
transmissivity, in square feet per day (or square meters
per day).
After steady-shape conditions have developed, the
drawdowns at an observation well begin to fall along a
straight line on semilogarithmic graph paper, as Figure
53 shows. Before that time, the drawdowns plot below
the extension of the straight line. When a time-draw-
down graph is prepared, drawdowns are plotted on the
vertical (arithmetic) axis versus time on the horizontal
(logarithmic) axis.
The slope of the straight line is proportional to the
pumping rate and to the transmissivity. Jacob derived
the following equations for determination of trans-
missivity and storage co-efficient from the time-draw-
down graphs:
f= 2-3Q
4nAs
_ 2.25 Tt0
r2
(32)
(33)
where Q is the pumping rate, As is the drawdown
across one log cycle, t0 is the time at the point where
the straight line intersects the zero-drawndown line,
and r is the distance from the pumping well to the
observation well.
Equations 32 and 33 are in consistent units. Thus, if
0 is in cubic meters per day and s is in meters, T is in
square meters per day. S is dimensionless, so that, in
equation 33, if T is in square meters per day, then r
must be in meters and r0 must be in days.
It is still common practice in the United States to ex-
press Q in gallons per minute, s in feet, t in minutes, r
in feet, and T in square feet per day. We can modify
equations 32 and 33 for direct substitution of these
units as follows:
T __ 2.3 Q _ 2.3 y gal x 1,440 min
4nAs 4n min d
ft2
7.48 gal
35 Q
s
(34)
(where T is in square feet per day, Q is in gallons per
minute, and s is in feet) and
S =
2.25 ff,
o _
1 d ft2 1,440 min
33
-------
(35)
(where T is in square feet per day, t0 is in minutes,
and r is in feet).
Distance-Drawdown Analysis
It is desirable in aquifer tests to have at least three
observation wells located at different distances from
the pumping well, as shown in Figure 54. Drawdowns
Observation wells
Pumping well
Depth to V\
i
*
i
S^*^-
o
•
/ater—
-- — -^
i
/ X / /
! t
•
•«
1
— r—
// /
\
//
^
/Static water level
/"^ ^Pumping wate
/ level
Confining bed
S / / / / /////////s///
Confined 1
aquifer P
Confining bed
r Datum Plane
Figure 54. Desirable Location for Observation Wells in
Aquifer Tests.
measured at the same time in these wells can be
analyzed with the Theis equation and type curve to
determine the aquifer transmissivity and storage coeffi-
cient.
After the test has been underway long enough,
drawdowns in the wells can also be analyzed by the
Jacob method, either through the use of a time-draw-
down graph using data from individual wells or
through the use of a distance-drawdown graph using
"simultaneous" measurements in all of the wells. To
determine when sufficient time has elapsed, see the
"Time-Drawdown Analysis" section of this chapter.
In the Jacob distance-drawdown method, draw-
downs are plotted on the vertical (arithmetic) axis ver-
sus distance on the horizontal (logarithmic) axis, as
shown in Figure 55. If the aquifer and test conditions
satisfy the Theis assumptions and the limitation of the
Jacob method, the drawdowns measured at the same
time in different wells should plot along a straight line.
The slope of the straight line is proportional to the
pumping rate and to the transmissivity. Jacob derived
the following equations for determination of the trans-
missivity and storage coefficient from distance-
drawdown graphs:
7 =
S =
2.3 Q
2nAs
2.25 Tt
(36)
(37)
where Q is the pumping rate, As is the drawdown
across one log cycle, t is the time at which the draw-
downs were measured, and r0 is the distance from the
pumping well to the point where the straight line inter-
sects the zero-drawdown line.
Equations 36 and 37 are in consistent units. For the
inconsistent units still in relatively common use in the
United States, equations 36 and 37 should be used in
the following forms:
T =
70 Q
As
(38)
(where T is in square feet per day, Q is in gallons per
minute, and s is in feet) and
Tt
640 rg
(39)
I—I I II 111 I 1—I I 111
f = 4 days
m 3 min-'( 2,455 gal
- ^= 30,000 m
Figure 55. Distance-Drawdown Graph
34
10 100 1000
Distance, in Meters
-------
(where T is in square feet per day, t is in minutes, and
r0 is in feet).
The distance r0 does not indicate the outer limit of
the cone of depression. Because nonsteady-shape con-
ditions exist in the outer part of the cone, before the
development of steady-state conditions, the Jacob
method does not apply to that part. If the Theis equa-
tion were used to calculate drawdowns in the outer
part of the cone, it would be found that they would
plot below the straight line. In other words, the
measurable limit of the cone of depression is beyond
the distance r0 _
If the straight line of the distance-drawdown graph
is extended inward to the radius of the pumping well,
the drawdown indicated at that point is the drawdown
in the aquifer outside of the well. If the drawdown in-
side the well is found to be greater than the drawdown
outside, the difference is attributable to well loss. (See
the "Single-Well Tests" section of this chapter.)
As noted in the section on "Hydraulic Conductiv-
ity," the hydraulic conductivities and, therefore, the
transmissivities of aquifers may be different in dif-
ferent directions. These differences may cause draw-
downs measured at the same time in observation wells
located at the same distances but in different directions
from the discharging well to be different. Where this
condition exists, the distance-drawdown method may
yield satisfactory results only where three or more
observation wells are located in the same direction but
at different distances from the discharging well.
Single-Well Tests
The most useful aquifer tests are those that include
water-level measurements in observation wells. Such
tests are commonly referred to as multiple-well tests. It
is also possible to obtain useful data from production
wells, even where observation wells are not available.
Such tests are referred to as single-well tests and may
consist of pumping a well at a single constant rate, or
at two or more different but constant rates (see the
"Well-Acceptance Tests and Well Efficiency" section of
this chapter) or, if the well is not equipped with a
pump, by "instantaneously" introducing a known
volume of water into the well. This discussion will be
limited to tests involving a single constant rate.
In order to analyze the data, it is necessary to under-
stand the nature of the drawdown in a pumping well.
Figure 56 shows that the total drawdown (Sf) in most, if
not all, pumping wells consists of two components. One
is the drawdown (sa) in the aquifer, and the other is the
drawdown (s^,) that occurs as water moves from the
aquifer into the well and up the well bore to the pump
intake. Thus, the drawdown in most pumping wells is
greater than the drawdown in the aquifer at the radius
of the pumping well.
The total drawdown (Sf) in a pumping well can be ex-
pressed in the form of the following equations:
— Sg + S|/y
6Q+CQ2
(40)
(41)
where Sa is the drawdown in the aquifer at the effective
radius of the pumping well, sw is well loss, Q is the
pumping rate, 8 is a factor related to the hydraulic
characteristics of the aquifer and the length of the
pumping period, and C is a factor related to the charac-
teristics of the well.
The factor C in equation 40 normally considered to
be constant, so that, in a constant rate test, CQ2 is also
constant. As a result, the well loss (sw) increases the
total drawdown in the pumping well but does not affect
the rate of change in the drawdown with time. It is,
therefore, possible to analyze drawdowns in the pump-
ing well with the Jacob time-drawdown method using
Static potentiometric surface
Confining bed
Figure 56. Two Components of Total Drawdown.
35
-------
semilogarithmic graph paper. (See the "Time-Draw-
down Analysis" section of this chapter.) As Figure 57
shows, drawdowns are plotted on the arithmetic scale
versus time on the logarithmic scale and transmissivity
is determined from the slope of the straight line through
the use of the following equation:
T =
2.3 Q
4nAs
(32)
Where well loss is present in the pumping well, the
storage coefficient cannot be determined by extending
the straight line to the line of zero drawdown. Even
where well loss is not present, the determination of the
storage coefficient from drawdowns in a pumping well
likely will be subject to large error because the effective
radius of the well may differ significantly from the
"nominal" radius.
I 10
Time, in Minutes
Figure 57. Drawdown Plot
sa - Aquifer loss
sw- Well loss
Pumping Rate, in
Cubic Meters per Minute
Figure 58. Relation of Pumping Rate and Drawdown
36
-------
In equation 41, drawdown in the pumping well is
proportional to the pumping rate. The factor 6 in the
aquifer-loss term (80) increases with time of pumping
as long as water is being derived from storage in the
aquifer. The factor C in the well-loss term (CQ2) is a
constant if the characteristics of the well remain un-
changed, but, because the pumping rate in the well-loss
term is squared, drawdown due to well loss increases
rapidly as the pumping rate is increased. The relation
between pumping rates and drawdown in a pumping
well, if the well was pumped for the same length of
time at each rate, is shown in Figure 58. The effect of
well loss on drawdown in the pumping well is important
both in the analysis of data from pumping wells and in
the design of supply wells.
Well-Acceptance Tests and Well Efficiency
Many supply-well contracts require a "guaranteed"
yield, and some stipulate that the well reach a certain
level of "efficiency." Most contracts also specify the
length of the "drawdown test" that must be conducted
to demonstrate that the yield requirement is met. For
example, many States require that tests of public-supply
wells be at least 24 hours. Tests of most industrial and
irrigation wells probably do not exceed about 8 hours.
Well-acceptance tests, if properly conducted, not only
can confirm the yield of a well and the size of the pro-
duction pump that is needed but can also provide infor-
mation of great value in well operation and main-
0
2
4
6
o> 8
I i I i i
. Constont - rote
1
i n 2.l5m5mm~'
\ » - - n nei7
\ s< 8.4m
\
\
\
\
"Water-level '-•-. '^"''rnin
"measurements ~^ '" —
j i i I i
I
test
m3 min"' m~'
., • 8.4 m -
1
~ (a) Pumping Rate Held Constant
|
1
0
2
4
6
8
'
\ Step
\
r- '•*..
1 0 ~"
r-:04m'm,r
No. 2
-lA-.OSrfn
No 3
~4-r = 0.24m
7.5
l i i i
Multiple-step test
No. 1 (Each step = 8 hr)
'm ' x Step No 2
^
)n-im-i \StepNo. 3 (
•-.._ 1
3 mm"1 m"'
i i i i
1
-
si" ' *
/'
/
-
-
_
1
50 5 10 15 20
Hours
(b) Pumping Rate Increased in Equal Steps
25
30
Figure 59. Pumping Rates for Well-Acceptance Tests
tenance. Such tests should, therefore, be conducted with
the same care as aquifer tests made to determine the
hydraulic characteristics of aquifers. A properly con-
ducted test will include:
1. Determination of well interference from nearby
pumping wells, based on accurate water-level
measurements made before the drawdown test.
2. A pumping rate that is either held constant during
the entire test, as shown in Figure 59(a), or in-
creased in steps of equal length, as shown in
Figure 59(b). The pumping rate during each step
should be held constant, and the length of each
step should be at least 2 hours.
Of these requirements, the constant, carefully regulated
pumping rate or rates and the accurate water-level
measurements are the most important. When a con-
stant-rate well-acceptance test has been completed, the
drawdown data can be analyzed to determine the
aquifer transmissivity. (See the "Single-Well Tests" sec-
tion of this chapter.)
Many well-acceptance tests are made with temporary
pump installations, usually powered with a gasoline or
diesel engine. Instead of maintaining a constant rate for
the duration of the test, the engine is frequently stopped
to add fuel or to check the oil level or for numerous
other reasons. The rate may also be increased and
decreased on an irregular, unplanned schedule or, more
commonly, gradually reduced during the test in an ef-
fort to maintain a pumping level above the pump in-
take. In such tests, the "yield" of the well is normally
reported to be the final pumping rate.
Determining the long-term yield of a well from data
collected during a short-period well-acceptance test is
one of the most important, practical problems in
ground-water hydrology. Two of the most important
factors that must be considered are the extent to which
the yield will decrease if the well is pumped continuous-
ly for periods longer than the test period and the effect
on the yield of changes in the static (regional) water
level from that existing at the time of the test.
When data are available only from the production
well and when the pumping rate was not held constant
during the acceptance test, the estimate of the long-term
yield must usually be based on an analysis of specific-
capacity data. Specific capacity is the yield per unit of
drawdown and is determined by dividing the pumping
rate at any time during the test by the drawdown at the
same time. Thus,
specific capacity = Piping rate = _Q (42)
drawdown st
Before the development of steady-state conditions, a
part of the water pumped from an aquifer is derived
from storage. The time required for steady-state condi-
tions to develop depends largely on the distance to and
characteristics of the recharge and discharge areas and
the hydraulic characteristics of the aquifer. The time re-
quired to reach a steady state is independent of the
pumping rate. At some places in some aquifers, a
37
-------
steady-state condition will be reached in several days,
whereas, in others, six months to a year may be re-
quired; in some arid areas, a steady-state condition may
never be achieved. Depending on the length of the well-
acceptance test and the period required to reach a
steady-state condition, it may be appropriate, in esti-
mating the long-term yield of a well, to use a specific
capacity smaller than that determined during the test.
Figure 60 shows the decline in a specific capacity with
time when a well is pumped continuously at a constant
rate and all the water is derived from storage in an iso-
tropic and homogeneous acquifer. For convenience in
preparing Figure 60, a value of 100 percent was as-
signed to the specific capacity 1 hour after the pump
was started. The rate at which the specific capacity
decreases depends on the decline of the water level due
to depletion of storage and on the hydraulic character-
istics of the aquifer. Differences in the rate for different
aquifers are indicated by the width of the band on
Figure 60. When withdrawals are derived entirely from
storage, the specific capacity will decrease about 40 per-
cent during the first year.
100
Hours
1000
10,000
Figure 60. Decline in Specific Capacity With Time at a
Continuous Pumping Rate
In predicting the long-term yield of a well, it is also
necessary to consider changes in the static water level
resulting from seasonal and long-term variations in
recharge and declines due to other withdrawals from the
aquifer. The long-term yield is equal to the specific
capacity, determined from the well-acceptance test, and
reduced as necessary to compensate for the long-term
decline discussed in the above paragraph, multiplied by
the available drawdown.
The available drawdown at the time of a well-
acceptance test is equal to the difference between the
static water level at that time and the lowest pumping
level that can be imposed on the well. The lowest
pumping level in a screened well is normally considered
to be a meter or two above the top of the screen. In an
unscreened (open-hole) well, it may be at the level of
either the highest or the lowest water-bearing opening
penetrated by the well. The choice of the highest or the
lowest opening depends on the chemical composition of
the water and whether water cascading from openings
above the pumping level results in precipitation of
minerals on the side of the well and on the pump in-
take. If such precipitation is expected, the maximum
pumping level should not be below the highest opening.
The yield of a well is not increased by a pumping level
below the lowest opening, and the maximum yield may,
in fact, be attained at a much higher level.
To predict the maximum continuous long-term yield,
it is necessary to estimate how much the static water
level, and thus the available drawdown, may decline
from the position that it occupied during the acceptance
test. Records of water-level fluctuations in long-term
observation wells in the area will be useful in this effort.
Well efficiency is an important consideration both in
well design and in well construction and development.
The objective, of course, is to avoid excessive energy
costs by designing and constructing wells that will yield
the required water with the least drawdown.
Well efficiency can be defined as the ratio of the
drawdown (sa) in the aquifer at the radius of the
pumping well to the drawdown (S() inside the well. (See
the "Single-Well Tests" section of this chapter.) Thus,
the equation
E = ±£x100
(43)
expresses well efficiency as a percentage.
Drawdowns in pumping wells are measured during
well-acceptance tests. Determining the drawdown in the
aquifer is a much more difficult problem. It can be cal-
culated if the hydraulic characteristics of the aquifer, in-
cluding the effect of boundary conditions, are known.
The difference between st and sa is attributed to head
losses as water moves from an aquifer into a well and
up the well bore. These well losses can be reduced by
reducing the entrance velocity of the water, which can
be done by installing the maximum amount of screen
and pumping at the lowest acceptable rate. Tests have
been devised to determine well losses, and the results
can be used to determine well efficiency. However,
these tests are difficult to conduct and are not widely
used. Because of difficulties in determining sa, well ef-
ficiency is generally specified in terms of an "optimum"
specific capacity based on other producing wells in the
vicinity.
Under the best conditions, an efficiency of about 80
percent is the maximum that is normally achievable in
most screened wells. Under less than ideal conditions,
an efficiency of 60 percent is probably more realistic.
38
-------
Specific Capacity and Transmissivity
As Figure 61 shows, the specific capacity of a well
depends both on the hydraulic characteristics of the
aquifer and on the construction and other features of
the well. Values of specific capacity, available for many
supply wells for which aquifer-test data are not avail-
able, are widely used by hydrologists to estimate trans-
missivity. Such estimates are used to evaluate regional
differences in transmissivity and to prepare transmissiv-
ity maps for use in models of ground-water systems.
The factors that affect specific capacity include:
1. The transmissivity of the zone supplying water to
the well, which, depending on the length of the
screen or open hole, may be considerably less than
the transmissivity of the aquifer.
2. The storage coefficient of the aquifer.
3. The length of the pumping period.
4. The effective radius of the well, which may be
significantly greater than the "nominal" radius.
5. The pumping rate.
The Theis equation can be used to evaluate the effect
of the first four factors on specific capacity. The last
factor, pumping rate, affects the well loss and can be
determined only from a stepped-rate test or an aquifer
test in which drawdowns are measured in both the
pumping well and observation wells.
The Theis equation, modified for the determination
of transmissivity from specific capacity, is
W(u) x Q
4rr S
7 =
(44)
where T is transmissivity, Q/s is specific capacity, Q is
the pumping rate, s is the drawdown, and W(u) is the
well function of u, where
rfS
4Tt
u =
(45)
where r is the effective radius of the well, S is the
storage coefficient, and t is the length of the pumping
period preceding the determination of specific capacity.
For convenience in using equation 44, it is desirable
to express W(u)l4n as a constant. To do so, it is first
Land surface
Potentiometric surface
Cone ~ o7 ~-
^epres •>
'on ~~~ ^
^
*\
Confining bed
///////////Z/Z/Z/Zs
a. Thickness of the producing zone
compared to the length of the
screen or open hole
~~ — — ___
"~ --
Producing Length of
zone screen
<
\
/ / /
"****
o
o
o
o
o
o
0
o
o
0
\J
1
1
1
==
•
Hi
=p=
3i
= 1 —
(nonpumping)
'
/
-
Drawdown in
the aquifer^ -"
^
/
/
Well
loss
__ __ — ~
b. Magnitude of the
well loss compared
to the drawdown in
the aquifer
/ z / z / / /y zy zz zy_y zy z / z ,
c. The
"
difference between the
nominal" radius and the
effective radius
"Nominal"
/radius _ —
*~ -•
0
o
§
o
o
gj^Effective
o| radius
Confined
aquifer
Confining bed
Figure 61. Factors Affecting Estimates of Transmissivity Based on Specific Capacity
39
-------
necessary to determine values for u and, using a table
of values of u (or 1/u) and W(u), determine the cor-
responding values for W(u).
Values of u are determined by substituting in equa-
tion 45 values of 7, S, r, and t that are representative of
conditions in the area. To illustrate, assume, in an area
under investigation and for which a large number of
values of specific capacity are available, that:
1. The principal aquifer is confined, and aquifer tests
indicate that it has a storage coefficient of about
2x 10~4 and a transmissivity of about 11,000
ft2d-'.
2. Most supply wells are 8 in. (20 cm) in diameter
(radius, 0.33 ft).
3. Most values of specific capacity are based on
12-hour well-acceptance tests (t = 0.5 d).
Substituting these values in equation 45, we obtain
r2S (0.33 ft)2 x (2 x 10 ~4)
47? 4x(11,000ft2d-1)x0.5d
2.22 x 10 ~5 ft2
u =
u =
2.2 x 104 ft2
= 1.01 x10~9
A table of values of W(u) for values of Mu is con-
tained in the section of this chapter entitled "Aquifer
Tests." Therefore, the value of u determined above
must be converted to 1/u, which is 9.91 X 10s, and this
value is used to determine the value of W(u). Values of
W(u) are given for values of 1/u of 7.69x 108 and
10 x 108 but not for 9.91 x 108. However, the value of
10 is close enough to 9.91 for the purpose of estimating
transmissivity from specific capacity. From the table,
we determine that, for a value of 1/u of 10 x 10s, the
value of W(u) is 20.15. Substituting this value in equa-
tion 44, we find the constant W(u)l4-n to be 1.60.
Equation 44 is in consistent units. However, transmis-
sivity is commonly expressed in the United States in
units of square feet per day, pumping rates are reported
in units of gallons per minute, and drawdowns are
measured in feet. To obtain an equation that is conve-
nient to use, it is desirable to convert equation 44 to
these inconsistent units. Thus
7.48 gal s
7 = 308— or 300— (rounded)
S S
(46)
Many readers will find it useful at this point to
substitute different values of 7, S, r, and t in equation
45 to determine how different values affect the constant
in equation 46. In using equation 46, modified as neces-
sary to fit the conditions in an area, it is important to
recognize its limitations. Among the most important
factors that affect its use are the accuracy with which
the thickness of the zone supplying water to the well
can be estimated, the magnitude of the well loss in com-
parison with drawdown in the aquifer, and the differ-
ence between the "nominal" radius of the well and its
effective radius.
Relative to these factors, the common practice is to
assume that the value of transmissivity estimated from
specific capacity applies only to the screened zone or to
the open hole. To apply this value to the entire aquifer,
the transmissivity is divided by the length of the screen
or open hole (to determine the hydraulic conductivity
per unit of length), and the result is multiplied by the
entire thickness of the aquifer. The value of transmissiv-
ity determined by this method is too large if the zone
supplying water to the well is thicker than the length of
the screen or the open hole. Similarly, if the effective
radius of the well is larger than the "nominal" radius
(assuming that the "nominal" radius is used in equation
45), the transmissivity based on specific capacity again
will be too large.
On the other hand, if a significant part of the draw-
down in the pumping well is due to well loss, the trans-
missivity based on specific capacity will be too small.
Whether the effect of all three of these factors cancels
depends on the characteristics of both the aquifer and
the well. Where a sufficient number of aquifer tests
have been conducted, it may be feasible to utilize the
results to modify the constant in equation 46 to account
for the effect of these factors.
Quality of Ground Water
Water frequently is referred to as the universal solvent
because it has the ability to dissolve at least small
amounts of almost all substances that it contacts. Of
the domestic water used by man, ground water usually
contains the largest amounts of dissolved solids. As
Figure 62 illustrates, the composition and concentration
of substances dissolved in unpolluted ground water de-
pend on the chemical composition of precipitation, on
the biologic and chemical reactions occurring on the
land surface and in the soil zone, and on the mineral
composition of the aquifers and confining beds through
which the water moves.
The concentrations of substances dissolved in water
are commonly reported in units of weight per volume.
In the International System (SI), the most commonly
used units are milligrams per liter. A milligram equals
1/1,000 (0.001) of a gram, and a liter equals 1/1,000 of
a cubic meter, so that 1 mg/1 equals 1 gram m~3.2 Con-
centrations of substances in water were reported for
many years in the United States in units of weight per
weight. Because the concentration of most substances
dissolved in water is relatively small, the weight per
weight unit commonly used was parts per million
(ppm). In inch-pound units, 1 ppm is equal to 1 Ib of a
substance dissolved in 999,999 Ib of water, the weight
of the solution thus being 1 million pounds.
The quality of ground water depends both on the
substances dissolved in the water and on certain prop-
erties and characteristics that these substances impart to
2To put these units in possibly more understandable terms, 1 mg/1
equals 1 oz of a substance dissolved in 7,500 gal of water.
40
-------
the water. Table 6 contains information on dissolved in-
organic substances that normally occur in the largest
concentrations and are most likely to affect water use.
Table 7 lists other characteristics of water that are com-
monly reported in water analyses and that may affect
water use.
Atmosphere
Land Surface and soil zone
Shallow aquifers
Freshwater and saltwater
interfaces
Figure 62. The Chemical Characteristics of Ground Water are Determined by the Chemical and Biological Reactions in the
Zones Through Which the Water Moves
41
-------
Table 6. Natural Inorganic Constituents Commonly Dissolved in Water That are Most Likely to Affect Use of the Water
Substance
Major natural sources
Effect on water use
Concentrations of
significance (mg/iy
Bicarbonate (HCO3) and carbonate (CO3) . .
Calcium (Ca) and magnesium (Mg)
Chloride (Cl).
Fluoride (F)
Iron (Fe) and manganese (Mn)
Sodium (Na).
Sulfate(SO4).
Products of the solution of
carbonate rocks, mainly lime-
stone (CaCO3) and dolomite
(CaMgCO3), by water containing
carbon dioxide
Soils and rocks containing lime-
stone, dolomite, and gypsum
(CaSO^. Small amounts from
igneous and melamorphic rocks
In inland areas, primarily from
seawater trapped in sediments at
time of deposition, sition, in
coastal areas, from seawater in
contact with freshwater in produc-
tive aquifers.
Both sedimentary and igneous
rocks Not widespread in occur-
rence
Iron present in most soils and
rocks; manganese less widely
distributed.
Same as for chloride. In some
sedimentary rocks, a few hundred
milligrams per liter may occur in
freshwater as a result of ex-
change of dissolved calcium and
magnesium for sodium in the
aquifer materials
Gypsum, pynte (FeS), and other
rocks containing sulfur (S) com-
pounds
Control the capacity of water to
neutralize strong acids Bicar-
bonates of calcium and mag-
nesium decompose in steam
boilers and water heaters to form
scale and release corrosive car-
bon dioxide gas In combination
with calcium and magnesium,
cause carbonate hardness
Principal cause of hardness and
of boiler scale and deposits in
hot-water heaters.
in large amounts, increases corro-
siveness of water and, in com-
bination with sodium, gives water
a salty taste
In certain concentrations, reduces
tooth decay, at higher concentra-
tions, causes mottling of tooth
enamel
Stain laundry and are objec-
tionable in food processing, dye-
ing, bleaching, ice manufacturing,
brewing, and certain other indus-
trial processes
See chloride in large concentra-
tions, may affect persons with
cardiac difficulties, hypertension,
and certain other medical condi-
tions. Depending on the concen-
trations of calcium and mag-
nesium also present in the water,
sodium may be detrimental to cer-
tain irrigated crops
In certain concentrations, gives
water a bitter taste and, at higher
concentrations, has a laxative ef-
fect In combination with calcium,
forms a hard calcium carbonate
scale in steam boilers
150-200
25-50
250
07-1.22
Fe>03, Mn>0.05
69 (irrigation),
20-170 (health)3
300-400 (taste),
600-1,000 (laxative)
^A range in concentration is intended to indicate the general level at which the effect on water use might become significant
2Optimum range determined by the U S. Public Health Service, depending on water intake
3Lower concentration applies to dnnking water tor persons on a strict diet, higher concentration is for those on a moderate diet
Table 7. Characteristics of Water that Affect Water Quality
Characteristic
Principal cause
Significance
Remarks
Hardness ...
pH (or hydrogen-ion activity)
Specific electrical conductance. .
Total dissolved solids
Calcium and magnesium
dissolved in the water
Dissociation of water
molecules and of acids
and bases dissolved in
water
Substances that form ions
when dissolved in water.
Mineral substances
dissolved in water
Calcium and magnesium combine with
soap to form an insoluble precipitate
(curd) and thus hamper the formation of a
lather Hardness also affects the suit-
ability of water for use in the textile and
paper industries and certain others and in
steam boilers and water heaters
The pH of water is a measure of its reac-
tive characteristics Low values of pH,
particularly below pH 4, indicate a corro-
sive water that will tend to dissolve
metals and other substances that it con-
tacts. High values of pH, particularly
above pH 8 5, indicate an alkaline water
that, on heating, will tend to form scale
The pH significantly affects the treatment
and use of water
Most substances dissolved in water
dissociate into ions that can conduct an
electrical current. Consequently, specific
electrical conductance is a valuable indi-
cator of the amount of material dissolved
in water. The larger the conductance, the
more mineralized the water
Total dissolved solids is a measure of the
total amount of minerals dissolved in
water and is, therefore, a very useful
parameter in the evaluation of water qual-
ity. Water containing less than 500 mg/l is
preferred for domestic use and for many
industrial processes
USQS classification of hardness
(mg/l as CaGOs):
0-60- Soft
61-120: Moderately hard
121-180 Hard
More than 180 Very hard
pH values:
less than 7, water is acidic;
value of 7, water is neutral;
more than 7, water is basic.
Conductance values indicate the
electrical conductivity, in
micromhos, of 1 cm3 of water at
a temperature of 25°C
USGS classification of water
based on dissolved solids (mg/l):
Less than 1,000: Fresh
1,000-3,000: Slightly saline
3,000-10,000. Moderately saline
10,000-35,000- Very saline
More than 35,000: Briny
42
-------
Saltwater Encroachment
In coastal areas, fresh ground water derived from
precipitation on the land comes in contact with and dis-
charges into the sea or into estuaries containing brack-
ish water. The relation between the freshwater and the
seawater, or brackish water, is controlled primarily by
the differences in their densities.
The density of a substance is its mass per unit
volume; thus, the density of water is affected by the
amount of minerals, such as common salt (NaCl), that
the water contains in solution. In metric units, the den-
sity of freshwater is about 1 gm cm~3, and the density
of seawater is about 1.025 gm cm"3. Thus, freshwater,
being less dense than seawater, tends to override or
float on seawater.
isotropic aquifer in which the freshwater is static and is
in contact with a tideless sea or body of brackish water.
Figure 63 shows that tides cause saltwater to alter-
nately invade and retreat from the freshwater zone, the
result being a zone of diffusion across which the salinity
changes from that of freshwater to that of seawater. A
part of the seawater that invades the freshwater zone is
entrained in the freshwater and is flushed back to the
sea by the freshwater as it moves to the sea to
discharge.
Because both the seawater and the freshwater are in
motion (not static), the thickness of the freshwater zone
in a homogenous and isotropic aquifer is greater than
that predicted by the Ghyben-Herzberg equation. On
the other hand, in a stratified aquifer (and nearly all
aquifers are stratified), the thickness of the freshwater
aps?:;m\ i i1 I I ~/£
5f||##XX \ Freshwoter //£
Pumpi n g well
Land surface
Unconf med
aquifer
Confined
aquifer
Water;
Freshw ot e r
— Confming^bed ^--r^
_
Lateral "*/.'•'••.'•:':
encroachment /?:"!'• :'•'••:•
Figure 63. Freshwater Lens Floating on Saltwater
Figure 64. Two Aspects of Saltwater Encroachment
On islands, such as the Outer Banks of North
Carolina, precipitation forms a freshwater lens that
"floats" on the underlying saltwater, as illustrated in
Figure 63. The higher the water table stands above sea
level, the thicker the freshwater lens. This relation be-
tween the height of the water table and the thickness of
the freshwater lens was discovered, independently, by a
Dutchman, Badon Ghyben, and a German, B. Herz-
berg, and is referred to as the Ghyben-Herzberg rela-
tionship. This relation, expressed as an equation, is
(47)
where hs is the depth of freshwater below sea level, ef is
the density of freshwater, QS is the density of seawater,
and hf is the height of the water table above sea level.
On the basis of equation 47 and the differences be-
tween the densities of freshwater and seawater, the
freshwater zone should extend to a depth below sea
level (hs) equal to 40 times the height of the water
table above sea level (hf). The Ghyben-Herzberg relation
applies strictly, however, only to a homogenous and
lens is less than that predicted because of the head loss
incurred as the freshwater moves across the least
permeable beds.
When freshwater heads are lowered by withdrawals
through wells, the freshwater-saltwater contact migrates
toward the point of withdrawals until a new balance is
established, as shown in Figure 64. The movement of
saltwater into zones previously occupied by freshwater
is referred to as saltwater encroachment.
Saltwater encroachment is a serious problem in some
coastal areas. Upconing of salty water beneath pumping
wells is a more imminent problem than lateral encroach-
ment in most areas. One reason is that lateral encroach-
ment must displace a volume of freshwater much larger
than that displaced by upconing. As Figure 65 shows,
another reason is that approximately two-thirds of the
United States is underlain by aquifers that yield water
containing more than 1,000 mg/1 of total dissolved
solids. (See Table 7 in the "Quality of Ground Water"
section of this chapter.) In most places, these aquifers
are overlain by other aquifers that contain freshwater
and that serve as sources of water supply. However,
43
-------
where supply wells are drilled too deeply or are pumped
at too large a rate, upconing of the mineralized (salty)
water may occur.
In the design of supply wells in areas underlain by or
adjacent to salty water, consideration must be given to
the possibility of saltwater encroachment. This con-
sideration may involve selection of shallow aquifers or
small pumping rates to avoid upconing or involve mov-
ing wells to more inland locations to avoid lateral
encroachment.
EXPLANATION
Depth to saline ground
water in meters
Less than 150
150 to 300
More than 300
Not present
<150
Todd, Groundwater Hydrology, 2nd Ed., 1980
Approximate
0 700 40(J bOO Miles
l-r^-rW-—-'1
o mo 4oo 600 Kilometers
Figure 65. Depth to Ground Water Containing More Than 1000 mg/l of Total Dissolved Solids in the
Conterminous United States
Temperature of Ground Water
The temperature of ground water is one of its most
useful characteristics. Ground water has been used for
many years on Long Island, N.Y., and at other places
as a heat-exchange medium for air-conditioning sys-
tems. As a result of recent increases in energy costs,
ground water is also now becoming increasingly impor-
tant as a source of heat for "heat pumps."
The temperature of ground water responds to sea-
sonal variations in the heat received at the Earth's sur-
face from the Sun and by movement of heat from the
Earth's interior. Figure 66 shows that the seasonal
movement of heat into and out of the upper layers of
the Earth's crust causes a seasonal fluctuation in
ground-water temperatures to a depth of 10 to 25 m.
The fluctuation is greatest near the surface, amounting
to 5° to 10°C at depths of a few to several meters. In
the zone affected by seasonal fluctuations, the mean
annual ground-water temperature is 1 ° to 2°C higher
than the mean annual air temperature. Consequently, a
map showing the mean annual temperature of shallow
ground water can be prepared on the basis of mean
annual air temperature, as Figure 67 illustrates (based
on a map showing mean annual air temperature pre-
pared by the National Weather Service).
Figure 66 shows that movement of heat from the
Earth's interior causes ground-water temperatures to in-
crease with depth. This increase is referred to as the
geothermal gradient and ranges from about 1.8°C per
100 m in areas underlain by thick sections of sedi-
mentary rocks to about 3.6°C per 100 m in areas of re-
cent volcanic activity. The effect of the geothermal
44
-------
Degrees Celsius
-6-4-2 0 2 4 6 8
n i i i i i i i i i i i i i
U
2? 25
aj
a5
c
a>
o
£
| 50
S
0
3
i.
S 75
inn
Mean annual v/y/'fj Seasonal
air temperature Yy/i fluctuation
0
\
\o
\ <*
\ °
\ »
ll
I-
1 O
1 5.
1
1
1
I
1
1
1 1 1 I 1 ! 1 1 1 1 1 1 1
Figure 66. Changes in Ground-Water Temperature With Depth
gradient is not readily apparent in the zone affected by
seasonal temperature fluctuations.
Movement of ground water causes a distortion in iso-
therms (lines depicting equal temperatures). This effect
is most noticeable where ground-water withdrawal in-
duces a movement of water from a stream into an
aquifer. The distortion in ground-water temperature is
most pronounced in the more permeable zones of the
aquifer.
Protection of Supply Wells
Most, if not all, States have laws related to the loca-
tion and construction of public-supply wells. Figure 68
shows typical requirements for supply wells. These laws
and the rules and regulations developed for their admin-
istration and enforcement are concerned, among other
things, with protecting supply wells from pollution.
Pollution of the environment results from man's activ-
ities, and, consequently, except where deep wells or
mines are used for waste disposal, it primarily affects
the land surface, the soil zone, and the upper part of
the saturated (ground water) zone. Therefore, the pro-
tection of supply wells includes avoiding areas that are
presently polluted and sealing the wells in such a way as
to prevent pollution in the future.
Fortunately, most ground-water pollution at the pres-
ent time affects only relatively small areas that can be
readily avoided in the selection of well sites. Among the
areas in which at least shallow ground-water pollution
should be expected are:
1. Industrial districts that include chemical, metal-
Figure 67. Approximate Temperature of Ground Water, in Degrees Celsius, in the Conterminous United States at Depths of 10 to 25 m
45
-------
Cosing I m'+'above
pump pedestal
Concrete slab or wellhouse floor
3 ft from well and 4 m(+lin thickness
Land surface sloped
away from well
(-H A plus sign in parentheses
means distance or thickness
can be greater but not less
Figure 68. Typical Requirements for Supply Wells
working, petroleum-refining, and other industries
that involve fluids other than cooling water.
2. Residential areas in which domestic wastes are
disposed of through septic tanks and cesspools.
3. Animal feedlots and other areas in which large
numbers of animals are kept in close confinement.
4. Liquid and solid waste disposal sites, including
sanitary landfills, "evaporation ponds," sewage
lagoons, and sites used for the disposal of sewage-
plant effluent and solid wastes.
5. Chemical stockpiles, including those for salt used to
deice streets and highways and for other chemical
substances soluable in water.
In the selection of a well site, areas that should be
avoided include not only those listed but also the zones
surrounding them that may be polluted by movement of
wastes in response to both the natural hydraulic gra-
dient and the artificial gradient that will be developed
by the supply well.
Rules and regulations intended to prevent future
pollution include provision of "exclusion" zones
around supply wells, requirements for casing and for
sealing of the annular space, and sealing of the upper
end of the wells.
Many State regulations require that supply wells be
located at least 100 ft (30 m) from any sources or
potential sources of pollution. In the case of public-
supply wells, the well owner must either own or control
the land within 100 ft (30 m) of the well. In some
States, a public-supply well may be located as close as
50 ft (15 m) to a sewer if the joints in the sewerline
meet water-main standards.
Some State regulations require that all supply wells be
cased to a depth of at least 20 ft (6 m) and that the an-
nular space between the land surface and a depth of 20
ft (6 m) be completely filled with cement grout. The
casing of supply wells drawing water from fractured
bedrock must be seated and sealed into the top of the
rock.
Most regulations require that the casing of all supply
wells terminate above land surface and that the land
surface at the site be graded or sloped so that surface
water is diverted away from the well. Many States also
require that public-supply wells have a continuous-bond
concrete slab or concrete wellhouse floor at least 4 in.
(10 cm) thick and extending at least 3 ft (1 m) horizon-
tally around the outside of the well casing. The top of
the well casing must project not less than 6 in. (15 cm)
above the concrete slab or wellhouse floor. The top of
the well casing must also project at least 1 in. (2.5 cm)
above the pump pedestal. The top of the well casing
must be sealed watertight except for a vent pipe or vent
tube having a downward-diverted screened opening.
The regulations cited above provide, at best, only
minimal protection for supply wells. There are numer-
ous situations in which both the size of the exclusion
zone and the depth of casing are inadequate. Relative to
the radius of the exclusion zone, there are no arbitrary
limits, except the physical boundaries of an aquifer,
past which ground water cannot move. Relative to the
46
-------
minimum required casing, there are no vertical limits,
except for the impermeable base of the ground-water
system, past which polluted water cannot move.
On the other hand, there are geologic and hydrologic
situations in which these regulations may be unnecessari-
ly restrictive. An example is pollution in an unconfmed
aquifer down the hydraulic gradient from a supply well
drawing from a deep confined aquifer overlain by a
nonleaky confining bed.
Because of these factors, it is essential that officials
involved in regulating the location and construction of
supply wells be adequately trained in the fields of
ground-water geology and hydrology so that they can
protect the public health on the basis of scientific
knowledge and technical judgment rather than on
blind application of arbitrary regulations.
Well Records and Files
The collection and preservation of records on the
construction, operation, maintenance, and abandon-
ment of supply wells are an essential but largely
neglected activity. This responsibility rests largely on
the well owner or operator. The consequence of this
neglect is that it is not possible to identify and to
economically correct problems of declining yield or
deterioration in water quality, and the design of new
wells cannot incorporate past operational experience.
A file should be established on each supply well at
the time when plans for its construction are initiated.
From the initial planning to the final abandonment of
the well, the following records should be generated
and carefully preserved in this file:
1. Initial design, including drawings or written
specifications on diameter, proposed total depth,
position of screens or open hole, method of con-
struction, and materials to be used in construc-
tion.
2. Construction record, including the method of con-
struction and the driller's log and geophysical log
of the materials penetrated during construction,
the diameter of casings and screens, the slot size
and metallic composition of screens, the depths
of casing and screens, the total depth of the well,
and the weight of the casing. Records and logs
should also be retained for all test wells, in-
cluding those that were not successful because of
small yields.
3. Well-acceptance test, including a copy of the
water-level measurements made before, during,
and after the drawdown (pumping) test, a record
of the pumping rate or rates, copies of any
graphs of the data, and a copy of the hydrolo-
gist's report on the interpretation of the test
results. (See the "Well-Acceptance Tests and Well
Efficiency" section of this chapter.)
4. Pump and installation data, including the type of
pump, the horsepower of the motor, the depth to
the pump intake, a copy of the pump manufac-
turer's performance and efficiency data, and data
on the length of the air line or a description of
facilities provided for water-level measurements,
including a description of the measuring point.
5. Operating record, including data on the type of
meter used to measure the flow rate, weekly read-
ings of the flowmeter dial, weekly measurements
of the static and pumping water levels, and peri-
odic analyses of water quality.
6. Record of well maintenance, including the dates and
the activities instituted to increase the yield or to
improve the water quality and data showing the
results achieved.
7. Record of well abandonment, including the date that
use of the well was discontinued and description
of the methods and materials used to seal or plug
the well.
The type of forms used for the records described
above is not of critical importance. It is more impor-
tant that the records be collected, regardless of the
type of form that is used. It is important, however,
that the date and the watch time be noted with each
measurement of pumping rate and depth to water and
on each water sample collected for water-quality
analyses.
47
-------
Numbers, Equations, and Conversions
The preceding discussions of basic ground-water hydrology involve the use of equations and physical units with
which some readers may not be familiar. This discussion of numbers, equations,and conversion of units from one
system of measurement to another is included for the benefit of those readers and for others who need to refresh
their memories.
Expressing Large Numbers
1,000 = 10x10x10 = 1x103
1,000,000 = 10x10x10x10x10x10 = 1 x 106
The numbers 3 and 6 are called exponents and indicate the number of times that 10 must be multiplied by itself to
obtain the initial number.
Expressing Small Numbers
0.001 = —I— = —-— = 1 x 10~3
1,000 1x103
0.000001 = 1 = 1 = 1 x10~6
1,000,000 1x106
Exponents in the denominator acquire a negative sign when they are moved to the numerator.
Simplifying Equations
Symbols in equations have numerical values and, in most cases, units of measurement, such as meters and feet, in
which the values are expressed. For example, Darcy's law, one of the equations used in basic ground-water
hydrology, is
Q = KA
In metric units, hydraulic conductivity (K) is in meters per day, area (A) is in square meters, and hydraulic gradient
(dh/dl) is in meters per meter. Substituting these units in Darcy's law, we obtain
Q= meters Xmeters2)< meters = meters* = m4-i d-i = m3d-i
day meters meters day
Similarly, in inch-pound units, K is in feet per day, A is in square feet, and dh/dl is in feet per feet. Substituting
these units in Darcy's law, we obtain
Q = 2 fee*4
,
day feet feet day
The characteristics of exponents are the same, whether they are used with numbers or with units of measurement.
Exponents assigned to units of measurement are understood to apply, of course, to the value that the unit of
measurement has in a specific problem.
Conversion of Units
Units of measurements used in ground-water literature are gradually changing from the inch-pound units of
gallons, feet, and pounds to the International System of units of meters and kilograms (metric units). It is, therefore,
increasingly important that those who use this literature become proficient in converting units of measurement from
one system to another. Most conversions involve the fundamental principle that the numerator and denominator of a
fraction can be multiplied by the same number (in essence, multiplying the fraction by 1) without changing the value
of the fraction. For example, if both the numerator and the denominator of the fraction V4 are multiplied by 2, the
value of the fraction is not changed. Thus,
1X2 _ 2 _ 1 1 2 _ 1 1
_X-----orTx--Tx1 --
Similarly, to convert gallons per minute to other units of measurement, such as cubic feet per day, we must first
identify fractions that contain both the units of time (minutes and days) and the units of volume (gallons and cubic
feet) and that, when they are used as multipliers, do not change the numerical value. Relative to time, there are 1,440
minutes in a day. Therefore, if any number is multiplied by 1,440 min/d, the result will be in different units, but its
48
-------
numerical value will be unchanged. Relative to volume, there are 7.48 gallons in a cubic foot. Therefore, to convert
gallons per minute to cubic feet per day, we multiply by these "unit" fractions, cancel the units of measurement that
appear in both the numerator and the denominator, and gather together the units that remain. In other words, to
convert gallons per minute to cubic feet per day, we have
gallons _ gallons 1,44Qmin cubic feet
minute ~~ minute d 7.48 gal
and, canceling gallons and minutes in the numerators and denominators, we obtain
gallons =J,440ft3_ = 1925ft3c|_1
minute 7.48 d
which tells us that 1 gal min"1 equals 192.5 ft3 d~'.
We follow the same procedure in converting from inch-pound units to metric units. For example, to convert
square feet per day to square meters per day, we proceed as follows:
If = If X——— = m2 = 0.0929 m2d~1 = 9.29x10~
d d 10.76ft2 10.76 d
Table 8. Relation of Units of Hydraulic Conductivity, Transmissivity, Recharge Rates, and Flow Rates
Hydraulic Conductivity (K)
Meters per day Centimeters per second Feet per day
(md-1) (cms-1) (ftd-1)
1 1.16X10-3 3.28
8.64x102 1 2.83 x103
3.05x10-1 3.53X10-4 1
4.1 x10-2 4.73x10~5 1.34X10-1
Gallons per day
per square foot
(gal d~1 ft-2)
2.45 x101
2.12x10"
7.48
1
Transmissivity (T)
Square meters per day Square feet per day
(m2d-1) (ft2d-1)
1 10.76
.0929 1
.0124 .134
Gallons per day
per foot
(gal d~1 ft-1)
80.5
7.48
1
Recharge Rates
Unit depth Volume
peryear (m2d-1km-2) (ft2d-1mi-2)
(In millimeters) 2.7 251
(In inches) 70 6,365
(gal d~1 mi"2)
1,874
47,748
Flow Rates
(m3s-1) (m3min-1) (ft3s-1) (ft3min-1)
1 60. 35.3 2,120
.0167 1 .588 35.3
.0283 1.70 1 60
.000472 .0283 .0167 1
.000063 .00379 .0023 .134
(gal min~1)
15,800
264
449
7.48
1
49
-------
Bibliography
A large number of publications on ground-water hydrology were consulted in the preparation of this report. A
citation is shown in the text only where a publication was used as a specific source of tabular data.
The following list of principal references consulted is included to identify sources of specific information and for
the benefit of those who wish to obtain additional information.
General Bibliography
Bouwer, H., Groundwater Hydrology. McGraw-Hill,
New York, 1978.
Fetter, C. W., Jr., Applied Hydrogeology. Charles E.
Merrill, Columbus, 1980.
Freeze, R. A., and J. A. Cherry, Groundwater. Prentice
Hall, Englewood Cliffs, N.J., 1979.
Heath, R. C., and F. W. Trainer, Introduction to
Ground- Water Hydrology. Water-Well Journal
Publishing Co., Worthington, Ohio, 1981.
Todd, D. K., Groundwater Hydrology, 2d ed. John
Wiley, New York, 1980.
Walton, W. C., Groundwater Resource Evaluation.
McGraw-Hill, New York, 1970.
Section Bibliographies
A few publications were consulted in the preparation
of two or more sections. To save space, the complete
citation to a publication is shown only the first time
that it is mentioned.
Ground-Water Hydrology
L'vovich, M. I., World Water Resources and Their
Future (English translation, edited by R. L. Nace).
American Geophysical Union, Washington, D.C.,
1979.
Stratification and Unsaturated Flow
Palmquist, W. N., Jr., and A. I. Johnson, "Vadose
Flow in Layered and Nonlayered Materials," in Short
Papers in Geology and Hydrology. U.S. Geological
Survey Professional Paper 450-C, 1962.
Saturated Flow and Dispersion
Danel, P., "The Measurement of Ground-Water Flow,"
in Ankara Symposium on Arid Zone Hydrology, Paris
1953, Proceedings. UNESCO, pp 99-107, 1953.
Source of Water Derived from Wells
Theis, C. V., "The Source of Water Derived from
Wells, Essential Factors Controlling the Response of
an Aquifer to Development." Civil Engineering, v. 10
no. 5, pp 277-280, 1940.
Aquifer Tests
Stallman, R. W., "Aquifer-Test Design, Observations,
and Data Analysis." U.S. Geological Survey Tech-
niques of Water-Resources Investigations, Book 3,
Chapter Bl, 1971.
Underground Water
Meinzer, O. E., "The Occurrence of Ground Water in
the United States, With a Discussion of Principles."
U.S. Geological Survey Water-Supply Paper 489,
1923.
Hydrologic Cycle
L'vovich (1979)
Porosity
Meinzer (1923)
Specific Yield and Specific Retention
Meinzer (1923)
Hydraulic Conductivity
Lohman, S. W., et al, "Definitions of Selected Ground-
Water Terms — Revisions and Conceptual Refine-
ments." U.S. Geological Survey Water-Supply Paper
1988, 1972.
Analysis of Aquifer-Test Data
Jacob, C. E., "Determining the Permeability of Water-
Table Aquifers." U.S. Geological Survey Water-
Supply Paper 1536-1, pp 1245-1271, 1963.
Lohman, S. W., "Ground-Water Hydraulics." U.S.
Geological Survey Professional Paper 708, 1972.
Theis, C. V., "The Relation Between the Lowering of
the Piezo-Metric Surface and the Rate and Duration
of Discharge of a Well Using Ground-Water Storage."
Transactions of the American Geophysical Union, v.
16, pp 519-524, 1935.
Time-Drawdown Analysis
Jacob, C. E., "Flow of Ground-Water," in Rouse,
Hunter, Engineering Hydraulics. John Wiley, New
York, chapter 5, pp 321-386, 1950.
Distance-Drawdown Analysis
Jacob (1950)
50
-------
Quality of Ground Water Saltwater Encroachment
Hem, J. D., "Study and Interpretation of the Chemical Feth, J. H., et al, "Preliminary Map of the Conter-
Characteristics of Natural Water." U.S. Geological minous United States Showing Depth to and Quality
Survey Water-Supply Paper 1473, 1970. of Shallowest Ground Water Containing More Than
U.S. Environmental Protection Agency, "National 1,000 Parts per Million Dissolved Solids." U.S.
Interim Primary Drinking Water Regulations." Geological Survey Hydrologic Investigations Atlas 199,
EPA-570/9-76-003, 1977. scale 1:3, 168,000, two sheets, accompanied by 31 p.
text, 1965.
51
-------
Chapter 3
Classification of Ground-Water Regions
To describe concisely ground-water conditions in the
United States, it is necessary to divide the country into
regions in which these conditions are generally similar.
