EPA-600/9-77-014
June 1977
PROCEEDINGS OF THE THIRD NATIONAL GROUND
WATER QUALITY SYMPOSIUM
Robert S. Kerr Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This document is avaflable to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/9-77-014
June 1977
PROCEEDINGS OF
THE THIRD NATIONAL GROUND WATER
QUALITY SYMPOSIUM
Cosponsored by the
U.S. ENVIRONMENTAL PROTECTION AGENCY
and the
NATIONAL WATER WELL ASSOCIATION
September 15-17, 1976
Las Vegas, Nevada
Contract No. 68-03-2396
Project Officer
Jack W. Keeley
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 74820
-------
DISCLAIMER
This report has been reviewed by the Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
The Environmental Protection Agency was established to coordinate
administration of the major Federal programs designed to protect the
quality of our environment.
An important part of the Agency's effort involves the search for
information about environmental problems, management techniques, and new
technologies through which optimum use of the Nation's land and water
resources can be assured and the threat pollution poses to the welfare
of the American people can be minimized.
EPA's Office of Research and Development conducts this search through
a nationwide network of research facilities.
As one of these facilities, the Robert S. Kerr Environmental Research
Laboratory is responsible for the management of programs to: (a) investi-
gate the nature, transport, fate, and management of pollutants in ground
water; (b) develop and demonstrate methods for treating wastewaters with
soil and other natural systems; (c) develop and demonstrate pollution con-
trol technologies for irrigation return flows; (d) develop and demonstrate
pollution control technologies to prevent, control, or abate pollution from
the petroleum refining and petrochemical industries; and (f) develop and
demonstrate technologies to manage pollution resulting from combinations of
industrial wastewaters or industrial/municipal wastewaters.
This report contributes to the knowledge essential if the EPA is to
meet the requirements of environmental laws that it establish and enforce
pollution control standards which are reasonable, cost effective, and
provide adequate protection for the American public.
William C. Galegar u
Director
in
-------
ABSTRACT
The Third National Ground Water Quality Symposium was held in Las
Vegas, September 15-17, 1976, in conjunction with the annual convention
of the National Water Well Association.
The Symposium was dedicated to the late Thomas P- Ahrens, and the
keynote address was given by Charles C. Johnson, Jr., Chairman of the
National Drinking Water Advisory Council.
There were eight main sessions emcompassing 24 technical papers.
These were concerned with the disposal of waste on the land, the move-
ment of pollutants in the subsurface, and artificial recharge. A
special session was dedicated to ground water in the Las Vegas Valley.
The Transactions of this Symposium are submitted in fulfillment of
Contract No. 68-03-2396 by the National Water Well Association under
the sponsorship of the U.S. Environmental Protection Agency.
IV
-------
TABLE OF CONTENTS
2 Waste Disposal — A Most Pervasive Problem Charles C. Johnson, Jr.
9 Thomas P Ahrens - Father of Water Well Technology Michael D. Campbell
11 Wastewaters in the Vadose Zone of Arid
Regions: Hydrologic Interactions John F. Mann, Jr.
18 Wastewaters in the Vadose Zone of Arid
Regions: Geochemical Interactions Donald D. Runnells
30 Zoning Aquifers for Tertiary Treatment of Wastewater Herman Bouwer
40 Potential Replacement of Septic Tank Drain Fields by
Artificial Marsh Wastewater Treatment Systems C. W. Fetter, Jr., W. E. Sloey & F. L. Spangler
47 Ground-Water Pollution Aspects of Land Disposal
of Sewage from Remote Recreation Areas Nils Johnson & Dean H. Urie
55 The Selection and Management of Feedlot Sites and
Land Disposal of Animal Waste in Boise Valley, Idaho L. L. Mink, C. M. Gilmour, S. M. Beck,
J. H. Milligan & R. L. Braun
70 Prediction of Future Nitrate Concentrations
in Ground Water C. P Young, D. B. Oakes & W. B. Wilkinson
83 The Contribution of Fertilizer to the
Ground Water of Long Island Joseph H. Baier & Kenneth A. Rykbost
93 Abatement of Nitrate Pollution in a Public-Supply Well
by Analysis of Hydrologic Characteristics L. A. Eccles, J. M. Klein & W. F. Hardt
99 Design and Optimization of Ground-Water Monitoring
Networks for Pollution Studies H. O. Pfannkuch & B. A. Labno
107 Tracing Sewage Effluent Recharge — Tucson, Arizona T. R. Schultz, J. H. Randall,
L. G.Wilson & S.N.Davis
116 Monitoring Cyclic Fluctuations in Ground-Water Quality Wayne A. Pettyjohn
125 The Lycoming County, Pennsylvania, Sanitary Landfill:
State-of-the-Art in Ground-Water Protection M. Todd Giddings, Jr.
135 Ground-Water Chemical Quality Management
by Artificial Recharge H. I. Nightingale & W. C. Bianchi
143 The Dashte-Naz Ground-Water Barrier
and Recharge Project Dennis E. Williams
152 Injection/Extraction Well System—A Unique
Seawater Intrusion Barrier N. Thomas Sheahan
171 A Nonstructural Approach to Control Salt
Accumulation in Ground Water Otto J. Helweg
178 Quantifying the Natural Flushout of Alluvial Aquifers J. S. Fryberger & W. H. Bellis
186 Improving the Sanitary Protection of Ground Water in
Severely Folded, Fractured, and Creviced Limestone E. E. Jones, Jr. & C. M. Murray
195 Land Disposal of Hazardous Wastes: An Example
from Hopewell, Virginia D. H. Walz & K. T. Chestnut, Jr.
201 Land and Water Use Impacts on Ground-Water
Quality in Las Vegas Valley Robert F. Kaufmann
210 Bull Session on Predicting Physical and Chemical Alteration of Land-Treated
Wastewater, and Land Disposal of Sewage
214 Bull Session on Controlling Pollution from Sanitary Landfills,
and Reduction of Nitrate Contamination
220 Bull Session on Monitoring the Flow of Polluted Ground Water,
and Artificial Recharge as a Solution to Pollution
224 Bull Session on Managing the Movement of Contaminants, and Protecting Mines, Wells, and Pits
230 Discussion Questions and Answers W. B. Wilkinson
v
-------
PROCEEDINGS OF THE THIRD NATIONAL
GROUND WATER QUALITY SYMPOSIUM
September 15-17, 1976
Las Vegas, Nevada
-------
WASTE DISPOSAL - A MOST PERVASIVE PROBLEM8
by Charles C. Johnson, Jr.b
I am most pleased to be here today to speak before The Third
National Ground Water Quality Symposium—in part because it is my
first opportunity to speak before such a large and distinguished audience
since becoming a member of the National Drinking Water Advisory
Council, and in part because of my belief that we must constantly be
aware of potential insults to all of the mediums—air, land, and water-
that comprise our environment. Of these, our water resources, more
perhaps than any other, illustrate the interaction of all parts of the
environment and particularly the recycling process that characterizes
every resource of the ecosystem. Everything that man himself injects
into the biosphere—chemical, biological or physical—can ultimately
find its way into the earth's water. And these contaminants must be
removed, by nature or by man, before that water is again potable.
But mostly, I am pleased because I believe that once again in this
decade the United States is on the verge of another environmental
awakening, and this time it is directly focused on man's insults to his
environment as they impact on the health and well-being of people.
This is evident by the sustained attention and debate by the public on
the proposed toxic substances' legislation; the broad citizen
acceptance of the principle of nondegradation of our air resources; the
firm stand evolving with respect to continued and forceful implementa-
tion of the water pollution control program; and the several court
decisions that support the rights of individuals to be protected from
potential as well as proven harmful effects attributable to their environ-
ments. Symposia such as this that bring together experts of proven
scientific competence and experience to exercise their mature judgment
in the light of all existing data, will prov-ide the fuel that is necessary to
sustain this continuing and expanding concern for problems associated
with the environment while contributing to the solutions of environ-
mental problems we so desperately need.
Close analysis may reveal that America's biggest problem, and
indeed maybe its biggest enemy, is its ever-growing and yet unmanaged
production and disposal of waste products. This apprehension is true
whether our concerns embrace social, political, or economic progress;
Presented at The Third National Ground Water Quality Symposium, Las
Vegas, Nevada, September 15-17, 1976.
bChairman, National Drinking Water Advisory Council, and Resident Manager
Malcolm Pirnie, Inc., Consulting Environmental Engineers, 1629 K St. NW,
Washington, D.C. 20006.
-------
ending of inflation and unemployment; increasing international trade
and producing a balance of payments; expanding agricultural production
or producing the goods desired for modern living; or creating the
energy independence so important to our survival as a Nation. The
production of unwanted waste products is part and parcel to every action
and reaction that accompanies proposed, accepted, and implemented
solutions to all of the societal problems that befall our communities
and our Nation. Our approach to these solutions affect us as a people
and affect our place in the world in which we live.
A discussion of the tremendous volumes of waste produced in
this country tends to boggle the mind. U.S. industries treat about
5,000 billion gallons per year of wastewater before discharging it into
the environment. Of this volume, about 1,700 billion gallons are pumped
to oxidation ponds or lagoons for treatment or as a step in the treatment
process. Another 5,375 billion gallons per year of municipal wastewater
is discharged to the environment, with an estimated 730 billion gallons
of this amount discharged to the land. In the United States, municipal
sludge production amounts to four million dry tons per year. Industrial
sludge production is believed to be many times this amount. Some
estimates place annual municipal solid waste production at more than
170 million tons per year and growing. To these figures must be added
the millions of tons of gaseous wastes that are produced annually; the
untold hundreds of millions of tons of mining tailings disposed of each
year; and the tremendous volumes (more than 24 million barrels per
day) or oil field brines produced each day.
As a result of all the actions and interactions that accompany and
follow the production of these wastes, we continue to create the kind of
environment that fouls our air, pollutes our streams, desecrates our
lands and endangers our people. In the process we create enmity between
persons, problems between communities, disputes between states, and
division between countries. We set people against industry, industry
against government; and place the government, seemingly, against both
industry and the people. These reactions are manifest in part by the
many court suits associated with efforts to regulate waste disposal
practices and to locate waste disposal sites at the local, regional, state,
and national levels. No one wants another's waste, and neither do we
want our own.
In our zeal to correct the problems we have created for ourselves,
we have almost legislated the Nation into a no-win situation. We have a
national air pollution control law that says you can't put the waste in the
air; a water pollution control law that says the Nation's goal is zero
-------
discharge; and a safe drinking water law that says you cannot contaminate
the domestic water supply. The only relatively unregulated medium
available for disposal of our unwanted waste products is the land.
There are those who argue, "Why allow the discharge of any waste?
Why not create the conditions under which all industries-public and
private-recycle or reuse all waste products?" You and I know that the
simple answer is that, sooner or later, there is an irreducible residual
which, even under theoretical conditions, must be disposed of or stored
in a safe repository. In actual practice geography, economics, waste
characteristics, and technology preclude total recycling and reuse of
many waste products. We have seen that the last repository not totally
regulated for the disposal of these wastes is the land. History has
demonstrated that waste products tend to gravitate for disposal
purposes to the point of least regulation.
This then brings me to the point I wish to make today: Our
continued failure to recognize, appreciate, deal with or accept the fact
that the environment in which we live is interconnected and interrelated
in every aspect of its being—its air, its water, and its land—can undercut
all of the good intentions manifest in the legislation and programs
that have been implemented to restore the integrity of the environment.
The evidence suggests that we are blindly engaged in activities
that could despoil two important and necessary resources—our land and
the ground-water aquifers that exist underneath it. In our eagerness to
make progress in dealing with the problems of the environment—and
we are making progress—we seem to be pursuing a course, where land
and waste products are concerned, of "out of sight, out of mind." We
are permitting—and even promoting in some circumstances—under the
guise of environmental conservation, or as accepted waste disposal
practice, or under the practice of zero discharge of liquid wastes, the
deposition of treated and untreated sewage and other hazardous wastes
onto and into our lands and over our ground-water aquifers with little,
and in some instances, no restraints.
I am told that in this country at least one-half of the population
depends upon ground water as a source of drinking water. Of the total
population, 29 percent use ground water delivered by community
systems, and another 19 percent have their own domestic wells. These
Americans would like to think—to believe—that because of the recent
concern for the environment, and the volume of legislation that has been
enacted to support this concern, their ground-water supplies are
protected against the insults of man-made pollution. Unfortunately,
this is not the case. While both the Water Pollution Control Act and the
Safe Drinking Water Act serve much needed purposes, neither provides
the kind of authority required to really protect our ground-water
supplies from possible degradation. Whether they should or not suggests
the basis for an interesting discussion.
The basic and primary and essentially sole interest of the Water
Pollution Control Act Amendments of 1972 is directed to control the
pollution of surface waters. In fact, sections of the Act that require
consideration of new and alternative waste treatment and disposal
methods are used to support and promote the use of land application
of sewage and other liquid wastes and their residuals as a method of
achieving zero discharge. When a land application methodology is used
no discharge permit is required unless runoff is collected for discharge '
from a point source to a surface stream. In the absence of a permit
there is no legal responsibility or authority in the Federal program to
require discontinuance of the practice, with one exception—the
Administrator may act under his emergency powers to protect the
public health. Few states and localities have standards and regulations
-------
to deal adequately with these practices.
What are the implications of this situation with respect to our
ground-water supplies? A recent publication of the EPA states, "The
ground-water quality and level will be affected by most land
application systems. Exceptions would be many overland flow, under-
drained, and pumped withdrawal systems. Wastewater constituents
that are not used by the plants, degraded by microorganisms, or fixed
in the soil may leach to ground water" (U.S.E.P.A., 1975). In a
literature review in 1974, EPA concluded, "Although the published
coverage on the various sources of ground-water pollution is uneven, it
appears that the number of locations where pollution occurs is very
large; however, the area affected is believed to be quite small" (U.S.
E.P.A., 1974).
Gerba, et al. (1975) of the Department of Virology and
Epidemiology at Baylor College of Medicine, within the year
published an article entitled "Viruses in Water: The Problem and Some
Solutions." In the article, the authors warn that "while land disposal
of domestic wastewater has been practiced on a large scale in Europe for
several decades, little is known about the fate of viruses during and
after application to the soil." They continue, "It has been feared that
without dilution in receiving waters, a greater threat will exist to
ground water if sewage is discharged directly onto the soil in large
quantities. It is also felt that the deep well injection of wastes may
pose similar problems if not controlled." Discussion and examples
included in the article are supportive of this concern. The authors
observed that between 1946 and 1960,61 percent of all waterborne
outbreaks of disease in the United States were caused by contaminated
ground water (Gerba et al., 1975).
Schif f man recently reported on a survey of hazardous waste
disposal in the State of Maryland and concluded, "It now appears that
all or nearly all existing discharges of industrial wastes to the ground
and underground waters are causing some environmental damage.
The basic question is the degree of impact" (Schif f man, 1975). The
report noted that in the State, approximately 2.4 million tons of
industrial waste are generated annually from the chemical industry
alone. It further noted that 75 percent of 530 million gallons of liquid
waste are disposed of to the ground at the place of generation. Under-
ground water is the supply source for over 40 percent of the people in
Maryland and is the source of stream base flow. What is the danger to
our underground aquifers if Maryland's experience is translatable to the
Nation's total industrial capacity?
Unfortunately, our concern cannot stop with the treatment and
disposal of the liquid fraction of wastewater flow. By 1985, municipali-
ties alone are expected to produce some 260 million wet tons of
sludge annually. This material cannot legally be placed in our streams or
dumped in the ocean, and its incineration is discouarged in most
localities. As a result, it must be placed on the land. By simple
deduction, sludge has to be as dirty as effluent is clean. It contains all
the elements which have been removed to protect our surface streams
and our oceans. Why then is its uncontained and uncontrolled
placement on the land such an improvement in the protection of the
environment?
The Safe Drinking Water Act was enacted in part to provide the
protection to our ground-water supplies that is not obtainable under the
water pollution control legislation. A reading of the House report easily
allows one to draw this conclusion. The report noted that "underground
injection of contaminants is clearly an increasing problem. Municipalities
are increasingly engaging in underground injection of sewage, sludge, and
-------
other wastes. Industries are injecting chemicals, byproducts and wastes.
Energy production companies are using injection techniques to increase
production and to dispose of unwanted brines brought to the surface
during production. Even government agencies, including the military,
are getting rid of difficult-to-manage waste problems by underground
disposal methods. Part C is intended to deal with all of the foregoing
situations insofar as they endanger underground drinking water sources'
(Safe Drinking Water Act, 1974). By way of identification, Part C in the
Act is entitled "Protection of Underground Sources of Drinking Water."
Without a doubt this is a very broad and laudable mandate. I for
one would like to believe that this indeed was the intention of Congress
when it enacted the legislation. I was further encouraged by the words in
the House report that defined the term "injection." They say, "The
definition of underground injection is intended to be broad enough to
cover any contaminant which may be put below ground level and
which flows or moves, whether the contaminant is in semi-solid, liquid,
sludge or any other form. . This definition is not limited to the
injection for disposal purposes."
What could be clearer? In the broadest possible context, in the
words of the Committee and subsequently the Congress, the primary
concern of this portion of the Safe Drinking Water Act is the protection
from contamination of the Nation's ground-water resources that are or
may be used for domestic water supplies. In the drafting of the
legislation, however, a problem arises. As the lawyers interpret the
operative words in the Act, the words "well" and "injection" are
regarded as a unit and as such limit the scope of authority inherent in
the Act to protect the ground water. Thus, subsequent regulations can
only regulate the practices whereby subsurface emplacement of wastes
occurs through "well injection."
By definition, pits, ponds, and lagoons—which many believe to be
the major contributors to ground-water pollution—and the practice of
land application of wastes under the impetus of zero discharge are not
included in the proposed underground injection control program. At the
Federal level, these practices can only be regulated by the emergency
powers granted the Administrator under the Act. If it is true that the
legal interpretation of the legislation precludes provision for the
protection that is required for underground-water sources, it is most
important that the public and the Congress be made aware of this
problem and that corrective legislation be enacted to clarify the situation.
We must constantly remind ourselves that contamination of
ground-water aquifers can occur easily and rapidly. Once they are
contaminated it may take many years and could cost millions of
dollars to purge the aquifer of the contaminant. Unless we are willing to
treat the water from these underground sources as we do that from our
surface streams, we must exert every effort to prevent their becoming
contaminated.
Lest we become too complacent in the progress that is being made
through our legislative and regulatory process and programs, it
should be recognized that even under the practices permitted by the
proposed underground injection regulations, insofar as protection of
the public's health is concerned, there is some risk, supported by
considerable ignorance. My reading tells me that there is much more
unknown than is known about the significance of the quality of waste
discharged to our underground aquifers and the ultimate fate of those
wastes once they have been placed in the ground. Some studies have
been conducted on the travel of some pollutants, under some
conditions. Few studies have fully addressed, and none have resolved
our current concerns for water reuse, ground-water recharge, discharg'e
-------
of highly concentrated effluents and sludges, or pretreatment require-
ments as related to the inherent relationship of these practices to long-
term exposure of people to the trace elements ultimately associated with
these wastes. We are told that the significance of the long-term ingestion
of low levels of pollutants of water is difficult to ascertain. We read
that their effects are insidious and are often screened by other factors
with other effects. We know that most of these elements are not good
for us; we just don't know when, and how bad, they are for us.
Okun (1976), writing in the American Journal of Public Health,
referenced a London study that indicated that persons drinking water
from a safe underground source experienced lower cancer mortalities
than persons in another population group in the same city provided
with treated water from a polluted source. In the same article, he noted
that "studies made in Holland showed the municipalities receiving their
drinking water from polluted rivers had a higher cancer death rate than
those taking their water from purer underground sources." These
references suggest that once a water supply is polluted, even treatment
of the water does not assure the level of protection we desire for the
public compared to the use of a protected and unpolluted source. Why
then should we even consider a disposal methodology for situations
that have the potential for contamination of naturally safe ground
water?
Many years ago, a decision was made to use our streams, our
rivers, our oceans, and the air over our lands for the disposal of the
waste products of our progress. That decision was made in ignorance
and supported by convenience. Waste volumes were small and disposal
costs were low. Today we are paying the price for our lack of knowledge
and absence of forethought. Notwithstanding the untold illnesses and
deaths that some relate to these past activities, the many billions of
dollars allocated from tax dollars and by industries to reclaim our
streams and our air is testimony to the folly of our past actions. Doesn't
this history teach us something? Shouldn't we avoid repeating our past
mistakes? Will we take the steps that are necessary to preserve our land
for the production of crops and our underground-water resources for
domestic water-supply needs?
Earlier on, I made the statement that, in addressing our
requirements for cleaning up our environment, we had almost legislated
ourselves into a no-win situation—don't discharge wastes to the air, don't
discharge wastes to the water. Now it appears that I am suggesting a
zero discharge to the land. Lil Abner, Al Capp's famous comic strip
character, probably would say, "As any fool can plainly see, it's got to
go some place." And it must. Under the circumstances, we people have
some hard choices to make. These choices concern beneficial use, risk
analysis and priorities associated with our knowledge of and
appreciation for the interrelationship of air, land, and water in our
total environment. We must decide what waste will be tolerated
in the environment and in what amounts. We must determine under
what circumstances it is better for us to deposit our waste products
on the land, in the air or in the water. We must recognize that in the
disposal of wastes, convenience, expediency, and low costs may be
shortcuts to environmental deterioration and endangerment of human
lives. We must accept the fact that knowledge and understanding of
the forces of the environment influence the sum total of our actions as
a Nation and are essential to the protection of our people and the
preservation of the earth.
I would not want to leave the impression that no waste materials
can or should be placed on, in, or under the ground surface. Given the
proper hydrogeological conditions and using appropriately designed
-------
facilities, there are situations when selected wastes can be disposed of
into the ground without appreciably modifying the quality of the
potable ground water. Disposal of brine from petroleum exploration
and development is one example. Another situation may require the
use of a nonpotable water to enhance the quality of an unusable supply
or to protect the quality of a potable supply, as in the case of
agricultural production or salt-water intrusions, when the required
results warrant the implied risk. Further, a decision could be made that
an underground aquifer's most beneficial use is for the disposal of
waste products. In each of these situations, the choice is deliberate. It is
made with full knowledge of its subsequent effects, either actual or
potential. When we proceed in this manner, we bring order to our
actions and our environmental control reguations can then be designed
to have both direction and purpose. To do less is to risk repeating the
history of pollution that is so indelibly associated with our surface
waters and our air.
The results of your deliberations during the course of this
Symposium can make a substantial contribution to the Nation's quest
for knowledge and in the shaping of decisions which will guide future
activity in this subject area. I wish you well as you undertake the
earnest debate that will shape your sessions, as you attempt to resolve
honest differences of opinion that are sure to arise, and as you labor to
reach the common concensus which undergirds our democratic process.
References
Gerba, C. P., C. Wallis, and J. L. Melnick. 1975. Viruses in water: the problem,
some solutions. Environmental Science and Technology, v. 9, no. 13, pp.
1122-1126, December.
Okun, D. A. 1976. Drinking water for the future. American Journal of Public
Health, v. 66, no. 7, pp. 639-643, July.
Safe Drinking Water Act. 1974. Report of the Committee on Interstate and Foreign
Commerce. House of Representatives, U.S. Congress. Report No. 93-1185,
July-
Shiffman, A. 1975. Disposal of hazardous and industrial wastes in Maryland. Report
to the Legislative Council, Maryland General Assembly. Maryland Department
of Natural Resources, Water Resources Administration, November.
U.S. Environmental Protection Agency. 1974. Polluted ground water: a review of
the significant literature. Environmental Monitoring Series, EPA-600 4-74-001,
March.
U.S. Environmental Protection Agency. 1975. Evaluation of land application
systems. Technical Bulletin EPA-430/9-75-001, March.
-------
Thomas P. Ahrens — Father of
Water Well Technology1
by Michael D. Campbellb
On May 1, 1976, Tom Ahrens died of lung
cancer at the age of 69 after a long illness. In my
opinion, no worker in the field of ground water
contributed more to the practical aspects of water
well design, construction and ground-water
development than Tom Ahrens. His absence is
keenly felt by all who knew him—his family and
friends, as well as his professional associates both
in the United States and overseas, where most of
his efforts were directed after his retirement from
the U.S. Bureau of Reclamation in 1972.
Tom Ahrens was born on February 14, 1907,
in Chicago. His father died when Tom was 4 years
old, leaving his mother to raise Tom, two older
brothers and an older sister. In his early teens, the
family moved to White Plains, New York where
his mother was a registered nurse, subsequently
becoming Director of Nursing at the hospital in
White Plains. Tom's childhood and teens were happy
years in a closeknit family where all the children
worked after school to help support the family.
Tom put himself through Georgetown University
in Washington, D.C., where he met and married
Arietta Phillips in the early 30's. He went on to
Harvard for a Master's degree in Geology, while
Arietta, being multilingual, worked for a number of
embassies in Washington, D.C.
It was at Harvard that Tom's interest in
ground water first developed, and it continued to
grow later on during courses at Northwestern and
other universities that offered engineering and
ground-water-related topics. But in those days, the
field of ground water had not yet received the
attention it was to enjoy in later years. In the early
30's, only a handful of professionals were familiar
with the rudimentary concepts of ground-water
flow and even fewer knew of the applications to
aMemorial address presented at The Third National
Ground Water Quality Symposium, Las Vegas, Nevada,
September 15-17, 1976.
t>Senior Consulting Geologist, Department of
Geology, Rice University, Houston, Texas 77001.
water well design and operation. As late as the mid
30's, C. F Tolman at Stanford had not yet com-
pleted his monumental text Ground Water, C. V
Theis was still working on the significance and
nature of the cone of depression, and Morris
Muskat, J. King Hubbert and C. E. Jacob had yet
to make their individual contributions to the field.
Armed with the works of pioneers like Darcy,
Dupuit, Meinzer, and others, Tom was one of the
first to apply the principles of ground water to the
real world, wherever water supplies were needed. It
was at this stage of Tom's life that his wife, Arietta,
played such a significant role in supporting Tom's
attitude toward his professional goal ... to help
people, wherever they were, get water where it was
urgently needed for human existence and
development. They were prepared to go wherever
the need was greatest, which at times turned out to
be in the most far-off and desolate parts of the
earth.
In 1936 Tom joined the U.S. Department of
Agriculture as a water planning analyst and was
located first in Washington, D.C., and then Mesilla,
New Mexico and Denver, Colorado. With the onset
of World War II, he was called back to Washington
to work for the Military Geology Branch of the
U.S. Geological Survey. In 1945, Tom went to
work for the U.S. Bureau of Reclamation where, as
he once expressed it, "he could deal more directly
with the critical field problems of water well
construction and operation than with any other
outfit." During this same period, Tom helped to
organize the National Water Well Association and
his indelible impact can still be felt. He was the
first professional geologist to stress the point that
technical people must learn to serve as technical
translators of the principles of ground-water
geology for the water well contractors, if they
are to be expected to do their jobs effectively.
His wife, Arietta, continued her enthusiastic
support, being a dynamo of enthusiasm in NWWA
activities. Arietta died in 1957, leaving Tom and
an adopted son, Pat. In the years before her death,
during their foreign assignments, she and Tom were
-------
avid collectors of South American art, acquiring
numerous articles of Indian dress and ceremonial
objects. They donated most of the pieces to the
Denver Museum of Natural History, where they
are still on display today.
A few years later, Tom met and married
Helen Seep, an occupational therapist with degrees
from Santa Fe State and Columbia University.
Having also recently lost a special spouse, Helen
found unusual compatibility and comfort in the
relationship. Tom had found a soul-mate. After my
first extended visit with Tom and Helen in Denver,
it was very apparent that their honeymoon had
never really ended.
By the time Tom retired from the Bureau in
1972, he and Helen were seasoned world
travelers who had represented the Bureau and the
United States with a dignity and honor that did
justice to our highest diplomatic ideals. In their
own country as well, they were gracious hosts to a
constant stream of associates and friends from all
over the world. Theirs was an open-door policy of
true hospitality. Guests were always welcome in
their Denver home—drilling contractors, geologists
and engineers alike—as those of you who knew them
can attest.
In March of 1975, during one of his many
post-Bureau overseas consulting assignments, Tom
contracted pneumonia in Pakistan, which began a
long period of decaying health. He refused to
give up his work commitments, however, and a few
months later, in May, he took on a series of lectures
at the Indian Reservation at Window Rock, Arizona
where a serious attack brought his active contribu-
tion to an end. Cancer, surgery, radiation treatment
and chemotherapy occupied the painful 12 months
to follow. In spite of all, he maintained his wonder-
ful sense of humor, his courage and his remarkable
patience to the very end.
Although Tom produced numerous reports
and publications, most of his energies were directed
toward the technical and personal development of
many of his younger associates, one of whom was
myself. His philosophy was that of a practical and
far-sighted man of action. One of Tom's beliefs
was that an error-free individual is an unproductive
individual, and that without mistakes real progress
cannot be achieved, either professionally or
personally. Tom felt that many scientists,
especially earth scientists, are so restricted in their
thinking that many of their careers are spent
avoiding mistakes rather than accepting their
inevitability and profiting from them. He believed
that all individuals must be innovative to be
productive and that the fear of failure must not
paralyze the will to achieve.
Tom gave his time, his thoughts and his
energies unselfishly. He was the guiding light for
many of the projects and programs in the ground-
water field, especially within the U.S. Bureau of
Reclamation and the NWWA. It was Tom Ahrens
who began the dialogue between the well driller
and the technical man. It was he who urged, years
ago, an engineering approach to the design of
water wells, especially in cementing and grouting
and casing selection. And it was Tom who quietly
pressed for effective water well construction
standards and efficient rural water systems.
Tom was one of the early ground-water
salesmen, but where most good salesmen tend to
exert uncomfortable pressure, he was low-key and
gentle, but at the same time would put forth a
strong case for his beliefs. He preferred to be part
of a crowd until something important to him would
stir him to speak up. The confidence, experience
and well-considered position that would come
forth were always rewarding to the listener.
Tom was a Registered Professional Engineer
in the State of Colorado, a member of the American
Geophysical Union, National Water Well Association
(Honorary Life Member), American Water Resources
Association, Geological Society of America, Rocky
Mountain Society of Geologists, American Society
of Professional Engineers, and chief consultant to
the NWWA Research Facility since 1972. He
published ground-water reports on Somalia, East
Africa, Chile, India, and gave numerous
presentations to societies, commissions and
associations. He was awarded a special service
citation by the U.S. Department of the Interior
and numerous other merit awards by scientific
and industrial groups.
Tom's contribution to his field will be felt
for many years to come. He maintained a natural
comradeship with most everyone. He placed
integrity, personal as well as scientific, above all
else. The field of ground water could not afford
to lose him and those who knew him well will long
miss him. He was, indeed, the father of water well
technology, and it is fitting and proper that this
Third National Ground Water Quality Symposium
is dedicated in the memory of Thomas P. Ahrens
to record our collective appreciation of his
professional contribution to the field of ground
water and our admiration of his personal
contribution to the many people who were guided
toward a better understanding of our science and
of ourselves.
10
-------
Wastewaters in the Vadose Zone of Arid
Regions: Hydrologic Interactions"
by John F. Mann, Jr.
ABSTRACT
In truly arid regions there is essentially no direct
penetration of rainfall. Recharge to ground water is not
only infrequent, but extremely localized, occurring only
where surface runoff has been channelized or ponded. Over
those vast desert areas covered by sparse xerophytic
vegetation, the scant rainfall has little or no chance of
becoming ground-water recharge. Such water is quickly
dissipated by capillarity-assisted evaporation, or through
rapid evapotranspiration by short-lived annuals. Where
perennial xerophytes cover the ground surface, the
extensive shallow root systems quickly utilize all of the
rainfall stored in the soil. Beneath the infrequently
moistened soil zone is the lower part of the vadose zone,
extending to water tables which are usually at depths of
tens to hundreds of feet. Almost always these vadose zones
have moisture contents well below field capacity.
Regardless of the cause, these dry vadose zones are
capable of holding additional water, at least up to field
capacity. And no water-carried pollutants can reach the
water table from the ground surface until a pre-wetted
path has been formed for the entire vertical distance. A
practical use of this water-holding capacity can be made
in the design of wastewater tailings ponds, with predictable
safety and with great economic benefit. However, only in
predictable geologic conditions, and in limited amounts can
the use of this water-holding capacity be recommended.
On the other hand, to make no use whatsoever of these
great natural dry sponges would be an economic waste.
INTRODUCTION
The term "vadose water" was originally used
by Posep.ny (1894) to include all of the water in the
zone of aeration. Later, Meinzer (1939) divided the
zone of aeration into three parts and proposed a
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
Consulting Geologist and Hydrologist, 945 Reposado
Drive, La Habra, California 90631.
series of terms of Greek origin to apply to the
contained waters. For all of the water in the zone
of aeration he proposed the term "kremastic water."
He proposed "rhizic water" for the water in the
soil zone, "anastatic water" for the water in the
capillary fringe, and "argic water" for the water in
the intermediate zone. Unlike "phreatophyte"
which has become a commonly used term among
hydrologists, Meinzer's adaptations from the
Greek for waters in the zone of aeration have never
caught on, and few ground-water geologists normally
use them. As regards research activities, the inter-
mediate portion of the zone of aeration must be
considered a true orphan (Bean et al., 1967). The
surficial materials to the depth of root penetration
(soil, hydrologically speaking) have been the
subject of intensive research by soil scientists.
Some agriculture-oriented hydrologists such as
Blaney (1933) have systematically calculated deep
penetration of rainfall primarily from a soil-
moisture approach. The water which is not disposed
of by evapotranspiration or stored in the soil zone
is presumed to reach the water table. The soils
engineer, to whom soil may be anything which is
not hard rock, is concerned mainly with strength
characteristics, but does collect samples of the
intermediate portion of the vadose zone and makes
moisture determinations. Actually, a great deal of
soil moisture information has been collected by
soils engineers in arid regions, but remains in their
files because the information was not relevant to
anticipated problems. Thus we find the vadose zone
of arid regions an inadvertent victim of non-
emphasis, with considerable data existing, but with
few organized attempts to study and understand its
hydrologic aspects. From laboratory studies.
Palmquist and Johnson (1962) have demonstrated
the mechanism of movement of water in vadose
zones, and suggested the possible use of arid
11
-------
vadose zones for nuclear waste disposal. Winograd
(1974) raised the possibility of storing high-level
radioactive wastes within containers in the thick
unsaturated zones of arid regions.
THE VADOSE ZONE IN ARID REGIONS
It is generally recognized that there is no
direct recharge by rainfall through the soil zones
of arid regions. Xerophytic vegetation, by defini-
tion, grows in an environment of sparse soil
moisture. Following a rainstorm, such vegetation
quickly removes the water stored in the soil until
the plant's capacity to remove water is equalled
by the soil's tenacity for holding water. In this
environment of great competition for soil
moisture, the perennial xerophytes are typically
widely spaced and shallow rooted. In addition,
short-lived annuals quickly remove moisture from
the shallow soils in places where they can compete
successfully against capillarity-nourished evapora-
tion and the far-flung root systems of the
perennial xerophytes. Many xerophytes are capable
of utilizing more water than reaches them as direct
rainfall. Examples can be seen in the Mojave
Desert of California where the most abundant
xerophyte is creosote bush (Larrea divaricata).
Individual plants growing close to paved roads are
commonly two to three times as large as those
growing on the normal alluvial surfaces. The paved
roads typically have shallow drainage ditches on
both sides into which is directed the rain which
falls on the impervious pavement. Those creosote
bushes whose roots are able to reach the
augmented rainfall in the soil below the ditches
grow with unusual vigor. As yet, no research has
been done on the question of whether there is
penetration of rainfall beyond the rooting depth
in these ditches. Quantitatively, such recharge to
ground water, if any, would be small.
Studies of recharge in arid regions, particularly
by the United States Geological Survey in Arizona
and Nevada show that the main recharge to the
alluvial basins occurs from the beds of flowing
streams within a short distance after those streams
have left the mountains. Winograd (1974) has
provided us with the following summary:
"Hydrologic, geomorphic, and pedologic evi-
dence suggest that little or no recharge (namely,
infiltration of precipitation to the water table)
occurs beneath interfluves (inter-stream areas)
in the arid and semiarid portions of the
Southwest under present climatic conditions."
In truly arid regions the above generalization seems
to hold, but as pointed out by the writer previous-
ly (Mann, 1957) there is considerable penetration
of rainfall (especially in very wet years) in
semiarid regions. As average annual rainfall
increases, a point should be reached where direct
penetration of rainfall may be expected to start.
Many other variables such as the permeability of
the soil, angle of slope, and nature of vegetation
would be involved. In a study of the Mojave River
basins in California, Chun and Weber (1967) used
an average annual rainfall of 8 inches as the upper
limit of the zone of no direct rainfall penetration.
The same figure of 8 inches was used by Malmberg
(1967) in studies in Pahrump Valley.
In those areas where there has been no
downward passage of infiltrating rainfall for
unknown thousands of years—perhaps since the
ending of the last Pluvial period—moisture levels
have dropped far below field capacity. This condi-
tion has been confirmed in numerous borings
supervised by the writer in the Mojave Desert area.
Where air rotary rigs are used, these dry vadose
zones are characterized by lost circulation problems
and by dust blowing out of the hole. The
mechanism by which the vadose zone drops from
the field capacity condition it is assumed to. have
had at the time of the last infiltrating water to its
present low-moisture condition is poorly under-
stood. Evaporation by air movement accompanying
changes of barometric pressure is probably a
factor. The production of hydrated weathering
products is another. Minor, but perhaps reversible,
changes may be due to vapor transfer.
The depth of the dry vadose zone is highly
variable. Winograd (1974) speaks of depths of 100
to 300 meters in local areas. Regardless of the
depth, there are in places thick zones with moisture
levels far below field capacity. And in considering
the applications of liquids to interfluve soils we may
liken these vadose zones to huge dry sponges.
Before liquids can reach the water table they must
travel a path which has been pre-wetted to field
capacity. The volumetric pellicular film capacity
which can be suspended above the water table
could be viewed as a permanent storage reservoir
for certain liquid wastes.
MOVEMENT OF WATER IN
DRY VADOSE ZONES
Some interesting field tests bearing on the
mechanism of movement of water in dry vadose
zones were conducted during 1970 at the site of a
coal-fired generating plant in southern Nevada. One
of the wastes to be disposed of was coal ash. Prior
to the start of operation of the plant, a question
was raised as to whether rain, falling upon ash
12
-------
.PROPERTY LINE
Scale In feet
01234 5000
INFILTRATION TEST PLOTS
(No Scale)
Fig. 1. Locations of infiltration test plots near coal-fired
generating plant (after Glenn A. Brown).
which had been spread on the ground surface,
would cause pollution of the underlying ground
waters, which in turn were thought to be tributary
to the Colorado River. It was known from labora-
tory tests that ash leachate had a total dissolved
solids content in the range of 3000 to more than
5000 parts per million. Prior to disposal of the ash,
three infiltration test plots were constructed
(Figure 1). Each plot was about 10 feet by 10 feet
(3 m X 3m) and about 18 inches (0.5 m) high. At
each plot three holes were drilled and equipped
with 1 Ys-inch (4.13 cm) shelby tubes. The central
holes were 40 feet deep, and plywood sheets were
placed around the holes to prevent disturbance to
the natural soil by the drilling rig. At each plot 2
holes were drilled in the berms to depths of 25
feet and also equipped with 15/8-inch (4.13 cm)
shelby tubes.
The ash disposal area is in a dissected alluvial
surface extending easterly from the Newberry
Mountains. Several generations of terrace deposits
overlie an extensive sequence of silty and clayey
lake beds.
Prior to the flooding of the test plots, the
natural soil moisture was measured in each hole
by means of a Nuclear-Chicago Model P19
moisture probe. Such neutron-scattering moisture
meters have been used by soil scientists for many
years (Nixon and Lawless, 1960) and have been
confirmed as acceptable substitutes for the older,
more laborious methods of determining soil
moisture (de Boodt, Moerman and de Boever,
1969). Both the neutron probe measurements and
standard laboratory tests on core samples indicated
that the moisture conditions in the vadose zone,
above the water table at a depth of about 85 feet
(26 m), were considerably below field capacity.
More specific results are shown for Hole A-l
(Figure 2). The materials encountered in the
entire 40 feet of hole consisted of fine to coarse
sand with 10 to 20 percent gravel. A 6-inch (15
cm) plug of concrete was poured around the tube
to prevent direct entry of water down the outside
of the tube. Water was introduced to the plot once,
to a depth of 5.25 inches (13.3 cm). Its downward
progress was then monitored by means of the
neutron probe. On the first day, the top 4 feet
(1.2 m) were nearly saturated. By the second day,
some of the water from the top 4 feet (1.2 m) had
moved down to 10 feet (3 m). There appears to be
a translocation of gravity water from the shallow
zones and a downward progression of a pellicular
front in the deeper zones. The rate of reduction of
PERCENT MOISTURE
5 IO 15 20
5/31/70
INITIAL MOISTURE CONTENT
Fig. 2. Progress of wetting front in Hole A-1 (after Buena
Engineering).
13
-------
PERCENT MOISTURE
5 10 15 20
25
o
m
13
5 -
10 -
INITIAL MOISTURE CONTENT
Fig. 3. Changes of moisture content in Hole A-2 (after
Buena Engineering).
moisture content in the shallow zones slows with
time as there is downward drainage. Almost none
of the applied water reached to depths greater than
12 feet (3.7 m), below which the natural moisture
conditions prevailed. From this diagram, in a
relatively homogeneous geologic situation, one
could suggest that the field capacity was 10 or 11
percent, but that factors operating in the arid
vadose zone had reduced the moisture content to
about 5 percent.
The curves for Hole A-2 (Figure 3) show a
vastly different picture. This shows some lateral
movement beneath the berm, but only to a depth
of 7 feet (2.1 m). The low percentages of
moisture suggest that capillarity, rather than
gravity saturation, is the dominant wetting
mechanism.
A comparison of Figures 2 and 3 indicate a
downward vertical movement of the water, with
only minor lateral movement. In this relatively
homogeneous geologic material, such a movement
would be expected from the laboratory experi-
ments of Palmquist and Johnson (1962). Departing
from this homogeneous and permeable situation, it
is possible to predict the effects of certain types
of geologic heterogeneities. If there are nearly
horizontal low-permeability layers, perching is to
be expected, along with greater lateral spread of
the saturated zones. Such low-permeability layers
will not only slow the downward rate of movement
but there is a strong possibility that a greater
percentage of water will be required to bring the
lower-permeability materials up to field capacity.
DISPOSAL OF WASTES OF
DRY VADOSE ZONES
The capacity of dry vadose zones to tie up
liquid wastes as pellicular films and wedges can
be utilized to great economic advantage and with
predictable safety. Figure 4 illustrates the simpli-
fied concept of controlled leakage. At the
prospective site of a tailings pond for liquid wastes
which are not chemically acceptable for direct
recharge to the water table, a drilling and
coring program would have to be undertaken to
determine natural geologic and moisture conditions.
From this information the amount of leakage
permitted could be calculated, of course with an
adequate safety factor.
Industrial tailings ponds are often of huge
dimensions—hundreds of acres in extent. The costs
of low permeability liners are enormous. Within
this economic framework, there is justification for
TAILINGS POND
DIK
DIKE
WATER TABLE
Fig. 4. Diagrammatic distribution of moisture beneath a
tailings pond under a controlled leakage approach.
14
-------
considerable testing and experimentation based
on sound geologic and hydrologic principles.