Because the presence and availability of ground water
depends primarily on geologic conditions, ground-water
regions also are areas in which the composition, ar-
rangement, and structure of rock units are similar.1
To divide the country into ground-water regions, it is
necessary to develop a classification that identifies
features of ground-water systems that affect the occur-
rence and availability of ground water. The five
features pertinent to such a classification are: (1) the
components of the system and their arrangement, (2)
the nature of the water-bearing openings of the domi-
nant aquifer or aquifers with respect to whether they
are of primary or secondary origin, (3) the mineral
composition of the rock matrix of the dominant
aquifers with respect to whether it is soluble or insolu-
ble, (4) the water storage and transmission
characteristics of the dominant aquifer or aquifers, and
(5) the nature and location of recharge and discharge
areas.
The first two of these features are primary criteria
used in all delineations of ground-water regions. The
remaining three are secondary criteria that are useful in
subdividing what might otherwise be large and un-
wieldy regions into areas that are more homogenous
and, therefore, more convenient for descriptive pur-
poses. Table 9 lists each of the five features together
with explanatory information. The fact that most of
the features are more or less interrelated is readily ap-
parent from the comments in the column headed
"Significance of Feature".
Ground-Water Regions of the United States
On the basis of the criteria listed above the United
States, exclusive of Alaska and Hawaii can be divided
into 11 ground-water regions.
Figure 69 shows the boundaries of these 11 regions.
A special area, region 12, which consists of those
segments of the valleys of perennial streams that are
underlain by sand and gravel thick enough to be
hydrologically significant (thicknesses generally more
than about 8 meters), is shown in Figure 70.
The nature and extent of the dominant aquifers and
their relations to other units of the ground-water
system are the primary criteria used in delineating the
regions. Consequently, the boundaries of the regions
generally coincide with major geologic boundaries and
at most places do not coincide with drainage divides.
Although this lack of coincidence emphasizes that the
physical characteristics of ground-water systems and
stream systems are controlled by different factors, it
does not mean that the two systems are not related.
Ground-water systems and stream systems are intimate-
ly related, as shown in the following discussions of
each of the ground-water regions.
1. Western Mountain Ranges
(Mountains with thin soils over fractured rocks,
alternating with narrow alluvial and, in part,
glaciated valleys)
The Western Mountain Ranges, shown in Figure 71,
encompass three areas totaling 708,000 km. The largest
area extends in an arc from the Sierra Nevada in
California, north through the Coast Ranges and
Cascade Mountains in Oregon and Washington, and
east and south through the Rocky Mountains in Idaho
and Montana into the Bighorn Mountains in Wyoming
and the Wasatch and Uinta Mountains in Utah. The
second area includes the southern Rocky Mountains,
which extend from the Laramie Range in southeastern
Wyoming through central Colorado into the Sangre de
Cristo Range in northern New Mexico. The smallest
area includes the part of the Black Hills of South
Dakota in which Precambrian rocks are exposed.
As would be expected in such a large region, both
the origin of the mountains and the rocks that form
them are complex. Most of the mountain ranges are
underlain by granitic and metamorphic rocks flanked
by consolidated sedimentary rocks of Paleozoic to
Cenozoic age. The other ranges, including the San
Juan Mountains in southwestern Colorado and the
Cascade Mountains in Washington and Oregon, are
underlain by lavas and other igneous rocks.
The summits and slopes of most of the mountains
consist of bedrock exposures or of bedrock covered by
a layer of boulders and other rock fragments produced
by frost action and other weathering processes acting
53
-------
Table 9. Features of Ground-Water Systems Useful in the Delineation of Ground-Water Regions
Feature
Aspect
Range In Conditions
Significance of Feature
Component of the
system
Unconfined aquifer
Confining beds
Confined aquifers
Presence and
arrangements of
components
Thin, discontinuous, hydrologically
insignificant.
Minior aquifer, serves primarily as a storage
reservoir and recharge conduit for under-
lying aquifer.
The dominant aquifer.
Not present, or hydrologically insignificant.
Thin, markedly discontinuous, or very leaky.
Thick, extensive, and impermeable.
Complexly interbedded with aquifers or
productive zones.
Not present, or hydrologically insignificant.
Thin or not highly productive.
Multiple thin aquifers interbedded with
nonproductive zones.
The dominant aquifer—thick and productive.
A single, unconfined aquifer.
Two interconnected aquifers of essentially
equal hydrologic importance.
A three-unit system consisting of an
unconfined aquifer, a confining bed, and
confined aquifer.
A complexly interbedded sequence of
aquifers and confining beds.
Affect response of the system to
pumpage and other stresses.
Affect recharge and discharge
conditions. Determine suscept-
ibility to pollution.
Water-bearing
openings of
dominant aquifer
Primary openings
Secondary openings
Pores in unconsolidated deposits.
Pores in semiconsolidated rocks.
Pores, tubes, and cooling fractures in
volcanic (extrusive-igneous) rocks.
Fractures and faults in crystalline and
consolidated sedimentary rocks.
Solution-enlarged openings in limestones
and other soluble rocks.
Control water-storage and trans-
mission characteristics. Affect
disperson and dilution of
wastes.
Composition of rock
matrix of
dominant aquifer
Insoluble
Soluble
Essentially insoluble.
Both relatively insoluble and soluble
constituents.
Relatively soluble.
Affects water-storage and trans-
mission characteristics. Has
major influence on water
quality.
Storage and
transmission
characteristics of
dominant aquifer
Porosity
Transmissivity
Large, as in well-sorted, unconsolidated
deposits.
Moderate, as in poorly-sorted unconsolidated
deposits and semiconsolidated rocks.
Small, as in fractured crystalline and
consolidated sedimentary rocks.
Large, as in cavernous limestones, some
lava flows, and clean gravels.
Moderate, as in well-sorted, coarse-grained
sands, and semiconsolidated limestones.
Small, as in poorly-sorted, fine-grained
deposits and most fractured rocks.
Very small, as in confining beds.
Control response to pumpage and
other stresses. Determine yield
of wells. Affect long-term yield
of system. Affect rate at which
pollutants move.
Recharge and
discharge
conditions of
dominant aquifer
Recharge
Discharge
In upland areas between streams, particu-
larly in humid regions.
Through channels of losing streams.
Largely or entirely by leakage across
confining beds from adjacent aquifers.
Through springs or by seepage to stream
channels, estuaries, or the ocean.
By evaporation on flood plains and in basin
"sinks."
By seepage across confining beds into
adjacent aquifers.
Affect response to stress and
long-term yields. Determine
susceptibility to pollution.
Affect water quality.
54
-------
2. Alluvial Basin
Glaciated
Central
region
Nonglaciated
Central
region
15. PUERTO RICO
AND
VIRGIN ISLANDS
800 Kilometers
Figure 69. Ground-Water Regions Used in This Report [The Alluvial Valleys region (region 12) is shown on figure 70)
*}W
^"{^•~
a 1 V 0
Figure 70. Alluvial Valleys Ground-Water Region
55
-------
CANADA
UNITED STATES
WASHINGTON
OREGON
San Juan Mountains
COLO.
3 100
1
1
200
1
300
1
1
400 5C
1
1 1 1
500 Miles
0 100 200 300 400 500 600 700 800 Kilometers
Figure 71. Western Mountain Ranges Region
56
-------
on the bedrock. This layer is generally only a few
meters thick on the upper slopes but forms a relatively
thick apron along the base of the mountains. The nar-
row valleys are underlain by relatively thin, coarse,
bouldery alluvium washed from the higher slopes. The
large synclinal valleys and those that occupy
downfaulted structural throughs are underlain by
moderately thick deposits of coarse-grained alluvium
transported by streams from the adjacent mountains,
as shown in Figure 72.
The Western Mountain Ranges and the mountain
ranges in adjacent regions are the principal sources of
water supplies developed at lower altitudes in the
western half of the conterminous United States. As
McGuinness2 noted, the mountains of the West are
moist "islands" in a sea of desert or semidesert that
covers the western half of the Nation. The heaviest
precipitation falls on the western slopes; thus, these
slopes are the major source of runoff and are also the
most densely vegetated. Much of the precipitation falls
as snow during the winter.
The Western Mountain Ranges are sparsely
populated and have relatively small water needs. The
region is an exporter of water to adjacent "have-not"
areas. Numerous surface reservoirs have been con-
structed in the region. Many such impoundments have
been developed on streams that drain the western flank
of the Sierra Nevada in California and the Rocky
Mountains in Colorado.
Melting snow and rainfall at the higher altitudes in
the region provide abundant water for ground-water
recharge. However, the thin soils and bedrock fractures
in areas underlain by crystalline rocks fill quickly, and
the remaining water runs off overland to streams.
Because of their small storage capacity, the
underground openings provide limited base runoff to
the streams, which at the higher altitudes flow only
during rains or snowmelt periods. Thus, at the higher
altitudes in this region underlain by crystalline rocks,
relatively little opportunity exists for development of
ground-water supplies. The best opportunities exist in
the valleys that contain at least moderate thicknesses of
saturated alluvium or in areas underlain by permeable
sedimentary or volcanic rocks. Ground-water supplies
in the valleys are obtained both from wells drawing
from the alluvium and from wells drawing from the
underlying rocks. The yields of wells in crystalline
bedrock and from small, thin deposits of alluvium are
generally adequate only for domestic and stock needs.
Large yields can be obtained from the alluvial deposits
that overlie the major lowlands and from wells com-
pleted in permeable sedimentary or volcanic rocks.
Consolidated
sedimentary rocks^JX
Water-bearing
fractures
Alluvial
deposits
Granitic and metamorphic
rocks
Figure 72. Topographic and Geologic Features in the Southern Rocky Mountains Part of the Western Mountain Ranges Region
57
-------
2. Alluvial Basins
(Thick alluvial deposits in basins and valleys
bordered by mountains and locally of glacial origin)
The Alluvial Basins region occupies a discontinuous
area of 1,025,000 km2 extending from the Puget
Sound-Williamette Valley area of Washington and
Oregon to west Texas. The region consists of an ir-
regular alternation of basins or valleys and mountain
ranges. In the Alluvial Basins region, basins and valleys
are the dominant feature. The principal exception is the
Coast Ranges of southern California which
topographically more closely resemble the Western
Mountain Ranges.
Most of the Nevada and all of the Utah parts of this
region are an area of internal drainage referred to as
the Great Basin. No surface or subsurface flow leaves
this part of the region and all water reaching it from
adjacent areas and from precipitation is evaporated or
transpired.
The basins and valleys range from about 85 m below
sea level in Death Valley in California to 2,000 m
above sea level in the San Luis Valley in Colorado. The
basins range in size from a few hundred meters in
width and a kilometer or two in length to, for the Cen-
tral Valley of California, as much as 80 km in width
and 650 km in length. The crests of the mountains are
commonly 1,000 to 1,500 m above the adjacent valley
floors.
The surrounding mountains, and the bedrock
beneath the basins, consist of granite and metamorphic
rocks of Precambrian to Tertiary age and consolidated
sedimentary rocks of Paleozoic to Cenozoic age. The
rocks are broken along fractures and faults that may
serve as water-bearing openings. However, the open-
ings in the granitic and metamorphic rocks in the
mountainous areas have a relatively small capacity to
store and to transmit ground water.
The dominant element in the hydrology of the region
is the thick (several hundred to several thousand
meters) layer of generally unconsolidated alluvial
material that partially fills the basins. Figures 73, 74,
and 75 illustrate this dominant element. Generally, the
coarsest material occurs adjacent to the mountains; the
material gets progressively finer toward the centers of
the basins. However, as Figure 74 shows, in most
alluvial fans there are layers of sand and gravel that ex-
tend into the central parts of the basins. In time, the
fans formed by adjacent streams coalesced to form a
continuous and thick deposit of alluvium that slopes
gently from the mountains toward the center of the
basins. These alluvial-fan deposits are overlain by or
grade into fine-grained flood plain, lake, or playa
deposits in the central part of most basins. The fine-
grained deposits are especially suited to large-scale
cultivation.
The Puget Sound and Williamette Valley areas differ
geologically from the remainder of the region. The
Puget Sound area is underlain by thick and very
permeable deposits of gravel and sand laid down by
glacial meltwater. The gravel and sand are interbedded
with clay in parts of the area. The Williamette Valley is
mostly underlain by interbedded sand, silt, and clay
deposited on floodplains by the Williamette River and
other streams.
The Alluvial Basins region is the driest area in the
United States, with large parts of it being classified as
semiarid and arid. Annual precipitation in the valleys
in Nevada and Arizona ranges from about 100 to 400
mm. However, in the mountainous areas throughout
the region, in the northern part of the Central Valley
of California and in the Washington-Oregon area, an-
Itnpermeable" .
bedrock/
Partly drained
tributary area
Figure 73.
813-G)3
Common Ground-Water Flow Systems in the Alluvial Basins Region (From U.S. Geological Survey Professional Paper
58
-------
Explanation
Gravel
Sand
Silt and clay
Geological Survey Professional Pai
nual precipitation ranges from about 400 mm to more
than 800 mm. The region also receives runoff from
streams that originate in the mountains of the Western
Mountain Ranges region. en-KA*ti«\
Because of the very thin cover of unconsolidated
material on the mountains, precipitation runs oft
Sly down the valleys and out onto the fans where
it infiltrates. The water moves through the sand and
gravel layers toward the centers of the basins. The
centers of many basins consist of flat-floored,
veSion-free areas onto which ground water may
Sw£ and on which overland runoff may collect
during mtense storms. The water that coUects in these
areas (playas), evaporates relatively quickly, leaving
Sh aSi deposi" of clay and other sediment and a
crust of the soluble salts that were dissolved in the
water, as Figure 73 illustrates.
Studies in the region have shown that the hydrology
of the alluvial basins is more complex than that
described in the preceding paragraph, which applies
only to what has been described as "undrmned f^l
Sis" As Figure 73 shows, water may move through
pSeable bedrock from one basin to another, arriv-
ing, ultimately, at a large playa referred tc> <* f
"sink" Water discharges from sinks not by sinking
into the graund, but by evaporating. In those parts of
Se rejo?drained by perennial watercourses ground-
water discharges to the streams from the alluvial
Sosit However, before entering the streams water
may move down some valleys through the aUuvud
osits for tens of kilometers. A reversal of this situa-
olrs along the lower Colorado River and at the
streams into "the alluvium to supply the needs of the
adiacent vegetated zones. .
Ground water is the major source of water in the
Alluvial Basins region. Because of the dry chmate
agriculture requires intensive irrigation. Most of the
ground water is obtained from the sand and gravel
deoosits in the valley alluvium. These deposits are
SeddTwHh finer grained layers of silt and clay
Sat are also saturated with water. When hydraulic
heads in the sand and gravel layers are lowered by
wHhdrawals, the water in the silt and clay begins to
move Sy into the sand and gravel. The movement,
which in some areas takes decades to become *&"*
cant is accompanied by compaction of the silt and clay
2S sSence of the land surface. Subsidence is most
Serein parts of the Central Valley, where it exceeds 9
m in one area, and in southern Arizona, where sub-
sidence of more than 4 m has been observed.
3. Columbia Lava Plateau
(Thick sequence of lava flows irregularly interbedded
with thin unconsolidated deposits and overlain by
thin soils)
As Figure 76 shows, the Columbia Lava Plateau
occupies an area of 366,000 km' in northeastern
California, eastern Washington and Oregon, southern
Idaho, and northern Nevada. As its name implies^ it is
basically a plateau, standing generaUy between 500 and
1,800 m above sea level, that is underlain by a great
59
-------
100 200 300 400 500 Miles
I I I I |
I I I I I I I I
0 100 200 300 400 500 600 700 800 Kilometers
Figure 75. Areas Underlain by Sand and Gravel in the Alluvial Basins Region
60
-------
Explanation
Chiefly sedimentary rocks
Chiefly volcanic rocks
Sedimentary and volcanic rocks
Major aquifers thin or absent
Figure 76. Generalized Distribution and Types of Major Aquifers of the Columbia Lava Plateau Region (Modified from U.S. Geological
Survey Professional Paper 813-S.)3
61
-------
thickness of lava flows irregularly interbedded with silt,
sand, and other unconsolidated deposits.
The great sequence of lava flows, which ranges in
thickness from less than 50 m adjacent to the bordering
mountain ranges to more than 1,000 m in south-central
Washington and southern Idaho, is the principal water-
bearing unit in the region. As Figure 77 shows, the
water-bearing lava is underlain by granite, metamor-
phic rocks, older lava flows, and sedimentary rocks,
none of which are very permeable. Individual lava
flows in the water-bearing zone range in thickness from
several meters to more than 50 m and average about 15
m. The volcanic rocks yield water mainly from
permeable zones that occur at or near the contacts be-
tween some flow layers. Parts of some flows are
separated by soil zones and, at places, by sand, silt,
and clay. These sedimentary layers, where they occur
between lava flows, are commonly referred to as "in-
terflow sediments". Gravel, sand, silt, and clay cover
the volcanic rocks and the older exposed bedrock in
parts of the area.
From the standpoint of the hydraulic characteristics
of the volcanic rocks, it is useful to divide the Colum-
bia Lava Plateau region into two parts: (1) the area in
southeastern Washington, northeastern Oregon, and
the Lewiston area of Idaho, part of which is underlain
by volcanic rocks of the Columbia River Group; and
(2) the remainder of the area shown on Figure 76,
which also includes the Snake River Plain. The basalt
underlying the Snake River Plain is referred to as the
Snake River Basalt; that underlying southeastern
Oregon and the remainder of this area has been divided
into several units, to which names of local origin are
applied.4
The Columbia River Group is of Miocene to
Pliocene (?) age and consists of relatively thick flows
that have been deformed into a series of broad folds
and offset locally along normal faults. Movement of
ground water occurs primarily through the interflow
zones near the top of flows and, to a much smaller ex-
tent, through fault zones and through joints developed
in the dense central and lower parts of the flows. The
axes of sharp folds and the offset of the interflow
zones along faults form subsurface dams that affect the
movement of ground water. Water reaching the in-
terflow zones tends to move down the dip of the flows
from fold axes and to collect updip behind faults that
are transverse to the direction of movement.5 As a
result, the basalt in parts of the area is divided into a
series of barrier-controlled reservoirs, which are only
poorly connected hydraulically to adjacent reservoirs.
The water-bearing basalt underlying California,
Older mountains
S\ /\ /\
River canyon
Explanation
Present soil zone
Interflow zone
'Silt and clay
'Silt and clay
-Cooling fractures
Figure 77. Topographic and Geologic Features of the Columbia Lava Plateau Region
62
-------
Nevada, southeastern Oregon, and southern Idaho is of
Pliocene to Holocene age and consists of small,
relatively thin flows that have been affected to a much
smaller extent by folding and faulting than has the Co-
lumbia River Group. The thin flows contain extensive,
highly permeable interflow zones that are relatively ef-
fectively interconnected through a dense network of
cooling fractures. Structural barriers to ground-water
movement are of minor importance. This is
demonstrated by conditions in the 44,000-square-
kilometer area of the Snake River Plain east of Bliss,
Idaho.
The interflow zones form a complex sequence of
relatively horizontal aquifers that are separated vertical-
ly by the dense central and lower parts of the lava
flows and by interlayered clay and silt. Hydrologists
estimate that the interflow zones, which range in
thickness from about 1 m to about 8 m, account for
about 10 percent of the basalt. MacNish and Barker6
have estimated that the hydraulic conductivity along
the flow-contact zones may be a billion times larger
than the hydraulic conductivity across the dense zones.
The lateral extent of individual aquifers is highly
variable.
The large differences in hydraulic conductivity be-
tween the aquifers and the intervening "confining
zones" result in significant differences in hydraulic
heads between different aquifers. These differences
reflect the head losses that occur as water moves ver-
tically through the system. As a result, heads decrease
with increasing depth in recharge areas and increase
with increasing depth near the streams that serve as
major lines of ground-water discharge. As Figure 78
shows, the difference in heads between different
aquifers can result in the movement of large volumes
of water between aquifers through the openhold (un-
cased) sections of wells.
Much of the Columbia Lava Plateau region is in the
"rain shadow" east of the Cascades and, as a result,
receives only 200 to 1,200 mm of precipitation annual-
ly. The areas that receive the least precipitation include
the plateau area immediately east of the Cascades and
the Snake River Plain. Recharge to the ground-water
system depends on several factors, including the
amount and seasonal distribution of precipitation and
the permeability of the surficial materials. Most
precipitation occurs in the winter and thus coincides
with the cooler, nongrowing season when conditions
are most favorable for recharge. Mundorff7 estimates
that recharge may amount to 600 mm in areas under-
lain by highly permeable young lavas that receive abun-
dant precipitation. Considerable recharge also occurs
by infiltration of water from streams that flow onto
the plateau from the adjoining mountains. These
sources of natural recharge are supplemented in
agricultural areas by the infiltration of irrigation water.
Discharge from the ground-water system occurs as
seepage to streams, as spring flow, and by evapotran-
spiration in areas where the water table is at or near
the land surface. The famous Thousand Springs and
O
•S
V)
•O
I
2. 100-
&
o.
o>
Q
Dense central
and lower parts
of flow
150-
200-
250 0 250
Hole Radius in Millimeters
500
750
Figure 78. Well in a Recharge Area in the Columbia River
Group (Modified from Luzier and Burt, 1974.)8
other springs along the Snake River canyon in southern
Idaho are, in fact, among the most spectacular displays
of ground-water discharge in the world.
The large withdrawal of water in the Columbia Lava
Plateau for irrigation, industrial, and other uses has
resulted in declines in ground-water levels of as much
as 30 to 60 m in several areas. In most of these areas,
the declines have been slowed or stopped through
regulatory restrictions or other changes that have
reduced withdrawals. Declines are still occurring, at
rates as much as a few meters per year, in a few areas.
4. Colorado Plateau and Wyoming Basin
(Thin soils over consolidated sedimentary rocks)
The Colorado Plateau and Wyoming Basin region
occupies an area of 414,000 km2 in Arizona, Colorado,
New Mexico, Utah, and Wyoming. It is a region of
canyons and cliffs of thin, patchy, rocky soils, and of
sparse vegetation adapted to the arid and semiarid
climate. The large-scale structure of the region is that
of a broad plateau standing at an altitude of 2,500 to
3,500 m and underlain by horizontal to gently dipping
63
-------
layers of consolidated sedimentary rocks. As Figure 79
shows, the plateau structure has been modified by an
irregular alternation of basins and domes, in some of
which major faults have caused significant offset of the
rock layers.
The region is bordered on the east, north, and west
by mountain ranges that tend to obscure its plateau
structure. It also contains rather widely scattered ex-
tinct volcanoes and lava fields.
The rocks that underlie the region consist principally
of sandstone, shale, and limestone of Paleozoic to
Cenozoic age. In parts of the region these rock units
include significant amounts of gypsum (calcium
sulfate). In the Paradox Basin in western Colorado the
rock units include thick deposits of sodium- and
potassium-bearing minerals, principally halite (sodium
chloride). The sandstones and shales are most prevalent
and most extensive. The sandstones are the principal
sources of ground water and contain water in fractures
developed both along bedding planes and across the
beds and in interconnected pores. The most productive
sandstones are those that are only partially cemented
and retain significant primary porosity.
Unconsolidated deposits are of relatively minor im-
portance in this region. Thin deposits of alluvium
capable of yielding small to moderate supplies of
ground water occur along parts of the valleys of major
streams, especially adjacent to the mountain ranges in
the northern and eastern parts of the region. In most
of the remainder of the region there are large expanses
of exposed bedrock, and the soils, where present, are
thin and rocky.
Recharge of the sandstone aquifers occurs where
they are exposed above the cliffs and in the ridges.
Average precipitations ranges from about 150 mm in
the lower areas to about 1,000 mm in the higher moun-
tains. The heaviest rainfall occurs in the summer in
isolated, intense thunderstorms during which some
recharge occurs where intermittent streams flow across
sandstone outcrops. However, most recharge occurs in
the winter during snowmelt periods. Water moves
down the dip of the beds away from the recharge areas
to discharge along the channels of major streams
through seeps and springs and along the walls of can-
yons cut by the streams.
The quantity of water available for recharge is small,
but so are the porosity and the transmissivity of most
of the sandstone aquifers. The water in the sandstone
aquifers is unconfined in the recharge areas and is con-
fined downdip. Because most of the sandstones are
consolidated, the storage coefficient in the confined
parts of the aquifers is very small. Even small rates of
withdrawal cause extensive cones of depression around
pumping wells.
The Colorado Plateau and Wyoming Basin is a dry,
sparsely populated region in which most water supplies
are obtained from the perennial streams. Less than 5
percent of the water needs are supplied by ground
water, and the development of even small ground-water
supplies requires the application of considerable
knowledge of the occurrence of both rock units and
their structure, and of the chemical quality of the
water. Also, because of the large surface relief and the
dip of the aquifers, wells even for domestic or small
livestock supplies must penetrate to depths of a few
hundred meters in much of the area. Thus, the
development of ground-water supplies is far more ex-
pensive than in most other parts of the country. These
Canyon
Extinct volcanoes
Ridges
Dome
Fault
Fault
Explanati
Fresh water
Salty water
Shale
on
** * "
^
^3
Sandstone
Limestone
Metamorphic
Figure 79. Topographic and Geologic Features of the Colorado Plateau and Wyoming Basin Region
64
-------
negative aspects notwithstanding, ground water in the
region can support a substantial increase over the
present withdrawals.
As in most other areas of the country underlain by
consolidated sedimentary rock, mineralized (saline)
water—that is, water containing more than 1,000 mg/1
of dissolved solids—is widespread. Most of the shales
and siltstones contain mineralized water throughout the
region and below altitudes of about 2,000 m. Fresh-
water—water containing less than 1,000 mg/1 of dis-
solved solids—occurs only in the most permeable sand-
stones and limestones. Much of the mineralized water
is due to the solution of gypsum and halite. Although
the aquifers that contain mineralized water are com-
monly overlain by aquifers containing freshwater, this
situation is reversed in a few places where aquifers con-
taining mineralized water are underlain by more
permeable aquifers containing freshwater.
5. High Plains
(Thick alluvial deposits over fractured sedimentary
rock)
The High Plains region occupies an area of 450,000
km2 extending from South Dakota to Texas. The
plains are a remnant of a great alluvial plain built in
Miocene time by streams that flowed east from the
Rocky Mountains. Erosion has removed a large part of
the once extensive plain, including all of the part adja-
cent to the mountains, except in a small area in
southeastern Wyoming.
The original depositional surface of the alluvial plain
is still almost unmodified in large areas, especially in
Texas and New Mexico, and forms a flat, impercep-
tibly eastward-sloping tableland that ranges in altitude
from about 2,000 m near the Rocky Mountains to
about 500 m along its eastern edge. The surface of the
southern High Plains contains numerous shallow cir-
cular depressions, called playas, that intermittently con-
tain water following heavy rains. As Figure 80 shows,
other significant topographic features include sand
dunes, which are especially prevalent in central and
northern Nebraska, and wide, downcut valleys of
streams that flow eastward across the area from the
Rocky Mountains.
The High Plains region is underlain by one of the
most productive and most extensively developed
aquifers in the United States. The alluvial materials
derived from the Rocky Mountains, which are referred
to as the Ogallala Formation, are the dominant
geologic unit of the High Plains aquifer. The Ogallala
ranges in thickness from a few meters to more than
200 m and consists of poorly sorted and generally un-
consolidated clay, silt, sand and gravel.
Younger alluvial materials of Quarternary age overlie
the Ogallala Formation of late Tertiary age in most
parts of the High Plains. Where these deposits are
saturated, they form a part of the High Plains aquifer;
in parts of south-central Nebraska and central Kansas,
where the Ogallala is absent, they comprise the entire
aquifer. The Quarternary deposits are composed largely
of material derived from the Ogallala and consist of
Platte River
^
O ofo d
Explar
Sand
Gravel
lation
jr-lr:
Clay
Sandstone
Figure 80. Topographic and Geologic Features of the High Plains Region.
65
-------
gravel, sand, silt, and clay. An extensive area of dune
sand occurs in the Sand Hills area north of the Platte
River in Nebraska.
Other, older geologic units that are hydrologically
connected to the Ogallala include the Arikaree Group
of Miocene age and a small part of the underlying
Brule Formation. The Arikaree Group is predomi-
nantly a massive, very fine to fine-grained sandstone
that locally contains beds of volcanic ash, silty sand,
and sandy clay. The maximum thickness of the
Arikaree is about 300 m, in western Nebraska. The
Brule Formation of Oligocene age underlies the
Arikaree. In most of the area in which it occurs, the
Brule forms the base of the High Plains aquifer.
However, in the southeastern corner of Wyoming and
the adjacent parts of Colorado and Nebraska, the
Brule contains fractured sandstones hydraulically inter-
connected to the overlying Arikaree Group; in this area
the Brule is considered to be a part of the High Plains
aquifer.
In the remainder of the region, the High Plains
aquifer is underlain by several formations, ranging in
age from Cretaceous to Permian and composed prin-
cipally of shale, limestone, and sandstone. The oldest
of these underlies parts of northeastern Texas, western
Oklahoma, and central Kansas and contains layers of
relatively soluble minerals including gypsum, anhydrite,
and halite (common salt), which are dissolved by cir-
culating ground water.
Prior to the erosion that removed most of the
western part of the Ogallala, the High Plains aquifer
was recharged by the streams that flowed onto the
plain from the mountains to the west as well as by
local precipitation. The only source of recharge now is
local precipitation, which ranges from about 400 mm
along the western boundary of the region to about 600
mm along the eastern boundary. Precipitation and
ground-water recharge on the High Plains vary in an
east-west direction, but recharge to the High Plains
aquifer also varies in a north-south direction. The
average annual rate of recharge has been determined to
range from about 5 mm in Texas and New Mexico to
about 100 mm in the Sand Hills in Nebraska. This
large difference is explained by differences in evapora-
tion and transpiration and by differences in the
permeability of surficial materials.
In some parts of the High Plains, especially in the
southern part, the near-surface layers of the Ogallala
have been cemented with lime (calcium carbonate) to
form a material of relatively low permeability called
caliche. Precipitation on areas underlain by caliche
soaks slowly into the ground. Much of this precipita-
tion collects in playas that are underlain by silt and
clay, with the result that most of the water evaporates.
It is only during years of excessive precipitation that
significant recharge occurs and this, as noted above,
averages only about 5 mm per year in the southern
part of the High Plains. In the Sand Hills area about
20 percent of the precipitation (or about 100 mm an-
nually) reaches the water table as recharge.
Figure 81 shows that the water table of the High
Plains aquifer has a general slope toward the east.
Gutentag and Weeks9 estimate, on the basis of the
average hydraulic gradient and aquifer characteristics,
that water moves through the aquifer at a rate of about
0.3 m (1 ft) per day.
Natural discharge from the aquifer occurs to
streams, to springs and seeps along the eastern bound-
ary of the plains, and by evaporation and transpiration
in areas where the water table is within a few meters of
the land surface. However, at present the largest
discharge is probably through wells. The widespread
occurrence of permeable layers of sand and gravel,
which permit the construction of large-yield wells
almost any place in the region, has led to the develop-
ment of an extensive agricultural economy largely
dependent on irrigation. Most of this water is derived
from ground-water storage, resulting in a long-term
continuing decline in ground-water levels in parts of
the region of as much as 1 m per year.
The depletion of ground-water storage in the High
Plains is a matter of increasing concern in the region.
However, from the standpoint of the region as a
whole, the depletion does not yet represent a large part
of the storage that is available for use. Weeks and
Gutentag10 estimate, on the basis of a specific yield of
15 percent of the total volume of saturated material,
that the available (usable) storage in 1980 was about
4x 102 m3 (3.3 billion acre-ft). Luckey, Gutentag, and
Weeks11 estimate that this is only about 5 percent less
than the storage that was available at the start of
withdrawals. However, in areas where intense irrigation
has long been practiced, depletion of storage is severe.
6. Nonglaciated Central Region
(Thin regolith over fractured sedimentary rocks)
As Figure 82 shows, the Nonglaciated Central region
is an area of about 1,737,000 km2 extending from the
Appalachian Mountains on the east to the Rocky
Mountains on the west. The part of the region in
eastern Colorado and northeastern New Mexico is
separated from the remainder of the region by the
High Plains region. The Nonglaciated Central region
also includes the Triassic Basins in Virginia and North
Carolina and the "driftless" area in Wisconsin, Min-
nesota, Iowa, and Illinois where glacial deposits, if
present, are thin and of no hydrologic importance.
The region is geologically complex. Most of its
underlain by consolidated sedimentary rocks that range
in age from paleozoic to Tertiary and consist largely of
sandstone, shale, limestone, dolomite, and con-
glomerate. A small area in Texas and western
Oklahoma is underlain by gypsum. Figure 83 shows
that throughout most of the region the rock layers are
horizontal or gently dipping. Principal exceptions are
the Valley and Ridge section, the Wichita and Arbuckle
Mountains in Oklahoma, and the Ouachita Mountains
in Oklahoma and Arkansas, in all of which the rocks
have been folded and extensively faulted. As Figure 84
66
-------
40=
35C
105°
-7—
WYOMING
100°
SOUTH DAKOTA
TEXAS \.~v-.
Explanation
•300— Altitude of the water table in
meters, winter 1978
Contour interval 300 meters
Datum is National Geodetic
Vertical Datum of 1929
50
100
200 Miles
i
50 100
I
200 Kilometers
Figure 81. Altitude of the Water Table of the High Plains Aquifer.
67
-------
Triassic
basin
P
-4^—'
Arbuckle ,
Mountains I Ouachita
'Mountains
Figure 82. Location of Geographic Features Mentioned in the Discussions or Regions Covering the Central and Eastern Parts of the
United States.
shows around the Black Hills and along the eastern
side of the Rocky Mountains the rock layers have been
bent up sharply toward the mountains and truncated
by erosion. The Triassic Basins in Virginia and North
Carolina are underlain by moderate to gently dipping
beds of shale and sandstone that have been extensively
faulted and invaded by narrow bodies of igneous rock.
The land surface in most of the region is underlain
by regolith formed by chemical and mechanical
breakdown of the bedrock. In the western part of the
Great Plains the residual soils are overlain by or inter-
mixed with wind-laid deposits. In areas underlain by
relatively pure limestone, the regolith consists mostly of
clay and is generally only a few meters thick. Where the
limestones contain chert and in the areas underlain by
shale and sandstone, the regolith is thicker, up to 30 m
or more in some areas. The chert and sand form
moderately permeable soils, whereas the soils developed
on shale are finer grained and less permeable.
As Figure 83 shows, the principal water-bearing
openings in the bedrock are fractures, which generally
occur in three sets. The first set, and the one that is
probably of greatest importance from the standpoint of
ground-water as well yields, consists of fractures
developed along bedding planes. The two remaining
sets are essentially vertical and thus cross the bedding
planes at a steep angle. The primary difference between
the sets of vertical fractures is in the orientation of the
fractures in each set. The vertical fractures facilitate
movement of water across the rock layers and thus
serve as the principal hydraulic connection between the
bedding-plane fractures.
In the parts of the region in which the bedrock has
been folded or bent, fractures range from horizontal to
vertical. Fractures parallel to the land surface, where
present, are probably less numerous and of more
limited extent than in areas of flat-lying rocks.
The openings developed along most fractures are less
than a millimeter wide. The principal exception occurs
in limestones and dolomites. Water moving through
these rocks gradually enlarges the fractures to form, in
time, extensive cavernous openings or cave systems.
Many large springs emerge from these openings.
Recharge of the ground-water system in this region
occurs primarily in the outcrop areas of the bedrock
aquifers in the uplands between streams. Precipitation
in the region ranges from about 400 mm per year in
the western part to more than 1,200 mm in the eastern
part. This wide difference in precipitation is reflected in
recharge rates, which range from about 5 mm per year
in west Texas and New Mexico to as much as 500 mm
per year in Pennsylvania and eastern Tennessee.
68
-------
Discharge from the ground-water system is by springs
and seepage into streams and by evaporation and
transpiration.
The yield of wells depends on: (1) the number and
size of fractures that are penetrated and the extent to
which they have been enlarged by solution, (2) the rate
of recharge, and (3) the storage capacity of the bedrock
and regolith. Yields of wells in most of the region are
small, in the range of 0.01 to 1 m3 min"1 (about 2.5 to
about 250 gallons per minute), making the
Nonglaciated Central region one of the least favorable
ground-water regions in the country. Even in parts of
the areas underlain by cavernous limestone, yields are
moderately low because of both the absence of a thick
regolith and the large water-transmitting capacity of the
cavernous openings which quickly discharge the water
that reaches them during periods of recharge.
The exceptions to the small well yields are the caver-
nous limestones of the Edwards Plateau, the Ozark
Plateaus, and the Ridge and Valley section. Figure 82
Regolith
--~-~ - -
fractures, . .,. . 4' " • •' '•)£,
Fresh water
Salty water
Figure 83. Topographic and Geologic Features of the Nonglaciated Central Region
Explanation
7?*^
Fresh water
Salty water
•^-^
-^_-_-
-"^^•^— — v
-X_x — •
•j Sandstone
Shale
Metamorphic rocks
Figure 84. Topographic and Geologic Features Along the Western Boundary of the Nonglaciated Central Region.
69
-------
shows the location of these areas. The Edwards Plateau
in Texas is bounded on the south by the Balcones Fault
Zone, in which limestone and dolomite up to 150 m in
thickness has been extensively faulted, which facilitated
the development of solution openings. This zone forms
one of the most productive aquifers in the country.
Wells of the City of San Antonio are located in this
zone; individually, they have yields of more than 60
m3min'1.
As Figures 83 and 84 show, another feature that
makes much of this region unfavorable for ground-
water development is the occurrence of salty water at
relatively shallow depths. In most of the Nonglaciated
Central region, except the Ozark Plateaus, the
Ouachita and Arbuckle Mountains, and the Ridge and
Valley section, the water in the bedrock contains more
than 1,000 mg/1 of dissolved solids at depths less than
150m.
7. Glaciated Central Region
(Glacial deposits over fractured sedimentary rocks)
Figure 82 shows the Glaciated Central region which
occupies an area of 1,297,000 km2 extending from the
Triassic Basin in Connecticut and Massachusetts and
the Catskill Mountains in New York on the east to the
northern part of the Great Plains in Montana on the
west. Figure 85 shows that the Glaciated Central region
is underlain by relatively flat-lying consolidated
sedimentary rocks that range in age from Paleozoic to
Tertiary. The bedrock is overlain by glacial deposits
which, in most of the area, consist chiefly of till, and
unsorted mixture of rock particles deposited directly by
the ice sheets. The till is interbedded with and overlain
by sand and gravel deposited by meltwater streams, by
silt and clay deposited in glacial lakes, and, in large
parts of the North-Central States, by loess, a well-
sorted silt believed to have been deposited primarily by
the wind.
On the Catskill Mountains and other uplands in the
eastern part of the region, the glacial deposits are
typically only a few to several meters thick. In much of
the central and western parts of the region, the glacial
deposits exceed 100 m in thickness. The principal ex-
ception is the "driftless" area in Wisconsin, Min-
nesota, Iowa, and Illinois where the bedrock is overlain
by thin soils. This area, both geologically and
hydrologically, resembles the Nonglaciated Central
region and is, therefore, included as part of that
region.
The glacial deposits are thickest in valleys in the
bedrock surface. In most of the region westward from
Ohio to the Dakotas, the thickness of the glacial
deposits exceeds the relief on the preglacial surface,
with the result that the locations of valleys and stream
channels in the preglacial surface are no longer discer-
nible from the land surface. Figure 85 shows that the
glacial deposits in buried valleys include, in addition to
till and lacustrine silts and clays, substantial thicknesses
of highly permeable sand and gravel.
Ground water occurs both in the glacial deposits and
in the bedrock. Water occurs in the glacial deposits in
pores between the rock particles and in the bedrock
primarily along fractures.
Large parts of the region are underlain by limestones
and dolomites in which fractures have been enlarged by
solution. On the whole, caves and other large solution
openings are much less numerous and hydrologically
Loess
Fresh water
Salty water
Figure 85. Topographic and Geologic Features of the Glaciated Central Region
70
-------
much less important in the Glaciated Central region.
The glacial deposits are recharged by precipitation on
the interstream areas and serve both as a source of
water to shallow wells and as a reservoir for recharge
to the underlying bedrock. Precipitation ranges from
about 400 mm per year in the western part of the
region to about 1,000 mm in the east. On sloping
hillsides underlain by clay-rich till, the annual rate of
recharge, even in the humid eastern part of the region,
probably does not exceed 50 mm. In contrast, relatively
flat areas underlain by sand and gravel may receive as
much as 300 mm of recharge annually in the eastern
part of the region.
Ground water in small to moderate amounts can be
obtained any place in the region, both from the glacial
deposits and from the bedrock. Large to very large
amounts of ground water are obtained from the sand
and gravel deposits and from some of the limestones,
dolomites, and sandstones. The shales are the least pro-
ductive bedrock formations in the region.
Because of the widespread occurrence of limestone
and dolomite, water from both the glacial deposits and
the bedrock contains as much as several hundred
milligrams per liter of dissolved minerals and is
moderately hard. Concentrations of iron in excess of
0.3 mg/1 is a problem in water from some of the sand-
stone aquifers in Wisconsin and Illinois and locally in
glacial deposits throughout the region. Sulfate in excess
of 250 mg/1 is a problem in water both from the
glacial deposits and from the bedrock in parts of New
York, Ohio, Indiana, and Michigan.
As is the case in the Nonglociated Central region
mineralized water occurs at relatively shallow depth in
bedrock in large parts of this region. The thickness of
the freshwater zone in the bedrock depends on the ver-
tical hydraulic conductivity of both the bedrock and
the glacial deposits and on the effectiveness of the
hydraulic connection between them. Both the
freshwater and the underlying saline water move
toward the valleys of perennial streams to discharge.
As a result, the depth to saline water is less under
valleys than under uplands. At depths of 500 to 1,000
m in much of the region, the mineral content of the
water approaches that of seawater (about 35,000 mg/1).
At greater depths, the mineral content may reach con-
centrations several times that of seawater.
8. Piedmont Blue Ridge Region
(Thick regolith over fractured crystalline and
metamorphosed sedimentary rocks)
The Piedmont and Blue Ridge region is an area of
about 247,000 km2 extending from Alabama on the
south to Pennsylvania on the north. The Piedmont part
of the region consists of low, rounded hills and long,
rolling, northeast-southwest trending ridges. The Blue
Ridge is mountainous and includes the highest peaks
east of the Mississippi.
The Piedmont and Blue Ridge region is underlain by
bedrock of Precambrian and Paleozoic age consisting
of igneous, and metamorphosed igneous, and sedimen-
tary rocks. The land surface in the Piedmont and Blue
Bedrock outcrops
Best well sites indicated with X's
Figure 86. Topographic and Geologic Features of the Piedmont and Blue Ridge Region.
71
-------
Ridge is underlain by clay-rich, unconsolidated material
derived from in situ weathering of the underlying
bedrock. This material, which averages about 10 to 20
m in thickness and may be as much as 100 m thick on
some ridges, is referred to as saprolite. In many
valleys, especially those of larger streams, flood plains
are underlain by thin, moderately well-sorted alluvium
deposited by the streams. While the distinction between
saprolite and alluvium is not important, the term
regolith is used to refer to the layer of unconsolidated
deposits.
As Figure 86 shows the regolith contains water in
pore spaces between rock particles. The bedrock, on
the other hand, does not have any significant in-
tergranular porosity. It contains water, instead, in
sheetlike openings formed along fractures. The
hydraulic conductivities of the regolith and the bedrock
are similar and range from about 0.001 to 1 m day"1.
The major difference in their water-bearing
characteristics is their porosities, the porosity of
regolith being about 20 to 30 percent and the porosity
of the bedrock about 0.01 to 2 percent. Small supplies
of water adequate for domestic needs can be obtained
from the regolith through large-diameter bored or dug
wells. However, most wells, especially those where
moderate supplies of water are needed, are relatively
small in diameter and are cased through the regolith
and finished with open holes in the bedrock. Although,
the hydraulic conductivity of the bedrock is similar to
that of the regolith, bedrock wells generally have much
larger yields than regolith wells because, being deeper,
they have a much larger available drawdown.
All ground-water systems function both as reservoirs
that store water and as pipelines that transmit water
from recharge areas to discharge areas. The yield of
bedrock wells in the Piedmont and Blue Ridge region
depends on the number and size of fractures penetrated
by the open hole and on the replenishment of the frac-
tures by seepage into them from the overlying regolith.
Thus, the ground-water system in this region can be
viewed, from the standpoint of ground-water develop-
ment, as a terrane in which the reservoir and pipeline
functions are effectively separated. Because of its larger
porosity, the regolith functions as a reservoir that slow-
ly feeds water downward into the fractures in the
bedrock. The fractures serve as an intricate intercon-
nected network of pipelines that transmit water either
to springs or streams or to wells.
Recharge of the ground-water system occurs on the
areas above the flood plains of stream, and natural
discharge occurs as seepage springs that are common
near the bases of slopes and as seepage into streams.
With respect to recharge conditions, it is important to
note that forested areas, which include most of the
Blue Ridge and much of the Piedmont, have thick and
very permeable soils overlain by a thick layer of forest
litter. In these areas, even on steep slopes, most of the
precipitation seeps into the soil zone, and most of this
moves laterally through the soil and a thin, temporary,
saturated zone to surface depressions or streams to
discharge. The remainder seeps into the regolith below
the soil zone, and much of this ultimately seeps into
the underlying bedrock.
The Piedmont and Blue Ridge region has long been
known as an area generally unfavorable for ground-
water development. This reputation seems to have
resulted both from the small reported yields of the
numerous domestic wells in use in the region that were,
generally, sited as a matter of convenience and from a
failure to apply existing technology to the careful selec-
tion of well sites where moderate yields are needed. As
water needs in the region increase and as reservoir sites
on streams become incresingly more difficult to obtain,
it will be necessary to make intensive use of ground
water.
9. Northeast and Superior Uplands
(Glacial deposits over fractured crystalline rocks)
The Northeast and Superior Uplands region is made
up of two separate areas totaling about 415,000 km2.
The Northeast Upland encompasses the Adirondack
Mountains, the Lake Champlain valley, and nearly all
of New England. The Superior Upland encompasses
most of the northern parts of Minnesota and Wiscon-
sin adjacent to the western end of Lake Superior.
Bedrock in the region ranges in age from Precam-
brian to Paleozoic, and as Figure 87 shows, consists
mostly of intrusive igneous rocks and metamorphosed
sedimentary rocks. Most have been intensively folded
and cut by numerous faults.
As Figures 87 and 88 show, the bedrock is overlain
by unconsolidated glacial deposits including till and
gravel, sand, silt, and clay. The thickness of the glacial
deposits ranges from a few meters on the higher moun-
tains, which also have large expanses of barren rock, to
more than 100 m in some valleys. The most extensive
glacial deposit is till. In most of the valleys and other
low areas, the till is covered by glacial outwash con-
sisting of interlayered sand and gravel, ranging in
thickness from a few meters to more than 20 m.
Ground-water supplies are obtained in the region
from both the glacial deposits and the underlying
bedrock. The largest yields come from the sand and
gravel deposits, which in parts of the valleys of large
streams are as much as 60 m thick. Water occurs in the
bedrock in fractures similar in origin, occurrence, and
hydraulic characteristics to those in the Piedmont and
Blue Ridge region.
Recharge from precipitation generally begins in the
fall after plant growth stops. It continues intermittently
over the winter during thaws and culminated during the
period between the spring thaw and the start of the
growing season. Precipitation on the Northeast
Upland, about 1,200 mm per year, is twice that on the
Superior Upland, with the result that recharge is largest
in the Northeast. The glacial deposits in the region
serve as a storage reservoir for the fractures in the
underlying bedrock.
72
-------
Figure 87. Topographic and Geologic Features on the Northeast and Superior Uplands Region.
Water supplies in the Northeast and Superior
Uplands region are obtained from open-hole drilled
wells in bedrock, from drilled and screened or open-
end wells in sand and gravel, and from large-diameter
bored or dug wells in till. The development of water
supplies from bedrock, especially in the Superior
Upland, is more uncertain than from the fractured
rocks in the Piedmont and Blue Ridge region because
the ice sheets that advanced across the region removed
the upper, more fractured part of the rock and also
tended to observe many of the fracture-caused depres-
sions in the rock surface with the layer of glacial till.
Most of the rocks that underlie the Northeast and
Superior Uplands are relatively insoluble, and conse-
quently, the ground water in both the glacial deposits
and the bedrock generally contains less than 500 mg/1
of dissolved solids. Two of the most significant water-
quality problems confronting the region, especially the
Northeast Upland section, are acid precipitation and
pollution caused by salts used to de-ice highways.
Much of the precipitation falling on the Northeast in
1982 had a pH in the range of 4 to 6 units. Because of
the low buffering capacity of the soils derived from
rocks underlying the area, there is relatively little op-
portunity for the pH to be increased. One of the results
of this is the gradual elimination of living organisms
from many lakes and streams. The effect on ground-
water quality, which will develop much more slowly,
has not yet been determined. The second problem—
that of de-icing salts—affects ground-water quality ad-
jacent to streets and roads maintained for winter
travel.
10. Atlantic and Gulf Coastal Plain
(Complexly interbedded sand, silt and clay)
The Atlantic and Gulf Coastal Plain region is an
area of about 844,000 km2 extending from Cape Cod,
Massachusetts, to the Rio Grande in Texas. This region
does not include Florida and parts of the adjacent
States.
The topography of the region ranges from extensive,
flat, coastal swamps and marshes, 1 to 2 m above sea
level, to rolling uplands, 100 to 250 m above sea level,
along the inner margin of the region.
The region is underlain by unconsolidated sediments
that consist principally of sand, silt, and clay. These
sediments, which range in age from Jurassic to the
present, range in thickness from less than a meter near
the inner edge of the region to more than 12,000 m in
southern Louisiana. The sediments are complexly inter-
bedded to the extent that most of the named geologic
units into which they have been divided contain layers
of the different types of sediment that underlie the
region. These named geologic units dip toward the
coast or toward the axis of the Mississippi embayment,
with the result that those that crop out at the surface
form a series of bands roughly parallel to the coast or
to the axis of the embayment, as shown in Figure 89.
Although sand, silt, and clay are the principal types
of material underlying the Atlantic and Gulf Coastal
Plain, there are also small amounts of gravel interbed-
ded with the sand, a few beds composed of mollusk
shells, and small amounts of limestone present in the
region. The most important limestone is the semi-
73
-------
CO
c
8
CD
CD
c
I
Q.
01
a
o
C5
CD
O
O
CO
"
CO
I
1
Explanation
Areas occupied by lakes during the glacial period
Areas underlain by glacial deposits
-------
Figure 89. Topographic and Geologic Features of the Gulf Coastal Plain
consolidated Castle Hayne Limestone of Eocene age
which underlies an area of about 26,000 km2 in eastern
North Carolina, is more than 200 m thick in much of
the area, and is the most productive aquifer in North
Carolina. A soft, clayey limestone (the chalk of the
Selma Group) of Late Cretaceous age underlies parts
of eastern Mississippi and western Alabama, but in-
stead of being an aquifer, it is an important confining
bed.