In the Mojave Desert of California, studies
were made for the installation of a 129-acre
tailings pond for the evaporation of highly saline
wastes from a large chemical plant. Local clays
were available to construct a compacted low
permeability liner. Laboratory tests indicated that
a permeability of 0.002 feet (0.06 cm) per year
could be achieved. The clay layer (compacted in
two lifts and 0.83 feet [25 cm] thick) would be
covered by a permeable protective blanket to avoid
desiccation of the clay layer if the ponds were
emptied. Pre-flooding with fresh water was
planned to provide a reservoir of fresh water
within the protective blanket which could saturate
the clay and negate capillary forces within the
clay. The saline tailings water would then follow
the fresh water through the clay. Under a "zero
leakage" concept there is no provision for pellicular
storage in the underlying dry vadose zone; on the
contrary, the useful life of the pond is considered
to end when the tailings water reaches the
bottom of the base (clay) layer. Pond life as a
function of total pond head, using a design
permeability of 0.002 feet (0.06 cm) per year and
an effective porosity of 0.33 is shown in Figure 5
for different thicknesses of the impervious clay
layer. A pond built with a clay layer 0.83 feet
(25 cm) thick and a head of 10 feet (3 m) is
expected to have a useful life of about 12 years,
at which time a new tailings pond would have to
be completed and ready to receive tailings water.
However, by adopting a controlled leakage
approach, the useful life of the pond could be
increased. As this pond receives solid wastes (slurry)
as well as liquids, it will progressively lose its
capacity for seasonal storage of liquids. Its useful
life could be extended further if the dikes could
be raised to permit a higher theoretical head.
Actually, during the deposition of the water-
carried solids, there are laid down nearly
horizontal clay layers, so that the effective head
operating on the base layer may be less than would
be indicated by Figure 5.
Under the design criteria used, Figure 6
shows the relationships among total pond head and
base (clay) thickness as it relates to the rate at
which water will move through the base (clay)
layer. Using such data, along with information as
to moisture content of the underlying vadose zone,
it is possible to predict the downward movement
of the wetting front and to stop it at a safe distance
above the water table.
\
BASE!THICKNESS\i
Tfeet)
™
!T
50
40
30
20
UJ
I
10 15 20 30 40 60 80 100
POND LIFE (years)
Fig. 5. Relationships among pond liner thickness, total
pond head, and pond life under a zero leakage concept
(after Moore and Taber).
EVAPORATIVE POND TRANSMISSION RATE
.02 .04 .06 .08 .10 .12
UNIT TRANSMISSION RATE
(cubic f eet/yeor/square foot)
Fig. 6. Rate of leakage through clay layers of various
thicknesses at various total pond heads (after Moore and
Taber).
15
-------
MONITORING
The key to acceptance of a controlled leakage
approach is monitoring. Classical monitoring of the
water table downgradient from the tailings pond
may be unacceptable if such monitoring results in
the detection of a pollution problem only after it
has reached unmanageable proportions. The facts
demonstrating direct downward vertical movement
suggest monitoring wells within the pond itself, or
such monitoring of perched lateral flow as would
be indicated by the geologic conditions. With such
wells and a neutron probe, the downward
movement of the wetting front could be followed,
and discharge of the effluent could be stopped
where desired.
ACKNOWLEDGEMENTS
For making available basic data used in this
paper, and for many helpful suggestions, the writer
would like to thank Mr. Glenn A. Brown, Consult-
ing Geologist, Los Angeles; Mr. R. F. Moore, of
Moore and Taber, Consulting Engineers and
Geologists, Anaheim, California; and Buena
Engineering, Ventura, California. However, neither
their cooperation nor helpfulness should be
construed as agreement with all of the writer's
conclusions, or any of them.
REFERENCES
Bean, R. T. et al. 1967. Methods and techniques of ground-
water investigations. UNESCO Water Resources
Series, no. 33, pp. 63-66.
Blaney, H. F. 1933. Rainfall penetration in Ventura County
investigation. California Division of Water Resources.
Bull. no. 46, pp. 82-90.
Chun, R.Y.D. and E. M. Weber. 1967. Mojave River
ground-water basins investigation. California Depart-
ment of Water Resources. Bull. no. 84, p. 38.
de Boodt, M., P. Moerman, and J. de Boever. 1969.
Comparative study of the water balance in the
aerated zones with radioactivity methods and
weighable lysimeter. Water in the Unsaturated Zone.
Proceedings of the Wagenigen Symposium. UNESCO-
IASH Joint Publication.
Isaacson, R. E., L. E. Brownell, R. W. Nelson and E. L.
Roetman. 1974. Soil moisture transport in arid site
vadose zones. Isotope Techniques in Groundwater
Hydrology. International Atomic Energy Agency,
Vienna.
Malmberg, G. T. 1967. Hydrology of the valley-fill and
carbonate rock reservoirs, Pahrump Valley, Nevada-
California. U.S. Geol. Survey Water-Supply Paper
1832.
Mann, J. F., Jr. 1957. The significance of the sub-alluvial
outcrop in arid and sub-arid regions. XX International
Geological Congress. Section IV, pp. 55-73.
Meinzer, O. E. 1939. Discussion of question no. 2 of the
International Commission on Subterranean Waters.
Definitions of the different kinds of subterranean
water. Trans. Am. Geophys. Union, part 4, pp. 674-
677.
Nixon, P. R. and G. P. Lawless. 1960. Translocations of
moisture with time in unsaturated soil profiles.
Journal of Geophysical Research, v. 65, no. 2, pp.
655-667.
Palmquist, W. N., Jr. and A. I. Johnson. 1962. Vadose
flow in layered and nonlayered materials. U.S. Geol.
Survey Prof. Paper 450-C. pp. C142-C143.
Posepny, F. 1894. The genesis of ore deposits. Trans. Am.
Inst. Min. Engrs. v. 23, p. 213.
Winograd, I. J. 1974. Radioactive waste storage in the arid
zone. EOS, Trans. Am. Geophys. Union, v. 55, no. 10,
pp. 884-894.
DISCUSSION
The following questions were answered by John F. Mann,
Jr., after delivering his talk entitled "Wastewaters in the
Vadose Zone of Arid Regions: Hydrologic Interactions."
Q. by G. F. Hendricks. Please redefine the term "field
capacity. " How is it determined?
A. Field capacity describes the condition of a soil after
gravity drainage. It is a term most often used in agricultural
circles, and similar to specific retention.
Q. by Ron Barto. Do you think that the low field capacity
is due to evaporation?
A. Evaporation is a possibility. On a longer-term basis,
hydration as feldspars are weathered to clays, is another.
Q. In the case of the Nevada Ash Pit, at about what level of
natural rainfall would the vadose zone reach field capacity?
A. There are some suggestions that average annual rainfall
as little as 8 inches per year would result in some deep
penetration of rainfall on coarse permeable soils. Tighter
soils and hot season rainfall would probably place this
value higher.
Q. by Jack E. Sceva. How do you evaluate the potential for
the development of perched ground-water zones and the
lateral movement of leakage away from the site?
A. By thorough coring of the site and determination of
permeabilities. Low permeability layers may be expected
to cause perching and lateral spread.
16
-------
Q. by K. Childs. Shouldn't the integrity of the suggested
leaking pond be verified by a ground water and soil zone
monitor network, or don't you feel it is necessary?
A. I would suggest both neutron probe monitoring within
the pond and water-table monitoring downgradient.
Q. by Dan Kimball./lrf there any ground-water observation
wells located around the ash pond at the Mojave plant and,
if so, have they shown any changes in ground-water quality?
A. No changes in ground-water quality in the monitor wells
have been related to the disposal of ash.
Q. by Bill Bellis. With all the core samples - what is the
mineralogy?
A. There has been visual logging of the cores and cuttings
but no detailed mineralogical study.
Q. by Ed Meiser. What is the danger or likelihood of
localized, presetted paths developing as routes of direct
communication between the source of wastewater
application and the water table?
A. In usual alluvial environments, this should be no
problem. Concern should be expressed in areas of cracking
due to compaction or tectonics. A thorough subsurface
investigation must be made.
Q. by Jan Turk. In your "no leakage " example, fresh
water was used to saturate the clay lining before the highly
saline waters were allowed to pass through the clay. Have
you noticed any shrinkage of the clay when the brine
arrives (i.e. osmotic desiccation)?
A. The brine is discharged as a slurry, which causes the
bottom to be covered by deposited clays. Furthermore, the
clay is covered by a permeable layer, pre-wetted with fresh
water. Under these conditions, desiccation is unlikely.
Q. by Jack Robertson. Does the clay lining of ponds produce
any ion filtration effects?
A. No liquid has shown in the underdrains, so it hasn't been
possible to check this.
Q. by D. G. Boyer. What is your type of clay? Bentonite or
non-expanding clay? What is the concentration and make-up
of your saline fluid?
A. The clay is not highly expansive. The saline fluid is a
high sodium brine with 50,000 to 100,000 parts per million
total dissolved solids.
Q. by C. W. Fetter. Homo sapiens have existed, so far,
through a number of climatic changes. During the late
Pleistocene, the "arid"southwest had heavy precipitation.
Does waste disposal in the arid zone present potential
problems to future generations if the climate shifts to a
humid one? Long-forgotten contaminants might become
mobile.
A. The possibilities of a climatic change to wetter (or drier)
over a span of thousands (rather than hundreds) of years are
a consideration. Covering with a low permeability blanket
has been suggested as a means of preventing future leaching
by infiltrating rainfall or flood flows.
17
-------
Wastewaters in the Vadose Zone of Arid
Regions: Geochemical Interactions1
by Donald D. Runnells
ABSTRACT
Because of increasingly stringent laws governing
discharge of fluid wastes to surface waters, the alternative
of discharge to the subsurface has become attractive. The
physical-chemical processes that prevail in the subsurface
are not well understood, but they are clearly not identical
to processes of purification in surface waters. For example,
in the subsurface the process of oxidation may be of little
value in significantly reducing the concentration of
discharged contaminants; in contrast, oxidation plays an
important role in purifying surface waters. Eleven physical-
chemical processes can be identified as having potential
value for purifying wastes discharged to the subsurface, as
follow: dilution, buffering of pH, precipitation by reaction,
hydrolysis, oxidation or reduction, filtration, volatilization,
biological assimilation, radioactive decay, membrane
filtration, and sorption.
Discharge to the vadose zone may be a safe means
of disposal of wastes in arid regions. But it is necessary to
carefully test the suitability of a particular site for a
particular waste. Processes of purification in the vadose
zone can be incorporated into a workable plan of discharge
if adequate studies and safeguards are employed. Regulations
governing subsurface discharge should take into account
the physical-chemical processes that may act to purify the
waste fluids. In one set of experiments, a soil from Sulfur
Springs, New Mexico was capable of removing large
quantities of dissolved molybdenum and copper from
a synthetic mill water, and the soil was able to quantitatively
retain the copper during subsequent leaching by fresh and
metal-free mill waters. Such studies permit rational plans of
discharge to be developed.
INTRODUCTION
Until fairly recent times we have generally
assumed that ground water was well protected from
contamination. And in many areas this seemed to
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
bprofessor of Geology and Geochemistry, Department
of Geological Sciences, University of Colorado, Boulder,
Colorado 80309.
be a fairly realistic view. Now, however, we know
that the apparent purity of ground water was the
result of our failure to look for pollution. With
increasing pressures of population, and with
incessant prodding from new regulations, water
scientists have begun to fully recognize the
widespread occurrences of contamination of ground
water. Table 1 shows chemical components that
have been demonstrated as being present in
documented cases of contamination of ground
water throughout the world (summarized from
several sources, including Furiman and Barton,
1971; Scalf and others, 1973; Miller and others,
1974; Cole, 1974; van der Leeden and others,
1975). Each of these cases demonstrates that the
protective capacity of subsurface materials is finite,
and that this protective barrier can be overwhelmed
by excessive discharge of virtually any contaminant.
The means by which contaminants may enter
ground water are many and varied. Figures 1 and 2
illustrate some of the more obvious sources and
pathways of contamination of ground water. The
details of the hydrology of movement of
Table 1. Chemical Components Present in Documented
Cases of Contamination of Ground Water
Throughout the World
chromium cyanide
cadmium copper
zinc selenium
lead chloride (and hydrochloric acid)
fluoride sulfate (and sulfuric acid)
iron nitrate
barium detergents
manganese radium and other radioactive wastes
nickel phenols
silver alcohol
molybdenum gasoline
boron leachate from landfills
uranium ' pesticides
mercury herbicides
aluminum solvents
lithium
18
-------
POSSIBLE SOURCES OF GROUNDWATER POLLUTIO
SPREADERS
OR SPRAYERS
DUSTS.
SALTS.
SLUDGES
SUBSURFACE
STORAGE
DISPOSAL
WELLS
WATER
"TABLE
DUMPS IN
GROUNDWATER
GROUNDWATER FLOW
Fig. 1. Summary of some obvious pathways by which contaminants may enter ground water.
POSSIBLE SOURCES OF GROUNDWATER POLLUTION
PONDS. LAGOONS.
LEAKING TANKS
WASTES ON
SURFACE
f BURIED
; SOLIDS OR
I SLUDGES
ATER TABLE
OUNDWATER FLOW
Fig. 2. Additional illustrations of possible means by which ground water can be contaminated.
19
-------
contaminants in the subsurface is treated in other
papers in this Symposium and will not be discussed
here.
Despite the fact that we are aware of hundreds
of documented cases of ground-water contamina-
tion, the sobering fact is that there certainly exist
thousands of similar cases of which we remain
blissfully unaware. Miller and others (1974) made
the important observation that most instances of
ground-water contamination become known
because of obvious leaks or spills at the surface, or
because of complaints from subsequent users.
Contamination of ground water is generally not
discovered because of monitoring or routine
chemical analysis. Many States in the past have not
required analyses of water supplies for many toxic
substances, such as barium, selenium, and silver,
despite the listing of such substances in the 1962
Drinking Water Standards of the U.S. Public Health
Service.
POSSIBLE PROCESSES OF PURIFICATION
Introduction
Having now emphasized the fact that ground
water can surely become contaminated, let us look
at some of the processes which offer protection to
ground water, even in the face of significant
discharges of potential contaminants to the
subsurface. In Table 2 are listed eleven physical-
chemical processes that may, under favorable
circumstances, afford significant protection to the
chemical purity of ground water. These eleven
processes are listed together for the sake of
completeness, but all are surely not equally
effective.
The possible role and importance of each of
these processes is discussed briefly in the following
sections. Special emphasis is placed on their role in
purifying wastes that might be discharged to the
vadose zone in arid regions.
Table 2. Eleven Physical-Chemical Processes That May
Operate in the Subsurface to Purify Fluid Wastes
1. Dilution.
2. Buffering of pH.
3. Precipitation by reaction of wastes with indigenous
waters or solids.
4. Precipitation due to hydrolysis.
5. Removal due to oxidation or reduction.
6. Mechanical filtration.
7. Volatilization and loss as a gas.
8. Biological assimilation or degradation.
9. Radioactive decay.
10. Membrane filtration.
11. Sorption.
Dilution
The first process listed in Table 2, dilution, is
well understood as it occurs in surface waters.
However, we cannot rely on dilution to be
effective over the short term in the vadose zone,
especially in an arid region. And in any location, we
must expect dilution of wastes below the water
table to take place far more slowly than in surface
streams. Indeed, the one outstanding characteristic
of contamination of ground water is the slow rate
at which it is diluted and dispersed by the natural
flow system. Fryberger (1972) has presented
convincing data and arguments to show how
expensive and difficult it may be to reclaim a
contaminated aquifer, even considering natural
attenuation due to flow and dilution.
Buffering of pH
Natural ground waters in arid regions generally
exhibit pH values between about 6 and 9. The
upper limit is established in nature by the reaction
between carbon dioxide gas and either limestone or
caliche in the soil. The lower limit is an educated
guess at the pH that might be found in shallow
ground water from non-reactive rocks, such as
quartz sandstone, or in ground water issuing from
a mountain meadow. Trost (1974) found a range in
pH from 6.0 to 9.0 in 1392 samples of ground water
from southern Arizona.
The pH is a critical factor in many reactions
involving contaminants, including processes that
affect the stability of solid minerals and precipitates
in the subsurface. If a discharged fluid has a pH
outside of the range of 6 to 9, it is possible that
drastic chemical reactions may occur, including
the dissolution of such natural solids as calcium
carbonate, iron oxyhydroxide, and manganese
oxyhydroxide. Such reactions can exacerbate
problems of contamination of ground water.
Precipitation by Reaction
This process could be quite important in
purification of discharged fluids. A perusal of any
handbook of chemistry will show that in theory it
is possible to precipitate virtually any dissolved
contaminant if the appropriate precipitating agent
is present. In nature, however, the necessary
precipitants do not commonly exist in sufficient
quantities or favorable locations to cause removal
of toxic substances to safe levels. For example, it
is theoretically possible to reduce the concentration
of dissolved sulfate in a discharge by mixing it with
a natural barium-rich ground water. However,
barium-rich ground waters are not common, so
20
-------
this particular theoretical control will generally
not operate in nature. In natural ground waters and
in the enclosing solids, only certain species are
normally present and available for reaction in
meaningful concentrations. These species include
dissolved calcium, magnesium, sodium, potassium,
bicarbonate, sulfate, chloride, and silica. In the
surrounding porous solids we find the additional
abundant components of iron, aluminum, and
possibly carbonate and phosphate. Dangerous or
undesirable contaminants that might be reduced
to safe concentrations by reaction with these
species include the following: arsenic (precipitation
by iron, aluminum, and calcium), barium (by
sulfate and carbonate), cadmium (by sulfate and
carbonate), copper (by carbonate), fluoride (by
calcium and phosphate), cyanide (by many metals,
including iron), iron (by silica), lead (by carbonate,
sulfate, and phosphate), mercury (reduced form
precipitated by carbonate and chloride),
molybdenum (by iron and aluminum at proper
pH), silver (by chloride), zinc (by carbonate and
silica), and radium (by sulfate and carbonate).
Evidence that such precipitation reactions do occur
and may control concentrations in ground water
has been given by many researchers. For example,
Dutt and others (1968) found that the concentra-
tions of iron and zinc in the ground waters of the
Tucson Basin are apparently controlled by
dissolved silica and the pH. However, despite the
numerous possibilities for purification offered by
precipitation reactions, a great many examples of
contamination of ground water exist involving
species that might be expected to precipitate; such
examples demonstrate the failure of precipitation
to protect ground water in many cases.
Hydrolysis and Precipitation
This type of reaction occurs when a dissolved
contaminant reacts with water, with the release of
either a hydrogen ion or a hydroxyl ion. The best
known example involves the hydrolysis of
dissolved ferric ion with precipitation of ferric
hydroxide. On the acid side, molybdenum is
thought to hydrolyze to bimolybdate ion and
precipitate with compounds of iron and aluminum
under acid conditions (LeGendre and Runnells,
1975). In an aerated environment, such as the
vadose zone in an arid region, the concentration of
dissolved iron can be held to very low values by
hydrolysis and precipitation of ferric oxyhydroxide
(Hem, 1970). However, in oxygen-deficient zones,
such as commonly exist in the phreatic zone, high
concentrations of iron can occur in the ferrous
(Fe2+) form. High rates of infiltration or flooding
by waters into the vadose zone, such as might occur
beneath a disposal lagoon, may produce anaerobic
conditions and lead to solubilization of some
metals.
In addition to ferric (Fe3+) iron, several other
toxic metals form highly insoluble hydroxides in
the range of pH of 6 to 9. These include copper
(above a pH of about 6.5), chromium (the +3 form
only, above a pH of about 6), nickel (above a
pH of 9), and zinc (above pH 7.5). For example,
Dutt and others (1968) found that the
concentration of copper in ground water of the
Tucson Basin seems to be controlled by either
cuprous (Cu+) or cupric (Cu2+) hydroxide.
The apparent importance of hydrolysis in
both the vadose and phreatic zones is the result
of the simple requirement that water be present,
in contrast to other precipitation reactions that
demand the presence of specific precipitating
agents.
Precipitation Due to Oxidation or Reduction
Relatively insoluble oxides of certain
contaminants are known for copper, iron,
manganese, mercury, and nickel. Aerobic
conditions favorable for the development of these
oxides will normally be found in the vadose zone
and possibly in the upper parts of the phreatic
zone. Deep in the phreatic zone or in swampy or
flooded vadose environments, anaerobic conditions
may prevail and lead to the mobilization of some
species. Dutt and others (1968) in their work in
the Tucson Basin found that rapid infiltration of
domestic sewage caused a loss of oxidizing condi-
tions in the vadose zone and led to increased
movement of copper, zinc, manganese, nickel,
and lead through the soil.
In some instances reducing conditions are
most favorable for removing possible contaminants
from water. One example is that of chromium,
which is highly soluble and mobile in the oxidized
state (+6, chromate or dichromate ion), but quite
insoluble as the solid oxide or hydroxide of the
reduced form (+3). Natural reducing conditions
can also theoretically cause the formation of such
native elements as arsenic, copper, mercury,
selenium, silver, and lead, each of which is quite
insoluble. We do find significant quantities of
native copper, mercury, and silver in mineral
deposits, and native selenium appears to be a stable
phase in some soils (Goering and others, 1968). In
reducing environments bacteria can convert
dissolved sulfate to sulfide and dissolved nitrate to
21
-------
ammonia or nitrogen gas. Reaction of the sulfide
with certain metals will produce highly insoluble
precipitates, such as the sulfides of arsenic,
cadmium, copper, iron, lead, mercury,
molybdenum, nickel, silver, and zinc.
In terms of usefulness and reliability for
purification, the processes of oxidation and
reduction can be of significant value if the
necessary environmental conditions are maintained
and managed. Considerable reliance can probably
be placed on the process of oxidation for purifica-
tion of wastes in the vadose zone. Processes of
reduction may be more risky because some
undesirable species can be mobilized, such as iron
(Apgar and Langmuir, 1971) and manganese. One
well-known benefit of the process of reduction
is the denitrification of dissolved nitrate from
sewage effluent, with significant removal of
nitrogen possible under a carefully programmed
mode of disposal (Bouwer, 1974).
Mechanical Filtration
The sixth process listed in Table 2, mechanical
filtration, will generally be of little help in
removing dissolved contaminants. Exceptions to
this might occur following the formation of
precipitates by means of one of the processes
mentioned earlier. Such precipitates would be
filtered out mechanically as the water moves
through the porous medium. Iron is the species
most likely to exhibit this behavior; it is well
established that much of the total iron in aerated
waters is present as p articulate ferric hydroxide
(Hem, 1970). Manganese probably behaves in a
similar fashion. If these possible particulates sorb or
include other dissolved species as they form, the
filtration would remove these other species as well.
Of course, most particulates originally present in the
discharge would be filtered out quickly during
downward movement through the vadose zone.
Volatilization and Loss as a Gas
Some inorganic species can be effectively
removed from discharged waters by volatilization.
Perhaps the best known example is that of the
bacterial reduction of dissolved sulfate to hydrogen
sulfide gas, with loss of the hydrogen sulfide to the
atmosphere (Kellogg and others, 1972). Mercury in
solution can also be volatilized in anaerobic
environments (Lagerwerff, 1972) or by reaction
with dissolved humic acids (Alberts and others,
1974). Several organic compounds of arsenic are
volatile, and escape of arsenic as a gas has been
demonstrated for both aerobic and anaerobic soils
(Woolson and others, 1971). Based on its similarity
to sulfur, we might also expect selenium to be
subject to volatilization (Lakin, 1973). And of
course, the microbial reduction of nitrate to
ammonia and nitrogen gas is well documented
(Bouwer, 1974), although the failure of this
mechanism to protect ground water is demonstrated
by numerous examples of pollution of ground water
by nitrate (Minear and Patterson, 1973).
Volatilization as a means of purification is
poorly understood, and we probably run consider-
able risk in relying upon it to any great extent. The
exceptions to this would be the conversion of
dissolved sulfate and nitrate to gases; these processes
are known to be effective if managed properly.
Biological Degradation
Biological degradation or assimilation, the
eighth process, is very important in removal of
organic and biologic contaminants. Many organic
substances would be removed or attenuated by
biologic activity in the subsurface, especially in
the oxidizing environment of the normal vadose
zone. The biologic involvement of sulfate and
nitrate have already been mentioned. In addition,
arsenic, cyanide, mercury, and selenium are likely
candidates for biologic fixation or volatilization.
Molybdenum is strongly assimilated and concen-
trated by plants that are nitrogen-fixers (Johnson,
1966). And in their study in the Tucson area, Dutt
and others (1968) found that grass assimilated
and removed significant quantities of metals from
infiltrating sewage effluent.
Biological processes could be of great value
in managing discharges to the subsurface, but at
present we know so little of the principles
involved that each case must be studied and
evaluated on its own.
Radioactive Decay
This mechanism is of value in the management
of radioactive wastes by means of storage in the
subsurface. Winograd (1974) has recently discussed
the attractiveness of storage of high-level radio-
active wastes in the vadose zone of arid environ-
ments. It seems clear that storage in this environ-
ment is possible with a high degree of safety
for periods of time from thousands to hundreds of
thousands of years, during which time the wastes
would lose much of their activity through processes
of decay. There is a substantial history of disposal
of radioactive wastes to the subsurface near nuclear
facilities in the United States, and in humid regions
serious problems of contamination of ground
22
-------
water can be expected. It is the lack of recharge
to move contaminants to the ground water that
makes disposal to the vadose zone in arid regions
so attractive.
Membrane Filtration
The tenth possible process of purification,
membrane filtration, is a topic that is greatly in
vogue at the present time in the literature of the
geochemistry of subsurface waters. A summary
of the phenomenon, given by Hanshaw (1972),
points out that some observed data on the pressure
and salinity of formational waters cannot be
explained by simple gravitational flow of water or
by dissolution of minerals. In such cases, the
anomalous data may be related to osmotic pressures
and filtration of salts across beds of shale. Pressure
differentials of several hundred atmospheres can
theoretically be generated across a membrane if
fresh water exists on one side and a saturated brine
on the other (Hanshaw, 1972).
In terms of purification of wastes in the
subsurface, it is theoretically possible that a
discharged wastewater could be diluted by osmotic
transfer of water across a clay-rich aquitard.
Conversely, one can speculate on the possibility
of forcing waste fluids through a clay liner on the
bottom of a disposal pond, anticipating some
removal of dissolved salts and passage of a
relatively more pure fluid. However, at this stage in
development of the technology of disposal,
careful pilot studies would be required to test the
efficacy of such a scheme.
Sorption
Finally, and importantly, ground water may
be protected by various processes of sorption. Clays,
metallic oxides and hydroxides, and organic matter
can all be good substrates for sorption of various
dissolved species. With the exception of fractured
shale or siltstone, consolidated bedrock will
generally not be very effective as a sorbent.
Virtually every ionic species and many non-electro-
lytes will be sorbed and removed to some extent
as ground water moves through an aquifer; of the
ions, only chloride, and to a lesser extent sulfate
and nitrate, seem to pass through soils and
alluvium without significant sorption.
The troubling aspect of sorption as a means of
purifying waste waters in the vadose zone is that
the process can be highly specific in its action,
both as to the dissolved substance and the solid
substrate. Molybdenum offers an interesting
example. Dissolved molybdenum is sorbed strongly
by most soils that exhibit an acidic reaction with
water, but molybdenum will move freely through
alkaline soils (Katz and Runnells, 1974).
Similarly, fluoride is sorbed much more
extensively by acidic soils than by alkaline soils
(Bower and Thatcher, 1967). Perhaps the most
revealing study of the effect of pH on sorption of
ions is that by Griffin and others (1976), in which
they determined the extent of removal by clay
minerals of a suite of heavy metals from synthetic
solutions and natural leachates from sanitary
landfills. They found a very great variation in the
extent of sorption, depending on the metal involved
and the pH-of the solution.
It is clear that the processes of sorption
depend on the type of contaminant and on the
physical and chemical properties of both solution
and porous medium. Under favorable circumstances
sorption can be wonderfully effective in purifying
waters, as evidenced by the extensive use of
ion-exchange media in water treatment. One should
keep in mind, however, that when a contaminant
ion is sorbed, some other change must also take
place to compensate for loss of the charged species
from solution. In ion-exchange reactions, a
different ion is released by the solid to the solution.
The ions released to water from a previously
uncontaminated soil or alluvium will almost
certainly be less harmful than most of the
contaminants of concern here.
We can place considerable reliance on the
processes of sorption to help us protect ground
water, but each case will be different and tests must
be run to assure an adequate degree of sorption
and retention.
Summary
Eleven processes have been discussed, each of
which is capable, under favorable circumstances, of
contributing to the purification of liquid wastes
discharged to the subsurface. However, hundreds
of documented examples of ground-water
contamination show that in specific instances the
defensive mechanisms have been overwhelmed by
discharged contaminants. It is reasonable to make
an attempt to use the purifying properties of the
subsurface as part of a plan of discharge, especially in
the vadose zone in arid regions. However, such
a plan of discharge must incorporate careful studies
of the capacity and efficiency of the vadose zone
for removing contaminants, and of possible later
remobilization of the toxic substances.
23
-------
USING NATURAL PROCESSES OF
PURIFICATION IN A PLAN OF DISCHARGE
Introduction
Because of the potential value for purification
and long-term storage of wastes in the vadose zone
of arid regions, it seems reasonable to include the
use of this zone in plans of discharge. In fact,
Winograd (1974) presents a detailed analysis of
such a plan for storage of radioactive wastes. As
water scientists, we can envision similar useful
applications in the discharge of other wastes.
Bouwer (1974) has given an example of proper
management of disposal of sewage effluent to the
vadose zone in Arizona in order to maximize
removal of dissolved nitrate. At the time of this
writing, the New Mexico Water Quality Control
Commission is considering, for possible adoption,
new regulations for protection of ground water.
These proposed regulations, developed chiefly by
the New Mexico Environmental Improvement
Agency, would permit a discharger to make use of
a fraction of the purifying capacity of the vadose
zone as part of his plan of discharge to the
subsurface. The concept of using the vadose zone
in this manner was suggested by the present writer,
as a consultant to the New Mexico Environmental
Improvement Agency. It is gratifying that most
parties concerned, from industry on the one hand
to environmental groups on the other, seem to
agree that the vadose zone in arid regions can offer
a safe repository for contaminants, provided that
very careful study and monitoring are incorporated
into the plan of discharge.
Potential Capacity of the Vadose Zone
for Purification
The capacity of the vadose zone can be large
for removal and storage of contaminants. As an
example of this large theoretical capacity, let us
consider a hypothetical discharge of dissolved
cadmium from a metal-plating plant. The
assumptions involved in this example are
summarized in Table 3. The concentration of
3.7 mg/1 dissolved cadmium is that actually
reported by Lieber and others (1964) for the
well-known example of ground-water contamination
on Long Island. It would be dangerous to allow a
discharger to use the total capacity of the vadose
zone for sorption of contaminants from his
wastewater, so for purposes of this example we will
assume that State regulations allow him to use 25
percent of the estimated total sorptive capacity of
the unconsolidated mantle beneath his discharge
pond. Based on the data listed in Table 3, the
24
Table 3. Assumptions Used in Hypothetical Example of
the Capacity for Sorptive Removal of Dissolved Cadmium
from Discharge to the Vadose Zone in an Arid Region
Assumptions:
1. Unconsolidated mantle with 10 weight percent clay.
2. Cation exchange capacity of 50 milliequivalents per 100
grams (0.22 pounds) of clay.
3. Density of 2 grams/cc (124.9 Ibs/cubic foot).
4. Thickness of mantle = 100 feet (30.5 meters).
5. Area of disposaL= 1 acre (0.405 hectare).
6. Efficiency of sorption is 100 percent.
7. Only 25 percent of the total sorptive capacity is utilized.
discharger could store approximately 380,000
pounds (172,400 kg) of cadmium in the vadose
zone. This is a lot of cadmium. At a concentration
of 3.7 mg/1, this amount corresponds to about
38,000 acre-feet (4.7 X 1010 liters) of wastewater
that could be discharged from the one-acre pond.
Of course, the preceding simplistic example
ignores competition or enhancement of sorption by
other ions, other chemical reactions, possible
channelized flow of the wastewater through the
vadose zone, and a great many other complicating
factors. The example does point out, however,
that a significant degree of purification of waste-
water and storage of dissolved contaminants is
possible under the conditions that exist in an arid
region.
Dangers and Necessity for Study
Having emphasized the great potential value of
the vadose zone for discharge, we must temper our
enthusiasm by recognizing that the vadose zone
has a finite capacity for purification, and that
there is always a danger of leakage and contamina-
tion of ground water. Referring back to Table 1,
we can see that it is possible to overwhelm the
protective mechanisms in the subsurface with any
dissolved contaminant. This knowledge warns us
of the necessity for thorough and competent
studies of the hydrologic and chemical character-
istics of a proposed site before discharge begins. A
proposed plan of discharge must also allow for a
wide margin of error in the measurements
required, including such parameters as heterogeneity
in the hydraulic properties of the earth materials
at the site. Finally, an adequate scheme for monitor-
ing must be included, together with contingency
plans'in the event that the discharged contaminants
do not behave as expected.
Necessary Measurements and Testing
Our knowledge of the specific physical-
chemical processes that obtain in the subsurface
-------
is primitive. We cannot predict from theory what
will happen when a particular contaminant in a
particular fluid matrix is discharged into a
specific type of soil. We do have some general
knowledge of the principles involved, as discussed
in an earlier section of this paper, but we remain
ignorant of many of the specifics. Now that the
Safe Drinking Water Act is a reality, we can
expect to see a rapid increase in research and
knowledge concerning the movement of con-
taminants through earth materials. In fact, it
would be possible to list several new publications
that deal specifically with this subject. By far
the most complete and satisfying such report
known to this writer is that by Weir and others
(1975), prepared for the Electric Power
Research Institute. The purpose of the work
reported by these workers was to determine if
toxic ions would be leached from ash and sludge
discharged into ponds by electric utilities, and
to determine if such ions would move through
underlying soils into ground water. Another
excellent study, with greater emphasis on the
hydrologic factors, is that by Papadopulos and
Winograd (1974).
In order to have confidence in the probable
efficacy and safety of a plan of discharge to the
vadose zone, a great many factors must be
considered. For example, one must have a
fairly complete knowledge of the hydrogeologic
conditions and homogeneity of the earth materials
at the proposed site of disposal. In order to obtain
such information it will certainly be necessary to
conduct fairly extensive field studies, possibly
including drilling, to determine the rate and paths
of movement of the fluid discharge. An example of
one aspect of the work that may be necessary can
be found in the study of seepage of effluent from
septic tanks, published by Bouma and others
(1972). Hajek (1969) has given a good summary
of some of the technical aspects of the tests and
calculations that should be done to understand
and predict the chemical interactions of waste-
waters with soils. And finally, a potential
discharger must look to the future, realizing that
"purification" of wastewaters by the vadose zone
really represents storage, either long-term or short-
term. For some chemical species, such as phosphate,
zinc, and copper, the storage may be permanent
because these ions can become fixed in the
structure of minerals and not be available for later
release by percolating solutions (Ellis and Knezek,
1972). In other cases, contaminants that have been
discharged and stored in the vadose zone may be
readily available for leaching and remobilization
by the next soaking rainfall, or by a subsequent
change in the chemical composition of the
discharged fluid. Tests must therefore be run, not
only to determine the extent of removal of
contaminants from the discharge, but also to
determine the possibility of later remobilization
and flushing to ground water. With regard to
storage, Winograd (1974) has discussed the
selection of a site for disposal to best avoid
exhumation by the normal processes of erosion or
remobilization by climatic change over periods of
hundreds of thousands of years.
Discharge of any kind may pose a long-term
threat to the environment, and at present, our
knowledge of interactions in the vadose zone does
not permit us to make unequivocal predictions
from theoretical principles. Much work must be
empirical, and teams of scientists and engineers
must be involved in devising the plan of discharge.
Perhaps most importantly, each type of discharge
and potential site must be treated individually, on
a case-by-case basis. No one can claim that the job
of insuring non-destructive disposal of wastes to the
vadose zone will be easy, but neither can they
claim that the job is impossible.
An Example of Geochemical Testing
As part of the task of developing possible
regulations to govern discharges to the subsurface
in New Mexico, this writer was asked to present an
example of how the geochemical portion of the
testing might be conducted. The work was not
intended to be complete, but was designed instead
to illustrate the minimum information that would
be appropriate. A very brief summary is presented
here.
The soil chosen was from Sulfur Springs, New
Mexico. It was dark brown in color, contained
rootlets, and represented a composite of the upper
six inches of the profile. A 1:2 soil:water slurry
yielded a potentiometric pH of 6.4. The soil was
air-dried and sieved to (-) 2 mm to remove the
coarse, non-reactive fragments. A small column of
soil [approximately 23.2 cc (1.42 cubic inches)]
was prepared in a chromatography tube. The test
solution was gravity-fed into the column at
various rates, from about 0.5 to 1.6 ml/min
(corresponding to a velocity of about 4 to 12
feet/day). Saturated conditions were maintained,
together with a relatively high rate of flow,
because of limitations of time. An actual case
would require slower rates of flow and an
unsaturated column. Aliquots of the effluent were
25
-------
collected in volumes of about 9.5 ml (0.0025
gallons) by means of an automatic fraction
collector. The porosity of the soil column was
estimated in two ways: (1) addition of water to
saturation and reweighing in the column and
(2) measurement of volume of water withdrawn
from a graduated cylinder by suction into the
soil column. Density was determined by weighing
and measuring the volume of soil column. The
porosity was 48 + 4 percent, and the density was
0.88 ± 0.03 gm/cc (55 ±2 Ibs/cubic ft). The pore
volume of the column was 11 ± 1 cubic centimeter
(0.00039 cubic feet). The errors involved in this
crude study are obvious, such as the disturbance
and sieving of the soil, and the rough estimate of
porosity and density. For an actual plan of
discharge it would probably be necessary to obtain
cores of the soil; ASTM methods would be used to
measure the porosity and density of the core
material.
The test solution was made up to simulate the
aqueous discharge from a metallurgical flotation
mill. The aqueous matrix was a saturated solution
of gypsum, initially spiked with 1.0 ppm (mg/1)
dissolved copper and 2.1 ppm (mg/1) dissolved
molybdenum. The initial concentration of copper
was made considerably higher than might be
expected in an actual mill effluent; this was done
because initial tests indicated that the soil had a
large capacity for removal of copper from the
test solution. The initial solution and the aliquots
of effluent were analyzed for copper and
molybdenum by atomic absorption and a thio-
cyanate colorimetric procedure, respectively.
Standard additions were employed to test for
sensitivity and interference in the copper analysis.
The results for molybdenum are shown in
Figure 3. The results for copper are not shown
because copper was generally below the limits of
detection in the effluent during most phases of
the study (the detection limit for dissolved
copper was 0.01 parts per million).
As shown in Figure 3, an initial period of
leaching by distilled water (simulated rain water)
failed to release detectable amounts of
molybdenum (less than 15 parts per billion
dissolved). The concentration of dissolved copper
was approximately 0.05 ppm in the effluent during
this initial leaching, which continued for 378 ml
(34 pore volumes). This initial period of leaching
shows that neither the concentration of copper
nor molybdenum released by heavy applications
of fresh water to Sulfur Springs soil poses a threat
to ground water. Next, in Figure 3, the column
-
-
-
-
_
-
_
.
-DISTILLED
- WATER
-
-
-
SAT'D
GYPSUM
SOL'N
m=mmMlt
H
I'
A
t
/
I
1
I SAT D
• GYPSUM
i SOLUTION
} WITH
/I 0 mg/1 Cu
AND
f 2.1 mg/1 Mo
/
\
{
1 —
\
\
1
I DISTILLED
\ WATER
*
Is T T
SATURATED
GYPSUM
SOLUTION
L . -, . _> .-
SAT'O
GYPSUM
SOL'N
WITH
2 0 mg/1
Cu
AND
20 mg/i
Mo
0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600
MILLILITERS OF EFFLUENT
Fig. 3. Results of experimental study of leaching, removal,
and remobilization of molybdenum using synthetic mill
water and a New Mexico soil. Analytical error shown by
vertical bars; error is approximately the diameter of the
circles along the abscissa.
was leached with approximately 26 pore volumes
(288 ml) of metal-free synthetic mill water. In
this case, dissolved copper in the effluent
averaged less than 0.01 ppm and molybdenum less
than 15 ppb. Again, these results show that
extensive leaching with metal-free mill water fails
to leach significant quantities of either metal from
the soil.
At a volume of 661 ml in Figure 3 the
metal-spiked water was introduced into the column,
and 72 pore volumes were passed through. No
detectable copper (less than 0.01 ppm) ever
appeared in the effluent. Molybdenum was first
detected after 66 ml (6 pore volumes) of spiked
water had flowed through the soil. Fifty percent
breakthrough of molybdenum occurred after
passage of 32 pore volumes. The capacity of the
soil to remove molybdenum was completely
exhausted after passage of 72 pore volumes of
mill water (792 ml). The difference in behavior
of these two metals is caused by their ionic form
in solution; molybdenum occurs as an anion
(MoO4~~), whereas copper is present as a cation
(Cu++) under these conditions. This soil had a very
large capacity for removal of copper.
In Figure 3, after exhaustion of the column
for molybdenum, the soil was again leached with
simulated rainfall and metal-free mill water to test
for desorption and remobilization. Of the 792
micrograms of copper introduced, leaching by
165 pore volumes of distilled water and synthetic
mill solution released only about 19 micrograms;
this means that the removal of copper by this
soil approaches being irreversible. The initial
leaching of the soil, prior to the introduction of
26
-------
the metal-rich mill water, released as much copper.
In contrast, leaching of the loaded column with
83 pore volumes of distilled water remobilized
nearly 60 percent of the approximately 700
micrograms of molybdenum previously added.
The rate of release of molybdenum was
decreasing exponentially with continued leaching
by the distilled water, as shown in Figure 3,
indicating that a very large volume of water would
be required to remove all of the added molybdenum.
Interestingly, upon initiation of leaching of the
loaded soil with metal-free mill water, the
concentration of molybdenum in the effluent fell
to about 10 parts per billion, continuing to the end
of this portion of the experiment with an additional
82 pore volumes (about 3280 ml on Figure 3).
Calculations suggest that this decrease in desorbed
molybdenum was probably due to the precipitation
of solid CaMoO4 in the soil due to the high
concentration of dissolved calcium in the water.
Finally, in Figure 3, a new spiked mill water,
containing 2 ppm copper was passed through the
column in an attempt to cause breakthrough of the
dissolved copper; however, breakthrough did not
occur, even after a calculated addition of 256 ppm
copper to the solid phases in the soil column.
The results of this study show that the soil
from Sulfur Springs, New Mexico, has an
enormously high capacity to remove copper, and
that the removal is essentially irreversible. A
relatively high capacity for removal of molybdenum
is also exhibited, with 50 percent breakthrough at
32 pore volumes of spiked mill water. This type
of information would comprise an important
portion of a rational plan of discharge.
Space does not permit development here of
expressions to relate the data of Figure 3 to a real
situation. However, it can be shown from
considerations of the distribution of the ions
between the liquid and solid phases (Hajek, 1969;
Tamura, 1972; Wierenga and others, 1975) that
thousands of years may be required for the
molybdenum in this study to reach an assumed
water table at a depth of 30 meters (98 feet) in
the Sulfur Springs soil, depending on the rate of
discharge and infiltration. Copper would require
much longer to reach the ground water.
SUMMARY
In summary, I have tried to demonstrate that
there are valid chemical and physical reasons for
looking toward the subsurface, especially the
vadose zone in arid regions, for disposal of some
types of wastes. Processes that operate in this
environment can, under favorable circumstances,
greatly attenuate and detoxify some wastes. On the
other hand, we know of too many examples of
contamination of ground water to believe that the
overlying mantle of soil and sediment affords
absolute protection. With these facts in mind, we
must try to balance the need for protection of
ground water against the economic realities of the
use of water by industry, municipalities, and the
general public.
ACKNOWLEDGMENTS
My thanks to Helen Gram, Maxine Goad, John
Dudley, Charles Nylander, John Wright, and the
New Mexico Environmental Improvement Agency
for assistance and permission to publish. Thanks
also to John Mann and Jay Lehr for encouragement
and stimulation.
REFERENCES
Alberts, J. J., J. E. Schindler, R. W. Miller and D. E. Nutter,
Jr. 1974. Elemental mercury evolution mediated by
humic acid. Science, v. 184.
Apgar, M. A. and D. Langmuir. 1971. Ground-water
pollution potential of a landfill above the water table.
Ground Water, v. 9, no. 6.
Bouma, J., W. A. Ziebell, W. G. Walker, P. G. Olcott,
E. McCoy and F. D. Hole. 1972. Soil absorption of
septic tank effluent. Univ. Wisconsin Extens., Info.