From the standpoint of well yields and ground-water
use, the Atlantic and Gulf Coastal Plain is one of the
most important regions in the country. Recharge to the
ground-water system occurs in the interstream areas,
both where sand layers crop out and by percolation
downward across the interbedded clay and silt layers.
Discharge from the system occurs by seepage to
streams, estuaries, and the ocean.
Wells that yield moderate to large quantities of water
can be constructed almost anywhere in the region.
Because most of the aquifers consist of unconsolidated
sand, wells require screens; where the sand is fine-
grained and well sorted, the common practice is to sur-
round the screens with a coarse sand or gravel
envelope.
Withdrawals near the outcrop areas of aquifers are
rather quickly balanced by increases in recharge and
(or) reductions in natural discharge. Withdrawals at
significant distances downdip do not appreciably affect
conditions in the outcrop area and thus must be partly
or largely supplied from water in storage in the
aquifers and confining beds.
If withdrawals are continued for long periods in
areas underlain by thick sequences of unconsolidated
deposits, the lowered ground-water levels in the aquifer
may result in drainage of water from layers of silt and
clay. The depletion of storage in fine-grained beds
results in subsidence of the land surface. Subsidence in
parts of the Houston area totaled about 9 m as of
1978. Subsidence near pumping centers in the Atlantic
Coastal Plain has not yet been confirmed but is believed
to be occurring at a slower rate than along the Texas
Gulf Coast.
Depletion of storage in the aquifers underlying large
areas of the Atlantic and Gulf Coastal Plain is
reflected in long-term declines in ground-water levels.
These declines suggest that withdrawals in these areas
are exceeding the long-term yield of the aquifers.
Another problem that affects ground-water develop-
ment in the region concerns the presence of saline
water in the deeper parts of most aquifers. In some of
the deeper aquifers, the interface between freshwater
and saltwater is inshore, but in parts of the region, in-
cluding parts of Long Island, New Jersey, and
Mississipi, the interface in the most intensively
developed aquifers is a significant distance offshore.
Pumping near the interfaces has resulted in local prob-
lems of saltwater encroachment.
Another significant feature of the ground-water
system in this region is the presence of "geopressured"
zones at depths of 1,800 to 6,100 m in Texas and Loui-
siana which contain water at a temperature of 80°C to
more than 273°C. Water in these zones contains
significant concentrations of natural gas, and the water
in some zones is under pressures sufficient to support a
column of water more than 4,000 m above land sur-
face. Because the elevated temperature, natural gas,
and high pressure are all potential energy sources, these
zones are under intensive investigation.
75
-------
11. Southeast Coastal Plain
(Thick layers of sand and clay over semi-
consolidated
carbonate rocks)
Figure 90 shows the Southeast Coastal Plain, an area
of about 212,000 km2 in Alabama, Honda, Georgia,
and South Carolina. It is a relatively flat, low-lying
area. Much of the area, including the Everglades in
southern Florida, is a nearly flat plain less than 10 m
above sea level.
The land surface of the Southeast Coastal Plain is
underlain by unconsolidated deposits of Pleistocene age
consisting of sand, gravel, clay, and shell beds and, in
ALABAMA
W3°~
Explanation
-10— Altitude of the water level in wells in meters above
sea level. May 1980, Contour Interval 10 meters
^%% Principal recharge areas
Figure 90. Potent iometric Surface for the Floridan Aquifer (Adapted from Johnston, Healy, and Hayes, 1981.)13
76
-------
southeastern Florida, by semi-consolidated limestone.
In most of the region, the surficial deposits rest on for-
mations, primarily of middle to late Miocene age, com-
posed of interbedded clay, sand, and limestone. The
formations of middle to late Mioccene age or surficial
deposits overlie semi-consolidated limestones and
dolomites that are as much as 1,500 m thick.
The Tertiary limestone that underlies the Southeast
Coastal Plain constitutes one of the most productive
aquifers in the United States and is the feature that
justifies treatment of the region separately from the re-
mainder of the Atlantic and Gulf Coastal Plain. The
aquifer, which is known as the Floridan aquifer,
underlies all of Florida and southeast Georgia and
small areas in Alabama and South Carolina. The
Floridan aquifer consists of layers several meters thick
composed largely of loose aggregations of shells and
fragments of marine organisms interbedded with much
thinner layers of cement and cherty limestone. The
Floridan, one of the most productive acquifers in the
world, is the principal source of ground-water supplies
in the Southeast Coastal Plain region.
In southern Florida, south of Lake Okeechobee, and
in a belt about 30 km wide northward along the east
coast of Florida to the vicinity of St. Augustine, the
water in the Floridan aquifer contains more than 100
mg/1 of chloride. In this area, most water supplies are
obtained from surfacial aquifers. The most notable of
these aquifers underlies the southeastern part of Florida
and, in the Miami area, consists of 30 to 100 m of
cavernous limestone and sand and is referred to as the
Biscayne aquifer. The Biscayne is an unconfined
aquifer which is recharged by local precipitation and by
infiltration of water from canals that drain water from
impoundments developed in the Everglades. It is the
principal source of water for municipal, industrial, and
irrigation uses and can yield as much as 5 m3 min~J
(1,300 gal min"1) to small-diameter wells less than 25
m deep finished with open holes only 1 to 2 m long.
The surficial aquifers in the remainder of the region
are composed primiarily of sand, except in the coastal
zones of Florida where the sand is interbedded with
shells and thin limestones. These surficial aquifers serve
as sources of small ground-water supplies throughout
the region and are the primary sources of ground water
where the water in the Floridan aquifer contains more
than about 250 mg/1 of chloride.
The Floridan aquifer is the principal source of
ground water in the region. Ground water in the upper
part of the aquifer is unconfined in the principal
recharge areas in Georgia and in west-central Florida,
which are shown in Figure 90. In the remainder of the
region, water in the aquifer is confined by clay in the
Hawthorn Formation and in other beds that overlie the
aquifer.
Recharge occurs where the potentiometric surface of
•^
/t
v»V
Area of artesian
flow ..
»•''
,/'/
."/
/ y^~^ .<>uliiZ'
^^7?7
Potentiometric
surface
Sand3
jls
Limestone
spring ^
Ji
Jl>
w
'FloridanAouife,
JG ' '
^_l I / _/
J L
Figure 91. Topographic and Geologic Features of the Southeast Coastal Plain Region
77
-------
the Floridan aquifer is lower than the water table in the
overlying surficial aquifer. As Figure 90 shows, the
principal recharge areas include a broad area along the
west side of Florida extending from the central part of
the peninsula to south-central Georgia and an area ex-
tending from west-central Florida through southeast
Alabama into southwest Georgia. In these areas,
recharge rates are estimated to exceed 120 mm yr"1 (5
in. yr"1). Recharge occurs by infiltration of precipita-
tion directly into the limestone, where it is exposed at
the land surface, and by seepage through the permeable
soils that partly mantle the limestone in the outcrop
areas. Considerable recharge also occurs in the higher
parts of the recharge areas through permeable openings
in the confining beds, where these beds have been
breached by the collapse of caverns in the limestone
during the process of sinkhole formation. Figure 91 il-
lustrates this sinkhole formation. Thus, the land sur-
face in most of Florida north of Lake Okeechobee is
marked by thousands of closed depressions ranging in
diameter from a few meters to several kDometers. The
larger depressions are occupied by lakes generally re-
ferred to as sinkhole lakes.
Discharge from the Floridan aquifer occurs through
springs and by seepage to streams. Considerable
discharge also occurs by diffuse seepage across the
overlying confining beds in areas where the poten-
tiometric surface of the aquifer stands at a higher
altitude than the water table. In most of these areas
wells open to the aquifer will flow at the land surface.
The most spectacular discharge from the Floridan
aquifer is through sinkholes exposed along streams and
offshore.
Water supplies are obtained from the Floridan
aquifer by installing casing through the overlying for-
mations and drilling an open hole in the limestones and
dolomites comprising the aquifer. Total withdrawals
from the aquifer are estimated to have been about
13x 106 m3 day-l (3.5 billion gallons per day) in 1978.
Large withdrawals also occur from the other aquifers
in the region.
12. Alluvial Valleys
(Thick sand and gravel deposits beneath floodplains
and terraces of streams)
In the preceding discussions of ground-water regions,
streams and other bodies of surface water were men-
tioned as places of ground-water discharge. In most
areas ground-water systems and surface streams form a
water system so intimately interconnected that a change
in one causes a change in other. For example,
withdrawals from ground-water systems reduce
discharge to streams and thereby reduce streamflow.
The movement of water from streams into ground-
water systems in response to withdrawals is not a
significant feature in most areas because ground-water
withdrawals are dispersed over the uplands between
streams rather than concentrated near them. An excep-
tion to this occurs where stream channels and
floodplains are underlain by highly permeable deposits
of sand and gravel. The large yields of these deposits,
as well as the variability and availability of streamflow,
encourage the development of these sand and gravel
deposits as sources of ground water, and thus, en-
courage the concentration of withdrawals near streams.
From the standpoint of ground-water hydrology,
three criteria are used to differentiate alluvial valleys
from other valleys. These criteria are as follows:
1. The alluvial valleys contain sand and gravel deposits
thick enough to supply water to wells at moderate
to large rates. [Commonly, the water-transmitting
capacity of the sand and gravel is at least 10 times
larger than that of the adjacent (enclosing) rocks.]
2. The sand and gravel deposits are in hydraulic con-
tact with a perennial stream that serves as a source
of recharge and whose flow normally far exceeds
that demand from any typical well field.
2. The sand and gravel deposit occurs in a clearly
defined band ("channel") that normally does not
extend beyond the floodplain and adjacent terraces.
In other words, the width of the deposit is small or
very small compared with its length.
According to these criteria, the valleys of streams
that were not affected by glacial meltwater are not con-
sidered alluvial valleys. The floodplains in these valleys
are commonly underlain only by thin deposits of fine-
grained alluvium. These criteria also eliminate the
"burned" valleys of the glaciated area. Although the
water-transmitting capacity of the sand and gravel in
buried valleys may be large, the yield to wells in most
of them is small because of the limited opportunity for
recharge through the surrounding, less-permeable
materials.
The alluvial valleys are commonly underlain, in addi-
tion to sand and gravel, by deposits of silt and clay. In
many of the glaciated valleys in New York and New
England the land surface is underlain by a layer of
sand and gravel that ranges in thickness from 1 to 2 m
to more than 10 m. The bottom of this deposit ranges,
from one part of a valley to another, from a position
above the water table to several meters below the bot-
tom of streams. This surficial deposit of sand and
gravel is commonly underlain by interbedded silt and
clay which is, in turn, underlain by a discontinuous
"basal" layer of sand and gravel.
The sequence of deposits in the alluvial valleys
depends, of course, on the history of deposition in the
valleys. Figure 92 shows that the sand and gravel in the
valleys of major streams, such as those of the
Mississippi, Missouri, and Ohio, are commonly
overlain by deposits of clay and other fine-grained
alluvium deposited during floods since the end of the
glacial period.
Under natural conditions the alluvial deposits are
recharged by precipitation on the valleys, by ground
water moving from the adjacent and underlying
aquifers, by overbank flooding of the streams, and, in
some glacial valleys, by infiltration from tributary
78
-------
Mississippi Rh
and clay
Limestone
Figure 92. Topographic and Geologic Features of a Section of the Alluvial Valley of the Mississippi River
streams. Water in the alluvial deposits discharges to the
streams in the valleys.
The layers of sand and gravel in the alluvial valleys
are among the most productive aquifers in the country.
They have been extensively developed as sources of
water for municipalities, industries, and irrigation.
Some of the gravel layers have hydraulic conductivities
nearly as large as those of cavernous limestone. The
large yields of the sand and gravel depend not only on
their large water-transmitting capacity but also on their
hydraulic connection to the streams flowing in the
valleys. Large withdrawals from the deposits result in a
reduction in ground-water discharge to the streams
and, if large enough, cause infiltration of water from
the streams into the deposits.
-------
References
1 Heath, R.C. 1982. "Classification of Ground-Water
Systems of the United States", Ground
Water, V. 20, no. 4, July-August 1982.
2McGuiness, C.L. 1963. "The Role of Ground Water
in the National Water Situation." U.S.
Geological Survey Water-Supply Paper 1800.
3U.S. Geological Survey, Professional Paper 813,
"Summary Appraisals of the Nation's
Ground-Water Resources". Published as a
series of chapters based on the boundaries
established by the United States Water-
Resources Council for Water-Resources
Regions in the United States.
4Hampton, E.R. 1964. "Geologic Factors That
Control the Occurence and Availability of Ground
Water in the Forth Rock Basin, Lake County,
Oregon." U.S. Geological Survey Professional
Paper 383-B
5Newcomb, R.C. 1962. "Storage of Ground Water
Behind Subsurface Darns in the Columbia River
Basalt, Washington, Oregon, and Idaho." U.S.
Geological Survey Professional Paper 383-A.
6MacNish, R.D. and R.A. Barker. 1976. "Digital
Simulation of a Basalt Aquifer System,
Walla Walla River Basin, Washington and
Oregon." Washington Department of
Ecology, Water-Supply Bulletin 44.
7 Columbia-North Pacific Technical Staff. 1970.
"Water Resources" in Columbia-North
Pacific Comprehensive Framework Study of
Water and Related Lands. Vancouver,
Washington, Pacific Northwest River Basins
Comm., app. 5.
SLuzier, J.E. and R.J. Burt. 1974. "Hydology of
Basalt Aquifers and Deplation of Ground Water in
East-Central Washington." Washington Depart-
ment of Ecology, Water-Supply Bulletin 33.
9Gutentag, E.D. and J.B. Weeks. 1980. "Water
Table in the High Plains Aquifer in 1978 in
Parts of Colorado, Kansas, Nebraska, New
Mexico, Oklahoma, South Dakota, Texas,
and Wyoming." U.S. Geological Survey
Hydrologic Investigation Atlas 642.
10Weeks, J.B. and E.D. Gutentag. 1981. "Bedrock
Geology, Altitude of Base, and 1980
Saturated Thickness of the High Plains
Aquifer in Parts of Colorado, Kansas,
Nebraska, New Mexico, Oklahoma, South
Dakota, Texas, and Wyoming." U.S.
Geological Survey Hydrologic Investigations
Atlas 648.
nLuckey, R.R., E.D. Gutentag, and J.B. Weeks.
1981. "Water-Level and Saturated-Thickness
Charges Predevelopment to 1980, in the
High Plains Aquifer in Parts of Colorado,
Kansas, Nebraska, New Mexico, Oklahoma,
South Dakota, Texas, and Wyoming." U.S.
Geological Survey Hydrologic Investigations
Atlas 652.
12 U.S. Geological Survey. 1970. The National Atlas of
the United States of America.
13 Johnston, R.H., H.G. Healy, and L.R. Hayes. 1981.
"Potentiometric Surface of the Tertiary
Limestone Aquifer System, Southeastern
United States, May 1980." U.S. Geological
Survey Open-File Report 81-486.
Bibliography
A large number of publications were consulted, for
both general and specific information, in the prepara-
tion of this paper. Specific reference to these publica-
tions generally is omitted in the text, both to avoid in-
terruption of the discussions and to save space.
Publications that served as primary references are listed
below, under the categories of general references and
references to regional discussions. General references
include publications that were used both for
background information on the classification of
ground-water systems and for general information on
the regions. References to the regional discussions in-
clude publications that served as a source of additional
information on the individual regions.
General Bibliography
Fenneman, N.M., Physiography of Western United
States. McGraw-Hill, New York, 1931.
, Physiography of Eastern United
States.McGraw-Hill, New York, 1938.
Fuller, M.L., "Underground Waters of Eastern United
States." U.S. Geological Survey Water-Supply Paper
114, 1905.
Heath, R.C., "Classification of Ground-Water Systems
of the United States", Ground Water, v. 20, no. 4,
July-August 1982.
Mann, W.B., IV, et al, "Estimated Water Use in the
United States, 1980." U.S. Geological Survey Circular
1001, 1983.
McGuiness, C.L., "The Role of Ground Water in the
National Water Situation." U.S. Geological Survey
Water-Supply Paper 1800, 1963.
Meinzer, O.E., "The Occurrence of Ground Water in
the United States, With a Discussion of Principles."
U.S. Geological Survey Water-Supply Paper 489,
1923.
Shimer, J.A., Field Guide to Landforms in the United
States. Macmillan, New York, 1972.
Thomas, H.E., The Conservation of Ground Water.
McGraw-Hill, New York, 1951.
, "Ground-Water Regions of the United
States—Their Storage Facilities," v. 3 of The Physical
and Economic Foundation of Natural Resources.
U.S. 83d Cong. House Committee on Interior and
Insular Affairs, pp 3-78, 1952.
80
-------
U.S. Geological Survey, The National Atlas of the
United States of America. 1970.
, Professional Paper 813, "Summary
Appraisals of the Nation's Ground-Water
Resources." Published as a series of chapters based
on the boundaries established by the United States
Water-Resources Council for Water-Resources
Regions in the United States.
Bibliographies for Regional Discussions
2. Alluvial Basins
Harshbarger, J.W., D.D. Lewis, H.E. Skibitzke, W.L.
Heckler, and L.R. Kister, "Arizona Water" (rev. by
H.L. Baldwin) U.S. Geological Survey Water-Supply
Paper 1648, 1966.
Robinson, T.W. "Big Smoky Valley, Nevada," chap. 8
of Subsurface Facilities of Water Management and
Patterns of Supply-Type Area Studies, \. 4 of The
Physical and Economic Foundation of Natural
Resources. U.S. 83d Cong. House Committee of
Interior and Insular Affairs, pp 132-146, 1953
3. Columbia Lava Plateau
Columbia-North Pacific Technical Staff, "Water
Resources," in Columbia-North Pacific
Comprehensive Framework Study of Water and
Related Lands. Pacific Northwest River Basins
Comm., Vancouver, Washington, app. 5, 1970.
Hampton, E.R., "Geologic Factors that Control the
Occurrence and Availability of Ground Water in the
Fort Rock Basin, Lake County, Oregon." U.S.
Geological Survey Professional Paper 383-B, 1964.
Luzier, I.E. and R.J. Burt, "Hydrology of Basalt
Aquifers and Depletion of Ground Water in East-
Central Washington." Washington Department of
Ecology, Water-Supply Bulletin 33, 1974.
MacNish, R.D. and R.A. Barker, "Digital Simulation
of a Basalt Aquifer System, Walla Walla River Basin,
Washington and Oregon." Washington Department
of Ecology, Water-Supply Bulletin 44, 1976.
Nace, R.L., "Hydrology of the Snake River Basalt,"
Washington Academy of Science Journal, v. 48, no.
4, pp. 136-138, 1958.
Newcomb, R.C., "Storage of Ground Water Behind
Subsurface Dams in the Columbia River Basalt,
Washington, Oregon, and Idaho." U.S. Geological
Survey Professional Paper 383-A, 1962.
, "Geology and Ground-Water Resources of
the Walla Walla River Basin, Washington-Oregon."
Washington Division of Water Resources, Water-
Supply Bulletin 21, 1965.
4. Colorado Plateau and Wyoming Basin
Harshbarger, J.W., C.A. Repenning, and J.T.
Callahan, "The Navajo Country, Arizona-Utah-New
Mexico," chap. 7 of Subsurface Facilities of Water
Management and Patterns of Supply-Type Area
Studies, v. 4 of The Physical and Economic
Foundation of Natural Resources. U.S. 83d Cong.
House Committee on Interior and Insular Affairs,
pp 105-129, 1953.
Lohman, S.W., "Geology and Artesian Water Supply
of the Grand Junction Area, Colorado." U.S.
Geological Survey Professional Paper 451, 1965.
5. High Plains
Gaum, C.H., "High Plains, or Llano Estacado, Texas-
New Mexico," chap. 6 of Subsurface Facilities of
Water Management and Patterns of Supply-Type
Area Studies, v. 4 of The Physical and Economic
Foundation of Natural Resources. U.S. 83d Cong.
House Committee on Interior and Insular Affairs,
pp 94-104, 1953.
Gutentag, E.D., and J.B. Weeks, "Water Table in the
High Plains Aquifer in 1978 in Parts of Colorado,
Kansas, Nebraska, New Mexico, Oklahoma, South
Dakota, Texas, and Wyoming." U.S. Geological
Survey Hydrologic Investigations Atlas 642, 1980.
Lohman, S.W. "High Plains of West-Central United
States, General Aspects," chap. 4 of Subsurface
Facilities of Water Management of Patterns of
Supply-Type Area Studies, v. 4 of The Physical and
Economic Foundation of Natural Resources. U.S.
83d Cong. House Committee on Interior and Insular
Affairs, pp 70-78, 1953.
Luckey, R.R., E.D. Gutentag, and J.B. Weeks,
"Water-Level and Saturated-Thickness Changes,
Predevelopment to 1980, in the High Plains Aquifer
in Parts of Colorado, Kansas, Nebraska, New
Mexico, Oklahoma, South Dakota, Texas, and
Wyoming." U.S. Geological Survey Hydrologic Atlas
652, 1981.
Weeks, J.B., and E.D. Gutentag, "Bedrock Geology,
Altitude of Base, and 1980 Saturated Thickness of the
High Plains Aquifer in Parts of Colorado, Kansas,
Nebraska, New Mexico, Oklahoma, South Dakota,
Texas, and Wyoming." U.S. Geological Survey
Hydrologic Investigations Atlas 648, 1981.
7. Glaciated Central Region
Feth, J.H., et al, "Preliminary Map of the
Conterminous United States Showing Depth to and
Quality of Shallowest Ground Water Containing
More Than 1,000 Parts per Million Dissolved Solids."
U.S. Geological Survey Hydrologic Atlas 199, 1965.
Todd, D.K., Groundwater Hydrology, 2d ed. John
Wiley, New York, 1980.
8. Piedmont and Blue Ridge Region
LeGrand, H.E., "Ground Water of the Piedmont and
Blue Ridge Provinces in the Southeastern States."
U.S. Geological Survey Circular 538, 1967.
Le Grand, H.E., and M.J. Mundorff, "Geology and
Ground Water in the Charlotte Area, North
81
-------
Carolina." North Carolina Department of
Conservation and Development, Bulletin 63, 1952.
Stewart, J.W., "Infiltration and Permeability of
Weathered Crystalline Rocks, Georgia Nuclear
Laboratory, Dawson County, Georgia." U.S.
Geological Survey Bulletin 1133-D, 1964.
9. Northeast and Superior Uplands
Delaney, D.F., and A. Maevsky, "Distribution of
Aquifers, Liquid-Waste Impoundments, and
Municipal Water-Supply Sources, Massachusetts.'
U.S. Geological Survey Water-Resources
Investigations Open-File Report 80-431, 1980.
10. Atlantic and Gulf Coastal Plain
Back, W., "Hydrochemical Facies and Ground-Water
Flow Patterns in the Northern Part of Atlantic
Coastal Plain." U.S. Geological Survey Professional
Paper 498-A, 1966.
Brown, G.A., and OJ. Cosner, "Ground-Water
Conditions in the Franklin Area, Southeastern
Virginia." U.S. Geological Survey Hydrologic Atlas
538, 1975.
Cohen, P., O.L. Franke, and B.L. Foxworthy, "An
Atlas of Long Island's Water Resources." New York
Water Resources Commission Bulletin 62, 1968.
Gabrysch, R.K., "Approximate Land-Surface
Subsidence in the Houston-Galveston Region, Texas,
1906-78," 1943-78, and 1973-78. U.S. Geological
Survey Open-File Report 80-338, 1980.
LeGrand, H.E., and W.A. Pettyjohn, "Regional
Hydrogeologic Concepts of Homoclinal Flanks,"
Ground Water, v. 19, no. 3, May-June, 1981.
11. Southeast Coastal Plain
Cooper, H.H., Jr., and W.E. Kenner, "Central and
Northern Florida," chap. 9 of Subsurface Facilities of
Water Management and Patterns of Supply-Type
Area Studies, v. 4 of The Physical and Economic
Foundation of Natural Resources. U.S. 83d Cong.
House Committee on Interior and Insular Affairs,
pp 147-161, 1953.
Heath, R.C., and C.S. Conover, "Hydrologic Almanac
of Florida." U.S. Geological Survey Open-File
Report 81-1107, 1981.
Johnston, R.H., H.G. Healy, and L.R. Hayes,
"Potentiometric Surface of the Tertiary Limestone
Aquifer System, Southeastern United States, May
1980." U.S. Geological Survey Open-Fill Report
81-486, 1981.
Stringfield, V.T., "Artesian Water in Tertiary
Limestone in the Southeastern States." U.S.
Geological Survey Professional Paper 517, 1967.
12. Alluvial Valleys
Boswell, E.H., E.M. Gushing, and R.L. Hosman,
"Quarternary Aquifers in the Mississippi
Embayment." U.S. Geological Survey Professional
Paper 448-E, 1968.
Rorabargh, M.I., "Ground Water in Northeastern
Louisville, Kentucky." U.S. Geological Survey Water-
Supply Paper 1360-B, pp 101-169, 1956.
82
-------
Chapter 4
Ground Water-Surface Water Relationship
Introduction
The interrelations between ground water and surface
water are of great importance in both regional and
local hydrologic investigations and a wide variety of in-
formation can be obtained by analyzing streamflow
data. Most commonly the surface water investigator
deals with stream hydrographs, channel characteristics,
geomorphology, or flood routing. Although the hydro-
geologist may evaulate induced infiltration into a
streamside aquifer, he is generally more interested in
aquifer characteristics, such as hydraulic conductivity,
thickness, boundaries, and well yields. Many hydrolo-
gists tend to ignore the fact that, at least in humid
areas, ground-water runoff accounts for a significant
part of a stream's total flow.
The evaluation of the ground-water component of
runoff can provide important and useful information
regarding regional recharge rates, aquifer charac-
teristics, and ground-water quality, and can indicate
areas of high potential yield to wells. The purpose of
this chapter is to describe a number of techniques that
can be used to evaluate runoff to obtain a better
understanding and evaluation of ground-water
resources. In particular the following will be examined:
1. Ground-water runoff
2. Surface runoff
3. Regional ground-water recharge rates
4. Determination of areas of relatively high permeabil-
ity or water-yielding characteristics
5. Determination of the background concentration of
ground-water quality
6. Estimation of evapotranspiration
7. Determination of the percentage of precipitation
that is evaportranspired, becomes ground-water
runoff, or becomes surface water runoff.
The approaches taken, admittedly some highly sub-
jective, are based on: (1) short-term runoff events, (2)
long-term hydrographs, and (3) dry-weather flow mea-
surements. In the first approach a single event, such as
a flood wave of a few hours or few days duration, can
be analyzed, while the latter two approaches are based
on annual stream hydrographs, flow-duration curves,
or seepage runs. Short-term events may provide a con-
siderable amount of information for a local area, while
long-term events are most useful for regional studies.
Streamflow may consist of several components in-
cluding ground-water runoff, surface runoff, effluent,
and precipitation that falls directly into the channel.
Figure 93. Approximate Flow Pattern in Uniformly Permeable Material Between the Sources Distributed Over the Air-Water Interface
and The Valley Sinks (After Hubberl, 1940)
83
-------
MAM
^ Middle Loup River
Figure 94. Hydrographs of Two Nearby Streams
A S O
White River
The volume of water that is added by precipitation
directly into the channnel is relatively small compared
to the stream's total flow. The contribution by waste
effluent may or may not be significant, since it depends
on the activities that are occurring in the basin. In
permeable basins in humid regions, ground-water
runoff may account for 70 or 80 percent of the
stream's annual discharge. The remainder is surface
runoff, which originates as precipitation or snow melt
that flows directly into the stream channel. This
chapter is concerned largely with ground-water runoff
and surface runoff and the separation of these two
components.
In order to fully appreciate the origin and signif-
icance of ground-water runoff, it is first necessary to
examine the regional ground-water flow system. Figure
93 illustrates a typical flow pattern. Particularly in
humid and semi-arid regions, the water table generally
conforms with the surface topography. Consequently,
the hydraulic gradient or water table slopes away from
divides and topographically high areas toward adjacent
low areas, such as streams and rivers. Topographic
highs and lows, therefore, serve as recharge and
discharge areas, respectively.
Ground-water flow systems may be local, intermedi-
ate, or regional. As these terms imply, ground-water
flow paths may be short, amounting to a few yards at
one extreme to many miles in the regional case. In-
dividual flow lines are, of course, influenced by the
stratigraphy and, in particular, are controlled by
hydraulic conductivity.
As water inflitrates a recharge areas, the mineral
content is relatively low. The quality changes, however,
along the flow path and dissolved solids, as well as
several other constituents, generally increase with in-
creasing distances traveled in the subsurface. It is for
this reason that even nearby streams may be typified by
different chemical quality. A stream, seep, or spring in
a local discharge area may be less mineralized than that
issuing from a regional discharge zone because of the
increase in mineralization that takes place along longer
flow paths. It must be remembered, however, that
other conditions, such as soil type, solubility of the
enclosing rocks, surface drainage characteristics, and
waste disposal practices, may have a profound effect
on water quality at any particular site.
Even streams in close proximity may differ con-
siderably in discharge even though the size of the
drainage area and climatic conditions are similar.
Figure 94 gives the superimposed hydrographs of White
River in southwestern South Dakota and the Middle
Loup River in northwestern Nebraska which are good
examples. White River has a low discharge throughout
most of the year, but from May to September, flash
floods are common. The wide extreme in discharge is
characteristic of a flashy stream.
The flow of Middle Loup River is nearly constant,
although from late spring to early fall higher flows
may occur. These peaks, however, differ considerably
from those found in White River because the increase
in discharge takes place over a longer interval, the
stage does not range widely, and the recession occurs
more slowly. The differences in hydrographs of these
two nearby rivers is puzzling, until the geology and
topography of their respective basins are examined.
White River flows through the Badlands of South
Dakota, an area of abrupt changes in relief, steep
slopes, little vegetative cover, and rocks that consist
largely of silt and clay, both of which may contain an
abundance of bentonite. When wet, bentonite, a swell-
ing clay, increases greatly in volume. As a result of
these features, rainfall in the White River basin runs
84
-------
off quickly and there is little opportunity for inflitra-
tion and ground-water recharge to occur. Thus, intense
rainstorms cause flash floods, such as those that oc-
curred in June, August, and September.
The Middle Loup basin is carved into the undulating
grassland topography of the sandhills of Nebraska
where surficial materials consist of wind blown sand.
Since the low relief, grass-covered surface promotes in-
filtration, precipitation is readily absorbed by the
underlying sand. As a result, there is very little surface
runoff and a great amount of inflitration and ground-
water recharge. The ground water slowly migrates to
the river channel, thus providing a high sustained flow.
In a comparison of the hydrographs of these two
rivers, it is evident that the geologic framework of the
basin serves as a major control on runoff. This further
implies that in any regional hydrologic study, the in-
vestigation should begin with an examination of
geologic maps.
Gaining and Losing Streams
Although the discharge of most streams increases
downstream, the flow of some streams diminishes.
These streams are referred to as gaining or losing,
respectively. The hydrologic system, however, is even
more complex, because a stream that may be gaining in
one season, may be losing during another. Further-
more, various activities of man may also affect a
stream's discharge.
Under natural conditions a gaining stream is one
where the water table is above the base of the stream
channel. Of course the position of the water table fluc-
tuates throughout the year in response to differences in
ground-water recharge and discharge. Normally the
water table is highest in the spring, which is the annual
major period of ground-water recharge. From spring to
fall, very little recharge occurs and the amount of
ground-water in storage is slowly depleted as it seeps
into the stream. Eventually, the water table may
decline to the same elevation as a stream bottom, or
even below it, at which time stream flow ceases except
during periods of surface runoff. Following a period of
recharge, caused either by inflitration of rainfall or
seepage from a flood wave, the water table may again
rise and temporarily contribute ground-water runoff.
Figure 95 shows a generalized diagram of the
hydrology of a stream during two seasons of the year.
During the spring, the water table is high and the gra-
dient dips steeply towards the stream. If stream flow
was measured at selected points, it would be found that
the discharge increases downstream because of the ad-
dition of ground-water runoff. That is, it is a gaining
stream. In the fall when the water table lies at or below
the stream bottom, however, the same stream might
become a losing stream. During a major runoff event
the stage in the stream would be higher than the adja-
cent water table and water would migrate from the
stream into the ground. The stream would continue to
lose water until the water table and river stage were
equal. When the stage declined, ground-water runoff
would begin again.
In this case the stream changed from gaining to los-
ing and back again to gaining. Similar situations may
occur over longer intervals, such as during droughts.
As a drought continues, the water table slowly declines
as ground-water storage is depleted. A period of high
flows, such as release from a dam, may cause tremen-
Losing stream
(A-A')
Gaining in spring
Losing in fall
(B-B')
Gaining stream
(C-C')
Ephemeral
A
Perennial
C
Land surface
Water table
(S) in spring
(F) in fall
Figure 95. The Relation Between the Water Table and Stream Types
85
-------
dous amounts of water to flow from the channel back
into the ground, thus saturating the depleted stream-
side deposits. It may require weeks of high flow to
replenish the ground-water reservoir and until this is
accomplished, the stream will be losing.
Some streams, particuarly in arid and karst regions,
are nearly always losing. Examples include those chan-
nels that cross coarse-grained alluvial fans. Even during
flash floods, the great mass of flood water soon
spreads out over the fan or adjacent desert to inflitrate
or evaporate.
Because of the extensive network of solution open-
ings in karst terrain, the water table may consistently
lie below the bottom of all the streams. During a
period of runoff, the water may rapidly flow into sink
holes and solution openings or simply disappear into a
swallow hole in a stream channel, only to appear again
perhaps several miles downstream.
Gaining and losing streams can also be created ar-
tifically. Where well fields lie along stream channels
and induce water to flow from the stream to the well,
stream flow is diminished. In some cases stream deple-
tion by pumping wells has proceeded to such an extent
that the stream channels are dry throughout the year.
Conversely, in some irrigated regions, so much infiltra-
tion occurs that the water table rises to near land sur-
face. The underlying soil and ground water may
become highly mineralized by the leaching of soluble
salts. These highly mineralized waters may discharge
into a stream, increasing its flow but deteriorating the
chemical quality. In other places, municipal or in-
dustrial wastes may add considerably to a stream's
flow, also deteriorating its quality. In fact, at certain
times of the year, the entire flow may consist of waste
water.
Bank Storage
Figure 96 shows that, as a flood wave passes a par-
ticular stream cross-section, the water table may rise in
the adjacent stream-side deposits. The rise is caused by
two phenomena. First, the stream stage, which is
higher than the water table, will temporarily block
ground-water runoff, thus increasing the amount of
Land surface
(D
C
ID
O!
18
(3
13
11
ABC D E
I I I I
0 150
Began 1700 hours
13
«_ 11
CD
O)
O
200 400 600 800
Horizontal Distance, in Feet
Land Surface
1000
1200
_4
1 1 — >
3 2 1
Well number
Gage
Silt and clay
1 1 1 1 1
Sand
i i i i i
^xl
1
• I
Stage D
Stage E
0
200
400 600 800
Horizontal Distance, in Feet
Figure 96. Movement of Water Into and Out of Bank Storage Along a Stream in Indiana
1000
1200
86
-------
ground water in storage. Secondly, because of the in-
creased head in the stream, water will flow from the
channel into the ground, thus providing another com-
ponent of water added to storage.
Once the flood waves begin to recede, which may oc-
cur quite rapidly, the newly added ground water begins
to flow back into the channel, rapidly at first and then
more slowly as the gradient decreases. This temporary
storage of water in the near vicinity of the stream
channel is called bank storage.
The rising and recession limbs of a hydrograph of a
flood wave should provide clues concerning bank
storage and stream-side permeability. For example,
where stream-side deposits are of low permeability,
such as clay or shale, the rising limb should be quite
steep, but more gradual where the deposits are
permeable. Since there would be little or no bank
storage in the first case, recession curves also should be
steep, but the release from bank storage in a permeable
basin should reduce the slope of the recession curve.
Effect of the Geologic Framework on Stream
Hydrographs
Unfortunately, the discharge of ground water into a
stream is not always as simple as has been implied from
the above examples. As Figure 97 shows, an examina-
tion of the aquifer framework and its effect on a stream
hydrograph is enlightening. Notice in Figure 97a, that
the stream channel is deeply cut into a shale that is
overlain by sand. Ground water flows into the stream
along a series of springs and seeps issuing at the sand-
shale contact. During a runoff event the stream stage
rises but, even at its peak, the stage remains below the
top of the shale. In this case, the contribution of
ground water remains constant despite the rise in stage.
To separate the ground-water runoff component from
the stream hydrograph, one merely needs to draw a
straight line from the inflection points of the rising and
falling limbs.
In Figure 97b the stream channel is cut into a deposit
of sand that is underlain by shale. Ground water flows
into the stream, but as the stage rises, ground-water
runoff decreases and eventually stops. Surface water
then begins to flow into the ground where it is retained
as bank storage. As the stage declines, ground water
again starts to discharge into the channel eventually pro-
viding the entire flow. This is the classic case of bank
storage. Hydrograph separation is more difficult in this
case.
Figure 97c is a combination of the previous two ex-
amples. Ground water from a perched aquifer con-
tributes a steady flow, while bank storage is gained and
then released from the stream-side aquifer. Hydrograph
separation is even more difficult in this situation because
of the contribution from both aquifers.
The final example, Figure 97d, consists of three
aquifers—one perched, one in direct contact with the
stream, and one deeper, artesian aquifer. As the stream
rises, there is a decrease in the head difference between
the stream and the artesian aquifer. The decrease in
head difference will reduce upward leakage from the
artesian aquifer, the amount depending on the thickness
and vertical permeability of the confining bed and the
head difference.
Single-Event Hydrograph Separation Techniques
Following a runoff event, the water held as bank
storage begins to discharge into the channel. In the
beginning the rate of bank storage discharge is high
because of the steep water-level gradient, but as the gra-
dient decreases so does ground-water runoff. The reces-
sion segment of the stream hydrograph gradually tapers
off into what is called a depletion curve. To a large ex-
tent, the shape of the depletion curve is controlled by
the permeability of the stream-side deposits, although
soil moisture and evapotranspiration may play important
roles.
Depletion Curves
Intervals between surface runoff events are generally
short and for this reason, depletion curves are plotted as
a combination of several arcs of the hydrograph with
the arcs overlapping in their lower parts, as shown in
Figure 98. To plot a depletion curve, tracing paper is
placed over a hydrograph of daily flows and, using the
same horizontal and vertical scales, the lowest arcs of
the hydrographs are traced, working backward in time
from the lowest discharge to a period of surface runoff.
The tracing paper is moved horizontally until the arc of
another hydrograph coincides in its lower part with the
arc already traced; this arc is plotted on top of the first.
The process is continued until all the available arcs are
plotted on top of one another.
The upward curving parts of the individual arcs are
disregarded because presumably they are affected by
channel storage or surface runoff, or both. The
resulting continuous arc is a mean or normal depletion
curve that presumably represents the hydrograph that
would result from ground-water runoff alone during a
protracted dry period.
Even for the same stream, there may be appreciable
differences in the shape of the depletion curve at dif-
ferent times of the year. This is largely due to evapora-
tion, transpiration, and temperature effects. In cases
such as these, a family of depletion curves may be con-
structed. One curve should represent winter when there
is little or no evapotranspiration, another curve should
represent the summer when evapotranspiration is at its
maximum, and perhaps a third curve should be
prepared to represent intermediate conditions.
Depletion curves are the basis for estimating ground-
water runoff during periods of surface runoff. They also
shed a great deal of light on the characteristics of a
ground-water reservoir.
Hydrograph Separation
A flood hydrograph is a composite hydrograph con-
sisting of surface runoff superimposed on ground-water
87
-------
oo
oo
d.
Figure 97. The Aquifer Framework in the Vicinity of a Stream Plays a Major Role in Ground-Water Runoff and Hydrograph Separation
-------
40 60
Time, in Days
Figure 98. Ground-Water Depletion Curves have Different
Shapes that Reflect the Seasons14
runoff. When attempting to separate these two com-
ponents of flow, however, some problems generally oc-
cur. Whatever method is employed, there is always some
question as to the accuracy of the division. One can on-
ly say that in any given case ground-water runoff is
probably not less than about "x" or more than about
"y". Keeping in mind the complexities of a runoff
hydrograph brought about by variable parameters, and
particularly the geology of the basin, an attempt will be
made to develop some logical methods for hydrograph
separation.
Using the flood hydrograph in Figure 99a, we can see
that point A represents the start of surface runoff. Us-
ing a previously prepared depletion curve, the original
recession can be extended to B. The area below AB
represents the ground-water runoff that would have oc-
curred had there been no surface runoff. Point D
represents the end of surface runoff. A depletion curve
can be matched with the recession limb, extending it
from D to C. A partial envelope has now been formed
that shows the upper and lower limits between which a
line may reasonably be drawn to separate the two com-
ponents of runoff. This assumption ignores possible ef-
fects brought about by difference in the geologic
framework. This envelope forms a basis for the most
commonly used separation methods which are described
below.
Method 1. Using a depletion curve and starting at D in
Figure 99b, extend the recession curve back to a line
drawn vertically through the peak of the hydrograph
(C). A second line is then extended from A, the start of
surface runoff, to C. This method is more likely to be
valid where ground-water runoff is relatively large and
reaches the stream quickly.
Not uncommonly, the end of surface runoff is dif-
ficult to determine, but point D can be estimated by
means of the equation
N = A0-2
(48)
where N = the number of days after a peak when sur-
face runoff ceases and A = the basin area in square
miles. The distance N is measured directly on the
hydrograph.
Method 2. In this example in Figure 99b, separation is
accomplished merely by extending a straight line,
originating at the start of surface runoff (A), to a point
on the recession curve representing the end of surface
runoff (D). This method of separation is certainly the
simplist and is justifiable if little is known about the
aquifer framework.
Method 3. In this example, also in Figure 99b, the pre-
runoff recession line is extended from A to a point
directly under the hydrograph peak (B). From this point
a second line is projected to D, the end of surface
runoff.
The separation technique to be employed should be
based on knowledge of the hydrogeology of the basin,
keeping in mind the affect of the geologic framework
on the hydrograph.
Separation of Complex Hydrographs
Commonly runoff events occur at closely spaced in-
tervals and there is insufficient time for the recession
curve to develop before runoff again increases. This
complicates hydrograph separation.
Figure 99c shows two methods that can be used to
determine ground-water runoff under a complex
hydrograph, which represents two storms.
Method 1. The recession curve preceding the first runoff
event is continued to its intersection with a line drawn
through the first peak (A-B). The distance N is
calculated and measured. The recession limb of the first
event is continued to its intersection with the N-days
line (C-D). Line B-D is then constructed. The first
recession trend is continued to its intersection with a line
drawn through the peak of the second runoff event (C-
D-E). From this point (E), the line is extended N days
to F.
89
-------
a.
Peak
b.
Figure 99. Separation of the Stream Hydrograph15'16
90
Method 2. As Figure 99c shows, the easiest method is to
project a straight line from A to F. Although by far the
simplest, this technique is not necessarily any less ac-
curate than Method 1.
Hydrograph Separation by Chemical Techniques
Generally ground water is more highly mineralized
than surface runoff. During baseflow the streams
natural quality is at or near its maximum concentration
of dissolved solids, but as surface runoff reaches the
channel and provides an increasing percentage of the
flow, the mineral content is diluted. Following the
discharge peak, surface runoff diminishes, ground-water
runoff increases, and the mineral content again
increases.
Several investigators have used the relation between
runoff and water quality to calculate the ground-water
contribution from one or more aquifers or to measure
streamflow. This method of hydrograph separation,
which requires the solution of a series of simultaneous
equations, is based on the concentration of a selected
chemical parameter that is characteristics of ground
water and surface runoff.
The basic equations, which may take several forms,
are as follows:
+ QS =
(49)
C QQ + CSQS = CQ (50)
where Q , Qs, and Q are ground-water runoff, surface
runoff, and total runoff, respectively; and C , Cs, and
C represent the concentration of dissolved mineral
species or specific conductance of ground water, surface
runoff, and total runoff, respectively. Usually specific
conductance is used as the C parameter because of the
relative ease of obtaining it.
If Cg, Cs, C, and Q are known we can determine the
quantity of ground-water runoff as follows:
Qa = Q(C-Cs)/(Ca-Cs)
(51)
C is determined by measuring the specfic conductance
in a well, in a series of wells, or during baseflow. The
quality of surface runoff, Cs, is obtained from analysis
of overland flow or, possibly in the case of small
streams, at the period of peak discharge when the entire
flow consists of surface runoff. It is assumed Cg and Cs
are constant. C and Q are measured directly.
Visocky17 used continuous recording equipment to
measure specific conductance and stage (water level) in
the Panther Creek Basin in north-central Illinois. By
using the equations given above, he calculated the
ground-water runoff component of the stream on the
basis of the relationship between discharge and specific
conductance. He also calculated and compared ground-
water runoff as determined from a ground-water rating
curve and found that the chemical method provided a
lower estimate under normal conditions than did the
rating curve technique. On the other hand, the
chemical method indicated more ground-water runoff
following storms that were preceeded by extended dry
periods, which had caused considerable declines in
water level in nearby observation wells.
-------
Q + Q d = Q
CQa + CQ°d = CQ
Qa = ? Ca=50
Q=18 C =43
50Qa+10Qsd =
-10Qa-10Qsd= -180
50Qa + 10Qsd = 774
400=594
Q, = 14.85 fs
Q =
* "CT^Q
sd
OR Q? = 18(43-10) = 594 =1485cfs
50-10 40
Figure 100. Contribution to Econfina Creek During a Period of
Dry Weather Flow When the Stream Discharge was 18 cfs and
the Dissolved Solids Concentration was 43 mg/l:
From Sand Aquifer = 3.15 cfs
From Limestone Aquifer = 14.85 cfs
Surface runoff
Glacial till
Figure 101. Four Mile Creek, Iowa
During baseflow, the quantity of ground-water
discharge from surficial sand and from limestone in the
Floridan artesian aquifer into Econfina Creek in north-
west Florida was distinguished by Toler.18 In this case,
as Figure 100 shows, the artesian water had a dissolved
solids content of 50-68 mg/l, while that from the sur-
ficial sand was only 10-20 mg/l. The artesian water
discharged through a series of springs along the central
part of the basin and amounted to 70 to 75 percent of
the stream's baseflow. The equation used for this
analysis is as follows:
= (C-Csd)/(Ca-Csd)Q
(52)
runoff and C^, Ca,
where Qa = artesian runoff, Q =
and C represent the dissolved solids in water from the
sand, the artesian aquifer, and during any instant in
91
-------
the stream, respectively. Of course,
Q - Qa = Qsd
(53)
Continuous streamflow and conductivity
measurements were collected at a gaging station on
Four Mile Creek in east-central Iowa by Kunkle.19 The
basin above the gage, which contains 19.5 square miles,
consists largely of till that is capped on the uplands by
loess. As Figure 101 shows, the stream lies in a valley
that contains as much as 30 feet of permeable
alluvium. Ground water from the alluvium and loess,
as well as the stream during low flow, has an average
specific conductance of 520 micromhos (Cg) while sur-
face runoff is about 160 micromhos (Cs).
Figure 102 shows continuous record of discharge and
conductivity representing a storm in September 1960.
Instantaneous ground-water runoff during this event
was calculated for several points under the hydrograph
by using the following formulas:
Qg +
CQ
gg
CSQS = CQ
(49)
(50)
where Q. = ground-water runoff, Qs = surface
runoff, Q = runoff and Cg, Cs, and C = specfic con-
ductance of ground-water, surface runoff, and runoff,
respectively. As determined from the graphs in Figure
1000
Ground-water runoff
computed from
conductivity
22 23 24 25 26 27 28
SEPTEMBER
29 30
1 2 3
OCTOBER
Figure 102. Hydrographs Showing Water Discharge, Specific
Conductance, and Computed Ground-Water Runoff in Four Mile
Creek near Traer, Iowa, September and October 1963.19
102, where Q = 2.3 cfs, C = 410; Cg = 520 and Cs
= 160, Q can be calculated as follows:
Q
Qs = Q
-160Qg - 160 Qs = -368
520 Q + 160 Qs = CQ
520Qg+160Qs= 943
360g 575
Qg = 1.6 cfs
Therefore, when the stream's discharge (Q) was 2.3 cfs,
ground-water runoff was 1.6 cfs. This calculation pro-
vides one point on the hydrograph. Several other
points need to be determined so that a separation line
can be drawn.
7503230800 DEER CHEEK BT MOUNT STERLING, OHIO
5 DRY FIXED INTERVBL
226.0 SO.MI.
»
TOTBL DISCHARGE 9.335E 9 CF OR 17.62
GROUND HBTER RUNOFF 5.225E 9 CF OR 9.66
GROUND HBTER HS 'I. 56.0
RECHRRGE RBIE 169000 GPD /SO. HI.
INCHES
INCHES
OHTS
7503230800 DEER CREEK BT MOUNT STERLING. OHIO
5 DBT SLIDING INTERVfll
226.0 SO.HI.
5-
TOTBL OISCHBRGE 9.335E 9 CF 8R 17.62
GHOUNO HRTER RUNOFF U.915E9 CF OR 9.3*4
GROUND HBTER BS 'I. 53.0
RECHBRGE SBTE imOOO GPO /SO. Ml.
INCHES
INCHES
—r 1 1 1 1 1 1 1
• M « M l*» im IM 111 1U
DflTS
7503230800 DEER CREEK BT HOUNT STERLING, OHIO
5 OBT LOCRL NINIMB
-5
-:S
228.0 SO."!.
r't
*•;•
TOTBL DISCHflRGE 9.33SE 3 CF OR 17.62 INCHES
GROUND WRTER RUNOFF U.318E 9 CF OH 8.21 INCHES
GROUND HBTER BS 1. 16.6
RECHBRGE RBTE 390000 GPO /SO. MI.
Figure 103. Deer Creek Hydrographs Separated by Three
Methods and Statistical Data20
92
-------
Computer Separation Programs
Various methods of hydrograph separation have
been described, all of which are laborious, time con-
suming, and commonly, open to questions of accuracy
and interpretation. In each case a mechanical technique
is used to provide a number of points on a hydrograph
through which a line can be drawn that separates
ground-water runoff from surface runoff. Once this
line is determined, one must measure, directly on the
hydrograph, the daily components of streamflow and
then sum the results.
Annual ground-water runoff divided by total dis-
charge provides the percentage of streamflow that con-
sists of ground water. Effective ground-water recharge
is that quantity of precipitation that infiltrates, is not
removed by evapotranspiration, and eventually dis-
charges into a stream.
Effective ground-water recharge rates can be easily
estimated with a computer program described by Petty-
john and Henning.20 This program separates the
hydrograph by three different methods, provides
monthly recharge rates and an annual rate, produces a
flow-duration table, and gives the operator the option
of generating with a line printer the separated hydro-
graph and a flow-duration curve; as illustrated in
Figure 103. The data base is obtained from annual
streamflow records, which are published by the U.S.
Geological Survey. The computer program will operate
on a mainframe or microcomputer.
Ground-Water Rating Curve
A widely used technique to measure streamflow is
the surface water rating curve, which shows the rela-
tion between stage and discharge. Figure 104 shows a
similar curve that shows the relation between the water
table and streamflow called a ground-water rating
curve. Prepared for those aquifer-stream systems that
are hydrologically connected, the ground-water rating
curve can be used to separate ground-water runoff
from a stream hydrograph.
To prepare the curve, synchronous water table and
stream discharge measurements are required. Ground-
water levels are obtained either from: (1) a series of
wells spread throughout the basin, (2) a series of wells,
each of which represents an area of similar geology, or
(3) a single near-stream well. Wells influenced by
pumping should not be used. If more than one well is
used, water levels, referred to some datum such as sea
level, must be averaged to form a composite hydro-
graph. Furthermore, measurements of both ground
water and stream stage should be made only during
rainless intervals when streamflow consists entirely of
ground-water runoff. Selected water-level measure-
ments are plotted on a graph with the mean daily
streamflow and a smooth curve is drawn through the
points.
The graph is used by determining, either from in-
dividual measurements or water-level recorder data, the
ground-water stage, reading across the graph to the
Mean Ground-Water Stage, in Feet Above Mean Sea Level
*-*••*••*»•.[*.&.*> *
,_. rou.^.^^^jo,^
•A
t
- ../_
/
10°
r
i
"V
,/f'"
A
.«,
\
^~*r-
o2 ^^^*°~
^^*u !
^
Explanation
• 1950
X 1951
0 1952
Numt>«f indicates calendar month January**]
•
^- —
10 20 30 40 50 60
Base Flow, in Cubic Feet per Second
Figure 104. Rating Curve of Mean Ground-Water Stage Compared with Base Flow of Beaverdam Creek, Maryland21
93
-------
0 Morgantown
Gneiss and granitic
to ultramafic rocks
Schist
Geology generalized from Bixom
and Siese (1932,1938)
Contact approximately
located
Stream-gaging station
O
Precipitation-gaging
station
Temperature-measuring
station
• Ch-3
Observation well
-^-Ch-12
Index well
Generalized boundary
of basin
Figure 105. Sketch Map of Brandywine Creek Basin, Showing Generalized Geology and Location of Hydrofogic and Meteorologic Sta-
tions Used in Report22
94
-------
o
CO
CD
Figure 106. Composite Hydrograph of Three Index Wells and the Discharge of Brandywine Creek22
curve, and then reading down to the stream discharge.
For example, in Figure 104 when the mean ground-
water stage is 44.5 feet, ground-water runoff is 10 cfs.
Any flow in excess of this amount is surface runoff.
Daily values of ground-water runoff are plotted on the
stream hydrograph, eventually forming a continuous
line throughout the period of record.
Although wells produce only limited yields from
crystalline rocks in the Piedmont Upland part of the
Delaware River Basin, streams have unusually high
base flows. Olmsted and Hely22 used a ground-water
rating curve to study this apparent inconsistency in a
287 square mile part of the Brandywine Creek basin in
southeastern Pennsylvania, as illustrated in Figure 105.
Bedrock units in the dissected upland basin, which
consists largely of folded Precambrian and Paleozoic
igneous and metamorphic rocks, have similar
hydrologic characteristics. Weathered material of
variable thickness mantles the area and the water table
lies largely within this zone. Precipitation averages
about 44 inches.
The 16 observation wells used in this study ranged
from 12 to 40 feet in depth; all tapped a weathered
schist aquifer. Six or seven wells were measured weekly
or immediately after storms and wells De-3, Ch-13,
Ch-14 were selected as index wells. The average depth
to water in all of the wells was 17.45 feet and the an-
nual fluctuation was 5.75 feet.
Figure 106 shows a composite hydrograph of the
three index wells and the discharge of Brandywine
Creek. The curves have similar trends, differing only in
amplitude following runoff events. This is to be ex-
pected because of the quick response time of a stream.
Certainly the closer an observation well is to a stream,
the more nearly the hydrographs will approach a
similar shape.
01
CO
c?
1
100
90
80
70
60
50
40
30
20
I
I
Apr. o
May +
June+ +Apr.
May O Mar.
Juneo
July +
0 100 200 300 400 500 600 700 800
Monthly average base flow, in cubic feet per second (Og)
Figure 107. Relation of Monthly Average Base Flow to Ground-
Water Stage in the Brandywine Creek Basin22
The rating curve in Figure 107 shows the relation
between ground-water runoff and ground-water stage
in the Brandywine Creek basin. Notice the elipitcal pat-
tern of the data, which approach a straight line from
October through March but then loops back during
spring, summer, and early fall. Although confusing at
first glance, the significance of the loop becomes readi-
ly apparent when the changes that occur in a ground-
water reservoir throughout a 12-month period in a
humid area are considered. From late fall to spring, the
ground-water stage rises because there are little or no
95
-------
losses to evapotranspiration, soil moisture may be at or
above field capacity, and ground-water recharge oc-
curs. The water table reaches its peak during the spring
runoff. From April to October, however, large quan-
tities of ground water are removed by evapotranspira-
tion, the soil moisture becomes so depleted there is lit-
tle or no recharge, and the quantity of water in storage
decreases because ground-water runoff exceeds
recharge. Thus, the elipitcal shape of the data on the
rating curve is controlled by evaportranspiration.
Using the rating curve, Olmsted and Hely22 separated
the Brandywine Creek hydrograph, shown in Figure
108, and found that over a six year period, ground-
water runoff accounted for 67 percent of the total
flow. This compares favorably with the 64 percent
determined for North Branch Rancocas Creek in the
coastal plain of New Jersey; 74 percent for Beaverdam
Creek in the coastal plain of Maryland21; 42 percent
for Perkiomen Creek, a flashy stream in the Triassic
Lowland of Pennsylvania; and 44 percent for the
Pomperaug River Basin, a small stream in
Connecticut23.
During certain times of the year, when the water
table lies at a shallow depth and large quantities of
water are lost by evapotranspiration, a single rating
curve cannot be used to separate a hydrograph with
any degree of accuracy. As Figure 109 shows, Schicht
and Walton24, in their study of Panther Creek basin in
Illinois, developed two rating curves. One is used when
evapotranspiration is very high and the other is used
when evapotranspiration is small. Double rating curves
also can be used to estimate evapotranspiration losses.
Evapotranspiration can also be calculated from the
graph used by Olmsted and Hely in the case cited
above. For example, when the ground-water stage was
80 inches, streamflow was expected to be about 550 cfs
in February and March but only 400 cfs in June. In
this case, the difference, about 150 cfs, is due to
evapotranspiration.
Seepage or Dry-Weather Measurements
Seepage or dry-weather measurements consist of flow
determinations made at several locations along a
stream during a short time interval. It is essential that
there be no surface runoff during these measurements.
Many investigators prefer to conduct seepage runs dur-
ing the stream's 90% flow duration, that is, when the
flow is so low that is is equaled or exceeded 90 percent
of the time.
It is often implied that the 90% flow duration is the
only time the flow consists entirely of ground-water
runoff. This is not necessarily the case. The 90% flow-
duration period, depending on geographic location and
climate, commonly occurs during the late summer and
fall when soil moisture is depleted, there is little or no
ground-water recharge, and the water table, having
declined to its lowest level, has a low gradient. Under
these conditions, ground-water runoff is minimal.
However the physical aspect of the system changes
following a recharge period and ground-water runoff
may account for a substantial portion of the stream's
flow. Hydrograph analyses, using techniques already
described, may readily show that ground water pro-
vides 50 to 70 percent or more of the runoff.
Therefore, the 90% flow may reflect only a small frac-
tion of the total quantity of ground-water runoff.
Seepage measurements permit an evaluation of
ground-water runoff (how much there is and where it
originates) and provides clues to the geology of the
basin as well. The flow of some streams increases
substantially in a short distance. Under natural condi-
tions this increase probably indicates deposits or zones
of high permeability in or adjacent to the stream chan-
nel. These zones may consist of deposits of sand and
gravel, fracture zones, solution openings in limestone,
or merely by local facies changes that increase
permeability. In gaining stretches, ground water may
discharge through series of springs and seeps, along
5000
Daily discharge, affected
by regulation at low flow
Estimated base flow, not including
effects of regulation
10
1953
Figure 108. Hydrograph of Brandywine Creek at Chadds Ford, Pennsylvania, 1952-5322
96
-------
Explanation
• Data for Periods When Evapotranspiration is Very Small
o Data for Periods When Evapotranspiration is Great
40
80 120 160
Ground-Water Runoff in Cubic Feet per Second
200
240
280
Figure 109. Rating Curves of Mean Ground-Water Stage Versus Ground-Water Runoff at Gaging Station 1, Panther Creek Basin24
valley walls or the stream channel, or seep upward
directly into the stream. Areas of significant ground-
water discharge may cause the formation of quicksand.
In areas where the geology and ground-water systems
are not well known, stream flow data can provide a
means of testing estimates of the ground-water system.
If the streamflow data do not conform to the estimates,
then the geology must be more closely examined.25 For
example, the northwest corner of Ohio is crossed by the
Wabash and Fort Wayne moraines between which lies
the St. Joseph River. As indicated by the Glacial Map
of Ohio26, the St. Joseph Basin consists mainly of till.
However, low flow measurements show that the
discharge of the river increases more than 14 cfs along
its reach in Ohio, indicating that the basin contains a
considerable amount of outwash. Thus, hydrologic
studies indicate the need for geologic map modification.
On the other hand, geologic maps may indicate
reasonable locations for constructing stream gaging sta-
tions for hydrologic monitoring networks. The main
stem of the Auglaize River in northwestern Ohio rises
from a mass of outwash that lies along the front of the
Wabash moraine. The southwest-flowing river breaches
the moraine near Wapakoneta and then flows generally
north to its confluence with the Maumee River at De-
fiance. A gaging station is near Ft. Jennings in a till
plain area and slightly above a reservoir on the Auglaize.
In reality this gage measures, at a single point, the flow
resulting as an end product of all causative hydrologic
factors upbasin (ground-water runoff, surface runoff,
slope, precipitation, use patterns, etc.)—it shows merely
inflow into the reservoir. Low-flow measurements,
however, indicate that nearly all of the baseflow is de-
rived from outwash along the distal side of the Wabash
moraine; there is no grain across the wide till-plain
downstream. It would seem that the most logical stream
97
-------
Columbus
167
.0500
o o
o -'
Explanation
Upper number is low flow, mgd.
Lower number is low flow, in mgd/sq. mi.
Area of surficial outwash; well yields
may exceed 1000 gpm.
Area of outwash covered by a few feet
of alluvium; well yields commonly
between 500 and 1000 gpm.
Generally fine-grained alluvium along
flood plain; well yields usually less
than 25 gpm.
Chillicothe
10
I
15
20
Scale in miles
Figure 110. Discharge Measurements in the Scioto River Basin, Ohio
98
-------
gage site for hydrologic evaluations would be at the
breech in the Wabash moraine just downstream from
the till-outwash contact.
Figure 110 shows a number of discharge
measurements made in the Scioto River basin, which lies
in a glaciated part of Central Ohio. The flow
measurements themselves are important in that they
show the actual discharge, in this case at about 90%
flow. The discharge is reported as millions of gallons per
day, instead of the usual cubic feet per second. The
discharge at succeeding downstream sites on the Scioto
River are greater than the sum of the flows immediate-
ly upstream. This shows that the river is gaining and
that water is being added to it by ground-water runoff
from the adjacent outwash deposits.
A paticularly useful method for evaluating streamflow
consists of relating the discharge to the size of the
drainage basin (cfs/sq.mi. or mgd/sq. mi. of drainage
basin). As Figure 110 shows, this technique can be used
to relate the flow index (cfs/sq.mi. or mgd/sq.mi.2) to
the geology and hydrology of the area. A cursory ex-
amination of the data shows that the flow indices can
be conveniently separated into three distinctive units.
These units are arbitrarily called Unit 1 (.01 to .020
mgd/mi.sq.), Unit 2 (.021 to .035 mgd/mi.sq.) and Unit
3 (.036 to .05 mgd/mi.sq.). The Olentangy River and
Alum and Big Walnut Creeks fall into Unit 1, Big Dar-
by and Deer Creeks into Unit 2, and the Scioto River,
Walnut Creek, and the lower part of Big Walnut Creek
into Unit 3. Notice that even though the latter water-
courses fall into Unit 3, the actual discharge ranges
widely—from 3.07 to 181 mgd.
Logs of wells drilled along the streams of Unit 1 show
a preponderance of fine-grained material that contains
only a few layers of sand and gravel and wells generally
yield less than 3 gpm. Along Big Darby and Deer Creek,
however, logs of wells and test holes indicate that
several feet of sand and gravel underlie fine-grained
alluvial material, the latter of which ranges in thickness
from 5 to about 25 feet. Adequately designed and con-
structed wells that tap these outwash deposits produce
as much as 500 gpm. Glacial outwash, much of it coarse
grained, forms an extensive deposit through which the
streams and rivers of Unit 3 flow. The outwash extends
from the surface to depths that exceed 200 feet. In-
dustrial wells constructed in these deposits, most of
which rely on induced inflitration, can produce more
than 1000 gpm. By combining the seepage data and well
yields with a map showing the areal extent of the
deposits characteristic of the stream valleys, the map in
Figure 110, that indicates potential well yields in the
area, was developed. The potential ground-water yield
map relies heavily on streamflow measurement, but
nonetheless, provides, with some geologic data, a good
first cut approximation of ground-water availability.
Stream reaches characterized by significant increases
in flow due to ground-water runoff, may also have
unusual quality characteristics. In northern Ohio the
discharge of a small stream, shown in Figure 111, that
drains into Lake Erie increases over a 3-mile stretch
from about 1 to more than 28 cfs and remains relatively
constant thereafter. The increase begins at an area of
springs where limestone, which has an abundance of
solution openings, approaches land surface and actually
crops out in the stream bottom. The till-limestone con-
tact declines downstream eventually exceeding 90 feet in
depth.
In the upper reaches of stream, baseflow is provided
by ground water that discharges from the adjacent till.
Since this water has been in the ground but a short
time, the mineral content is low. Where streamflow
begins to increase significantly, the limestone aquifer
provides the largest increment. Furthermore, the
bedrock water contains excessive concentrations of
dissolved solids, hardness, and sulfate, and in this
stretch calcite precipitates on rocks in the stream chan-
nel. The fish population in the upper reaches is ex-
ceedingly abundant until the stream reaches the
limestone discharge zone. At this point, the population
quickly diminishes and remains in a reduced state
throughout the remaining length of the stream. No
doubt the reduction in fish population is related to the
quality of the water.
In describing the hydrology of Wolf Creek in east-
central Iowa, Kunkle19 used seepage measurements and
water-quality data to determine the amount of ground-
water runoff provided by alluvium and limestone. As
Figure 112 shows, the 325-square mile basin is mantled
by till and underlain by limestone and shale, but the
valley itself contains about 40 feet of permeable
alluvium. Well data show that the stream is hydraulically
connected with the limestone aquifer along a 5-mile
stretch and baseflow is provided by discharge from both
the limestone and the alluvium. On either side of this
reach the limestone potentiometric surface is below the
stream bed.
Measurements were made at three stations during low-
flow conditions. The discharge 8 miles upstream from
the limestone discharge area was 16.4 cfs (.086
cfs/sq.mi.), midway along the reach was 29.8 cfs (.0997
cfs/sq.mi.), and 7 miles downstream was 37.0 cfs (.114
cfs/sq.mi.). Water from the limestone has an average
conductivity of 1330 micromhos, while that from the
alluvium and upstream-derived baseflow average 475
micromhos. After mixing the surface water had a con-
ductivity of 550 micromhos.
Using a slight modification of the equations given
previously, it is possible to calculate the amount of
ground-water runoff from the limestone in this reach
under the given conditions.
C, Q,
CaQa
CbQb = CQ0
Q, + CL + Qh = CL
(54)
(55)
where Qj, Qa, Qb, Q0 are the discharge from upstream
(inflow), from the alluvium, from the limestone, and
from the outflow respectively, and Cj, Ca, Cb and C
represent the conductivity of the inflow from upstream,
99
-------
Station Number
No. species
No. indiv.
D.O.
Q
Temp C
PH
Alk
C02
Cond.
1
10
980
10.8
0
21
8.27
285
4
638
Limestone with solution openings
pH t as CO2 t & CaCO3 I (travertine)
1 inch = 1 mile
Figure 111. Green Creek Drainage Basin (Seneca and Sandusky Co.'s, Ohio)
-------
C - 475
Q = 16.4 Cfs
8 miles
Qj
C = 550
Q = 29.8 cfs
5 miles
C = 550
Q0 = 37.0 cfs
Qr
7 miles
40ft
Qa
o &
O
1 ^x ° ° o °
Till
Potentiometric surface
in limestone
Conductivity of alluvium = 475;
limestone = 1330
Limestone C = 1330
Qj = inflow
Qb = bedrock inflow
475 Q, + 475 Qa + 1330 Qb = C Q0
Q, + Qa +
Qa = alluvium inflow
Q0 = outflow
Figure 112. Generalized Hydrogeology of Wolf Creek, Iowa
Qb =
from the alluvium, from the limestone, and from the
outflow water. Substituting:
475 Qj + 475 Qa + 1330 Qb = 550 X 37
-475 Qj - 475 Qa - 475 Qb = 475 X 37
855 Qb = 2775
Qb = 3.2 cfs
Thus in this particular stretch, the limestone was pro-
viding about 3.2 cfs of the stream's total flow of 37 cfs.
Carrying the analyses a bit farther, we could assume
that since the limestone provides, 3 to 4 cfs during
baseflow, wells tapping the limestone in this stretch
could provide a like amount without dewatering the
system. Since 1 cfs = 450 gpm, wells could produce a
total yield of 1350 to 1800 gpm. Using similar logic we
could predict the minimum yield of wells tapping the
alluvium, assuming that they would capture only the
ground water runoff.
Temperature Surveys
The temperature of shallow ground water is nearly
uniform, reflecting the mean annual air temperature of
the region. The temperature of shallow ground water
ranges from a low of about 37 degrees in the north-
central part of the U.S. to more than 77 degrees in
southern Florida. Of course at any particular site the
temperature of ground water remains nearly constant.
Surface water temperatures, however, range within wide
extremes—freezing in the winter in northern regions and
exceeding 100 degrees during hot summer days in the
south. Mean monthly stream temperatures during July
and August range from a low of 55 in the northwest to
more then 85 degrees in the southeast.
During the summer where ground water provides a
significant increment of flow, the temperature of surface
water in a gaining reach will decline. Conversely, during
winter the ground water will be warmer than that on the
surface and although ice will normally form, parts of a
stream may remain open because of the inflow of the
warmer ground water. In central Iowa, for example,
winter temperatures commonly drop below zero and ice
quickly forms on streams, ponds, and lakes. The
ground-water temperature in this region, however, is
about 52 degrees and, if a sufficient amount is discharg-
ing into a surface-water body, the temperature may re-
main above 32 degrees and the water will not freeze. In
the summer, the relatively cold ground water (52
degrees) mixes with the warm surface water (more than
79 degrees) producing a mixture of water colder than
that in non-gaining reaches.
Field examination of winter aerial photography may
show places where ice is either absent or thin. In the
summer it is possible to float down a river, periodically
measuring the temperature. Ground-water discharge
areas are detected by temperature decrease. A third
method of detection is by means of an aircraft-
mounted thermal scanner. This sophisticated instru-
ment is able to detect slight differences in temperature
and would probably be more accurate than ther-
mometry or low altitude aerial photography.
Flow-Duration Curves
A flow-duration curve shows the frequency of occur-
rence of various rates of flow. It is a cumulative fre-
quency curve prepared by arranging all discharges of
record in order of magnitude and subdividing them ac-
cording to the percentages of time during which
101
-------
CO
I
U)
I
0}
JZ
o
a>
O
1.0
O.I
.01
V
v\\>
^
s
\ \
\
\
\
\
\
\
\
\
Low
%
^
iky
\
\
\
\
\
\
\
\
er Extr
/
\
,\
\ \
\ \
A
\N|
\
\\
\ N
V
V
\
\
\
\
V
\
»
v
\
>
erne
Upi
\
\
\
V
V
^
^
\\
v
\
V
1
I
\
t
\
\
\
\
1
t
-"
>er
\
"X
\\
'Vv
, v>
\ xv
\
^
1
t
\
t
\
1
t
1 1_
\
Extr
K
\\
v\
K \
\\
^
^
k
\
\
\
\
\
\
\
\
3me
S-i
*>>.»
\
L^
\ ^
\
V
\\
\>
v
\
\
\
\
\
\
\
\
\
\
I
1
t
1
v
\
>\
\ -
v
\
\\
\\
\
\
\
I
1
1
I
I
1
V
^->
"^,
Little
-------
32°
31
90° 89°
Figure 114. Geologic Map of Area in Southern Mississippi Having Approximately Uniform Climate and Altitude
10
a
CT
W
1
O
5
O
\\
Explanation
1 Bogue Chitto near Tylertown
2 Bowie Creek near Hattiesburg
SHomochitto River of Eddicaton
4 Leaf River near Collins
• 5Oakohay Creek at Mize
6 Strong River at D lo
N
5 8 £ 8 S S 8 88 8SS88SS 88 S8SSS :
Percent of Time Indicated Discharge Was Equaled or Exceeded
Figure 115. Flow-Duration Curves for Selected Mississippi Streams, 1939-48
103
-------
specific flows are equaled or exceeded; all chronologic
order or sequence is lost.27 Flow-duration curves may
be plotted on either probability or semilog paper. In
either case, the shape of the curve is an index of
natural storage in a basin, including ground water.
Since dry-weather flow consists entirely of ground-
water runoff, the lower end of the curve indicates the
general characteristics of shallow aquifers.
Figure 113 shows several flow-duration curves for
Ohio streams. During low-flow conditions (the flow
equaled or exceeded 90 percent of the time), the curves
for several of the streams, such as the Mad, Hocking,
and Scioto River, and Little Beaver Creek trend toward
the horizontal, while Grand River, White Oak, and
Home Creeks all remain very steep.
Mad River flows through a broad valley that is filled
with very permeable sand and gravel. The basin has a
large ground-water storage capacity and, consequently,
the river maintains a high sustained flow. The Hocking
river locally contains outwash in and along its
floodplain, which provides a substantial amount of
ground-water runoff. Above Columbus, the Scioto
River crosses thin layers of limestone that crop out
along the stream valley and the adjacent uplands are
covered with glacial till; ground-water runoff from this
reach is relatively small. Immediately south of Colum-
bus, however, the Scioto Valley widens and is filled
with coarse outwash. The reason that Mad River has a
higher low-flow index than the Scioto River at
Chillicothe is because the Mad River receives ground-
water runoff throughout its entire length, while the
flow of the Scioto River increases significantly only in
the area of outwash south of Columbus.
White Oak and Home Creeks originate in bedrock
areas where relatively thin alternating layers of sand-
stone, shale, and limestone crop out along the hill
sides. The greater relief in these basins promotes sur-
face runoff and the rocks are not very permeable. Ob-
viously the ground-water storage characteristics and
potential yield of these basins are far less than those
filled or partly filled with outwash.
Figure 114 shows a geologic map of a part of southern
Mississippi. Notice that gaging stations at Sites 1, 2,
and 3 record the drainage from the Citronelle Forma-
tion, while stations 4, 5, and 6 represent the drainage
from the older rocks. Respective flow-duration curves,
shown in Figure 115, show that stations 1 and 2 have
high low-flow indices, with station 3 a relatively close
third. The high flow-duration indices indicate that the
Citronelle Formation has a greater ground-water
storage capacity, a higher rate of natural recharge, and
presumably would provide larger yields to wells than
the underlying strata. This formation consists of sand,
gravel, and clay, while the other strata are generally
composed of finer materials. Thus it would appear that
streamflow data can be used as an aid to a better
understanding of the permeability and infiltration
capacity, as well as facies changes, of geologic units.
Flow Ratios
Walton28 reported that grain-size frequency-
distribution curves are somewhat analogous to flow-
duration curves in that their shapes are indicative of
water-yielding properties of deposits. He pointed out
that a measure of the degree to which all of the grains
approach one size, and therefore, the slope of the
grain-size frequency distribution curve, is the sorting.
One parameter of sorting is obtained by the ratio
(D25/D75)1/2. Walton modified this equation by replac-
ing the 25 and 75 percent grain-size diameters with the
25 and 75 percent flow. In this case a low ratio is in-
dicative of a permeable basin or one that has a large
ground-water storage capacity.
The Q25 and Q75 data are easily obtainable from
flow-duration curves. Using the data from Figure 113,
Mad River has a flow ratio of 1.58 and the Scioto
River's ratio is 2.58, while Home Creek, typifying a
basin of low permeability, has the highest ratio which
is 5.16.
104
-------
References
14 Trainer, F.W. and F.A. Watkins. 1975.
"Geohydrologic Reconnaissance of the Upper
Potomac River Basin." U.S. Geogological Survey
Water-Supply Paper 2035.
15 Johnstone, D. and W.P. Cross. 1949. Elements of
Applied Hydrology. Ronald Press, New York.
16 Gray, D.M. (editior). 1970. Handbook of the Prin-
ciples of Hydrology. Water Information Center,
Inc.
17 Visocky, A.P. 1970. "Estimating the Ground-Water
Contribution to Storm Runoff by the Electrical
Conductance Method." Ground Water, vol. 8,
no. 2.
18 Toler, L.G. 1965. "Use of Specific Conductance to
Distinguish Two Base-Flow Components in Econ-
fma Creek, Florida." U.S. Geological Survey Pro-
fessional Paper 525-C.
19 Kunkle, G.R. 1965. "Computation of Ground-Water
Discharge to Streams During Floods, or to In-
dividual Reaches During Base Flow, By Use of
Specific Conductance." U.S. Geological Survey
Professional Paper 525-D.
20 Pettyjohn, W.A. and R.J. Henning. 1979.
"Preliminary Estimate of Ground-Water Recharge
Rates, Related Streamflow and Water Quality in
Ohio." Ohio State University Water Resources
Center, Project Completion Report 552.
21 Rasmussen, W.C. and G.E. Andreason. 1959.
"Hydrologic Budget of the Beaverdam Creek
Basin, Maryland." U.S. Geological Survey Water-
Supply Paper 1472.
^Olmsted, F.H. and A.G. Hely. 1962. "Relation
Between Ground Water and Surface Water in
Brandywine Creek Basin, Pennsylvania." U.S.
Geological Survey Professional Paper 417-A.
23 Menizer, O.E, and N.D. Stearns. 1928. "A Study
of Groundwater in the Pomerang Basin." U.S.
Geological Survey Water-Supply Paper 597-B.
^Schicht, R.J. and W.C. Walton. 1961.
"Hydrologic Budgets for Three Small Watersheds
in Illinois." Illinois State Water Survey Report of
Investigations 40.
25 LaSala, A.M. 1967. "New Approaches to Water-
Resources Investigations in Upstate New York,"
Ground Water, vol. 5, no. 4.
26 Goldthwait, R.P., G.W. White, and J.L. Forsyth.
1961. "Glacial Map of Ohio." U.S. Geological
Survey, Miscellaneous Geological Investigations
Map 1-316.
27 Cross, W.P. and R.E. Hedges. 1959. "Flow
Duration of Ohio Streams." Ohio Division of
Water Bulletin 31.
28 Walton, W.C. 1970. Groundwater Resource Evalua-
tion. McGraw-Hill Publications Co., New York.
105
-------
Chapter 5
Ground-Water Pollution
Long-Term Effects
For millennia, man has disposed of his waste prod-
ucts in a variety of ways. The disposal method might
reflect convenience, expedience, expense, or best
available technology, but in many instances, leachate
from these wastes have come back to haunt later gen-
erations. This is largely because we have not thought
out the consequences of our actions. Ground-water
pollution may lead to problems of inconvenience, such
as taste, odor, color, hardness, or foaming; but the
pollution problems are far more serious when patho-
genic organisms, flammable or explosive substances, or
toxic chemicals or their by-products are present, partic-
ularly when long-term health effects are unknown.
Individual polluted ground-water sites generally are
not large, but once polluted, ground water may remain
in an unusable or even hazardous condition for decades
or even centuries. The typically low velocity of ground
water prevents a great deal of mixing and dilution; con-
sequently, a contaminant plume may maintain a high
concentration as it slowly moves from points of
recharge to zones of discharge.
An oil-field brine holding pond was constructed adja-
cent to a producing well in central Ohio in 1968. Two
years later when the well was plugged, the holding pond
was filled, graded, and seeded. The chloride concentra-
tion in the ground water in the vicinity of the former
pond still exceeded 36,000 mg/1 some 10 years after the
operation began and 8 years after reclamation.
Scores of brine holding ponds were constructed in
central Ohio during an oil boom in 1964; many are still
in use. In 1978 a number of test holes were constructed
within 200 feet of one such pond. Within this area
shallow ground water contained as much as 50,000 mg/1
of chloride. Moreover, brine-contaminated ground water
provides part of the flow of many streams and this has
caused degradation of surface-water quality.29'30> 31
Documentation of the migration of leachate plumes
originating at garbage dumps and landfills is becoming
increasingly abundant. Data show that under certain
hydrologic conditions leachate plumes can move consid-
erable distances and degrade ground water throughout
wide areas. Furthermore, the problem is worldwide.
Exler32 described a situation in southern Bavaria, Ger-
many, where a landfill has been in operation since 1954.
The wastes are dumped into a dry gravel pit. As Figure
116 illustrates, data collected from 1967 to 1970 showed
the narrow lense-shaped plume had migrated nearly
2 miles.
Bavaria, Germany
1955
Landfill
Abandoned
pit
More than 2 miles
Figure 116. Leachate from a Landfill in Bavaria Has Migrated
More Than 2 Miles and the Ground Water Has Been Degraded
for Nearly 25 Years.
Keizer, Oregon
1945
Incompletely
processed Al ore
Aluminum
Processing
Plant
Figure 117. Thirty-Three Years After Disposal Began the
Leachate from Aluminum Ore and Mill Tailings is Still a Prob-
lem in Keizer, Oregon.
As Figure 117 illustrates, incompletely processed
aluminum ore was dumped into a borrow pit in Keizer,
Oregon from August 1945 to July 1946.33 The ore and
mill tailings had been treated with sulfuric acid and
ammonium hydroxide. When first recognized by local
residents in 1946, the ground water was contaminated
by more than 1,000 mg/1 of sulfate; many shallow
107
-------
domestic wells tapping the Recent alluvium were
contaminated. In the Spring of 1948 the waste was
removed from the borrow pit. Two wells, reportedly
capable of producing more than 700 gpm (gallons per
minute) were installed near the pit and the contaminated
groundwater was pumped to waste for several months.
By 1964 the contaminants had migrated more than a
mile. No doubt some of the contaminants are still in the
ground water at Keizer.
A well-documented study by Perlmutter and others34
showed that disposal of chromium and cadmium-rich
plating wastes from an aircraft plant on Long Island
during a 20-year period contaminated a shallow aquifer.
Figure 118 illustrates this study. The contamination was
first discovered in 1942, and by 1972 the degraded
ground water zone was about 4,200 feet long and 1,000
feet wide. The 1972 study demonstrated that the
chromium-cadmium enriched cigar-shaped plume "had
not only reached Massapequa Creek but was present in
the stream as well as in the beds beneath it."35
Long Island, New York 1942
Massapequa Creek
Metal
plating
wastes
/-J
1942
Aircraft
Factory
Figure 118. More Than 36 Years After Disposal of Plating Wastes
Began, the Ground Water Remains Polluted in South
Farmingdale.
London, England
1914
More than 1 mile
Figure 119. The Picric Acid, Which Has Been Found in the
Ground Water Near London for Decades, Originated at a World
War I Munitions Plant.
During the middle and late 1930's grasshopper infes-
tations were stripping the vegetation throughout wide
areas in the Northern Great Plains. In western Minne-
sota partial control was obtained by a grasshopper bait
consisting of arsenic, bran, and sawdust. Eventually the
leftover bait was buried. In May 1972, a contractor
drilled a well near his office and warehouse on the out-
skirts of a small town. During the next two and a half
months 11 of the 13 individuals employed at the site
became ill; two were hospitalized. They were suffering
from arsenic poisoning. One sample of water from the
well contained 21 mg/1 of arsenic. Analyses of soil from
the site revealed arsenic concentrations ranging from
3,000 to 12,000 mg/1. Apparently the well was drilled in
the vicinity of the grasshopper bait disposal site, which
had long been forgotten by the local residents.36
Wastes from munitions works include picric acid, a
toxic, intensely bitter, pale yellow substance. Picric acid
is not readily removed by traditional water treatment
methods and its migration through either the unsaturated
zone or the saturated zone does not appear to neutralize
it.
During the World War I years of 1914-1918, wastes
from the manufacture of explosives at a plant near the
Thames River just northeast of London, England, were
placed in abandoned chalk pits. Figure 119 illustrates
the migration of these wastes. In the early 1920's water
from a nearby well was first reported to have a yellow
tint.37 Additional water samples collected between 1939
and 1955 also contained a characteristic yellow picric
acid tint. Sampling ceased in 1955 when the pump was
removed. By 1942 the pollutants had migrated at least a
mile as indicated by another contaminated well. There is
no reason to believe that the picric acid has been
flushed from the aquifer. The ground water has certain-
ly been polluted for 40 years, quite probably for more
than 70 years, and very likely will be polluted for many
more years to come.
Because of high evaporation and low recharge, waste
disposal in arid regions can lead to long-lived ground-
water quality problems. In the first place, salts are con-
centrated by evaporation to form highly mineralized
fluids. Secondly, water supplies may not be readily
available and, therefore, every effort must be made to
protect existing sources.
Ground-water contamination in the desert environ-
ment near Barstow, California, was described by
Hughes.38 Beginning around 1910, waste fuel oil and
solvents from a railroad system were discharged to the
dry floor of the Mojave River near Barstow. The first
municipal sewage treatment plant was constructed in
1938; the effluent was discharged to the riverbed.
Sewage treatment facilities were enlarged in 1953 and
1968. Effluent disposal was dependent on evaporation
and direct percolation into the alluvial deposits.
At the U.S. Marine Corps base near Barstow, indus-
trial and domestic waste treatment facilities first became
operational in 1942; effluent disposal relied on direct
percolation and evaporation. Some of the effluent was
used to irrigate a golf course. Other sources of ground-
water contamination were two nearby mining and mill-
ing operations.
As Figure 120 shows, analysis of well waters collected
during the Spring of 1972 indicated the existence of two
zones of contaminated ground water in the alluvial
deposits of the Mojave River. The deeper zone, orig-
inating from the 1910 disposal area, exceeded 1,800 feet
in width and extended nearly 4!/i miles in a downgra-
dient direction. Its upper surface lies 60 or more feet
108
-------
below land surface. The second or shallow zone orig-
inates at the sewage treatment lagoon installed in 1938
and at the Marine Corps golf course. This zone consists
of two apparently separate plumes. The upgradient
plume extends nearly 2 miles downstream, while the
plume originating at the golf course is nearly a mile
long; the plumes are about 700 feet wide. Hughes esti-
mated that the pollution fronts are moving at a rate of
1 to 1.5 feet per day. The Marine Corps well field lies in
the path of these plumes; several domestic wells have
already been contaminated. In this instance poor waste
disposal practices, beginning nearly 75 years ago may
cause water-supply problems at the Marine Corps base
unless expensive corrective measures are undertaken.
Barstow, California
1910
Industrial wastes,
fuel oil, solvents
1910
Barstow waste
disposal ponds
1938
U.S.M.C.
golf course
sewage irrigation
1942
More than 4 miles
Figure 120. Waste Disposal Beginning Nearly 70 Years Ago at
Barstow, California is Now Threatening an Important Well Field
at the Nearby Marine Base.
London, England
1905
1905
Glassworks
Plant
Tar, acids, oil,
sulfate sludge
Figure 121. Ground-Water Pollution by Wastes from a
Gasworks Plant Near London Has Even Created a Fire Hazard.
From 1905 to 1967 wastes from a gasworks plant
were deposited in abandoned gravel pits along the Lee
River near Waltham Cross, a few miles northwest of
London, England.39 Figure 121 shows that the tar acids,
oils, and sulfate sludge infiltrated to contaminate the
ground water over a wide area. Apparently the pollution
was first detected in 1935, some 30 years after disposal
began. At that time oil, floating on the ground water,
emerged at land surface. Continual but slow accumula-
tion of oil on and near the land surface led to hazardous
conditions and, in 1943, the oil was ignited. Con-
taminated ground water was also encountered in new ex-
cavations where it appeared as high concentrations of
sulfate in 1958 and as oily waters in 1961. In 1965, oily li-
quids also seeped into Pymmes Brook and the Lee River
Navigation Channel following a substantial rise in the
water table after heavy rains. Additional surface-water
degradation occured in 1966 because of the discharge of
oil from streamside seepage zones.
Ground water in the surficial sand and gravel deposit
was contaminated over a wide area. Fortunately, most
water supplies in this region are pumped from an under-
lying chalk, which generally is separated from the gravel
by the London Clay. It is evident from this example
that waste disposal, which began 80 years ago, con-
tinues to be troublesome and that ground-water con-
tamination can indeed become a fire hazard.
All ground-water pollution is not necessarily bad.
Inhabitants of Crosby, a small village in northwestern
North Dakota, believed they produced the best coffee in
the State because the water from which it was made
contained "body". The rather highly mineralized water
(dissolved solids = 2,176 mg/1, sulfate = 846 mg/1,
chloride = 164 mg/1, and nitrate = 150 mg/1) used for
brewing the coffee was obtained exclusively from an old
dug well. The well, however, was constructed, probably
near the turn of the century, at the site of the local river
livery stable. Livestock wastes provided the
peculiar flavor so characteristic of the coffee made
in Crosby.40
The manufacture of soda ash, caustic soda, chlorine,
and allied chemicals began at Barberton, Ohio, shortly
before the turn of the century. The plant discharged a
mixture of calcium and sodium chlorides directly to the
Tuscarawas River and to retention ponds. The discharge
of chloride in 1966 averaged 1,500 tons per day.41 These
wastes have led to serious ground-water pollution prob-
lems in eastern Ohio and have necessitated abandon-
ment of streamside well fields at Barberton in 1926 and
at Massillon and Coshocton in 1953.
Municipal wells at Zanesville, more than 135 river
miles downstream from Barberton, have also been
adversely affected by the chloride induced into the
watercourse aquifer from the contaminated Muskingum
River. Due to high treatment costs Zanesville officials
considered abandoning their well field in 1963. At the
confluence of the Muskingum and Ohio Rivers, about
220 river miles below Barberton, is the city of Marietta.
Almost 30 years ago, Marietta officials were concerned
over the marked increase in chloride in municipal wells
during the preceding 10 years.42 The cause, of course,
was induced infiltration of the chloride-rich Muskingum
River water.29
It is evident that decades of poor waste-disposal prac-
tices at Barberton seriously impaired streamside aquifers
and well fields for a distance of over 200 river miles.
The soda ash plant at Barberton was closed in 1973 and
waste discharges substantially reduced. Presumably,
these water-quality problems will decrease in severity
over the next several years, after a history of 90 years or
more.
According to Mink and others43 mining operations in
109
-------
the Coeur d'Alene district of northern Idaho have been
continuous for more than 90 years. Unfortunately,
leaching of the ancient mining and milling wastes is now
affecting the chemical quality of ground water in several
areas, including Canyon Creek basin near Wallace. Here
high concentrations of zinc, lead, copper, and cadmium
occur in both ground water and soil samples.
In 1884 striking miners set fire to several deep coal
mines in the vicinity of New Straitsville, Ohio. Still
burning uncontrollably, the fires were started by dis-
gruntled workers who rolled burning wood-filled coal
cars into the shafts that honeycomb the ground under
the town. In the years since, many wells have become
contaminated, dried up, or produce water hot enough to
make instant coffee.
Disposal of domestic, industrial, and municipal
wastes, which probably began around 1872 through
wells and sinkholes tapping a permeable limestone
aquifer, was the birth of a contaminated area that now
encloses some 75 square miles. By 1919 the practice of
disposing of sewage at the northern Ohio town of
Bellevue was well established and many wells had been
contaminated. In the early 1960's some wells were
reported to yield easily recognizable raw sewage.44
This problem began more than a hundred years ago and
remains to this day.
A gasworks plant was built at Norwich, England, in
1815 and abandoned in 1830. Phenolic compounds,
originating from whale oil, infiltrated and remained in
the underlying chalk for at least 135 years when it con-
taminated a newly drilled well in 1950.40' 45 These
organic compounds, no doubt, are still there 170 or so
years later.
Sources of Ground-Water Contamination
As water moves through the hydrologic cycle, its
quality changes in response to differences in the phys-
ical, chemical, and biological environments through
which it passes. The changes may be either natural or
man-influenced; in some cases they can be controlled, in
other cases they cannot, but in most cases they can be
managed in order to limit adverse water-quality changes.
The physical, chemical, and biological quality of
water may range within wide limits even though there
are no man-made influences. In fact, it is often impos-
sible or at least difficult to distinguish the origin (man-
made or natural) of many water-quality problems. The
natural quality reflects the types and amounts of soluble
and insoluble substances with which the water comes in
contact. Surface water generally contains less dissolved
solids than ground water, although at certain times
(generally during low flow rates) in areas where ground-
water runoff is the major source of streamflow, the
quality of both surface water and ground water is
similar. During periods of surface runoff, streams may
contain large quantities of suspended materials and,
under some circumstances, a large amount of dissolved
solids. Most commonly, however, during high rates of
flow the water has a lower dissolved-mineral
concentration.
Although the chemical quality of water in surficial or
shallow aquifers may range within fairly wide limits
from one time to the next, deeper ground water is char-
acterized by nearly constant chemical and physical prop-
erties, at least on a local scale where the aquifer is un-
stressed by pumping. As a general rule, the dissolved-
solids content increases with depth and with the time
and distance the water has traveled in the ground. A
few uncommon water-quality situations exist throughout
the country, reflecting unusual geologic and hydrologic
conditions. These include, among others, thermal areas
and regions characterized by high concentrations of cer-
tain elements, some of which may be health hazards.
For centuries man has been disposing of his waste
products by burning, placing them in streams, storing
them on the ground, or putting them in the ground us-
ing various methods. Man-made influences on stream-
water quality reflect not only waste discharge directly
into the stream, but also include highly mineralized or
polluted surface runoff, which can carry a wide variety
of substances. Another major influence on surface-water
quality is related to the discharge of ground water into
the stream. If the adjacent ground water is polluted,
stream quality tends to deteriorate. Fortunately in the
latter case, the effect in the stream generally will not be
as severe as it is in the ground, due to dilution of the
pollutant. See Reference 31 for example.
The quality of ground water is most commonly af-
fected by waste disposal. One major source of pollution
is the storage of waste materials in excavations, such as
pits or mines. Water-soluble substances that are
dumped, spilled, spread, or stored on the land surface
or in excavations may eventually infiltrate to pollute
ground-water resources. Ground water is also polluted
by the disposal of fluids through wells and, in limestone
terrains, through sinkholes directly into aquifers. Like-
wise, infiltration of highly mineralized surface water has
been a major cause of underground pollution in several
places. Irrigation tends to increase the mineral content
of both surface and ground water. The degree of severi-
ty of pollution in cases such as these is related to the
hydrologic properties of the aquifers, the type and
amount of waste, disposal techiniques, and climate.
A major and widespread cause of ground-water qual-
ity deterioration is pumping, which may cause the
migration of more highly mineralized water from
surrounding strata to the well. The migration is directly
related to differences in hydrostatic head between adja-
cent water-bearing zones and to the hydraulic conductiv-
ity of the strata. In coastal areas pumping may cause
sea water to invade a fresh water aquifer. In parts of
coastal west Florida, wild-flowing, abandoned artesian
wells have salted, and consequently ruined, large areas
of formerly fresh or slightly brackish aquifers.
The list in Table 10 shows that man-influenced ground-
water quality problems are most commonly related to:
(1) water-soluble products that are placed on the land
surface and in streams; (2) substances that are deposited
or stored in the ground above the water table; and (3)
110
-------
material that is stored, disposed of, or extracted from
below the water table. Many of the pollution problems
related to these situations are highly complex, and some
are not well understood.
Table 10. Sources of Ground-Water Quality Degradation
Ground-Water Quality Problems that Originate on the Land Surface
1. Infiltration of polluted surface water
2. Land disposal of either solid or liquid wastes
3. Stockpiles
4. Dumps
5. Disposal of sewage and water-treatment plant sludge
6. De-icing salt usage and storage
7. Animal feedlots
8. Fertilizers and pesticides
9. Accidental spills
10. Particulate matter from airborne sources
Ground-Water Quality Problems that Originate in the Ground Above the
Water Table
1. Septic tanks, cesspools, and privies
2. Holding ponds and lagoons
3. Sanitary landfills
4. Waste disposal in excavations
5. Leakage from underground storage tanks
6. Leakage from underground pipelines
7. Artificial recharge
8 Sumps and dry wells
9. Graveyards
Ground-Water Quality Problems that Originate in the Ground Below the
Water Table
1. Waste disposal in well excavations
2. Drainage wells and canals
3. Well disposal of wastes
4. Underground storage
5. Secondary recovery
6. Mines
7. Exploratory wells
8. Abandoned wells
9. Water-supply wells
10. Ground-water development
Contaminated
Stream
Figure 122. Induced Infiltration of Contaminated Stream Water
May Lead to Ground-Water Pollution.
cases, the pollution originated from the disposal of
municipal or industrial waste directly into the stream,
which was induced by pumping into adjacent aquifers.
In hydrologic situations such as these, months or
perhaps years may be required for the polluant to ad-
vance from the stream into the well. Once at the well,
however, the aquifer between the well and the stream
may be completely polluted, requiring years to recover
once the source has been eliminated.
Land Disposal of Either Solid or Liquid Waste Materials. One
of the major causes of ground-water pollution is the
disposal of waste materials directly onto the land sur-
face. Examples include manure, sludges, garbage, and
industrial wastes. The waste may occur as individual
mounds or it may be spread out over the land. If the
waste material contains soluble products, they will in-
filtrate the land and may lead to ground-water pollu-
tion. Similar problems occur in the vicinity of various
types of stockpiles.
Stockpiles. Perhaps the prime example of ground-water
pollution caused by stockpiles is storage of de-icing salt
(sodium and calcium chloride) used for highway snow
and ice control. Not uncommonly, tons of salt are sim-
Ground-Water Quality Problems that Originate on the Land
Surface
Infiltration of Polluted Surface Water. The yield of many
wells tapping streamside aquifers is sustained by infiltra-
tion of surface water. In fact, Figure 122 shows that
more than half of the well yield may be derived directly
from induced recharge from a nearby stream, which
may be polluted. As the induced water migrates through
the ground, a few substances are diluted or removed by
filtration and sorption. This is especially true where the
water flows through filtering materials, such as sand
and gravel, particularly if these materials contain some
soil organic matter. Filtration is less likely to occur if
the water flows through large openings, such as those
that occur in carbonate aquifers. Many pollutants, for
example chloride, nitrate, and many organic com-
pounds, are highly mobile, move freely with the water,
and are not removed by filtration.
Examples of ground-water supplies being degraded by
induced recharge of polluted surface water are both
numerous and widespread. In the greatest number of
Figure 123. Leaching of Solids at the Land Surface. The Possi-
bility of Ground-Water Pollution Under These Conditions is
Rarely Anticipated.
Ill
-------
ply piled on the land surface awaiting use. As Figure
123 shows, the highly soluble material rapidly dissolves
and either infiltrates or runs off into streams. In recent
years, many highway officials have provided some pro-
tection for salt stockpiles by covering them with plastic
sheets or storing the salt in sheds. This is not necessarily
done to protect adjacent water resources, but merely to
preserve the salt.
Dumps. In the past few years, investigators have begun
to take a serious look at the environmental effects of
dumps. As rainwater infiltrates through trash in a
dump, it accumulates a wide variety of chemical and
biological substances. The resulting fluid, or leachate,
may be highly mineralized. As the leachate infiltrates,
some of the substances it contains are removed or
degraded. As Figure 124 shows, eventually the leachate
may reach the water table where it flows in the direction
of the regional ground-water gradient or toward a well.
(In some places "sanitary fills" or dumps lie below the
water table.)
Figure 124. Ground-Water Contamination Caused By Leachate
Infiltration from a Dump.
Disposal of Sewage and Water Treatment Plant Sludge. The
sludge from treatment plants presents not only a signifi-
cant waste disposal problem but one that is growing,
significantly. The wastes include lime-rich sludge from
water treatment plants as well as sewage sludge from
waste water treatment plants. In recent years, municipal
officials have attempted to solve the sewage-sludge
problem by spreading the sludge on the land surface or
filling abandoned strip mine pits with it. At first glance
this may seem to be an effective means of disposal, but
many exotic chemicals, derived from domestic, agricul-
tural, municipal, and industrial wastes may exist in the
sludge as soluble or relatively insoluble substances.
When the sludges are used as fertilizers, the soluble
compounds may infiltrate while the more insoluble
compounds, many of which may consist of toxic
metals, are removed and concentrated by plants. Much
needs to be learned about the chemical and biological
migration of numerous elements and compounds pres-
ent in sludges.
Salt Spreading on Roads. In recent years, particularly since
the construction of the interstate highway system, water
pollution due to wintertime road salting has become an
increasing problem. The effect is becoming even more
severe as salt usage increases with a concomitant
decrease in the use of sand. From a water quality view-
point, the salting brings about deterioration of stream
quality due to highly mineralized surface runoff, and
the infiltration of briney water causes ground-water
pollution.
Accidental Spills of Hazardous Materials. A large volume of
toxic materials are transported throughout the country
by truck, rail, and aircraft and are stored in above-
ground tanks; accidental spills of these hazardous
materials are not uncommon. There are virtually no
methods that can be used to quickly and adquately
clean up an accidental spill or spills caused by explo-
sions or fires. Furthermore, immediately following an
accident, the usual procedure is to spray the spill area
with water. The resulting fluid may either flow into a
stream or infiltrate the ground. In a few cases, the
fluids have been impounded by dikes, which lead to
even greater infiltration. In any case, water resources
may be easily and irreparably polluted from accidental
spills of hazardous materials.
Fertilizers and Pesticides. Increasing amounts of both
fertilizers and pesticides are being used in the United
States each year. Many of these substances are highly
toxic and, in many cases, quite mobile in the subsur-
face. Many compounds, however, become quickly at-
tached to fine-grained sediment, such as organic matter
and clay and silt particles; a part of this attached mate-
rial is removed by erosion and surface runoff. In many
heavily fertilized areas, the infiltration of nitrate, a
decomposition product of ammonia fertilizer, has gross-
ly polluted ground water. The consumption of nitrate-
rich water leads to a serious disease in infants com-
monly known as "blue babies" (methemoglobinemia).