Circ. 20, Madison, Wisconsin.
Bouwer, H. 1974. Design and operation of land treatment
systems for minimum contamination of ground water.
Ground Water, v. 12, no. 3.
Bower, C. and J. Thatcher. 1967. Adsorption of fluoride
by soils and minerals. Soil Sci. v. 103.
Cole, J. A. (ed.). 1974. Groundwater pollution in Europe.
Water Inform. Center, Inc., Port Washington, N.Y.
Dutt, G. R. and others. 1968. Trace and tracer elements
in ground water. Research Proj. Tech. Completion
Rept., Water Resources Res. Center, Univ. Arizona,
Tucson.
Ellis, B. G. and B. D. Knezek. 1972. Adsorption reactions of
micronutrients in soils, in Micronutrients in agriculture.
Soil Sci. Soc. Amer., Madison, Wisconsin.
Fryberger, J. S. 1972. Rehabilitation of a brine-polluted
aquifer. U.S. Environ. Prot. Agency, Environ. Prot.
Technology Series EPA-R2-72-014.
Furiman, D. K. and J. R. Barton. 1971. Ground-water
pollution in Arizona, California, Nevada, and Utah.
Environ. Prot. Agency, Water Poll. Control Res.
Series 16060ERU.
Goering, H. R., E. E. Gary, L.H.P. Jones and W. H. Allaway.
1968. Solubility and redox criteria for the possible
forms of selenium in soils. Soil Sci. Soc. Amer. Proc.
v. 32.
Griffin, R. A., R. R. Frost and N. F. Shimp. 1976. Effect of
pH on removal of heavy metals from leachates by
clay minerals. Preprint from Symposium on Hazardous
Waste Research, Residual Management Land Disposal
27
-------
(W. H. Fuller, Conf. Director). Tucson, Arizona,
Feb. 2-4, sponsored by Univ. Arizona and U.S.
Environ. Prot. Agency.
Hajek, B. F. 1969. Chemical interactions of wastewater in
a soil environment. Jour. Water Poll. Control
Feder. v. 41, no. 10.
Hanshaw, B. B. 1972. Clay membrane phenomena, in
R. W. Fairbridge (Ed.), The encyclopedia of geo-
chemistry and environmental sciences. Van
Nostrand-Reinhold Co., Princeton, N.J.
Hem, J. D. 1970. Study and interpretation of the chemical
characteristics of natural water, second edition. U.S.
Geol. Survey Water-Supply Paper 1473.
Johnson, C. M. 1966. Molybdenum, in Chapman, H. D. (ed.),
Diagnostic criteria for plants and soils. Homer D.
Chapman, Riverside, California.
Katz, B. G. and D. D. Runnells. 1974. Experimental study
of sorption of Mo by desert, agricultural, and alpine
soils, in D. D. Hemphill (ed.), Trace substances in
environmental health, VIII. Univ. Missouri, Columbia,
Missouri.
Kellogg, W. W., R. D. Cadle, E. R. Allen, A. L. Lazrus and
E. A. Martell. 1972. The sulfur cycle. Science.
v. 175, no. 4022.
Lagerwerff, J. V. 1972. Lead, mercury, and cadmium as
environmental contaminants, in Micronutrients in
agriculture. Soil Sci. Soc. Amer., Madison, Wisconsin.
Lakin, H. W. 1973. Selenium in our environment, in Kothny,
E. L. (ed.). Adv. in Chem. Series No. 123, Amer.
Chem. Soc., Wash., D.C.
LeGendre, G. R. and D. D. Runnells. 1975. Removal of
dissolved molybdenum from wastewaters by
precipitates of ferric iron. Environ. Sci. and Technol.
v. 9, August.
Lieber, M., N. M. Perlmutter and H. L. Frauenthal. 1964.
Cadmium and hexavalent chromium in Nassau County
ground water. Jour. Amer. Water Works Assoc. v. 56,
no. 6.
Miller, D. W., F. A. DeLuca and T. L. Tessier. 1974. Ground-
water contamination in the northeast states. Environ.
Prot. Agency, Environ. Protect. Technology Series
EPA-660/2-74-056.
Minear, R. A. and J. W. Patterson. 1973. Septic tanks and
ground-water pollution, in Ground-water pollution.
Underwater Res. Instit., St. Louis, Missouri.
Papadopulos, S. S. and I. J. Winograd. 1974. Storage of
low-level radioactive wastes in the ground—hydro-
geologic and hydrochemical factors. U.S. Geol.
Survey Open-File Report 74-344, Reston, Virginia,
for the U.S. Environ. Prot. Agency.
Scalf, M. R., J. W. Keeley and C. J. LaFevers. 1973. Ground-
water pollution in the south central states. Environ.
Prot. Agency, Environ. Prot. Technology Series
EPA-R2-73-268.
Tamura, T. 1972. Sorption phenomena significant in
radioactive waste disposal, in T. D. Cook (ed.),
Underground waste management and environmental
implications. Amer. Assoc. Petrol. Geologists Memoir
18, Tulsa, Oklahoma.
Trost, P. B. 1974. Groundwater geochemical exploration
programs—an overview. Abstract in water resources
problems related to mining, Symposium at Colo.
School of Mines, July 1-2, Golden, Colorado,
American Water Resources Assoc.
van der Leeden, Frits, L. A. Cerrillo and D. W. Miller.
1975. Ground-water pollution problems in the
northwestern states. U.S. Environ. Prot. Agency,
Environ. Prot. Technology Series EPA 660/3-75-018.
Weir, Alexander, Jr., S. T. Carlisle and John Morris. 1975.
The environmental effects of trace elements in the
pond disposal of ash and flue gas desulfurization
sludge. Report prepared by Southern California
Edison Company and Radian Corporation, for
Electric Power Research Institute, Palo Alto,
California.
Wierenga, P. J., M. Th. van Genuchten and F. W. Boyle.
1975. Transfer of boron and tritiated water through
sandstone. Jour. Environ. Quality, v. 4, no. 1.
Winograd, I. J. 1974. Radioactive waste storage in the arid
zone. Trans. Amer. Geophys. Union, v. 55, no. 10.
Woolson, E. A.,J. A. Axley and P. C. Kearney. 1971. The
chemistry and phytotoxicity of arsenic in soils:
I. Contaminated field soils. Soil Sci. Soc. Amer. Proc.
v. 35, no. 6.
DISCUSSION
The following questions were answered by Donald D.
Runnells after delivering his talk entitled "Wastewaters in
the Vadose Zone of Arid Regions: Geochemical
Interactions."
Q. by Leonard Wood. Rain water is not distilled water. Have
you used rain water in your column experiments?
A. No, I have not tried to exactly duplicate the chemistry of
rain water in the column studies. However, I do not believe
that the observed variations in the chemistry of rain water
would make a great deal of difference under most circum-
stances. The exception might be in regions of serious air
pollution in which sulfuric acid becomes the principal
control on the pH of rain water.
Q. by Mike Kaczmarek. Would substantial increases in the
Mo loading in the soil column result in comparable
increases of Mo in the leachate resulting from "rain water"
(distilled water)?
A. Yes, the greater the initial loading, the greater would be
the slug of Mo released from a soil when remobilization
due to leaching does take place.
Q. by E. E. Jones, Dale Ralston and Harry Nightingale.
What percent of pore space was saturated in your
experiments?
A. This particular column was filled from the bottom to
drive out air, then maintained in a fully saturated condition,
so the pores were nearly 100 percent filled with water.
28
-------
This was done because a limited amount of time was
available and I had to maintain high rates of flow. A
different experiment could be conducted under non-
saturated conditions, in which the rate of flow would be
substantially slower.
Q. by Harold Meiser. What experiments are you -planning
for the future?
A. We now have underway a program to test for leaching
of trace elements from spent oil shale, and the reactivity of
these elements with soils of the oil shale region in Colorado.
Q. by Neil Jaquet. How do the volumes of distilled water
used in your column studies compare to rates of precipita-
tion in the area of concern?
A. Experiments such as I described are designed to yield
basic information, such as the distribution coefficient, Kj.
Such information can then be used in equations involving
the specific rates of precipitation and infiltration for an
area. Thus, it is not necessary to model the amount of
rainfall and infiltration, just the chemistry and rates of flow.
Q. by Jon O. Nowlin. Assuming contaminant "removal" by
cropping, how is the crop to be disposed of?
A. The three advantages in cropping would be: (1) to convert
the dissolved contaminants into solid form, or (2) to
consume and transform such contaminants as nitrate, BOD,
etc., or (3) to concentrate the contaminants from a large
amount of fluid into a smaller volume of crop. However,
in some cases, cropping would still offer no advantages
because you still could not dispose of it.
Q. by K. E. Childs. Can these tests be used to evaluate
other areas, soils, and chemicals, or are the results too
qualitative for this?
A. The results are quantitatively useful for the particular
parameters chosen, but new tests would have to be run for
other discharges, soils, etc.
Q. by P. K. Saint. Can one quantify the sorption capacity
of the soil and predict the time of passage of an ion
through the vadose zone? How about desorption later on?
A. Yes, the results can be used for predicting migration,
and references are given in the paper to illustrate this.
Figure 3 in the paper shows the results of desorption
studies.
Q. by Logan Kuiper. What harmful effects would result if
the contaminants reached ground water in the case
considered?
A. Molybdenum at this concentration (2.1 mg/1) could be
toxic to ruminant animals, and the TDS (saturated gypsum
solution) would degrade the quality of the ground water.
Finally, the 1.0 mg/1 copper would probably not be
harmful to health, but would be undesirable because of
taste.
Q. How do different types of soils affect the leaching
process?
A. The processes of removal and leaching are strongly
dependent on the type of soil.
Q. by G. F. Hendricks. Any study of pick-up of copper by
plants?
A. There have been several such studies by others in the
past, but this was not part of our program.
Q. by Don Lundy. What mechanism retained the copper?
A. I really did not get into the study of specific mechanisms.
The experiments were designed only to give a general
indication of capacity for removal and retention.
Q. by Buck Steingraber. What happened to the sulfate in
your percolating water—did it come through in the original
concentration?
A. The specific conductance that was measured in the
effluent suggested that the sulfate passed through the soil
with little or no removal.
Q. What time equivalency in years of precipitation would be
required to equal the amount of solution put through the
soil column?
A. For a semi-arid region, with perhaps 2 centimeters of
recharge per year, the total volume of solution percolating
through each cubic centimeter of soil would correspond to
about 10 saturated pore volumes (assuming 20% porosity)
per year. In this study I passed about 300 pore volumes
through the saturated column, corresponding to 30 years
of recharge. However, the ions in the experiment were
retarded by the soil and would move much more slowly
than the water.
29
-------
Zoning Aquifers for Tertiary Treatment
o
of Wastewater
by Herman Bouwer
ABSTRACT
Soils and aquifers can function as effective and
economical filter systems for advanced treatment of
conventionally treated sewage and other wastewater. The
wastewater is applied to the land with low-rate or high-rate
infiltration systems. Physical, chemical, and biological
processes in the soil improve the quality of the wastewater
as it percolates through the vadose zone and into the
aquifer to become renovated water. The quality of the
renovated water, however, often is not as good as that of
the native ground water. To utilize the land for treatment
of wastewater, without trading a problem of surface-water
pollution for one of ground-water contamination, the
spread of renovated water in the aquifer must be restricted.
This can be accomplished by locating the system so that
the renovated water drains naturally into a stream or other
surface water, or by artificially removing renovated water
from the aquifer with wells or drains at some distance
from the application area. Examples are given of various
systems that utilize these principles, and general design
criteria are presented. Proper design involves analysis of
underground-flow systems for various system geometries.
Methods for measuring hydraulic conductivity, particularly
in the vadose zone, are briefly reviewed.
INTRODUCTION
Land application of wastewater in which the
pollutants are primarily of organic origin, like
conventionally treated sewage and effluents from
agricultural processing plants, can be an attractive
alternative to additional in-plant treatment. Interest
in land application of wastewater is increasing
because of:
presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
bDirector, U.S. Water Conservation Laboratory,
Agricultural Research Service, U.S. Department of Agri-
culture, 4331 East Broadway, Phoenix, Arizona 85040.
1. Legal restrictions on the discharge of waste-
water into streams, lakes, or other surface water
(Clean Water Bill, PL 92-500).
2. Favorable economic aspects, high reliability,
and lower energy requirements than comparable
in-plant treatment (including advanced techniques).
3. Opportunity for utilization of wastewater
for irrigation of crops and use of nitrogen,
phosphorus, and other "pollutants" in the water as
fertilizer.
4. Opportunity for filtration of wastewater
through soils and aquifers to produce renovated
water for ground-water recharge. This water
can then be reused for unrestricted irrigation,
recreation, and other purposes.
TYPES OF SYSTEMS
There are three main types of land application
systems: overland flow, low-rate infiltration
systems, and high-rate infiltration systems. Overland-
flow systems are used when soils are tight and
infiltration is slow. Improvement in water quality is
obtained primarily by flowing the wastewater in a
shallow sheet over a soil surface covered by grass or
other dense vegetation. Because of the low infiltra-
tion rates, overland-flow systems have little or no
effect on the underlying ground water. With low-
rate or high-rate systems, the wastewater infiltrates
into the soil and the water-quality improvement is
obtained as the water moves through the soil and
down to the ground water. In low-rate systems,
about 2 to 15 cm (1 to 6 inches) of wastewater are
applied weekly or every 2 weeks. All systems that
utilize wastewater for irrigation fall in this category.
The water may be applied with surface-irrigation
techniques (borders, furrows) or with sprinklers.
30
-------
Surface-irrigation techniques require relatively
smooth land, but they require less pumping (energy)
and cause less air pollution than sprinkler systems.
With high-rate systems, application rates vary from
about 50 to 500 cm (2 to 20 ft) per week. Such
amounts can be applied by over-irrigating agricul-
tural crops or with special recharge or infiltration
basins that generally are not vegetated. Intermittent
flooding is the key to successful operation of rapid
infiltration basins. Schedules range from 8-hour
flooding and 16-hour drying each day to flooding
and drying periods of several weeks each.
QUALITY IMPROVEMENT
The movement of wastewater through soil
greatly improves its quality through physical,
chemical, and biological reactions. Suspended
solids are essentially completely filtered out.
Bacterial activity in the soil results in almost
complete removal of the biodegradable waste
matter in the water, as expressed by the
biochemical oxygen demand (BOD). However, not
all organic compounds are biodegradable and some
refractory organics may be present in the renovated
water. For sewage effluent, for example, the
renovated water may contain about 5 mg/1
refractory organics, expressed as total organic
carbon. The numbers of bacteria and viruses in the
wastewater are greatly reduced, and often com-
pletely removed, as the water percolates through
soil and aquifer(s).
For low-rate systems, most of the nitrogen in
the wastewater is absorbed by crops, which can
remove almost 100% of the nitrogen. For high-rate
systems, the nitrogen is removed primarily by
denitrification in the soil, which can remove about
80% nitrogen and more from secondary sewage
effluent, as shown in laboratory studies (Lance
et al, 1976). Similar removal rates should be
possible for field systems under favorable
conditions and with careful management.
Otherwise, lower removal percentages can be
expected. Phosphates are also taken up by crops or
they can be immobilized in the soil, particularly
when the soil contains iron and aluminum oxides
or is calcareous and alkaline. Phosphates are more
mobile in sands with low pH. Immobilization of
metal ions in soil is stimulated by pH values
greater than 7 and by certain types of clay. Some
metal ions, however, may remain in the water
and move over long distances, especially when kept
in solution by refractory organics that act as
chelating agents. For a more detailed discussion of
the various physical, chemical, and biological
reactions in the soil and the chemical composition
of different types of wastewater and renovated
water, reference is made to Bouwer and Chancy
(1974).
Low-rate (irrigation) systems in dry climates
yield deep percolation water with a salt content
that is about 3 to 10 times that of the original
wastewater. Thus, unless the original wastewater
has a very low salt content, the deep percolation
water from those systems cannot be reused and
must be handled like deep percolation or return
flow from normal irrigated fields, including
removal by drainage and "disposal" into surface
water. For high-rate systems, the salt content of
the renovated water is about the same as that of
the original wastewater.
High-rate systems require relatively deep,
light soils, preferably in the sandy-loam to
loamy-sand range. Low-rate systems can be used
on a wider range of soil textures. Coarse sands
and gravels should be avoided. If such materials are
also relatively shallow and underlain by fractured
or cavernous rock, wastewater could penetrate to
ground water in almost unchanged condition.
Most of the quality improvement of wastewater
in both low- and high-rate systems occurs in the
top 1 m (3 ft) of soil. However, additional under-
ground travel and underground detention time
should be allowed for "polishing" treatment of the
wastewater, including taste-and-odor removal and
die-off of microorganisms (Gerba et al., 1975).
ZONING OF AQUIFERS
While the wastewater is greatly improved in
quality as it percolates through the soil in well-
managed land application systems on properly
selected sites, the quality of the resulting renovated
water may not be as good as that of the local
native ground water, whose quality is thus degraded.
To take advantage of the beneficial aspects of land
application without trading a surface-water
pollution problem for a ground-water contamination
situation, the spread of renovated water in the
aquifer must be restricted. This can be accomplished
by removing the renovated water from the aquifer
at some point away from the application area.
Sometimes this can be done by locating the land-
application system so that the renovated water
drains naturally into surface water (Figure 1).
Where such a system is not possible, the renovated
water can be collected at some distance from the
application system by drains if the aquifer is
shallow (Figure 2), and wells if the aquifer is deep
(Figure 3). The portion of the aquifer between
31
-------
IMPERMEABLE
LAYER
Fig. 1. Natural drainage of renovated water into surface
water.
the application system and the discharge point of
renovated water is then used as a natural filter
system for renovating wastewater. Such zoning of
part of an aquifer as a tertiary treatment facility
for wastewater is not new. For example, if a certain
separation is specified between a well and a septic
tank, cemetery, or landfill, the portion of the
aquifer in the zone of separation is in fact dedi-
cated to treatment and dilution of the leachate
from the pollution source. The difference
between this conventional zoning and zoning for
land-application systems is that, for the latter,
the wastewater is removed from the aquifer after
it has become renovated water. The renovated
water may be discharged into surface water
(through natural drainage or by pumping the
discharge from the drains or wells into surface
water) for indirect reuse, or it may be directly
reused for unrestricted irrigation, unrestricted
recreation, and certain industrial applications.
Using renovated water for drinking is not yet
encouraged, because the possible health effects
of the refractory organics and other compounds
present in low concentrations in the renovated
water are not yet completely understood.
Examples of land-application systems
where renovated water drains naturally into a
Mill1
1111
•IMPERMEABLE LAYER
Fig. 2. Collection of renovated water by horizontal drain.
32
INFILTRATION STRIPS
iwELL
WELL
1 1 1111
4 4,
11111
/
^^-IMPERMEABLE LAYER
Fig. 3. Collection of renovated water by wells.
stream are the high-rate infiltration systems of
Fort Devens, Massachusetts, and Bielefeld,
Germany. The wastewater for both systems
consists of treated sewage effluent. Renovated
water is collected by horizontal drains for the rapid
infiltration systems in the dunes of western
Holland, where polluted Rhine water is pretreated
for municipal water supply, and at Dortmund,
Germany, where a similar system is used to
pretreat relatively high-quality surface water from
the river Ruhr. An open ditch or trench is used
at the Santee, California, system to intercept
renovated sewage water for use in recreational
lakes. Renovated water is collected with wells in
the high-rate infiltration systems for Phoenix,
Arizona, to enable reuse of the sewage water for
unrestricted irrigation and recreation. Restricting
the spread of renovated water in the aquifer by
. drains or wells is easier and less costly for high-
rate than for low-rate application systems, because
the former require much less land area than low-
rate systems of the same capacity.
DESIGN
The design of a system for renovating waste-
water by land application and filtration through
soils and aquifers, and discharging the renovated
water from the aquifer at some distance from the
application system, should be based on the follow-
ing criteria:
1. Ground-water mounds (including perched
mounds) must not rise so high during application of
wastewater that they restrict the infiltration rate.
This can be accomplished by keeping the top of
-------
the capillary fringe above the ground-water mound
at least 0.5 m (2 ft) below the surface of the soil
or the bottom of the infiltration basin [the corres-
ponding water-table depth will then be about 1 to
2 m (3 to 7 ft), depending on soil texture].
Ground-water mounds should drop to a depth of
at least 1.5 m (5 ft) within 2 or 3 days after
cessation of wastewater application to permit
sufficient aeration of the soil during periods of no
infiltration (drying or resting).
2. The design lateral flow in the aquifer
between the area of application of wastewater and
that of discharge of renovated water should not
exceed the flow that the aquifer can handle on the
basis of its transmissivity and hydraulic gradients
(this applies primarily to systems as in Figure 1).
3. The water must have traveled a sufficient
distance underground and must have had sufficient
underground detention time before it leaves the
aquifer as renovated water. Desired distances and
times of underground travel depend on the
wastewater characteristics, the nature of the soil
and aquifer materials, and the desired quality of
the renovated water. For several systems, under-
ground travel distances and detention times are
about 100 m (300 ft) and 1 month, respectively
(see also Gerba etal, 1975).
Where renovated water discharges naturally
into surface water, the lateral flow from the
application area to the surface water can be
restricted by insufficient aquifer transmissivity. In
that case, the product WI of width of infiltration
area and infiltration rate must not exceed the
product KD H/L of aquifer transmissivity and slope
of the water table (Figure 1). When KD H/L is
relatively small, the wastewater-application area
should be long and narrow so that W and, hence,
IW are also small.
Where the renovated water is collected by
underground drains (Figure 2), the flow system
below the water table can be described by the
following equation (Bouwer, 1974):
W = width of application area,
K
Hc2 = Hd2 + IW(W + 2L)/K
where Hc = maximum height of water table
above impermeable layer,
(1)
= height of water table above imper-
meable layer at drains (height of
center of drains if drains are running
free with no back pressure),
= infiltration rate (average for entire
area of width W),
L
= hydraulic conductivity of aquifer,
and
= distance between edge of infiltration
area and drain.
Equation 1 is based on the assumption of
horizontal flow, which is valid if Hc is relatively
small as compared with W + L. The value Hc should
be taken at the outer edge of the infiltration area if
the drain is only on one side of the area. However,
if drains are on both sides of the infiltration area,
Hc refers to the water table below the center of the
application strip and W represents one-half the
width of the application area. Equation 1 can be
used to calculate L for various combinations of W,
I, and Hc and Hj so that the best geometry of
application area and drain location can be
selected. The transient situation, i.e. predicting the
rise and fall of ground-water mounds in relation to
time, can be handled with equations developed by
Marino (1974a and 1974b) for recharge areas bound
on one or both sides by surface water with constant
water level.
For deep aquifers, the renovated water is
more effectively intercepted by wells than by
horizontal drains. If several such wells are uniformly
spaced on a line midway between two parallel
infiltration strips (Figure 3), it is theoretically
possible to collect essentially all the water that
entered the soil as wastewater. To achieve this, the
wells should completely penetrate the aquifer (or
at least the upper, active region if the aquifer is
deep enough to develop active and passive regions).
In addition, infiltration and pumping rates should
be managed so that ground-water levels below the
outer edges of the infiltration strips are not affected
by the infiltration-and-pumping system and remain
at the same level as the ground water in the aquifer
adjacent to the renovation system. For this purpose,
observation wells for monitoring ground-water
levels should be installed at the periphery of the
system.
To obtain the optimum design of a system
such as in Figure 3 (width and length of infiltration
strips, distance between strips, distance between
wells, and capacity of wells), the water-table drop
from the outer edge of the recharge system to the
wells should be calculated for various designs so
that the best system geometry can be selected. This
water-table drop can be predicted by superimposing
the ground-water mound formed by infiltration
from the recharge strips on the drawdown due to
the pumped wells. The rise and fall of ground-water
33
-------
mounds in aquifers of large lateral extent can be
predicted with equations or graphs presented by
Hantush (1967), Bianchi and Muckel (1970),
Hunt (1971), and Singh (1972). Drawdowns
around wells can be calculated with conventional
well-flow theory. Steady-state solutions for the
flow system between the recharge areas and the
wells for the geometry of Figure 3 were presented
by Bouwer (1970). Equations developed on the
basis of horizontal-flow theory should be used with
caution where the thickness of the aquifer is
larger than the width of the infiltration area,
because the lower portions of the aquifer then do
not contribute much to the recharge flow system
and are essentially stagnant. In those cases, the
effective transmissivity of the aquifer for recharge
is less than the total transmissivity of the aquifer
(Bouwer, 1962).
MEASUREMENT OF HYDRAULIC
CONDUCTIVITY
To predict underground-flow systems in
connection with land-application systems for
wastewater with discharge of renovated water
from the aquifer, values of the hydraulic con-
ductivity K in the aquifer and the vadose zone
must be known. Such knowledge of K also enables
prediction of maximum infiltration rates. Local
experimentation, however, is usually required to
determine the effect of clogging, biological
activity, and weather on infiltration rates, so that
a realistic estimate of the infiltration rates can
be obtained.
Values of K should preferably be measured
in place. The K-values of aquifers can be obtained
with pumped-well techniques, like the Theis
pumping test, or with the slug test where a volume
of water is suddenly removed from a well and the
subsequent rise of the water level in the well is
measured for calculation of K. The theory for the
slug test, which originally was developed for
completely penetrating wells in confined aquifers
(Cooper, et al, 1967) has recently been extended
to partially penetrating wells in unconfined
aquifers (Bouwer and Rice, 1976). Point measure-
ments of K, as may be required to detect layers of
different K in the aquifer, can be obtained with
piezometer techniques (Bouwer and Jackson,
1974, and references therein). For the vadose zone,
K may have to be determined in vertical direction
to predict effects of restricting layers on perching
mounds and to estimate potential infiltration rates.
There may also be cases where K of the vadose
zone should be measured in horizontal direction,
for example, if infiltration will cause the water
table to rise and part of the vadose zone will
become saturated and contribute to lateral flow of
ground water.
The hydraulic conductivity in vertical direction
in the vadose zone can be determined with the
air-entry permeameter and infiltration-gradient
techniques (Bouwer and Jackson, 1974, and
references therein). The air-entry permeameter
(Figure 4) is essentially a covered infiltrometer
with standpipe and reservoir to let the water
infiltrate into the soil at high head. The infiltration
rate is calculated from the rate of fall of the water
level in the reservoir. When the wet front is
expected to have reached a depth of about 10 cm
(4 inches, equal to the depth of penetration of the
cylinder), a valve at the base of the pipeline
connecting the reservoir to the cylinder is closed.
This halts infiltration and causes the pressure of
the water inside the cylinder to decrease until the
air-entry value of the soil above the wetting front
is reached, after which the pressure of the water
inside the cylinder slightly increases. Measuring
the minimum water pressure inside the cylinder
with, for example, a vacuum gauge with memory
pointer, enables the air-entry value at the wetting
front to be evaluated. This value is then used to
estimate the pressure head at the wetting front
while water was still infiltrating. As soon as
minimum pressure is detected, the air-entry
VACUUM
GAGE
RESERVOIR
r
SUPPLY VALVE
DISK
AIR ESCAPE
VALVE
FRONT
Fig. 4. Schematic of air-entry permeameter.
34
-------
TO MANOMETER OR
/X PRESSURE TRANSDUCER
OUTER
TUBE '
.S.
INNER
TUBE
Fig. 5. Schematic of infiltration-gradient method.
permeameter is removed and the depth of the
wetting front is measured, for example, by
digging with a spade and observing how deep the
soil was wetted. Knowing the depth of the water
above the soil, the infiltration rate just before
closing the valve, the depth of the wetting front
at the time the valve was closed, and the pressure
head at the wetting front while it was advancing
downward, K in vertical direction of the wetted
zone can be calculated with Darcy's equation. The
method requires about 10 liters (3 gal) of water
per test and is usually completed in less than 0.5
hour. In its present form of construction, the
air-entry permeameter is a surface device. To use
the device for measuring K at a certain depth, pits
or trenches must be dug.
With the infiltration-gradient technique
(Bouwer and Jackson, 1974, and references
therein), an auger hole is dug and lined with a
steel tube. A second steel tube of smaller
diameter is placed concentrically in the hole and
pushed or driven about 5 cm (2 inches) into the
hole bottom (Figure 5). The tubes are filled with
water and constant, equal, water levels are
maintained in both tubes. Infiltration then creates
a wetted zone with positive pressures in the soil
below the auger hole. The infiltration rate from
the inner tube is measured while small, fast-reacting
piezometers are pushed into the auger-hole bottom
at increments of about 2 cm (1 inch) to determine
the vertical hydraulic gradient in the wetted zone.
The K-value in vertical direction below the hole
bottom then can be calculated with Darcy's
equation. Once the hole is dug and the tubes are
installed and filled with water, the test can
usually be completed in about 1 hour, depending
onK.
A third technique for measuring K in vadose
zones is the double-tube method (Bouwer and
Jackson, 1974, and references therein), which
also consists of two concentric tubes in an auger
hole with the inner tube penetrating the hole
bottom about 2 cm (1 inch). The tubes are covered
with a lid that has two standpipes—one connected
to the inner tube and the other to the annular
space between the inner and the outer tube in the
hole (Figure 6). The tubes are filled with water
and water levels are maintained at the top of the
standpipes to create a wetted soil region with
positive water pressures below the auger hole.
When this region has reached sufficient size, the
inflow into the inner tube is stopped and the fall of
the water level in the inner-tube standpipe is
measured. This is done while keeping the outer-
tube standpipe full and, in another measurement,
while adjusting the water level in the outer-tube
WATER.
SUPPLY J
Fig. 6. Schematic of double-tube method.
35
-------
standpipe so that it falls at the same rate as that
in the inner tube. These measurements then yield
the reduction in infiltration from the inner tube
when the pressure head in the inner tube is less
than that in the outer tube, which enables
calculation of K of the wetted zone. The resulting
value is affected by K in vertical and horizontal
direction, but it mostly reflects K in vertical
direction. The double-tube method requires about
2 to 5 hours to complete and about 100 liters
(30 gal) of water per test.
Values of K in horizontal direction in the
vadose zone can be determined with the shallow-
well pump-in method. This technique consists of
drilling an auger hole, filling it with water, and
maintaining a certain water level in the hole
until the outflow from the hole has become
constant (Bouwer and Jackson, 1974, and
references therein). Several days may be required
before the outflow approaches a constant value
and large quantities of water (hundreds of liters
or about 100 gal) may be necessary. The value of
K is calculated from the constant outflow rate and
the hole geometry. The resulting value primarily
reflects K in horizontal direction. A true measure
of K in horizontal direction can be obtained by
combining the double-tube method with the
infiltration-gradient method in the same hole
(Bouwer and Jackson, 1974, and references therein).
Where infiltration basins or other facilities
are already available, like an experimental or
pilot ground-water recharge system, the trans-
missivity, T, of the aquifer can also be determined
from the measured rate of rise of the ground-water
mound in response to infiltration (Bianchi and
Muckel, 1970). This procedure yields the ratio
T/f, where f is the fillable porosity. If f is known
(f can be evaluated as the difference between the
volumetric water contents above and below a
rising ground-water mound), T can be calculated.
When using this procedure, care should be taken
in case T is measured from the mound rise below a
narrow recharge strip and then used to predict the
mound rise below a much wider strip or basin. This
is because recharge flow systems in aquifers have
an upper, active zone and a lower, passive zone if
the height of the aquifer is larger than the width of
the recharge basin (Bouwer, 1962). Thus, the
effective transmissivity for a narrow recharge strip
could be less than the actual transmissivity for the
entire aquifer, and less than the effective trans-
missivity for a wider recharge strip. This would
overestimate the height of a ground-water mound
below a wide infiltration area if the height is based
on T calculated from the observed mound rise
below a narrow infiltration area.
CONCLUSION
Zoning aquifers for wastewater renovation is
possible if the renovated water is discharged from
the aquifer at some distance from the infiltration
system. Successful application of the technique
requires careful selection of the site, and careful
design and operation of the system.
REFERENCES
Bianchi, W. C. and D. C. Muckel. 1970. Ground-water
recharge hydrology. U.S. Dept. of Agric., Agric. Res.
Service Publ. ARS 41-161, 62 pp.
Bouwer, Herman. 1962. Analyzing ground-water mounds
by resistance network. Jour. Irrig. and Drain. Div.
Amer. Soc. Civil Eng. v. 88, pp. 15-36.
Bouwer, Herman. 1970. Groundwater recharge design for
renovating wastewater. Jour. Sanit. Eng. Div., Amer.
Soc. Civil Eng. v. 96, pp. 59-74.
Bouwer, Herman. 1974. Design and operation of land
treatment systems for minimum contamination of
ground water. Ground Water, v. 12, pp. 140-147.
Bouwer, Herman and R. L. Chancy. 1974. Land treatment
of wastewater. In Advances in Agronomy, N. C.
Brady, ed., Academic Press, Inc., New York. v. 26,
pp. 133-176.
Bouwer, Herman and R. D. Jackson. 1974. Determining
soil properties. In Drainage for Agriculture. J. van
Schilfgaarde, ed., Agron. Monograph, no. 17. Amer.
Soc. of Agronomy, Madison, Wisconsin, pp. 611-672.
Bouwer, Herman and R. C. Rice. 1976. A slug test for
determining hydraulic conductivity of unconfined
aquifers with completely or partially penetrating
wells. Water Resources Res. v. 12, pp. 423-428.
Cooper, H. H., Jr., J. D. Bredehoeft, and I. S. Papadopulos.
1967. Response of a finite-diameter well to an
instantaneous charge of water. Water Resources Res.
v. 3, pp. 263-269.
Gerba, C. P., C. Wallis, and J. L. Melnick. 1975. Fate of
wastewater bacteria and viruses in soil. Jour. Irrig.
and Drain. Div. Amer. Soc. Civil Eng. v. 101, pp.
157-174.
Hantush, M. S. 1967. Growth and decay of ground-
water mounds in response to uniform percolation.
Water Resources Res. v. 3, pp. 227-234.
Hunt, B. W. 1971. Vertical recharge of unconfined aquifer.
Jour. Hydraul. Div. Amer. Soc. Civil Eng. v. 97,
pp. 1017-1030.
Lance, J. C., F. D. Whisler, and R. C. Rice. 1976. Maxi-
mizing denitrification during soil filtration of sewage
water. Jour, of Environ. Qual. v. 5, pp. 102-107.
Marino, M. A. 1974a. Growth and decay of groundwater
mounds induced by percolation. Jour, of Hydrol.
v. 22, pp. 295-301.
Marino, M. A. 1974b. Rise and decline of the water table
induced by vertical recharge. Jour, of Hydrol. v. 23,
pp. 289-298.
Singh, R. 1972. Mound geometry under recharge basins.
Calif. State Univ., San Jose, Report No.
GK-18526 for National Science Foundation, 71 pp.
36
-------
DISCUSSION
The following questions were answered by Herman Bouwer
after delivering his talk entitled "Zoning Aquifers for
Tertiary Treatment of Wastewater."
Q. by Jeffrey A. Gilman, William A. Trippet II, Randy
Sweet, J. E. Biesecker, and William H. Walker. Is there any
trace metal contamination of the aquifer associated -with
spreading of the sewage sludge in the Bielefeld, Germany
project?
A. I do not know of situations where heavy metals in
sewage effluent had adverse effects on crops or soils in land
treatment or irrigation systems. Should such problems
develop, applications of lime may be effective in
immobilizing the metals in the soil. Source control of
metals would be the logical approach where heavy-metal
concentrations in sewage effluent are unusually high. Most
of the metals in sewage end up in the sludge. Unless metal
concentrations are abnormally high, metal problems in soils
and crops are not anticipated if sludge applications are
based on crop nitrogen requirements.
Q. by Jack Robertson. Has the movement of any organic
constituents been measured in the Phoenix project?
A. BOD is reduced to essentially zero in the renovated
water. Refractory organics were still present at levels of
4 to 7 mg/1 total organic carbon.
Q. by P. K. Saint, Bill Chase, and Don Runnells. What about
the fate of stable organics and viruses during their travel
through vadose zone and zone of saturation? Is there data
from European experiences?
A. The state of knowledge is incomplete but results so far
indicate that movement and detention of viruses in soil
depend on soil, climate, and ionic composition of water.
Fine-textured soils with a low pH effectively remove viruses
from sewage effluent with a relatively high TDS and
concentration of divalent ions. Rainfall during drying can
cause previously adsorbed viruses to move deeper into the
soil. Application of calcium chloride to the soil prior to
adding distilled water, however, prevented mobilization of
viruses in laboratory studies. A detailed article on these
phenomena will be published by J. C. Lance and coworkers
in the October 1976 issue of Applied and Environmental
Microbiology.
Q. by G. C. Slawson. Are the wastewaters you are addressing
chlorinated as part of the treatment before disposal? If so,
is there not a potential for introducing chlorinated hydro-
carbons into the environment? This would clearly have to
be considered for water reuse planning.
A. The Phoenix projects utilize unchlorinated secondary
effluent. Chlorination is a waste of money if effluent is
used for rapid-infiltration systems in soils and aquifers that
remove all pathogens anyway.
Q. by W. B. Wilkinson. If a rapid infiltration system recharges
water with nitrate-nitrogen in excess of 10mg/l to the water
table-will the nitrate degrade as it moves through the
saturated zone to a river or well? If not, should such
systems be used?
A. Denitrification requires organic carbon as energy source
for the bacteria. Aquifers usually are devoid of readily
available organic carbon. Hence, once nitrate has reached
an aquifer, it is usually not attenuated other than by
dispersion. Whether rapid infiltration should be used or not
depends on how much nitrogen can be removed and where
the renovated water will go. Under the proper hydro-
geological conditions and with good management, it may
be possible to remove one-half to two-thirds of the nitrogen
from the sewage water. Since some of the nitrogen will
remain as ammonium, this will bring the nitrates down to
near or below the limit of 10 mg/1 nitrate-N.
Q. by W. B. Wilkinson. How can one be certain that smearing
does not occur at the base of a permeability test auger hole-
leading to false permeability test results?
A. Smearing could be a problem for the reverse auger-hole
method, also called the shallow-well pump-in technique.
For the infiltration-gradient technique, smearing is no
problem because the vertical gradient is measured directly.
For the double-tube method, smearing is minimized by the
use of a special hole cleaner which takes a bite out of the
bottom and leaves an undisturbed surface with a clean
break. For the air-entry permeameter technique, the
surface is accessible and can be cleaned or brushed to
expose a natural surface.
Q. by Jim Braithwaite. What was the sample volume (per
sample) in the virology analyses?
A. For the renovated water, the target sample volume was
100 gallons. In practice, it varied from 46 to 120 gallons.
This volume was passed through filters (using pH-aojust-
ments) to concentrate the viruses. For secondary effluent,
viruses were concentrated from 1- to 5-gallon samples.
Q. by N. Jaquet. What percentage of consumptive use (via
evaporation) is being experienced in the Phoenix system?
A. Annual hydraulic loading rates are 200 ft to 300 ft.
Annual evaporation is about 6 ft. Thus, 2 to 3% of the
water is lost to evaporation.
Q. by G. F. Hendricks. Would septic tank effluent absorption
field be improved by phosphate removal?
A. Not as a rule. Since phosphate precipitates as apatite-
like compounds in some soils, the permeability could
eventually become smaller. This will take a very long time,
however. Where soils do not remove much phosphate (acid,
sandy soils, for example) and the effluent seeps into a lake
or other surface water, it may be desirable to precipitate
phosphates so as to reduce the pollution danger of the
surface water.
Q. by G. J. Thabaraj and Martin J. Allen. What is your
recommendation regarding the minimum distance that
should be maintained between a low^rate or bigh^rate land
application system (for secondary sewage effluents) and a
potable water-supply well?
A. Under favorable conditions (relatively deep and fine-
textured soil, relatively deep water table, granular aquifer
materials with no large pores, and scheduling drying periods
37
-------
to avoid large rainfall just after start of drying), a separation
distance of a few hundred feet between a potable-water-
supply well and a land treatment system may be adequate,
at least with respect to pathogenic microorganisms. The
greater the distance, the better. Water from any well that
may have the slightest chance of containing renovated
sewage effluent should be routinely chlorinated or
otherwise disinfected prior to drinking.
Q. by K. Childs and Ken Sylvester. The soil zone beneath a
disposal system is a fixed volume with a fixed capacity for
waste assimilation. Many States require only 5 feet of
aerated zone beneath sites. Doesn 't this imply that each
system has a fixed (limited) filter life? Does waste filtration
take place in the saturated zone as well as the aerated zone?
A. Most biological and other renewable renovation processes
take place in the top few feet of the aerated zone. Thus, 5
feet will be adequate for most cases. Additional renovation
will take place under saturated conditions, except in very
coarse or fractured materials where the water flows through
large pores or channels.
Q. by Pete Atkins. You started your presentation by
indicating that land disposal meets the goal of "zero
discharge. " This is not completely true unless the renovated
water is completely recycled. Land disposal does not meet
the goal of zero discharge. Do you agree?
A. Land treatment meets the goal of zero discharge into
surface water. If the wastewater eventually drains back into
surface water via underground movement, it will do
so as "renovated" water and cause much less pollution
than the original wastewater.
Q. by Don Lundy. Assuming you have unsaturated flow in
the vadose zone below an infiltration pond, don't you need
to adjust the vertical hydraulic conductivity determined in
your saturated flow field test where you used Darcy's Law?
A. The "saturated" hydraulic conductivity values of vadose
zones yielded by field tests give an upper limit of infiltration
rates that might be achieved. So if the average vertical
hydraulic conductivity (harmonic mean!) is 0.5 m/day, we
know that we cannot push more than 0.5 m/day into the
ground. In actuality, infiltration rates will be less than the
maximum limit because of clogging at the surface. In that
case, the underlying material will be unsaturated with the
unsaturated hydraulic conductivity numerically equal to
the infiltration rate (assuming unit hydraulic gradient).
Q. by R. S. Domenowske. What is your experience with the
role of carbon nitrogen ratios on chemical and biological soil
processes that might affect treatment capacity?
A. C/N ratios in sewage effluent are important for
denitrification, which requires about 1 mg organic carbon
per mg nitrate nitrogen denitrified. Secondary effluent does
not have enough organic carbon for complete denitrification
for its nitrogen, particularly since some organic carbon will
already be used up by the time ammonium has been
converted to nitrate. Primary effluent has a more favorable
C/N ratio and more denitrification has been obtained with
primary effluent than with secondary effluent.
Q. What kind and degree of treatment of the wastewater is
required before putting it in the infiltration basins in order
not to decrease the infiltration rate of the basins and
hydraulic conductivity of the aquifer?
A. Most sewage effluents will have had secondary treatment
when used for rapid infiltration. Pre-sedimentation before
letting the effluent flow into the infiltration basin may be
necessary if suspended-solids concentrations are too high
(for example, above 20 mg/1). Total detention time in pre-
sedimentation and infiltration basins should be sufficiently
short to prevent excessive growth of algae. Where the
in-plant and land-treatment systems are to be built from
scratch, it may be advantageous to apply primary treatment
only and use well-settled primary effluent for infiltration.
This may reduce infiltration rates, but the in-plant treatment
would be much cheaper and denitrification in the soil
would be stimulated because the primary effluent has more
organic carbon.
Q. by George Clark and Steven Cordiviola. Generally
speaking, how long can a given plot of land be utilized
effectively as a filter?
A. For the renewable processes (removal of BOD, suspended
solids, bacteria and viruses, denitrification), there will be
essentially no end to the useful life of a land treatment
system. Adsorptive processes, which can remove phosphate
and metals, on the other hand, are effective over a limited
time. Adsorbed phosphate, however, slowly reverts to
insoluble forms like calcium phosphates. Metals can also be
precipitated. The precipitates may eventually accumulate in
soil pores and on soil particles to such a level that soil
permeability is reduced. This will take a long time, however.
Useful life of land treatment systems at carefully selected
sites should be on the order of decades and possibly even
centuries.
Q. by Mike Apgar. In the humid German climates where
high^rate infiltration was mentioned, what was the rate of
"overapplication," its effect on crops, clogging problems,
and TOC concentrations and/or specific organics identified
in the renovated water?