In many irrigated regions, automatic fertilizer feeders
attached to irrigation sprinkler systems are becoming
increasingly popular. When the irrigation-well pump is
shut off, water flows back through the pipe system into
the well. This creates a partial vacuum in the lines that
may cause fertilizer to flow from the feeder into the well.
It is possible that some individuals are dumping fertil-
izers (and perhaps even pesticides) directly into the well
to be picked up by the pump and distributed to the
sprinkler system. In this case, it directly contributes
to ground-water pollution.
Animal Feedlots. Animal feedlots cover relatively small
areas but provide a huge volume of animal wastes.
These wastes have polluted both surface and ground
water with large concentrations of nitrate. Even small
feedlots and liveries have created local but significant
problems. In a few areas the liquid runoff from feedlots
is collected in lined basins and pumped onto adjacent
grounds as irrigation waters, providing a luxuriant
growth.
Paniculate Matter from Airborne Sources. A relatively minor
source of ground-water pollution is caused by the
fallout of particulate matter originating from smoke,
flue dust, or aerosols. Some of the particulate matter is
water-soluble and toxic. An example of this type of
112
-------
pollution is airborne chromium-rich dust that discharged
through the roof ventilators of a factory in Michigan
and accumulated on the downwind side of the plant. As
Figure 125 shows, the highly soluble chromium com-
pounds rapidly infiltrated and polluted a local municipal
drinking water supply. Along the Ohio River at Ormet,
Ohio, the airborne discharge of fluoride from an
aluminum processing plant has seriously affected dairy
operations and fluoride concentrations in ground water
at the plant exceeded 1,000 mg/1 in the mid 1970's.
Ground-Water Quality Problems that Originate in the Ground
Above the Water Table
Many different types of materials are stored, extracted,
or disposed of in the ground above the water table.
Table 10 shows that water pollution can originate from
many of these operations.
Septic Tanks, Cesspools, and Privies. Probably the major
cause of ground water pollution in the United States is
effluent from septic tanks, cesspools, and privies,
although each site is small, as shown in Figure 126.
Individually of little significance, these devices are im-
portant in the aggregate because they are so abundant
and occur in every area not served by municipal or
privately owned sewage treatment systems. The area that
each point source affects is generally small, since the
quantity of effluent is small, but in some limestone areas
effluent may travel long distances in subterranean
cavern systems.
Holding Ponds and Lagoons. The second major source of
ground-water pollution is holding ponds and lagoons.
As Figure 127 shows, these ponds and lagoons com-
monly consist of relatively shallow excavations that
range in surface area from a few square feet to many
acres. In some places they are euphemistically called
"evaporation" ponds. Such ponds were commonly used
to hold oil-field brines, and when the pond floors
became sealed, the operators would disc them to in-
crease infiltration. Holding ponds are also used to store
municipal sewage as well as large quantities of wastes
including a host of industrial chemicals. The latter are
generally characterized by highly concentrated solutions
that may contain toxic compounds.
Special problems develop with holding ponds and
lagoons in limestone terrain where extensive near-surface
solution openings have developed. In Florida, Alabama,
Missouri, and elsewhere, municipal sewage lagoons have
collapsed into sinkholes draining raw effluent into wide-
spread underground openings. In some cases the sewage
has reappeared in springs and streams several miles
away. Wells producing from the caverns could easily
become polluted and lead to epidemics of water-borne
diseases.
Holding ponds are commonly considered to be liquid-
tight but the vast majority leak. Although rarely
reported, ground-water pollution caused by leaking
holding ponds at a large number of industrial sites has
been so extensive that all of the water supplies on the
plant property are unusuable for many purposes without
treatment. As a result, expensive treatment plants have
been required. Moreover, the ground water may be so
Windblown chrome-laden dust
Sand and gravel aquifer ;::.::.:.--nHUi
Figure 125. Air Pollution Can Lead to Ground-Water Pollution.46
113
-------
-------
polluted that it exceeds surface water effluent standards
and it cannot be pumped into adjacent streams.
Oil-field brines, a highly mineralized salt solution, are
particularly noxious and without doubt they have locally
polluted both surface and ground water in every state
that produces oil. The brine, an unwanted byproduct, is
produced with the oil. In many states it is disposed of
by placing it in holding ponds from which it infiltrates
into the ground. Commonly the oil well has been long
abandoned before it becomes apparent that the adjacent
ground water is polluted. This, in turn, may leave no
possibility for recovery of damages by the landowner.
Sanitary Landfills. Sanitary landfills generally are con-
structed by placing wastes in excavations and covering
the material daily with soil—thus the term "sanitary" to
indicate that garbage and other materials are not left ex-
posed to produce odors or smoke or attract vermin and
insects. Even though a landfill is covered, however,
leachate may be generated by the infiltration of precip-
itation and surface runoff. Fortunately many substances
are removed from the leachate as it filters through the
unsaturated zone, but leachate may pollute ground
water and even streams if it discharges at the surface as
springs and seeps.
At one site, rejected transformers and capacitors con-
taining polychlorinated biphenyls from an industrial
plant were disposed of in a municipal landfill. A number
of stillbirths and birth defects soon occurred in cattle
that drank water from a nearby stream. Analyses of the
water showed large concentrations of PCB, the origin of
which was, without question, the landfill.
Waste Disposal in Excavations. Following the removal of
clay, limestone, sand, and gravel, or other material, the
remaining excavations are traditionally left unattended
and often are used as unregulated dumps. The quantity
and variety of materials placed in dumps and excavations
are almost limitless. Excavations also have been used for
the disposal of liquid wastes, such as oil-field brines and
spent acids from steel mill operations. Many other ex-
cavations serve as disposal sites for snow removed from
surrounding streets and roads—snow that commonly
contains a large amount of salt. Disposal of these and
other wastes in excavations may lead to ground-water
pollution.
Leakage from Underground Storage Tanks. A growing prob-
lem of substantial potential consequence is leakage from
storage tanks and from pipelines leading to such tanks.
Gasoline leakage has caused severe hazardous pollution
problems throughout the nation. Gasoline floats on the
ground-water surface and leaks into basements, sewers,
wells, and springs, causing noxious odors, explosions,
and fires. A single-wall steel tank has a life expectancy
of 18 years and costs about $1 per gallon to replace. A
cleanup operation will generally exceed $70,000.
Leakage from Underground Pipelines. Literally thousands of
miles of buried pipelines crisscross the United States.
Leaks, of course, do occur, but it may be exceedingly
difficult to detect and locate them. Leaks are most likely
to develop in lines carrying corrosive fluids. An example
occurred in central Ohio where a buried pipeline carried
oil-field brine from a producing well to a disposal well.
The corrosive brine soon weakened the metal pipe,
which then began to leak over a length of several tens
of yards. The brine infiltrated, polluting the adjacent
ground water, then flowed down the hydraulic gradient
and discharged into a stream. During the ensuring
months, nearly all of the vegetation between the leaking
pipeline and the stream was killed. The leaking area of
the pipe was detected only because of the dead vegeta-
tion and salty springs.
A vexing problem of chromium compounds that pol-
luted several shallow wells in Michigan was traced to a
leaky sewer transporting metal finishing wastes. Radio-
active materials have also leaked from pipelines. The
several leaks reported at the Hanford A.E.C. Works
came about as a result of loaded, underground tanks
settling differentially into the subjacent earth materials,
causing the pipelines carrying radioactive waste to break
at joints.
Artificial Recharge. Artificial recharge includes a variety
of techniques used to increase the amount of water infil-
trating an aquifer. It consists of spreading the water
over the land or placing it in pits or ponds, from which
the water will seep into the ground, or pumping water
through wells directly into the aquifer. As water
demands continue to increase, there is no doubt that
artificial recharge will become more popular as a
ground-water management tool.
Waters used for artificial recharge consist of storm
runoff, excess irrigation water, stream flow, cooling
water, and treated sewage effluent, among others. Obvi-
ously the quality of water artificially recharged can have
a deleterious effect on the water in the ground under
certain conditions.
Sumps and Dry Wells. Sumps and dry wells may locally
cause some ground-water pollution and, in places where
these structures are adjacent to a stream, bay, lake, or
estuary, may pollute such surface-water bodies and
perhaps lead to a proliferation of algae and other water
weeds. These structures are commonly used to collect
runoff or spilled liquids, which will infiltrate through
the sump. Sumps and dry wells are typically installed to
solve surface drainage problems, so they may transmit
whatever pollutants are flushed into the well to ground
water.
Graveyards. Leachate from graveyards may cause
ground-water pollution, although cases are not well
documented. In some of the lightly populated glaciated
regions in the north central part of the United States,
graveyards are commonly found on deposits of sand
and gravel, because these materials are easier to exca-
vate than the adjacent glacial till and, moreover, are
better drained so that burials are not below the water
table. Unfortunately, these same sand and gravel
deposits may also serve as a major source of water sup-
ply. Graveyards are also possible sources of pollution in
115
-------
many hard rock terrains where there are sinkholes or a
thin soil cover.
Ground-Water Quality Problems that Originate in the Ground
Below the Water Table
Table 10 lists a number of major causes of ground-
water pollution produced by the use and misuse of
space in the ground below the water table.
Waste Disposal in Wet Excavations. Following the cessation
of various mining activities, the excavations are com-
monly abandoned; eventually they may fill with water.
These wet excavations have been used as dumps for
both solid and liquid wastes. The wastes, being directly
connected to an aquifer, may cause extensive pollution.
Furthermore, highly concentrated leachates may be
generated from the wastes due to seasonal fluctuations
of the water table. In the late 1960's at a lead-zinc mine
in northwestern Illinois, processing wastes were
discharged into an abandoned mine working. The
wastes, moving slowly in the ground water, polluted
several farm wells. Analyses of water from several of
the polluted wells showed high concentrations of dis-
solved solids, iron, sulfate, and, more importantly,
heavy metals and cyanide.
Drainage Wells and Canals. Where surficial materials con-
sist of heavy clay, flat-lying land may be poorly
drained and contain an abundance of marshes and
ponds. Drainage of this type of land is generally accom-
plished with field tiles and drainage wells. As Figure 128
shows, a drainage well is merely a vertical, cased hole in
the ground or in the bottom of a pond that allows the
water to drain into deeper, more permeable materials.
The pond water may be polluted which, in turn, leads
to deterioration of water quality in the receiving aquifer.
Deepening of stream channels may lower the water
table. Where the fresh-saltwater interface lies at shallow
depths, lowering of the water table (whether by channel-
ization, pumping, or other causes) may induce upward
migration of the saline water; it may even flow into the
deepened channel and pollute the surface water, as
Figure 129 shows. Under these circumstances, reduction
of the depth to fresh water can result in a rise in the
level of saline water several times greater than the dis-
tance the fresh water level is lowered.
In some coastal areas, particularly in Florida, the
construction of extensive channel networks has per-
mitted tidal waters to flow considerable distances
inland. The salty tidal waters infiltrate, increasing
the salt content of the ground water in the vicinity of
the canal.
Well Disposal of Wastes. For decades, man has disposed
of liquid wastes by pumping them into wells. Since
World War II, a considerable number of deep well
injection projects have come into existence, usually at
industrial sites. Industrial disposal wells range in depth
from a few tens of feet to several thousand feet. The in-
jection of highly toxic wastes into come of these wells
has led to several water-pollution problems. The prob-
lems are caused by the pollution of fresh water due to
direct injection into the aquifer as well as leakage of
pollutants from the well head, through the casing, or
via fractures in confining beds. Injection of liquid
wastes near Denver by means of deep well disposal ap-
parently caused an increase in the frequency of local
earthquakes. Deep well injection in the vicinty of Sar-
nia, Ontario, caused several long-abandoned brine wells
in Michigan to flow because of the greatly increased
aquifer pressure.
Exclusive of oil-field brine, most deep well injection
operations are tied to the chemical industry. Well depths
range from 1,000 to 9,000 feet and average 4,000 feet.
The deepest wells are found in Texas and Mississippi.
Original surface of pond Drained part of pond
^>;°;.- .••»•••:•.• v:-:>\-:v$
&£*.•• vXY£L; ;'"&• ££$
:*i
t*t«?
0.°%.°-^\;AV»><»^0- ^
LM'
**.
s^ New piezometric sufrace — —— —— —•
•Old piezometric surface •
Unr-
-*1
:•::?:.•:.:••:.%•:•:>:.<•{::;•..
f':'.;-:.?::?:./.':/.':."^.-.'
A.
»*fc:
, » o.
• • o
— Aquiclude ~
Figure 128. Diagram Showing Drainage of a Pond into an Aquifer Through a Drain Well.47
116
-------
Original channel
«;'."« '-*,—. ___
! Dredged channel'^
Figure 129. Diagram Showing Migration of Saline Water Caused by Lowering of Water Levels in an Effluent Stream and Streamside
Aquifer Hydraulically Connected to an Underlying Saline Water Aquifer.48
As of October 1983, EPA reported the existence of at
least 188 active hazardous waste injection wells in the
United States. There was an additional 24,000 wells used
to inject oil-field brine.
Properly managed and designed underground injec-
tion systems can be effectively used for storage of
wastes deep underground and may permit recovery of
the wastes in the future. Before deep well disposal of
wastes is permitted by EPA, however, there must be an
extensive evaluation of the well system design and instal-
lation, the waste fluids, and the rocks in the vicinity of
the disposal well.
Underground Storage. The storage of material under-
ground is attractive from both economic and technical
viewpoints. Natural gas is one of the most common sub-
stances stored in underground reservoirs. However, the
hydrology and geology of underground storage areas
must be well understood in order to insure that the
materials do not leak from the reservoir and degrade
adjacent water supplies.
Secondary Recovery. With increased demands for energy
resources, secondary recovery, particularly of petroleum
products, is becoming even more important. Methods of
secondary recovery of petroleum products commonly
consist of injection of steam or water into the producing
zone, which either lowers the viscosity of the hydro-
carbon or flushes it from the rocks, enabling increased
production. Unless the injection well is carefully moni-
tored and constructed, fluids can migrate from a leaky
casing or through fractures in confining units.
Mines. Mining has caused a variety of water pollution
problems. These problems are caused by pumping of
mine waters to the surface, by leaching of the spoil
material, by waters naturally discharging through the
mine, and by milling wastes, among others. Literally
thousands of miles of stream and hundreds of acres of
aquifers have been polluted by highly corrosive mineral-
Figure 130. Diagram Showing Migration of Saline Water
Caused by Dewatering in a Fresh-Water Aquifer Overlying a
Saline-Water Aquifer.'®
117
-------
Potentiometric Surface of
Saline-Water Aquifer
water table
corroded
casing
Potentiometric
surface
Water table
Figure 131. Upward Leakage and Flow Through Open Holes.
Some Important Aquifers Have Been Ruined by Improper Drill-
ing Practices.
ized waters originating in coal mines and dumps in
Appalachia. In many western states, mill wastes and
leachates from metal sulfide operations have seriously
affected both surface water and ground water.
Many mines are deeper than the water table, and in
order to keep them dry, large quantities of water are
pumped to waste. If salt water lies at relatively shallow
depths, Figure 130 shows that the pumping of fresh-
water for dewatering purposes may cause an upward
migration of the salt water, which may be intercepted
by the well. The mineralized water most commonly is
discharged into a surface stream.
Exploratory Wells and Test Holes. Literally hundreds of
thousands of abandoned exploratory wells dot the coun-
tryside. Many of these holes were drilled to determine
the presence of underground mineral resources (seismic
shot holes, coal, salt, oil, gas, etc.). The open holes per-
Si* Contami nated >:::•:•
:&•?:?: aquifer ^S-SSj^Sxi^ix:::::::::
Figure 132. Downward Leakage. Contamination of One Aquifer
Can Affect Others in a Multi-Aquifer System.
mit water to migrate freely from one aquifer to another.
As Figures 131 and 132 show, a freshwater aquifer
could thus be joined with a polluted aquifer or a deeper
saline aquifer, or polluted surface water could drain into
freshwater zones.
Abandoned Wells. Another major cause of ground-water
pollution is the migration of mineralized fluids through
abandoned wells. In many cases when a well is aban-
doned the casing is pulled (if there is one) or the casing
may become so corroded that holes develop. This per-
mits ready access for fluids under higher pressure to
migrate either upward or downward through the aban-
doned well and pollute adjacent aquifers. In other cases,
improperly cased wells allow high-pressure artesian
saline water to spread from an uncased or partly cased
hole into shallower, lower-pressure aquifers or aquifer
zones, resulting in widespread salt intrusion.
River in Flood
Contaminated
>^ Water -**s
s^^a^'gsgg'Ea^Bg^s^giigasi^iS-^-'^'-^^^ —-^
L
Gravel Pack Too
Close to Surface
Wafer tab!
Aquifer
Aquiciude
Figure 133. Diagram Showing Flood Water Entering a Well Through an Improperly Sealed Gravel Pack.
118
-------
Water-Supply Wells. Improperly constructed water-supply
wells may either pollute an aquifer or produce polluted
water. Dug wells, generally of large diameter and
shallow depth, and poorly protected, are commonly pol-
luted by surface runoff flowing into the well. As Figure
133 illustrates, other pollution has been caused by infil-
tration of water through polluted fill around a well or
through the gravel pack. Still other pollution has been
caused by barnyard, feedlot, septic tank, or cesspool ef-
fluent draining directly into the well. Many pollution
and health problems can arise because of poor well
construction.
Ground-Water Development. In certain situations pumping
of ground water can induce significant water-quality
problems. The principal causes include interaquifer leak-
age, induced infiltration, and landward migration of sea
water in coastal areas. In these situations the lowering
of the hydrostatic head in a freshwater zone leads to
migration of more highly mineralized water toward the
well site. Undeveloped coastal aquifers are commonly
full, the hydraulic gradient slopes towards the sea, and
freshwater discharges from them through springs and
seeps into the ocean, as shown in Figure 134. Extensive
pumping lowers the fresh-water potentiometric surface
permitting sea water to migrate toward the pumping
center. Figure 135 shows a similar predicament which
occurs in inland areas where saline water is induced to
A. Natural Conditions
flow upward, downward, or laterally into a fresh water
aquifer due to the decreased head in the vicinity of a
pumping well. Wells drilled adjacent to streams induce
water to flow from the streams to the wells. If the
stream is polluted, induced infiltration will lead to
deterioration of the water quality in the aquifer.
B. Sea-water Intrusion
Figure 134. Sea-Water Intrusion is Caused by Overpumping of
Coastal Aquifers.
Figure 135. Diagram Showing How a Pumping Well Can Induce
Highly Mineralized Water to Flow from a Saline Aquifer into a
Fresh-Water Aquifer.
Natural Controls on Ground-Water Contamination
As Deutsch47 clearly pointed out, there are four ma-
jor natural controls involved in shallow ground-water
contamination. The first control includes the physical
and chemical characteristics of the earth materials
through which the liquid wastes flow. A major attenuat-
ing effect for many compounds is the unsaturated zone,
which has been called the "living filter". Many chemical
and biological reactions in the unsaturated zone lead to
contaminant degradation, precipitation, sorption, and
oxidation. The greater the thickness of the unsaturated
zone, the more attenuation is likely to take place. Below
the water table, the mineral content of the medium
probably becomes more important because various
clays, hydroxides, and organic matter take up some of
the contaminants by exchange or sorption. Many of the
other minerals may have no effect on the contaminants
with which they come into contact.
The second major control includes the natural proc-
esses that tend to remove or degrade a contaminant as it
flows through the subsurface from areas or points of
recharge to zones or points of discharge. These proc-
esses include filtration, sorption, ion-exchange, disper-
sion, oxidation, and microbial degradation, as well as
dilution.
The third natural control relates to the hydraulics of
the flow system through which the waste migrates,
beginning with infiltration and ending with discharge.
The contaminant may enter an aquifer directly, by flow-
ing through the unsaturated zone, by interaquifer leak-
age, by migration in the zone of saturation, or by flow
through open holes.
The final control is the nature of the contaminant.
This includes its physical, chemical, and biological char-
119
-------
acteristics and, particularly, its stability under varying
conditions. The stability of the more common constit-
uents and the heavy metals are fairly well known
although more complex than commonly realized. On the
other hand, the stability of organic compounds, particu-
larly synthetic organic compounds, has only recently
come under close inspection and actually little is known
of their degradation and mobility in the subsurface.
This fact has been brought clearly to the attention of
the general public by the abundance of reported inci-
dences of ground-water contamination by EDB, TCE,
and DBCP.
To a large extent, it is the aquifer framework that
controls the movement of ground water and contam-
inants. Of prime importance, of course, is the hydraulic
conductivity, both primary and secondary. In the case
of consolidated sedimentary rocks, primary permeabil-
ity, or that which came into being with the formation of
the rock, in many respects is more predictable than
secondary permeability, which came later and includes
fractures and solution openings, among others. In sedi-
mentary rocks similar units of permeability tend to
follow bedding planes or formational boundaries, even
if the strata are inclined. Permeable zones are most
often separated by layers of fine-grained material, such
as clay, shale, or silt, which serve as confining beds.
Although leakage through confining beds is the rule
rather than the exception, both water and contaminants
are more likely to remain in a permeable zone than to
migrate through a thickness of several units of different
permeability. The movement of ground water and con-
taminants through larger openings, such as fractures,
complicates the assumed picture. Figure 136 illustrates
this movement. Not only can the velocity change dra-
matically, but in fracture flow, much of the attenuation
capacity is lost, and it is difficult to predict local direc-
tions of flow.
The geologic framework, in conjunction with surface
topography, also exerts a major control on the config-
uration of the water table and the thickness of the
unsaturated zone. Generally speaking, a deposit of per-
meable surficial sand and gravel would be characterized
by a water table that is relatively flat. In contrast, a
covering of glacial till, which is typically fine-grained,
would be characterized by a water-table surface that
more closely conforms to the elevation of the land sur-
face. The position of the water table is important not
only because it is the boundary between the saturated
and unsaturated zones, but also because it marks the
bottom and, therefore, the thickness of the unsaturated
material.
In many, if not most, contaminated areas, the water
table has been or is intermittently affected by pumping.
The resulting cone of depression on the water table
changes both the hydraulic gradient and ground-water
velocity resulting in flow to a discharging well. A
change in gradient and velocity also occurs in the vicin-
ity of recharge basins (lagoons, pits, shafts, etc.),
because the infiltrating water forms a mound in the
water table. As Figure 137 shows, the mound causes
radial flow and, therefore, contaminants will move in
UJjMSn
*$£^g!S&^-J
Legend:
Flow Direction of Leachate
Enriched Ground Water
Leachate Enriched
Ground Water
Figure 136. Movement of Ground-Water and Contaminants
Through Large Openings.
directions that are different than the regional hydraulic
gradient, at least until the mounding effects are over-
come and the regional gradient is reestablished.
Ground-water or interstitial velocity is controlled by
the hydraulic conductivity, gradient, and effective
porosity. Water movement through a permeable gravel
120
-------
Plan View
• Unsaturated zone :;••:::
•'•' of oercolation •••'•'••::•'•'
::; Original water table \H.l7
Figure 137. Diagrams Showing Lines of Flow from a Recharge Mound on a Sloping Water Table.
with a gradient of 10 feet per mile averages about 60
feet per day, but in a clay with the same gradient and
no secondary permeability the water movement would
be only about 1 foot in 30,000 years. In most aquifers,
ground-water velocity ranges from a few feet per day to
a few feet per year.
Carlston and others49 determined that the mean
residence time of ground water in a basin in Wisconsin
was about 45 days and in New Jersey about 30 days.
This study shows that ground water may discharge into
closely spaced streams in humid areas within a few days
to a few months. On the other hand, in less permeable
terrains ground water and contaminants may remain in
the ground for years or even decades.
121
-------
Leachate
The causes of ground-water pollution are many, but it
is the source that needs special consideration. For exam-
ple, an accidental spill from a ruptured tank may pro-
vide a considerable volume of liquid with an extremely
high concentration that is present only during a very
short time span, but leachate generated from a landfill
may consist of a large volume with a low concentration
that spans a period of many years. Once it reaches the
water table, the accidental spill might move as a con-
servative contaminant because of its high concentration
despite the fact that it might be degradable in smaller
concentrations. The leachate is more likely to be at-
tenuated by microbial degradation, sorption, dilution,
and dispersion.
In the case of landfills, leachate is a liquid that has
formed as infiltrating water migrates through the waste
material extracting water-soluble compounds and partic-
ulate matter. The mass of leachate is directly related to
precipitation, assuming the waste lies above the water
table. Much of the annual precipitation, including
snowmelt, is removed by surface runoff and evapo-
transpiration; it is only the remainder that is available
to form leachate. Since the landfill cover, to a large
extent, controls leachate generation, it is exceedingly
important that the cover be properly designed, main-
tained, and monitored in order to minimize leachate
production.
Schuller and others50 described the effect of regrad-
ing, installation of a PVC topseal and revegetation of a
landfill in Windham, Connecticut. As Figures 138 and
139 illustrate, field data clearly indicate that the cover
reduced infiltration and leachate generation, which
caused a reduction in the size and concentration of the
leachate plume.
According to an EPA estimate51, a disposal site con-
sisting of 17 acres with 10 inches/year of infiltration
could produce 4.6 million gallons of leachate each year
for 50 to 100 years. This estimate, of course, is site
dependent.
Feen and others52 described a landfill in Cincinnati,
Ohio, and estimated leachate generation using a water
balance technique. The data in Table 11 show that per-
colation through the landfill was calculated to occur
only during January through April and in December. In
this 50 acre landfill, the leachate generation averaged
about 11,394,000 gallons/year, 949,475 gallons/month,
or 31,216 gallons/day. Considering only the calculated
values for the months when precolation was assumed to
occur, leachate generation averaged about 113,900
gallons/day in January, 120,740 in February, 96,400 in
March, 31,700 in April, and 17,500 gallons/day in
December, in contrast to the annual daily average of
31,216 gallons.
The Cincinnati calculations appear unusual because
the spring runoff in Ohio normally occurs during March
and April and, since there is a soil-moisture excess at
that time, most ground-water recharge takes place dur-
ing this interval. Moreover, ground-water recharge may
occur any time there is rain. Therefore, one must use
Pond 3
c^
Key
• Control Point
Specific Conductance Contour
(micromhos/cm)
Figure 138. Distribution of Specific Conductance, May 19,1981.
Key
• Control Point
Specific Conductance Contour
(micromhos/cm)
Figure 139. Distribution of Specific Conductance, November 12,
1981.
caution when applying water balance techniques to esti-
mate leachate generation. The method may provide a
good estimate or long-term average, but it likely pro-
duces an estimate of the total volume of recharge that is
too low.
The physical, chemical, and biological characteristics
of leachate are influenced by: (1) the composition of the
waste, (2) the stage of decomposition, (3) the microbial
activity, (4) the chemical and physical characteristics of
the soil cover and of the landfill, and (5) the time rate
of release (recharge). Since all of the above can range
within extremely wide limits, it is possible to provide
only a general range in concentration of leachate constit-
uents, as Table 12 shows.
It is also important to account for the fact that mate-
rials placed in landfills may range widely depending on
the season. For example, many municipal landfills are
used to dispose of snow and ice, which may contain
large concentrations of calcium, sodium, and chloride
from de-icing salts. This could lead to the generation of
leachate that varies seasonally, particularly in regard to
the chloride concentration. It is also important to
remember that leachate collected from a seep at the base
of a landfill should be more highly mineralized than the
122
-------
Parameter*
PET
P
CR/O
R/O
1
I-PET
IN EG (I-PET)
ST (Table C)
AST
AET
PERC
J
0
80
0.17
14
66
+66
150
0
0
+ 66
F
2
76
0.17
13
63
+ 61
150
0
2
+ 61
M
17
89
0.17
15
75
+ 58
150
0
17
+57
A
50
82
0.17
14
68
+ 18
(0)
150
0
50
+ 18
M
102
100
0.17
17
83
-19
-19
131
-19
102
0
J
134
106
0.13
14
92
-42
-61
99
-32
124
0
J
155
97
0.13
13
84
-71
-132
61
-38
122
0
A
138
90
0.13
12
78
-60
-192
41
-20
98
0
S
97
73
0.13
9
64
-33
-225
33
-8
72
0
O
51
65
0.13
8
57
+ 6
39
+ 6
51
0
N
17
83
0.13
11
72
+ 55
94
+ 55
17
0
D
3
84
0.17
14
70
+ 67
150
+ 56
3
+ 11
Annual
766
1025
154
872
+ 106
658
213
'The parameters are as follows: PET, potential evapotranspiration; P, precipitation; C^Q, surface runoff coefficient; R/O, surface runoff, I,
infiltration; ST, soil moisture storage; AST, change in storage; AET, actual evapotranspiration; PERC, percolation. All values are in
millimeters (1 inche = 25.4 mm).
Table 11. Water Balance Data for Cincinnati, Ohio.
Constituents
BOD6, mg/l
COD, mg/l
Ammonia-N, mg/l
Hardness, mg/l as
CaCO3
Total Iron, mg/l
Sultate, mg/l
Specific Conductance,
mmhos
Operating Landfill
1,800
3,850
160
900
40.4
225
3,000
Abandoned Landfill
18
246
100
290
2.2
100
2,500
Table 12. Comparison of Chemical Characteristics of Leachate
from an Operating Landfill and a 20-Year-Old Abandoned Land-
fill in Southeastern Pennsylvania.53
leachate present in the underlying ground water, which
is diluted.
Cyclic Changes in Ground-Water Quality
It is commonly assumed and often reported that
natural ground-water quality is nearly constant at any
particular site. Field data substantiate this assumption,
and logic leads to the same conclusion if the aquifer is
not submitted to a new stress. On the other hand, mul-
tiple samples from a single well are very likely to show
slight changes in concentrations of specific constituents,
due to differences in sample collection, storage, and
analytical technique.
Conversely, investigators are finding with increasing
frequency that ground-water quality, at least in shallow
or surficial aquifers, can change significantly and rather
rapidly, perhaps as much as an order of magnitude
within a few hours or days, even though there is no
source of man-made contamination.
Deeper or confined aquifers generally are character-
ized by a nearly constant chemical quality that, at any
particular site, reflects the geochemical reactions that
occurred as the water migrated through confining layers
and aquifers from its recharge area to the point of col-
lection or discharge.
The quality of deeper water can change, but generally
not abruptly, in response to stresses on the aquifer
system. Changes in hydrostatic head brought about by
pumping, for example, may cause leakage of more
highly mineralized water from adjacent units into the
producing zone. This leakage may be due to fluid
migration along the well casing or gravel pack or by
leakage through confining beds or abandoned wells or
exploration holes. Another cause of chemical change is
waste disposal, particularly through well injection.
Surficial or shallow aquifers, however, are not well
protected from chemical changes brought about by
natural events occurring at the land surface or from
man-induced pollution. Surficial aquifers are, in fact,
the most susceptible to rapid and sometimes dramatic
changes in quality. Some of the changes are related to
man's activities; others are not.
The Concept of Cyclic Fluctuations
Several years ago Pettyjohn29-54 described cyclic fluc-
tuations of ground-water quality. The mechanisms that
Explanation
A
Evaporation pit
A
Filled evaporation
pits
864-
Altitude of
ground-water
surface
~* ~
Monitoring wells
Scale in feet
Figure 140. Water-Level Surface and Monitoring Wells at the
Contaminated Site.
123
-------
lead to cyclic fluctuations will be discussed in greater
detail here because both the cause and effect can have a
significant impact on: (1) ground-water quality monitor-
ing and determination of background quality, (2) trans-
port and fate of organic and inorganic compounds, as
well as bacteria and viruses, and (3) monitoring well
design and installation.
The site that Pettyjohn54 used to develop the concept
of cyclic fluctuation was a particularly well-instrumented
area in central Ohio. The contamination area, shown in
Figure 140, lies on the floodplain of the Olentangy
River and is bounded on the east and south by small
intermittent streams and on the west by the river.
Underlain by shale, the alluvial deposits consist of sand,
gravel, silt, and clay, that range from 15 to 35 feet in
thickness. The water table lies from 1.5 to 5 feet below
land surface and fluctuates a maximum of a foot or so
throughout the year. Precipitation averages about 38
inches per year.
Oil production began at this site in mid-1964, but by
July 1965, all wells had been plugged. Ground-water
contamination occurred because of leakage of oil-field
brine from three holding ponds. One pond (Skiles) was
used from July 1, 1964 to June 30, 1965; about 126 bar-
rels of brine were placed in it. Two other ponds
(Slatzer), received 110,000 barrels of brine containing
about 35,000 mg/1 of chloride over approximately the
same time period.
When samples were first collected from 23 monitoring
wells in July 1965, the aquifer locally contained more
than 35,000 mg/1 of chloride. At this time the ponds
were abandoned and the two Slatzer pits were filled
with previously excavated materials that had formed
surrounding embankments. The fact that the ground
water contained higher concentrations of chloride than
the original brine is most likely due to increased concen-
trations in the holding ponds brought about by
evaporation.
As Figure 141 shows, an interesting relationship
becomes apparent when examining the area! extent of
ground-water contamination with time. Note that the
area enclosed by the 1,000 mg/1 isochlor during 1965-66
changed monthly but the changes did not necessarily
encompass smaller areas. The 1969 data show a similar
phenomenon. This suggests that a linear flushing rate
did not exist.
Of particular importance in the monitoring of this site
are three adjacent wells, one screened at a depth of 9
feet and another at 23 feet, while a third is gravel-
packed through much of its length (23 feet) and receives
water from the entire aquifer. Figure 142 illustrates
these three wells. The locations of these wells are shown
in Figure 140 at the position marked A. It is assumed
that the first two wells represent the quality that exists
at depths of about 7 to 9 feet and 21 to 23 feet, respec-
tively, and that the gravel-packed well provides a
composite sample of the reservoir. It is also assumed
that when the composite well had a higher concentration
than both the deep and shallow wells, the most highly
mineralized water was between 9 and 23 feet.
Figure 143 shows the chloride fluctuations in the three
wells during 1965, 1966, and 1969. Notice that at certain
times the highest concentrations occur at the shallowest
depths, at other times at the greatest depth, and at still
Sept., 1965-Jan., 1966
\
1966
1966
1969
1969
1969
Figure 141. Areal Extent of the Contaminated Area Enclosed by
the 1,000 mg/l (1965,1966) and the 500 mg/l Isochlors During
Selected Months. Contours Based on Data from Monitoring
Wells.
124
-------
other times the greatest concentration must lie some-
where in the middle of the aquifer. The only means for
accounting for the variable distribution is intermittent
reintroduction of the contaminant, which is puzzling in
view of the fact that oil-field activities ceased in June
1965 before any of these samples were collected.
Figures 144 and 145 show another technique for illus-
trating the temporal-vertical distribution of chloride in
the aquifer. These illustrations are based on monthly
data obtained from the three adjacent wells. Concentra-
tions at depths of 9 and 23 feet were measured; inter-
pretation of the chloride distribution between these
points was based on data from the fully penetrating
well. In October and November 1965, the highest
chloride concentrations were present at the shallowest
depth but from December 1965 to April 1966 the
highest concentrations were near the bottom of the
aquifer. Furthermore, in November and December, the
water in the middle of the aquifer was less mineralized
than that above or below.
•TTT Land surface TTTH
•^ Water level *
liiGravel' pack
Screen
Figure 142. Completion Details of Three Closely Spaced Moni-
toring Wells.
30,000
g 25,000
2
£ _ 20,000
£.£ 15,000
•g
1 10,000
O
5,000
1600
1400
1200
1000
800
600
400
200
0
JASONDJFMAMJJASONO
1965 1966
Jan.
May Sept.
1969
Figure 143. Fluctuation of Chloride Content in Closely Spaced Wells of Different Depths During 1965,1966, and 1969.
Although greatly reduced, in January 1969 the largest
concentration of chloride was again at the shallowest
depth, but the situation was reversed during April and
May. During February and March the central part of
the aquifer was less mineralized than adjacent parts. By
August there was only a slight chloride increase with
depth, but in September and October the greatest
concentrations again appeared in the central part of
the aquifer.
The chloride fluctuations that occurred during
1965-66 and 1969 are shown schematically in Figure
146. The October 1965 samples apparently were col-
lected shortly after a recharge event, which leached salt
from the ground and formed a highly concentrated
mass. This slowly sinking mass (1) was subsequently
replaced with less mineralized water. A month later, the
first mass had reached and was migrating along the bot-
tom of the aquifer when another recharge event oc-
curred (2). By December, the second mass had reached
the bottom of the aquifer and was moving toward the
river. Recharge events also occurred in January 1966 (3)
and in February 1966 (4). The February event represented
the spring runoff when evapotranspiration was minimal
and the soil-moisture content exceeded field capacity over
a wide area. This major period of recharge caused a large
influx of salt and by March, the aquifer was con-
taminated throughout a wide area. This mass eventually
discharged into the river.
In spite of the fact that brine disposal ceased by
mid-1965, Figure 146 shows that the aquifer was recon-
taminated several times during 1969. Following an
established pattern, small recharge events took place in
January, February, and March 1969.
This study indicates that water soluble substances on
the land surface or in the unsaturated zone may be
intermittently introduced into a shallow aquifer for
many years. The introduction of these contaminants is
dependent upon the chemistry of the waste and the soil
and upon the frequency of the recharge events. These
recharge events are controlled by evapotranspiration, by
the rate, duration, and intensity of precipitation, and by
soil-moisture conditions.
125
-------
Chloride concentration in mg/l
o
o
o
o o
o o
o o
O O
8 8
o
o
o
o
o
o
8 8:
8
m in in
T- CM
m m m
i- CM
o
o
o
in"
CM
10
3
OT
•o
C
§15
.c
Q.
3 20
25
Oct.
Nov.
1965
Dec.
Jan.
Feb. Mar.
1966
April
Figure 144. Vertical Distribution of Chloride at the Contaminated Site from October 1965 to April 1966
Chloride concentration in mg/l
§00 o o
o o o o
o o_ o. p,
c\f o T-" of o T-" cvf
o o o o
o o o o
o o o o
o T-" CM" o T-" CM" o
eg o
of o T-" c\T
* 5
10
•a
c
!
I 15
a.
0
Q
20
25
Jan. Feb. Mar. April May Aug. Sept. Oct.
1969
Figure 145. Vertical Distribution of Chloride at the Contaminated Site During 1969
126
-------
October 1965
January 1969
February 1969
Figure 146. Schematic Diagram Showing the Cyclic Movement of
Masses of Contaminated Water Through the Aquifer During
Selected Months in 1965, 1966, and 1969. Stippled Areas for
1965-66 Represent Concentrations in Excess of 20,000 mg/l and
for 1969 Represent Concentrations Greater than 500 mg/l.
Throughout most of the year in humid and semiarid
regions, the quantity of water that infiltrates and the
amount of contaminants that are washed into an aquifer
are relatively small. On the other hand, during the
spring recharge period significant quantities of contam-
inants may be flushed into the ground over wide areas.
Therefore, the major influx of contaminants occurs on
an annual basis, although minor recharge events may
occur at any time.
In order to account for cyclic fluctuations in ground-
water quality it is assumed that: (1) the unsaturated
zone may store a considerable volume of water-soluble
substances for long periods of time, and (2) the main
paths along which contaminants move through the
unsaturated zone to the water table consist largely of
fractures and macropores.
Large concentrations of many water-soluble
substances are stored within the unsaturated zone.
Nitrate storage has been extensively studied. For exam-
ple, in parts of western Kansas, the top 3 to 4 feet of
unplowed prairie soils have as much as 10,000 to 20,000
pounds of nitrogen per acre.55 A particularly interesting
case in Texas was described by Kreitler.56 In addition to
natural substances, such as nitrate, many man-related
contaminants are stored for years in the unsaturated
zone and their presence may lead to continual recontam-
ination of aquifers, as was the case adjacent to the
Olentangy River in Ohio.
The long-term effects of oil-field brine holding ponds
in central Ohio were investigated for several years by
geology students at The Ohio State University. One
pond adjacent to a producing well was excavated in
1968. Two years later, when the well was plugged, the
holding pond was filled, graded, and seeded. Figure 147
shows that the chloride concentration in the ground a
few inches below the water table in the vicinity of the
former pond was 36,000 mg/l some 10 years after oper-
ation began and 8 years after pond reclamation. This
particular area is characterized by a thick and very dense
glacial till of low primary permeability.
Figure 147. Chloride Concentrations in Shallow Ground Water 8
Years After Reclamation of an Oil-Field Brine Holding Pond
A few miles to the north, three ponds were con-
structed. Apparently they had not been used for many
months when the area was investigated since the
chloride concentration in them ranged from 100 to
3,200 mg/l. In the vicinity of these ponds, 18 holes were
hand augered to the water table where chloride concen-
trations ranged from 100 to 16,900 mg/l, as shown in
Figure 148. Most of the wells contained less than 2,000
mg/l chloride. In contrast to the area which contained
the one pond mentioned above, this three-pond site con-
sists of glacial till which contains several thin layers of
sand and gravel. The higher permeability presumably
accounts for the overall lower chloride concentration.
The preceding examples illustrate that substantial vol-
umes of both naturally occurring and man-related
chemical substances are stored in the unsaturated zone.
The quantity of water-soluble substances in storage
probably increases with decreasing grain size.
It is commonly assumed that ground-water recharge
cannot occur until there is a soil-moisture excess, that is
until field capacity is exceeded. Increasing evidence
clearly shows, however, that the concept of a distinct
wetting front or pistonlike displacement flow through
the main matrix of the soil is unlikely in a great many
127
-------
1900
100
100
650
100
I
' Scale, in feet
Figure 148. Chloride Concentration in Shallow Ground Water in
the Vicinity of Three Inactive Oil-Field Brine Holding Ponds.
cases.57"67 In fact, in many situations, much of the flow
occurs rapidly through macropores and fractures. In a
study of the Missouri Ozarks, Aley68 showed that water
entering macropores contributed five times as much
recharge as did diffuse flow. Aubertin69 found that in
sloping forested lands, macropores conducted large
quantities of water to depths of 30 feet or more. In the
majority of cases, at least during parts of each year,
ground-water recharge probably is a function both of
flow through macropores and fractures as well as
displacement flow.
Some macropores and fractures can be exceedingly
permeable. While drilling a test hole in north-central
North Dakota with a rotary drilling rig, circulation was
completely lost at a depth of about 20 feet. Even after
mixing two bags of bentonite and a bag of bran in the
drilling fluid, it was still not possible to regain circula-
tion in a fine sand. At this time a curious circumstance
was noticed—about 30 feet from the rig a fountain of
muddy water was flowing from the trunk of an oak tree
that had been cut down years before. Apparently, the
drill bit had cut through the remains of a large decayed
root of the dead tree.
In another case in the same general area, circulation
was lost at a depth of about 60 feet while drilling
through a thick section of glacial till. After mixing three
bags of bentonite with the drilling fluid, circulation was
finally regained but not before a spring had formed
about 20 feet from the rig. In this case, the drill bit
encountered a fracture at some unknown depth that was
more permeable than the cutting- and mud-filled
annulus of the hole.
The summer of 1980 in north-eastern Oklahoma was
exceedingly hot and dry. Large, extensive desiccation
crack systems were common throughout the finer tex-
tured soil areas within the state, including the author's
backyard. Figure 149 shows one of several such fracture
systems which was measured on August 7, 1980. This
fracture system was about 72 feet long, was as much as
2 inches wide, and ranged from 12 to 14 inches deep. A
metal probe could be pushed into the crack to a depth
of 34 inches.
Two garden hoses, which provided a combined flow
of 6 gpm, were placed in two separate locations in the
fracture system. Water flowed into the fracture for 132
minutes (792 gallons). At no time did water overflow
the fracture. When the water supply was shut off, the
water level dropped 2 inches in 15 seconds. This prim-
itive test clearly illustrates that fracture systems can col-
lect, store and transmit large volumes of water. Further-
more, where extensive desiccation cracks exist, it would
be unlikely that much, if any, overland flow could oc-
cur except, perhaps, in response to a heavy precipitation
event from a convective storm.
Neither macropores nor fractures are commonly as
large as those described above. Most may be barely
detectable without a close examination. Ritchie and
others60 suggested that the interfaces between adjacent
soil peds also serve as macropores. Moreover, these
openings need not extend to the land surface in order
for flow to occur in them.66 Nonetheless, water can
flow below the root zone in a matter of minutes.
Fractures as much as 2" wide,
12—14" deep.
792 gallons during 132 minute period.
Water level dropped 2" in 15 seconds
at end of test.
Injection rate = 6 gpm.
Scale, Feet
August 7,1980
Stillwater, OK
Figure 149. Sketch Map of a Desiccation Fracture in North-
eastern Oklahoma
128
-------
Thomas and Phillips57 suggested that this type of flow
does not appear to last more than a few minutes or per-
haps, in unusual cases, more than a few hours after
"cessation of irrigation or rain additions".
The preceding discussion demonstrates that the unsat-
urated zone can store a large volume of water-soluble
substances and that macropores and fractures can serve
as highly permeable connecting routes between the land
surface and the water table. The next consideration is
the movement of salts within and from the unsaturated
zone to the water table.
When a soil dries, desiccation cracks may form.
Moisture on fracture walls may quickly evaporate and a
strong capillary potential is developed that draws water
from the adjacent soil matrix to the fracture walls. The
soil water may be rather highly mineralized, either with
natural substances, such as nitrate, or with contam-
inants that may have been spilled or disposed of on the
land surface. Once the soil water reaches fracture or
macropore walls and then evaporates, the salts remain
as a water soluble lining.
During a period of rain or irrigation, Figure 150
shows that water may flow into these openings, dissolve
the water-soluble lining, and rapidly flow downward.
This may result in a highly concentrated solution that
quickly reaches the water table and substantially
degrades ground-water quality. Of course, some of the
fluids and salts migrate back into the soil matrix
because of head differences and the now reversed capil-
lary potential.
Figure 150. Fractures and Macropores Accumulate Water-
Soluble Linings as a Result of Evaporation (Upper). These
Linings are Dissolved and Flushed to the Water Table During
Recharge Events
Even though there may be a considerable influx of
contaminants through macropores and fractures to the
water table following a rain, the concentration of
solutes in the main soil matrix may change little, if at
all. This is clearly indicated in studies by Shuford and
others67 and again shows the major role of large open-
ings. On the other hand, in the spring, when the soil-
moisture content exceeds field capacity, some of the
relatively immobile or stagnant soil water may percolate
to the water table transporting salts with it. A similar
widespread recharge period may occur in some places
during the fall as a result of decreasing temperature and
evapotranspiration and of wet periods that might raise
the moisture content above field capacity.
In summary, the preceding suggests the following.
During the spring, a large quantity of water-soluble
substances may be leached from storage in the unsat-
urated zone over a wide area. The substances eventually
reach the water table and cause significant change in
ground-water quality. Although the quantity of leached
substances is larger than at any other time during the
year, the change may occur more slowly and the result-
ing concentration in ground water may not be at a max-
imum because of the diluting effect brought about by
the major influx of water over a wide area. A similar
but less dramatic change in quality may occur in the fall.
During summer months, ground-water quality
changes would be expected to occur more rapidly,
perhaps in a matter of hours, because of the large size
and abundance of the macropores and fractures. These
changes, however, may occur only over a relatively
small area because of the local nature of convective
storms.
Conclusions
Ecologic conditions in fractures and macropores
should be quite different from those in the main soil
matrix largely because of the greater abundance of ox-
ygen. Resulting, one might well expect different micro-
bial populations (types and numbers) and chemical
conditions in macropores and fractures than in the soil
matrix. Coupled with their far greater fracture permea-
bility, this may help to explain why some biodegradable
organic compounds or those that should be strongly
sorbed actually may reach the water table and move
with the ground water. This environment cannot be ade-
quately examined by means of column studies.
Some public health investigators have reported70 that
waterborne diseases seem to increase in the spring and
fall. This might conform to and be the result of ground-
water recharge through macropores and fractures.
In order to detect and evaluate cyclic fluctuations in
ground-water quality and determine background concen-
trations as well, it will be necessary to install monitoring
wells that can be used to measure vertical head differ-
ences and collect water samples from discreet sections of
the aquifer. Moreover, it will be necessary to collect
data frequently, perhaps weekly or even daily, until a
pattern can be established.
129
-------
Prediction of Contaminant Migration
In any ground-water pollution study it is essential to
obtain the background concentration of a wide variety
of chemical constituents, particularly those that might
be common both to the local ground water and a con-
taminant. As mentioned previously, the water in shallow
or surficial aquifers can undergo substantial fluctuations
in chemical quality. Therefore, it is not always a simple
task to determine background concentrations, particu-
larly of the more conservative constituents, such as
chloride or nitrate. In general, samples should be col-
lected during dry periods and not during or within a
week following a period of rain. Throughout much of
North America the major period of ground-water
recharge occurs in wetter periods of the year (generally
in the spring), while minor recharge events occur during
or immediately after a rain. These recharge events may
flush water-soluble compounds from the unsaturated
zone to the water table and may substantially change
the chemical quality of the ground water. Since the
quality of shallow ground water may fluctuate within
fairly wide limits during short intervals, it is essential to
determine background concentrations statistically by col-
lecting several samples at different times and from dif-
ferent depths.
The severity of ground-water pollution is partly
dependent on the characteristics of the waste or leach-
ate, that is, its volume, composition, concentration of
the various constituents, time rate of release of the con-
taminant, the size of the area from which the contam-
inants are derived, and the density of the leachate,
among others. Data describing these parameters are dif-
ficult to obtain and are lumped together into the term
"mass flow rate", which is the product of the contami-
nant concentration and its volume and recharge rate, or
leakage rate.
Once a leachate is formed it begins to migrate slowly
downward through the unsaturated zone where several
physical, chemical, and biological forces act upon it.
Eventually, however, the leachate may reach saturated
strata where it will then flow primarily in a horizontal
direction as defined by the hydraulic gradient. From this
point on, the leachate will become diluted due to a
number of phenomena, including filtration, sorption,
chemical processes, microbial degradation, dispersion,
time, and distance of travel. Figure 151 illustrates some
of these phenomena.
Filtration removes suspended particles from the water
mass, including particles of iron and manganese or
other precipitates that may have been formed by chem-
ical reaction. Dilution by sorption of chemical com-
pounds is caused largely by clays, metal oxides and
hydroxides, and organic matter, all of which function as
sorptive material. The amount of sorption depends on
the type of pollutant and the physical and chemical
properties of the solution and the subsurface material.
Chemical processes are important when precipitation
occurs as a result of excess quantities of ions in solu-
tion. Chemical processes also include volatization as
well as radioactive decay. In many situations, particu-
Porosity
Effective
porosity
C6
C5
C4 Dispersion
C2
C3 Sorption
Biological
Decay
Figure 151. Phenomena Which Can Dilute a Leachate
larly in the case of organic compounds, microbiological
degradation effects are not well known. It does appear,
however, that a great deal of degradation can occur if
the system is not overloaded and appropriate nutrients
are available.
Dispersion of a leachate in an aquifer causes the con-
centration of the contaminants to decrease with increas-
ing length of flow. It is caused by a combination of
molecular diffusion, which is important only at very low
velocities, and dispersion or hydrodynamic mixing,
which occurs at higher velocities in laminar flow
through porous media. In porous media, different
macroscopic velocities and flow paths that have various
130
-------
Direction of
Ground-Water Flow
in.
,.r*pll
<"*••::!•:•:•:
-i06'Milesji
5
in
a) Chloride Plume, Inel, Idaho
Transverse dispersivity; 450 feet
Time: 16 years
Disposal
fT.-x
A
Direction of
Ground-Water Flow
r
IliiPi
If
w
b) Chromium Plume, Long Island
Transverse dispersivity; 14 feet
Time- 13 years
Figure 152. Effect of Differences in Transverse Dispersivity on
Shapes of Contamination Plumes.
lengths are to be expected. Leachate moving along a
shorter flow path or at a higher velocity would arrive at
an end point sooner than that part following a longer
path or a lower velocity, which results in hydrodynamic
dispersion. Dispersion can be both longitudinal and
transverse and the net result is a conic form downstream
from a continuous pollution source. As Figure 153
shows, the concentration of the leachate is less at the
margins of the cone and increases toward the source.
Because dispersion is directly related to ground-water
velocity, a plume or slug will tend to increase in size
with more rapid flow within the same period of time.
Since dispersion is affected by velocity and the config-
uration of the aquifer's pore spaces, coefficients must
be determined experimentally or empirically for a given
aquifer. There is considerable confusion regarding the
quantification of the dispersion coefficient and many of
the published values are fitted values that cannot be
transferred.
Selection of dispersion coefficients that adequately
reflect conditions that exist in an aquifer is a problem
that can not be readily solved and herein lies one of the
major stumbling blocks of chemical transport models.
Often confused with the term dispersion (Dx = longitu-
dinal dispersion and D = transverse dispersion) is disper-
sivity («x, Oy). Dispersion includes velocity: to transform
from one to another requires either division or multi-
plication by velocity.
The rate of advance of a contaminant plume can be
retarded if there is a reaction between its components
and ground-water constitutents or if sorption occurs.
This is called retardation (Rd). The plume in which
sorption and chemical reactions occur generally will ex-
pand more slowly and the concentration will be lower
than the plume of an equivalent nonreactive leachate.
Hydrodynamic dispersion affects all solutes equally
while sorption and chemical reactions can affect various
constituents at different rates. As Figure 154 shows, a
leachate source that contains a number of different
solutes can have several solutes moving at different rates
due to the attenuation processes.
The areal extent of plumes may range within rather
wide extremes depending on the local geologic condi-
tions, influences on the hydraulic gradient, such as
pumping, ground-water velocity, and changes in the time
rate of release of contaminants.
The many complex factors that control the movement
of leachate and the overall behavior of contaminant
plumes are difficult to assess because the final effect
represents several factors integrated collectively. Like-
wise, concentrations for each constituent in a complex
waste are difficult to obtain. Therefore, predictions of
concentration and plume geometry, at best, can only be
used as estimates, principally to identify whether or not
a plume might develop at a site and, if so, to what ex-
tent. Models can also be used as an aid in determining
potential locations for monitoring wells and to test
various renovation or restoration schemes.
A graphical solution (nomograph) was developed to
provide a simple computational tool for the prediction of
leachate plume movement and corresponding concentra-
tion.71 It is often necessary to estimate the potential
distance of travel or length of time required for a plume
to migrate some distance in the saturated zone from a
point directly below a contaminant source. The concen-
tration of conservative elements in the plume, such as
chloride, can be estimated for some selected point in
space and time along a flow path that extends directly
downgradient from the source.
The nomograph is intended as a rapid means for
obtaining an approximate solution. It also aids in under-
standing the model. It is one-dimensional (restricted to a
line).
The Wilson-Miller72 equation was reformulated to
131
-------
0 500 1000 1500
i ii i
Scale in Feet
Figure 153. Ground-Water Velocity Exerts a Major Control on Plume Shape. Upper Plume V = 1.5 ft/day and Lower Plume V = .5
ft/day.
900-
450-
0-
450-
900-
900
1800
I
2700
I
3600
4500
1. Chlorobenzene Rjj = 35
2. Unknown = 15
3. Chloroform = 3
4. Chloride = 1
Plumes after 2800 days
Figure 154. Plume of a Leachate Containing Several Different Solutes
132
-------
introduce scale factors and to provide the basis for the
nomograph as follows:
C= QCQ exp [(x/Xp Vr^fiOerfc (0)]
4 QD VTTX/XD (56)
where:
x/XD-t/TD
Application allows mapping the center-line concentra-
tions of the plume, with respect to time, in one direction
(x distance) that is directly downgradient from the
source. Dilution-dispersive mixing and retardation
parameters are included in the solution. The equation
and the nomograph apply to only one chemical constit-
uent, such as chloride or dissolved solids, at a time.
Three scale factors are used in the nomograph as ratios
with the primary variables x (distance), t (time), and
QC0, (mass flow rate from source) in the forms of x/XD,
t/TD, and QC0/QD. The y distance is set to zero. The
scale factors are:
T -
D
(58)
(59)
QD = nm VDxDy (60)
Two of the three ratios are computed directly and the
third is then found using the nomograph. The factors
provide two conveniences. First, the ratios are dimen-
sionless except for concentration. Second, the scale fac-
tors combine the constant parameters, which makes it
easier to repeat computations of concentration (C) for
various positions along the x axis (x) or for different
times (t).
The nomograph in Figure 155 is designed to provide a
simple technique to estimate one of the following:
Application 1: The concentration (C) at a selected
distance (x) and time (t).
Application 2: The distance (x) where a selected
concentration (C) will occur at a
given time (t).
Nomograph for
Plume Center-Line
Concentration
10-8-
10-3~
10-"-
10-3-
10-2-
_QCo
QD
(Ib/ft3)
10 -
102-
103-
104-
105-
r10-2
-10-1
QC0
1 QD
(mg/l)
,,J
-102
-103
-105
-106
-107
-108
-109
-10-2
-10-1
C
(mg/l)
' 1
100
1,000
10,000
100,000
-102
-103
L104
Figure 155. Nomograph for Solutions of Time, Distance, and Concentration for Any Point Along the Principal Direction of Ground-
Water Flow.
133
-------
Application 2b:
Application 3: The time (t) when the concentration
will reach a selected concentration (C)
at a predetermined location.
As time passes, the concentration in a given area ap-
proaches a constant pattern known as steady state.
Results for steady-state conditions can be determined
for the first two applications as follows:
Application Ib: The maximum concentration (C) that
would occur at a selected distance
after a long time period.
The maximum distance (x) at which a
selected concentration (C) will occur
after a long time period.
The advantage of applications Ib and 2b is that it is
possible to predict, for example, the maximum distance
of plume migration for a given concentration threshold
or limit. Such concentrations might be those that have
been established as standards for the safe drinking water
by EPA. Alternatively, it is also possible to predict the
maximum leachate concentration that could be reached
at a specified distance from the landfill or lagoon.
The estimate of distance (x), time (t), and concentra-
tion (C) may require an adjustment of the concentration
value to correct for significant background concentra-
tions. In applications 1 and Ib the estimated concentra-
tion (C) must be added to the background concentra-
tion. In applications 2, 2b, and 3 the concentration
value used must be the remainder after subtracting the
natural background concentration.
The nomograph provides a visual representation of
the plume concentration. The solution can be found
easily for various locations, times, and concentrations.
This leads to gaining a "feel" for the nature of the
plume. As time passes, the concentration at a given
location reaches steady state. The steady state value for
concentration can be useful, for example, as a "worst
case" scenario (maximum concentration reached in the
infinite time). The upper line on the nomograph repre-
sents the time and distance at which steady state is
reached. It is easy to see that before steady state, small
changes in location or time correspond to large changes
in concentration. In a sense, the steepness of the non-
steady state time lines indicates that the "leading edge"
of the plume is relatively narrow and, therefore, passes
a given location in a relatively short period of time.
Behind the leading edge, the concentration remains con-
stant at the steady-state value represented by the steady-
state line on the nomograph.
Example Problem
The ground-water contamination case that is used in
this example occurred in South Farmingdale, Nassau
County, New York, and was described by Perlmutter
and Lieber.7^ Most of the data required for the solution
described below were obtained directly from their
report.
Contamination was caused by cadmium- and hexava-
lent chromium-enriched electroplating wastes that infil-
trated from disposal basins into a shallow glacial
aquifer. Apparently disposal began in 1941 and con-
tinued intermittently for several years. By the early
1960's a leachate plume originating at the disposal
ponds extended downgradient about 4,300 feet, and was
as much as 1,000 feet wide and as much as 70 feet
thick. Figure 156 shows that the plume extends to the
headwaters of Massapequa Creek, a small stream that
serves as a natural drain for part of the contaminanted
water.
Disposal Ponds
Plume
Massapequa Creek
N
1000 0 1000 2000
•••
Feet
Figure 156. Leachate Plume at South Farmingdale, New York
The surficial or upper glacial deposits, which are
Pleistocene in age, extend from the land surface to a
depth of 80 to 140 feet and lie on the Magothy Forma-
tion, a unit of stream deposits of Late Cretaceous age.
The water table in the surficial deposits lies from 0 to
about 15 feet below land surface.
The aquifer is in dynamic equilibrium and receives
about 22 inches or about 1 mgd (million gallons per
day) per square mile of recharge from precipitation. The
water-table gradient averages about 1 foot in 500 feet
and the water table undergoes an annual fluctuation of
2 to 3 feet. The direction of ground-water flow is south-
ward from the disposal ponds toward Massapequa
Creek. Estimates of ground-water velocity for the area
range from 0.5 to 1.5 feet per day. Reportedly, the
average velocity for the area is about 1 foot per day.
Chemical analyses of ground water in the South
Farmingdale area indicate that the background concen-
tration of hexavalent chromium is less than 0.01 mg/I.
Likewise, the concentration is also less than 0.01 mg/1
in Massapequa Creek upstream from the area of the
leachate plume. Along that stretch where the plume dis-
charges into the stream, the concentration of chromium
is substantially greater.
134
-------
As much as 200,000 to 300,000 gallons per day of ef-
fluent (equivalent to 52 pounds per day of chromium)
were discharged during the early 1940's to three disposal
pits, which have a combined area of about 15,470 square
feet. Since 1945 the volume of the waste stream has been
reduced substantially and eventually a treatment plant
was constructed. On two occasions the chromium con-
centration in the raw effluent was 28 and 29 mg/1.
The relatively clean nature (free of clay or organic mat-
ter) of the materials forming the surficial aquifer pre-
cluded any significant reduction in the chromium load in
the plume. That is, ion-exchange during movement was
negligible. Maximum chromium concentrations in the
plume ranged from about 40 mg/1 in 1949 to about 10
mg/1 in 1962.
The plume is about 200 feet wide at its origin at the
disposal ponds. It reaches a maximum length of about
4,300 feet and increases in width to about 1,000 feet.
Assuming a velocity of 1.5 feet per day, this plume has
an average longitudinal dispersion (Dx) of 105 ft2/day, a
transverse dispersion (D ) of 21 ft2/day, and dispersivities
of 70 ft (ax) and 14 ft. (a;v), respectively. Table 13 shows
a summary of the required data. Application of the
three methods to this example follows.
Table 13. Summary of Data for Example 1
thickness:
porosity:
velocity.
dispersion.
retardation:
volume flow rate1
source concentration.
mass flow rate:
or
m
n
V
R*
Q
Q°CO
QC°
= 110 feet
= 0.35
= 1.5 ft/day
= 105 ft2/day
= 21 ft2/day
= 1
= 26,800 ft3/day
= 31 mg/l
= 26,800 ft3/day x 31
= 52 Ib/day
mg/l
Application 1, illustrated in Figure 157:
To find concentration (C) for a distance (x) of 4,200
feet from the source and time (t) of 2,300 days,
calculate:
_x_ - 4,200 ft = 60 (Locate at A)
XD 70 ft
_t_ = 2,300 days = 50 (Locate curve E)
TD 46.7 days
,-10-2
Nomograph for
Plume Center-Line
Concentration
.-10-2
-10-1
C
(mg/l)
-10
-102
I-1Q3
L104
A 100
1,000
10,000
100,000
XD
Figure 157. Applications 1a and 1b: Using the Nomograph to Estimate the Concentration Given Values of Distance and Time
135
-------
QC0 _ (26,800 ft3/day) (31 mg/l)
QD 1,800ft3/day
= 460 mg/l (Locate at D)
°2e_ = 52 lb/daV = .029 Ib/ft3 (Locate at D)
QD 1,800ft3/day
Using Figure 157 draw a line vertically from A to the
intersection with the t/TD curve B, then horizontally
from B to C and from C through the scale D to E, giv-
ing a concentration (C) of 2.6 mg/l.
Application Ib, illustrated in Figure 157:
To find the maximum concentration for a given
distance for large time, use the steady-state line instead
of curve (t/TD). Proceeding as above, a concentration
(C) of 20 mg/l is read at F.
Application 2, illustrated in Figure 158:
To find distance (x) where a concentration (C) of 2.6
mg/l will occur at a time (t) of 2,300 days, calculate:
_t_ = 2,300 days = 50 (Locate at D)
TD 46.7 days
QC0 ^ (26,800 ft3/day) (31 mg/l)
QD 1,800ft3/day
= 460 mg/l (Locate at B)
Using Figure 158, locate the selected concentration at A.
Draw a line from A through B to C, then horizontally
from C to curve D, and vertically to E, giving:
Multiply by XD to determine distance (x):
x= (JL)(XD) = (60)(70 ft) = 4,200 feet
Application 2b, illustrated in Figure 158:
To find the maximum distance at which a selected con-
centration will reach a given value for large time, use the
steady-state line instead of curve (t/TD) in Figure 158.
f-10-2
10-8-T
-10-1
C
(mg/l)
Nomograph for
Plume Center-Line
Concentration
1,000 F 10,000 100,000
-109
L 10,10
: 1
-10
-102
.103
Figure 158. Application 2a: Using the Nomograph to Estimate the Distance for Given Values of Concentration and Time
136
-------
Application 3, illustrated in Figure 159:
To find time (t) when the concentration (C) will reach
2.6 mg/1 at location (x) of 4,200 feet, calculate:
x = 4,200 ft = 60 (Locate at D)
XD 70 ft
QC. _ (26,800 ft3/day) (31 mg/l)
Qr
1,800ft3/day
= 460 mg/l (Locate at B)
Using Figure 159, locate the selected concentration at A,
draw a line from A through B to C, then horizontally
from C to an intersection with a vertical line from D,
giving at E:
_L=50
Multiply by TD, to determine time (t):
t = (_L) (TD) = (50) (46.7 days) = 2,300 days
TD
If the lines intersect above the steady-state line, the con-
centration will not reach the given value at that location.
Nomograph for
Plume Center-Line
Concentration
-10
-102
v*-
10 -
102-
103-
10*-
105-
-10-2
-10-1
-1
QC0
QD
(mg/l)
-103
-104
-105
-106
-107
-108
-109
-101°
-10-2
-10-1
C
(mg/l)
" I
-10
-102
-103
L.10"
D 100
1,000
10,000
100,000
Figure 159. Application 3: Using the Nomograph to Estimate Time Given Values of Concentration and Distance
137
-------
References
29Pettyjohn, W.A. 1971. "Water Pollution by Oil-Field
Brines and Related Industrial Wastes in Ohio."
Ohio Jour. Sci., v. 71, no. 5, pp 257-269.
30Pettyjohn, W.A. 1973. "Hydrologic Aspects of
Contamination by High Chloride Wastes in
Ohio." Jour. Water, Air and Soil Poll., v. 2,
no. l,pp 35-48.
31Pettyjohn, W.A. 1975. "Chloride Contamination in
Alum Creek, Central Ohio." Ground Water, v. 13,
no. 4, pp 332-339.
32Exler, H.J. 1974. "Defining the Spread of Ground-
water Contamination Below a Waste Tip in
Groundwater Pollution in Europe." Water Infor-
mation Center, Port Washington, New York,
pp 215-241.
33Price, D. 1967. "Rate and Extent of Migration of a
"One-Shot" Contaminant in an Alluvial Aquifer in
Keizer, Oregon." U.S. Geological Survey Profes-
sional Paper 575-B, pp B217-B220.
34Perlmutter, N.M., M. Lieber, and H.L. Frauen-
thal.1963. "Movement of Waterborne Cadmium
and Hexavalent Chromium Wastes in South Farm-
ingdale, Nassau County, Long Island, New York."
U.S. Geological Survey Professional Paper 475-C,
pp C179-C184.
35Perlmutter, N.M. and, M. Lieber. 1970. "Dispersal
of Plating Wastes and Sewage Contaminants in
Ground Water and Surface Water, South
Farmingdale-Massapequa Area, Nassau County,
New York." U.S. Geological Survey Water-
Supply Paper 1879-G.
3&American Water Works Association. 1975. "Status of
Water-Borne Diseases in the U.S. and Canada."
Journal of American Water Works Association, v.
67, no. 2, pp 95-98.
37Essex Water Co. 1974. "Enduring Pollution of
Groundwater by Nitrophenols in Groundwater
Pollution in Europe." Water Information Center,
Port Washington, New York, pp 308-309.
38Hughes, J.L. 1975. "Evaluation of Ground-Water
Degradation Resulting from Waste Disposal to
Alluvium Near Barstow, California." U.S.
Geological Survey Professional Paper 878.
39Toft, H.P. 1974. "Pollution of Flood Plain Gravels
by Gas Works Waste in Groundwater Pollution in
Europe." Water Information Center, Port
Washington, New York, pp 303-307.
40Pettyjohn, W.A. 1972. "Good Coffee Water Needs
Body." Ground Water, v. 10, no. 5, pp 47-49.
41Rau, J.L. 1975. "Effects of Brining and Salt By-
Products Operations on the Surface and Ground-
Water Resources of the Muskingum Basin, Ohio."
Fourth Symposium on Salt, Northern Ohio
Geological Society, pp 369-368.
42Parker, G.G. 1955. "The Encroachment of Salt
Water into Fresh." Water, Yearbook of
Agriculture, pp 615-635.
43Mink, L.L., R.E. Williams, and A.T. Wallace. 1972.
"Effect of Early Day Mining Operations on Pre-
sent Day Water Quality." Ground Water, v. 10,
no. 1, pp 11-26.
^Ohio Division of Water. 1961. "Contamination of
Underground Water in the Bellevue Area." Ohio
Department of Natural Resources, Mimeo Report,
1961.
45Wood, B.C. 1962. "Pollution of Ground Water by
Gasworks Waste." Proc. Soc. Water Treatment
and Examination, v. 11, pp 32-33.
^Deutsch, M. 1961. "Incidents of Chromium Con-
tamination of Ground Water in Michigan." PHS
Technical Report W-61-5, pp 98-104.
47Deutsch, M. 1965. "Natural Controls Involved in
Shallow Aquifer Contamination." Ground Water,
v. 3, no. 3, pp 37-40.
48Deutsch, M. 1963. "Ground-Water Contamilation
and Legal Controls in Michigan." U.S. Geological
Survey Water-Supply Paper 1691.
49Carlston, C.W. 1964. "Tritium-Hydrologic Research,
Some Results of the U.S. Geological Survey
Research Program." Science, v. 143, no. 3608,
pp 804-806.
50Schuller, R.M., A.L. Dunn, and W.W. Beck. 1983.
"The Impact of Top-Sealing at the Windham
Connecticut Landfill." Proceedings 9th Annual
Research Symposium on Land Disposal of Hazar-
dous Waste, EPA-600/9-83-018, pp 334-342.
51Brown, M.H. 1980. Laying Waste, Pantheon Books,
New York, N.Y.
52Fenn, D.G., K.J. Hanley, and T.V. DeGeare. 1975.
"Use of the Water Balance Method for Predicting
Leachate Generation from Solid Waste Disposal
Sites." U.S. Environmental Protection Agency,
EPA/530/SW-168.
53Wu, J.S. and R.C. Ahlert. 1976. "State of the Art
Review—Non-Point Source Pollution." Water
Resources Engineering Technical Report 76-3,
Rutgers University.
54Pettyjohn, W.A. 1976. "Monitoring Cyclic Fluctua-
tions in Ground-Water Quality." Ground Water,
v. 14, no. 6, pp 472-480.
55Kansas Department of Health and Environment.
1981. "Summary of Seminar on Nitrates in Kan-
sas." Kansas Department of Health and
Environment.
56Kreitler, C.W. 1975 "Determining the Source of
Nitrate in Ground Water by Nitrogen Isotope
Studies." Texas Bureau of Economics Geological
Report of Investigation 83.
"Thomas, G.W., and R.E. Phillips. 1979. "Conse-
quences of Water Movement in Macropores."
Journal Environmental Quality, v. 8, no. 2,
pp 149-152.
58Elrick, D.E. and L.K. French. 1966. "Miscible
Displacement Patterns on Disturbed and Un-
disturbed Soil Cores." Soil Sci. Soc. Am. Proc.,
v. 30, pp 153-156.
138
-------
59Wild, A. 1972. "Nitrate Leaching Under Bare Fallow
at a Site in Northern Nigeria." Jour. Soil Set., v.
23, pp 315-324.
^Ritchie, J.T., D.E. Kissel and E. Burnett. 1972.
"Water Movement in Undisturbed Swelling Clay
Soil." Soil Sci. Soc. Am. Proc., v. 36, pp 874-879.
61Thomas, G.W., R.L. Blevins, R.E. Phillips, and
M.A. McMahon. 1973. "Effect of Killed Sod
Mulch on Nitrate Movement and Corn Yield,"
Jour. Agron., v. 65, pp736-739.
62Thomas, G.W., R.E. Phillips, and V.L. Quisenberry.
1978. "Characterization of Water Displacement in
Soils Using Simple Chromotographic Theory."
Jour. Soil Sci., v. 29, pp 32-37.
63Blake, G., E. Schlichting, and U. Zimmerman. 1973.
"Water Recharge in a Soil with Shrinkage
Cracks," Soil Sci. Soc. Am. Proc. N. 37,
pp 669-672.
^Ehlers, W. 1975. "Observations on Earthworm Chan-
nels and Infiltration on Tilled and Unfilled Loess
Soil." Soil Sci., v. 119, pp 242-249.
65Quisenberry, V.I. and R.E. Phillips. 1976. "Percola-
tion of Surface Applied Water in the Field." So/7
Sci. Soc. Am. Jour., v. 40, pp 484-489.
Quisenberry, V.I. and R.E. Phillips. 1978. "Displace-
ment of Soil Water by Simulated Rainfall." So/7
Soc. Am. Jour., v. 2, pp 675-679.
67Shuford, J.W., D.D. Fritton, and D.E. Baker. 1977.
"Nitrate-Nitrogen and Chloride Movement
Through Undisturbed Field Soil." Jour. Environ.
Qua!., v. 6, pp 736-739.
68Aley T. 1977. "A Model for Relating Land Use and
Ground-Water Quality in Southern Missouri'' in
Dilamarter, R.R. and S.C. Csallany, Hydraulic
Problems in Karst Regions. Western Kentucky
University, pp 232-332.
69Aubertin, G.M. 1971. "Nature and Extent of
Macropores in Forest Soils and Their Influence on
Subsurface Water Movement." USDA, For. Serv.
Res. Paper NE-192.
70Gerba, C. 1981. University of Arizona, Personal
Communication.
71Kent, D.C., W.A. Pettyjohn, T.A. Prickett, and
F.E. Witz. 1982. "Methods for the Prediction of
Leachate Plume Migration." Proceedings. 2nd
Nat. Symp. of Aquifer Restoration and Ground-
Water Monitoring, National Water Well Associa-
tion, pp 246-261.
72Wilson, J.L. and P.J. Miller. 1978. "Two-
Dimensional Plume in Uniform Ground-Water
Flow." Jour. Hydraulic Div. Assn. Soc. CivilEng.
Paper No. 13665, HY4, pp 503-514.
73Perlmutter, N.M. and M. Lieber. 1970. "Dispersal of
Plating Wastes and Sewage Contaminants in
Ground Water and Surface Water, South
Farmingdale-Massapequa Area, Nassau County,
New York." U.S. Geological Survey Water-Supply
Paper 1879-G.
139
-------
Chapter 6
Management Alternatives
Introduction
All of the available case studies of ground-water con-
tamination were reviewed to collect the various strategies
used by local public water supply managers. These case
studies showed that the Water Utility Manager is a very
small player in major contamination incidents. Often he
controls few of the resources available and the drinking
water solution is considered a low priority by the various
agencies involved. From the local manager's point of
view, the immediate problem is providing safe water to
the public in sufficient quantity and at a price the com-
munity can afford. Other agencies and interests may
have other objectives. What is the bottom line? For the
manager there are two main facts: (1) prevention is pro-
bably all a community can afford and (2) safe drinking
water presents a cluster of problems, issues, and objec-
tives which must be addressed by the whole community.
Hindsight and the Rubick's Cube
Although you have no current problem, you should
be doing a great deal about ground-water contamina-
tion. The easiest way to understand why is to see what
has happened to other communities and to ask a
number of "what if questions. Unfortunately, most
people have learned what they should have done to
prepare and prevent contamination by discovering con-
taminated wells.
EDB Found in Well Water At Southside Va.
School*
By Tom Sherwood
Washington Post Staff Writer
RICHMOND, Feb. 10—Gov. Charles S. Robb said
today that the suspected cancer-causing agent ethylene
dibromide (EDB) has been found in the well water of a
rural elementary school in Southside Virginia's Halifax
County.
The 215-pupil school, located in the farming com-
munity of Turbeville near the North Carolina border,
immediately suspended most uses of the contaminated
well pending more tests by state officials and the federal
Environmental Protection Agency, according to acting
School Superintendent S. Dail Yeatts.
The EDB announcement was the second action taken
in Virginia since the EPA announced an emergency ban
on agricultural uses of the pesticide chemical almost two
weeks ago. Last week, Virginia officials announced that
two prepared muffin mixes were being removed from
grocery shelves.
An EPA spokesman in Washington said his agency
had not yet received a report on the well water, but said
initial indications are that the Virginia school contamina-
tion "is relatively low." He added: "But we're con-
cerned about any level of EDB contamination."
Two state test samples showed a contamination level
of .027 parts per billion of water, Robb's office said.
The minimum detection level for EDB is .02 parts per
billion.
The school's well was the only one that showed EDB
contamination in tests of 87 public water supply wells in
17 counties and one city of Southside and southeastern
Virginia, Robb's office said.
Those heavily farmed areas were identified by the
Virginia Department of Agriculture as the most likely to
be contaminated by EDB.
School Superintendent Yeatts said in a telephone in-
terview that Halifax officials were informed of the con-
tamination late today. He said a freshwater tank was
sent to the school for drinking.
"We're taking portable tanks of water to wash dishes
in the cafeteria. We're going to use paper plates and
plastic utensils," Yeatts said. "We're doing everything
we can to protect the children."
Yeatts said officials were uncertain when the well was
established, but said it was the only source of drinking
water for the school.
He said the school, built in 1932, is located in an
isolated area about 20 miles from the town of Danville.
It is surrounded by farmland that produces tobacco,
grain, and canteloupes, a major crop in the community
that sponsors an annual canteloupe festival in the
spring.
Robb's office said long-term solutions to the well
problem included the possibility of drilling another well
or adding a carbon filtration system.
A state task force is continuing its test of grain prod-
ucts, officials said.
*Reprinted with permission of The Washington Post
141
-------
The case studies we have read and the ones we have
heard contain many lessons. The lessons can all be
boiled down to show that responding to contaminated
public wells is not a technical problem alone. It is a
technical problem, requiring professional knowledge of
hydrogeology, engineering, and other disciplines; but it
is also a financial problem, a legal problem, a public af-
fairs and public information problem, a health effects
and risk assessment problem, and in the final analysis,
fundamentally a community problem because only the
whole community can decide what the future of the
community should be. This is why the problem is like a
Rubick's cube; you must coordinate all sides of the cube.
When you discover contamination or a spill is
reported, the questions from your customers and the
press do not come in logical order. Everyone wants to
know: "Where is the contamination coming from?; Is
the water safe to drink?; Have we already been af-
fected?; What are you doing about it?"
A newspaper article, reporting EDB in the well of a
small school in rural southern Virginia, is reproduced
on the previous page. This article points out the prob-
lems which hit the manager of a water supply. The
acting superintendent is called by a major newspaper
the same day the governor's office announces the
results of the sampling. Concurrently there are ques-
tions about adequate protection for the children, an
alternate water supply, and the health effects. All these
areas must be addressed immediately by a local school
board and superintendent.
Where do you start? As superintendent Yeatts did,
you start with temporary measures to remove health
risks and, equally important, to control fear. Whether
the children were in any immediate danger or not, their
parents and the public needed to be reassured. Then the
local community can work out a longer temporary
solution, giving them time to decide the future of the
school and its water supply.
Perhaps this is an extreme example, but it does seem
typical for several reasons. The incident was unexpected
and apparently there was no emergency plan. In a case
like this, you usually wish that you had taken the time
to collect information about your wells and aquifer
before the contamination was discovered.
The previous chapters serve to illustrate how often
clean water can be found with contamination close by.
There are methods to protect an aquifer from an
approaching plume. You must take the time to find out
which hydrogeologists have studied your aquifer and to
collect and evaluate any results of samples taken from
private wells. You should put together a good map of
your aquifer. You also should calculate the influence of
all the cones of depression as the various wells are
turned off and on. This would tell you the real gradient
and allow you to calculate the travel time of a
hypothetical slug to your well. Using these methods,
you can collect and analyze a lot of free information
about your wells and aquifers and calculate various
possible contamination problems. You should also get
more than one opinion about the local hydrogeology.
Hydrogeologists do not always agree, and amazingly,
sometimes are wrong. In some cases, large amounts of
money have been spent only to discover that the slurry
wall did not work because the aquifer was deeper in
places than calculated. So get as many opinions as you
can.
Legal Issues
Before you discover that you have contaminated
wells and all kinds of questions need answers, you
should take the time to look into Federal laws, but
more importantly, into your own State law. Can the
customers sue? What is your obligation to inform the
public if the contaminant is not regulated? Who is
responsible for the contamination? Can we prove it? Do
we have to sue to get them to clean it up? Do we
qualify for Superfund help? You may already know
there are cases of utilities being sucessfully sued even
when no negligence was involved. What is the situaiton
in your State? We also have heard of cases where
Superfund came in and provided emergency help and
then pulled out because the immediate threat to public
health was over.
Financial Issues
You probably have had enough trouble getting the
water rates up to a level where you could pay for opera-
tion and maintenance. There is no operating reserve.
What happens when you discover contamination? What
can you afford to do? What does bottled water cost?
Can you reduce pumping? Can you hook up to a
neighboring utility? Can you get other well owners to
cooperate? What will it cost? Will your customers be
willing to pay? What about treatment? What do
granular activated carbon and airstripping cost? How
long will it take to find out what you need? Who can
tell you? You must put together some contingency plan
to deal with raising money and negotiate the options
with the community. Once you discover contamination
there's no time to arrange for loans or bonds, and
Federal funds seem a very distant possibility. What
happens if the health department orders you to take
steps you cannot afford? What happens to the
community if you go bankrupt?
Public Information and Health Effects
You now have a report from a lab confirming the
presence of a contaminant in the water from your wells
at a specific concentration. Most likely the contaminant
is a man-made chemical, a suspected carcinogen, and
present in trace amounts; e.g., parts per billion. What
do you tell your customers, reporters, the public?
It's very late, certainly too late to start educating your
customers about what "safe" means. One thing is cer-
tain, they are not going to be reassured by statements
concerning risks of ten to the minus sixth. EPA is trying
to come to grips with this problem. The December 1984
EPA Journal is devoted to the problems of risk assess-
142
-------
ment and risk management and the public's role in these
areas. EPA recognizes that in the end it is the local
community which must decide how to deal with risk
management. This is why it is impossible to separate the
issue of pubb'c information from the issue of risk
management.
When referring to contamination, your customers
most certainly will tell you, "take it all out." They want
no risk at all. Do they understand how much it will cost
and the consequences for the community of spending all
that money on drinking water? Will it discourage new
industry? Will employment suffer? Will people just
move away from high water rates? What is the benefit
of treating to zero? You can see that it is very late to
start asking the citizens such questions when there is
publicly reported contamination.
The public's first question when contamination is
reported is, "Is it safe to drink?" You have to give
them the truth as soon as you have it or any trust you
have built will be destroyed. What if you, and the state,
and EPA do not honestly know the health effects for
sure? Then you have to say so, although as Mr.
Ruckelshaus says in his interview in the December EPA
Journal, people will tend to think we really do know
and are not telling them. So what is the problem—
educating the public or determining the status of
research on health effects? Both are the problem. Once
you have contamination you have to start a crash course
for the public on the risk involved with the particular
chemical at the specific level detected. The public may
not be satisfied because they are unaware of how such
risks are evaluated, and therefore, will be unsatisfied
with your statistical talk and lack of clear affirmative
statements. Too bad you did not try to educate them
sooner. Now you will have to present them with a plan
involving costs, and perhaps allowing certain levels of
the contaminants to remain in their water. How clean is
clean?
The remainder of this chapter provides some
guidelines for developing a technical plan to prevent and
to respond to contamination.
Just because a water-quality problem is not known to
exist at the present time does not mean that one may
not appear tomorrow or at some other time in the
future. Contingency plans should be developed before a
problem develops so that sufficient time is available for
rational decisions and planning. Several approaches can
be followed, any of which should be dictated by the
particular political, economic, and technical situations
that exist. The contingency plans need not be expensive
nor should they necessarily follow traditional methods.
Some of the best ideas are generated by individuals with
little or no scientific training, but these same people are
characterized by a great deal of common sense and a
need to quickly and inexpensively solve a problem.
Although any management plan must be flexible, a
number of steps can be followed that should make the
plan easier to implement. Certainly not all inclusive, at
least the following steps could be taken: (1) determine
where the supply originates and what problems might be
associated with it, (2) learn the system, (3) locate poten-
tial sources of contamination, (4) develop a system of
self monitoring, (5) consider alternate sources of perma-
nent or temporary supply, (6) locate and evaluate ex-
isting laws and regulations on waste disposal, (7) develop
an aquifer sensitivity model, and (8) develop emergency
response plans.
Where Does the Supply Originate
A short distance from its border with Kansas,
Oklahoma's Cimarron River contains more than 50,000
mg/1 of dissolved solids during low flow conditions.
Scores of miles down stream, near its confluence with
the Arkansas River, the Cimarron still contains more
than 2,000 mg/1 of dissolved solids despite the dilution
from several major tributaries. In this case any wells
drilled in the flood plain that are dependent on induced
infiltration soon would be contaminated. The source for
the calcium sulfate and sodium chloride in the river is
natural, being derived from a series of saline springs and
seeps.
At Minot, North Dakota, two muncipal wells pro-
duce water that contains higher concentrations of
chloride than do the other wells in the field. The two
wells are also about 50 feet deeper than the average. In
this case a buried interglacial river valley trends through
the center of town; it had cut several tens of feet into
the underlying bedrock. One of the bedrock formations
that subscrops along the buried valley walls is the Can-
nonball Formation, a unit that contains salty water. Ap-
parently the high chloride wells are screened near the
subscrop of the Cannonball and when pumped induce
salty water to flow from the Cannonball, mix with the
fresher water in the glacial sand and gravel, and even-
tually reach the municipal wells. This problem also is
natural and perhaps the most practical control is to
blend the water with that from other wells.
Several years ago, in an industrialized city in
Michigan, a plant water manager decided to dredge the
adjacent river in order to increase the yield from their
induced infiltration supply. Not realizing the river con-
tained high concentrations of a variety of industrial
wastes, the river was dredged and within days the
chemical quality of the well water deteriorated
dramatically. It was then recognized that waste papermill
products had sealed the river bottom, providing a last
line of defense between the contaminated river and the
well field.
It is evident from the above that a knowledge of the
origin of the water-supply system can serve as a starting
point in the development of management alternatives.
Learn the System
For the most part, geologic and hydrologic evalua-
tions of the subsurface are based on an analysis of logs
of wells and test holes. Unfortunately, these data com-
monly are not readily available, although experience has
shown that they are likely to be stored in a file some
where. If they do not turn up, it might be possible to
obtain copies from the original driller, contractor, or
143
-------
consultant. Well construction details, such as depth,
length of screen, etc., also can be of considerable value.
The geologic data can be used to construct a number of
maps and cross section of the aquifer system, as shown
in Figure 160. Cross sections should provide an idea of
how much protection the aquifer and confining units
provide against contamination.
: Clay ;
Figure 160. Aquifer A Has No Natural Protection and is Highly
Subject to Contamination from the Surface. Aquifer B is Pro-
tected to Some Degree by the Overlying Layer of Clay
Notice in Figure 160 that the shallow or surficial
aquifer (A) consists entirely of permeable material that
extends from the land surface to the base of the water-
bearing unit. Consequently, this aquifer has practically
no natural protection other than the thickness of the
unsaturated zone. It could be easily contaminated by a
spill or nearly any type of waste disposal scheme. On
the other hand, the deeper aquifer (B) is covered by a
confining unit of low permeability, one that might re-
quire years for a contaminant to penetrate. In this case
the deeper aquifer has some degree of natural protection
and, in the case of a spill, time likely would be available
to develop plans to overcome a potential contamination
problem.
Well production/acceptance or aquifer test results
also can be very useful as aids in understanding how the
ground-water system functions. In both production and
aquifer tests, water levels are measured in a pumping
well and in one or more observation wells during a time
period that commonly exceeds eight hours. These data,
plotted as drawdown versus time or as drawdown in
several wells versus distance from the pumped well, can
be used to determine an aquifer's hydraulic conductivity
or transmissivity, as well as the storativity. These
parameters are necessary to calcuate ground-water
velocity, among other things (see Chapter 2).
Nearly as important as geologic information are
records of well discharge (rate and time interval) and
water-level fluctuations, the latter indicating how the
aquifer acts under stress. Of particular concern is the
size of the cone of depression around a pumping well.
As described in Chapter 2, the radius of the cone of
depression is controlled by the aquifer properties and
the discharge rate. In an unconfined aquifer, the radius
of the cone may be in the order of a few hundred feet,
but in a confined aquifer it may extend outward for
miles. Furthermore, the drawdown caused by overlap-
ping cones of depression is greater than that caused by a
single well. Additionally, horizontal and vertical varia-
tions in aquifer properties, pumping schedules and rates,
and well interference will tend to distort the shape of
the composite cone of depression in a well field.
The shape and areal extent of the cone of depression
are important in contamination studies for two major
reasons. First, the cone represents a change in the
hydraulic gradient, which steepens as it approaches the
well. This, in turn, increases the velocity of ground-
water flow. Second, contaminants that reach the aquifer
and are within a cone will migrate toward the pumping
well. Therefore, what ever happens within the radius of
influence of a well is of concern to the water manager
and this is why a knowledge of the size of the cone of
depression is so important.
Likewise, the size of the area of influence of a well
field is important because this is the area that should be
protected, as shown in Figure 161. Although the area of
influence might well exceed several square miles, the
velocity of the ground water near the outer margin
should be relatively low, as compared to the velocity in
the vicinity of a well. If the aquifer were contaminated
in this region, it might require several months or even
years for the contaminant to appear at a well. In the
meantime, the contaminant might be diluted or degrad-
ed to such an extent that it would not be of concern to
the plant operator.
Locating Potential Sources of Contamination
In order to develop a mangement plan, local sources
or potential sources of contamination must be known.
These include, in addition to the more obvious ones,
such things as the location of railroads, major highways,
gasoline stations, and small industrial or service plants,
particularly those small concerns that might be operated
in someone's garage, basement, or outbuilding. The lat-
ter are not likely to be well known to regulatory agen-
cies nor are they likely to have discharge permits.
In order to develop a data base for potential sources
of ground-water pollution, several waste surveys could
be conducted. These should include: (1) an industrial
waste survey, (2) a municipal waste survey, and (3) a
state and federal property survey. The surveys could be
as simple as examining a map to locate highways,
railroads, industrial sites, disposal sites, etc., or as com-
plex as interviewing a large number of people. Surveys
of this nature could become both time consuming and
expensive, particularly if done inhouse.
One approach, which could be both comprehensive
and inexpensive, would be to contact a number of local
service clubs, such as the Lion's Club, and request their
members to provide information. Other than the ob-
vious advantages of this method, one important con-
sideration is that the members of local service clubs
commonly represent a wide spectrum of the population
that, as a whole, might well have a detailed knowledge
of the area. The data base that could be developed
144
-------
Control point
Number is altitude of water level, in feet above mean
sea level
Piezometric contour
Shows altitude to which water will rise in wells
Dashed where approximately located Contour
interval, 5 feet Datum is mean sea level
0 1
Figure 161. Water-Level Contour Map of The Main Aquifer Showing a Large Cone of Depression and Steep Hydraulic Gradient
would potentially be comprehensive, inexpensive, and
flavored with community pride.
Potential Contaminants and Travel Times
Once the location of potential sites for contamination
are located, it will be possible to estimate the time re-
quired for a contaminant to migrate to a well. It must
be remembered, however, that what ever method is used
to predict travel time, it will be only an estimate. Three
methods that can be used are a simple equation, a
nomograph, and a computer model.
As discussed in Chapter 2, ground-water velocity is
controlled by the aquifers hydraulic conductivity (K), ef-
fective porosity(n), and hydraulic gradient(I), that is,
= KI/7.48n.
(61)
The hydraulic gradient is measured along a flow line
that originates at the contaminant source and continues
to the closest well in the downgradient direction. Of
course, the flow line must be drawn so that it intersects
the water-level contours at right angles. The easiest way
to determine the average gradient is to subtract the
water level at the well from the water level at the con-
tamination site and then divide this number by the
distance between the two points as measured along the
flow line.
Let us assume that a spill occurs at point A as shown
in Figure 161. The spill consists of leakage of 2,000
gallons a day for three days of salty waste water that
contains 30,000 mg/1 of chloride. The distance between
the source and the nearest downgradient well, as
measured along a flow line is about 1,585 feet. The
aquifer, which averages 50 feet in thickness, consists of
gravel and sand with an average hydraulic conductivity
of about 2,000 gpd/ft2 and an average effective porosity
of about .25. The difference in water level from the
source to the well is 15 feet (1,500 ft-l,485ft). Therefore,
the gradient is 15ft/l,585ft = .009 and the velocity is
about 9.6 feet per day.
V=KI/7.48n = (2,000 gpd/ft2)(.009)/(7.48 gal/ft3)(.25) =
9.6 ft/day
It would require about 165 days (l,585ft/9.6ft/day)
for the center of mass of the contaminant to reach the
145
-------
well under the prevailing conditions. The rate could be
slowed by reducing the gradient, that is, turning off the
well or reducing its discharge. It should also be pointed
out that the plume formed by the spill would arrive at
the well prior to 165 days; it is only the center of mass
of the plume, which is its highest concentration, that
would arrive at the time calculated by this method.
An actual situation very likely would be far more
complex and the travel time would be substantially less
than that calculated in the above example. Nonetheless,
this simple approach provides at least an estimate of
what might happen.
Another approach is to use some type of graphical
technique or computer model, many of which are
available. It must be remembered, however, that a com-
puter simulation can only be as valid as the equation
and data on which it is based and no matter how
sophisticated or expensive this simulation might be, the
results can only be used as an estimate. The greatest ad-
vantage of computers is that they can provide a solution
very quickly and permit one to readily change the input
values in order to get a feel for the way the aquifer
reacts to different stresses. Additionally, the nomograph
described in Chapter 5 and computer models permit one
to more accurately represent the aquifer and the manner
in which it functions, particularly in regard to
hydrodynamic dispersion and retardation of
contaminants.
The problem described above can be solved using a
nomograph or computer model, although additional
data are required. These include the retardation factor
(Rd) and longitudinal and transverse dispersivity (ax,
ay). Since the contaminant in this case is chloride, a
conservative chemical that is neither sorbed nor degrad-
ed, then Rd = 1. Dispersivity (ax, ay) is much more dif-
ficult to estimate. (Recall from Chapter 5 that
longitudinal dispersion, Dx, equals axV and transverse
dispersion, Dy, equals ayV). In this case, however, the
distance between the contaminant source and the well is
relatively short, the velocity of the ground water is high,
and the water is converging in all directions toward the
well bore. As a consequence, advection or simple
ground-water flow is more important than dispersion
and the latter either can be ignored or small numbers
can be used in the equation. In the few studies
available, the ratio between longitudinal and transverse
Steady State
Nomograph for
Plume Center-Line
Concentration
1,000
2,000
1475
5,000
10,000
20,000
50,000
10
100
1,000^
(-D
10,000
= 1585ft.
100,000
1-1010
-10-2
-10-1
r-1
-10
L-1Q2
1-104
Figure 162. Nomograph Solution to the Example Problem
146
-------
dispersivity ranges between 1 and 10. For this example
let us assume a ratio of 5 and a longitudinal dispersivity
of 1, which requires that transverse dispersivity is 0.2.
Table 14 shows the data required for the nomograph
solution and computer simulation.
Table 14. Summary of Data for Example
Distance from source to well (x)
Aquifer thickness (m)
Effective porosity (n)
Velocity (V)
Longitudinal dispersivity (ax)
Transverse dispersivity (cO
Retardation (RJ
Volume flow rate (Q)
Source concentration (CJ
*Mass flow rate (QCJ
1,585 feet
50 feet
.25
9.6 feet/day
1 foot/day
.2 feet/day
1
2,000 gallons/day
30,000 mg/1
502 Ib/day
'Mass flow rate = .134QCo/16,019 = (.134) (2,000) (30,000yi6019
= 502 Ib/day
Using the nomograph equations provided in Chapter
5, calculate the following:
XD=DX/V = (ax)(V)/V = (1)(9.6)/9.6 = 1
TD RdDx/V2 = (1) (9.6)/(9.6)(9.6) = 9.6/92 = .1
"
= nm
= 53 ft3/day
= (.25) (50)
To determine the time when the leading edge of the
plume, say a concentration of 1 mg/1, reaches the well
calculate:
X/XD = 1,585/1 = 1,585 (Locate at D)
QC0/QD = 502 lb/day/53 ft3/day = 9.5 Ib/ft3
(Locate at B)
or
QC0/QD = (267 ft3/day) (30,000 mg/1)/53
= 151,132 mg/1 (Locate at B)
Using Figure 162, locate the selected concentration (1
mg/1) at A, draw a line from A through B to C, then
horizontally from C to the intersection with a vertical
line from D, giving at E:
i/JD = 1,470
Multiply by TD to determine time(t), in days or
t = (1,470) (.1) = 147 days
Thus, under the given conditions, the margin of the
plume represented by a chloride concentration of 1
mg/1 should reach the production well in about 147
days or 18 days sooner than the center of mass
calculated earlier. It is essential to remember, however,
that this calculation also is only an estimate. On the
other hand, it does suggest that dispersion plays an im-
portant role in chemical transport and that the contami-
nant will migrate faster than anticipated even with very
small dispersivity values.
One of the limitations of the nomograph is that it can
be used only to calcuate concentration or time directly
down gradient from the source, that is, along a flow line
that goes directly through the source. Another limiting
factor is that the mass flow rate is constant, that is,
there is no method available to stop the source from
leaking, despite the fact that in the example case leakage
occurred only during a 3-day period.
Several advantages over the nomograph are provided
by a computer model, even though both are based on
the same equation. The computer model will generate a
concentration distribution map of the entire plume, not
just along a single flow line. Furthermore, in addition to
being fast, the model will allow the operator to insert
multiple sources and vary the mass flow rate from each
source.
Figures 163 and 164 show examples of computer
generated maps, based on the data listed in Table 14. In
Figure 161 the spill was allowed to discharge only 3
days. Notice that there is a distinct plume moving
toward the well and that the concentration shown on
the map in Figure 163 must be multiplied by 10. It can
be calculated that, at this time (147 days), the concen-
tration at the well is only 0.5 mg/1 because the plume
has not yet reached it. In Figure 164, the source was not
shut off but was allowed to discharge 2,000 gpd for the
entire time of the simulation (147 days), thus providing
a two-dimensional view of the concentration distribution
as compared to the single point nomograph solution. A
BASIC COEFFICIENTS
NO.
DESCRIPTION
VALUE
1.
2.
3.
4.
5.
6.
7.
SOURCE
AT
0,0
0,0
VELOCITY
LONGITUDINAL DISPERSIVITY
TRANSVERSE DISPERSIVITY
RETARDATION COEFFICIENT
HALF-LIFE
POROSITY
AQUIFER THICKNESS
INJECTION START
RATE TIME
502 0
-502 3
9.6 ft/day
1ft
.2ft
1
2,000 years
.25
50ft
LENGTH
100
100
PLUME AFTER 147 DAYS
SCALE. 10MG/L
5601 + + + + + + + + + + + + + + + + H
3201 + + + + + + + + + + + + + + + +
-80 1 + + + + + + + + + + + + + + + +
-1601 + + + + + + + + + + + + + + + +
-2401 + + + + + 4 + + + + + + + X + +
-3201 + + + + + + + + + + + + + + + +
-4001 + + + + + + + + + + + + + + + +
-4801 + + + + + + + + + + + + + + + +
-5601 + + + + + + + + + + + + + + + +
h + + + +
+ + t
+ + +
+ + +
19 18 2
+ + +
+ — + — + 1
081234456788911111111
06420864208601223456
0000000000042086420
00000000
SYMBOLS: X - ANSWER INACCURATE, # - ANSWER ABOVE 1000 MG/L
Figure 163. Computer Generated Map (Example 1)
147
-------
map of this type is very misleading because it implies
that high concentrations are continuing to appear be-
tween the source and the well despite the fact that the
leak occurred only during a 3-day period. Notice also
that the concentrations listed on the map in Figure 164
must be multiplied by a factor of 100.
BASIC COEFFICIENTS
NO.
DESCRIPTION
VALUE
1.
2.
3.
4.
5.
6.
7.
VELOCITY
LONGITUDINAL DISPERSIVITY
TRANSVERSE DISPERSIVITY
RETARDATION COEFFICIENT
HALF-LIFE
POROSITY
AQUIFER THICKNESS
9.6 ft/day
1ft
.2ft
1
2,000 years
.25
50ft
SOURCE
AT
0,0
INJECTION
RATE
502
START
TIME
LENGTH
100
PLUME AFTER 147 DAYS
SCALE 100MG/L
400 1 + +
240 1 + +
80 1 4 4
0 1 X 47
-80 1 4 4
- 160 I 44
-240 1 4 4
- 320 I 44
-480I 4 4
560 1 4 4
— H +
4444
4444
4444
33 27 24 21
+ 444
+ 444
+ 44 +
+ 44 +
+ 4 + 4
-t +t- + -_++
'
+
+
+
19
+
+
4
+
4
_-
4 +
+ 4
4 4
+ +
18 17
4 4
4 4
4 4
+ +
+ 4
H 4
+
+
+
16
+
+
+
+
+
— 4
+ +
4 4
+ +
15 14
+ +
+ +
+ +
+ +
+ +
L. -
+
+
+
14
4
4
+
+
+
- +
+
4
+
13
4
4
+
+
+
— 4
+ 4 +
+
+
+
13 1
+
+
+
+
+
+
2 12
4
4
4
+ + +
444
— 4 — 4
44 + 4
4
4
4
10
+
+
4
+
4 4
4 4
4 +
+ +
+ +
+ 4
4 4
+ +
+ 4 + 4
—4—4—4
081234456788911111111
06420864208601223456
0000000000042086420
00000000
SYMBOLS. X - ANSWER INACCURATE, # - ANSWER ABOVE 1000 MG/L
Figure 164. Computer Generated Map (Example 2)
Situation Monitoring
Situation monitoring should cover two main
categories: 1) monitoring of the existing water-supply
system and plant and (2) monitoring of other local situa-
tions. The former can and should be accomplished in-
house, while the latter can be carried out by interviews,
the news media, and local agencies.