A. Annual application rates were 8 to 15m. Crop effects
must have been beneficial, because more farmers wanted to
get in on the project. Quality data on the renovated water
were not available.
Q. How does the cost of the percolation-pumping scheme for
Phoenix compare with that for chlorination and direct
discharge to canals? What about energy consumption? How
does percolation and pumping, etc. compare with other
means of removing bacteriological constituents?
A. Chlorination would probably be cheaper than renovation
by rapid infiltration. However, chlorination of secondary
effluent does not kill the viruses inside suspended solids.
Also, BOD and suspended solids contents of the chlorinated
effluent would be too high to meet State requirements for
unrestricted irrigation. In addition, it is desirable to remove
at least half of the nitrogen (to avoid adverse effects of
high-nitrate irrigation water on crops), more than 90% of the
phosphorus (to make the renovated water suitable for
recreational lakes), and to make the water aesthetically
acceptable for reuse in a densely populated area (renovated
water from a land treatment system looks crystal clear and
is free from taste and odor).
Q. by L. A. Swain. Can you conclusively say that in long-
term operation of the high-rate system the nitrate will not
38
-------
be taken out by the unsaturated zone up to a certain limit
and be leached out with subsequent infiltration at a higher
concentration than the inflowing water initially was?
A. Nitrates not denitrified during drying are flushed out by
the newly infiltrating water when a new flooding period is
started. Some of this nitrate may then be denitrified as it
mixes with the sewage water. The remaining nitrate then
moves downward with the renovated water and forms a
nitrate peak. Maximum nitrate concentrations in this peak
have exceeded total N concentrations in the sewage effluent.
However, with additional underground movement and lateral
flow, the peaks are attenuated to low "plateaus."
Q. by Roger Clissold. What is the population of Bielefeld
and what quantity of effluent is involved? Do you feel there
is an upper practical size for such a system ?
A. The population of Bielefeld was 260,000 in 1967. Water
use was 150 1 per person, so that effluent production
probably was around 35,000 m3/day or about 10 mgd. The
infiltration area initially was 50 ha (125 acres) with plans to
expand it to 300 ha (750 acres). Generally land treatment
systems are relatively small, but there are also some very
large systems like the Melbourne, Australia, system (20,000
acres serving 2 million people) and the Muskegon County,
Michigan, system (10,000 acres).
Q. by Roger Clissold. Phoenix system — does the sand and
gravel below the ponds become saturated from the bottom of
the pond down to the original water-table level at any time?
A. Saturation of the soil material below the infiltration
basins is unlikely because infiltration is restricted by a
clogged layer on the surface. The only way that saturation
could occur is through perching layers.
39
-------
Potential Replacement of Septic Tank
Drain Fields by Artificial Marsh
Wastewater Treatment Systemsa
by C. W. Fetter, Jr.b, W. E. SIoeyc and F. L. Spanglerc
ABSTRACT
Individual subsurface liquid waste disposal has been
cited as a source of ground-water contamination. Waste-
water treatment systems using emergent marsh vegetation
planted in a gravel substrate in a plastic-lined trench could
be used to treat septic tank effluent. A pilot plant treating
unchlorinated primary municipal effluent achieved the
following reductions in mass: BOD5-77%; COD-71%;
orthophosphate—35%; total phosphorus—37%; nitrate—22%;
coliform bacteria—99.9%. While such treatment is possible
only during the growing season, it could be useful at summer
cottages, camping areas, resorts and roadside rest areas.
Marsh treatment systems are inexpensive to operate and
virtually automatic.
ACKNOWLEDGEMENTS
This research was partially funded by the U.S.
Environmental Protection Agency under grant
number R-803794-01-0. The contents do not
necessarily reflect the views and policies of the EPA,
nor does mention of trade names or commercial
products constitute endorsement or recommenda-
tion for use. Additional funding was provided by
the Wisconsin Department of Natural Resources,
the East Central Wisconsin Regional Planning
Commission, and the University of Wisconsin/
Oshkosh. Laboratory analyses were performed
under the supervision of Kathleen Garfinkel.
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
b Associate Professor of Geology, University of
Wisconsin/Oshkosh, Oshkosh, WI 54901.
cAssociate Professor of Biology, University of
Wisconsin/Oshkosh, Oshkosh, WI 54901.
INTRODUCTION
Research activities in problems of small-scale
liquid waste disposal have tended to focus on either
improving the efficiency of subsurface disposal
systems or applying conventional treatment
methods (Bouma and others, 1972). Our research
has concentrated on developing a new method of
wastewater treatment using emergent aquatic
vegetation for the treatment of liquid domestic
wastes. Such treatment could be useful in summer
cottages, camping areas, resorts, and roadside rest
areas.
GROUND-WATER POLLUTION FROM
SEPTIC TANKS
The most common method of small-scale
liquid waste disposal is via septic tanks and soil
absorption drain fields or cesspools. The
septic tank removes solids from the waste by
settling and grease by flotation. Anaerobic
biological decomposition of the settled sludge
produces gas and releases soluble organic matter.
The effluent is high in BOD, bacteria, organic and
ammonia nitrogen, and phosphorus (Dudley and
Stephenson, 1973). It is similar in composition to
primary sewage effluent.
Soil absorption of effluent tends to remove
organic matter and bacteria by filtration, forming a
crust of organic matter, colloids and micro-
organisms below the seepage bed (Bouma and
others, 1972). Phosphate is generally removed by
adsorption on the soil or chemical precipitation
(Syers and Armstrong, 1971). Nitrogen is basically
conservative, with the ammonia and organic forms
being converted to the nitrate form by nitrifying
40
-------
bacteria although under some conditions denitrifica-
tion may occur (Lance, 1972).
Bouma and others (1972) list several condi-
tions under which soil absorption fields either fail
or could not be constructed: Formation of a thick
crust beneath the drain field which tends to partially
seal it. The soil surrounding the seepage bed has a
natural permeability which is too low. Bedrock,
either impermeable or creviced, is too close to the
surface for a drain field to be constructed. The
water table is too close to the surface. In addition,
soils which are too highly permeable (gravels) or
chemically unreactive may also not properly remove
wastewater constituents.
Under these conditions septic systems are
either not feasible or may fail and result in ground-
water contamination. There are numerous examples
of ground-water contamination due to soil absorp-
tion field failure (e.g. Fetter, 1974). In many areas
rural development is limited by poor soil conditions.
In some Wisconsin counties, more than 95% of the
land area is underlain by soil conditions which
preclude the use of drain fields.
Holding tanks which are pumped and the
effluent taken to a treatment plant are an alterna-
tive to soil absorption fields, but disposal by this
method is quite expensive.
ARTIFICIAL MARSH TREATMENT SYSTEMS
Aquatic vegetation has been used for waste-
water treatment in Europe with varying degrees of
success (Seidel, 1966, 1971a, 1971b, 1976; Althaus,
1966; Kickuth, 1969; Kok, 1974; de Jong, 1976).
However, it was not known if plant species found
in the north-central United States could be used.
Accordingly, our research initially focused on
selection of suitable North American species. We
also experimented with plant propagation methods
and evaluated various substate materials in terms
of suitability for plant growth.
As a result of these initial studies, we deter-
mined that the bulrush Scirpus validus could be
propagated from rhizomes and grew well when
planted in a shallow gravel substrate (Spangler,
Sloey and Fetter, 1976a).
Experimental Facilities
An experimental facility was constructed in
Seymour, Wisconsin (population 2,257) next to
the municipal wastewater treatment plant.
Artificial marshes were made by lining shallow
trenches with PVC plastic, filling the trenches
with gravel and hand planting vegetation. The
experiments described in this paper were made in a
pilot plant with a surface area of 85 m2. It was in a
trench 19.3 m long, 3.05 m wide at the bottom, and
5.8 m wide at the top. The trench had plywood
ends, was lined with 20 mil PVC plastic, and filled
with 15 cm of sand at the bottom, then 30 cm of
coarse gravel and 30 cm of pea gravel on the top
(Figure 1).
There were provisions for withdrawing either
unchlorinated primary or secondary effluent from
the municipal treatment plant, and adding metered
amounts to the basin through a pipe which allowed
a free fall of effluent.
The water from the artificial marsh was with-
drawn from the gravel substrate via a drain tile
across the lower end. Marsh effluent was also
metered. A smaller basin which had an area of
about 10m2 was used as a control. It had a gravel
substrate but no emergent vegetation (Spangler,
Sloey and Fetter, 1976b).
Several nests of small plastic pipes were
inserted vertically into the gravel to depths of 15,
30, 45 and 60 cm. These enabled the withdrawal
of water quality samples from various locations and
levels in the pilot plant (Figure 1).
Experimental Methods
Unchlorinated primary effluent was applied
to the pilot plant for the period August 21-
November 4, 1975. The vegetation and microbiota
of the marsh had been previously established by
two seasons of experiments with secondary
effluent.
Effluent was applied at an average rate of
2.5 m3/day which yields a loading factor of
29 1/m2 of surface area and a retention time of
about 10 days. Grab samples were collected once a
week from the marsh outlet and from the small
sampling wells in the gravel. A twenty-four hour
composite sample of the marsh influent was taken
by an automatic sampler. Prior experiments had
shown through intensive sampling that while there
PILOT PLANT
Metered |
Outlet .
•JL — PVC Liner
^ i m
V'
3" Drain tile
'across end
«— Typical
Plant
(Ty
vt
TWv Pea
) |)v Gravel
^Coarse
Gravel
Sand
Bulrush
Metered
Inlet — >,==
Sampling Tubes
|
1U 30cm
ul
| 60 cm
75cm
Fig. 1. Schematic diagram of artificial marsh pilot treatment
plant. (Not to scale. Total length 19.3 m.)
41
-------
was considerable chemical variation in the marsh
influent (wastewater treatment plant effluent),
there was little variation in the marsh effluent.
Samples were refrigerated and analyzed within 24
hours of collection by Environmental Protection
Agency (1971) methods.
Results
Summary results of the effectiveness of the
pilot plant are given in Table 1. The values given are
for concentrations of solute in mg I"1 Tables 2
and 3 give the water quality of influent and effluent
water respectively. During this period a total of
188.3 m3 of water drained from the marsh. Evapo-
transpiration amounted to 37.9 m3 during the
period or 0.44 m depth during the 75-day period.
The effect of evapotranspiration is to concentrate
the solute in the marsh effluent. However, as
rainfall was almost equal to evapotranspiration
during this period, this tendency was counteracted
by the dilution of the rainfall.
The following reductions in mass (corrected
for evapotranspiration) were achieved: BOD5—77%;
COD-71%; orthophosphate-35%; total
phosphorus-37%; nitrate-22%; coliform bacteria-
99.9%.
The design of the pilot plant tended to reduce
the effectiveness of the bulrush treatment. Much of
the wastewater passed through the coarse gravel
layer at the 30 to 60 cm depth. The bulrush roots
did not generally extend to that depth. The bulrush
exudes oxygen through its roots into the sediments
surrounding them. In experimental basins with
15 cm gravel, there was always dissolved oxygen in
the marsh effluent. In the pilot plant with the layer
of coarse gravel, there was zero dissolved oxygen
in the deeper gravel layers. However, the surface
gravels were aerobic, and therefore no noxious
odors developed. With a shallower gravel substrate,
greater treatment would be anticipated since the
sediments would be aerobic.
The deep layer of coarse gravel created
chemical stratification in the gravel. Figure 2
shows the mean concentrations of BOD5, COD,
orthophosphate, and total phosphorus at various
locations within the gravel. Concentrations are seen
Table 1. Effectiveness of Pilot Plant in Treating Primary Effluent - Summer 1975
Parameter
BOD
COD
Orthophosphate
Total phosphorus
Conductivity (/imhos)
Turbidity (JTU)
Nitrate N
Coliform (log col. 100 ml"1)
PH
Total solids
Suspended solids
Dissolved solids
Source
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Influent
Effluent
Mean
317
72
401
117
26
17
27
17
1330
1353
48
38
0.27
0.21
7.31
5.07
7.28
7.27
973
1156
42
32
932
1124
Std. Dev.
133
36
158
94
6
3
6
4
174
160
16
5
0.27
0.09
6.72
5.17
0.12
0.23
113
19
47
14
84
34
Number
10
9
12
9
12
9
12
9
12
9
12
9
10
9
7
6
12
9
3
2
3
2
3
2
Percent
reduction
—
77
_
71
—
35
—
37
—
-2
—
21
_
22
—
99.9
—
-
—
-19
—
23
_
-21
Mean values were not corrected for evapotranspiration.
" All values expressed as mg 1 * unless otherwise indicated in parentheses.
42
-------
Table 2. Water Quality of Primary Sewage Added to the Pilot Plant
1975 BOD" COD8 O-Pa
26 August >324 309.68 24.30
28 August >154 354.42 23.33
4 September 250 172.48 12.0
9 September 295 328.6 31.32
16 September 245 284.2 28.61
24 September 393 791.0 28.86
30 September 255 395.5 31.87
7 October 480 309.6 22.84
15 October 345 362.8 31.97
21 October 570 554.8 30.48
28 October 320 487.3 24.11
4 November 215 457.84 26.01
Mean 316.8 401.52 26.32
Standard Deviation 133.48 157.94 5.64
a concentrations in mg I"1
b micromhos
C J.T.U.
d log units
t-Pa
22.72
25.07
12.52
29.52
30.99
27.50
31.08
24.04
32.42
32.19
27.99
30.44
27.21
5.63
Cond.b
1080
1260
1310
1060
1440
1450
1425
1200
1300
1600
1180
1550
1330
174
Turbc
44
43
26
39
36
80
39
31
42.5
55.5
67
67
47.5
16.37
Colif.d
7.29
7.204
7.267
7.505
7.301
7.301
7.27
7.314
6.715
N03a
1.00
.04
0.22
0.22
0.34
0.10
0.24
0.21
0.25
0.00
0.36
0.271
0.267
pH
7.15
7.25
7.40
7.15
7.25
7.50
7.20
7.2
7.2
7.4
7.4
7.25
7.28
0.12
Table 3. Water Quality of Effluent from Marsh Treatment System Pilot Plant
1975 BODa CODa O-Pa
26 August 230.00 272.44 3.37
28 August NS NS NS
4 September NS NS NS
9 September 62 79.20
16 September 70 72.5 16.90
24 September 50 76.8 13.80
30 September 40 49.92 12.18
7 October 81 128.0 19.45
15 October 72 105.6 18.32
21 October 53 64.6 15.16
28 October 65 119 16.34
4 November 162 356.96 22.93
Mean 72.8 116.95 16.76
Standard Deviation 35.7 93.61 3.19
concentrations in mg 1
b micromhos
c J.T.U.
" log units
NS = No Sample
to decrease along the flow path from the influent
to the effluent end of the marsh. They also tend to
be higher in the stratum below the bulrush roots.
Other experiments indicated that the control
basin without emergent vegetation performed
almost as well as basins planted with bulrush in
terms of removal of BOD5, COD, total phosphorus,
orthophosphate, and coliform bacteria. Decomposi-
tion is primarily due to microflora and fauna
growing in the substrate as well as physicochemical
processes.
The bulrushes are important primarily for
aesthetic reasons and odor control, although they
do increase treatment efficiency by several percent.
t-Pa
3.41
NS
NS
16.97
13.45
11.83
19.67
19.67
14.85
15.47
23.59
16.68
3.68
E
* .30
0.
a .45
E
£.30
a.
a .45
Cond.b
1540
NS
NS
1050
1390
1400
1280
1400
1350
1400
1260
1650
1353
159
*±liL)ULJLu,
^Ms^H?
i?fif
- (130 M01 " >
- 157 181
i
Turbc
83
NS
NS
44
39
45
43
34
36
34
31.0
37
38.11
4.96
jllji
1^Htl"f™
i/f
87 |
'
134 '
a
BOD mg. I'1
-^LjULJUs,
^raW^ml
- Y25 " 18 M H
26 22
Orthophosphate m
Colif.d
5.158
NS
NS
4.477
5.097
5.491
5.580
5.301
4.301
5.249
5.168
^40^ |[ |
^"SiiiitljWj
:- fW
- .30- Y231 '1
3
3 .45- 418
N03a
.37
NS
NS
0.28
0.39
0.18
0.20
0.22
0.19
0.08
0.18
0.22
.094
JLJlJjL
Hfl^'ySli
159 VI I 123J
541 238
COD me. I'1
rw7
5 '
15 I h
1 0
u
J
rwswIUU
' /J-^mv^.V^
!- j fff
- .30[- ^25 'M
i
i
.45- 25
^JUJQUL
refly^M
If (
18 V \ 16 '|
25 19
a
pH
7.05
NS
NS
7.05
7.20
7.10
7.12
7.3
7.10
7.78
7.48
7.30
7.27
0.23
4*-
117
i
u>
17
1
g. 1" Total Phosphorus mg. 1
Fig. 2. Mean concentrations of BOD5, COD, orthophosphate
and total phosphorus, in influent, effluent and at 30 and 45
cm depth in artificial
marsh pilot treatment
plant receiving
primary municipal sewage effluent.
43
-------
It is important to design marsh treatment systems
such that the water level is below the gravel surface.
If standing water is not present, mosquitoes and
similar insect pests cannot breed.
Wisconsin winters tend to be severe and our
marsh systems were frozen from about December
into March. During these periods the green parts of
the bulrushes died. In spring new shoots grew from
the rhizomes of the plant. Each season the number
of bulrush shoots increased as the plants spread by
growth of the rhizomes.
Since the artificial marshes are frozen during
the winter, they cannot function as treatment
systems. However, there are many situations where
lack of treatment capacity during the winter is not
critical. Indeed, in areas which do not experience
killing frosts, research might reveal that marsh
treatment systems could function all year.
POTENTIAL FOR SMALL-SCALE
WASTE TREATMENT
Septic tank effluent has undergone mechanical
settling and some anaerobic decomposition. It is
chemically and bacteriologically similar to the
primary sewage effluent used in our experiments. A
small artificial marsh should yield results similar to
our pilot plant (Table 1). Table 3 gives the chemical
and bacteriological characteristics of the marsh
effluent.
An artificial marsh suitable for a single-family
summer home in Wisconsin would require a sloping
trench lined with 20 mil plastic and filled with
15 cm of pea gravel. Initial studies indicate this
should be 5 m wide and 30m long although further
research should be done to determine if the size
could be reduced. The degree of slope would be
determined by the hydraulic conductivity of the
gravel. If the loading rate is 1.5 mVday (400 gal/
day) and the conductivity is 41 m/day (1000 gal/
day/ft2), then the required slope would be 0.05.
Bulrush rhizomes would be planted about 0.5 m
on center. A sample design is shown in Figure 3.
A wet well would collect septic tank effluent
and a sump pump would periodically flood the
marsh with effluent. Evapotranspiration would
reduce the volume of the wastewater. In our
experiments this amounted to about 1.16 cm/day
during July 17-August 21. This amount varies with
weather, and is offset by precipitation into the
trench. During peak evapotranspiration periods,
a 5 m X 30 trench could lose up to 1.7 m3/day.
Under these conditions discharge from the marsh
could fall to zero.
Following marsh treatment the water could be
PRODUCTION PRETREATMENT
TREATMENT
DISPOSAL
Bulrushes m soil absorption field
Gravel r> A §
Fig. 3. Proposed design of artificial marsh treatment system
for small-scale waste treatment.
discharged into the subsurface through conventional
drain fields or cesspools with much less danger of
ground-water contamination. In sandy soils part of
the trench might be unlined and the effluent simply
allowed to seep into the ground.
One of the significant ground-water pollutants
is nitrate. The primary effluent we were treating
was very low in nitrogen. While the marsh effluent
had only 0.22 mg I"1 nitrate, it also had only 0.29
mg I"1 nitrate in the influent. Complete nitrogen
analyses were not made and this would be a
necessary area of future research.
Much more research is needed on small-scale
marsh treatment systems to thoroughly evaluate
their potential. The next step should be the
construction and testing of some actual systems
using septic wastes.
Marsh systems have the potential to provide
an inexpensive individual treatment system which
is virtually automatic and has few maintenance
requirements. They could significantly reduce
ground-water contamination from individual waste
disposal systems.
REFERENCES
Althaus, H. 1966. Biological waste water treatment with
bulrushes. Das Gasund Wasserfach. v. 107, pp.
486-488.
Bouma, J., W. A. Ziebell, W. G. Walker, P. G. Olcott, E.
McCoy and F. D. Hole. 1972. Soil absorption of
septic tank effluent. Wisconsin Geological and
Natural History Survey. Information Circular 20,
235pp.
Dudley, J. G. and D. A. Stephenson. 1973. Nutrient
enrichment of ground water from septic tank disposal
systems. University of Wisconsin. Inland Lake Renewal
and Shoreline Management Demonstration Project
Report, 131 pp.
Environmental Protection Agency. 1971. Methods for
chemical analysis of water and wastes. U.S. Environ-
mental Protection Agency, 312 pp.
Fetter, C. W., Jr. 1974. Water quality and pollution — South
44
-------
Fork of Long Island, New York. Water Resources
Bulletin, v. 10, pp. 779-788.
de Jong, Joost. 1976. The purification of wastewater with
the aid of rush or reed ponds in Biological Control
of Water Pollution. J. Tourbier and R. W. Pierson,
Jr. Ed. Univ. of Pennsylvania Press, pp. 133-140.
Kickuth, R. 1969. Higher water plants and water purifica-
tion: Ecochemical effects of higher plants and their
functions in water purification. Schriften-reihe der
Vereiningung Deutscher Gerwasserschutz VDG. v. 19.
Kok, T. 1974. The purification of sewage from a camping
site with the aid of a bulrush pond. H2O. v. 7,
pp. 536-544.
Lance, J. C. 1972. Nitrogen removal by soil mechanisms.
Journal Water Pollution Control Federation, v. 44,
p. 1352.
Seidel, Kathe. 1966. Reinigung von Gewass durch hohere
pflanzen. Naturwissenschaften. v. 53, no. 12,
pp. 289-297.
Seidel, Kathe. 197 la. Macrophytes as functional elements
in the environment of man. Hydrobiologia. v. 12,
pp. 121-30.
Seidel, Kathe. 197Ib. Wirkung hohere pflanzen auf
pathogene keime in gewassern. Naturwissenschaflen.
v. 58, pp. 150-51.
Seidel, Kathe. 1976. Macrophytes and water purification
in Biological Control of Water Pollution. J. Tourbier
and R. W. Pierson, Jr. Ed. Univ. of Pennsylvania
Press, pp. 109-121.
Spangler, F. L., W. E. Sloey and C. W. Fetter, Jr. 1976a.
Experimental use of emergent vegetation in the
biological treatment of municipal wastewater in
Wisconsin in Biological Control of Water Pollution.
J. Tourbier and R. W. Pierson, Jr. Ed. Univ. of
Pennsylvania Press, pp. 161-171.
Spangler, F. L., W. E. Sloey and C. W. Fetter, Jr. 1976b.
Artificial and natural marshes as wastewater treatment
systems in Wisconsin in Freshwater Wetlands and
Sewage Effluent Disposal. Symposium Proceedings.
University of Michigan, Ann Arbor, pp. 215-240.
Syers, J. K. and D. E. Armstrong. 1971. Chemical
characteristics of phosphorus in soils and water in
Proceedings of conferences on farm animal wastes,
nitrates and phosphates in rural Wisconsin ecosystems.
T. J. Brevik and M. T. Beatty, ed. Univ. of Wisconsin-
Extension, Madison, p. 138.
DISCUSSION
The following questions were answered by C. W.
Fetter after delivering his talk entitled "Potential
Replacement of Septic Tank Drain Fields by
Artificial Marsh Wastewater Treatment Systems."
Q. What were the typical input and output coliform
densities per milliliter? Was the marsh effluent ever examined
for bacterial pathogens, virus or stable organics?
A. The average influent value was 7,314 log coliform units
and the effluent had an average of 5.249 log coliform units.
The effluent value is lower than the average coliform count
of a stream which we studied in northeastern Wisconsin.
The stream received a chlorinated secondary effluent and
had an average of 4.76 log coliform units above the outfall
and 5.54 log coliform units 100 meters below the outfall.
Incidentally, following passage through 1800 meters of
marsh, the coliform count in the stream then dropped to
4.68 log units.
We made no analyses of pathogenic bacteria, virus or
stable organics. Some published reports indicate some species
of Scirpus can degrade phenol (Kickuth, R., 1969).
Q. What caused the increase in total dissolved solids?
A. We don't know, but we noticed a similar increase in
dissolved solids in a study of a natural marsh. This could
be due to an increase in dissolved organic acids from the
marsh plants.
Q. Was the pilot plant operated under saturated conditions?
A. Yes, water levels were kept at the gravel surface.
Q. What is the life expectancy for a marsh treatment system
for a single family dwelling? How long would the PVC liner
last?
A. The life expectancy of a marsh treatment system depends
upon the suspended solids which accumulate in the gravel.
We have no data on how long this process might take.
The PVC lining the basin was necessary for the
experimental design since we didn't want unmetered water
seeping from the experimental basin. In soil types which
are not highly permeable, a basin liner would not be
necessary. The bulrushes should not have the roots dry for
extended periods, so the basins should be able to retain some
water. A clay or silt lining would be better than a PVC
lining for home systems since it would aid in phosphorus
removal.
Q. Are there other plants which could function like
bulrushes? Could these be effective in arid climates?
A. There are several species of bulrushes—S. lacustris, S.
validus, S. acutus and S. fluviatilis. The cattail, Typha,
functions in a similar manner, but is difficult to
transplant. Dr. Kathe Seidel (1976) of the Max Planck
Institute of Krefeld, West Germany, has experimented with
a number of European plants. There has also been some
work at the N.A.S.A. National Space Technology
45
-------
Laboratories using water hyacinth (Eicbhornia crassipes)
and alligator weeds (Alternanthera philoxeroides) in
biological treatment systems. If these plants are maintained
with the roots in a saturated environment, they would grow
in arid climates.
Q. What is the impact of heavy metals on the system?
A. We did not study heavy metal uptake. The wastewater
source did not contain any industries which produce a
heavy-metal-rich effluent. Other studies have shown that
water hyacinths and alligator weeds do remove heavy
metals and could be used in a heavy-metal treatment
system if they are harvested (Wolverton, Barlow and
McDonald, 1976, Chapter 17, in Biological Control of
Water Pollution, University of Pennsylvania Press).
Q. Could winter treatment be continued in northern
climates if warm effluent is used?
A. Not if there is a long retention time. A marsh system
with standing water could be made much smaller, and
enclosed in a greenhouse. This type of system would no
doubt function, but we don't know the treatment efficiencies
with lower temperatures and shorter daylight periods.
Q. Could you recommend this system for institutions, such
as schools, which have intermittent loads?
A. Yes, if the bulrush roots are kept moist during periods
when the system is not in operation.
Q. Was there a control bulrush group which did not have
effluent applied?
A. Not as such, although my coworkers have done some
studies of growth rates in natural, unpolluted marshes.
Q. Does this system differ from a sewage lagoon treatment
system?
A. Yes, since sewage lagoons often have anaerobic areas
with associated odor problems.
46
-------
Ground-Water Pollution Aspects of Land Disposal
of Sewage from Remote Recreation Areas"
by Nils Johnson0 and Dean H. Uriec
ABSTRACT
Local ground-water pollution problems are often
associated with disposal of sewage from remote recreational
areas. Ground water in these areas is routinely used for
domestic purposes without treatment. Sewage treatment
facilities are often prohibitively expensive. Transporting
sewage to municipal sewage treatment facilities is often
equally costly, or adequate facilities are not available.
A soil incorporation method was tested at two field
sites in Michigan's Upper Peninsula to evaluate its impact
on ground-water quality. Liquefied campground sewage
was injected at 15 cm (6 in.) depths using a liquid manure
system. Average dosage levels were 2.7 metric tons per
hectare (1.2 ton/acre) of dry sewage solids. The application
rate was equivalent to a fertilization rate of 116 kg/ha
(104 Ib/a) of total nitrogen. Field tests were conducted on
adjacent strips of Kalkaska sand soils in 1973, 1974, and
1975.
Changes in ground-water quality in the zone
immediately beneath the water table were evaluated by
analyzing samples from randomly located wells. Test wells
were located directly beneath the treated area and up to
30.5 m (100 ft.) away in the direction of ground-water flow.
Nitrate levels in ground water at 3.6 m (12 ft.) depths were
higher than control levels from the treated zone to the
limit of the test well sampling. Levels of nitrate did not
exceed the limits for potable water (10 mg/1 NO3-N). No
fecal coliform organisms were detected in ground-water
samples. Laboratory tests of the filtration capability of
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
bporest Hydrologist, Hiawatha National Forest,
Escanaba, Michigan 49829.
cPrincipal Hydrologist, North Central Forest Experi-
ment Station, USDA-Forest Service, East Lansing, Michigan
48823.
the A and B horizons indicated a high level of filtration of
polio virus.
Fertilization with sewage increased the biomass of
native herbaceous vegetation by 410 percent. Nitrogen
content of the treated vegetation was 63 percent above
controls. Approximately 48 percent of the added nitrogen
fertilizer was incorporated in herbaceous foliage at the end
of the first growing season.
Site selection guidelines for use by National Forests
include remote location of incorporation sites, no potable-
water sources within one-half mile downgradient from the
site, infiltration rates between 5 and 25 inches per hour,
slopes less than 5 percent, and dosage rates which will not
exceed 50 kg/ha (56 Ib/a) of mineralized nitrogen per year.
Injection depths should provide complete soil coverage
but sewage should be placed above the B horizon for
maximum use of nutrients by plants.
The soil incorporation method has been approved for
selected sites by the Michigan Department of Natural
Resources and the U.S. Environmental Protection Agency,
with ground-water monitoring required at each site.
INTRODUCTION
The simplest and oldest way to dispose of
sanitary wastes is to drop them on the soil and let
nature take its course. This fact is derived from the
most casual observations of animal and primitive
man. When modern civilizations result in concen-
tration of wastes, the mechanics become much
more complex and the environmental hazards more
serious.
Disposal of human waste on the Hiawatha
National Forest in northern Michigan required a
method that was inexpensive and yet gave satis-
factory renovation. The volume of waste from
campground vault toilets located throughout the
Forest is approximately 15 to 20 thousand gallons
per year.
Vault toilets were installed on the Hiawatha
47
-------
National Forest because of the extremely sandy
soils located in the area. These soils have subsoils
with very rapid percolation rates and little capacity
to renovate wastes. To protect ground water and
lake quality, sealed vaults were installed on the
Forest. These vault toilets on the Hiawatha
National Forest are sealed concrete tanks, having a
volume of approximately 500 to 1,000 gallons. A
building containing a toilet is placed on top of the
vault. This dry system requires periodic pumping
with a standard septic tank pumper to empty the
vaults.
Several alternative methods were available to
dispose of the vault sewage. The wastes could be
spread on the ground surface where biodegradation
would occur. This would leave human waste
exposed however. U.S. government land adminis-
tered by the U.S. Forest Service is open to public
access. For public health reasons, this alternative
was rejected.
The waste could be taken to municipal waste
treatment plants located adjacent to the Forest.
This alternative was rejected for two reasons. First,
the plants do not have holding lagoons, and a single
dose application of concentrated waste would upset
the biological operation of the plants. Second, this
method of disposal would require expensive, long
hauls because of the remote locations of
campgrounds.
The third alternative was to build a waste-
water treatment facility to dispose of toilet wastes.
This is a viable alternative; however, initial invest-
ment is high and the system would require an
operator and continuing maintenance.
GENERAL DESCRIPTION OF THE AREA
The two incorporation sites are located in
Alger and Delta Counties of Michigan's Upper
Peninsula. The aquifers affected by the tests are
sandy drift deposits of Wisconsin Age that are
approximately 15m (50 ft.) thick overlying
limestone of Ordovician Age. Ground-water levels
are between 3-5 m (10-16 ft.) below the ground
surface. Soils in the test areas are sandy, mixed,
frigid, Typic Haplorthods and have been used for
agriculture. Soil cation exchange capacities are
about 5-10 m.e./lOO g. The average rainfall is
813 mm (32 in.) and is evenly distributed through-
out the year. The climate is continental with a
moderating lake effect.
Many ground-water wells in the area are
shallow sand points and thus protection of the
quality of ground water in the shallow aquifer is
of paramount importance.
METHODS
A system referred to as soil incorporation
was initiated for waste disposal. Wastes were
injected at a depth of about 15 cm (6 in.) below
the ground surface. Renovation of the waste is by
micro-organisms present in the soil, filtration and
adsorption by soil materials above the ground-
water table, and nutrient uptake by the vegetation
present on the incorporation site.
A series of test wells consisting of 3.18 cm
(1.25 in.) 80 mesh well screens were placed just
below the water table. Porous ceramic samplers
were also placed in the upper three feet of the
saturated zone. These wells and samplers were
periodically sampled for analysis of ground-water
quality. All water samples were preserved and
analyzed according to Standard Methods (APHA,
1971).
Soil samples were collected by the use of a
punch auger. Control and treatment samples were
composited and analyzed by micro-kjeldahl and
emission spectrolysis.
Vegetation for analysis was collected using a
929 cm2 (1 ft.2) sampler. Samples were dried and
ground for analysis. Analysis of total N was by
micro-kjeldahl. Cations were analyzed by
emission spectrolysis.
Ground-water flow direction was determined
by the construction of a flow net using four
observation wells. Wells for analysis of ground-
water quality were located at incremental distances
in the direction of ground-water flow. Lateral
locations were randomly selected within the
interior 50 percent of the zone of expected
fertilizer effect.
KENTUCKY CCC SITE-CRITICAL TEST
A preliminary test began in 1972 on the
Munising Ranger District of the Hiawatha National
Forest in northern Michigan. The site presented a
relatively high risk of ground-water contamination
because of (1) highly permeable sand soil of the
Kalkaska series, and (2) water tables at only 1.8-2.4
m (6-8 ft.) depths.
A 0.08 ha (0.2 acre) plot was injected with
5,300 1 (1,400 gal.) of campground vault wastes in
August 1972, at the rate of 6.2 1/m (0.5 gal/ft.) of
furrow. Furrows were spaced 0.69 m (2.25 ft.)
apart. The sewage was injected at a 15 cm (6 in.)
depth using a Clay Honeywagon tank trailer
equipped with two injector shoes which opened
two furrows, injected the sewage and closed the
furrows in one operation. (Mention of product
names does not indicate endorsement by the U.S.
48
-------
Table 1. Composition and Application Rates of Sewage
Vault Wastes Injected at 65,500 I/ha at Kentucky CCC
+N0250 —
mg/1
Total solids
Total N
Total P
Concentration
2.86%
13. 17% (dry wt.)
2.0% (dry wt.)
Dosage Rate
1869 kg/ha (1668 Ib/a)
224 kg/ha (200 Ib/a)
34kg/ha(301b/a)
Forest Service.) Composition and dosage rates of
the vault sewage are shown in Table 1.
RESULTS OF NUTRIENT STUDY
Monitoring wells were installed (Figure 1).
After two years of monitoring, no elevated nitrogen
levels have been found at either FC-2 or W-3 located
6.1 m and 12.2 m (20 and 40 ft.) down the
ground-water gradient, respectively. No significant
phosphorus increases have been detected in any
ground-water samples.
High nitrogen values are found only directly
under the plot and adjacent to the plot. Increased
nitrogen values have not been detected in wells
located further down the ground-water gradient.
Nitrate nitrogen rose to a peak of about
50 mg/1 in wells directly under the treated plot
following snow melt in 1973, but dropped
sharply within six months. Nitrate and nitrite
nitrogen accounted for practically all the total
nitrogen as would be expected under conditions of
nitrification. The increase in nitrogen levels
following treatment is attributed to the fact that
the (1) water table is only 1.8-2.4 m (6-8 ft.) below
the surface, (2) Kalkaska sand soil has a rapid
infiltration rate, and (3) nitrogen content analysis
of the waste was 1.5 times higher than average
W-WELL POINT
FC-FILTER CANDLE
INFILTROMETER
TEST
SOIL
INCORPORATION,
TEST
§ FC!S,5,IO,IS ft
(1.5,3,6 M)
50 100ft
H 1
W-l LOCATED WITHIN TREATMENT AREA
W-Z LOCATED ON EDGE OF TREATMENT PLOT
IN DIRECTION OF GROUNDWATER FLOW
FC-2 LOCATED 6-OM FROM EDGE OF PLOT
IN DIRECTION OF GROUNDWATER FLOW
W-3 LOCATED I2CMFHOM EDGE OF PLOT
IN DIRECTION OF GROUNDWATER FLOW
C CONTROL WELL
BECAUSE OF A LARGE AMOUNT
OF DATA ONLY REPRESENTATIVE
VALUES ARE SHOWN
Fig. 1. Location of sewage treatment tests and ground-
water quality sampling points, Kentucky CCC test.
Fig. 2. Nitrite plus nitrate nitrogen levels in ground water
following sewage injection at Kentucky CCC Site.
values measured in subsequent treatments. Much
of the horizon development of the soil has been
disturbed during occupancy of the site by a
Civilian Conservation Corps Camp during the
1930's. The relatively high nitrate levels could be
attributed to poorer soil-nutrient relationships
resulting from this disturbance.
BACTERIOLOGICAL RESULTS
Through a cooperative study with Michigan
Technological University, by Tluczek (1974), the
movement of the fecal bacterial indicators was
monitored.
The raw waste was characterized as follows:
5.52 X 107 Total Coliforms (TQ/100 ml
1.74 X 107 Fecal Coliforms (FC)/100 ml
3.28 X 107 Enterococci (Streptococcus
facalis) (EC)/100 ml
A circular 2.52 square meter (25 ft.2) single
ring infiltrometer was placed over two injection
furrows and irrigated with 161 mm (6.35 in.) of tap
water at the rate of 380 mm (15 in.) per hour.
Under the artificial rainfall condition, total
and fecal coliform and enterococci organisms were
detectable at 0, 1.5, 3.1, 4.6 m (0, 5,10, and 15 ft.)
down the ground-water gradient directly after
treatment. Enterococci dropped to zero one day
after injection and remained at zero at the two
most distant wells. Enterococci were detected at
low levels for one month after treatment at 0 and
1.5 m (0 and 5 ft.). No enterococci were detected
in any ground-water samples after this date.
Fecal coliforms were detected at 0, 1.5, 3.1,
and 4.6 m (0, 5, 10, and 15 ft.) down the ground-
water gradient directly after treatment (August
1972) through the first week of June 1973. Counts
ranged from approximately 1 to 9 organisms per
100 milliliters. Highest counts occurred directly
under and adjacent to the plot (0-1.5 m).
Total coliform organisms were detected at
49
-------
\
H-Z
35 M (775ft)
ALIGNMENT OF FURROWS
/
XS?S I973VAULT WASTE
^S? TREATMENT AREA 0 Oo i
/ "^ °0b /£
III -FC /£
fflXn 1
W-3 / <
oCW-l
/I65M (550ft)
£ olSMb(Sftb) OI-^STO«/g \2IOM(700ftJ
5 06 .OM $ \
? 0l2M(40ft) (20ff)/£ CW-4
030M (100ft.)
Table 2. Composition and Dosage Rate of Vault Wastes
Injected at 65,500 I/ha at Birch Farm
Dosage Rate
Concentration Ibs/acre (kg/ha)
Total solids
Total N
NH3-N
Total P
K
Ca
Mg
2.02%
8.83% (dry wt.)
7.63% (dry wt.)
1.37% (dry wt.)
7.7% (drywt.)
2.6% (drywt.)
0.1% (drywt.)
1,173
104
90
16
90
31
1
1,314
117
101
18
101
35
1
0 8 15 22 30 M
I I I I I
0 25 5O 75 100ft
Fig. 3. Sewage treatment test area and location of ground-
water quality sampling points. Birch Farm test.
all treatment wells directly after treatment on
August 17, 1972, until the middle of December
1972.
Elution and localization studies of
reconstructed soil columns showed that bacteria
were mechanically filtered from the effluent by
the somewhat finer-textured soil horizons ( A & B
horizon).
Longevity of fecal coliform and enterococci
organisms in the soil at the Kentucky CCC Site
showed that in approximately one year, fecal
coliforms were down to 42 organisms per gram of
soil and enterococci were zero.
BIRCH FARM PILOT TEST
During the summer of 1973, a project sized
test using vault-toilet waste from three ranger
districts was begun.
The procedure was similar to that used at the
Kentucky CCC Site. A large open field, Birch Farm,
occupied by soil of the Kalkaska sand series with
the water table at approximately 3.6 m (12 ft.)
below the surface was used. The site received
approximately 53,000 1 (14,000 gal.) of waste on a
0.81 ha (2 acre) tract during June through
September 1973, and approximately 18,900 1
(5,000 gal.) on 0.4 ha (1 acre) during June through
September of 1974, and a similar amount in the
summer of 1975. Monitoring of the site was
accomplished by the design shown in Figure 3.
RESULTS OF BIRCH FARM
NUTRIENTS STUDY
Dosage Rates
Analysis of the sewage has shown a high
variability in the total solids content from 3.2 to
0.4 percent. This variability is due to the amount
of water added to pick up the sewage from the
sewage vaults, and the presence or absence of
ground-water seepage among the various camp-
ground locations. Average concentrations and
nutrient dosage rates are shown in Table 2.
Total nitrogen and nitrite plus nitrate as
nitrogen rose slightly directly under and
adjacent to the plot. There have been no increases
detected at 40 feet from the edge of the plot down
the ground-water gradient (Figure 4). Although
nitrate plus nitrite levels rose to a maximum of
Mg/l
10
8
fi
4
2
0
CONTROL
BENEATH
AREA
— ^
s-'^^ t
^"" \ J
' 1 1
TREATED
^ .
N
1
_ 1973
10
8
6
4
2
0
10
8
6
4
2
0
1974
1975
1976
1.5 M(5.Oft)
3.0M(IO.Oft)
_ 1973
1974
1975
1976
6.OM( 20.0ft)
I 2.0M (40.0ft)
30.0M(IOO.Oft)
1973
1974
1975
1976
Fig. 4. Concentration of total nitrogen in ground-water
samples removed from wells at Birch Farm test site.
50
-------
1.5 r-
1.0
• SEWAGE APPLIED •
CONTROL
BENEATH TREATED
AREA
\5
,_ 1975
Mg/1
.5
\5
1.0
1974
1975
1976
I. 5 M (5.0f1)
3.0 M (1.0ft)
r- 1973
1974
1975
1976
• 6.0 M 020.0ft.)
12 OM (40.0ft.)
30.0 M (100.0ft)
1973
1974
1975
1976
Fig. 5. Total dissolved phosphorus in ground-water samples
removed from wells at Birch Farm test site.
8.9 mg/1 as nitrogen, they did not reach the U.S.
Public Health Service Standard of 10 mg/1. Total
dissolved phosphorus as PO4 has remained near
background levels as would be expected on sand
soils with a high ability to absorb phosphorus
(Figure 5).
BIRCH FARM BACTERIOLOGICAL RESULTS
The waste contained the following bacterial
population ranges:
Total Coliforms (TC)/100 ml
Fecal Coliforms (FC)/100 ml
Enterococci (St. facalis)
(EO/100 ml
Measurement of fecal and enterococci
organisms indicated that no enteric organisms
reached the ground water at any of the observation
wells.
The absence of bacteria in ground water at the
Birch Farm Site versus positive findings at the
Kentucky CCC Site can be attributed to a number
of variations in the sites and experimental technique.
The water table at Birch Farm compared to
the Kentucky CCC Site is 3.0 m and 1.8 m (10 ft.
and 6 ft.) respectively. This provided a thicker
l.OX 105-2.4X 106
1.08 X 104-2.6X 104
2.8 X 105-4.0X 10s
filter for bacteria removal at Birch Farm. The soil
at Birch Farm has more fine-textured particles and
a higher organic matter content allowing for more
bacteria to be mechanically filtered in the upper
horizons of the soil.
Comparing the two sites, the Kentucky CCC
Site was artificially irrigated and the Birch Farm
Site was not. Thus more downward movement of
water allowed bacteria to reach the ground-water
table at the Kentucky CCC Site. The Kentucky
CCC Site had much soil disturbance in the past;
the Birch Farm Site has not been disturbed. Soil-
fauna-nutrient relations are thus much better at
the Birch Farm Site.
VIROLOGY STUDY
Laboratory studies of virus-innoculated 48 cm
soil columns showed that soil columns flushed with
one large volume of water were able to retain
greater than 99.99% of the T4 bacteriophage and
99.9% of the poliovirus.
The study concluded that there is only a
low-grade danger of virus contamination of ground
water. The studies were conducted in reconstructed
soil columns only 48 cm in length. Under actual
site conditions where there is approximately 3
meters of soil above the water table, contamination
results would be expected to be proportionally
lower than in laboratory studies.
VEGETATIVE RESPONSE
The operation of the system is highly
dependent on utilization of the nutrients by the
vegetation that is present on the site. Table 3 shows
the levels of foliar elements of the treatment versus
the control plots in August 1975.
Total nitrogen concentrations in plant tissues
increased significantly following fertilization with
vault wastes:
July 1974
August 1974
Birch Farm
Treated Control
1.88% 1.46%
1.24% 0.76%
Kentucky CCC
Treated Control
1.92% 0.98%
1.06% 0.94%
The volume of dry matter on the site in above-
ground herbaceous vegetation increased an average
of 410 percent on injected areas. During July 1974,
approximately 48 percent of the nitrogen added as
vault wastes was incorporated in the increased
biomass.
SOIL FERTILITY EFFECTS
The Kentucky test site with its disturbed
topsoil condition presented extremely variable
51
-------
Table 3. Concentration of Nutrient and Mineral Elements in Plant Tissue
on Soil Incorporation Plots (Mean of 5 Samples)
Phosphorus %
Potassium %
Calcium %
Magnesium %
Aluminum ppm
Iron ppm
Zinc ppm
Copper ppm
Manganese ppm
Boron ppm
Kentucky CCC
Treated Control "t"
.184**
1.48**
2.11
.098*
33*
109*
44.7*
6.44
152.5
4.22*
.136
.72
.221
.076
134
221
52.6
9.18
270.1
6.24
6.16
9.8
.37
2.85
2.45
2.34
2.38
1.31
1.95
2.67
Birch Farm
Treated Control
.177**
1.65**
.388
.128
119
162**
55.1*
4.96
61.4
6.80
.118
.80
.453
.115
406
321
74.8
8.8
198.3
9.34
"t"
4.3
4.4
0.8
.89
1.75
4.2
2.83
1.5
1.15
1.08
Significance — * 5% level
** 1% level
materials for analysis. The cation exchange
capacity was also very low (Table 4). Injection of
vault wastes generally resulted in increased total
soil nitrogen values. Potassium values were
consistently higher and bases were reduced,
probably due to slightly lower pH values and
displacement by ammonia from the injected
sewage.
DISCUSSION
The soil incorporation of vault toilet wastes
has proved to be a reliable, low-cost method for
the disposal of concentrated wastes from
recreation areas on the Hiawatha National Forest.
The annual cost of disposing of the 57,000 1
(15,000 gal.) of wastes from the three ranger
districts is approximately $2,000 per year.
Bacteria and nutrient results at the Kentucky
CCC Site indicated that bacterial and nutrient
contamination of the ground water can be
expected under extreme conditions. These condi-
tions are (1) poor soil development due to past
disturbances resulting in less than optimum uptake
of available nutrients from the waste, (2) relatively
shallow water tables, (3) high soil flushing rates
with water after incorporation of wastes.
Virus contamination of ground water resulting
from soil incorporation is a relatively remote
hazard when the guidelines in the following section
are followed.
Bacteria and nutrient studies from a pilot
study conducted at Birch Farm gave results that
indicated when a site is selected that has (1) a deep
ground-water table, (2) good soil development,
and (3) operates under normal precipitation
conditions, the water quality of the ground water
will not be appreciably affected.
Good soil-nutrient relationships are a must if
the site is to renovate the waste and the vegetation
is to utilize the resulting available forms of
nitrogen. An indication of the nutrients that are
made available from the incorporation of waste, is
the great increase in vegetation growth that has
been observed following incorporation of waste at
a site.
The system depends on filtration by the soil
material above the ground-water table and nutrients
from the waste being used by existing or planted
vegetation. With these thoughts in mind and the
results of the studies on the Hiawatha National
Table 4. Soil Nutrient Analysis in Sewage Injection and Control Sites — Kg/ha
Date
Kentuck
8/74
4/75
10/75
Total N Total P Potassium Calcium Magnesium C.E.C. pH
- Kg/ha - meq./l 00 g.~--
Treat Cont Treat Cont Treat Cont Treat Cont Treat Cont Treat Cont Treat Cont
yCCC:
835 610
860 1230
990
Birch Farm:
8/74
4/75
10/75
1900 1560
1820 1790
2170 1635
78 72
43 58
47 81
39 44
54 27
27 14
81 54
92 49
307 460
156 460
461 615
435 409
25 37
12 52
25 12
38 38
3.3
14.4 16.2
5.6 5.9
5.4 5.9
5.2
5.9 5.9
5.6 5.6
52
-------
Forest, the following Guidelines for Use are
proposed to guide future and existing soil
incorporation projects.
GUIDELINES FOR USE
1. Injection depth should be between 10 to
20 cm (4 to 8 inches). This depth is required to
take advantage of favorable chemical and
bacteriological relationships in the surface soil
horizons necessary for renovation of vault wastes,
and to provide covering of the waste.
2. The site is to be located at least 180 m
(200 yd.) from any residence or regular travel
lane. This is the distance from residences required
for disposal of wastes from septic tanks in Michigan.
This is a minimum and in actuality should be a
minimum of a half mile. The ability of this system
to be used in remote areas is an advantage over
other systems.
3. Application rate is not to exceed 6.2 1
(1.6 gal.) of waste per meter (3.3 ft.) of furrow at
0.69 m (2.25 ft.) spacing. This was the rate and
spacing used on the Hiawatha National Forest and
proved to be a workable dosage. These values
could possibly vary in other areas based on
equipment used and soil capabilities.
4. Site should be nearly level with no
drainage ways within the area or immediately
adjacent to the injection area. General topography
should have a maximum slope of 5.0% and should
not have erratic undulations. This is so overland
flow, if present, will not carry contaminants into
surface waterways.
5. Ground-water levels must be at least 3.0 m
(10 ft.) below the existing ground level. Experi-
ments on the Hiawatha National Forest have
shown this to be the most acceptable depth for
waste renovation.
6. The area used should never be subject to
flooding and have bedrock at least 3 m (10 ft.)
below the surface.
7. The site should have less than 10% stone
over 2 cm (5 in.) diameter in the surface foot. This
is a requirement based on equipment limitations
and could be changed if stronger equipment were
available.
8. Infiltration rates are important when
considering the need for the liquid faction of the
waste to move into the soil without oozing
occurring at the surface. Profiles should be
homogeneous without restrictive pans or banding
within the surface 3 m (10 ft.) and have infiltration
rates of 12.7 cm to 63.5 cm (5 to 25 in.) per hour.
9. Soil texture is also an important consider-
ation when rating soils due to the fact that finer
soils are better suited for filtration of bacteria
and virus and for the renovation of chemical
constituents of vault waste. A listing of soil
textures that fulfill the requirements for
renovation and infiltration in decreasing order of
preference are:
a. Well-drained sandy loams
b. Well-drained fine sandy loams
c. Well-drained loamy sands
d. Well-drained sands
e. Well-drained loams
10. Ground water shall be monitored at all
sites for the following chemical and biological
parameters:
a. Chemical parameters:
1. Total nitrogen
2. Nitrate and nitrite — N
3. Total dissolved phosphorus
4. Ammonia nitrogen
5. Specific conductance
b. Biological parameters:
1. Total coliform organisms
2. Fecal colif orm organisms
11. The site must be well vegetated with
grasses and forbs or be vegetated through seeding
to effectively utilize and recycle nutrients
imparted to the soil by the injected vault waste.
12. The area shall be fenced and posted to
exclude entrance by persons who happen upon
the site.
13. Loading rates of waste should not exceed
50 kg/ha (56 Ib/a) of mineralized nitrogen per year
to insure protection from nitrate buildup in the
ground water.
REFERENCES
American Public Health Association, American Water Works
Association and Water Pollution Control Federation.
1971. Standard methods for the examination of
water and waste water, 13th Ed. New York, Am.
Public Health Assoc.
Buller, Richard S. 1974. Potential viral hazards of
land disposal of human wastes. Unpublished M.S.
thesis—Michigan Technological University, Houghton,
Michigan.
Tluczek, Louis J. M. 1974. Public health considerations of
the soil incorporation of campground sanitary wastes.
Unpublished M.S. thesis—Michigan Technological
University, Houghton, Michigan.
53
-------
DISCUSSION
The following questions were answered by Nils Johnson
after delivering his talk entitled "Ground-Water Pollution
Aspects of Land Disposal of Sewage from Remote Recreation
Areas."
Q. Do you think that the reduction of nitrates with time is
caused by increased plant growth?
A. Yes. At the Kentucky Site it is because of vegetative use
of the nitrogen and the nitrogen source (waste) being lost,
partially by leaching.
Q. Did you determine TDS in ground water?
A. Indirectly, by the use of specific conductance and then
using a correction factor to estimate TDS. A periodic
evaluation of cations and anions showed that nitrogen was
the major added substance in the ground water.
Q. by Joe Dix. Did you harvest the vegetation?
A. No, because we did not have a use for it. Future applica-
tions should consider this as it would serve as a method of
nutrient removal from the site and as a source of useful
animal fodder.
Q. In terms of human activity, what happens to the site
when abandoned?
A. The pilot area is dedicated to wildlife management and
will be allowed to revert to this use.
Q. What is the natural rainfall-longevity intensity at the test
sites?
A. The 100-year, 24-hour storm is 6.35 inches,
the 50-year, 24-hour storm is 4.25 inches,
the 10-year, 24-hour storm is 3.25 inches,
the 5-year, 24-hour storm is 2.90 inches,
the 2-year, 24-hour storm is 2.25 inches,
the 1-year, 24-hour storm is 1.95 inches.
Q. by Martin J. Allen. Did you examine whether your
injection practices resulted in increased bacterial densities
(non-coliforms) within the ground water since excessive
bacterial population can desensitize the bacteriological
methods used to isolate total or fecal coliforms?
A. Total plate counts were not performed; however, selective
mediums and the MPN method were used which tend to
decrease this problem. Wherever the MF technique was
used as a check later in the study, overgrowths, which are
often prevalent when high concentrations of non-coliform
organisms are present, were not seen on the plates.
Q. by Mike Apgar. How much confidence do you have that
samples downgradient from the infiltrometer site were
really intercepting contaminant flow? Was bypassing of the
sampling points possible?
A. Wells were approximately 5 feet or deeper below the
water table. Samples were always taken during or directly
after recharge period (i.e. spring, fall) and periodically at
other times. Therefore, we are quite confident of our
sampling. It is possible that nutrients could have passed the
well during major rainfall events but they would have been
of lesser concentrations than those reported.
Q. by Jim Kimball. What did you use for odor control in
the holding tank?
A. Nothing. The vault toilets have a draft ventilation system.
Q. Did beer cans and other trash often disposed of in
vaults plug the pumper truck?
A. No. The septic tank pumper did not usually pick up this
material. Whenever foreign material plugged the system,
the positive pressure created by the injector was sufficient
to clear the lines.
Q. by William A. Trippet. Are the inorganic (NO3-PO4)
peakings over the years of data collection due to rainfall?
A. No. Peaks are usually related to recharge periods (i.e.
snow melt, vegetative dormancy, etc.). Controls do not
show the amplitude in peaks that the treated area does, thus
indicating that peaks are related to the treatment.
Q. by Charles P. Vanderlyn. Have you investigated alterna-
tive waste treatment systems; for instance, chemical units,
biological aeration systems, or incineration of sanitary
wastes?
A. Yes. We use these systems in some instances. We offer
this method as one that is compatible with the land
resource, builds soil properties, does not create an end
product that must be disposed of, and is environmentally
sound. It is only applicable with dry human waste systems.
Q. by K. Childs. What do you do with your sewage during
the winter months? Don't your site requirements exceed
those presently required by the State of Michigan for spray
irrigation systems?
A. Our recreation areas are not open during the winter
months and, thus, no waste is generated during this time.
We use this system between April and October.
This system is not a high leaching system nor is the
disposal material a treated effluent, as with a spray irrigation
system, so these requirements are not applicable.
Q. by Clinton Whitmer. How long before the area may be
impacted a second time?
A. At the Birch Farm Site we reinjected the 1972 site after
three years. We felt the land system could withstand
reapplication because of the increased vegetative cover
resulting from the initial application.
54
-------
The Selection and Management of Feedlot Sites
and Land Disposal of Animal Waste
in Boise Valley, Idahoa
by L. L. Minkb, C. M. Gilmourc, S. M. Beckc,
J. H. Milliganc and R. L. Braund
ABSTRACT
Environmental effects of feedlot location and related
land disposal operations can be minimized if proper
knowledge of hazardous conditions are known and
appropriate protective steps taken. Common guidelines
often do not apply because of differing physical character-
istics of local areas.
Analyses of various parameters within a soil profile
beneath a feedlot revealed none of the chemical constituents
present in high concentrations below the 23-foot depth. In
areas where shallow ground water was less than 5 feet from
the surface, the ground water was found to be affected by
the feedlot. Other hazardous areas in feedlot location are
flood-prone areas, areas of surface bedrock, and areas of
excessive slope.
For land disposal operations, loading rates and
frequency of application of feedlot waste should be
adjusted in accordance with soil permeability, depth to
ground water, and irrigation practices to minimize
detrimental effects on ground-water quality.
INTRODUCTION
Early in 1970 the question of point source
pollutants with respect to cattle feedlot locations
and the disposal of feedlot waste became a major
issue in Idaho. Of prime concern was the Boise
River Valley located in southwest Idaho (Figure 1).
It has been estimated that within the Boise Valley
there are over 350,000 head of beef cattle being
finished annually. Because of economic and
environmental factors, feedlot waste is now viewed
as a beneficial product rather than a waste to be
disposed of at a minimum of effort and expense.
Within the Boise Valley the practice of applying
aSupported by U.S. Army Corps of Engineers, Walla
Walla District. Presented at The Third National Ground
Water Quality Symposium, Las Vegas, Nevada, September
15-17, 1976.
t>Hydrogeologist, Boise State University, Boise, Idaho
83725.
cMicrobiologists and Sanitary Engineer, respectively,
University of Idaho, Moscow, Idaho 83843.
dEnvironmental Engineer, Idaho Department of
Health and Welfare, Boise, Idaho 83705.
manure to crops is gaining acceptance because of
the nutrient value and soil building characteristics
of the manure.
Within the Boise Valley a major portion of the
surface waters are diverted and used for irrigation.
The irrigation return flow back to the surface-
water system and the infiltration of irrigation
waters into the shallow ground-water system
presents a potential problem of surface-water and
ground-water contamination due to feedlot
location and/or management. From this analysis
of the problem, the U.S. Army Corps of Engineers
contracted with the University of Idaho and the
Idaho Bureau of Mines and Geology in cooperation
with local governing agencies to undertake a two-
year study of feedlot locations and associated land
disposal of feedlot wastes.
LOCATION
Boise Valley is a broad alluvial valley which
lies at elevations ranging from 3000 feet in the
upper (SE) portion of the valley to 2400 feet in the
lower (NW) portion of the valley. Several low river
terraces occur on both the north and south sides
of the valley but are more pronounced to the south.
The valley is bounded on the north by what
is termed the Boise Ridge and the south by a
rolling upland plain. The mountains comprising the
Boise Ridge area rise to elevations of over 6000
feet and are composed of granitic rocks of the
Idaho batholith. Quaternary-tertiary sediments
form a low foothills area between the granites of
the Boise Ridge and the Boise River Valley.
The Boise River is the major surface-water
system of the area draining several tributary
valleys from the north. Much of the water from
the Boise River is diverted and used for irrigation
within lowlands of the valley. This diversion water
and resulting irrigation have raised the water table
over large areas of the valley to within a few feet of
the surface.
The climate of Boise Valley is arid to semiarid.
Summer months are dry and warm with cool
55
-------
V -3^<^***>W N Wr^a V^oc?££9!\
£. .T^sjtj i. ^j--1 . tT -1 •- pdrktf=Sv«/.'<»vfc'j»';'>$'J N <•
Is ~ V v«dJcr^^s%iS!
• :-X^ >>flm^; ^Px -^ ^»
> v^^^'^^ll^^' '• '."":4
A ,» 'krfe4|v^^'^r^|-iH'''--"•• ^••-.
\ -^ & *" "™ *! ' ^ f' \Pt*i
Modified from AAPG Geologic Highway Map
Study Area
Fig. 1. Boise Valley feedlot and land disposal study area.
nights while winter months are generally moist
and cold. Average summer daily maximum
temperature is 84° F (29° C) and average winter
temperature is 30° F (-1°C) for the valley. A
majority of the precipitation occurs during winter
months with average annual precipitation
averaging 10-15 inches annually at the lower
elevations.
56
-------
ANALYSIS
Feedlot waste is generally considered an
agronomic resource material, containing substantial
quantities of nitrogen, phosphorus, potassium and
other elements well established as valuable
fertilizer material. It also is beneficial in increasing
water retention and increasing organic matter
content within soils. In all, the estimated value of
feedlot waste if applied to crops is $3.36/ton
(Table 1).
Several constituents of feedlot waste may
become hazardous if proper disposal procedures
are not followed. Often the nitrate nitrogen
component and the coliform bacteria found in
feedlot waste find their way into ground water or
surface waters and create hazardous situations.
Feedlot Characteristics
Chemical composition of feedlot waste used
in land disposal operations near Caldwell is
summarized in Table 2. Hrubant et al. (1972)
reported assorted species of entero-bacteria to
number from 4.4 to 6.8 X 107 per gram of dry
feedlot waste. To determine the effect of feedlot
waste on the subsurface environment, two feedlots
within the Boise Valley were studied in detail.
Feedlot A was located in the Boise River Flood-
plain where ground-water levels are less than 5 feet
from surface. Feedlot B was located on an alluvial
terrace where depth to ground water was over
80 feet.
Density measurements within Feedlot A
found average dry densities of the feedlot surface
to be 52.1 lb/ft3 (0.84 g/cm3) and average densities
of the soil immediately beneath the manure pack
to be 120.6 lb/ft3 (1.93 g/cm3). Volatile solids
content of the manure on a dry-weight basis was
46.4% while the volatile solids of the soil was
only 2.8%. The high density of the soil indicates
compaction of a feedlot surface is not limited
only to the manure layer, but extends into the
upper portions of the soil horizon. It also
indicates one may expect lower permeabilities and
infiltration rates in the soil due to the lower
porosity resulting from compaction.
Infiltration tests were conducted using both
ring infiltrometers and rainfall simulator tests on
both feedlots during the summer and fall of
1974. In all short-term tests (less than 6 hours),
infiltration rates decreased with time resulting in
a final intake rate of less than 0.08 in/hr. In the
long-term infiltration tests (7 days), final
infiltration rates averaged 0.0033 in/hr with all
rates below 0.005 in/hr. At the termination of
Table 1. Potential Value of Applied Beef Feedlot Wastes
Fertilizer Benefit
Nitrogen/ton
Phosphate/ton
Potash/ton
Content/
Ton Manure
4 Ibs. N
4 Ibs. P2OS
9 Ibs. K2O
Price/
Pound
$.325
.256
.087
Value/
Ton
$1.300
1.024
0.783
Fertilizer Value Per Ton Beef Feedlot Manure $3.107
Physical Benefit
Increased Water Retention Rate/Ton
Increased Organic Matter Content/Ton
$ .007
.250
Physical Value Per Ton Beef Feedlot Manure $0.257
TOTAL POTENTIAL VALUE OF APPLIED
BEEF FEEDLOT MANURE/TON $3.364
the 7-day tests the infiltration rings were removed
and the manure and soil directly beneath the rings
examined for moisture content. The manure was
saturated but the underlying soil had a moisture
content averaging 14.6%. This moisture content
was actually less than the moisture content of
surrounding soil which averaged 16.4%. This
indicates vertical percolation beneath the manure
packs will be under unsaturated flow conditions.
Simulated rainfall tests were conducted to
sample all feedlot surface conditions and to
determine the effect of slope on surface runoff.
Although average water application rates as low as
0.06 in/hr were maintained, runoff occurred. This
is consistent with the low rates found utilizing
the ring infiltrometers. Another purpose of rainfall
simulation was to calculate the total amount of
precipitation absorbed or abstracted by the
manure pack. For initial dry conditions the volume
of water retained by the manure pack was 0.56 inch
and for initially wet conditions the amount
averaged 0.23 inch. On two occasions simulated
rainfall was repeatedly applied for two days. Total
water-holding capacity of the manure was
Table 2. Composition of Caldwell Beef Cattle
Feedlot Wastes (from Sikora, 1973)
Constituent
Organic matter
Organic carbon
Organic nitrogen
Soluble calcium, Ca
Soluble magnesium, Mg
Soluble sodium, Na+
Soluble potassium, ¥C
Soluble chloride, Cl"
Soluble ammonium nitrogen, NH4-N
Soluble nitrate nitrogen, NO3-N
Soluble nitrite nitrogen, NO2-N
Concentration
51.
29.
0
130
133
3,990
774
22,800
202
137
2.8
.06%
.69%
.77%
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
57
-------
estimated by summing the cumulative rainfall and
allowing for evaporation and runoff. The average
water-holding capacity of the manure pack was
1.33 inches which calculated to be 0.44 inch of
water per inch of manure thickness. The moisture
penetrated approximately 0.4 inch for a single
application on a dense manure pack. Where
water was applied repetitively on a dense manure
pack, moisture penetration was no greater than
4 inches. For single application on a poorly formed
manure pack over sandy soil, a penetration depth
was found to be 3 inches or less.
Slope up to 12% had little apparent effect on
infiltration rates and water-holding capacity of
the manure pack, but slopes over 12% caused
substantial reduction in water-holding capacity
and increased surface runoff. This was found to
be consistent with several other reported
investigations.
Ground-Water Quality
Ground-water samples beneath three
established feedlot pens and a nearby pasture
were collected and analyzed for the following:
pH, electrical conductivity, sodium absorption
ratio, calcium, sodium, chloride, bicarbonate,
orthophosphate, nitrite, nitrate, ammonium,
chemical oxygen demand, fecal coliform,
sulfate, potassium and carbonate.
General trends were found to exist among
several of the parameters. Calcium, bicarbonate,
orthophosphate and nitrate followed similar
trends during the sampling period from June 1974-
January 1975, indicating the sample locations were
in a common ground-water flow system.
Two pens were higher in electrical con-
ductivity, calcium, chloride, bicarbonate, nitrate,
ammonium and chemical oxygen demand than a
third pen or the pasture (Table 3). The pen
concentration difference could be attributable to
age since two of the pens have been used for 15-20
years, while the third pen is relatively new. Also the
two pens with highest concentrations have had the
largest stocking rate. Sulfate, potassium and
carbonate concentrations were low in both the
pens and pasture with no significant difference
between pens and pasture. Concentrations of
ammonium and chemical oxygen demand were
found to be significantly high in all of the pen
samples for December 1974 (Table 4). This could
be attributed to the high rainfall commonly
occurring during November which would be
sufficient to saturate the manure pack and result
in increased infiltration into the ground-water
system. Subsequent natural sealing of the manure
pack may have caused the decrease for the month
of January. The soil beneath the feedlot is sandy
which could allow the passage of the ammonium
ion. The high chemical oxygen demand may have
been caused by soluble organic matter being
carried by the ground water. Nitrate concentrations
were fairly consistent (below 10 ppm) and suggest
sources other than the feedlots. Sodium and
salinity hazards of the ground water beneath the
feedlots would be considered low with respect to
agricultural standards for irrigation waters.
Bacteria counts were high initially but
reduced to nearly zero towards the end of the
testing (Table 5). This indicated a possible
contamination problem in the installation of the
piezometers during the initial stage of the
project.
Ground water beneath Feedlot A was only
2-3 feet below surface in a gravelly-sand river
bottom material. On the basis of these short-term
data it appeared the feedlot had an effect on the
quality of the ground water in the immediate
vicinity of the feedlot. Since transmissivity and
Table 3. Mean Chemical Concentration of Ground Water Beneath Feedlot A (June 1974—January 1975)
Chemical Parameter
pH
Electrical Conductivity (m-mho's/cm)
Sodium Absorption Ratio
Calcium (ppm)
Sodium (ppm)
Chloride (ppm)
Bicarbonate (ppm)
Orthophosphate (ppm)
Nitrite (ppm)
Nitrate (ppm)
Ammonium (ppm)
Chemical Oxygen Demand (mg/l)
Pen 5
(4 samples)
7.6
0.74
2.2
44.0
62
80
246
0.12
0.002
3.2
3.9
80.9
Pen 2
(5 samples)
7.5
1.11
4.1
48.0
112
91
381
0.17
0.037
3.9
4.5
4.1
Pen 6
(5 samples)
7.5
0.52
1.6
33.0
41
34
224
0.10
0.002
2.3
1.5
2.8
Pasture
(1 0 samples)
7.8
0.67
2.9
27.0
72
56
243
0.23
0.028
2.8
0.3
1.6
58
-------
Table 4. Ammonium Concentrations and Chemical Oxygen Demands of Ground Water
Beneath Feedlot A (December 1974 and January 1975)
Ammonium (ppm)
Chemical Oxygen Demand (COD) (ppm)
Pen
Dec.
14.8
321.0
1
Jan.
0.1
0.1
Pen 2
Dec.
13.3
8.5
Jan.
3.8
2.4
Pen 3
Dec.
6.5
9.5
Jan.
0.2
0.7
Pasture
Dec.
0.5
1.2
Jan.
0.2
0.6
thickness of the aquifer were not determined
accurately, total contribution of pollutants could
not be evaluated.
Within Feedlot B soil profiles were taken to
establish vertical movement of chemical
constituents. Values given were only those
parameters which were solubilized in a soil-
moisture extract. The investigators felt that these
water soluble parameters were the most significant
in consideration of ground-water contamination.
Analyses of nitrate, nitrite, ammonium, and
chloride indicated maximum concentrations in the
upper portion of the soil profile (Figures 2-5).
Nitrates were found in concentrations of up to
66 ppm in the upper 10 feet of the profile but
decreased to 20 ppm in the 6-7-foot level
indicating denitrification was occurring in the
profile and significant nitrogen had not penetrated
deeply into the soil profile (Figure 2). Nitrites
reduced from concentrations of 3-8 ppm in the
upper foot to less than 1 ppm below the 3-foot
depth (Figure 3). Ammonia concentrations
were from 45-88 ppm in the upper 5 feet of
profile but reduced to less than 3 ppm at the
5-foot depth (Figure 4). Chloride concentration
decreased from highs of over 500 ppm in the
upper 1 foot of soil profile to less than 200 ppm
at the 8-foot depth and below (Figure 5). One pen
had chloride concentration significantly higher
(100-200 ppm chloride) than the other pen
sampled (less than 50 ppm chloride). This was
possibly the result of several factors: (1) difference
in age of pens, one being 30 years older than the
other; (2) different stocking rate; (3) different
management practices between the pens; and
(4) different soil and hydraulic conditions.
Similar decreases with depth were noticed
Table 5. Coliform Concentrations in Ground-Water Samples
Beneath Feedlot A (June 1974^January 1975)
Sample
Pen5
Pen 2
Pen 6
Pasture 1
Pasture 2
June
0
3950
100
0
500
July Aug. Sept.
0 - -
2200 -
17 -
0 - -
56 -
Oct.
-
10
0
0
0
Nov. Dec.
- 0
65
0
0
0
Jan.
0
0
0
0
0
for soluble cations including sodium, potassium,
magnesium and calcium (Figure 6). The highest was
the pH value of 8.3 in the upper 2 feet of the
profile but was found to decrease to a pH value
of 7.0 at the 4-foot depth (Figure 7).
Since ground-water levels within this feedlot
were over 80 feet from surface and ground-water
samples were not analyzed, the results of the soil
chemistry indicated the probability of ground-
water degradation was low. None of the various
chemical constituents analyzed were present in
high concentrations in the soil below the 23-foot
depth.
0>
5"! 4
o
I 8
20
22
24
_ Pan 404
D 6 Profiles
10 20 30 40 50 60 70
Nitrate-N, ppm (Air Dry Basis)
Fig. 2. Nitrate nitrogen in soil profile sample extracts
(Feedlot B).
59
-------
1234567
Nitrite -N, ppm (Air Dry Basis)
Fig. 3. Nitrite nitrogen in soil profile sample extracts
(FeedlotB).
0>
r 3
0.4
«>
Q
5
Field
2 Profiles
10 20 30 40 50 60 70 80 90 100
Ammonium-N, ppm (Air Dry Basis)
Fig. 4. Ammonium nitrogen in soil profile sample extracts
(FeedlotB).
Fig. 6. Total soluble cation content of soil profile sample
extracts (Feedlot B).
Land Disposal of Feedlot Wastes
A portion of the study was to determine the
effect of land disposal on soil conditions and
ground water under different loading rates. An area
was chosen near Feedlot A which was typical of
the entire floodplain area adjacent to the Boise
River. The soil was a coarse sandy loam with water
table at a depth of 3-4 feet below surface.
Feedlot manure from Feedlot A was applied
to each of 16 separate plots according to Figure
8. The field was irrigated with a furrow-type system
also typical of irrigation practice in the area.
Piezometers were installed on plots 1-4 and 9-12
and at strategic points surrounding the field to
monitor background-water quality. With this
Or
100 200 300 400
Chloride, ppm (Air Dry Basis)
Fig. 5. Chloride content of soil profile sample extracts
(FeedlotB).
500
^ I
*-
0)
* 2
CL
0>
Q
6.0
70
PH
Pen 315
5 Profiles
Pen 404
6 Profiles
Field
2 Profiles
8.0
Fig. 7. Observed pH values of profile sample extracts
(Feedlot B).
60
-------
Table 6. Ground-Water Cation-Anion Concentrations Beneath Land Disposal Area*
Cations (ppm)
Plot
1
2
3
4
9
10
11
12
Background
Pl-0
Pl-14
Dram
D-2
Ca++
16
17
27
29
38
36
27
27
26
14
26
MS*
10
9
14
15
20
19
14
15
14
7
14
Na*
49
54
97
66
172
267
107
114
92
33
108
*r
34
39
28
8
32
7
3
8
5
3
8
HCOj
260
260
337
277
417
423
304
307
267
173
352
cr
26
25
50
39
96
128
64
73
50
19
37
Anions (ppm)
sol
26
32
44
44
77
116
42
58
57
3
27
pH
8.0
8.0
7.9
8.0
7.9
8.0
8.0
8.0
7.8
8.0
8.0
EC
m mhos cm
0.55
0.53
0.80
0.67
1.30
1.59
0.84
0.89
0.74
0.37
0.81
* Samples = averages of 18-20 samplings taken from August 1973 through October 1974.
network of sampling points, it was possible to
contrast the influence of manure loading at each
disposal site with over-all ground-water quality of
the site. Samples were taken from August 1973
through October 1974 and data represent mean
values based on at least 18 analyses (Table 6).
High sodium values in several of the plots
correlated to initial high Na values for these plots
prior to application of manure. The high E.G. and
HCO3 values correlated well with the corresponding
sodium values. The over-all cation-anion data did
not indicate undue contamination of the ground
water when compared to the background samples.
The Cl and Na data indicated a direct linkage with
the ground-water system but compared to the
quality of the background samples (PL-0 and
PL-14), the ion levels did not indicate massive
transfer of ions to the ground-water system.
Nitrogen and phosphorus concentration in the
ground water (Table 7) indicated low levels of NOz,
NO3, and P over most of the plots. NH3 values
were constant over the plot and indistinguishable
from background samples.
Concentrations of NO~2 and NO3 above back-
ground concentrations were detected below plots
9 and 10. The practice of furrow irrigation
together with soil type may have had an effect on
the distribution of these anions. The high P values
in plots 1 and 2 were probably the result of the
highly permeable sandy-gravel soil beneath these
plots.
Fecal coliform data taken before and after
irrigation indicated a direct transfer of fecal
coliform into the ground-water system, especially
where the highly permeable sandy soil was con-
Table 7. Observed Nitrogen and Phosphorus Levels in
Ground Water Beneath Land Disposal Area*
A - (.round-Water Sampling Sites m 40' X 60' Plots with 20' Borders
Fig. 8. Plot layout design for land disposal area (Feedlot A).
Plot
1
2
3
4
9
10
11
12
Background
Pl-0
N0~2
.01
.01
.01
.01
0.15
0.11
.01
.01
.01
A/O3
3.35
2.35
3.83
1.94
10.17
11.50
3.20
2.79
3.36
NH3
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
P
1.64
1.30
0.59
0.05
0.20
0.24
0.08
0.09
0.35
* Mean values obtained on 18-20 samples from August
1973 through October 1974.
61
-------
Table 8. Influence of Irrigation Water Flow
on Fecal Coliform Levels in Ground Water Beneath
Land Disposal Area (August 1974)
No. Fecal Colifarms/100 ml
Before After
Plot Irrigation Irrigation
2.5 X 104
1.0 X 10
0
0
(Irrigation make-up water contained from 1.85 - 7.4 X 102
fecal coliforms per 100 ml.)
cerned (plot 1) (Table 8). A much lower fecal
coliform count was found with the heavier silty
clay loam test site (plot 9).
Furrow irrigation also had an influence on the
transfer of selected anions NO3, PO4, and Cl into
the ground-water system (Table 9). In the sandy
soil (plot 1) all the anions showed an increase in
ground-water concentration following irrigation.
Plots with heavier soils (plot 11) exhibited little
anion movement to the ground-water system.
Mapping
A program was initiated to map the Boise
Valley and recommend areas where feedlots and
associated land disposal operations could be safely
located based upon the previously discussed data on
feedlot and land disposal loading rates.
Feedlots
Several factors must be considered in location
of feedlots. Some of the physical parameters are
slope of the land, flood-prone areas, and areas of
near-surface bedrock. These parameters can be
defined fairly accurately and mapped accordingly.
The figures in the following text present
various generalized zones to be used as planning
tools in selection of feedlot sites. They are not
hard-fast boundaries but depict areas where
problems of various magnitudesrmay occur.
Figure 9 of this section shows the following:
Table 9. Influence of Irrigation Water Flow on
Ion Movement to Ground Water Beneath
Land Disposal Area (September 1974)
Before Irrigation (ppm) After Irrigation
No~3 POP cr No~3 PO\ cr
Plot
P0\
1
9
3
11
2.10
7.20
3.70
3.00
1.06
0.20
0.22
0.01
7
84
71
59
5.20
6.00
3.10
3.00
8.28
0.02
0.28
0.01
41
84
56
56
(1) slopes over 12%, (2) floodplain area, and
(3) bedrock within 5 feet of surface.
Excessive Slope
Excessive slopes may cause problems to
feedlot operations in several ways. In areas of steep
slopes (over 12%), the impermeable organic mat
described earlier will not develop. The organic
material will be washed or carried to the bottom of
the slope rather than developing on the slope itself.
This will allow infiltration of nutrients and
bacteria possibly resulting in contamination of a
ground-water supply.
Excessive slopes also involve the operation of
the catchment basins located downslope from a
feedlot which are required for containment of
surface runoff. The steeper slopes generally make
the design and construction of catchment basins
more expensive and management of these basins
more difficult.
Slopes in excess of 12% were considered to
be poorly suited for reasons previously mentioned.
Figure 9 shows generalized areas in the Boise River
Valley where excessive slopes occur. Excessive
slopes occur north of the Boise River in both Ada
and Canyon Counties. Scattered areas of excessive
slopes also may be found in southern Canyon and
Ada Counties along terraces and valleys near the
Snake River.
Floodplain
Location of feedlot sites within a floodplain is
hazardous because of possible inundation by flood
waters and subsequent transport of pollutants into
surface waters. Floodplain locations should not be
considered for any long-range operation. The
expense and problems, both physical and environ-
mental, caused by flooding can be monumental.
The floodplain plates were prepared utilizing
data from the United States Geological Survey.
Figure 9 shows the approximate area which could
be flooded in a probable one-hundred-year event.
This area covers most of the lower valley area of
the Boise River and certain tributaries of the Boise
River. Channel alteration and artificial structures
such as bridges could change the flood hazard area.
An important factor relating directly to
surface water is the location of a feedlot near a
stream. A feedlot should be designed so the
catchment basins and settling ponds can be
constructed between the surface stream and the
62
-------
Fig. 9. Cattle feedlot and land disposal hazard areas-excessive slope, near-surface bedrock and flood-prone areas.
-------
feedlot itself to prevent direct runoff and ground-
water seepage from entering the stream. This
distance is variable depending on the nature of the
stream, terrain, soil types, and feedlot design.
Bedrock Areas
Figure 9 shows near-surface bedrock areas.
Outcrops of bedrock, mainly basalt (lava rock),
can be hazardous because of the incomplete
formation of the impermeable organic mat over
the bedrock. If this mat is not formed, surface
waters carrying organic and bacterial pollutants
may flow into the fractures of the bedrock and
could ultimately contaminate a ground-water
supply.
The Figure depicts bedrock within 5 feet of
the surface. Most of the area of concern which has
shallow bedrock is located in southern Ada and
Canyon Counties. Because of the scale, the Figure
should be used as a guide with detailed geologic
investigation performed on specific sites within the
shaded area.
Land Disposal of Feedlot Waste
Land disposal of feedlot waste should be
given serious attention when locating the feedlot
site. First the site should be selected where the
economics of disposal provide maximum benefit
and where the problems of disposal are best
resolved. Analysis of land disposal operations
indicates that four factors are of primary
importance: (1) depth to ground water, (2) soil
permeability, (3) near-surface bedrock or caliche,
and (4) proximity to high population area.
Depth to Ground Water and Soil Permeability
Depth to ground water has been found to be
one of the most serious limiting factors in disposal
operations when viewed in conjunction with soil
permeability (Figures 10 and 11). Depths to ground
water reach a minimum in the Boise River Valley in
mid-July and persist through mid-October. The rise
in ground water in the Boise Valley is due to
recharge of the ground-water system by irrigation.
In planning a land disposal operation, ground-water
depths during the late summer and fall months of
high water table should be considered.
Shallow ground-water depths over much of
the Boise River floodplain are less than 5 feet, and
in numerous places the water table is within a few
inches of the surface. The shallow ground-water
system is also found in several terrace and bench
areas. One extensive area extends in an east-west
direction from the western edge of Boise to
Caldwell. Other shallow zones are located between
Caldwell and Nampa and around Lake Lowell
(Figure 10).
The shallow ground-water areas should be
avoided for land disposal if soils in the area have a
medium to high permeability. Highly permeable
and medium permeable soils have been mapped and
are shown in Figure 10. These soils have a
permeability of greater than 6.3 in/hr for the
highly permeable soils and between 2.0 and 6.3
in/hr for the medium permeable soils. Where
water-table depths are less than 5 feet from the
surface during late summer and fall, areas of high
and medium permeability soils should be
avoided. Where water-table depths are between
5 and 10 feet, the areas of high permeability soils
should be avoided. In both the highly permeable
soils and medium permeable soils careful water
management should be practiced.
Low permeability soils (0.2 in/hr) are limiting
because of the amount and type of waste which
can be applied. These soils will absorb only a
minimum amount of liquid waste material before
surface runoff occurs. For this reason they should
be reviewed with respect to the size and type of
land disposal operation being considered and
irrigation water carefully regulated to prevent
surface runoff from occurring.
The soil permeability maps are generalized
and should be used only for planning purposes,
and it should be understood that soil types are
complex and often discontinuous. For this reason,
any given land disposal site should have a qualified
soil scientist or geologist check the soil permeability
and ground-water levels prior to the initiation of
the disposal operation.
Hardpan
Hardpan zones near the surface may limit
the amount and kind of waste material which may
be applied in a land disposal operation. The hardpan
zones should be considered if the waste material is
in a liquid form or if irrigation water is applied to the
disposal site. Since hardpan zones are essentially
impermeable, the water cannot percolate downward
through the zone, and the soil above the hardpan
layer may become saturated resulting in surface
runoff. This could result in a serious limitation on
the amount of waste or the kind of waste which is
being considered for a land disposal operation.
Hardpan areas are found throughout the Boise
Valley; therefore, subsurface investigations for
hardpan should precede any land disposal operation.
64
-------
^ Wni-r l.ihk- S 10 I mm Surlji.
H VV.im l.iMc V iToin Surt.ni
;
-• L
Fig. 10. Cattle feedlot and land disposal hazard areas-depth to water.
-------
D
So
Fig. 11. Cattle feedlot and land disposal hazard areas-soil permeability.
-------
All areas ot surface bedrock should be
avoided for land disposal sites. Waste material can
be transmitted through fractures in the rock to the
ground water without any purification by the soil,
thus resulting in a seriously contaminated ground-
water aquifer (Figure 9).
Urbanization
Population pressure is probably one of the
most unpredictable and potentially serious
problems in feedlot location and related land
disposal operations within the Boise River Valley.
Odor and general nuisance complaints stand as
the problems. With increased urbanization, through
growing cities and scattered high-density subdivi-
sions, odor and general nuisance complaints
become a serious matter in feedlot and land disposal
location.
Odor from livestock production can be at a
maximum during times of manure disposal on land.
In order to minimize odor production during
disposal operations, it is important that attention
be given to weather conditions which may affect
the transportation of odors. Of particular concern
is the temperature inversion condition which
prevents normal air movement and ventilation. The
National Weather Service in many areas prepares
an Air Pollution Dispersal Forecast which indicates
the degree of inversion existing or expected.
Feedlot operators should curtail manure disposal
operations when the dispersal forecast is rated
as poor.
SUMMARY AND CONCLUSIONS
Many factors influence the siting of a feedlot
and land disposal operation. With adequate
knowledge of the factors and limiting conditions,
many of the hazards associated with feedlot
operations can be eliminated or greatly reduced. It
is also important to realize each sector of the
country possesses different characteristics which
make application of common environmental
standards difficult. Different characteristics can
often be found even within State and local
boundaries causing difficulties in the use of
common guidelines. The previous discussion and
following recommendations relate the semiarid
Boise Valley region and are based on an interaction
of several parameters.
Within the feedlot the manure pack and
especially the semipermeable organic mat should
be maintained to reduce infiltration and minimize
ground-water degradation. Slopes of from 2-3%
are recommended to minimize ponding within
the feedlot, but slopes over 12% should be avoided
because of erosion and incomplete formation of
the organic mat. Basalt outcrops should be avoided
because of potential ground-water contamination
through open fractures. Feedlot location on
floodplain areas may be hazardous to animals and
property although in certain circumstances, feedlot
design can alleviate problems due to flood waters
entering the feedlot area.
Land disposal of feedlot waste should be
avoided in areas where there exists soil of high
permeabilities in conjunction with high water
tables (less than 5 feet). In areas of deeper water
tables the manure should be incorporated into the
soil as soon after application as possible. Sprinkler
irrigation is recommended over furrow irrigation
in all land disposal operations. Yearly waste
loading rates of 40-50 tons/acre are recommended
and should be adjusted in accordance with soil
permeability, depth to ground water, and irrigation
practices. Applications should be made during late
fall and the surface allowed to dry prior to
incorporation into the soil to decrease potential
ground-water pollution due to fecal coliform. If
spring application is needed, application should
be carried out 3-4 weeks prior to the first
irrigation.
Surface runoff from a feedlot area should be
controlled through the use of narrow, steep,
fast-flowing ditches and channeled to lined
settling ponds prior to entering the storage area.
The contaminant ponds should be emptied as
soon as possible to prevent pond discharge and to
reduce odor problems. Disposal of settled solids
from the pond should be accomplished in
accordance with land disposal guidelines. For
both feedlots and land application operations,
sites should be selected outside areas of projected
future urbanization with careful consideration
given to effects of odor and esthetics.
SELECTED REFERENCES
Ada Council of Governments. 1971. Land use analysis,
Ada County-1970. Boise, Idaho.