It is surprising that so few operators, particularly
those involved with small systems, are aware of the
chemical quality of their well supplies. Even if routine
chemical analyses are carried out periodically, it is
unlikely that samples will be scanned for the more exotic
compounds, such as heavy metals or organics. This is
understandable in view of the cost. On the other hand,
without background data it is commonly difficult, if not
impossible, to detect many contaminants or locate a
source, especially if proof is required in a legal action.
The costs of chemical analyses must be accepted by the
operator as a part of the business practice.
Another part of the survey that should be conducted
by the water utility personnel includes an examination of
their facilities addressing such items as: (1) possibility of
back siphonage, (2) cross-connection, (3) distribution
system deficiencies, and (4) poor well construction or
location. Are there, for example, potential sources of
contamination, such as fuel tanks or sewer lines, adja-
cent to the well or well house? In the fall of 1971 at a
trailer court in Anchorage, Alaska more than 80 in-
dividuals became ill due to consumption of sewage con-
taminated well water obtained from a semi-public sup-
ply. The system consisted of 2 wells, about 242 feet
deep, enclosed in a block well house. A soft plug
formed in a borough sewer causing raw sewage to back-
up, eventually to flow from a drain in the floor of the
well house and, when reaching a foot in depth, to flow
directly into and down the well casing. Subsequently,
sewage contaminated water was pumped into the water
supply of the trailer park.
Another part of situation monitoring involves the col-
lection and evaluation of information in the communty
or area within the influence of the cone of depression of
the well field. For example, have there been any fires
that might have resulted in the release of chemicals that
could reach the aquifer? Have there been any spills from
truck or train wrecks? Might new construction produce
a hazard? Are plans being developed for the placement
of hazardous waste storage or disposal sites? In other
words, the purpose of surveys of this nature, which
must be continuous, is to keep in touch with the
community.
Alternate Sources of Supply
A common solution to a water-quantity problem is to
deepen a well and to a contamination event is to offset
and drill another well. Unfortunately, such potential
solutions, although simple, are rarely viable. It may not
be possible to deepen a well and merely offsetting a
contaminated well will only solve the problem for a few
hours or days. It appears that human nature is such that
we tend to procrastinate, hoping that life will continue
uninterrupted. The far thinking individual, however, will
consider alternatives, formulate cost estimates, and
develop plans, both for design and obtaining the
necessary funds for construction, for other sources of
supply.
Is there a source of surface water available nearby
that will meet water quality standards after treatment? If
so, is a site available and what are the potential costs of
constructing intake structures and conveyance facilities
and of treating the water, and how much time would it
require to actually provide the water? Is the supply
dependable, or contaminated, or can water rights be
obtained?
Sometimes it might be possible to use nontraditional
concepts to develop a surface supply. One method
might be to use collection gallaries, particularly if the
available streams are small. In this case a ditch could be
cut across the stream into which a gravel bed is placed.
A well screen attached to a suction line can be placed
148
-------
Well
Permeable
Sand and gravel
Figure 165. A Subsurface Dam of Clay Impeded the Flow of Ground Water in a Pond and Gravel-Filled Channel in the Vicinity of
Glenburn, North Dakota.
on the bed and the remainder of the ditch filled with
gravel. This is virtually a horizontal well whose supply
depends on infiltration of surface water through the
gravel pack. Although the filter pack might well reduce
turbidity, it would have little or no effect on many
chemical contaminants. Nonetheless, this technique of-
fers a simple and relatively inexpensive altenative.
Another method is the subsurface dam. The village of
Glenburn in north-central North Dakota had a difficult
time supplying sufficient water for their needs. They
overcame this deficiency with an unusual and inexpen-
sive artificial recharge technique. Most of the surficial
rocks in the Glenburn area consist of clay, but nearby
there is a 30 foot wide channel, usually dry, that con-
tains 7 to 8 feet of coarse gravel and sand. Upstream
the deposit widens and there is an abandoned gravel pit.
During the spring runoff a considerable amount of
water infiltrates the gravel and the water table rises
dramatically. Because the deposits are very permeable,
the ground water flows down gradient quickly, however,
and the water table soon declines as the aquifer is
drained. The gravel channel has a considerable capacity
for storage but no natural controls to prohibit rapid
drainage.
As Figure 165 illustrates, this problem was solved by
excavating a ditch, 4 or 5 wide, across the channel and
entirely through the gravel deposit. The excavation was
backfilled with readily available clay forming a subsur-
face dam. A perforated culvert, serving as a well, was
installed on the upstream side. A diversion ditch was ex-
cavated from the intermittent stream to the abandoned
gravel pit, which served as a recharge basin. During
periods of runoff, some of the surface water flows into
the gravel pit, where it infiltrates, and part of the re-
mainder infiltrates along the stream bottom. Thus, dur-
ing wet periods a considerable amount of water is col-
lected in the underground storage reservoir. The subsur-
face dam impedes the flow of the ground water down
the gravel-filled channel and the water table remains at a
high level, permitting increased water usage. It should
also be mentioned that this was a community project ac-
complished by volunteer labor and equipment. The total
cost was minimal.
Water supply problems in arid regions are particularly
vexing because of scanty rainfall and the high rate of
evaporation. In some situations, it may be possible to
augment supplies by constructing artificial aquifers. Ar-
tificial aquifers, by necessity, can store only moderate
quantities of water, but they are labor intensive and,
therefore, can be built at a modest equipment cost.
The design and construction of an artificial aquifer at
the Santa Clara Indian Reservation, New Mexico, was
described by Helweg and Smith.74 As Figure 166 shows,
a small gulley, several yards wide, was cleared of vegeta-
tion, deepened, and sloped. Spoil material was used to
construct an earthen dam across the gulley. A trench
was cut adjacent to and parallel with the dam into
which was installed a slotted plastic pipe. The slotted
pipe was connected, at a right angle, to a second pipe,
extending through the dam in the low point of the
gulley. The second or discharge pipe was laid on a slight
downslope and installed prior to dam construction.
Once the gulley was shaped, the pipes installed, and
the dam built, plastic lining was placed on the floor of
149
-------
Gravel mulch
Plastic liner
r Plastic sheet —_^-_^:
~ Collector pipe'.
Gravel pack ~CT. Discharge line
f)
,-„/
Figure 166. Schematic of an Artificial Aquifer
the structure. The gulley was then backfilled with
uniform sand (gravel could be used) and topped off
with gravel mulch.
During the rainy season or periods of surface runoff,
water flows down the gulley and infiltrates through the
gravel mulch to the artificial aquifer. (In some cases it
might be necessary to construct a spillway to avoid ex-
cessive erosion of the dam.) Water is removed from the
reservoir via the discharge pipe, the rate being controlled
by a valve.
Another management alternative is to consider
developing another aquifer or a different part of a con-
taminated aquifer. In the latter case, care would be re-
quired in the well field design to insure that the new
system would not be contaminated due to changing
hydraulic gradients.
Generally, new well fields require considerable time
and financing to achieve a proper design and adequate
construction. The first question to be addressed should
be, "Is there an aquifer available that will supply the re-
quired needs and what are its characteristics?" If one is
available, what is the quality of the water it contains
and what is an estimate of the treatment requirements
and costs? How many wells will be required and how
much will they cost?
Probably one of the most unusual and farsighted
ground-water systems was designed in Kalamazoo,
Michigan, which in the 1930's was suffering from a
severe water shortage. One individual who worked for
the city, Mr. Al Sabo, took a leave of absence to work
for a drilling firm in order to learn the trade. After
several months he returned to his previous position and
convinced city officials that a drilling rig should be pur-
chased. A crew then began to drill test holes throughout
a wide area, paying particular attention to swampy
regions. In this glaciated terrain, streams are commonly
small but they flow throughout the year, that is, the
flow is sustained by ground-water runoff. Furthermore,
swampy areas characterize places where the water table
is at or near the land surface and in the Kalamazoo
drilling program it was found that these swampy areas
were underlain by considerable thicknesses of saturated
sand and gravel. If the test hole encountered a substan-
tial thickness of sand and gravel, a well screen and cas-
ing were installed and an aquifer test conducted. If the
yield was satisfactory the pump was removed and a plate
welded to the top of the casing until the well was
needed. The swampy area was then purchased by the ci-
ty and the sand and gravel sold, which more than paid
for the property and wells. After the sand and gravel
were mined, the excavation served as an artificial
recharge basin. In a few years, the city of Kalamazoo
had more than enough water to supply all of their
needs, plus all the water required by several surrounding
communities.
Periodically various regions suffer from prolonged
droughts, streamflow decreases or may even cease, and
water rationing becomes the rule. It is interesting to
note that in many of these areas it is only the surface-
water supply that is decreasing, while billions of gallons
of ground water remain untapped in naturally occurring
underground reservoirs that remain hidden from view.
This, of course, was the case in Kalamazoo in the early
1930's.
Legal Controls on Waste Disposal
A variety of laws, regulations, and rules exist to con-
trol waste disposal. In addition to the often quoted
federal laws, there are laws established by state
legislatures and regulations formulated by state agencies.
Local zoning ordinances may play an important role for
the water-plant operator. These need to be researched,
understood, and modified as the need arises.
Development of an Aquifer Protection Plan
The basis of an aquifer protection plan is at least a
general knowledge of the aquifer system and the manner
in which it functions. Where are the inherent
weaknesses of the physical system, where are the strong
points, and where is a contamination problem most like-
150
-------
ly to occur? These questions can best be answered or at
least evaluated by means of a series of maps.
The maps shown in Figures 167 to 170 are modifica-
tions of illustrations published some years ago in a
report by the U.S. Geological Survey.75 Many similar
maps, reports, and books have been published by
federal, state, and local agencies. In cases where none
are available, the utility operator may be required to
prepare his own, hopefully with the aid of an ac-
complished hydrogeologist. The data base, of course,
consists largely of well logs and water-level
measurements.
The map in Figure 167 shows a large cone of depres-
sion around a well field and the arrows indicate the
general direction of ground-water flow. Notice that the
contours indicate that water is flowing into the pumping
center from all directions. Consequently, a large area
should be protected or at least monitored because con-
taminants reaching the aquifer in any part of the area of
pumping influence could eventually reach a well.
The map in Figure 168 plots potential sources of con-
tamination. Information of this type can be obtained
from the waste surveys discussed earlier. All fixed
sources are plotted, as well as the location of major
transportation routes, such as railroads and highways.
Figure 169 outlines the degree of natural protection
afforded the aquifer. The map, based on well logs, in-
dicates that the eastern part of the valley-fill aquifer is
protected by a considerable thickness of clay through
which contaminants are not likely to flow. To the west,
however, the valley fill consists of sand and gravel that
extends from the land surface to bedrock, a distance of
more than 100 feet. Any contaminants entering the
ground here could quickly reach the aquifer. In the cen-
tral part of the aquifer the major water-bearing zone has
some overlying protection in the form of alternating
layers of clay and sand. Although a contaminant even-
tually could reach the aquifer in this area, it would re-
quire a substantial amount of time and probably the
contaminant would be degraded or sorbed to some ex-
tent and certainly diluted as it migrated through several
tens of feet of clay and sand.
Figure 170 shows an aquifer sensivity map. Based
largely on well logs and the previous map, it rather
1486
1487 "22 1484
Control point
Number ts altitude of water level, in feet above mean
sea level
Piezometric contour
Shows altitude to which water will rise in wells
Dashed where approximately located Contour
interval, 5 feet Datum is mean sea level
0 1
Figure 167. Water-Level Contours indicate that the Cone of In-
fluence Surrounding the Well Field Extends Outward for Miles
in the Valley and that Water is Flowing into the Cone of
Depression from all Directions.
151
-------
Figure 168. Potential Sources of Contamination Include
Highways, Airports, Railroads, and Streams, as well as Com-
mercial Sites.
Explanation
Railroad
Highway
Buried Valley
Potential sources
of contamination
Potential line
source of contamination
Source of natural
contamination
EXPLANATION
Railroad
Highway
Figure 169. Areas of Recharge to the Aquifer. Area A, Direct Infiltration of Precipitation and Runoff. Area B, Infiltration by Way of
a Shallow Aquifer to the Main Aquifer. Area C, Area Underlain by Thick Layers of Clay Resulting in Only Small Quantities of
Recharge.
152
-------
Explanation
Railroad
Highway
Figure 170. An Aquifer Sensitivity Map Indicates the Range of Natural Protection of the System.
clearly indicates those areas of most and least concern.
The eastern part of the aquifer is moderately safe, cau-
tion should be exercised in the central part, and the
western part should be carefully protected and
monitored. The latter critical area should be brought to
the attention of city officials and an attempt made to
protect it by local zoning ordinances.
Emergency Response Plans
This section addresses what action a water utility may
be required to take as a result of contamination of the
ground-water source used to provide public drinking
water. It must be emphasized, however, that the preven-
tion of ground-water contamination, through proper
management of the resource, is the least expensive ap-
proach. The old adage of "an ounce of prevention is
worth a pound of cure" is certainly applicable to this
vital resource. During this seminar series, we have not
seen many examples of planning for prevention of
ground-water contamination or, for that matter, of a
planned course of action for loss of the primary water
supply.
When ground-water contamination is confirmed,
there are two immediate problems: (1) notifying the
consumers and (2) providing a safe drinking-water sup-
ply. For most utilities, the task of notifying the public
that their drinking water has been contaminated will be
the most difficult activity in responding to the con-
tamination. In transmitting its message to the public, the
utility must answer the technical questions of how it will
respond to the situation as well as the more personal
question of the consumer of, "What does this problem
mean to me?"
There is no sure-fire plan which can be followed that
will prevent an adverse response from the public to this
message. There are a few points that have been borne
out through the case histories we have heard in this
series. The most important of these points are: (1) tell
the public the truth as you know it, (2) tell them as
soon as the presence of contamination has been con-
firmed, and (3) continue to keep them abreast of the
situation and any remedial actions you are taking.
Only through these three steps do the people dealing
with the situation have any hope of maintaining
credibility with the public. Any attempt to hide or gloss
over the situation is an invitation to an adverse public
response and potential litigation.
The next problem—continuing to provide a safe
potable drinking-water supply—presents the utility
several alternatives based on the particulars of the situa-
tion. The possibility of a timely and logical response is
greatly enhanced by prior planning for dealing with loss
of the ground-water source. This response plan should
be developed in sufficient detail to be implemented with
only fine tuning to address the specifics of the particular
incident. Agreements between the various parties, which
outline the responsibilities of each party, should be laid
out in advance and not postponed until the emergency
is at hand. Response to the ground-water contamination
may range through a broad spectrum from issuing a
boil-water order to providing an alternative drinking-
153
-------
water supply. In considering a response, the entire pic-
ture must be addressed. For example, "Will a boil water
order be a good interim solution or will exposure to the
contaminants through inhalation or skin contact be as
great a health threat as ingestion?" The use of an alter-
native source may create problems within the distribu-
tion system as a result of the different chemical
characteristics of the new source. If water is provided
through the use of water trailers or bulk haulers, efforts
to maintain the potable quality of the water must be
taken.
Once the public notification and the continued provi-
sion of safe drinking water has taken place, the utility
may then address the long-term problem of remedial ac-
tion. The specific response that a utility takes will be im-
pacted by a number of factors among which are: (1) the
type and characteristics of contaminant; (2) the
hydrogeologic conditions; and (3) the political,
economic, and legal concerns. Therefore, how a utility
should respond cannot be detailed in a presentation
such as this. However, there is some commonality in the
steps which can be taken when considering alternatives.
These steps do not necessarily occur sequentially and, in
fact, may occur simultaneously. The steps may not be
entirely accomplished by the utility, but it is important
for the utility to understand the approach taken by its
consultants or the authorities addressing this problem.
Activities can be placed in 4 groups: (1) gathering and
evaluating available information, (2) developing the field
investigation to provide unavailable information, (3)
developing and selecting remedial alternative(s), and (4)
dealing with state and federal regulatory authorities and
the public.
The first two areas, gathering and evaluating available
information and developing the field investigation, are
closely linked. The full development of the field in-
vestigation should only occur after the available infor-
mation has been gathered and evaluated. This will allow
maximum use of the finite funds available for gathering
information.
In gathering and evaluating available information, we
are looking to estimate the extent, nature, and direction
of movement of the contamination. This estimate can
be based on readily available information from such
agencies as the U.S. Geological Survey and state
geological organizations. Topographic and hydrologic
maps will provide insight into geologic units and in-
fluence of landforms and surface waters on ground
water. The characteristics of the contaminant(s) and a
knowledge of industrial/commercial activities in the area
may help in locating the source of contamination.
Development of a map overlay depicting potential
source(s) such as gasoline stations, industrial sites,
highways, and railways will assist in this search. We fre-
quently limit our concern to industrial/commercial ac-
tivities; however, we should not overlook the impact of
domestic activities and disposal practices. Cases involv-
ing contamination of private wells with materials such as
chlordane have created widespread contamination of an
aquifer. State regulatory and health agencies can be a
source of information on both industrial and domestic
activities which may be of concern.
By evaluating the available information we can
develop the questions which will be answered by the
field investigation. Developing the field investigation in a
logical manner rather than undertaking a haphazard
program of installing monitoring wells provides for the
best use of the finite monies available.
During the information gathering and field investiga-
tion we are trying to answer the following questions: (1)
What is the source of contamination? (2) What is the
nature of the contamination? and (3) How extensive is
the contamination?
We are trying to develop a three-dimensional picture
of the area. By understanding the interplay between the
geologic framework, the topography, the hydrology, the
contaminant characteristics, the source location, and the
drinking water well field location a program for remedial
action can be developed. A word of caution about this
information cycle; there comes a point when you have
to act based on the information available. The cycle of
gathering of information, evaluating it, determining if
sufficient information is available upon which to base
remedial action, and continuing to develop information
if there is insufficient data for remedial action cannot be
repeated indefinitely. There have been case histories of
the cost of data gathering, in particular the field in-
vestigation, exceeding the cost of the remedial action.
The cost of gathering additional information and any
resultant delay in implementing remedial action must be
balanced with any benefit that might be gained by hav-
ing that information.
Based on the knowledge of the contaminated area
determined in the information gathering phase, a
remedial response can be developed. What is an ap-
propriate response? The remedial program will be
unique for each site because of the influence of the type
of contaminant, the geology, the economics, the politics
and the legal aspects of that specific site. The final selec-
tion will be based on development of alternatives,
analysis of those alternatives, and selection of the best
overall alternative for the situation. In the development
of alternatives, a number of alternatives, which upon
initial evaluation appear applicable, are selected. After
preliminary assessment of each of these alternatives, the
list of alternatives is narrowed.
The alternatives remaining after this screening should
be developed in more detail. The construction specifica-
tions should be stated and detailed cost estimates should
be made. Additional data gathering which includes
laboratory pilot scale studies may be required. Based on
this detailed information the best final selection can be
made. It is important to understand that the best overall
alternative may not be the "best" technical solution
because of the impact of political, economic, and legal
constraints.
Some of the alternatives which may result are:
(1) No response — the ground-water contamination is
not projected to be a threat. Additional monitoring of
the situation may be all that is required.
154
-------
(2) Inground Treatment/Containment
(3) Onsite Treatment
(4) Alternative Sources
The areas of inground treatment/containment and
onsite treatment will be developed later in the publica-
tion. However, a brief comment on the use of alter-
native water sources is in order here. Although an alter-
native water source is generally an immediate solution to
the contamination of a well field, it is frequently the
final alternative selected. The reason for this can be: (1)
economic constraints and (2) technological constraints.
The response to Federal, state, and local agencies may
be an immediate concern and will last throughout the
rest of the problem. It should be realized that in some
states as many as eight agencies have responsibility for
some aspects of ground water. When you add the
groups within the Federal EPA responsible for various
regulations such as the Resource Conservation and
Recovery Act, the Toxic Substances Control Act, and
the Safe Drinking Water Act, you can begin to under-
stand the number of people that may be interested in
your ground water contamination problem. You should
also realize that your objective of providing a safe,
potable drinking-water supply may be only one concern
among many in responding to the problem. Concerns
for cleaning up the site, pursuing enforcement action,
securing the spill site, cleaning up surface waters and soil
may be among the other objectives. In dealing with
these organizations, a number of items must be ac-
complished which include: (1) establishing who is
responsible for what, (2) developing a relationship of
trust among the players, and (3) for the water utility,
assuring that provision of a safe drinking water remains
one of the top objectives.
In conclusion, it should be pointed out again that the
prevention of ground-water contamination is a better
program than responding to the immediate and
sometimes gut-wrenching requirements of a con-
taminated well field.
References
74 Helweg, O.J. and G. Smith. 1978. "Appropriate
Technology for Artificial Aquifers". Ground Water,
Vol. 16, No. 3, pp 144-148.
75 Pettyjohn, W.A. 1967. "Geohydrology of the Souris
River Valley in the Vicinity of Minot, North Dakota".
U.S. Geological Survey Water-Supply Paper 1844.
155
-------
Chapter 7
Controlling Volatile Organic Compounds in Ground Water
Used for Drinking
Introduction
While numerous contaminants may be present in
ground water, e.g., inorganic, microbiological, organic,
radiological, this paper will discuss only organic con-
taminants and give an overview of the options available
to a water utility to lower the concentrations of these
contaminants to acceptably low levels for drinking. Fur-
ther, this paper will discuss only the volatile organic
compounds (VOCs) listed in Table 15. Some of these
VOCs have been found throughout the United States in
ground waters.76 In 1982, the first step of the regulatory
process was undertaken to limit exposure of these com-
pounds in drinking water.77 This paper relates to drink-
ing water sources which typically contain less than 1,000
micrograms per liter (/tg/1) of total VOCs as distin-
guished from hazardous waste disposal sites where VOC
concentrations may exceed several milligrams per
liter (mg/1).
Table 15. Volatile Organic Compounds Found in Ground Water
Used for Drinking76
trichloroethylene* cis-1,2-dichloroethylene*
tetrachloroethylene* trans-1,2-dichloroethylene*
carbon tetrachloridet benzenett
1 ,1 ,1-trichloroethanet chlorobenzeneft
vinyl chloride* dichlorobenzene(s)tt
1,1-dichloroethylene*
1,2-dichloroethanet
methylene chloridet
trichlorobenzene(s)tt
*alkene (unsaturated compound)
talkane (saturated compound)
ttaromatic
Mention of commercial products in this chapter does not constitute
acceptance or endorsement by U.S. EPA.
For the sake of brevity, acronyms are sometimes sub-
stituted for VOC names. For example, TCE is a recog-
nized name for trichloroethylene, but, unfor-
tunately,TCE has sometimes been used in newspaper ar-
ticles and some engineering reports to refer to
tetrachloroethylene or 1,1,1-trichloroethane. Similarly,
DCE is confusing because it has been used to refer to
any of the three isomers of dichloroethylene. This chapter
will use the entire VOC name to avoid any confusion.
When a water utility is confronted with VOC contam-
ination it should systematically approach a resolution of
the problem. This involves characterizing the problem,
analyzing the data describing the problem, and consider-
ing a site-specific solution to the problem.
Characterization of the Problem
When VOC contamination is discovered in a distribu-
tion system sample, the first step the water utility should
take is to determine which, and particularly how many,
contaminants were identified in the water sample. The
reason is that a single compound, such as trichloroethyl-
ene or tetrachloroethylene, found in a distribution sys-
tem sample can be a materials problem. For example,
vinyl-toluene-lined asbestos cement pipe has been used
in certain parts of the country; tetrachloroethylene
leached out of these liners and into the drinking
water.78 Trichloroethylene has been detected in joint
solvent used for reservoir liners.79 Certain organics from
contaminated soil can penetrate plastic pipe.80 On the
other hand, a distribution system water sample that con-
tains several VOCs is generally a good indication that
the water source is contaminated. The Ground Water
Supply Survey77 found that an affected ground water
typically contains several VOCs, with trichloroethylene
and tetrachloroethylene being the most common. Com-
parison of the VOCs found in a contaminated well and
those used or previously used in the community may
lead to clean up at the source. Identification of a mate-
rials problem may lead to material changes, rather than
to implementation of water treatment processes.
Table 16. Hypothetical Well Water Analysis
Total Organic Carbon (TOC) 0.75 pg/1
Trichloroethylene 500 ^g/1
Tetrachloroethylene 200 /ig/1
cis-l,2-Dich!oroethylene 100 ^g/1
1,2-Dichloroethane 10 /*g/l
Total Organic Halogen (TOX) 700 /tg/1 a
Iron 1.0 mg/1
When a problem is discovered in distribution system
samples, all the wells in the system should be sampled
and, if possible, a hydrologic investigation should be
undertaken to define movement of VOCs within the
aquifer and their relationship to each well. A contam-
157
-------
Cl
Cl
\
c = c
H
\
Cl
Trichloroethylene
MW = 131.5
Trichloroethylene = 500
Carbon: 2x12 x 500 = 91
131.5
Hydrogen: 1x1 x 500 =
131.5
Chlorine: 3 x 35.5 x 500 = 405
131.5
TOX = 405 pg/L
TOC = 91 ng/L = 0.091 mg/L
Figure 171. Example of Calculating Organic Carbon and Organic Halogen Contributions for Trichloroethylene
inated well should be sampled not only for VOCs, but
also for other parameters. Table 16 is a hypothetical
well water analysis showing a typical blend of VOCs
and other parameters pertinent to the design of treat-
ment processes.
Analysis of the Data
It is important to analyze the water-quality data
describing the problem. The total organic carbon (TOC)
concentration is a very good overall indication of
organic quality. Well water typically contains 1 milli-
gram per liter (mg/1) or less of background organic
carbon, but some highly colored ground waters have a
TOC concentration above 10 mg/1. Waters with high
TOC may decrease the capacity of granular activated
carbon (GAQ to adsorb VOCs. Note that the units for
TOC and iron are milligrams per liter, whereas the vola-
tile organic compounds are expressed in micrograms per
liter 0*g/l). Iron can precipitate on GAC and on the
packing material in aerators and foul their operation.
It is useful when interpreting such data to determine
how much of the TOC and total organic halogen (TOX)
can be accounted for with specific compounds. Figure
171 is an example of this calculation using trichloro-
ethylene. Table 17 is a summary of the accountable car-
bon and halogen. Note, only about 20 percent of the
background organic carbon can be accounted for in this
water sample. On the other hand, the accountable
organic chlorine is slightly less than the TOX concentra-
tion. This might signal further investigation for halogen-
ated compounds because the TOX concentration would
be expected to be approximately 25 percent less than the
sum of the chlorine concentration of the individual
compounds81 if the only organic halide contribution
were the VOCs.
When analyzing the data, high concentrations of
trihalomethanes (THMs) might be viewed with suspi-
cion. Chlorinated well waters are typically low in THMs
and THMs seldom occur in well water itself. It is not
uncommon for laboratories to misidentify VOCs as
THMs because of similar gas chromatographic retention
times.
Table 17. Summary of Accountable Organic Carbon and
Organic Halogen
voc
Trichloroethylene
Tetrachloroethylene
cis-l,2-Dichloroethylene
1 ,2-Dichloroethane
Concentration, jig/1
500
200
100
10
Calculated Total
Measured
TOC
mg/1
0.091
0.029
0.025
0.002
0.147
0.75
TOX
mA
405
171
73
7
656
700
Solutions to the Problem
Of the two solutions offered for contending with
VOC contaminated ground water, non-treatment tech-
niques will be discussed first.
Non-Treatment Alternatives
Take the well out of potable service. The word potable is
italicized because wells have been pumped to waste to
prevent contamination from spreading to other wells in
the aquifer. Shutting down a well may cause the con-
tamination to move to uncontaminated wells.
Locate and stop the source of the VOCs. This is, of course,
desirable, but pinpointing the source is generally dif-
ficult and often impractical. Complicating factors are
biological or abiotic transformations. A solvent such as
trichloroethylene may have been the original organic
chemical that was spilled, discarded, or disposed of, but
158
-------
because of certain biological conditions within the soil
or because of chemical hydrolysis reactions82, transfor-
mation products that ultimately reach the well may be
vinyl chloride or cis-l,2-dichloroethylene.83~85
Blend well water. All water purveyors are familiar with
this alternative to improve inorganic quality by moderat-
ing hardness or diluting the total dissolved solids from
one or more wells. VOC and non-VOC contaminated
wells have been successfully blended both with and
without hydraulic changes in the distribution systems.
Alternate Sources. Generally, the first course of action
taken by a utility when VOC contamination is dis-
covered is to drill a new well. A review of the pros and
cons of drilling a new well is given by Gaston.86 In
cases cited by Gaston, a new well could mean investing
several hundred thousand dollars with no guarantee that
the new well would remain uncontaminated after devel-
opment. Again, hydrogeologic investigation may reveal
whether drilling is a viable alternative.
Interconnection. A connection to a neighboring distribu-
tion system or source exists at many utilities. If not,
making such arrangements for emergency preparedness
is prudent. This alternative is similar to blending.
Bottled water. This alternative can provide some imme-
diate action while a permanent solution is being investi-
gated. However, this alternative suffers from the same
shortcomings as some point-of-use devices in that VOC
inhalation exposure from showering is neglected.
Treatment Alternatives
Conventional Water Treatment. In general, conventional
water treatment (i.e., coagulation, flocculation,
sedimentation, and filtration) is not effective for
removal of VOCs. Following spills, VOCs were found
to pass through surface water treatment plants with little
removal.87' 88 Any losses were likely the result of evap-
oration from open basins.
Oxidation and Ultraviolet Irradiation. The U.S. EPA's
Drinking Water Research Division (EPA-DWRD) found
no reduction in VOC concentration following treatment
with chlorine, chlorine dioxide, or hydrogen peroxide.89
Using permanganate ion or ferrate ion, no control of
the alkanes or the aromatics listed in Table 15 was
observed. The alkenes listed in Table 15, however, were
oxidized by these ions. EPA-DWRD found oxidation
decreased as halo-substitution of the alkene increased.
Tetrachloroethylene, for example, was unaffected by
permanganate, but trans-l,2-dichloroethylene and vinyl
chloride were rapidly oxidized by 2 mg/1 permanganate.
No attempts were made to identify end products which
may also be halo-substituted.89 The reaction between
alkenes and permanganate is well known.90
EPA-DWRD found ozone to be ineffective for oxida-
tion of alkanes, but effective for control of aromatics
and alkenes.89 The end products of these oxidation reac-
tions also remain unknown.
Ultraviolet irradiation was found by EPA-DWRD to
reduce VOC concentrations. Under similar conditions,
better removals of alkenes and aromatics than alkanes
were observed. When ultraviolet irradiation was com-
bined with hydrogen perioxide oxidation, little or no
improvement was observed when compared to ultravio-
let treatment alone.89 Several commercial processes are
available combining ultraviolet irradiation and ozona-
tion. The mechanisms and end products of these proc-
esses are not well defined.
In summary, certain VOCs are not influenced by ox-
idants, yet others are to a degree. VOCs are probably
not oxidized to carbon dioxide, halide, and water;
rather, end products result that, as yet, are uncharac-
terized. These processes are likely more costly than other
alternative, e.g., aeration or adsorption, and may not
be suited to a typical blend of VOCs which includes
alkanes.
Reverse Osmosis. EPA-DWRD found a spiral-wound
cellulose-acetate membrane to be ineffective for VOC
removal. A hollow fiber, polyamide membrane per-
formed somewhat better; it rejected about 25 percent of
certain VOCs.91 VOC rejection tests, however, on thin-
film composite membranes demonstrate better control.89
It was felt that if VOCs were discovered at sites where
reverse osmosis was already being used to control inor-
ganic compounds, the best alternative may be simply to
change to a thin-film composite membrane. Unfortu-
nately, EPA-DWRD discovered that a short-term
experimental protocol can be deceptive. Note in Table
18 that several VOCs diffused through a thin-film com-
posite membrane after a few hours. Carbon tetrachlo-
ride and 1,1,1-trichloroethane were well rejected. Subse-
quent studies of other thin-film composite membranes
showed similar patterns of time-dependent performance
and better control of some VOCs than others.89
In summary, this process may be useful for treating
certain VOCs. Prior testing should evaluate the mem-
brane over several weeks time. This process may be
more costly than other alternatives. Finally, considera-
tion must be given to proper handling and disposal of
the concentrated reject stream.
Table 18. Removal of Volatile Organic Compounds* by a Thin-
Film Composite Reverse Osmosis Membrane91
Average
Organic Influent
Compound cone., /ig/1
Benzene
Carbon tetrachloride
Chlorobenzene
1,2-Dibromomethane (EDB)
m-Dichlorobenzene
o-Dichlorobenzene
1,2-Dichloroethane
cis- 1 ,2-Dichloroethylene
Tetrachloroethylene (PCE)
1,1, 1-Trichloroethane
Trichloroethylene (TCE)
39
109
23
8
26
36
50
65
95
90
97
Percent Removal
1 hr.
52
98
98
88
95
93
49
47
99
98
89
3hrs.
20
98
%
34
91
88
34
6
98
98
66
6hrs.
18
98
85
39
88
85
29
24
97
98
35
9hrs.
18
98
72
15
85
83
29
12
74
98
25
21 hrs.
19
98
50
15
66
70
37
20
68
98
30
•Spiked in distilled water.
Aeration. One inquiry showed approximately three
dozen water utilities in the United States were using
159
-------
aeration to alleviate problems with VOCs in their
ground water.92 As noted, the treatment alternatives
presented above are more costly than certain types of
aeration and adsorption. A more extensive discussion on
these technologies is therefore warranted. Much of the
following aeration discussion is taken from reference 92.
Aeration can be broadly categorized in two ways: (1)
putting air through water and (2) putting water through
air. In the air-through-water systems, mass transfer
takes place at the bubble surface and the practical upper
limit for the air-to-water ratio is probably around 20 to
1 (volume-to-volume). The water-through-air systems
create water droplets or a thin layer of water to facili-
tate mass transfer. Although air-to-water (A/W) ratios
as high as 3000 to 1 (vol-to-vol) have been reported93,
the A/W ratio is generally less than 100 to 1.
Air-Through-Water Aeration Systems
Diffused-Air Aeration. As Figure 172a illustrates, in
diffused-air aeration compressed air is injected into
water through perforated pipes or porous plates. These
are sometimes called injection or bubble aerators and
are more commonly used for transferring oxygen into
waste water. Figure 173 shows the results of a one-year
pilot-scale study at a contaminated ground-water site
where the aerator removed approximately 95 percent of
the total volatile organic compounds. Figure 173 shows
the results of a one-year pilot-scale study at a contam-
inated groundwater site where the aerator removed
approximately 95 percent of the total volatile organic
compounds. Figure 173 is typical of any type of aera-
tion device in that VOCs are generally not removed to
"below detection" levels. One advantage of diffused-air
aeration is that a system can be put into operation very
quickly using existing facilities. Reservoirs, caisson
wells, and even well casings and well bores have served
as temporary aeration basins. Figure 172b illustrates the
latter scheme which is currently (1985) being evaluated
in Pennsylvania.95
Air-Lift Pump. As Figure 172c shows, this approach com-
bines air stripping with pumping. Early in the 20th cen-
tury, air-lift pumps were commonly used in the United
States. These pumps are simple devices with only two
pipes in the well. One pipe introduces compressed air in-
to the open bottom of the other pipe, called an eductor.
In the eductor, air mixes with water. The mixture, being
less dense than the surrounding water, rises. The air and
the VOCs are separated before the water is pumped into
the distribution system. Because of poor pumping effi-
ciency, typically 35 percent, air-lift pumps fell into
disuse with the introduction of submersible pumps. A
utility in Pennsylvania is evaluating air-lift pumping to
determine whether resurrecting this technology is less
costly (trading poor efficiency for reduced capital con-
struction) than other treatment alternatives.95 A
preliminary assessment showed 40 to 70 percent removal
of trichloroethylene with A/W ratios below 10 to 1.
This technology, therefore, may be sufficient for utilities
needing only modest removal efficiencies to achieve
their water-quality goals.
Mechanical Surface Aeration. Figure 172d illustrates this
alternative. Like diffused-air aeration, mechanical sur-
face aerators are commonly used in waste-water treat-
ment. Roberts and Levy96 see these devices as offering
many advantages for air stripping of VOCs from drink-
ing water. Several types are available. They can be
mounted on platforms or bridges or supported by col-
umns or pontoons. A motor-driven, impeller-like tur-
bine creates turbulent mixing of air and water. These
devices combine both features of putting air through
water and, the next category, water through air.
Water-Through-Air Aeration Systems
Packed Tower Aeration. Approximately two-thirds of the
aerators installed within the United States for VOC con-
trol are packed-tower aerators.92 Figure 174a illustrates
this device which consists of a column, typically 1 to 3m
in diameter and 5 to 10 m in height, that is filled with
packaging material. This packing can be glass, ceramic,
or plastic and is available in numerous geometrical
shapes to create a water film on a large surface area,
thus, enhancing transfer of VOCs to the air phase. The
inside wall of the aeration column has several redistri-
butors that force the water to flow over the packing and
thus prevent the water from simply running down the
walls. Figure 174a shows a fan at the top; this device is
called an induced-draft packed tower. Figure 174a also
shows the more common forced-draft packed tower
where a blower positioned at the bottom of the tower
forces air up through the packing. Several excellent
references97-102 are available that describe these devices,
discuss their theory and design, and present data that
demonstrate 98 percent or more removal of the VOCs
found in ground water. Reference 92 presents similar
performance data for this type of aerator.
Tray Aeration. In a tray aerator, water is pumped to the
top of the device, distributed over the cross-sectional
area by a perforated plate, and allowed to trickle down
and over trays which, in Figure 174b, are redwood slats.
Water splashing on the slats, which are staggered to pre-
vent the water from short circuiting the device, creates
the air-water interface for mass transfer. Air is drawn
into the tower by fans or blown in as described earlier
for packed towers. The air travels upward, counter-
current, to the water flow. These devices are commonly
used for controlling inorganic compounds (iron, manga-
nese, hydrogen sulfide, carbon dioxide) in drinking
water and units have been modified (additional blower
and slats) to treat VOCs.103 Cooling towers are also in-
cluded in this category. For cooling towers, however,
Figure 174c shows that the air movement might be
cross-current, rather than counter-current.
Spray and Venturi Draft Aeration. According to
Cheremisinoff104, these devices, illustrated in Figure
174d, evolved from spray ponds that were surrounded
by louvered walls. By spraying water through nozzles,
small droplets having a large interfacial surface area are
produced. In a natural-draft spray tower, air movement
is dependent on the atmospheric conditions and the
160
-------
VOCs
Air
VOCs
Air
Water
Water-
'—-O
Compressor
Water
Pump
a. Diffused-Air Aerator
Pump-
•Air
Compressor
• — Airline
(7— Diffuser
b. In-Ground Aeration
VOCs
Air
Air + VOCs /=, Drive Unit
Water
•-U-A
'ump |
Separator
Eductor-
^
~
*-
=H h
Compre
"Airline
•Air
Air
Water + VOCs
Impeller
Draft Tube
Water + VOCs
* •
c. Air Lift Pump
d. Mechanical Surface Aerator
Figure 172. Aeration Devices that Put Air through Water
161
-------
1000
- Influent
New Jersey groundwater
Aeration: 4:1 (vol. to vol.)
c
CO
6
0)
C
g
|
w
10
1
Aerator Effluent
Concentration, /^g/l
Time a b c d e
A 261 118 78 7 2.
B 20 9 2 1 0.2
A' 264 99 179 7 3.
B' 14 5 11 0.2
I
464
32
552
21
a - 1,1,1-trichloroethane
b - tetrachloroethylene
c - 1,1-dichloroetnylene
d - 1.1 plus 1,2-dichloroethane
e - trichloroethylene
_L
I
J_
1
_L
I
0 10 20 30 40 50 60
Time in Service, weeks Diffused-Air
Figure 173. Removal of Volatile Organic Compounds by Diffused-Air Aeration (Pilot-Scale Study)94
70
aspirating effect of the nozzles. Spray towers are gener-
ally placed side by side at right angles to the prevailing
winds. In a natural-draft device air is induced as a result
of water injected through spray nozzles into a venturi-
shaped plenum. The Calgon Corporation* has incorporated
one of these devices with an activated carbon adsorber
to combine aeration with carbon adsorption.
Catenary Grid Scrubber® f. This is a proprietary device, il-
lustrated in Figure 174e, that was developed as an air
scrubber105, but may have an application for removing
VOCs from water. A fluidized air-water contact is
created by balancing the air-flow velocity profile
through catenary-shaped grids. These grids, generally
placed in pairs, are stainless steel wire mesh with 0.6 to
2.5 cm openings. The manufacturer reports a very low
air pressure drop and speculates that this device would
be perhaps one-half the height of a packed tower.
•Pittsburg, Pennsylvania. Interphase®Ground-Water Treatment
System
tChem-Pro Corporation, 17 Daniel Road, P.O. Box 1248, Fairfleld,
New Jersey
162
-------
VOCs
Air
vP s$
3_ A;, Water 1 -El— _ _r-:=-
Pump f —
Redwood
Slats
— — • Air
— rV»- Water
Pump
Aerator
b. Redwood Slat Aerator
VOCs i
Air f
1- ' 00 1 1
^-, " " r '•• ^_ vc
V / / t / — i A
\ / / / / f "
1 / / / , f T
* J // //; / 1 f+~^
\ ' / //// / Ajr F=Z\
* \ ""^ / *~ /
* \ / *~ H^l
. Water 1 f - —
{-/ *" \V^~~r^/
r r T PumP W f
c. Cooling Tower
/ocs | .:;.i.:.;.
^•^,
c_* ^s
* 4 ( A'r + VOCs
Nozzle / ^a^liul
v "*~w J
J _.
1 -^ A, .»— \f\/,:l... ^^.. ,
J v j Water
DCs
r
\
Water
Catenary
Grids
Jyr Forced
j^ Air
Pump e. Catenary Grid
Spray Tower Scrubber®
Figure 174. Aeration Devices that Put Water through Air
163
-------
Operational Aspects (Secondary Effects) of Aeration
Air Emission. Aeration is not a technology that destroys
or alters VOCs, it simply transfers them to the ambient
air where they are dispersed, diluted and, perhaps,
photochemically degraded. This is, perhaps, the most
important secondary effect of aeration to consider.
An EPA-DWRD project in the Los Angeles basin is
identifying the off-gas dispersion pattern, and quantify-
ing, if possible, the VOC concentrations emitted from a
packed tower aerator.106 The state of California has
rigid constraints on air emissions from point sources
and a discharge permit was required before this aeration
research project could proceed from pilot-scale opera-
tion to full-scale design. Umphres, et a/.,107 documented
the effort that was required to receive the air discharge
permit and discussed the state's (January 1983) require-
ment for "best available technology" for air pollution
control.
The state of Michigan, requires "best available tech-
nology" be applied to VOC treatment devices discharg-
ing to the air. Further, "carcinogens" must not be
transferred to ambient air. Thus, the state requires gran-
ular activated carbon (GAC) treatment of the off-gas
stream from aerators used to remove VOCs from drink-
ing water or from aerators used to remove VOCs from
aquifers.
Designers must take care to ensure that state and
local ambient air-quality standards are not exceeded in
the vicinity of the aerator. The EPA-DWRD project in
the Los Angeles basin has, as an objective, a laboratory
examination of the gas-phase adsorption performance of
GAC for off-gas treatment as a function of relative
humidity.106 The addition of off-gas treatment technol-
ogies, i.e., GAC and humidity control, can significantly
increase the cost of aeration technologies.108
Participates. Aerators are, in effect, large air scrubbers
and the drinking water is in contact with 100,000 m3 or
more of air each day of operation. Ground-water treat-
ment rarely includes filtration. Thus, an increase in par-
ticulates may be possible in an area with high air particu-
lates. EPA-DWRD's contractors have not been able to
demonstrate any measureable water-quality deterioration
from particulates after aeration92.
Corrosion. Aeration will strip carbon dioxide from water
and introduce modest increases in pH. Additionally,
dissolved oxygen concentrations will increase slightly.92'
95, %,109 A survey of full-scale aerators revealed these
conditions have not produced distribution system
materials problem.92 EPA-DWRD's Los Angeles project
has as an objective, an examination of corrosion rates
in aerator effluent waters106. One area of corrosion con-
cern, however, lies with in-well aeration, shown in
Figure 172b and c, where, because of the pressure head,
dissolved oxygen concentrations can reach saturation95
which may be corrosive in some distribution systems.
Bacteriological Quality. Increases in standard plate count
and, occasionally, in total coliform densities, occur with
aeration.109 An aerator is a "break-in-the-system" and
disinfection should be incorporated following treatment.
Legionella species have not been isolated in aerated
water.109
Operational Problems. Cold weather does not have a
demonstrable effect on aerator performance but some
utilities have elected to house their aerator to protect
against ice damage should a pump fail during inclement
weather. A far greater potential problem is the precip-
itation of iron on the packing material within an
aerator. Fronk and Love109 described a significant loss
of VOC removal efficiency when iron precipitation
clogged an aerator. These devices, therefore, should have
an access port so the internals can be routinely inspected
and cleaned. Manganese and copper can also precipitate
following aeration.
Estimating Aeration Performance. Several investigators have
reported the efficiency of aerating VOCs and some have
attempted to relate molecular properties to performance.
McCarty and coworkers110 concluded Henry's Law
Constant (H), or Henry's coefficient, is the most useful
indicator of relative ease for air stripping VOC. H can
be defined as the ratio of the concentrations of a
volatile compound in air and in water at equilibrium. H
is expressed either as atmospheres (atm), atm-mVmole,
or dimensionless. By expressing H as dimensionless, the
reciprocal, 1/H (called the partition coefficient) is the
theoretical optimum air-to-water ratio for removing a
volatile compound by air stripping. To express H as
dimensionless:
multiply H expressed as atm by 0.00056 (5.6x Kr4)
multiply H expressed as atm-mVmole by 42
Knowing H, organic compounds can be ranked as is
shown in Figure 175. Figure 175 reveals to important
points about aeration: (1) all VOCs are strippable—some
are more so than others and (2) the use of aeration to
remove one VOC will remove all other VOCs to some
extent. Figure 175 does not show ease of stripping to be
a function of VOC concentration; for VOC concentra-
tions typical to ground waters, aeration performance is
independent of concentration.
Aeration needs can be simplistically estimated knowing
H. Trichloroethylene, for example, has a Henry's Law
Constant of 0.48 at 20°C.in The partition coefficient,
therefore, is 1/0.48 or 2.1. Thus, at equilibrium, 2.1
volumes of air are theoretically necessary for each volume
of water to strip out trichloroethylene. In practice, the
theoretical optimum air-to-water ratio is not sufficient for
99 percent removal. For the hypothetical blend of VOCs
described in Table 16, the Henry's coefficients and strip-
ping factors are given in Table 19. From experience,
EPA-DWRD has found that the actual air-to-water ratio
averages approximately 20 times the theoretical value for
diffused-air aeration.94 This, then, gives an estimate of
aeration requirements. Table 20 illustrates the result of a
selected air-to-water ratio on the control by diffused-air
aeration of the Table 16 VOCs. At an air-to-water ratio
of 20:1, tetrachloroethylene effluent levels will be approx-
imately 2 fig/I; the others having lower Henry's coeffi-
cients, will not have reached 99 percent removal. At a
ratio of 400:1, 1,2-dichloroethane removal will be 99 per-
164
-------
Ease of stripping _
0.1 1.0
1 1
&••••
L
_.!...„
_ Concentration in air, /4J/I
Concentration in water. »n/l
10 100
..E: .......
J «-T-« < 4 4* *
, ' > I I *- Carbon Tetrachloride
I Tetrachloroethylene
* r~o
1000
I
t
Vinyl Chloride
1,1,1 -Trichloroethane
Trichloroethylene
Chlorobenzene
Benzene
Cis-1, 2-Dichloroethylene
1 , 1 -Dichloroethylene
Trans-1 ,2-Dichloroethylene
1 ,4-Dichlorobenzene
" 1,3-Dichlorobenzene
_ Methylene Chloride
T 1,2-Dichloroethane
1 ,2-Dichlorobenzene
1,2.4-Trichlorobenzene
Figure 175. Comparison of Henry's Law Constants for Selected Organics
110,111
cent; the others, having higher Henry's coefficients will
have removals better than 99 percent. For packed-tower
aeration, the actual-to-theoretical ratio is appreciably bet-
ter, i.e., approximately 3. Recall that this exercise is a
simple estimation of aeration needs. The importance of
proper design based on pilot testing should not be under-
estimated. References 97 to 102 should be consulted.
Table 19. Estimating Aeration Needs for Diffused-Air Aeration
VOC
Trichloroethylene
Tetrachloroethylene
cis-1 ,2-Dichloroethylene
1,2-Dichloroethane
H
0.5
1
0.3
0.05
99% Removal
Air: Water Ratio
Theoretical Actual
2 40
1 20
3 60
20 400
Table 20. Water Quality Assuming 99 Percent Removal
Efficiency by Diffused-Air Aeration fcg/L)
AirWater = 20:1 Air:Water = 400:1
Trichloroethylene
Tetrachloroethylene
cis-1,2-Dichloroethylene
1,2-Dichloroethane
>5
2
<5
<2
0.1
Adsorption
Synthetic Resins. In 1976, the Rohm and Haas Company
introduced an experimental synthetic resin called Amber-
sorb XE-340. This material was designed specifically for
removing low molecular weight halogenated compounds
because, at that time, chloroform and other
trihalomethanes were the topic of intense research in the
United States. Although this material looked very
promising because of its high adsorption capacity, in
1982 the manufacturer announced that XE-340 would
not be commercially produced. The future of resins for
controlling VOCs is therefore questionable.
Powdered Activated Carbon. Although powdered activated
carbon (PAC) is rarely employed in groundwater treat-
ment where clarification is not required, it has been
monitored in surface-water plants treating for VOCs. At
the doses at which PAC can be applied, typically for
taste and odor control, it has not been found to be cost-
effective.94
Granular Activated Carbon. Figure 176, a compilation of
data from several laboratory studies with activated car-
bon, illustrates two important points: (1) the organic
compounds found in ground water cover a wide range
of adsorption capacities and (2) treatment by adsorption
to remove one VOC will removal all VOCs to some ex-
tent. In contrast to aeration, where performance was in-
dependent of concentration, the adsorption capacity
varies directly with the equilibrium concentration of the
organic contaminant.