Ada Council of Governments. 1973. Ada County ecology.
Environmental 9. Boise, Idaho.
Ada Council of Governments. 1973. Background informa-
tion for Ada County, Idaho. Environmental Planning
Report 3. Boise, Idaho.
American Public Health Association, American Water Works
Association, and Water Pollution Control Federation.
1971. Standard methods for the examination of
water and wastewater. 13th ed. Washington, D.C.
Cornell, Howard, Hayes and Merryfield/Hill. 1972.
Planning for Canyon County. Canyon County
Planning and Zoning Commission, Boise, Idaho.
Davis, S. N. and R.J.M. DeWiest. 1966. Hydrogeology.
67
-------
New York: John Wiley and Sons, Inc.
Dion, Norman P. 1972. Some effects of land use changes
on the shallow groundwater system in the Boise-
Nampa area, Idaho. Idaho Dept. of Water Adm.,
Water Information Bull. 26.
Federal Water Pollution Control Administration. 1969.
Proceedings of animal waste management conference.
Kansas City, Mo.
Federal Water Pollution Control Administration. 1969.
Second compendium of animal waste management.
Kansas City, Mo.
Gilbertson, C. B., T. M. McCalla, J. R. Ellis, O. E. Cross,
and W. R. Woods. 1970. The effect of animal density
and surface slope on characteristics of runoff, solid
wastes and nitrate movement on unpaved beef
feedlots. Bulletin SB 508, Nebraska Agricultural
Experiment Station, Lincoln.
Gilmour, C. M., S. M. Beck, J. H. Milligan, L. L. Mink,
A. A. Araji, R. L. Reid, and R. L. Braun. 1975.
User's manual for the selection and management of
feedlot sites and land disposal of manure in Boise
Valley, Idaho. Army Corps of Engineers Advance
Report.
Gilmour, C. M., S. M. Beck, J. H. Milligan, L. L. Mink, A. A.
Araji, R. L. Reid, and R. L. Braun. 1976. Development
of basic information for feedlots and land disposal
system in the Boise Valley. Army Corps of Engineers
Advance Report.
Hollenbaugh, K. M. 1973. The evaluation of geologic
processes in the Boise Foothills that may be hazardous
to urban development. Ada County Council of
Govts., Boise, Idaho.
Hrubant, G. I., R. V. Daughtery, and R. A. Rhodes. 1972.
Entero-bacteria in feedlot waste and runoff.
Applied Microbiology. 24:378-383.
Idaho Water Resource Board. 1973. Comprehensive rural
water and sewerage planning study Canyon County.
Boise, Idaho.
Keener, K. B. 1920. Report on rise of ground water—Boise
project. Unpublished report, U.S. Dept. of Reclamation.
Mohammad, D.M.J. 1970. Hydrogeology of the Boise Ridge
area. Unpublished M.S. Thesis, University of Idaho,
Moscow, Idaho.
Pacific Northwest River Basins Commission, Meteorology
Committee. 1969. Climatological handbook Columbia
Basin states, v. 1 and 2.
Powers, W. L., G. W. Wallingford, and L. S. Murphy. 1975.
Research status on effects of land application of
animal wastes. Office of Research and Development,
U.S. Environmental Protection Agency, Corvallis, Ore.
Priest, T. W., C. W. Case, J. E. Witty, R. K. Preece, Jr.,
G. A. Monroe, H. W. Biggerstaff, G. H. Logan, L. M.
Rassmussen, and D. H. Webb. 1972. Soil survey of
Canyon area, Idaho. Soil Conservation Service,
USDA, Washington, D.C.
Ralston, D. R. and S. L. Chapman. 1970. Groundwater
resource of southern Ada and western Elmore
Counties, Idaho. Idaho Dept. of Reclamation, Water
Information Bulletin No. 15.
Savage, C. N. 1958. Geology and mineral resources of Ada
and Canyon Counties. County Report No. 3, Idaho
Bureau of Mines and Geology, Moscow, Idaho.
Sikora, L. J. 1973. Soil as an animal waste disposal system.
Ph.D. Dissertation, University of Idaho Graduate
School, Moscow.
Stevens, P. R. 1962. Effect of irrigation on groundwater in
southern Canyon county, Idaho. U.S.G.S. Water-
Supply Paper 1585, Boise, Idaho.
Toron, Praphat. 1964. Groundwater in the Boise area,
Idaho. Unpub. M.S. Thesis, University of Idaho,
Moscow.
U.S. Department of Commerce, National Oceanic and
Atmospheric Administration. 1974. Environmental
data service, climatological data, Idaho, v. 75, no. 7.
U.S. Environmental Protection Agency. 1972. Cattle feedlots
and the environment. Control Guidelines. Seattle,
Washington.
U.S. Environmental Protection Agency. 1974. Methods for
chemical analysis of water and wastes. National
Environmental Research Center, Analytical Quality
Control Laboratory, Cincinnati, Ohio.
Williams, R. E., D. D. Eier, and A. T. Wallace. 1969.
Feasibility of re-use of treated wastewater for
irrigation, fertilization, and ground-water recharge
in Idaho. Idaho Bureau of Mines and Geology, Boise,
Idaho, Pamp. 143.
Williams, R. E., and A. T. Wallace. 1970. Hydrogeological
aspects of the selection of re-use disposal sites in
Idaho. Idaho Bureau of Mines and Geology, Boise,
Idaho, Pamp. 145.
DISCUSSION
The following questions were answered by L. L. Mink
after delivering his talk entitled "The Selection and Manage-
ment of Feedlot Sites and Land Disposal of Animal Waste
in Boise Valley, Idaho."
Q. What is the nitrate level of the ground water under those
cultivated areas next to the feedlots? Are they higher than
those waters beneath the feedlots?
A. Mean nitrate-nitrogen concentrations of ground water
beneath a pasture and a cultivated field near feedlot A was
2.8 mg/1 and 3.36 mg/1 respectively. Mean concentrations
of nitrate-nitrogen of ground water beneath three feedlot
pens were 3.2 mg/1, 3.9 mg/1, and 2.3 mg/1. This indicates
nitrate-nitrogen in the ground water beneath the feedlots
was not significantly different from the nitrate-nitrogen
beneath the cultivated areas.
Q. by H. Bouwer. Your results seem to contradict others
where old feedlots that were not cleaned contributed less
nitrate to ground water than feedlots regularly cleaned.
This was due to ammonium accumulation which increased
pH. This in turn decreased nitrification and increased
ammonia volatilization. Comment?
A. All feedlots within this investigation were being utilized
prior to and during the investigation. The slightly greater
amounts of nitrates found in the older pens is thought to
be due to accumulation of nitrates over time. If a pen is
cleaned to a degree the organic mat is disturbed or removed,
increased infiltration would result and a definite increase in
nitrates could be expected. Although the data indicate
higher concentrations of ammonia beneath the older pens
there was no significant difference in pH in the ground water
beneath the pens. This could be due to the low concentra-
68
-------
tions of both nitrates and ammonia, therefore not creating
the conditions of increased ammonia volatilization and
decreased nitrification.
Q. With respect to nitrates in soil systems beneath feedlots,
studies in other areas have shown high nitrate concentra-
tions subsequent to feedlot abandonment and deterioration
of the manure pack. What have been your observations in
Montana subsequent to feedlot abandonment?
A. We did not specifically look at abandoned feedlots during
this investigation. Observations since the investigation indi-
cate that upon abandonment the impermeable organic mat
deteriorates allowing surface water to infiltrate through the
organic material. The percolating waters carry substantial
amounts of nitrates which cause high concentrations in
ground waters beneath the feedlot.
Q. by W. Turner. How long does it take to develop an
effective manure pack ?
A. No testing was performed during the study but observa-
tions on newly established feedlots indicate a period of
from 3 to 6 weeks is required to establish an effective
organic mat.
Q. by Don Runnells. Were the analyses for the components
below the "dry" feedlot performed on the soils or on
vadose interstitial waters? In either case, how did you sample
the materials?
A. The chemical analysis in the vadose zone was performed
on air-dried soil samples and the values reported are those
parameters which were solubilized in a soil moisture extract.
The samples were taken utilizing a small bucket-type hand
auger with the samples collected in plastic bags and taken
directly to the laboratory for analyses.
Q. by P. K. Saint. Your feasibility map would recommend
feedlot location on areas with thick vadose zones. But these
are zones of regional recharge and would you not contami-
nate large areas of ground water? Low water-table areas are
ground^water discharge areas, and would these not be more
suitable?
A. The recommended areas for feedlots did not consider
depth to water. As long as an effective organic mat is
maintained, data indicate ground-water pollution from
feedlots will be minimal. Soil profile data indicates the
cation-anion concentrations decrease with depth so at a
ten-foot depth the concentrations are at or near background
levels. Since we did find above background amounts of
several constituents beneath feedlots where the water table
was from 2 to 5 feet below surface, we recommend
feedlots be located in areas where water-table depths are at
least 5 feet from surface and preferably over 10 feet from
surface. To locate feedlots in ground-water discharge areas
would cause problems with surface-water systems since the
ground water would be discharging in surface-water systems
within a short distance.
Q. Have you considered the rising trend of the water table
in the area in delineating the feedlot locations and selecting
the recommended depth to the water table?
A. Yes, the data used to delineate areas of high water tables
were gathered during late summer and early fall. This is
during the period when the water table is at its maximum
elevation due to irrigation in the Boise Valley.
Q. by K. Childs. How important is the ability of differing
soil types for waste assimilation? Was this considered? Don't
low permeabilities increase retention time for wastes and
consequently increase treatment of waste? Are waste
constituents removed with the zone of saturation?
A. As evidenced by the data from the land disposal portion
of the study, soil types play an important role in waste
assimilation. High concentrations of bacteria, nitrates,
nitrites, and chlorides were detected in the ground waters
beneath plots which had high permeable sandy soils but
no significant increase of these parameters was
detected in lower permeable silt loam soils. Within the
study we did not attempt to determine the waste assimilative
capability of the zone of saturation.
Q. by Maxine Goad. What consideration did you give to the
availability of suitable sites for impoundments to catch
surface runoff? Did you investigate the possibility of
ground-water contamination caused by seepage from such
impoundments?
A. We recommended areas of steep slopes be avoided
because of problems associated with design and maintenance
of catchment facilities. We also recommended an adequate
area be maintained between a surface stream and the feedlot
for surface runoff catchment facilities. The study did not
address the problem of seepage from surface-water
impoundment facilities. Within the feedlots investigated
such facilities were lined with an impermeable liner.
Q. by Clinton C. Whitmer. What percentage of the Boise
Valley was left available to feedlot use?
A. A large area within the Valley appears to be suitable for
feedlots as far as physical restrictions. Only flood-prone
areas near the Boise River and areas of excessive slopes along
the Boise Ridge are not suitable. The major problem is that
of urbanization—considering the scattered development
which is now occurring in the Valley one may conclude there
is no acceptable safe location.
Q. by Don Runnells. Is there any legal or regulatory
machinery in Idaho to implement (or impose) the recom-
mendations of work such as yours? Are there any significant
efforts to preserve your agricultural lands?
Q. Within Idaho the protection of agriculture lands is in the
hands of the county zoning agencies. Within the Ada-
Canyon Counties comprising the Boise River Valley a
serious attempt is being undertaken to protect agricultural
lands from urbanization but at the present time is plagued
by local politics and special interest groups. The planning
and zoning agencies are also working to implement the
recommendations of the study. We are very pleased the
local cattle feeders association has expressed considerable
interest in the results and recommendations of the study.
"The Lycoming County, Pennsylvania, Sanitary
Landfill: State-of-the-Art in Ground-Water Protection,"
by M. Todd Giddings, Jr. appears on page 125.
69
-------
Prediction of Future Nitrate Concentrations
in Ground Water"
by C. P. Young, D. B. Oakes and W. B. Wilkinsonc
ABSTRACT
Over the last few years rises in the nitrate content of
ground water from wells and springs in the principal aquifers
of the United Kingdom have been observed. In a number
of cases the concentrations have exceeded the WHO lower
recommended limit. In order to determine the reason for
the rise, to assess whether it will continue and the eventual
nitrate levels, the Water Research Centre has undertaken
an extensive programme of drilling and sampling on the
Chalk and Bunter Sandstone, and by August 1976,
twenty-two sites had been examined. This work has
established that high nitrate concentrations (peaks up to
60 mg/1 NOs-N have been observed) are present in the
unsaturated aquifers at fertilized arable/ley sites. At
unfertilized grassland sites nitrate concentrations are low
(less than 4 mg/1 NO3-N) and below fertilized established
grassland values are in the intermediate range. At one farm
site near Winchester, models to predict the rate of
movement of nitrate through the unsaturated and saturated
Chalk have been developed. These suggest that the nitrate
levels at this site will remain at an essentially constant
value of about 4 mg/1 NO3-N until the late 1970's when
they will rise progressively to about 14 mg/1 NOs-N. The
models have been checked against tritium data and the
approach is now being extended to other sites.
1. INTRODUCTION
A rise in the concentration of nitrate in ground
water has been observed in many countries over the
last few years and is a cause of concern to health
authorities and water engineers. It has been
recognised for about 30 years (Comley, 1945) that
excessive quantities of nitrate in drinking water can
cause methaemoglobinaemia in infants under the age
of 6 months; the baby develops a grey-glue colour
which has been called "water well cyanosis." The
condition is extremely rare but if untreated,
permanent damage or death may result. There is
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
t>Water Research Centre, Medmenham Laboratory,
P.O. Box 16, Henley Road, Medmenham, Marlow,
Buckinghamshire SK7 2HD, England.
some evidence to suggest that increased
methaemoglobin levels in the blood of older
children may result from the ingestion of "high
nitrate" water (Petukhov and Ivanov, 1970) but
adverse effects on health have not been proved.
This paper describes some of the techniques being
used to estimate the future changes in the concen-
tration of nitrate in ground water in the United
Kingdom.
Ground water is a major source of public
supply in England and Wales where it meets about
30% of the demand. In some areas, particularly
southeast England, it is the major source. The
quality of ground water pumped from the two
principal aquifers in the United Kingdom, the
Chalk and Bunter Sandstone (Figure 1), is
Fig. 1. Sites of the Water Research Centre nitrate investiga-
tions.
70
-------
generally exceptionally good requiring only
minimal treatment. However, increasing nitrate
levels in both aquifers have been observed (Davey,
1970; Greene and Walker, 1970; Satchell and
Edworthy, 1972; Foster and Crease, 1974; Young
and Hall, 1976). At the present time ground water
from more than 100 public supply sources either
continuously or intermittently exceeds the World
Health Organisation lower limit for nitrate of
50 mg/1.
The WHO recommend in their European
Standards for drinking water that nitrate should
be less than 50 mg/1 but suggest that an upper limit
of 100 mg/1 is acceptable provided that health
authorities are notified if the lower limit is exceeded
so that problems arising with infants may be
readily recognized. (Nitrate analyses are usually
expressed in terms of nitrogen and from here on,
this is adopted in this paper. The equivalent of the
WHO limits of 50 and 100 mg/1 are 11.3 and 22.6
mg/1 nitrogen.)
The water industry is clearly concerned to
know whether
(a) the source of nitrate can be controlled,
(b) the trend of increasing nitrate concentra-
tions will continue and the level to which they will
eventually rise,
(c) currently acceptable ground-water sources
are also going to be affected,
so that alternative water sources may be planned,
treatment methods developed, or other appropriate
action taken as necessary.
Research was required in order to answer
these questions and in October 1974 the Water
Research Centre expanded its programme in this
area.
2. SELECTION OF SITES AND
INVESTIGATION TECHNIQUES
The principal objectives of the field studies
have been to
(1) measure the amount of nitrate contamina-
tion in the zone of aeration and in ground water at
typical sites on the Chalk and Bunter Sandstone;
(2) determine the relationships between the
above measurements and meteorological, hydro-
geological, historic land use, and agricultural factors;
(3) assess whether nitrate in moving down
through the zone of aeration and in the aquifer, is
subject to denitrification; and
(4) estimate the future concentrations of
nitrate in ground water, initially on a detailed
local level, but proceeding to a catchment scale.
The objectives are being met by an extensive
programme of drilling and sampling on the Chalk
and Bunter Sandstone, and to date 22 sites have
been investigated (Figure 1).
Sites have been selected to cover a range of
land-use situations. In order to ensure compatability
they have been located on outcrop areas free from
superficial deposits. An unsaturated zone depth
of at least 20 metres has been required so that the
rate of downward movement of contaminants
through this zone could be determined. Detailed
records of land use and fertilizer application rates
over a period of at least 10 years, but preferably
30 years, together with meteorological records
have been considered essential. The Ministry of
Agriculture, Fisheries and Food (MAFF) and the
water authorities have given invaluable assistance
with site selection and the supply of the basic data.
Great care has been taken in the choice of
drilling techniques so as to obtain fully representa-
tive samples of the rock and pore fluids. Procedures
employing a fluid to flush rock particles from the
drilling bit could not therefore be used. Dry auger-
ing and drive coring were the methods generally
adopted to obtain samples. However, under
certain adverse conditions in the Bunter Sandstone
air flush rotary drilling was employed (Gray et al.,
1976).
Continuous augering at 114 mm (4.5 in)
diameter at depths not exceeding 30 m (98 ft),
with sample collection from the auger flights, was
used at the preliminary stages of the investigations
to give a broad nitrate profile through the
unsaturated zone and to determine the suitability
of the site for more detailed study. If the site was
suitable, a drive coring or a rotary drilling method
was used to give 100 mm (4 in) diameter continuous
undisturbed core. These boreholes ranged in depth
up to 200 m (656 ft) and in some cases passed
through the unsaturated zone and deep into the
saturated aquifer. All samples, on removal from the
borehole, were immediately sealed in double
polythene wrappings and stored in insulated
containers for transport to the Centre's laboratory.
Pore water was extracted from the rock samples
using a high-speed centrifuge (Edmunds and Bath,
1975) and was analysed for nitrate and nitrite by
Auto Analyser techniques. Samples for tritium
analysis were passed to another organisation's
laboratory who undertook the preparation of the
samples and the measurement on a contract basis.
71
-------
3. GENERAL RESULTS OF FIELD
INVESTIGATIONS
The literature describing the leaching of
nitrogen from soil under a variety of land use and
meteorological conditions is profuse. However,
measurements in the United Kingdom and elsewhere
have generally been restricted to the soil zone
(Wild and Babiker, 1975). Nitrate profiles to a
depth of up to 30m (98 ft) have been measured
in granular aquifers in Israel and the United States
(Olsen et al, 1970; Adriano et al, 1971; Reinhorn
and Avnimelech, 1974; Smith and Young, 1975;
Mielke and Ellis, 1976; Lund et al, 1976), but as
far as the authors are aware the results of the
present study describe for the first time detailed
nitrate profiles in the unsaturated zones of
consolidated aquifers.
The 22 sites cover a range of land-use situations
that may be broadly grouped into three types:
(a) fertilized arable including temporary leys,
(b) fertilized long-term grassland, and
(c) unfertilized long-term grassland.
10
N03-N(mg/l)
20 30 40 50
60 70 80
RestW.L. Oct'75
40-
45 J
Fig. 2. Measured nitrate profile below a fertilized arable site
on Chalk.
0
0 10
N03-N(mg/l)
20 30 40
50 60
RestW.L. July'75
Fig. 3. Measured nitrate profile below a fertilized arable site
on Bunter Sandstone.
Nitrate profiles through the zone of aeration
beneath inorganically and organically fertilized
arable land on both the Chalk and Bunter
Sandstone are given in Figures 2 and 3. The profiles
show peak nitrate concentrations which are in
some cases up to five times greater than the WHO
lower recommended limit. The peaks are often
irregularly interspersed with zones of low concen-
tration. In contrast, profiles below permanent
unfertilized grass show comparatively low levels
which are almost always below 4 mg/1 NO3-N
(Figure 4). Fertilized long-term grassland sites give
profiles (Figure 5) that are intermediate between
the arable and unfertilized cases.
Deep drilling and sampling below the water
table has also been carried out at a number of sites.
At one site on the Bunter Sandstone adjacent to a
water-supply well, which has recently recorded an
increase in nitrate, a profile with high nitrate peaks
has been measured. These results are as yet
incomplete and so a full interpretation cannot be
made at this stage.
It is clear from the results described above
that a considerable quantity of nitrate is present in
the zone of aeration below fertilized sites on the
Chalk and Bunter Sandstone. However, in order to
72
-------
N03-N (mg/l)
0
0 1
5-
X10H
a.
LU
Q
20 J
Fig. 4. Measured nitrate profile below an unfertilized
permanent grassland site on Chalk.
attempt a prediction of the future concentration of
nitrate in ground water it is necessary to establish
quantitative information on
0
N03-N (mg/l)
10 20 30
5-
CL
QJ
15-
20-
25 -
computed
measured
WHO recommended
limits of 11-3,22-6mg/l
Rest water level
Oct'75
Fig. 5. Measured and computed nitrate profile below a
fertilized grassland site on the Chalk-Bridget's Farm
Borehole G, Maryland field.
(a) relationships between land use and nitrate
leaching loss from the soil,
(b) the rate of downward movement of
nitrate, and
(c) the amount of denitrification that is taking
place in the zone of aeration.
The results from all of the sites are being analysed
in order to establish these relationships. The
investigations and analyses at a site on the Chalk
near Winchester (Figure 1) are described in detail
in the following section.
4. CHALK SITE NEAR WINCHESTER
4.1. Land Use, Hydrogeology and Borehole
Locations
Investigations were undertaken on a Chalk site
at Bridget's Experimental Husbandry Farm between
October 1974 and January 1976 (Young et al,
1976). Detailed land utilisation and agricultural
practice records are available from 1948 when the
MAFF took over the farm.
The drilling and sampling procedures used in
the investigation were described in section 2.
The farm is situated on the south facing
slopes of the Itchen Valley (Figure 6). The lowest
fields border the River Itchen at a level of about
45 m (148 ft) Ordnance Datum from which the
land rises steadily over a distance of 1.5 km (1 ml)
to a level of 116 m (381 ft) O.D. at the northern
extent of the farm boundary.
The soil cover within the farm boundaries
and the surrounding area is generally thin and
with a low clay content. It rests directly on the
Chalk so that, except under conditions of very
heavy rainfall, surface runoff is negligible.
Estimates of infiltration from 1948 to the present
have therefore been made on the basis of rainfall
less actual evaporation.
Because the farm is experimental in nature,
changes in land use do not necessarily follow the
national trend which has shown a rapid increase
in arable acreage over the last 40 years. The local
increase for the Parish of Itchen Valley is shown
in Figure 7. There has also been a tenfold increase
in fertilizers over the country as a whole during
this period.
The ground-water catchment in which the
farm is situated is comprised of Upper Chalk and
it is estimated that its base lies at about
Ordnance Datum beneath the farm site. The
beds at the site are almost horizontal with a
regional dip of about 1° to the south.
73
-------
The Chalk is a fine-grained carbonate rock
traversed by horizontal and near vertical joints
and fissures. The jointing pattern is very irregular
with block sizes ranging from a few centimetres to
more than a metre. The size of the openings is
volues against these sites are
the nilrate nitrogen concentration in rng/l
Observation well (ft) ) measured on one ot the following Oates "•• 19 ?S
A - May • - 3 OctoOer C - IS November D - DeLpmDer
Grourxs-ater level I m 0 D I 60
Groundwater cati f»n - - - ITCHEN
Sub - tafchmeft sampled by boferioies A-J
Fig. 6. Location of Bridget's Farm, rest water levels,
ground-water catchment and nitrate concentrations in
ground water and springs in 1975.
35 1940 45 1950 55 1960 65 1970
Fig. 7. Changes in land use-ltchen Valley parish.
75
Fig. 8. Scanning electron micrograph of the Chalk matrix.
difficult to estimate. Horizontal discontinuities up
to 10 mm (0.4 in) in width have been reported by
Ward et al., 1968, but Foster and Milton, 1974,
have noted tightly closed joints at relatively
shallow depths. Near the water table these
discontinuities have been enlarged by solution and
this leads to the high transmissivity values that are
associated with the Chalk aquifer. The rock matrix
is composed of fossil debris ranging in size from
1 to 100 jum and pore diameters are generally less
than 5 jum (Figure 8). The intergranular permea-
bility is therefore low, of the order of 1 X 10"3
m/d (2 X 10"2 gals/d/ft2), but the porosity is high
at 0.3 to 0.4.
Ground-water levels in the Chalk of the area
are shown in Figure 6. These are based on the
information collected during the investigation and
data previously assembled by the Southern Water
Authority (1975). Long-term ground-water level
records are not available at the farm site but
fluctuations of up to 3 m (10 ft) have been
recorded at the head of the catchment.
Nitrate determinations were made on water
samples collected from boreholes and springs in
the area and these results are plotted in Figure 6.
The values generally lie between 3 and 7 mg/1
NO3-N. No trend with respect to location is
74
-------
apparent. The nearest public water-supply well is
on the south side of the Itchen Valley and this has
shown a rise in the nitrate concentrations from
about 3 mg/1 NO3-N in 1955 to 4.5 mg/1 NO3-N in
1975.
During the course of the investigations 8
boreholes, designated A to H in Figure 9, were
sunk by the Centre through the unsaturated and in
some cases into the saturated Chalk. At the same
time a 76-m (249-ft) deep observation borehole was
sunk at location J by MAFF in connection with
the experimental storage of animal waste slurry in
unlined Chalk pits (Chumbley et al., 1976).
Boreholes A to D were drilled during the
period October 1974 to January 1975 and
boreholes E to F in October 1975. All are situated
in Rhode Island field (Figure 9) which since 1948
has been rotated between cereals and kale with
periods of ley. Details of land usage together with
nitrogen fertilizer applications are given in Figure
10. Borehole F was located in a part of the field
which had only been cleared of woodland in 1951.
Borehole G (Figure 9) was drilled in Maryland
field during November 1975. This field was in
continuous use as rough grazing from 1949 to
200 400 600 800 1000m
i i ! i |
KEY. Boreholes
• A
'%M Fields for v.hich detailed land use and fertilizer records
ym: have been tabulated are shaded
Fig. 9. Location of investigation sites and boreholes on the
Bridget's Farm site.
Fig. 10. Changes in land use and fertilizer application
1948-1974, Rhode Island field-Boreholes A to F.
19iB i9 50 51 52 53 Si 55 56 57 56 59 60 61 62 63 61 65 66 67 68 69 70 71 72 73 7i
YEAR
Fig. 11. Changes in land use and fertilizer application
1948-1974, Maryland field-Borehole G.
!9i8 i9 50 51 52 53 54 55 56 57 58 59 60 61 62 63 6t 65 65 67 63 69 70 7! 72 73 7i 75
Fig. 12. Changes in land use and fertilizer application
1948-1974, New Hampshire north field-Borehole H.
1969 followed by ploughing and the growing of
spring barley in 1970 and 1971, since when it has
been returned to ley (Figure 11). Borehole H was
drilled in New Hampshire north field during
October 1975, land-use records being shown in
Figure 12.
4.2. Nitrate, Chloride and Tritium Profiles
The change in concentration of nitrate with
depth below ground surface is plotted for bore-
holes D, E, F, G and H in Figures 13, 14, 15, 5 and
16 respectively. All show peak values that exceed
the WHO 11.3 mg/1 NO3-N recommended limit.
Boreholes D and E were sited within 10 m of
each other in the same field and while the general
shape of their profiles is similar, the integrated
nitrate concentration to a depth of 29 m (95 ft)
in borehole E is 32% greater than that in
borehole D.
Chloride determinations were only made on
the pore water from borehole G (Figure 17). The
source of the chloride is attributable to the appli-
75
-------
N03- N (mg/IJ
10 20 30
Q-
LU
O
50J
WHO recommended
limits of 11-3,22-6mg/l
Rest water level
at 56m below ground
Fig. 13. Measured nitrate profile—Borehole D—Rhode
Island field.
cation of potassium chloride fertilizer in 1957,
1964 and 1965.
The concentration of tritium was measured
on chalk samples from boreholes A, B and C. All
show very similar profiles and as borehole C is
the most complete, it is presented in Figure 18.
4.3. Unsaturated Zone Flow Model
The characteristics of the Chalk have been
described in section 4.1. Moisture content
determinations on samples of unsaturated Chalk
Water Table 10/75
60J
Fig. 14. Measured and computed nitrate profile—Borehole
E-Rhode Island field.
76
-------
N03-N (mg/I)
0
computed^ [_ ^.^
-x-^__
WHO recommended
limits of 11-3,22-6mg/l
3CM
Fig. 15. Measured and computed nitrate profile—Borehole
F-Rhode Island field.
suggest that the intact blocks between the fissures
are almost fully saturated, the water being held by
surface tension forces. Water infiltrating from the
surface and moving downwards by intergranular
flow would migrate at a rate of 1 m/yr if typical
values of permeability and porosity are assumed
(1 X 10"3 m/d and 0.35). This is a similar rate to
that reported by Smith et al, 1970, for the
migration of tritium through unsaturated Chalk.
However, it is difficult to discount the multitude
of fissures that are so evident in Chalk exposures, as
having no place in the flow process. These fissures
are the main flow channels in the saturated zone,
are responsible for the aquifer's high trans-
missivity, and give a 1 to 2% effective porosity.
It has been suggested (Foster, 1975; Young
et al, 1976) that relatively rapid infiltration through
the unsaturated zone may take place along joints
and fissures and that solutes may diffuse between
the moving water in the joints and the relatively
static water in the matrix. This may account for
the observed relatively slow downward movement
of tritium and nitrate. It is uncertain whether there
would be sufficient time to allow effective diffusion
to take place but assuming this, it can be shown
that the unsaturated Chalk column behaves in a
similar manner to a chromatogram. When
equilibrium is established the solute ions divide
between the static pore water and the fissure
water in the ratio of their respective volumes. It
can also be shown that the ratio of the total
porosity to fissure porosity is equal to the ratio
of the velocity of the water in the fissures to the
downward velocity of the contaminant front. The
properties of the unsaturated Chalk are such that
the infiltrating water would move downwards at a
rate some 20 to 40 times faster than the associated
contaminants. Sharp input peaks would persist
during downward migration provided that rapid
equilibrium is attained but some blurring would
occur due to dispersion. If fissure flow is
sufficiently slow, then nitrate, chloride and
0
NO -N{mg/I)
10 J 20 30
0
•£10-1
Q_
LU
20-
measured
V computed
>
\
WHO recommended
limits of 11'3,22-6mg/l
25J
Fig. 16. Measured and computed nitrate profile-Borehole
H—New Hampshire north field.
77
-------
CHLORIDE (mg I/I)
10 20 30
5-
i-
Q_
LU
Q 15H
20-
25J
measured
computed
Rest water level
Oct'75
Fig. 17. Measured and computed chloride profile—Borehole
G-Maryland field.
tritium fronts would move at very similar rates.
Further experimental work is necessary
before the flow mechanism in the unsaturated
Chalk is fully understood. However, both the
intergranular flow and fissure flow concepts
suggest that a uniform velocity, downward-flow
model may be used to represent the movement of
contaminant inputs from the surface. This simple
model was therefore adopted. Some attenuation of
the contaminants was allowed for by applying a
vertical dispersion factor in the model.
Many wells in the Chalk give a rapid rise in
water levels following a period of rainfall. In
some cases it has been possible to calculate that
this represents 10 to 15% of the infiltration
resulting from the rain, and it is believed that the
rapid rise is due to flow through large fissures in
the unsaturated zone. It has therefore been assumed
in the model that 15% of the annual infiltration
together with any associated contaminants moves
rapidly to the water table.
4.4. Production of Nitrate in the Soil Profile
Crops assimilate nitrate during the growing
season and within limits the greater the application
of nitrate the greater the yield. A proportion of the
vegetable matter is removed on harvesting but the
residue of roots, weeds and microflora remains,
adding to the soil organic matter. In this organically
bound form the residual nitrogen is not available
to other plants but must be mineralised by specific
soil bacteria which are active from March to
December when the soil temperature exceeds 5°C.
The mineralised nitrogen is taken up by growing
crops during the period March to September but
mineralisation continues into early winter during
which period precipitation exceeds evaporation
so that infiltration through the soil and into the
unsaturated aquifer occurs. Nitrate is thus carried
downwards with the infiltrating water.
Kolenbrander (1975) estimated that for cereals
and root crops, 50% of the applied fertilizer
becomes available as organic matter for
mineralisation. This value was used in the model,
and it was further assumed that it could all be
leached from the soil. However, not all the material
is available in the year of application and on the
basis of Kolenbrander's work it was assumed that
mineralisation took place over a 3-year period. The
TRITIUM UNITS
100 200
10-
^§
I
20 H
LU
Q
30-
300
i
measured
Fig. 18. Measured and computed tritium profile—Borehole
C-Rhode Island field.
78
-------
amount of nitrate available for leaching as a
fraction of the application rate is given in Table 1.
Nitrifying bacteria are aerobic and are thus
normally restricted to the upper aerated rooting
zone of the soil. Compaction under permanent
grassland or leys may reduce the thickness of the
aerated zone and lead to anaerobic conditions at
depth in the soil profile. Anaerobic bacteria in this
zone may reduce the nitrate in the infiltrating
water to gaseous nitrogen. However, even under
permanent grass a small proportion of the mobile
nitrate will be leached through the soil zone.
Williams (1975) has reported leaching values of
about 4 mg/1 NO3-N from below established grass/
lucerne leys and this is in accordance with the
nitrate profiles obtained for unfertilized permanent
grassland sites and shown in Figure 4. This value of
4 mg/1 NO3-N was therefore adopted in the model
for the nitrate content in water leaching from
established grass.
If a soil is undisturbed, organic material may
accumulate and Kolenbrander (1975) has shown
that equilibrium between the inputs and losses
by mineralisation may only be achieved after a
period of between 10 and 40 years. Naturally
developed soil profiles are destroyed by ploughing,
and if this takes place at the end of the growing
season in autumn, the creation of aerated conditions
favours mineralisation and leads to nitrate loss. High
concentrations of nitrate in drainage water from
ploughed grassland have been reported (Cooke
and Williams, 1970; Kreitler and Jones, 1975;
Reinhorn and Avnimelech, 1974; Olsen et al.,
1970).
Table 1. Nitrogen Available from Arable Crops
[N (kg/ha) available as a fraction of the fertilizer
application rate N (kg/ha)] (1 kg/ha = 0.89 Ib/acre)
Table 3. Solute Migration Rates
Roots
Cereals
Year of
application
0.35
0.25
1 st year
following
application
0.10
0.15
2nd year
following
application
0.05
0.10
Table 2. Nitrogen Available from Ploughed Grassland
Years
in grass
prior to
ploughing
1
2
3
4 or more
N (kg/ha) released by ploughing
Total
30
60
170
260
Year of
ploughing
18
36
100
155
1st year
following
ploughing
9
18
50
80
2nd year
following
ploughing
3
6
20
25
Borehole
E
F
G
H
A and B
C
G
Solute
nitrate
nitrate
nitrate
nitrate
tritium
tritium
chloride
Migration rate
(m/yr)
1.05
0.84
1.00
1.05
1.05
0.80
1.00
The data on the rate of buildup of soil organic
matter below grassland is, however, rather sparse
but from, information based on the work of Smith
and Young, 1975, an estimate of the amount of
nitrogen released by ploughing grassland of various
ages was established and the results are shown in
Table 2.
4.5. Application of Model to the Profiles
The land-use records, fertilizer applications,
and estimates of infiltration rates for the period
1948 to 1974 were used together with the
nitrate release rules in the constant rate flow model
to predict the nitrate profiles in the unsaturated
zone. The measured and computed profiles for
boreholes E, F, G and H are given in Figures 14, 15,
5 and 16.
Using the same model the tritium profiles
from 1954 to 1975 were calculated for'boreholes
A, B and C. A comparison of the measured and
computed profile for borehole C is given in Figure
18.
An attempt was made to model the chloride
profile obtained from borehole G. The historical
information on the application of potassium
chloride fertilizer is not very reliable but for the
years 1957, 1964 and 1965 is approximately 60,
64, 54 kg/ha Cl. Chloride in rainfall at the Bridget's
Farm site was estimated to be 15 mg/1 Cl. This
data was used in the model and the measured and
computed profiles shown in Figure 17.
The downward migration rate for the con-
taminant used in the model was in all cases adjusted
so as to give the best match between the measured
and computed data and these values for each profile
are given in Table 3. The rates all fall in the range
0.8 to 1.05 m/yr (2.6 to 3.4 ft/yr) and there appear
to be no significant differences between solutes.
The authors are unaware of any experimenta-
tion on the relative mobility of various ions in the
Chalk. However, in other media, similar migration
rates for nitrate and chloride have been reported
from both field and laboratory studies (Kurtz and
79
-------
Melsted, 1973; Adriano et al, 1971; Garwood
and Tyson, 1973; Wild and Babiker, 1975).
The tritium profiles are particularly important
as in this case the surface inputs are accurately
known. An excellent comparison between the
computed and measured results for both the peak
position and size is obtained (Figure 18). The
tritium migration rates obtained are in close
agreement with those measured by Smith et al.
(1970) in the unsaturated Chalk of the Lambourn
area some 50 km (31 mis) from the Bridget's Farm
site.
The computed nitrate profiles, due to the
uncertainties concerning inputs, do not match
the measured profiles so well as those for tritium.
No attempt was made to simulate the measured
profile in borehole E below a depth of 28 m (92 ft)
as this corresponded to a pre-1948 land-use activity
for which data is not available. The results from
boreholes F and G suggest that the rules governing
the release of nitrate from ploughed land require
some refinement.
There are anomalies between the profiles in
boreholes D and E. The holes are adjacent
to each other and are located in the same field.
Nitrate peaks occur at the same depth but the
magnitude of the peaks in borehole D are
generally less than those in borehole E. The
position of the peaks supports the hypothesis of a
uniform downward migration rate for nitrate but
the differences in peak magnitude suggest that the
nitrate inputs from the soil may not be uniform in
any one field. This may be due to differences in
the soil structure or local irregularities in the
Chalk. However, the over-all behaviour of the
model in matching the nitrate, chloride and
tritium profiles is generally good.
4.6. Future Concentrations of Nitrate in
Ground Water — Predictive Models
The concentration of nitrate at any point in
the saturated aquifer is a function of the nitrate
content of the water draining from the unsaturated
zone above that point and of additions from ground
water flowing into the area. Vertical mixing in the
saturated zone is likely to be slow and inputs of
differing concentrations along a flow path may
remain distinct for some distance. However, wells
and boreholes effectively mix the water over their
depth of penetration so that, unless specialised
sampling techniques are employed, an average
sample is obtained. Pore water samples collected
from cores below the water table will reflect the
composition at various levels. This may be seen
from the profiles of boreholes E and G (Figures
14 and 5) which penetrate 4 and 2.3 m (13 and 8
ft) respectively below the water table. Borehole E
in particular shows a series of peaks increasing to
10 mg/1 NOs-N at a depth of 3 m (10 ft) and these
are possibly discreet inputs that have occurred at
an upgradient situation from borehole E some
time in the past. This concept is shown diagram-
matically in Figure 19.
If the downward flow model for the
unsaturated zone is accepted as being valid, it
may be used to predict the rate at which nitrate
already in the profile and future surface inputs
would enter the ground water. This approach was
applied to borehole E by assuming that the past
agricultural practice and infiltration at this site will
be repeated cyclically in the future. Figure 20
shows the model predictions from 1950 to 2030.
From 1950 to 1994 the 4 mg/1 NO3-N entering the
ground water is due entirely to rapid fissure flow
from the surface. From 1994 onwards the high
nitrate concentrations currently present in the
profile reach the water table and levels rise to
peak concentrations of nearly 30 mg/1 NO3-N
in the year 2010. It must be stressed that the
results presented in Figure 20 represent the
conditions only under a single area of one field.
In reality there would be a range of migration
rates (Table 3), and this will have the effect of
smoothing the nitrate inputs to the water table by
lowering the peaks and raising the troughs of the
curve shown in Figure 20.
Estimates of the expected nitrate concentra-
tions in ground water were obtained from a
catchment model in which infiltrating water and
nitrate were routed through the saturated zone of
the aquifer. The sub-catchment used in this study
is shown in Figure 6, and encompasses most of the
area of Bridget's Farm. The sub-catchment was
divided into five zones, based on the available
records of land use. In three of the zones, land use
was identified as being similar to that practised in
the Rhode Island field (Figure 10), and in one other
zone land use was similar to that practised in the
Fig. 19. Sketch showing postulated mechanism responsible
for variations in nitrate concentration with depth below the
water table.
80
-------
Fig. 20. Model predictions for period 1950 to 2030.
(a) Nitrate concentrations entering water table from
below site E.
(b) Average nitrate concentrations in ground water
below farm site.
New Hampshire north field (Figure 12). The fifth
zone comprised woodland. Mean depth to water
table was estimated for each zone. The vertical
flow model was used to estimate the nitrate flux
across the water table in each zone from. 1950 to
2030 with the assumption that past agricultural
practice and infiltration rates would be repeated
cyclically in the future. It was assumed that there
would be a range of downward migration rates
through the Chalk in each zone, which results in
smoothing of the nitrate inputs to the water
table. Analysis of the available data suggested the
migration rates were normally distributed with
mean 1 m/yr (3.3 ft/yr) and standard deviation
0.07 m/yr (0.2 ft/yr), and the nitrate flux across
the water table was calculated accordingly.
A simple catchment model, assuming
complete mixing of water and nitrate in each zone,
was used to calculate mean concentrations of
nitrate in ground water beneath borehole E, and
in the spring discharge at the River Itchen. Times
of transit through the aquifer were estimated to
be small compared with times of transit through
the unsaturated zone, having a maximum value of
about 5 years. The catchment model predictions
of nitrate in ground water beneath borehole E
are shown in Figure 20. The estimate for 1975 is
5.7 mg/1 which compares well with a measured
value of 5.6 mg/1 (see Figure 6). The peak value
of 15.5 mg/1 occurs in 2008 with a long-term
average of about 13 mg/1. The steady rise in
concentrations from 1970 onwards results from
contributions from upgradient zones where the
unsaturated thickness is less than at borehole E
and the delay times are correspondingly less. The
predicted concentrations in the spring discharge
rise progressively from 6.7 mg/1 in 1975 to about
14 mg/1 from 2000 onwards. The measured
concentration in 1975 was 6.9 mg/1.
4.7. Denitrification
The predictive models described in section
4.6 assumed that denitrification does not occur
once the nitrate has left the soil and entered the
unsaturated aquifer. Little direct evidence has
been obtained to suggest that significant biological
action takes place at depth. However, the carbo-
hydrate content of the Chalk in borehole E has
been measured at 0.1 g/kg dry Chalk. This
corresponds to a carbon content of 120 mg/1 of
pore water which, if available as an energy source,
could enable denitrification to the extent of about
100 mg/1 NO3-N (Young, et al, 1976) to take
place. If reliable predictions are to be made of the
future concentrations of nitrate in ground water, it
is important to establish whether significant
denitrification within the aquifer is taking place,
and work on this objective is continuing at the
Centre.
5. CONCLUSIONS
1. High concentrations of nitrate, in excess of
the WHO upper recommended limit, have for the
first time been measured in the interstitial water
of the unsaturated and saturated zones at a number
of sites on the two principal aquifers in the United
Kingdom.
2. Quantitative relationships have been
established between the amount of nitrate in the
profiles and the historic land use and fertilizer
applications. Fertilized arable/ley sites have high
concentrations, fertilized permanent grassland
sites intermediate concentrations, and
unfertilized grassland sites low concentrations of
nitrate in their profiles.
3. Detailed analysis at a site on the Chalk near
Winchester has led to the formulation of a flow
model for the unsaturated zone which, on the
basis of the historic land use, predicts
nitrate profiles which are in good agreement with
those measured. The validity of the model has
been checked against measured tritium and
chloride profiles.
4. A series of such models has been integrated
into a catchment model which predicts that the
nitrate concentrations in the Chalk ground water
at the Winchester site will rise progressively from
its present value of about 4 mg/1 N03-N, reaching
14 mg/1 NO3-N during the early 2000's. The models
have assumed that little or no denitrification takes
81
-------
place within the aquifer. The modelling studies
are being extended to other sites.
6. ACKNOWLEDGEMENTS
The authors gratefully acknowledge the
facilities made available to the Centre's field studies
programme by Mr. P. J. Jones, Director, and
Mr. A. L. Francis, Deputy Director of Bridget's
Farm. We also wish to thank the water authorities
for supplying data and Mr. K. Gostick of MAFF
for providing data on land use and nitrogenous
fertilizer.
The authors are grateful to Dr. R. G. Allen,
Director of the Water Research Centre for
permission to publish this paper.
7. REFERENCES
Adriano, D. C., P. F. Pratt and S. E. Bishop. 1971. Nitrate
and salt in soils and ground water from land disposal
of dairy manure. Soil Sci. Soc. Am. Proc. 35, pp.
759-762.
Chumbley, C., E. Gray and M. Appleton. 1976. An investi-
gation into groundwater pollution from a cow slurry
lagoon. Proceedings—Water Research Centre Conf.
on Groundwater Quality—Measurement, Prediction
and Protection, University of Reading.
Comley, H. H. 1945. Cyanosis in infants caused by nitrates
in well water. Journ. American Medical Assoc. 129,
p. 112.
Cooke, G. W. and R.J.B. Williams. 1970. Losses of nitrogen
and phosphorus from agricultural land. Water Treat.
Exam. 19(3)253-276.
Davey, K. W. 1970. An investigation into the nitrate
pollution of the Chalk borehole water supplies. North
Lindsey Water Board, Scunthorpe.
Edmunds, W. M. and A. H. Bath. 1975. Centrifuge
extraction and chemical analysis of interstitial waters.
Institute of Geological Sciences, Report WD/ST/75/
11 (now in publication).
Foster, S.S.D. 1975. The Chalk ground-water tritium
anomaly—a possible explanation. J. Hydrol. 25,
159-165.
Foster, S.S.D. and R. I. Crease. 1974. Nitrate pollution of
Chalk ground water in East Yorkshire. A hydro-
geological appraisal. J. Inst. Wat. Engs. 28, pp.
178-194.
Foster, S.S.D. and V. A. Milton. 1974. The permeability
and storage of an unconfined Chalk aquifer. Hydrol.
Sci. Bulletin. 19, pt. 4, pp. 485-500.
Garwood, E. A. and K. C. Tyson. 1973. Losses of nitrogen
and other plant nutrients to drainage from soil under
grass. J. Agr. Sci. Camb. 80, pp. 303-312.
Gray, Mrs. E., Mrs. J. Holland, Miss J. Breach and C. D.
Rowland. 1976. Nitrate in ground water—procedures
in the collection and preparation of rock and water
samples. Proceedings—Water Research Centre Conf.
on Groundwater Quality—Measurement, Prediction
and Protection, University of Reading.
Greene, L. A. and P. Walker. 1970. Nitrate pollution of
Chalk waters. Wat. Treat. Exam. 19 (2), pp. 169-182.
Kolenbrander, G. R. 1975. Nitrogen in organic matter and
fertilizer as a source of pollution. In. Proc. Int. Ass.
Wat. Poll. Res. Specialised Conf. Nitrogen as a Water
Pollutant, Copenhagen.
Kreitler, C. W. and D. C. Jones. 1975. Natural soil nitrate:
the cause of the nitrate contamination of ground
water in Runnels County, Texas. Ground Water.
13 (l)pp. 53-61.
Kurtz, L. T. and S. W. Melsted. 1973. Movement of
chemicals in soil by water. Soil Science. 115 (3)
pp. 231-239.
Lund, L. J., A. L. Page and C. O. Nelson. 1976. Nitrogen
and phosphorus levels in soils beneath sewage
disposal ponds. J. Environ. Qual. 5 (1) pp. 26-31.
Mielke, L. N. and J. R. Ellis. 1976. Nitrogen in soil cores
and ground water under abandoned cattle feedlots.
Journ. Environmental Qual. 5 (1) pp. 71-75.
Olsen, R. J., R. F. Hensler, O. J. Attoe, S. A. Witzel and
L. A. Peterson. 1970. Fertilizer nitrogen and crop
rotation in relation to movement of nitrate nitrogen
through soil profiles. Soil Sci. Soc. Am. Proc. 34
pp. 448-452.
Petukhov, N. I. and A. V. Ivanov. 1970. Investigation of
certain psychophysiological reactions in children
suffering from methaemoglobinaemia due to nitrate
in water. Hyg. Sanit. 35, 29-32.
Reinhorn, T. and Y. Avnimelech. 1974. Nitrogen release
associated with the decrease in soil organic matter
in newly cultivated soils. J. Environ. Qual. 3 (2)
pp. 118-121.
Satchell, R.L.H. and K. J. Edworthy. 1972. Artificial
recharge: Bunter Sandstone, Trent Research Pro-
gramme, v. 7. Water Resources Board, Reading.
Smith, D. B., P. L. Wearn, H. J. Richards and P. C. Rowe.
1970. Water movement in the unsaturated zone of
high and low permeability strata by measuring
natural tritium. Proc. Symp. Isotope Hydrology.
IAEA. Vienna, pp. 73-87.
Smith, S. J. and L. B. Young. 1975. Distribution of
nitrogen forms in virgin and cultivated soils. Soil
Science. 120 (5) pp. 354-360.
Southern Water Authority. 1975. Itchen groundwater
regulation scheme, 2nd Progress Report on Caudover
Pilot Scheme, Southern W.A. November.29 pp.
Ward, W. H., J. B. Burland and R. W. Gallois. 1968. Geo-
technical assessment of a site at Mundford, Norfolk,
for a large proton accelerator. Geotechnique. 18,
pp. 339-431.
Wild, A. and I. A. Babiker. 1975. Winter leaching of
nitrates at sites in southern England. Paper no. 7,
ADAS/ARC Conf. Sutton Bonnington, 16 pp.
Williams, R.J.B. 1975. The chemical composition of rain,
land drainage and borehole water from Rothamsted,
Saxmundham and Woburn experimental stations.
In. Techn. Bull. Ministr. Agric. Fish, Fd., London,
no. 32.
Young, C. P. and E. S. Hall. 1976. Investigation into factors
affecting the nitrate content of groundwater.
Proceedings—Water Research Centre Conf. on
Groundwater Quality—Measurement, Prediction and
Protection, Reading.
Young, C. P., E. S. Hall and D. B. Oakes. 1976. Nitrate in
groundwater—studies on the Chalk near Winchester,
Hampshire. Water Research Centre, Technical
Report, 56 pp.
82
(Mr. Wilkinson's Discussion- Questions and Answers appear on page 230.)
-------
The Contribution of Fertilizer to the
a
Ground Water of Long Island
by Joseph H. Baier and Kenneth A. Rykbostc
ABSTRACT
In 1973, the Suffolk County Department of Environ-
mental Control in cooperation with Cornell University
began a study on nitrogen (N) fertilization of potatoes and
turfgrasses. The research and demonstration project has
shown that current practices result in substantial N losses;
and, in many cases, excessive use of N reduces crop yields
and turf quality. Annual N losses of 50 Ibs. per acre
(55.5 kg-N/ha) are sufficient to cause a concentration in
the aquifer's surface layer of 10 mg/1 nitrate-N (New York
State Drinking Water Standard).
The eastern portion of Long Island supports a
productive agricultural industry whose main crop is
potatoes. Ground-water surveys have shown that the
aquifer system of this area is contaminated with nitrate
nitrogen. The average potato grower applies 200 to 250
Ib-N/a (222 to 278 kg-N/ha) at planting time; and depending
upon a number of factors, N recovered in harvested tubers
varies from 75 to 150 Ib-N/a (83 to 167 kg-N/ha). Losses
to the ground water could vary from 50 to 175 Ib-N/a
(55.5 to 194 kg-N/ha). The study has shown that the
application of 150 Ib-N/a (167 kg-N/ha) can still maintain
maximum potato yields and keep the N loss to ground
water below 50 Ib-N/a (55.5 kg-N/ha) by improving
nitrogen-use efficiency. This is done through splitting N
applications so that one-third to one-half is applied at
planting and the remainder is applied prior to the period of
rapid crop growth and nutrient uptake. On-farm demonstra-
tion plots are being used to convince growers to reduce N
rates and adopt more efficient application methods.
The western portions of Long Island are highly
urbanized and turfgrasses may be fertilized at rates up to
350 Ib-N/a (389 kg-N/ha). Potential leaching losses are high
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
DChief, Fresh Water Resources Section, Suffolk
County Department of Environmental Control, 1324
Motor Parkway, Hauppauge, New York 11787.
°Formerly Project Coordinator and Research
Associate, Long Island Vegetable Research Farm, affiliate
of Cornell University, Ithaca, New York; presently Crop
Scientist, McCain Foods Ltd., Florenceville, New Brunswick,
Canada.
for turfgrass as N is not recovered in harvested plant
materials. Experiments are underway to establish rates
of biomass N buildup under several fertilization regimes.
Preliminary results indicate that N-use efficiency increases
with more frequent but smaller N applications. Encouraging
the use of low maintenance turf species appears to be the
best long-term solution.
INTRODUCTION
Long Island is a rapidly growing area with
fixed land and water resources. Water for all uses
(industrial, commercial, agricultural and municipal)
is obtained from ground water. The quality of this
resource is most important and is of particular
concern on the eastern portion of Long Island
where a limited quantity of water exists.
Agricultural areas are pointed to as having nitrate
concentrations approaching the drinking water
standard of 10 mg/1 nitrate-nitrogen (Holzmacher
. . . 1970). A later study for Southold Township
concluded: "It is perfectly clear that high nitrates
in Southold Township are due primarily to
fertilizers" (Hol/macher . . . 1972). Some typical
nitrate results from wells in the area are shown in
Table 1.
The hydrogeology of the farming areas (to be
discussed later) limits the amount of water
available. This, together with the fact that only
expensive long-distance water transmission could
Table 1. Typical Nitrate (N) Concentrations in
Eastern Long Island Ground Water
Well Location Date Sampled NO3-N (mg/l)
Mattituck
Greenport
Southold
East Mattituck
Test Plot (Demarest)
Test Plot (Dickerson)
Test Plot (Halsey)
Test Plot (Babinski)
10/22/75
10/14/75
11/03/75
09/16/75
03/26/75
03/26/75
03/26/75
03/26/75
16.3
8.2
12.0
13.1
23.0
21.0
20.0
17.0
83
-------
alleviate the drinking water quality problem,
indicates that to preserve and protect the eastern
Long Island water resource the source of the
nitrates, including fertilizers, must be limited or
reduced.
Suffolk County ranks first in New York State
in total agricultural sales. Potatoes, the principal
field crop, accounts for about 25,000 acres
(10145 ha) of the total of over 60,000 acres
(24420 ha) in agricultural production. Other
important agricultural activities include production
of ducks, sod, nursery products, and a wide variety
of fruits and vegetables (Suffolk County CES,
1973).
Potatoes and vegetable crops require intensive
management, including relatively high rates of
fertilization, for maximum production. In the past,
the consequences of over-fertilization have been
minor in contrast with yield losses associated
with under-fertilization, and a trend of over-fertiliza-
tion has developed. Environmental concerns and
drastic increases in fertilizer prices have resulted
in the need for a reappraisal of fertility management.
Results of a potato fertilization survey
(Rykbost, 1976) indicate that application rates
range from 125 to 310 Ib-N/a (A) (139 to 344
kg-N/ha) with a median of 225 Ib-N/a (250 kg-N/ha).
Most of this is applied at planting. A significant
portion of the N applied at planting is susceptible
to leaching to the ground water prior to the period
of maximum plant uptake. Nitrogen removal in the
harvested tubers account for 75 to 150 Ib-N/a (83 to
167 kg-N/ha) (Meisinger, 1976) depending
primarily on crop yield.
Plant tops may contain up to 100 Ib-N/a
(111 kg-N/ha) at harvest, but this goes into the
soil organic nitrogen pool which supplies
approximately an equivalent amount to the next
year's crop.
Since soil conditions are not favorable for
denitrification or volatilization processes, it seems
likely that significant quantities of nitrate-N
may be leached to the aquifer under present
fertilization practices.
Recharge of ground water on Long Island
averages about 23 inches per year (58.4 cm).
A loss of 50 Ib-N/a (55.5 kg-N/ha) via leaching will
be sufficient to maintain this amount of recharge at
10 ing/1 NO3-N, assuming no other nitrogen source.
A change in potato fertilization practice is necessary
to restore the eastern Long Island ground water to
an acceptable level of nitrate. However, this must
be accomplished without jeopardizing the potato
industry.
Fertilization of home lawns and other turf-
grass areas can also serve as a source of ground-
water nitrates. A 1973 survey of fertilization
practices for Long Island turfgrass indicated a
range in fertilization rates from zero to about
350 Ib-N/a (389 kg-N/ha) a year. While much of
the N may be taken up by turf, N removal from
the biosphere is nil, even in the case of removal of
clippings since these are composted or used for
mulch and eventually decay. Volatilization of N
from turfgrass may be an important N sink;
however, leaching of nitrates from heavily
fertilized lawns is undoubtedly a source of
ground-water contamination (Snow, 1976).
During the spring of 1973, the Suffolk County
Department of Environmental Control (SCDEC),
Cornell University, acting through the Vegetable
Crops, Agricultural Engineering, and Agronomy
Departments and the Long Island Vegetable
Research Farm (VRS), and the Suffolk County
Cooperative Extension Service (CES) began
discussion on how the fertilizer question could be
investigated. By April 1974, a contract was signed
by SCDEC and Cornell University to perform a
four-year study of N fertilization of potatoes and
turfgrass.
The study was designed to formulate and
evaluate alternative N fertilization schemes which
would reduce nitrate leaching losses while
maintaining potato yields and turfgrass quality.
Initial stages of the study were to be conducted
on the VRF facility. Promising alternative
fertilization practices were to be demonstrated and
evaluated on commercial potato fields and typical
turfgrass areas. Finally, a public information
program would be formulated by CES to encourage
adoption of recommended practices.
The duration of the study is presently June
1978 which will allow for three full growing
seasons at the demonstration sites. Final
recommendations will be based on five full years
of intensive research efforts. The importance of
longevity in a study of this nature cannot be
overemphasized.
The results thus far are of sufficient
importance that the authors felt that an interim
report should be presented to the scientific
community before the entire study has been
completed.
BACKGROUND
Long Island is the easternmost part of New
York State and extends northeastward parallel to
the continental coastline (Figure 1). It is bounded
84
-------
Fig. 1. Map of Long Island.
on the north by Long Island Sound, on the east and
south by the Atlantic Ocean, and on the west by
New York Bay and the East River. Politically, Long
Island is divided into four counties: Kings, Queens,
Nassau and Suffolk Counties. Kings and Queens
Counties are part of New York City.
The topographical features of Long Island are
the result of the glacial advance which occurred
during the Pleistocene epoch that ended approxi-
mately 10,000 years ago. The major physiographic
features (Figure 2) are: the terminal moraines that
form the northern and central hills and "forks" of
the island; the gently sloping outwash plain that
extends southward from the hills; the deeply
eroded headlands along the north shore; and the
barrier beaches along the south shore. The total
length of the island is 120 miles (193 km); its
maximum width is about 23 miles (37 km); and
its area is 1,400 square miles (3,626 sq km)
(Suffolk County ..., 1976).
The ground-water system of Long Island
consists of five major geologic units (Figure 3). The
uppermost formation is termed the Glacial aquifer
consisting of sand and gravel of moderately high
Fig. 2. Major landforms of Long Island, New York.
permeability and having an approximate maximum
thickness of 400 feet (122 m). The Gardiner's
clay is a formation which appears most extensively
in the southern portion of the island and has an
approximate maximum thickness of 150 feet
(46 m). The Magothy aquifer is the largest forma-
tion consisting of coarse to fine sand of moderate
permeability and an abundance of silt and clay of
very low permeability. The approximate maximum
thickness of this aquifer is 1,000 feet (305 m).
Next is the Raritan clay which is a clay of
relatively low permeability and an approximate
maximum thickness of 300 feet (91 m). The final
formation which rests on bedrock is the Lloyd
aquifer consisting of sand and gravel of moderate
permeability which may be as thick as 300 feet
(91 m). These formations are saturated with water
to an uppermost level which forms the water table
(New York . . . 1968).
The water budget of Long Island may be
generally represented by the equation:
R = P - (r + E)
where R = natural recharge to the aquifer,
Solid Hni whir* opproilmottly
hnovn, dothcd llni whtf» kiftrrtd
HORIZONTAL SCALE
VERTICAL EXAGGERATION ABOUT 20X
Fig. 3. Hydrogeologic cross section of Long Island.
85
-------
P = precipitation,
r = direct runoff to streams and estuaries,
and E = evapotranspiration.
E is generally taken as 45 percent of P and r as
5 percent of P.
For Nassau and Suffolk Counties, ground water
is the sole source of water and rainfall is the only
replenishment.
Annual average precipitation is 44 inches
(112 cm) (New York . . . 1968) with long-term high
and low of 70 inches (178 cm) and 26 inches (66
cm) respectively (Suffolk County . . . , 1976).
The farming activity on Long Island is
primarily centered in eastern Suffolk County and,
more specifically, on the north and south forks.
These two areas differ hydro logically from the
main body of Long Island in that their aquifers are
underlain by salt water. This limits the amount of
water available for any type of use and causes
pollution sources to be of greater concern.
The nitrogen budget for potato production
involves a large number of input and output
components. After an exhaustive analysis of this
budget, Meisinger (1976) indicates that for Long
Island the simple equation: NF = NT + NL can be
used.
NF is N fertilizer applied in the crop year, NT
is N removed in the harvested tubers, and NL is
nitrate-N leached to the ground water. Since NT is
nearly a linear function of crop yield (Meisinger,
1976), leaching losses will be determined by
fertilizer N application rates and crop yields.
The above relationship suggests two
alternatives for reducing leaching losses. The first
and most obvious is to reduce N application rates.
The second is to increase the nitrogen-use
efficiency (NUE) by providing fertilizer N at the
time it is needed and in a form which is available
to the crop but not subject to leaching.
Applying most or all N at planting opens the
door for substantial losses during major storm
events in spring months. Under Long Island
conditions, a split-application scheme should
reduce spring leaching losses while maintaining
or improving yields, provided the timing can be
planned to coincide with crop needs. It was felt
that by improving the NUE and reducing NF, the
yields could be maintained and the nitrogen
leaching loss could be reduced.
Research conducted at the VRF over the past
two decades has shown that in most years most
potato varieties achieve maximum yields at rates of
150 to 160 Ib-N/a (167 to 178 kg-N/ha) (S. Dallyn,
86
unpublished data). In years when excessive
precipitation during the period from April through
May leaches significant amounts of N past the crop
root zone, yield responses to rates up to 200 Ibs. N
have been observed. Frequently, yield depressions
have occurred in response to rates above 160 Ib-N/a,
due to delayed maturity and/or salt injury to
seedlings. These data, coupled with grower practices
identified (Rykbost, 1976) suggest that reducing N
rates would not only reduce leaching losses but in
many cases would result in increasing crop yields
and dollar returns. Furthermore, crop quality may
be improved as excessive N fertilization frequently
results in an immature crop susceptible to bruising
during harvest and decay in storage (Smith, 1968).
Leaching losses from N fertilization of
turf grass areas are difficult to quantify. Fertilization
practices are varied and few records are kept. N
usage by turf varies greatly depending on type of
grass, mowing practices, irrigation practices, age of
the turf, frequency, rate and timing of N applica-
tion, and other factors. A 1973 survey of Long
Island turf fertilization practices (Rykbost,
unpublished data) indicated a range in N application
rates from 0 to 350 Ib/a (0 to 390 kg/ha).
Recent literature on turfgrass fertilization has
been reviewed by Snow (1976), and a glaring lack of
research relevant to the fate of N applied to turf is
evident. Substantial volatilization N losses may
occur from fertilizer as well as from decaying
clippings. During turf establishment, N is rapidly
incorporated into plant tissues and stable organic
debris. However, after an initial period of biomass
buildup, which may require 20 years or more, no
obvious N sink can account for N additions. Thus,
in established turf areas leaching losses from
fertilization of turfgrass may outweigh the losses
from fertilization of crop lands on Long Island.
POTATO EXPERIMENTS
Experiment I, conducted in 1972-1975, was
designed to determine the timing of crop N uptake,
and yield and NUE response to split applications.
Results (Schippers and Rykbost, 1975) indicated
that single N applications at planting were no better
than split applications for crop yield except during
seasons when heavy spring rains take place [11
inches (28 cm) of rain in June 1972], when
substantial leaching losses occurred from single
applications.
Split applications significantly increased tuber
N removal for the farm year average. Weekly
sampling of tops, tubers, and roots showed slow N
uptake (30 Ib-N/a - 33 kg-N/ha) until the 6-inch
-------
Table 2. Total Yield and Tuber N Removal for Five Varieties and Four N Rates in 1974 and 1975
N-Rate
Ib-N/a (kg-N/ha)
Variety
K
C
H
S
A
Avg.3
Avg.
Avg.
Avg.
19741
140(155)
140(155)
140(155)
140(155)
140(155)
80(89)
120(133)
160(178)
200(222)
19752
125 (139)
125(139)
125 (139)
125(139)
125(139)
50(56)
100(111)
150(167)
200(222)
Total
Yield4
lOOlbs/a (kg/ha XlO2)
1974
394 (437)bc
454 (504)a
400 (444)b
374(415)cd
365 (405)d
338(375)c
398 (442)b
427 (474)a
427 (474)a
1975
363 (403)b
415 (461)a
374(415)b
344(382)°
305 (339)c
328 (364)b
355 (394)ab
380(422)a
378 (420)a
Tuber N Removed*
Ib/a (kg/ha)
1974
112(124)a
118(131)a
108 (120)a
115 (128)a
100(lll)a
88 (98)b
113 (125)a
120(133)a
121 (134)a
1975
120(133)b
142(158)a
120(133)b
118(131)b
93 (103)c
93 (103)c
111 (123)b
132(147)a
138(153)a
1 Averaged over N rates of 80, 120, 160, and 200 Ibs/a (89, 133, 178, and 222 kg-N/ha).
2 Averaged over N rates of 50, 100, 150, and 200 Ibs/a (56, 111, 167, and 222 kg-N/ha).
3 Averaged over five varieties.
4 Means followed by the same letter are not significantly different at the 5 percent level, based on the Duncan's multiple
range test (Steel and Torre, 1960).
(15-cm) plant height stage. During the next three
weeks, over 50 percent of total seasonal N uptake
occurred. Thus, split applications must be carefully
timed to insure that the N is available to the plant
when needed.
Experiment II, conducted in 1973-75,
compared N rates from 0 to 200 Ib/a (0 to 223
kg/ha) single versus split applications and the
nitrification inhibitor "N-Serve" [2-chloro-6
(trichloromethyl) pyridine] versus no inhibitor
(Meisinger, 1976). A split N application with
50 Ib/a (55.7 kg/ha) applied at planting and 100
Ib/a (111 kg/ha) side dressed at the 6-inch (15.2-
cm) height stage consistently achieved maximum
yield. Split N applications significantly increased
the total three-year tuber N yield, hence NUE.
N-Serve delayed nitrification long enough to avoid
early season leaching losses, increased total three-
year tuber N yields, and increased residual
soil N. Meisinger (1976) concluded that the split
application at 150 Ib-N/a (167 kg-N/ha) would
produce maximum yields while maintaining
leaching losses at acceptable levels.
Experiments I and II were conducted with the
Katahdin potato variety which accounts for
two-thirds of the Long Island crop. Other varieties
will respond differently to a given fertilization
program. Experiment III was conducted in 1974-
1976 to evaluate five varieties at a range of N rates
using split applications. Included were three full-
season varieties: Katahdin (K), Cascade (C),
and Hudson (H), and two early maturing varieties:
Superior (S) and Alamo (A). In 1974, 40 Ib-N/a
(44 kg-N/ha) were applied at planting. At the
6-inch (15-cm) stage, side dress rates of 40, 80,
120 and 160 Ib-N/a (44, 89, 133, 178 kg-N/ha)
were applied.
In 1975, 50 Ib-N/a (56 kg-N/ha) were applied
at planting and side dress rates of 0, 50, 100 and
150 Ib-N/a (0, 56, 111, and 167 kg-N/ha) were
applied at the 6-inch (15-cm) stage. The inter-
actions between variety and N rate were not
statistically significant in either year for tuber yield
or tuber N removal. Table 2 shows average yields
and tuber N removals for both years. The most
important conclusions to be drawn from these
data are:
(1) Tuber yields were not increased in
response to the last increment of N.
(2) Tuber N removal increased only slightly
in response to the last two N increments, hence
NUE decreased rapidly.
(3) Yield and NUE are variety dependent.
(4) Based on the equation NF = NT + NL,
leaching losses exceeded 50 Ib-N/a (55.5 kg-N/ha)
in both years for the 200 Ib-N/a (222 kg-N/ha)
rate.
TURFGRASS EXPERIMENTS
A blend of four Kentucky Bluegrass varieties
was seeded in May 1974 (Snow, 1976). Nitrogen
rates of 0, 2, 4 and 8 lbs/1000 sq ft (0, .0098,
.0195 and .039 kg/m2) were applied in 0, 1, 2, 4
or 8 applications per year. Clippings were sampled
weekly in 1975 to determine dry matter production
and N removal. Biomass samples were taken in
November 1974 and March, September and
November 1975 to assess biomass N accumulation.
87
-------
By November 1975, total N accumulated in the
biomass where clippings were returned was 2.5, 3.3,
3.8, and 4.8 lbs/1000 sq ft (.0122, .0161, .0185,
.0231 kg/m2) for two-year total N applications of
0, 4, 8 and 16 lbs/1000 sq ft (0.0, .0195, .039, and
.0781 kg/m2) respectively. Where clippings were
removed, biomass N was somewhat less, 2.2, 2.8,
3.6 and 4.0 lbs/1000 sq ft (.0107, .0136, .0175 and
.0195 kg/m2) respectively, for each application.
These data indicate that a small percentage of
fertilizer N is incorporated in plant material during
the period of turf establishment. This percentage
can be expected to diminish as the turf matures.
Snow (1976) concluded that:
(1) Where clippings are returned, N fertiliza-
tion above a rate of 2 lbs/1000 sq ft (.0098 kg/m2)
may lead to ground-water contamination above
lOmg/lnitrate-N.
(2) Where clippings are removed, a large part
of the applied N is removed from the system, but
may eventually reach ground water. This is due to
the fact that the clippings are normally composted
by the homeowner or they decompose in the
local landfill. The nitrogen still returns to the
ground water.
(3) N recovery will decrease as the turf
matures, and high rates of N on mature turf may
lead to serious ground-water contamination.
(4) At low total N rates, more frequent
applications will provide for more uniform growth
and quality throughout the growing season.
(5) The use of low N-requiring varieties of
turf grass should be encouraged.
In September 1974, 18 Kentucky bluegrasses,
10 red fescues, 8 perennial ryegrasses, and 8
blended mixes were seeded under a minimum
fertility maintenance schedule of 2 lb-N/1000
sq ft (.0098 kg-N/m2) applied in V4 Ib (.22 kg)
increments in September and October 1974, and
April and May 1975. The replicated plots are
under routine observation to evaluate turf
performance and isolate those varieties which can
maintain acceptable visual quality. A minimum of
three years of evaluation will be needed to select
low maintenance turf grass. As expected, a wide
range in quality response has been observed to
date, with several varieties performing very well.
DEMONSTRATION PLOTS
The objectives of the potato fertilization
demonstration trials were to:
(1) Verify results from research conducted
at the Research Farm on commercial growers'
fields;
(2) Demonstrate alternative fertilization
practices which should result in equivalent yields,
reduced costs for fertilizer, and reduced levels of
nitrate nitrogen in ground waters underlying
agricultural fields;
(3) Monitor the impact of nitrogen
fertilization practices on ground water; and
(4) Encourage growers to adopt new practices.
Four demonstration sites were selected in
January 1975. Growers involved agreed to conduct
the trials for at least two years. Shallow
skimming wells (two wells per site) were installed
to allow monitoring of ground-water nitrogen
under "grower practice" and "recommended
practice" treatments.
The demonstration plots were divided in half
with the "recommended practice" of 1,670 Ib/a
(186 kg/ha) of 3-18-9 fertilizer and 300 Ib/a
(334 kg/ha) of aeroprills (34-0-0) applied at the
4-6-inch (10-15.2-cm) plant height, side dressed
in bands. The "grower practice" varied from farm
to farm (see Table 3).
The skimming wells were all suction well,
2-inch (5.08-cm) diameter, with a 5-foot (1.52-m)
stainless steel well screen. Top of screen was set
1 foot (.304 m) below the water table. Initial
samples were taken before planting and two-week
sampling intervals were followed thereafter. Samples
were collected after pumping approximately 50
Table 3. Demonstration Plot Information, 1975
Plot
No.
1
2
3
4
Location
North Fork
North Fork
South Fork
South Fork
Size
a (ha)
2.0 (.81)
1.6 (.64)
3.5(1.41)
2.5 (1.01)
Soil Type
Haven
Haven
Bridgehampton
Plymouth
Depth
to Water
ft(m)
3.3 (1.0)
16.5 (5.0)
11.0(3.3)
4.1(1.3)
Elev. ofGWT
at MSL, 3/76
ft(m)
0.7 (.2)
3.9(1.18)
4.0(1.21)
4.4(1.34)
Recommended
N-Rate
Ibs/a (kg/ha)
133 (148)
154(172)
149(167)
148 (166)
Grower
N-Rate
Ibs/a (kg/ha)
200 (224)
217(243)
270(302)
185 (207)
-------
Table 4. Demonstration Plot Results
Site 1- G.P.1
Orient Point R.P.2
Site 2- G.P.
Southold R.P.
Site 3- G.P.
Wainscott R.P.
Site 4- G.P.
Water Mill R.P.
Applied N
Ibs/a (kg/ha)
200(223)
133 (148)
217(241)
154(171)
270(301)
149(166)
185 (206)
148 (164)
Yield
Cwt/a (kg/ha)
384 (42800)
323 (36000)
259(28900)
256(28530)
259(28900)
225 (25000)
290(32300)
313 (34900)
Initial
Nitrate-N (mg/l)
3/20/75
16.3
25.0
21.8
20.8
8.3
8.0
23.4
24.9
Mid-Sampling
NO3-N(mg/l)
10/3/75
20
22.8
16.5
19.2
6.2
12.3
20.0
24.0
1-Year Sampling
NO3-N (mg/l)
3/1 5/76
14.2
18.9
16.4
21.0
11.9
5.9
13.5
14.1
1 Grower Practice.
2 Recommended Practice.
gallons (189 liters) and a micro-Kjeldahl steam
distillation procedure was used to determine
ammonium, nitrite and nitrate nitrogen.
Throughout the well sampling program,
ammonium and nitrite nitrogen levels in all wells
have been negligible in comparison to nitrate
nitrogen contents. Ground-water nitrogen in the
discussions will refer only to nitrate nitrogen.
The results from Site 1 are shown in Table 4.
Only 75% of the recommended second N applica-
tion was actually applied, and the method applica-
tion used was top dressing as opposed to the more
efficient side dressing method. These factors were
important in contributing to a lower potato yield
on the r.p. treatment.
The ground-water nitrate concentrations are
plotted in Figure 4. The one-year's sampling shows
a decrease in ground-water nitrate on both plots,
but the recommended rate shows a greater
decrease.
Site 2 was irrigated three times during the
growing season and no problems were encountered.
TEST PLOT NO I
Fig. 4. Nitrate-N results from test plots 1 and 2.
The yields were equivalent with a 63 Ib-N/a (70
kg-N/ha) fertilizer reduction.
The ground-water nitrate concentrations are
plotted in Figure 4. They show the ground-water
quality in the growers' field to be consistently
lower throughout the season (see comments
further on).
Site 3 was not irrigated during the growing
season, and the fact that the "recommended
practice" area did not receive the same weed
killing treatment as the "grower practice" area
contributed to the yield differential. Figure 5
shows a ground-water improvement in the
"recommended practice ".area after one year of
operation.
At Site 4 the yields were satisfactory and serve
to demonstrate that the reduced fertilizer applica-
tions can produce equivalent and even greater crop
yield. Nitrogen levels in both areas decreased on
both treatments (see Figure 5).
Figures 4 and 5 present graphs of the nitrate
nitrogen concentrations by site number, treatment,
and sampling date. Wide variations were observed
between sites, between treatments on a given site,
and between sampling periods for a given well.
Some generalizations can be made from the data.
(1) At each site, nitrate nitrogen contents in
excess of 10 ppm (and frequently more than double
or triple that value) were observed.
(2) Since these fields all have an agricultural
history, the major source of ground-water nitrogen
is fertilizer.
(3) As the season progressed, differences in
nitrate levels between treatments on a given site
gradually decreased. On June 12, 1975, the average
nitrate concentration was 9.0 ppm higher on the
"recommended practice" treatment. This difference
89
-------
Fig. 5. IMitrate-N results from test plots 3 and 4.
gradually dropped and on March 1, 1976, levels
were the same on both treatments.
(4) A statistically significant trend towards
lowering nitrate levels on the "recommended
practice" treatment can be seen in Figure 6.
However, an additional year's data will be needed
to confirm this apparent trend (VRF, 1976).
(5) A period of time is needed to allow for
soil and ground-water reaction and adjustment. This
time period appears to be at least a growing season.
A review of the 1975 ground-water nitrogen
data has resulted in the expansion of the ground-
water sampling network at each site from two
wells to six wells per site. In addition, detailed
ground-water directions were obtained and soil
analyses performed. Data collected for the 1976
growing season has shown that the nitrate levels
on all four fields is lower in the "recommended
CONG RECOMMENDED PRACTICE MINUS CONC.
60 I2O 180 240
X ' TIME FROM 3/20/75 (DAYS- 10)
3OO 36-0
Fig. 6. Nitrate-N regression line.
practice" than the "grower practice." Additionally,
at two of the sites, the quality of the ground water
is below 10 mg/1 nitrate-N standard (authors'
observation of 1976 data).
CONCLUSIONS
The following represent the important results
thus far:
(1) NUE decreases rapidly when fertilizer
rates exceed 150 Ib-N/a (167 kg-N/ha).
(2) Two fertilizer applications—one-third at
planting, two-thirds at the 4- to 6-inch (10-15.2-cm)
plant height—will result in maximum NUE and still
produce economical yields.
(3) Reducing fertilizer application to 150
Ib-N/a (167 kg-N/ha) with split applications appears
to result in equivalent crop yields, optimum NUE
and a reduction in nitrogen leaching, to a point
which will maintain ground-water quality below
10mg/l-nitrate-N.
(4) Nitrogen/fertilizer is a major source of
nitrate in ground water beneath agricultural areas.
(5) Immediate ground-water response to
changes in nitrogen applied to the soil cannot be
expected.
(6) The ground-water quality is changing in
direct relation to the amount of fertilizer applied.
(7) N fertilization of turfgrass has the potential
of contributing more to the ground-water deteriora-
tion than fertilization of potatoes.
If the study continues providing successful
crop yields at low fertilization rates and the
ground-water quality continues to show improve-
ment, a very serious ground-water pollution
problem will be solved and restoration of the
aquifer's quality can begin.
ACKNOWLEDGEMENTS
Thanks to Dr. A. Schippers for statistical and
scientific information and to various members of
the Department of Environmental Control for their
assistance in the preparation of the text.
REFERENCES
Frizzola, John A., and Joseph H. Baier. 1974. Contaminants
in rainwater and their relation to water quality.
Presented at the Fall Meeting of the New York State
Section of the American Water Works Association,
September 10-13, 1974, McAfee, New Jersey.
Holzmacher, McLendon and Murrell. 1970. Comprehensive
public water supply study, Suffolk County, New York.
90
-------
Holzmacher, McLendon and Murrell. 1972. Proposed
Cutchogue-Mattituck water district, Town of Southold,
New York.
Klein, J.V.N. 1973. Suffolk County farmlands preservation
program. Report to the Suffolk County Legislature.
Suffolk County Center, Hauppauge, New York.
Meisinger, J. J. 1976. Improving the efficiency of fertilizer
use for potatoes. Ph.D. thesis, Cornell University,
Ithaca, New York.
New York Water Resources Commission Bulletin 62. 1968.
An atlas of Long Island's water resources.
Rykbost, K. A. 1976. A survey of Long Island potato
fertilization practices and some comments. Suffolk
County Agricultural News. March.
Schippers, P. A., and K. A. Rykbost. 1975. Influence of
nitrogen fertilization on potato yields and nitrogen
use efficiency on Riverhead sandy loam soil. Presented
at the 67th Annual Meeting of the American Society
of Agronomy. Knoxville, Tenn. In preparation for
publication.
Smith, O. 1968. Potatoes: production, storing, processing.
The AVI Publishing Company, Inc. Westport,
Connecticut.
Snow, J. T. 1976. The influence of nitrogen rate and
application frequency and clipping removal on
nitrogen accumulation in Kentucky Bluegrass Turf.
M.S. Thesis. Cornell University, Ithaca, New York.
Steel, R., G., D. and J. H. Torre. 1960. Principles and
procedures of statistics. McGraw-Hill Book Company,
Inc., New York.
Suffolk County Cooperative Extension Service. 1973.
Comparative importance of Suffolk County, New
York agriculture. Unpublished mimeo. Suffolk
County Cooperative Extension Service.
Suffolk County Department of Environmental Control.
1976. Environmental assessment statement—lateral
sewer construction grants—Suffolk County Sewer
District No. 3—Suffolk County Department of
Environmental Control, 1976.
Vegetable Research Farm. 1976. Summary of 1975 fertility
experiments. William Selleck, Director, Riverhead,
New York.
DISCUSSION
The following questions were answered by Joseph H. Baier
after delivering his talk entitled "The Contribution of
Fertilizer to the Ground Water of Long Island."
Q. by K. Childs. Your presented results suggest that lower
nitrate concentrations in the ground water correspond to
higher application rates^why?
A. A certain period of adjustment appears to be taking place
within the soil mantle before nitrogen equilibrium is reached
and the effects of reduced fertilizer application can be
shown by the ground water. 1976 QW data indicate that
nitrate in ground water under recommended fertilizer plots
is lower than that of grower practice. Finally, we have no
historical data from the demonstration sites that could help
us review the seasonal fluctuations.
Q. by Lawrence C. Eccles. You assume that nitrogen not
used by the crop is lost to the ground water. Did you
evaluate losses by denitrification also ?
A. The Long Island outwash plain sands and gravels are of
such a nature that solubilized nitrates once through the root
zone pass directly to the ground water without change.
The entire nitrogen balance (input and output) is
summarized by Meisinger in his Ph.D. thesis at Cornell
University, "Improving the Efficiency of Fertilizer Use
for Potatoes."
Q. by Charles P. Vanderlyn. How were you able to separate
the source of the nitrogen in the ground water between that
applied as fertilizer versus the ammonia leached from the
proliferation of septic tanks on Long Island? What was the
difference between N readings in surface water and ground
water?
A. The area is rural, and the only septic tank influence
would be the occasional farm house. However, wells were
put upgradient of the farms and sample results have shown
the nitrogen concentration to be extremely low.
In the areas where the farming communities lie, there
are virtually no surface-water bodies which would receive
runoff.
Q. by Mike Apgar. Are the wells located on, or downgradient
from the test plots? How far?
A. There are no wells downgradient of the test plots—only
upgradient. The upgradient wells were immediately
adjacent to the sites.
Q. by Charles Kreitler. What are the different nitrogen
forms predominantly used in the fertilizer in Suffolk
County? Is denitrification or ammonia volatilization
occurring?
A. The main nitrogen form of fertilizer is nitrate. Some,
but very little amounts of urea are used. Even less are the
amounts of inhibitors used.
Denitrification and volatilization may be occurring;
but in the over-all nitrogen budget, they amount to very
small losses. Meisinger in his thesis at Cornell University
("Improving the Efficiency of Fertilizer Use for Potatoes")
has presented an exhaustive nitrogen budget analysis.
Q. by Abe Kreitman. How is rainfall taken into account (or
supplementary irrigation) subsequent to fertilizer application,
and what were the antecedent conditions?
A. Little to no irrigation was experienced on four plots.
This can be said of the Long Island potato farming in
general. Heavy spring rains often play a part in the fertiliza-
91
-------
tion practice of the farmers in that they will go back a
second time and apply more fertilizers if they feel that
spring rains have washed away the initial application.
Q. by Jeffrey Gilman. Can you attribute higher NO$ con-
centrations in your recommended practice to your recom-
mended method of application (side dressing)?
A. Side dressing fertilization allows the fertilizer to be
placed almost directly on top of the planted seed where it
can be used immediately. Side dressing is used by some
farmers, but no study was ever conducted to try to compare
the yields of side dressing versus top dressing.
Q. by P. O. Seman. Did you consider nitrification in your
budget to get the input to ground water? If not, could you
estimate the error?
A. Yes, the nitrification of organic material from plowed-
under plants is a consistent nitrogen source throughout the
year. This has been quantified as approximately 100 Ibs/
acre/year.
Q. by N. W. Johnson. What was the form of the nitrogen
fertilizer applied? Was volatilization considered in your
mass balance?
A. Fertilizer applied was in the nitrate form. Second, the
mass balance did volatilize but it was negligible in comparison
to the quantity of fertilizers being applied and/or lost to
the ground water.
Q. Has anyone done a historical land-use study on Long
Island in relationship to all agricultural practices and
ground-water quality?
A. This has been done in piecemeal fashion, and it is
presently being redone in all agricultural practices. Feedlot
and animal wastes are being quantified under the Federal
208 study.
Q. What assurance do you have that the sampling done in
those years represent the different practices?
A. The areas used in the demonstration plots have all been
farmed for many years before the demonstration plots were
used. By reviewing the amounts of fertilizer applied by
the growers, we have ascertained that these are average
applications.
92
-------
Abatement of Nitrate Pollution in a Public-Supply
Well by Analysis of Hydrologic Characteristics3
by Lawrence A. Eccles, John M. Klein, and William F. Hardr
ABSTRACT
High concentrations of dissolved nitrate in the ground
water in the Redlands, California, vicinity threaten public
water supplies. Dissolved nitrate-nitrogen concentrations in
water from wells, which frequently exceed 10 mg/1
(milligrams per litre), are attributed to the previous applica-
tions of large quantities of commercial nitrogen fertilizer on
citrus crops. In a city of Redlands public water-supply well
field, wells 1 and 2 normally produce water with dissolved
nitrate-nitrogen concentrations of about 18 and 30 mg/1,
respectively. Well 1 is a large-capacity well capable of
yielding 3,700 gal/min (gallons per minute) [233 1/s (litres
per second)] and is the major source of water in the well
field.
The very permeable unconfined alluvial aquifer is
composed of sand, gravel, boulders, and discontinuous
clayey deposits. Well 1 is 742 feet (226 metres) deep and
is perforated throughout most of the zone of saturation. A
major clayey interval from 425 to 480 feet (130 to 146
metres) effectively separates the aquifer into an upper and
lower zone. Well 2 is 426 feet (130 metres) deep and is
perforated throughout most of the upper zone of saturation.
At the well field the static water level (February 1976) was
180 feet (55 metres) below land surface.
Independent tests were made on wells 1 and 2 to
evaluate aquifer characteristics and to determine the
sources of the high-nitrate water. Chemical analyses of
water collected from well 1 during the first 48-hour test
showed an increase in dissolved nitrate-nitrogen from
4.1 mg/1 to a maximum of almost 20 mg/1. Dissolved
nitrate-nitrogen concentrations during the 48-hour test of
well 2 stayed constant at 30 mg/1. Interpretation of the
chemical data and the results of previous studies indicate
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
bHydrologists, U.S. Geological Survey, Laguna
Federal Building, 24000 Avila Road, Laguna Niguel,
California 92677.
that the upper zone of saturation is the higher in concentra-
tion of dissolved nitrate.