165
-------
Mean Adsorption Capacity, mg/gm @ Equilibrium Concentration = 500
1.0 10 100
Methytene Chloride
Trans-1,2-Dichloroathylene •
1,2-Dichloroeth«ne
0
Benzene
Carbon Tetrachloride
Tt
Trichloroethylene
I
Chlorobenzene
Cia-1,2-Dlchloroethylene
u 1.1,1-Trichloroethane
_ 1.1-Oichloroethylene
-if \
1,2,4-Trichlorbenzene
1,4-Oichlorobenzene
T 1.2-Oichlorobenzene
1.3-D ichlorobenzene
L Tetrachloroethylene
Figure 176. Comparison of Isotherm Adsorption Capacities on Activated Carbon94
1000
100 _
to
o
'c
«
E>
O
£
to
o
B
o
I
3
w
0
3
1
(
— J l 1 1 i ' ' t '
- A /*? M ir^
/ o
— Influent
New Jersey groundwater f -
Adsorption on Witearb* 950 4 P"*-"
_ Empty Bed Contact Time V-
~ =18 min. T
1
J
I '
1
1
- 1
1
1
- #
Activated Carbon Effluent '
B .'
* i
1 ! 1 ! I -4 1 1
I 5 1 1 i -
A A" -
«ys B"
Concentration, M9/I
a b c d « I
A 261 118 76 7 2 464
B <0.1 <0.1 <0.1 <0.1 <0.1 <0.1
A* 264 99. 179 7 3 552
B' 95 0.1 <0.1 2 <0.1 97
A" 259 100. 190 21 7 572
B" 281 0.7 16 9 <0.1 3O6
a • 1,1.1-trichloroethane
b - tetrachloroethyiene
c • 1,1-dichloroethyiene
d - 1.1 plu* 1,2-dichloroethane
e - trichloroethyiene
I ! I I I
) 10 20 30 40 50 60 70
Time in Service, weeks
Figure 177. Removal of Volatile Organic Compounds by Adsorption on Granular Activated Carbon (Pilot-Scale Study)94
166
-------
Figure 177 summarizes data collected from an EPA-
DWRD pilot-scale adsorption study. For approximately
30 weeks (17,000 bed volumes), the contaminants were
removed to "below detection" (0.1 micrograms per
liter, /tg/1) by the GAC. This is typical of adsorption
systems and is unlike aeration systems, as Figure 173
shows. When breakthrough occurred, the principal
organic compounds were the chloroethanes, which is
consistent with the order of adsorption capacities given in
Figure 176. At point B" the 1,1,1-trichloroethane concen-
tration in the effluent slightly exceeded the influent in-
dicating the carbon was exhausted for control of the
compound.
Adsorption can be affected by the amount of back-
ground organic carbon, generally measured as TOC.
High background organic content (i.e., over 1.5 to 2
milligrams per liter) can affect adsorption capacities.
Competitive adsorption and the influence of high con-
centrations of background organic carbon are recog-
nized, but not well understood. Table 21 demonstrates
how, in one study, increasing TOC decreased the
adsorption capacity of trichloroethylene112. This again
demonstrates the importance of defining TOC.
Table 21. Effects of Competitive Adsorption on
Trichloroethylene112
Freundlich Isotherm Coefficients
1/n
TOC
mg/l
K
(Capacity at 1.0 mg/l)
mg/gram
1.97
8.39
0.48
0.44
0.44
56.0
31.6
24.4
Operational Issues of Adsorption
Spent Adsorbent. At some point following breakthrough,
effluent concentrations of a contaminant will exceed
effluent quality levels and the carbon must be replaced
or reactivated. Typically ground-water utilities have not
employed onsite reactivation, but have employed an
adsorption service. Two manufacturers of activated car-
bon (Calgon Corporation of Pittsburgh, Pennsylvania,
and ICI Americas of Wilmington, Delaware) offer
adsorption services and approximately a dozen water
utilities have employed this treatment option. In general,
the service consists of characterizing the water, sizing,
constructing and installing pressure adsorption vessels,
and replacing the spent GAC with virgin GAC material
when the concentration of a contaminant exceeds a pre-
determined effluent quality. The adsorbent is then
returned to the manufacturer, reactivated, and resold
for industrial use. These adsorbers typically contain up
to 9,000 kg (20,000 Ib) of activated carbon. Utilities not
employing an adsorption service must plan for disposal
of the spent GAC. One firm (Carbon Services of Min-
neapolis, Minnesota) offers gravity adsorbers that hold
approximately 2,500 kg (6,000 Ib) of activated carbon.
These adsorbers are off-the-shelf refuse dumpsters, that
have been lined for corrosion protection and can be
conveniently hauled onsite, leveled, and put into service.
When effluent quality concentration has been exceeded,
the dumpsters can be lifted onto a truck and
the GAC transported for disposal, reactivation, or
replacement.
Pressure adsorbers are normally operated in series
with empty bed contact times typically ranging from 8
to 18 minutes. The time of operation varies because at
this writing (1985), no national maximum contaminant
levels (MCLs) exist and the action levels of the various
states are not uniform. Pressure adsorbers are typically
designed to operate 8 or more months. Operating in
series allow each adsorption vessel to remain on-line
until the activated carbon is fully exhausted before it is
replaced. This two-stage, series operation and the steps
involved are shown in Figure 178. In contrast to adsor-
bers operating in parallel or alone, the series operation
uses a significantly larger portion of the adsorbent.
Desorption. When an adsorbant is exhausted, a decrease
in the influent concentration will upset the equilibrium
between the adsorbed contaminant and the contaminant
in solution. Under these conditions, or whenever there is
a significant decrease in the influent concentration,
desorption can occur. An example of this concentration
is seen at point B" in Figure 177 for 1,1,1-trichloro-
ethane. In addition to equilibrium, desorption can occur
as a result of displacement by a more competitive com-
pound. It is important to sample not only the GAC ef-
fluent to monitor VOC concentrations reaching the con-
sumer, but also the GAC influent to better understand
GAC contactor behavior.
Bacteriological Quality. Increases in standard plate count
and total coliform densities are well known across GAC
contactors treating surface waters, especially in summer
months when water temperatures rise. GAC will elim-
inate the free chlorine residual and adsorb nutrients on-
to its surface. While bacterial densities in ground waters
are generally several logs lower than surface waters and
while ground-water temperatures are relatively constant,
bacterial densities can increase across GAC contactors
treating ground water. Thus, a GAC contactor should
be followed by a disinfection process prior to water
distribution.
Estimating Adsorption Performance. Certain molecular prop-
erties may be useful for preliminary estimates of ad-
sorption capacity when confronted with a newly iden-
tified contaminant. Love and Miltner113 correlated
adsorption capacity with molecular weight, density, vapor
pressure, solubility, Henry's Law Constant, parachor,
and molar refraction. The parachor relates molar volume
(ratio of molecular weight to density) to surface tension.
The parachor of a compound can be estimated from the
sum of the atomic and structural parachor equivalents
given in Table 22.
Table 22. Atomic and Structural Parachor Equivalents114
Carbon 4.8 Chlorine 53.8 Carbon double bond 23.2
Hydrogen 17.1 Bromine 68.0 Carbon triple bond 46.6
Oxygen 20.0 Iodine 91.0 Aromatic ring 6.1
167
-------
Water from well «.,
Contaminant Wave Front
rC^.To Conaumer
. \ Vaaaals A and B contain
virgin granular activated
carbon and ara operated in
I
I
I
I
I
I
! Adaorfaant in Vaaaal A
becomes exhausted;
contaminant wawa front
movas through B.
Raptaea or
Regenerate
Activated
Carbon
k Vessel B removed from
service and adaorbam
raplacad. Cycle than
repeated.
Replace or
Regenerate
Activated
Carbon
//Exhausted
^Activated
% Carbon
Vesaal A removed from
I service and adsorbent
replaced.
I Adsorbent in Vessel B
becomes exhausted;
contaminant wave front
. moves through A.
Figure 178. Two-Stage Series Operation to Maximize Use of Activated Carbon
The molar refraction of a compound describes the
relationship between molar volume and the compound's
refractive index. Like the parachor, molar refraction can
be approximated by adding the appropriate atomic and
structural refraction equivalents shown in Table 23.
Table 23. Atomic and Structural Refraction Equivalents114
Carbon 2.42 ml
Hydrogen 1.10
Chlorine 5.97
Bromine 8.87
Iodine 13.90
Carbon double bond 1 .73 ml
Carbon triple bond 2.40
Carbonyl oxygen 2.21
Hydroxyl oxygen 1.53
Ether oxygen 1.64
Molecular weight, density, and Henry's Law Constant
produced poor correlations with adsorption. In Figure
179, three arbitrary ranges of adsorption capacity
(poor= < 10 mg/g; fair= 10-50 mg/g; good= >50
mg/g) are shown with corresponding fitted curves
(y = bx°) for solubility, vapor pressure, molar refraction,
and parachor. Assume, for example, that a newly iden-
tified organic compound is known to have a low solubil-
ity and vapor pressure, and the parachor and molar
refraction are calculated to be relatively high. That
organic compound is likely to be well adsorbed on gran-
ular activated carbon. This approach is not intended to
replace gathering the actual adsorption data, but might
be useful in obtaining preliminary information for
adsorption capacity where empirical data are lacking.113
168
-------
Molar Refraction (R3 = 0.85)
Vapor Pressure (R3 = 0.81)
100
H-
1500
-t-
Solubility (mg/l)
80
H-
Vapor Pressure (mmHg)
200
Parachor 1—
25
Molar Refraction « \
260
33
Figure 179. Use of Molecular Properties to Estimate Adsorption of Volatile Organic Compounds on Activated Carbon113
Adsorption performance can be roughly estimated
starting with the Freundlich equation:
Table 24. Adsorption Capacity Based on Isotherm Data
X_=
M
(62)
where:
X/M = adsorption capacity, mg VOC adsorbed/gram
activated carbon
Ce = equilibrium concentration of the VOC, mg/l
K and 1/n are empirical constants characteristics of
compound, the water, and the activated carbon.
The K and 1/n values are published for certain com-
pounds112' 113> 115 or can be obtained through labora-
tory adsorption isotherm studies. Some of these data
give the K and 1/n values for trichloroethylene as 36
mg/g and 0.5, respectively, in distilled water. Therefore,
at an equilibrium concentration (Ce) of 500 jtg/1 (0.5
mg/l) the adsorption capacity would equal:
x/M = KCe1/n = 36(0.5)°-5 = 26 mg/g
Similarly, the adsorption capacities for the VOCs in
Table 16 are shown in Table 24.
VOC
Trichloroethylene (TCE)
Tetrachloroethylene
cis-1 ,2-Dichloroethylene
1,2-Dichloroethane
K, mg/g
36
134
8
5
1/n
0.5
0.5
0.5
0.6
X/M, mg/g
26
60
3
0.4
Assuming a well with this blend of VOCs is pumped
continuously at 200 gal/min, and that 20,000 pounds of
granular activated carbon are to be used, the estimated
time to exhaustion may be determined:
Estimated TCE loading per day:
200 gal x 3.78 liter x 1,440 min x 0.5 mg TCE = 544,300 mg
min gal day liter day
Estimated Adsorption capacity:
20,000 Ib GAC x 454 £ x 26 mg TCE = 2.4 x 108 mg TCE
Ib g GAC
169
-------
Estimated Time to exhaustion:
2.4x108mgTCE = 430 days
544,300 mg TCE/day
From experience, EPA-DWRD has found that break-
through occurs at approximately half the time to
exhaustion at typical loading rates and GAC contactor
configurations. An example of this can be seen in
Figure 177. Breakthrough for trichloroethylene, there-
fore, can be estimated to be 215 days.
From Table 25, calculated in a similar manner, the
estimated life of this adsorption system would be about
four months, if reactivation or replacement were re-
quired shortly after breakthrough of cis-l,2-dichloro-
ethylene or 1,2-dichloroethane. Note that the contam-
inant in highest concentration does not necessarily con-
trol. A similar approach might be used if the situation
was: "I want to be able to run one year. How much
adsorbent would I need?" Recall that adsorption
isotherm capacities are generally reported on individual
compounds and often in distilled water, so the back-
ground total organic carbon concentration in the ground
water must be low (< 1 mg/1) to have much confidence
in such estimates. As the TOC increases, so does the
potential for competition for adsorption sites; therefore,
the adsorption capacity for specific compounds might
decline.
It must be stressed that this example of estimating ad-
sorbtion performance is, like that for aeration, a simpli-
fication and must be followed up with onsite, pilot-scale
investigation.
Table 25. Estimated Time for Breakthrough on Granular
Activated Carbon
voc
Trichloroethylene
Tetrachloroethylene
cis-1 ,2-Dichloroethylene
1,2-Dichloroethane
Days
215
1,430
130
110
Combination of Treatment Processes. Two or three water
utilities in the United States have preceeded adsorption
with aeration and have found this combination to be ex-
tremely effective for producing a water with volatile
organic concentrations below a detection limit of 0.1
j*g/l. The aeration step reduces the organic load to the
adsorbent and may remove compounds competing for
adsorption sites. The Calgon Corporation has a process
called Interphase® that incorporates a venturi-draft
aerator with a gravity-flow adsorber.
Relative Costs of Treatment. The extent of treatment pro-
vided will be governed by the desired effluent quality and
the associated costs. Cost estimates vary significantly
because each process has a spectrum of effectiveness.
From the available estimates116' 117 some generalizations
can be made on relative costs. Packed-tower aeration is
considerably less expensive than diffused-air aeration or
adsorption on activated carbon. The cost of packed-
tower aeration (capital plus operation and maintenance)
ranges from 0.03 to 0.1 dollars per cubic meter (4 to
13C/1,000 gallons in 1982 dollars). Depending on the size
of the aerator, capital costs typically range from $60,000
to $200,000 (1982 dollars). Similar cost comparisons are
not available for the other aeration devices. Note that
these aeration costs do not include any expenditures that
may result from secondary effects. For example, off-gas
control has been estimated to double the cost of aera-
tion.108 Adsorption, on the other hand, typically costs
two to three times more than aeration, depending on the
contaminant of concern.108' 1I7
Treatment in the Home
Boiling. One of the first questions generally asked by
health department personnel or homeowners after dis-
covering VOC contamination is: "Can the organics be
removed by boiling?" This is probably because histor-
ically contaminated water contained pathogenic bacteria
and boil-water orders come to mind as the first course of
action. VOCs can be removed by boiling. Table 26 con-
tains a summary of relevant points and precautions.
Table 26. Boiling to Remove Volatile Contaminants94
• All have boiling points or azeotrophic boiling points less
than 100°C.
• The depth of water in the pan affects efficiency; better
removals occur in shallow pans.
• Boiling must be vigorous.
• A timer is necessary; 5 minutes is sufficient to give 95 to
99+% removal.
• A range hood should be used to vent the VOCs out of
doors.
• Inorganics will be concentrated.
• VOC inhalation exposure from showering will not be
eliminated.
Point-of-Use Adsorption Units. In some circumstances, such
as where a central treatment system is not feasible, or in
emergency situations, point-of-use treatment devides us-
ing activated carbon adsorption have been successful.118
Point-of-use devices which use varying amounts of
adsorbents are available. The smallest units attach to the
spigot and provide little contact time. Larger units treat
water prior to the spigot. Reference 119 summarizes
studies of microbial levels across point-of-use
devices—primarily of the larger size. Finally, whole-
house units are available. The state of New York120
recommends only whole-house units because of the pos-
sibility of inhalation exposure to VOCs during and after
showers in the confines of a bathroom. The exposure
issue has been studied121 but is still a debated topic.
A plumbing problem that must be avoided when
installing and operating whole-house units is the intro-
duction of untreated water when the adsorption units
are being backwashed. This permits adsorption of
organics in the bottom of the activated carbon bed.
When returned to service, the organics desorb into the
treated water. Perhaps a more significant problem with
170
-------
point-of-use devices is insuring that they receive the care
necessary to maintain a desirable water quality. The Na-
tional Sanitation Foundation122 has evaluated the
surveillance problems related to providing the mainte-
nance and attention these units require.
Summary
Organic solvents of industrial origin occur in ground-
water supplies and both non-treatment and treatment
actions are being taken by water utilities. Aeration,
adsorption on granular activated carbon, or a combina-
tion of these processes are the technologies of choice.
The following steps are suggested when faced with an
organics problem:
1. Characterize the Problem. Attempt to locate and stop
the source of VOC contamination making certain the
problem is not from materials used to treat, transport,
or store the drinking water. Each well in the system
should be sampled, as should several locations in the
distribution system. If possible, conduct a hydrogeo-
logic investigation.
2. Analyze the Data. The background total organic car-
bon (TOC) and total organic halogen (TOX) concen-
trations are important supplements to knowing specific
organic concentrations. Be certain that the VOCs have
been properly identified.
3. Solve the Problem. What are the non-treatment alter-
natives? What are the treatment alternatives? What are
the budget constraints? What finished water quality is
desired or required? Be sure the solution does not sub-
stitute for the organic problem a situation of microbio-
logical or inorganic deterioration, or increased corro-
sion potential, or of prohibitive air emissions.
171
-------
References*
76Westrick, J., J.W. Mello, and R.F. Thomas. 1984.
"The Groundwater Supply Survey," Journal
American Water Works Association (JAWWA),
Vol. 76, No. 5, 1984, pp 52.
77"National Revised Primary Drinking Water Regula-
tions, Volatile Synthetic Organic Chemicals in
Drinking Water." 1982. Federal Register, Vol. 47,
No. 43, March 4, pp 9350.
78Larson, C.D., O.T. Love, Jr., and G.B. Reynolds.
1983. "Tetrachloroethylene Leached from Lined
Asbestos Cement Pipe into Drinking Water,"
JAWWA, Vol. 75, No. 4, pp 184-188.
79Yoo, R.S. 1984. "Water Quality Problems Associ-
ated with Reservoir Linings and Coatings,"
Proceedings, AWWA Annual Conference, Dallas,
Texas, June 1984.
80"Plumbing Material and Drinking Water Quality."
1984. U.S. EPA, Office of Drinking Water/Office
of Research and Development Technology
Transfer Seminar, Cincinnati, Ohio, May 16-17,
1984
81Stevens, A.A., R.C. Dressman, R.K. Sorrell, and
H.J. Brass. 1984. "TOX, is it the Non-Specific
Parameter of the Future?" Proceedings, Pre-
conference Seminar "Non-Specific Organic
Analysis for Water Treatment Process Control and
Monitoring", AWWA Annual Conference, Dallas,
Texas, June 1984.
82Mabey, W.R., V. Barich, and T. Min. 1983.
"Hydrolysis of Polychlorinated Alkanes," Pro-
ceedings, Division of Environmental Chemistry,
American Chemical Society, 186th National Meet-
ing, Washington, D.C., Vol. 23, No.2, August 28-
September 2, 1983, pp 359.
83Parsons, F., G. Lage, and R. Rice. 1983."Trans-
formation of Chlorinated Organic Solvents in
Ground Water Environments in Southern
Florida," Ibid, p 286.
^Bouwer, E.J. 1983. "Laboratory and Field Evidence
for Transformation of Trace Halogenated Organic
Compounds," Ibid, p 291.
85Parsons, F., P.R. Wood, and J. DeMarco. 1984.
Transformations of Tetrachloroethylene and Tri-
chloroethylene in Microcosms and Ground
Water," JAWWA, Vol. 76, No. 2, p 56
86Gaston, J.M. 1984. "The Dilemma of New Wells
Versus Treatment," Proceedings, Preconference
Seminar: Experiences with Groundwater Con-
tamination, AWWA Annual Conference, Dallas,
Texas, June 1984.
""Unpublished reports and sponsored project informa-
tion available from Director, Drinking Water Research
Division, Water Engineering Research Laboratory, Of-
fice of Research and Development, USEPA, 26 W. St.
Clair Street, Cincinnati, Ohio 45268.
87Seeger, D.R., CJ. Slocum, and A.A. Stevens.
1978. "GC/MS Analysis of Purgeable Contam-
inants in Source and Finished Drinking Waters,"
In: Proceedings: 26th Annual Conference on Mass
Spectrometry and Applied Topics, St. Louis,
Missouri, May 1978.
88Ohio River Valley Water Sanitation Commission.
1980. "Water Treatment Process Modifications
for THM Control and Organic Substances in the
Ohio River," EPA-600/2-80-028, Office of
Research and Development, Drinking Water
Research Division, Cincinnati, Ohio, March 1980.
89U.S. EPA, Drinking Water Research Division, Task
105, Project Officer, R.J. Miltner, 26 W. St. Clair
Street, Cincinnati, Ohio.
^Reusch, W.A. 1977. An Introduction to Organic
Chemistry. Holden-Day, San Francisco, California.
91Sorg, T.J. and O.T. Love, Jr. 1984. "Reverse
Osmosis Treatment to Control Inorganic and
Volatile Organic Contamination," Proceedings,
Preconference Seminar: Experiences with Ground
Water Contamination, AWWA Annual Confer-
ence, Dallas, Texas, June 1984.
^Love, O.T., Jr., W.A. Feige, J.K. Carswell, R.J.
Miltner, and C.A. Frank. 1984. "Aeration to
Remove Volatile Organic Compounds from
Ground Water," Draft Report, U.S. EPA,
Drinking Water Research Division, 26 W. St.
Clair Street, Cincinnati, Ohio, August 1984.
93McCarty, P.L. 1983. "Removal of Organic Sub-
stances from Water by Air Stripping," Control of
Organic Substances in Water and Waste Water,
EPA-600/8-83-011, B.B. Berger, editor. U.S.
EPA, Office of Research and Development, Cin-
cinnati, Ohio, April, 1983.
^Love, O.T. Jr., R.J. Miltner, R.G. Eilers, and C.A.
Fronk-Leist. 1983. "Treatment of Volatile Organic
Compounds in Drinking Water," EPA-600/
8-83-019, U.S. EPA, Drinking Water Research
Division, Cincinnati, Ohio, May 1983.
95U.S. EPA Research Cooperative Agreement
CR809758, "Removal of Volatile Organics from
Ground-Water Supplies by In-Well Aeration,"
Project Officer: R.J. Miltner, Drinking Water
Research Division, Cincinnati, Ohio, 1982-1985.
^Roberts, P.V., and J.A. Levy. 1983. "Air Stripping
of Trihalomethanes," In: Proceedings, AWWA
Seminar, "Controlling Trihalomethanes," Annual
AWWA Conference, Las Vegas, Nevada, June
1983.
^Kavanaugh, M.C. and P.R. Trussell. 1980. "Design
of Aeration Towers to Strip Volatile Contaminants
from Drinking Water," JAWWA, Vol. 72, No.
12.
98Cummins, M.D. and J.J. Westrick. 1982
"Packed Column Air Stripping for Removal of
Volatile Compounds," In: Proceedings, National
Conference on Environmental Engineering,
American Society of Civil Engineers, Minneapolis,
Minnesota, July 1982.
172
-------
"Gossett, J.M. 1983. "Packed Tower Air Stripping of
Trichloroethylene from Dilute Aqueous Solution,"
ESL-TR-81-38, U.S. Air Force Engineering and
Services Laboratory, Tyndall Air Force Base,
Florida, April 1983.
100Termaath, S.G. 1982. "Packed Tower Aeration for
Volatile Organic Carbon Removal," Presented at:
Seminar on Small Water Systems Technology,
U.S. EPA, Cincinnati, Ohio, April 1982. (Tele-
vision video cassette available on loan).
101Cummins, M.D., 1983. "Trichloroethylene Removal
by Packed Column Air Stripping: Field Verified
Design Procedure," In: Proceedings, National
Conference on Environmental Engineering,
American Society of Civil Engineers, Boulder,
Colorado, July 1983.
102Cummins, M.D., "Economic Evaluation of Tri-
chloroethylene Removal from Contaminated
Ground Water," Technical Support Division, Of-
fice of Drinking Water, U.S. EPA, Cincinnati,
Ohio (In Press).
103Silvobitz, A.M., 1982. "Removal of TCE from a
Drinking Water Supply—A Case Study," Pro-
ceedings, ASCE National Conference, New
Orleans, Louisiana, October 1982.
104Cheremisinoff, N.P. and P.N. Cheremisinoff. 1981.
Cooling Towers-Selection, Design and Practice,
Ann Arbor Science Publishers, Inc., Ann Arbor,
Michigan.
105Hesketh, H.E., K.C. Schifftner, and R.P. Hesketh.
1983. "High Efficiency-Low Pressure Drop Wet
Scrubber," Paper No. 38-55.1, 76th Annual
APCA Meeting, Atlanta, Georgia, June 1983.
106U.S. EPA Research Cooperative Agreement
CR809974, "Evaluation of Packed Tower Aera-
tion for Removal of Volatile Organics from Drink-
ing Water," Project Officer: R.J. Miltner, Drink-
ing Water Research Division, Cincinnati, Ohio,
1982-1985.
107Umphres, M., C.H. Tate, J.A. Wojslaw, and J.H.
Van Wagner. 1984. "Feasibility of Counter Cur-
rent Packed Tower Aeration for Removal of Vola-
tile Organic Contaminants from Ground Water,"
Proceedings, AWWA National Conference, Dallas,
Texas, June 1984.
108Environmental Science and Engineering, Inc.,
1985. "Techniques and Costs for the Removal of
VOCs from Potable Water Supplies," ESE. No.
84-912-0300, Draft, Gainesville, Florida, March
1985.
109Fronk-Leist, C.A. and O.T. Love, Jr. 1983. "Eval-
uating the Quality of Water Treated by Redwood
Slat Tower Aerators," Proceedings, AWWA
Water Quality Technology Conference, Norfolk,
Virginia, December 1983.
110McCarty, P.C., K.H. Sutherland, J. Graydon, and
M. Reinhard. 1979. "Volatile Organic Con-
taminants Removal by Air Stripping," Pro-
ceedings, AWWA Seminar "Controlling Organics
in Drinking Water," Annual AWWA Conference,
San Francisco, California, June 1979.
luDilling, W.L., 1977. "Interphase Transfer Processes.
II. Evaporation Rates of Chloro Methanes,
Ethanes, Ethylenes, Propanes, and Propylenes
from Dilute Aqueous Solutions. Comparisons with
Theoretical Predictions," Environmental Science
and Technology, Vol. 11, No. 4, p. 405.
112Miltner, R.J. and O.T. Love. 1984. "A Comparsion
of Procedures to Determine Adsorption Capacity
of VOCs on Activated Carbon," Proceedings:
AWWA Water Quality Technology Conference,
Denver, Colorado, December 1984.
U3Love, O.T. Jr., and R.J. Miltner. 1982. "Removal
of Volatile Organic Contaminants from Ground
Water by Adsorption," Proceedings, Atlantic
Workshop on Organic Chemical Contamination of
Ground Water, AWWA/IWSA, Nashville, Ten-
nessee, December 1982.
114Glasstone, S. and D. Lewis. 1960. Elements of
Physical Chemistry. D. Van Nostrand Company,
Inc., Princeton, New Jersey, pp 265-269.
115Dobbs, R.A. and J.M. Cohen. 1980. "Carbon
Adsorbtion Isotherms for Toxic Organics," EPA
600/880-023, Office of Research and Develop-
ment, Wastewater Treatment Division, Cincinnati,
Ohio, April 1980.
116Love, O.T., Jr. and R.G. Eilers. 1982. "Treat-
ment of Drinking Water Containing
Trichloroethylene and Related Industrial
Solvents," JAWWA, Vol. 74, No. 8, pp 413-425.
117Clark, R.M. and R.G. Eilers. 1982. "Treatment of
Drinking Water for Organic Chemical Contamina-
tion: Cost and Performance," Proceedings, Atlan-
tic Workshop on Organic Chemical Contamination
of Ground Water, AWWA/IWSA, Nashville,
Tennessee, December 1982.
118Gonshor, L., 1982. "Point-of-Use-Systems—House-
hold Units." Presented at Seminar on Small Water
Systems Technology, U.S. EPA, Cincinnati, Ohio,
April 1982, (Television video cassette available on loan).
119Geldreich, E.E., R.M. Taylor, J.C. Blannon, and
D.J. Reasoner. 1985. "Bacterial Colonization of
Point-of-Use Water Treatment Devices,"
JAWWA, Vol. 77, No. 2, p 72.
120New York State Department of Health, Division of
Environmental Protection. 1982. "Point-of-Use
Activated Carbon Treatment Systems," Interim
Report by Ad Hoc Committee on Removal of
Synthetic Organic Chemicals from Drinking
Water, George Stasko, Chairman, Albany, New
York, December 1982.
I21Couch, A.F., 1984. "Assessing Human Inhalation
Exposure to Trichloroethylene Volatilization from
Contaminated Residential Water Supplies," M.S.
Thesis, University of Pittsburgh, Pittsburgh, Penn-
sylvania, June 1984.
122U.S. EPA Research Cooperative Agreement
CR8092480, "Point-of-Use Removal of Volatile
Halogenated Organic Contaminants from Drinking
Water," Project Officer: S.W. Hathaway, Drink-
ing Water Research Division, Cincinnati, Ohio,
1982-1984.
173
-------
Chapter 8
Inground Treatment, Restoration, and Reclamation
Prevention of ground-water contamination is far
more logical, simple, and cost effective than attempting
to correct a problem—a problem that may have been in
existence for years. A great deal of time, effort, and
money are presently being expended to develop remedial
measures to counteract the effects of contaminated
aquifers and public water supplies. These include tradi-
tional as well as innovative construction techniques,
water management, and research initiatives. As far as
the public or private water purveyor is concerned, how-
ever, only a few of these methods are within their realm
of possibility, owing largely to cost considerations.
Assuming, beyond a reasonable doubt, that the
ground-water supply system is actually contaminated,
one has several options: (1) provide inground treatment/
containment, (2) provide above ground treatment, (3)
remove or isolate the source of contamination, (4) aban-
don the source of supply, or (5) ignore the problem.
Generally, several techniques are coupled in order to
achieve the desired results. None of the options, how-
ever, are entirely satisfactory, but in order to maintain
public confidence, something generally will need to be
done and, most likely, in an open manner.
Inground Treatment/Containment
In some situations it might be reasonable to attempt
to degrade or immobilize a contaminant or to contain it
within a specified or general area. In this manner, un-
contaminated wells could still be operated but pumping
schedules and rates would need to be maintained in such
a way that the cones of depression would not interfere
with the containment. Inground treatment/containment
methods appear attractive in two widely different situa-
tions: (1) where the contaminant covers a relatively
small area and can be immobilized and (2) where con-
taminants cover a broad area in which the surrounding
strata are amenable to the emplacement of barrier walls
or the installation of discharge or recharge wells.
In Situ Encapsulation. This technique has proven suc-
cessful in areas where the contaminant can be immobi-
lized chemically or by means of a grout. One interesting
example, reported by Williams,123 involved a 4,200
gallon leak of acrylate monomer from a corroded pipe-
line at a small plant in Ohio. The contaminant migrated
through a layer of fill, consisting largely of cinders, and
then downward through a storm sewer trench into a
thin sand and gravel aquifer. A test boring and soil
sampling program delineated the plume and indicated
that the contaminant was slowly beginning to undergo
polymerization and, therefore, immobilization. To in-
crease the rate of reaction, 2-inch diameter perforated
PVC pipe was buried, about 2 feet below land surface,
in four narrow trenches that trended across the plume.
A riser and manifold header connected each pipe to
solution tanks containing a catalyst in one and an acti-
vator in the other. Both solutions contained a wetting
agent. A total of 8,000 gallons of solution were injected
during the two treatment operations and 1,000 gallons
had been injected previously during the investigative
phase. On the basis of pre- and post-treatment soil bor-
ings, it was estimated that 85 to 90 percent of the liquid
monomer contaminant was solidified and in some places
it exceeded 99 percent polymerization. It is assumed that
the remaining material will polymerize naturally.
Depending on the hydrogeologic situation and the
chemical characteristics of the waste, it would appear
that similar techniques could be used elsewhere. Where
chemical immobilization can not be achieved, it might
be possible to grout in order to limit the potential for
leaching and movement.
Microbial Degradation. Many, if not most, organic com-
pounds can be degraded, at least to some extent, by
microorganisms that occur naturally in the subsurface.
Aquifer restoration by means of microbial degradation
is still in its infancy, techniques are crude at best, and
most results have not been entirely satisfactory. In
general, microbial restoration techniques revolve around
the injection of oxygen and inorganic nutrients, such as
nitrogen and phosphorus sources, into the subsurface in
order to encourage population increases of the indig-
enous microbial flora. Great numbers of microbes occur
naturally underground, but apparently in reduced
numbers because of limitations brought about by the
scanty food supply. Introduction of outside sources of
energy may cause rapid increases in the population and
these, in turn, can degrade a fairly wide range of
organic compounds. When the external source of energy
is removed, the population will die back.
Several problems exist with the concept of microbial
restoration. First, there must be available microorgan-
isms that are compatible with the waste; that is, they
175
-------
must be able to degrade the waste so that it can be used
as a source of carbon. Second, variations in concentra-
tion of the waste may kill or starve the organisms.
Third, methods to evenly distribute the nutrients and
oxygen, particularly where subsurface conditions are
complex, are not well advanced. Furthermore, consider-
able care and monitoring are required to ensure that the
injected air does not force the contaminant to migrate
away from the zone of treatment and further compound
the problem and to ensure that the degradation products
that result from transformation are not more toxic than
the original compound. Additionally, there appears to
be a practical limit, both technically and economically,
to the degree of restoration. Attempting to reduce the
concentration of the contaminants to a zero level would
likely required expenditures of both time and money
that far exceed the benefits derived.
Fortunately, a great deal of research is presently being
conducted on the role and potential use of microorgan-
isms in aquifer restoration. Early results appear prom-
ising. On the other side of the coin, however, this tech-
nology is so new, the data base so small, and our igno-
rance so large that it will probably require many years
before these methods become routine. The path between
laboratory experiment and successful field implementa-
tion is generally tortuous, filled with blind alleys, and
rutted with bad experiences and lack of practicality.
Barrier Walls. The emplacement of a wall of very low
permeability across the path of a contaminant plume,
around a contaminated source or area, or around an
uncontaminated area, has been used with considerable
success for several years. Barrier or cutoff walls, most
commonly, are used in conjunction with well systems.
The purpose of the wells is to withdraw the con-
taminated fluid so that it can be treated and, perhaps,
reinjected. Obviously, the construction of barrier walls
is expensive, but perhaps the most important considera-
tion is the long-term cost of pumping and water treat-
ment, which may require decades.
Barrier walls generally are constructed by one of three
different methods: (1) slurry-wall, (2) grout curtain, or
(3) sheet piling. In the slurry-wall system a trench is
excavated in the proper location and filled with a mix-
ture of bentonite and soil or, perhaps, cement or special
additives. The foundation should lie on, or preferably
in, an underlying unit of low permeability so that
contaminants do not flow under the wall. Slurry walls,
under proper conditions, can be constructed to depths
of 100 feet or so.
Grout curtains are formed by driving or placing
grouting pipe in drilled holes that extend downward to a
unit of low permeability. Cement or chemical grout is
injected in a number of primary holes under high pres-
sure as the pipe is withdrawn. The next set of holes are
placed midway between the primary injection points and
the process is continued until the area will take no more
grout.
Sheet pile cutoff walls have been used for many years
for excavation bracing and dewatering. Where condi-
tions are favorable, depths of 100 feet or more can be
achieved. The system consists of driving interlocking
sheet piles down to a unit of low permeability. Unfortu-
nately, sheet piling is seldom water tight and individual
plates can move laterally several to several tens of feet
when they are being driven. Acidic or alkaline solutions,
as well as some organic compounds, can reduce the ex-
pected life of the system.
Two separate slurry walls have been constructed along
parts of the margin of the Rocky Mountain Arsenal
near Denver in order to contain plumes that originate
on the plant property.124- 125> 126 Along the north
boundary, where unconsolidated sand and gravel occur
with a thickness that averages about 30 feet, the slurry
wall, about 2 feet thick, is 6,800 feet long. On the
upgradient side are a series of 35 12-inch diameter
discharging wells on 200 foot centers that pump con-
taminated ground water into a treatment facility. After
flowing through a carbon filtration system the water is
then reinjected into 50 6-inch diameter recharge wells on
100 foot centers on the opposite side of the barrier.
Along the northwest boundary is a recently completed
bentonite slurry barrier that is 1,425 feet long, extending
southwestward from a bedrock high. The wall, ex-
cavated with the bentonite slurry trench method, is 30
inches wide and extends 3 feet into the underlying
bedrock. The barrier contains about 7,000 cubic yards
of backfill, obtained from a borrow pit and blended
with the bentonite prior to emplacement. The barrier
was constructed where the saturated thickness of the
permeable material is less than 10 feet. Paralleling the
downgradient side of the barrier is a series of 21
recharge wells, stretching nearly 2,100 feet along the
Arsenal boundary. Directly behind the barrier and ex-
tending into the thicker part of the surficial aquifer are
15 discharge wells. The contaminated ground Water is
pumped to a treatment plant and then reinjected into
the recharge wells, thus forming a hydraulic barrier.
Farther southeast along the boundary is another
hydraulic barrier system, about 1,500 feet long, that
consists of two parallel rows of discharge wells with 15
wells per row and, downgradient, a row of 14 recharge
wells. The contaminanted water, originating from a
spill, is pumped, treated, and then reinjected. This
system and the one along the north boundary was put
into operation in late 1981 and the system along the
northwest boundary began operation in 1984. The cost
of the construction of the systems as well as the operat-
ing expenses are unknown.
A modified barrier design was described by Arlotta
and others.127 Called Envirowall, this system consists of
a trench that is lined with 100 mil high density polyeth-
ylene and backfilled with sand. It is installed with the
slurry trench construction method. The system was
tested in New Brunswick, New Jersey, in the fall of
1982.
Brunsing and Cleary128 described a method of com-
plete isolation by slurry-induced ground displacement or
block displacement. Demonstrated in Whitehouse,
Florida, a slurry wall was constructed around a small
area, 60 feet in diameter, to a depth of 23 feet in
unconsolidated material. Injection wells were then used
176
-------
to force a soil bentonite slurry outward along the bot-
tom of the cell. Subsequent test holes indicated that the
new floor of the cell contained 5 to 12 inches of slurry.
In a case such as this, wells and a treatment system
would be required to keep leachate from overflowing
the cell unless an effective cover that totally prohibited
infiltration was placed over the unit.
Cutoff Trenches and Ground-Water Dams. A cutoff trench is
merely a drain that extends a sufficient distance below
the water table to intercept contaminated ground water.
Construction can be simple to complex depending on
soil and subsurface conditions and on the thickness of
the unsaturated zone. Since a cutoff trench is designed
to serve as a ground-water discharge line, in a manner
similar to a perennial stream, hydrologic and topo-
graphic conditions are critical to its success. That is, it
can be effective where the water table is near land sur-
face, the contaminants are moving at a relatively
shallow depth, and the geologic conditions permit exca-
vation. Cutoff trenches are effective in both high and
low permeability materials, although the latter would
likely provide less water, which in turn would require
less storage and treatment costs. Furthermore, they are
particularly suited to immiscible contaminants that tend
to float. The trenches can remain open, filled with
gravel, or be fitted with perforated underdrain pipes.
A ground-water dam installation was described by
Giddings.129 In this case, a landfill that began as a
burning dump, was found to be discharging leachate
both to the surface and the ground water, much of
which eventually flowed into an adjacent river. A
leachate interceptor trench was constructed on the
downgradient side of the disposal area as shown in
Figure 180. In the trench on the upgradient side was
placed a perforated pipe in a gravel envelope that was
covered with permeable material. The remainder of the
trench on the downgradient side was then backfilled
with fine-grained material as shown in Figure 181.
Leachate from the landfill flows into the filled trench,
seeps into the perforated pipe, and then is pumped back
into the landfill. An application to discharge to a
regional sewer line was under consideration at the time
Giddings' report was published. In this case, the main
Upgradient Monitoring
Well
Downgradient
Monitoring Well
Figure 180. Site Layout129
177
-------
V
.
* • ' '. '• \
River
t^ . • >.
?.'.-• • • :. •'.
Perforated'
Pipe
Gray Till
Figure 181. Site Cross Section129
purpose of the ground-water dam was to prohibit water
originating in the adjacent river from flowing into the
trench, which would have substantially increased the
volume of wastewater.
Gradient Reversals. Controlling the movement of contam-
inated ground water by means of recharge and discharge
wells has been practiced for several years. The major
disadvantages include the commonly long pumping
periods, well construction and maintenance costs, and
the fact that the subsurface geology dictates system
design. The cone of depression around a pumping well
can be controlled by the discharge rate and thus change
ground-water flow directions as well as velocity. Man-
agement of the cone or cones permits the operator to
capture contaminants, which can then be diverted to a
treatment plant. Well placement is particularly impor-
tant since proper spacing and pumping rate are required
to capture the contaminants. Moreover, well placement
should be optimized so that as little uncontaminated
water as possible is produced in order to reduce treat-
ment costs.
Recharge wells are used to develop a hydraulic bar-
rier. In this way they can be used to force the contam-
inant plume to move in preferred directions, such as
toward a drain or discharging well.
The design of well systems is, in large part, based on
trial and error methods coupled with experience. Herein
also lies one of the most useful exercises of computer
simulations because one can easily evaluate different
schemes and estimate costs. Of course as in all simula-
tions, the output can never be more valid than the input
data.
Gradient reversal techniques are used at a great
number of contaminated sites and nearly always play
some role in containment methods, as is the case at the
Rocky Mountain Arsenal.
Surface Sealing. Nearly all of the leachate that is gen-
erated from solid waste is derived by infiltrating rain
and surface runoff that slowly percolate through the
waste removing the water-soluble products. Clearly, one
type of control to reduce the quantity of leachate
generation is to stop or greatly reduce the volume of
water that flows through the waste. One practical
method is to provide a cover or cap that promotes sur-
face runoff and prohibits infiltration. The cover, which
would require continual maintenance and monitoring,
could be constructed of clay, perhaps blended with ben-
tonite or other additives, and used in combination
with any of a number of liners sandwiched between clay
layers.
Monitoring and maintenance of the cover are impor-
tant to ensure that water cannot be impounded as the
waste settles, that cracks caused by wetting and drying
do not serve as a path through the cover, and that bur-
rowing organisms do not penetrate the seal. Although
cover monitoring and maintenance are commonly ig-
nored after construction, the time and costs that they
involve can be infinitely smaller than a cleanup opera-
tion. The substantial decrease in the size and concentra-
tion of a leachate plume brought about by surface seal-
ing of the Windham, Connecticut, landfill is briefly
discussed elsewhere in this report.
Source Removal. In some cases it is justifiable to remove
a source of contamination and backfill the excavation
178
-------
with clean material. On the other hand, removal might
be dangerous in itself, a secure site must be available to
dispose of the excavated material, it will certainly be ex-
pensive, and the process might create additional water
contamination problems because removal is likely to in-
crease the rate of decomposition. As described else-
where, once the unsaturated zone becomes contami-
nated, it might require years or even decades for the
material to be slowly flushed from the zone and,
throughout this time, it may continually or intermit-
tently supply a leachate. Of particular concern are those
sites where the source material extends below the water
table or where the annual fluctuation of the water table
causes cyclic wetting and drying.
Source removal is certainly a viable option, but it
must be predicated on a sound technical and economic
understanding of all of the potential consequences.
Would it be more sound, for example, to provide an
effective seal than to remove it? Or is it so toxic that
there is no other choice? In many past restoration at-
tempts removal was accomplished with little or no prob-
lem, other than cost.
Noncontainment
Sometimes a contamination problem is more apparent
than real, but under emergency conditions or public/
regulatory pressure little thought may be given to the
consequences. As an example, let us assume there is a
derailment and tank cars, containing 300,000 gallons of
a special drilling fluid, are ruptured and the 30,000 mg/1
chloride brine is spilled along the railroad embankment.
At the end of two days all of the brine has infiltrated.
The derailment occurred in the middle of the wide flood
plain of a midwestern river, a river that flows generally
no more than three or four months each year. The river
channel in the vicinity of the spill lies along the north
edge of the flood plain, as shown in Figure 182. In this
rural area the nearest public water supply is at a recrea-
tional site a mile downgradient from the spill. What
should be done?
Only limited hydrogeologic data are likely to be avail-
able, but many conditions can be inferred from a topo-
graphic map and a basic understanding of ground water
geology. The few water wells in the flood plain indicate
that the average thickness of the alluvium is about 50
feet. Drillers logs of the wells and an examination of
borrow and sand pits indicate that most of the alluvium
is a medium to coarse sand, which should have an effec-
tive porosity of about 20 percent. An estimate of
ground-water velocity, 1.5 feet/day, is based on aquifer
grain size and the gradient of the flood plain, which
should be about the same as the water table. Under
these conditions, an estimate of the longitudinal and
transverse dispersivities should be about 70 and 15 ft.
Since the major contaminant is chloride, it will not be
retarded by the aquifer material, but rather will flow at
the same velocity as the ground water.
If the above estimates are used as input for a simple
computer transport model or the nomograph described
earlier, one can obtain a general impression of the con-
taminant movement with time and then develop some
guidelines for reclamation, if necessary. A simulation
representing the contaminant plume one year after the
spill is shown in Figure 183. Notice that the plume has
migrated downgradient at least 1,500 feet and the largest
N
Margin of Floodplain
Derailment
Recreation Site Well
Sand pit
Margin of Flood Plain
0
1000
I
Scale, in feet
Figure 182. Map View of Railroad Spill Area
179
-------
PLUME AFTER 365 DAYS
1500 + .
1250 + .
1000 + .
500 + .
250 + 2
0 +14
-250 + 2
-1250+ .
-1500+ .
17 63120 95 33 5
X496805627220 34 2 ,
17 68120 95 33 5
.....................
-025711112222333344445
2 50502570257025702570
5 00005050505050505050
0 00000000000000000
Figure 183. Chloride Distribution, in mg/l, 365 Days After the Spill
PLUME AFTER 1825 DAYS
1750 + .
1500 + .
1250 + .
1000 + .
750 + .
500 + .
250 + .
0 + .
-250 + .
-500 + .
-750 + .
-1000+ .
-1250+ .
-1500+ .
-1750+ .
. . . 1
. . . 1
. . . 1
+__+__+„+
X
2
4
2
1
•— +
1
2 5
7 16
10 23
7 16
2 5
. i
+ +„
i
10
32
47
32
10
1
— +
345554
18 26 33 35 32 25
55 82103111101 79
81120151162148115
55 82103111101 79
18 26 33 35 32 25
345554
j ^ j j j ^
+__+__+__+„+ +
?
17
52
76
52
17
2
,_+.
1
9
29
43
29
9
1
— +
1
4
14
20
14
4
i
+.
2
6
8
6
2
._+
1
2
3
2
1
+.
1 .
1 .
1 .
.-+— +
-02571 1 1 12222333344445
2 50502570257025702570
5 00005050505050505050
0 00000000000000000
Figure 184. Chloride Distribution, in mg/l, 1,825 Days After the Spill
180
-------
concentration is 805 mg/1. If the river had been peren-
nial, the plume would have moved toward it and even-
tually provided part of its flow. The concentration dis-
tribution of chloride in the plume 5 years after the spill
is shown in Figure 184. The leading edge of the plume
has moved nearly 5,000 feet from the spill but the max-
imum concentration is only about 162 mg/1.
This example illustrates that a few minutes with a
simple computer model or nomograph clearly indicated
that an accidental spill of the type described above
really presented no problem at all because the aquifer
diluted the contaminant to levels below drinking water
standards. Furthermore, the limit on chloride in drink-
ing water was established only for taste considerations.
On the other hand, had the spill consisted of organic
compounds the aquifer might not have been able to suf-
ficiently dilute or retard the plume and corrective action
might have been warranted.
In a great number of cases only one or two wells in
the supply system are contaminated and these only with
barely detectable concentrations. Depending on the
health hazard of the contaminants, its concentration in
the wells, and the volume pumped by the remaining
wells, as well as on regulatory constraints, it might be
possible to blend the water from all of the wells to
reduce the concentration below detectable limits. There
appears to be some philosophical differences to this
solution among regulatory agencies. Some believe that if
a contaminant is below detection limits that it would
present no health hazard or at least an acceptable
hazard. Others apparently have the opinion that if a
part of the supply is known to be contaminated then it
should not be used regardless of the concentration.
Some public water-supply systems contain excessive
concentrations of ions that are derived from natural
sources. These are largely inorganic substances that are
likely to be more troublesome than hazardous, such as
chloride or hardness. Excessive concentrations of these
constituents might be the result of salt water intrusion
or interaquifer leakage. In such cases the best control is
probably a reduction of pumping, if possible, and
blending with other supply sources that are less
mineralized.
In places where the water supply system or the
aquifer is grossly contaminated, it may be impossible to
restore it despite the expenditure of great sums of
money and time. Where this occurs, it might be
necessary to simply abandon the system as a supply
source. Monitoring and perhaps some reclamation
would be required in order to reduce the potential for
contamination of other sources of water.
181
-------
References
123 Williams, E.B. 1982. "Contaminant Containment
by In Situ Polymerization." Proceedings 2nd Na-
tional Symposium on Aquifer Restoration and
Ground-Water Monitoring, National Water Well
Association, pp 38-44.
124Shukle, R.J. 1982. "Rocky Mountain Arsenal
Ground-Water Reclamation Program." Pro-
ceedings 2nd National Symposium on Aquifer
Restoration and Ground-Water Mentoring, Na-
tional Water Well Association, pp 366-368.
125Pendrell, D.J. and J.M. Zeltinger. 1983. "Contam-
inated Ground-Water Containment/Treatment
System at the Northwest Boundary, Rocky Moun-
tain Arsenal, Colorado." Proceedings 3rd National
Symposium on Aquifer Restoration and Ground-
Water Monitoring, National Water Well Associa-
tion, pp 453-461.
126Hager, D.G., C.E. Smith, C.G. Loren, and D.W.
Thompson. 1983. "Ground-Water Decontamina-
tion at Rocky Mountain Arsenal." Proceedings 3rd
National Symposium on Aquifer Restoration and
Ground-Water Monitoring, National Water Well
Association, pp 123-134.
127Arlotta, S.V., G.W. Druback, and N. Cavalli. 1983.
"The Envirowall Vertical Cutoff Barrier." Pro-
ceedings 3rd National Symposium on Aquifer
Restoration and Ground-Water Monitoring, Na-
tional Water Well Association, pp 23-27.
128Brunsing, T.P. and J. Cleary. 1983. "Isolation of
Contaminated Ground-Water by Slurry Induced
Ground Displacement." Proceedings 3rd National
Symposium on Aquifer Restoration and Ground-
Water Monitoring, National Water Well Associa-
tion, pp 28-36.
129Giddings, T. 1982. "The Utilization of a Ground-
Water Dam for Leachate Contaminant at a Land-
fill Site." Proceedings 2nd National Symposium on
Aquifer Restoration and Ground-Water Monitor-
ing, National Water Well Association, pp 23-29.
182
•ftU.S. GOVERNMENT PRINTING OFFICE: 1985-559-111/20696
------- |