To reduce the concentration of dissolved nitrate in
water from well 1, an inflatable packer was placed in the
casing at 480 feet (146 metres) to coincide with the
bottom of the clayey interval. The packer-sealed off the
upper part of the well and, as determined from a final test
of well 1, reduced dissolved nitrate-nitrogen concentrations
from 20 to 4 mg/1 while only reducing well yield from
3,700 to 2,600 gal/min (233 to 164 1/s).
INTRODUCTION
Redlands, California (Figure 1), is in a semiarid
area in which nearly a century of irrigated citrus
culture has contributed to dissolved-nitrate con-
centrations in water from wells frequently exceeding
the recommended nitrate criterion for public
water supplies of 10 mg/1 (milligrams per litre)
nitrate as nitrogen (Environmental Protection
Agency, 1972). The citrus groves (Figure 2) are
gradually being replaced by urban development
resulting in increased use of ground water for
li;DOS' B 2 »
Fig. 1. Index map of Redlands and vicinity.
93
-------
KITRATE-NltflOGEN III WATER FROM
BELLS 111 MILLIGRAMS PER LITRE
-. 9
Fig. 2. Land use and nitrate distribution.
domestic and industrial purposes. The California
Department of Health requires that remedial action
be taken to meet the 10 mg/1 criterion for public
water supplies. In the city of Redlands Texas Street
well field (Figure 3), wells 1,2, and 3 produce
water with dissolved nitrate as nitrogen concentra-
tions of 18, 30, and 26 mg/1, respectively. Well 1 is
the largest capacity well and the major source of
water in the well field.
As part of an ongoing investigation in
cooperation with the San Bernardino Valley
Municipal Water District, the California Department
of Health, and the city of Redlands, the U.S. Geo-
logical Survey is making an appraisal of the hydro-
geologic system to aid in an attempt to reduce the
concentration of dissolved nitrate in well 1. The
objectives of the cooperative investigation
summarized herein were to determine the hydro-
logic characteristics of the aquifer near the well
field in order to suggest certain modifications to
the construction of well 1 that might satisfactorily
reduce the dissolved-nitrate concentration in the
water while minimizing the reduction in well yield.
In this report, nitrate concentrations are
expressed in terms of an equivalent amount of
elemental nitrogen (NO3) = 4.43 (N).
HYDROGEOLOGY
Redlands is in the upper Santa Ana River
basin at the east end of the San Bernardino Valley
near the base of the San Bernardino Mountains
(Figure 1). The valley is filled with alluvium
consisting of sand, gravel, and boulders interspersed
with lenticular deposits of clay and silt. The maxi-
mum depth to bedrock is about 1,000 feet (305
metres) (Burnham and Dutcher, I960; Dutcher and
Garrett, 1963).
The alluvial aquifer within the study area is
considered to be unconfined, although discontinu-
ous intervals of clay can cause short-term
semiconfined conditions. The depth to water below
land surface in the Redlands vicinity ranges from
20 to 300 feet (6.1 to 91 metres). At the Texas
Street well field the static water level was 180 feet
(55 metres) below land surface in February 1976.
The general gradient for both land surface and
ground-water surface is from east to west (Figure
3). Ground water is replenished from natural and
imported flow in the Santa Ana River, artificial
recharge, irrigation return, and direct infiltration
of precipitation.
As evidenced by Figure 2, no clearly defined
pattern for the areal distribution of dissolved nitrate
in ground water exists in the Redlands area. A study
by L. A. Eccles and W. L. Bradford (U.S. Geological
Survey) in 1975 gave some indication of the areal
and vertical distribution of dissolved nitrate so that
a conceptual model could be designed. Their
findings showed that some differences in dissolved-
nitrate concentrations (Figure 3) correlated with
(1) the different deptns and perforated intervals of
the wells, (2) the location of semiconfining clay
intervals, and (3) to some extent the depth of the
pump intake.
Well 1 (Figure 4) was drilled to 742 feet
(226 metres) and is perforated from 240 to 700
feet (73 to 213 metres) below land surface, with
the exception of two zones unperforated opposite
Well 4 Well I Well 2
Well 3
250-
£ 500 —
750-
Land
Waler ?—
'r^
7 ^^-J
7 _» -
? -£?__
Clay
^=?
-T — 7
.— 7
=v
sur f ace
1 a D 1 e T^Z^
1 ? —
IPERFORATEO 7_^ —
/INTFRVll '
J
\
n i 1 13 1 1 oca 1 1 on
ol pump
C— jJ7
-7
_^_f
-0
-50
— 100 "
-200
6B6
956
Fig. 3. Schematic illustration showing concentration of
dissolved nitrate and perforated interval in wells along A-A'.
, FEET
209 ""89'" 291 ^METRES ~'
Fig. 4. Schematic illustration of well field.
94
-------
sandy-clay intervals at 302-360 and 423-490 feet
(92-110 and 129-149 metres). Well 1 contains two
significant zones: the upper zone of saturation
from 240 to 425 feet (73 to 130 metres) and a lower
zone, below the lowermost clay interval from 480
to 700 feet (146 to 213 metres).
Well 2 is 285 feet (87 metres) from well 1
and 426 feet (130 metres) deep, with perforations
throughout the upper saturated zone. The upper
sandy-clay intervals encountered in well 1 were not
reported in well 2, although the bottom of the well
coincides with the top of the lower sandy-clay
interval. Well 3, which is 956 feet (291 metres)
from well 2, was drilled to 500 feet (152 metres)
and perforated from 200 to 482 feet (61 to 147
metres), although the zone from 285 to 355 feet
(87 to 108 metres) was reported by the driller to
contain a significant clayey fraction.
In addition to the public water-supply wells
in the well-field vicinity, several nearby large
irrigation wells (Figures 2 and 3) pump water
containing dissolved-nitrate concentrations of as
much as 25 mg/1 nitrate as nitrogen.
Nearby public-supply well 3 5, approximately
1 mile (1.6 kilometres) northeast of the Texas
Street well field (Figure 3), is pertinent to this
study as it was drilled to 926 feet (282 metres)
and is perforated from 500 to 809 feet (152 to
247 metres), or only in the lower part of the
aquifer. The dissolved nitrate in water from this
well is generally about 4 mg/1 nitrate as nitrogen,
which is probably indicative of nitrate concentra-
tions in the lower part of the aquifer.
A preliminary mass-balance calculation
indicated that the upper zone contributes 60
percent of the water and 90 percent of the dissolved
nitrate while the lower zone contributes 40 percent
of the water and only 10 percent of the dissolved
nitrate in well 1. It was expected that after
modification, well 1 would produce water from
the lower zone containing dissolved nitrate similar
to the concentration found in the water from
well 35.
APPLICATION OF AQUIFER TESTS TO
NITRATE ABATEMENT
An analysis of the hydrologic characteristics
of the alluvial aquifer coupled with the hydraulics
of the public-supply wells for the city of Redlands
is essential in determining the behavioral pattern
and possible means for abatement of ground-water
nitrate pollution. Controlled tests in which wells
are pumped to note aquifer response and changes
in chemistry of the water pumped from the well
are one method to help develop a conceptual model
of the hydrochemical flow system in the aquifer.
Two independent tests (1976) were made in
wells 1 and 2 to evaluate aquifer characteristics,
changes in water quality, and sources of high-
nitrate water. A third and final test was performed
in well 1 after it had been modified, on the basis
of interpretation of the results of the two prior
tests.
The first test consisted of pumping well 1 for
48 hours at a rate of 3,700 gal/min (gallons per
minute) or 233 1/s (litres per second) with a
drawdown of nearly 17 feet (5.2 metres). Specific
capacity was computed to be 220 gal/min per foot
(45 1/s per metre) of drawdown. Water-level
measurements were also made in nearby observa-
tion wells 2, 3, and 4. The pump was set at 360 feet
(110 metres) with all the water entering the pump
column at the end of the suction pipe at 392 feet
(119 metres) below land surface. Chemical analyses
of water collected from well 1 during the test
showed an increase in dissolved nitrate as nitrogen
from 4.1 mg/1 to a maximum of almost 20 mg/1
(Figure 5a).
Because of a 274-foot (84-metre) difference
in the depths of wells 1 and 2 and the effects of
partial penetration in analyzing aquifer trans-
missivity by the Theis and Jacob methods, a
second test was conducted by pumping the
shallower well 2 to better define the upper zone.
This well was pumped for 48 hours at a rate of
2,270 gal/min (143 1/s) with a drawdown of about
13 feet (4 metres). Specific capacity was computed
to be 174 gal/min per foot (36 1/s per metre) of
drawdown. Water-level measurements were made
in nearby observation wells 1 and 3. Dissolved
nitrate as nitrogen in water from well 2 stayed
constant at 30 mg/1 during the test (Figure 5b).
This would indicate that the upper part of the
aquifer as defined by the depth of well 2 is high
in dissolved nitrate.
Results of the two tests from the two wells of
different depths indicate a range in aquifer trans-
missivity that can be largely related to differences
between the zones in the aquifer. Transmissivity
of the upper aquifer zone was computed to be
40,000-50,000 feet squared per day (3,720-4,640
metres squared per day).
Aquifer transmissivity as determined by
pumping well 1 indicated values of 64,000-94,000
feet squared per day (5,950-8,730 metres squared
per day). These values include both the upper and
lower saturated zones. By subtracting transmissivity
values determined by test 2 from test 1, values of
95
-------
TEST I. WELL 1
10 100 1000
(A) TIME. IN MINUTES
TEST 2, WELL 2
10.000 -
2
I 10 100 1000
(B) TIME. IN MINUTES
TEST 3. WELL I
35
30
25
20
15
10
5
0
10.000
NITRATE
15
(C)
10 100 1000
TIME. IN MINUTES
0
10.000
Fig. 5. Drawdown and dissolved nitrate (a) test 1, well 1,
(b) test 2, well 2, (c) dissolved nitrate, test 3, well 1.
artesian conditions, but the long-term effects
indicate water-table conditions.
Interpretation of data from the two tests and
of data from well 3 5 was used to design a third
test to determine if the high-nitrate problem in
well 1 could be alleviated. An inflatable packer was
installed in well 1 at 480 feet (146 metres), at the
base of the lower sandy-clay interval reported in the
driller's log and confirmed by a natural gamma-
radiation log. The pump was reset at 500 feet
(152 metres) with the water intake at 532 feet
(162 metres) below land surface (Figure 6). Well 1
was then pumped at a rate of 2,600 gal/min (164
1/s) for 48 hours. Water-level response was not
measured below the packer. Throughout the test,
the dissolved nitrate as nitrogen concentration
remained at 4 mg/1 (Figure 5c), the same as in the
water from well 35. Thus, as a result of the well
modifications, nitrate concentration decreased
about 80 percent while the well yield decreased
only about 25 percent, in comparison with the
yield from the full aquifer thickness tested in the
first test.
The water level in the well, representing
aquifer head in the upper zone, declined about 2
feet (0.6 metres) during the third test, although the
total water-level decline in the well must have been
much greater, based on water-level measurements
from the previous tests. This minimal decline
indicates the lower sandy-clay interval and the
packer in the well retarded the downward move-
ment of high-nitrate water from the upper zones of
the aquifer. The results of this third test suggest
that permanent modifications to well 1 necessary
to seal off the perforations above 480 feet (146
24,000-54,000 feet squared per day (2,230-5,020
metres squared per day) were computed for the
lower zone. Thus, aquifer transmissivity values
indicate a high-yielding lower zone. These values
compare favorably with data from the nearby
public-supply well 35, which shows a yield of
1,417 gal/min (89 1/s) with 12 feet (3.7 metres) of
drawdown, for a specific capacity of 118 gal/min
per foot (24 1/s per metre) of drawdown. This
translates into an aquifer transmissivity of about
31,500 feet squared per day (2,930 metres squared
.per day).
Data analyses indicate a hydraulic conductivity
of the aquifer, in the vicinity of the pumped wells,
of about 115 feet per day (35 metres per day),
with the lower zone somewhat less permeable than
the upper zones. All test results indicate semi-
o —
250 —
500-
750-
Well
Clay
New
locat i on
of pump
— 0
-50
-100
-150
— 200
-250
Fig. 6. Well 1 as modified for test 3.
96
-------
metres) would be useful to satisfactorily reduce the
dissolved-nitrate concentrations in the water while
maintaining an acceptable well yield.
CONCLUSION
Results of this evaluation and the ensuing
solution for reducing the dissolved nitrate in water
pumped from wells has applicability throughout
the Redlands and other hydrologically similar areas
where there has been heavy nitrogen fertilization
with resultant pollution of ground water used for
a public supply.
Present practices of indiscriminant pumping
from various producing zones in an aquifer and
from wells of different depths and perforated
intervals can result in mixing, throughout a well
bore, native water containing low concentrations
of dissolved nitrate with water containing high
concentrations of dissolved nitrate derived from
surface sources. Modification of wells to limit
public-supply production to the lower parts of the
aquifer, particularly below clay beds, may result in
temporary abatement of the high concentrations
of dissolved nitrate in water pumped from public-
supply wells, but long-term abatement of
dissolved nitrate in ground water would require a
more sophisticated management program. Such a
long-term abatement procedure could be
accomplished by (1) minimizing the downward
movement of water in an aquifer and, thus, avoid-
ing the progressive deeper mixing of the polluted
water with the native water; (2) avoiding the
indiscriminant perforating of wells in all water-
bearing zones penetrated; and (3) pumping several
wells at lower rates rather than a single well or a
few wells at maximum rates.
REFERENCES
Burnham, W. L., and L. C. Dutcher. 1960. Geology and
ground-water hydrology of the Redlands-Beaumont
area, California, with special reference to
ground-water outflow. U.S. Geol. Survey open-file
rept., 352 pp.
Dutcher, L. C., and A. A. Garrett. 1963. Geologic and
hydrologic features of the San Bernardino area,
California, with special reference to underflow across
the San Jacinto fault. U.S. Geol. Survey Water-Supply
Paper 1419, 114pp.
Environmental Protection Agency, Environmental Studies
Board. 1972. Water quality criteria, 1972-A report of
the Committee on Water Quality Criteria. Washington,
Government Printing Office, 594 pp.
DISCUSSION
The following questions were answered by Lawrence A.
Eccles after delivering his talk entitled "Abatement of Nitrate
Pollution in a Public-Supply Well by Analysis of Hydrologic
Characteristics."
Q. Was the preliminary mass balance calculation and the
indirectly calculated transmissivity value for the lower part
of the aquifer confirmed by other methods?
A. The purpose of the preliminary mass balance calculation
for the water produced from well 1 was to determine if
there was significant production from the lower part of the
aquifer. Additional development of the lower zone in well 1
occurred when the pump was relocated and packer installed.
Direct calculation for a transmissivity value for the lower
part of the aquifer was not made because the device
intended to measure drawdown below the packer failed.
Transmissivity for the lower zone of well 1 compares
favorably with the estimated transmissivity for well 35.
Q. Was there vertical leakage and if so what are the effects
of vertical leakage in the aquifer?
A. Results of test 3 indicated vertical leakage during
pumping. Vertical leakage may also occur in unused wells
that are perforated in both upper and lower parts of the
aquifer. The upper "sandy-clay" intervals reported in the
drillers' logs for wells 1 and 3 were probably mostly sand.
That would explain why no "sandy-clay" intervals were
reported in well 2 and why well 3 was perforated in a
sandy-clay interval. The vertical leakage through the upper
part of the aquifer was expected and nitrate data from
test 2 indicated that it had occurred. The change in nitrate
concentrations during test 1 probably represents stabiliza-
tion of the distribution of production between upper and
lower zones and an increase in the rate of leakage induced
by pumping the well. Vertical leakage can be reduced by
pumping several wells at lower rates rather than a single
well or fewer wells at higher rates.
Q. How long will the solution last?
A. The solution could be long-term depending on manage-
ment of withdrawal and recharge in the basin. Irrigation
return is only one of several sources of recharge. Future
prospects are for less irrigation return in comparison to
97
-------
artificial recharge and for incentives to water producers to
withdraw irrigation supplies from the upper part of the
aquifer and domestic supplies from the lower part.
Q. How much did this study cost?
A. This hydrochemical study of a local well field was not a
separately conceived or funded project, but rather a small
part of other regional nitrate pollution and flow digital
model studies. These interrelated studies kept costs at a
minimum. This test presented an opportunity to obtain
badly needed aquifer characteristics of transmissivity and
storage, and time dependent nitrate analyses in the Redlands
area. In addition, it was helpful to the cooperator in
alleviating a high nitrate problem in a public-supply well.
The cost of installing the packer and relocating the pump
was $3,000 and chemical analysis for major constituents in
water about $600. The cost of the pump or aquifer test
including interpretation and administration was estimated to
be about $3,000. Thus, the total cost is estimated to be
about $6,600. Without the previously mentioned regional
interrelated studies as a source of background knowledge,
costs could conceivably have been much higher.
Q. On the final test with the packer, did you monitor water
levels in the observation wells and if so, did the data
confirm "T" computed from previous tests?
A. Yes, the water levels were measured in observation wells
2 and 3 during the final test with the packer installed in
well 1. The water-level declines in well 2 were dampened
by the sandy-clay interval located above the packer and
pump in well 1. Well 3 is perforated to 482 feet and water-
level declines were similar to declines in test 1, prorated to
the difference in well yield between tests 1 and 3. Water
levels were not measured below the packer in the pumped
well. An air line was set below the packer, but it leaked.
Supporting data from nearby well 3 5 (perforated in the
lower part of the aquifer only) indicated the previously
determined transmissivity values were reasonable.
98
-------
Design and Optimization of Ground-Water
Monitoring Networks for Pollution Studies
by H. O. Pfannkuchb and B. A. Labnoc
ABSTRACT
The successful design and operation of a ground-water
monitoring and surveillance system are based on a stepwide
process of obtaining hydrogeologic information.
Because of the inherent uncertainty and inhomoge-
neity of natural hydrogeological systems, the true monitor-
ing network cannot be specified before some basic
knowledge about system configuration and dynamics is
known.
Forethought, planning and incorporation of design
criteria as part of the initial phase of project management
establishes the monitoring network strategy. Optimization
of the monitoring network takes place through the comple-
tion of the following five phases: (1) preliminary network
design and information gathering, (2) initial installation and
testing, (3) completion and verification, (4) operational,
and (5) project termination.
Experience gained from monitoring the conditions
at the University of Minnesota's chemical and special waste
disposal site resulted in the design and optimization
procedure. Concern for possible ground-water contamina-
tion led to analysis of surface and subsurface, physical and
chemical conditions. Subsequently a monitoring system
was established to meet project objectives. No degradation
of the ground water was found during the five-year study.
INTRODUCTION
For any type of ground-water management
operation, design and optimization of a ground-
water monitoring network is vitally important to
the success of the project. Since project objectives
and the general nature of the study area will vary
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
b Associate Professor, Department of Geology and
Geophysics, University of Minnesota, Minneapolis,
Minnesota 55455.
cHead, Envirotechnical Department, Pfeifer and
Shultz/HDR, Inc., 1433 Utica Avenue South, Minneapolis,
Minnesota 55416.
from project to project, forethought, planning and
incorporation of design criteria at the beginning are
essential to limiting additional costs and problems
later.
The successful design and operation of a
ground-water monitoring network are based on a
stepwise process of obtaining hydrogeologic
information. This, in turn, is used as input for a
more rational design of the network to meet
project objectives. The functions of the ground-
water monitoring network are threefold. First, it
needs to establish the basic structure, geometry and
hydrologic characteristics of the flow field. Second,
it needs to establish the dynamic reaction of the
system under natural and stressed conditions. It is
only after these two steps have been determined,
and after a decision has been made of what critical
quantities to monitor, that the third step, the final
network density and configuration, can be defined.
Because of the inherent uncertainty and
inhomogeneity of natural hydrogeological
systems, the true monitoring network cannot be
specified before some basic knowledge about
system configuration and dynamics is known.
Under the usual budgetary constraints, an optimum
design for an observation network would be one
where network points obtained in the first two
steps can be efficiently used in the over-all and final
configuration.
Ground-water pollution studies in general
deal with problems of identifying pollution sources
and containment of spreading from known sources,
or elimination of the stress conditions. The other
class of problems seeks to protect ground-water
resources from pollution or from adverse effects
due to overuse. Due to the nature of hydrogeologic
systems, basic information is usually scarce and
cost for new information is relatively high. The
changing nature of geologic materials and their
99
-------
inherent inhomogeneities raise questions about
network density and representativity of monitor
samples.
These difficulties should underline the
importance of proper and efficient design of
monitor networks, since every sample and every
point have to count by holding maximum informa-
tion content.
The purpose and scope of this paper is to
provide a step-by-step thought process for
developing a monitor network. It defines and
describes five successive phases of the monitoring
process. The information gained through monitor-
ing in one phase serves as basis for the rational
design of the next phase. The general principles are
illustrated and reinforced by an actual example of a
monitoring design for a waste disposal operation,
and experience gained from monitoring the condi-
tions at the University of Minnesota's chemical and
special waste disposal site.
MANAGEMENT CONSIDERATIONS
AND OBJECTIVES
Before developing the principles of monitoring
system design, it is necessary to show how the
information gathering and monitoring process inter-
act with and depend on the over-all management
scheme. The information gathering and the monitor-
ing processes are an integral part of management
because they enter into the decision-making process.
In this context, information has value since
increased information allows more efficient manage-
ment, and it tends to reduce risk (hence loss) when
a choice has to be made between several alternatives
(Pfannkuch, 1975).
Because of this interrelation and interde-
pendence, two sets of goals or objectives can be
defined. One is the set of objectives formulated
for the over-all management plan, while the other
deals with the particular objectives of the monitor-
ing system. It remains clear, however, that the
objectives of the monitoring system design are a
subset of the over-all management objectives.
Four considerations form the basis on which a
rational design for a monitoring system can be
achieved. They are the over-all management
objectives, the particular monitoring system
objectives as mentioned above, and the constraints
imposed by the physical and financial conditions.
Obviously, it would be impractical to formulate
design objectives that are physically impossible to
meet or that would require an unreasonably high
proportion of the allocated project funds. The
interrelation between these factors shown in
MONITORING SYSTEM
OBJECTIVES
Fig. 1. Evolution of monitoring system design.
Figure 1 indicates how these four aspects result in a
monitoring system design strategy.
With the establishment of a design strategy,
an incremental stepwise design and implementation
of successive phases of the information-gathering
process is developed. The findings or outputs of
each phase serve as input and design information
for the following one. All steps are interrelated
by feedback loops, that permit the maximization
of the information base by multiple and inter-unit
correlation and corroboration methods. For this
general discussion five different phases of the
monitoring system design and operation are
identified (Figure 2).
DESIGN STRATEGY FOR
MONITORING NETWORK
1. Preliminary Network Design and Information-
Gathering Phase
During the formulation of objectives and the
development of an operational strategy, it is
necessary to review the physical and chemical
conditions and the dynamics of the system in terms
of the subsurface movement of any possible
pollutants. The pathways and mechanisms of
transport and dispersal that should enter into the
design considerations are the following:
a. Infiltration from a pollution source to the
subsurface.
b. Percolation through the unsaturated zone.
c. Transport and dispersion of disposed
substances within the ground-water flow system.
In order to study the fate of pollutants in the
100
-------
Table 1. Sources of Information
Maps (Areal Information)
1. Topographic
2. Aerial Photographs
3. Geologic
4. Soils
Reports (Comprehensive
Data)
1. Government Publications
2. Journals
3. Theses
4. University Research
5. Environmental Reports
Records (Point Sources)
1. Well Logs
2. Soil Borings
3. Climatologic
4. Stream Data
Miscellaneous
1. Personal Communications
2. Current Research
a. Universities
b. Interest Groups
c. Government Agencies
3. Field Surveys
subsurface, the relative importance of each one of
these zones has to be properly assessed from a fluid
motion point of view, and secondly from the
perspective of possible importance of their
environmental impact. This basic information is
necessary to serve as framework to define and
clarify the objectives, and to formulate feasibility
and strategy considerations for the design and
operation of the monitor system.
After the mechanisms have been considered
and identified in a general way, the hydrogeologic
setting has to be evaluated. Much of the general
geology, meteorology and hydrology and a certain
amount of site specific studies can be carried out
at low-intensity, low-cost levels on the basis of
pre-existing information. It is at this point that
preliminary objectives, choice of methods and tools,
and planning of the subsequent steps is carried
out. An overview of sources of information and
tools for independent information generation are
listed in Tables 1 and 2.
2. Initial Installation and Testing Phase
Here the minimum number of observation
and test wells are installed to give information
about the general ground-water flow field. The
information base is increased by intercorrelation of
project well hydrographs to determine the quality
of observations and the degree of confinement of
the ground-water system. Intercorrelation gives
some further idea about the persistence of the
flow field configuration with time, and therefore
serves as a basis for critical area designation. The
bulk response of the system to precipitation and
infiltration events can be estimated by correlation
of well hydrographs with general climatologic
data. Finally, correlation of project wells will give
some idea about the lateral extent of the flow
field and the possible vertical interconnection with
deeper aquifer systems. For example, preliminary
analyses can be carried out from sampling points
that lie downstream along the principal axis of
flow and dispersion. At this point close, inter-
mediate and distant observation wells are adequate
to assess the approximate extent of any existing
contamination.
From drillers' logs and drill samples, a number
of hydrogeological parameters can be estimated
or determined in relatively inexpensive laboratory
tests. These deal with flow properties such as
permeability, relative permeability of immiscible
phases, and retention capacity as well as adsorption
and ion exchange properties of the soil material for
the downward migration.
3. Completion and Verification Phase
In this phase, the observation well network is
completed on the basis of information obtained
Table 2. System Evaluation Tools and Methods
Direct:
Indirect:
Direct:
Indirect:
Geologic System
1. Drilling
2. Grab Samples
3. Lab Analysis
1. Geophysical (Shallow)
Hydrogeologic System
1. Observation Wells/Piezometers
2. Pumping Wells
3. Suction Lysimeters
4. Laboratory Analysis
5. Aquifer Testing
1. Flow Net Analysis
2. Analog and Digital Models
Hydro logic System
1. Climatology Monitoring
2. Streamflow Monitoring
1. Water Budget Modeling
2. Watershed Modeling
Chemical System
1. Water Samples
2. Analysis of Samples
3. Ion Exchange and Adsorptive
Capacity of Media
1. Isoconcentration Map
2. Geophysical (Density)
101
-------
Fig. 2. Flow diagram of network design process.
during the first two phases in order to meet project
specifications. At the same time intercorrelation
studies of Phase 2 are continued so that the existing
data base can be refined.
The network density is increased (i.e. more
wells) to give an adequate resolution to solve the
problem at hand, such as more precisely defining
the spatial extent of a pollution enclave.
A crude model of the system as formulated
on the basis of information obtained in Phase 2 is
now verified by checking whether the additional
information obtained in Phase 3 corroborates the
model. Furthermore, the quality and accuracy of
the collected data is verified by the continued
correlation of the monitoring network. On the
basis of this information, sampling frequencies'
used in the next two phases can be established.
The principal consideration in the first three
phases is to complete each step (or substep) with
minimum expenditures in order to minimize
redundant information.
4. Operational Phase
During the first three phases, the greatest
frequency and number of tests, sampling operations
and readings occur. These data are then used to
establish the final operation program designed to
give maximum efficiency by selecting only the
most critical points at the lowest frequencies
compatible with project objectives. During this
operational phase, contingency funds must be
available to increase or modify the monitoring
network in case long-term observations change the
initial short-term findings. It is during the
operational Phase 4 that the importance of proper
objective formulation, adequate system and process
identification and most efficient monitoring
network design and operation become obvious.
The total number of observations and analyses are
determined and thereby the cost of operations.
5. Project Termination Phase
An integral project phase that is often
overlooked in the planning stage is the termination
of the project. It is necessary to determine when
sufficient information has been acquired. Too often
neglect of this phase tends to prolong data
collection beyond its managerial or engineering
usefulness.
It is therefore necessary to consider this
phase when objectives are formulated. At this early
point the criteria should be established to deter-
mine when further monitoring is unnecessary.
APPLICATION OF PRINCIPLES
The basis for the preceding monitoring
network development strategy results from over
five years of monitoring and data acquisition of
the surficial and ground-water conditions
surrounding the University of Minnesota's
chemical and special wastes disposal facility. For
well over ten years, chemical and biological
laboratory wastes were collected and transported
to a disposal facility on relatively isolated University
property, 25 miles south of Minneapolis and St.
102
-------
Fig. 3. Waste disposal site.
Paul (Figure 3). With the advent of federal and
state regulations concerning the environment, the
University found it necessary to reassess its disposal
policies and evaluate the potential impact of the
facility on the environment. Of particular concern
was whether any of the waste had reached the
ground water and if so, what was the extent of
degradation.
The study that resulted was designed to
address these problems. The early information-
gathering phases, the incremental drilling and the
maximization of data led to the methodology used
for the design of a ground-water monitoring
network. The advantages and shortcomings of the
data collection and testing procedures used during
the course of the study helped determine the
inadequacies of the original methodology and
led to the refined version discussed here.
THE WASTE DISPOSAL SITE
The disposal process for the chemical and
special wastes involved collecting the wastes from
the University laboratories once a week. The
flammable and nonexplosive wastes were carefully
removed from the truck and dumped into a 15-foot
(4.5-m) deep disposal pit. A flare was then thrown
into the pit from behind a protective barrier to
ignite the-flammable materials. Much of the
flammable chemicals were volatilized by the
process; however, residual material and
nonflammable chemicals remained on the pit
bottom ready for percolation of leachates into the
subsurface.
PROJECT OBJECTIVES AND
PRELIMINARY NETWORK DESIGN
On the basis of the disposal method and
information on the type of wastes, certain project
objectives were outlined. First, and of primary
concern, was whether the residual wastes from the
disposal areas had percolated through the 50-foot
(15-m) thick unsaturated zone to the underlying
drift aquifer. Second, if pollutants had entered the
ground-water system, then determine the extent of
the pollution plume and where it was going. Third,
determine the rate of entrance of the pollutants
into and their dispersal within the ground-water
system. Finally, evaluate the interconnectivity
between the drift aquifer and the more regional
bedrock aquifer and establish if a potential regional
hazard existed.
Given these objectives, the diverse nature of
the wastes and the disposal practices, design of a
ground-water monitoring network had to be
flexible and capable of serving a number of
functions. Specifically, the network had to be
properly aligned to intercept a pollution plume
and allow water-quality sampling. It also had to be
structured to allow hydrogeologic analysis of the
underlying flow system. To properly design the
network, background information about the
geology, and climatic conditions had to be
collected. A list of tests and equipment needed to
meet the project's objectives had to be formulated.
In essence, the preliminary network design and
information-gathering phase had to be completed.
GEOLOGIC SETTING AND
PRIMARY HYDROLOGIC UNITS
The waste disposal site is located near the
edge of the St. Croix moraine on the Rosemount
outwash plain formed during the retreat of the
Superior Lobe of the Wisconsin glacier (Ruhe and
Gould, 1954). Having an areal extent of over 100
square miles, (260 m2), this outwash plain serves
as a drift aquifer of major proportions. In the
vicinity of the waste disposal site, the outwash
plain varies in thickness from 30 to 80 feet
(9 to 24 m) and is composed of interbedded layers
of fine to coarse sands and gravels. The surface is
flat to gently rolling and pockmarked with
depressions (possibly kettle holes). No defined
drainage courses exist on the outwash plain. A
clay layer of variable composition underlies the
outwash plain and rests on the bedrock surface
within the area surrounding the waste disposal
site (Labno, 1973).
The bedrock surface has been dissected by
preglacial streams and then filled with glacial debris
forming an intricate network of buried valleys
throughout the Minneapolis-St. Paul region. One of
the larger buried valleys, cut nearly 400 feet (122
103
-------
NW
1000-
WASTE
DISPOSAL
SITE
SE
BURIED VALLEY
' / '•',', i ',',':'• '• :'.','- jv^o"; "PSEMOUNT ^OUT WASH'•/'/; '^.TVV^^^'-J^'-'^ Lil
JORDAN SANDSTONE
0 .5 I MILE
HORIZONTAL SCALE
Fig. 4. Geologic setting.
m) into the bedrock, lies to the northeast of the
study area and apparently controlled the preglacial
drainage. As a result a series of smaller buried
valleys are incised into the bedrock as tributaries of
the larger valley. The waste disposal site lies on
the flanks of one of these valleys (Figure 4).
The Prairie du Chien dolomite underlies the
waste disposal site and is found at the base of the
local buried valley. St. Peter sandstone overlies the
Prairie du Chien dolomite and is found along the
upper portions of the buried valley walls.
The Prairie du Chien dolomite and underlying
Jordan sandstone are hydrologically connected and
form a major bedrock aquifer in the region. The
regional dip of the bedrock formations is towards
Minneapolis and St. Paul forming a structural basin.
This places the waste disposal site at the higher
elevations of Prairie du Chien-Jordan aquifer and
therefore near one of the aquifer's recharge areas
(Norvitch etal., 1973). The interconnection
between the surficial Rosemount Outwash aquifer
and the underlying Prairie du Chien-Jordan aquifer
becomes important since the contamination of
the drift aquifer could affect the more regional
bedrock aquifer.
DEVELOPMENT OF THE
MONITORING NETWORK
With the knowledge gained from preliminary
investigations of the regional and local geology, a
general idea of the ground-water flow direction in
both the drift and the Prairie du Chien-Jordan
aquifer could be estimated. Three observation wells
were installed and water levels confirming the flow
directions were taken. Plans were then delineated
for the installation of four other wells.
Each of the wells was designed and located to
satisfy certain project objectives. Since the
determination of the extent if any of a pollution
enclave was of primary importance, certain wells
had to be located downgradient from the burning
and burial pits so that water-quality samples could
be taken. To meet this specification, 7 four-inch
(10-cm) diameter wells with two-foot (.61-m)
long suitably sized well screens were drilled and
equipped with hand pumps and water-level
measuring ports. One bedrock well (number 5)
was installed to provide a correlation between the
drift and bedrock aquifer.
In order to facilitate aquifer testing, one of
the drift wells was equipped with a submersible
pump and a continuous water-level recorder. The
location of this well with respect to the others had
to be carefully planned so as to obtain the best
results during an aquifer test. In addition, a U.S.
Weather Service Standard recording rain gauge was
installed adjacent to the well capable of measuring
rainfall intensity and duration. This data plus the
continuous barometric and other meteorological
data recorded at the University's Agricultural
Station, located one mile west, were of importance
to the total ground-water resource evaluation.
Once the layout of the wells and the type of
equipment was established, the remaining four wells
were installed. Each well was developed with the
hand pump, and water levels were noted. After a
few weeks, a monitoring program was determined.
Periodic review of the accumulated data helped
determine the adequacy of the program and the
development of any problems.
During the completion and verification stage,
biweekly water-level measurements were taken and
plotted. Seasonal water-level fluctuations were
noted and all wells, with the exception of two,
fluctuated in a similar manner suggesting a
homogeneity of the flow system. To more closely
monitor the changes of the drift aquifer, three
continuous well records were used to note
instantaneous changes. Shortly after their
installation, it became readily apparent that the
drift aquifer was homogeneous. The instantaneous
hydrographs (Figure 5) matched very closely,
signifying an intercorrelation of project wells and
a unified response to changes in the ground-water
system.
In contrast, wells 3 and 5 fluctuated erratically
Fig. 5. Water-level fluctuations, March 1973: wells 1 and 6.
104
-------
on a daily and weekly basis, but for the most part
followed the established seasonal variations of the
other wells. Examination of the instantaneous
hydrographs demonstrated that erratic fluctuations
were common and that they could be attributed
to barometric fluctuations. The silty sandy matrix
surrounding well number 3 apparently was
sufficient to confine the well, thereby increasing its
barometric efficiency to a point of being quite
similar to the bedrock well number 5. In the latter
well, the 20 feet (6 m) of overlying clay material
was of sufficient size and thickness throughout the
study area to confine the Prairie du Chien-Jordan
aquifer to create a very noticeable barometric
response.
The importance of this intercorrelation of
wells becomes apparent when a flow net analysis
of the system is conducted. The drift aquifer is
considered homogeneous, therefore implying a
constant flow direction. However, the inclusion of
well number 3 water levels causes a change in the
flow net construction and ultimately the flow
direction. This would then suggest a greater
potential spread of pollutants than actually existed.
In addition to the intercorrelation of wells,
tests measuring the aquifer response to meteorologic
events were conducted. Such factors as aquifer
response to precipitation, snow melt and barometric
changes could be noted with the continuous well
recorders. Analysis of the instantaneous well
hydrographs taken throughout the year suggested
that a direct response to a rainfall event could be
noted during periods of recharge while no response
to rainfall was noted during ground-water recession
periods. This, when combined with information
gained on aquifer recharge, indicated when infiltra-
tion was the greatest and possibly when the
greatest amount of leachate might reach the ground
water.
Throughout the testing phase, water samples
were taken monthly and checked for evidence of
a pollution enclave. The results from these
preliminary samples proved to be encouraging. No
signs of a pollution enclave were noted. Water
samples taken during an aquifer test using well
number 6 similarly showed no significant change in
water quality. The water-quality monitoring
program was then geared to these results.
As was previously discussed, long-term
monitoring will sometimes alter initial findings.
The lack of evidence for a pollution enclave raised
the question as to what depth had leaching
occurred. The 50 feet (15 m) of unsaturated sand
and gravel apparently was a sufficient buffer for
BURNING
,-FENCE
I l/m tiiaiti/i
LYSIMETERS
POSSIBLE
EXTENT OF
-LEACHATE' '
r ' c ' •
' FLOW
CLAY LAYER •
/ / / / / / / /
V / V DOLOMITE V V / V
Fig. 6. Lysimeter placement.
attenuation of leachate to the ground water, yet
due to the nature of the porous media, some
movement of pollutants should occur.
To evaluate the situation, soil samples were
taken from various depths below the burning pit
and suction lysimeters installed in the holes.
Chemical analysis of the soil and water samples
demonstrated that the wastes had leached from
the burning pit and decreased in concentration with
depth. Additional lysimeters placed outside of the
pit (Figure 6) at an angle indicated that the
pollutants were fanning out in a bell-shaped
manner. Field and laboratory studies by Saint
(1973) discussed the potential factors affecting
leachate movement and dispersion below the
burning pit and ultimately concluded that no
serious threat to the ground-water system had been
detected.
Monitoring of the quality and the hydro-
geologic conditions of the aquifer continued for a
number of years and eventually was terminated.
No adverse impacts on the ground-water system
were noted.
THE PROJECT IN RETROSPECT
Application of the five phases developed at
the beginning of this paper to the waste disposal
study is easy in retrospect as are those portions of
the study that could have been better planned. For
example, the pumping test and preliminary flow
net analysis were conducted before the inter-
correlation of wells had been completed. Although
the data was not unusable, it did result for a while
in some erroneous conclusions. In another case, it
was nearly two years after the project began that
the well number 1 measuring point elevation was
found to be off by nearly one foot. This gave
anomalous flow direction 70° off the actual flow.
The above are problems that were rectified
and had no effect on the conclusions of the study.
105
-------
However, in other studies, such may not be the
case. The need for proper planning in ground-water
pollution studies should be apparent. Faulty and
unusable data collected from a monitoring system
can do more damage than lost capital. It is the
responsibility of the hydrogeologist to insure that
a ground-water monitoring network is properly
designed and optimized.
POSTSCRIPT
In 1975 the University of Minnesota closed its
waste disposal facility after finding other practical
and safer means of disposal. The residual wastes in
the burning and burial pits will remain and continue
to leach into the environment. However, based on
the research conducted by the various University
departments, a serious hazard is not expected.
Should the wastes eventually leach to the ground
water, the dispersive characteristics of the aquifer
are sufficient to eliminate any major problems. The
remote nature of the site is another plus factor.
The nearest well downgradient is over a mile and
a half away. Finally, the clay layer overlying the
Prairie du Chien dolomite is known to be quite
extensive locally and can prevent direct connection
with the bedrock aquifer.
REFERENCES
Labno, B. A. 1973. The hydrogeological implications of
chemical waste disposal in a glaciated terrain,
Rosemount, Minnesota. M.S.Thesis, University of
Minnesota.
Norvitch, R. F., T. G. Ross, and A. Brietkrietz. 1973.
Water resources outlook for Minneapolis-St. Paul
Metropolitan area, Minnesota. Publ. Metropolitan
Council, St. Paul.
Pfannkuch, H. O. 1975. Study of criteria and models
establishing optimum level of hydrogeologic informa-
tion for ground-water basin management. Water
Resources Research Center, University of Minnesota,
Bull. 81.
Ruhe, R. V., and L. M. Gould. 1954. Glacial geology of the
Dakota County area, Minnesota. Geol. Soc. Am. Bull.
v. 65, pp. 769-792.
Saint, P. K. 1973. Effects of landfill disposal of chemical
wastes on ground-water quality. Ph.D. Thesis,
University of Minnesota.
(Mr. Labno's Discussion Questions were lost.)
106
-------
Tracing Sewage Effluent Recharge
Tucson, Arizona*
by Thomas R. Schultzb, Jeffery H. Randallb,
L. G. Wilson0 and Stanley N. Davisb
ABSTRACT
Dry washes or river beds are often used by south-
western communities to dispose of treated sewage effluent.
Because many of these communities rely on ground water
as a water supply, there is concern that this disposal
practice may contaminate local aquifers. This has led to
implementation of monitoring and tracing programs to
quantify effluent and ground-water interactions and to
development of efficient, easily used predictive models.
The treated sewage effluent from the City of Tucson
treatment plant has historically been used for irrigation
and/or discharged to the normally dry Santa Cruz River.
Numerous sampling programs have been undertaken to
quantify the chemical quality, temperature, and micro-
biological activity of the ground water in the area near the
Santa Cruz. Ground-water regions with high chloride and
nitrate concentrations tend to be associated with areas
irrigated with sewage effluent. Quality degradation due to
channel recharge is not as evident because the effluent
recharge is restricted by fine materials plugging the channel
deposits. Recharging water tends to mound near the
contact between the Recent and Fort Lowell formations
spreading laterally more rapidly than downward.
A new tracer, trichlorofluoromethane (trade name
Freon 11, C13CF) with applications similar to environ-
mental tritium is being evaluated. C13CF enters the hydro-
logic cycle when it is partitioned between the gas and
liquid phases during raindrop formation. C13CF in water
aPresented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
^Graduate Associates in Research, and Head,
respectively, Department of Hydrology and Water
Resources, College of Earth Sciences, University of Arizona,
Tucson, Arizona 85721.
cHydrologist, Water Resources Research Center,
University of Arizona, Tucson, Arizona 85721.
samples is separated and quantitatively measured by a gas
chromatograph with pulsed electron-capture detector.
Preliminary C13CF analyses of ground water along the
Santa Cruz do not correlate with nitrate values because
mixing and increasing atmospheric C13CF concentrations
were not accounted for. However, the presence of C13CF
in the ground water indicates recent recharge. Predictive
modeling will be implemented using C13CF and a finite-state
mixing model.
INTRODUCTION
It is a common practice in southwestern
communities to utilize dry washes or river channels
for the disposition of wastewaters. The City of
Tucson, Arizona has, for example, discharged
secondary treated effluent to the normally dry
Santa Cruz River since 1951. Because Tucson relies
completely on ground water, withdrawing about
five times the natural recharge rate (Arizona Water
Commission, 1975), there is concern that this
effluent disposal practice may lead to ground-water
contamination. Consequently, local water managers
are currently (1) examining monitoring and tracer
techniques to quantify the interactions of effluent
and ground water, and (2) evaluating the role of
efficient and easily-implemented models to predict
such interactions.
The area discussed in this paper is located in
the northwest corner of the Tucson Basin, covering
about 60 square miles (155 square kilometers) and
includes 25 river miles (40 kilometers) of the Santa
Cruz River (Figure 1). At present, this particular
reach of the River receives the entire discharge of
the City of Tucson treatment plant, amounting to
31,000 acre-feet (38 million cubic meters) in 1974
(Dye, 1974). The main downstream ground-water
uses are for irrigation and individual water supplies,
107
-------
Fig. 1. Santa Cruz River study area, northwest corner of
Tucson Basin (inset).
A
3000'-
500
A1
Fig. 2. Diagrammatic cross section of the Santa Cruz River
study area (adapted from Davidson, 1973).
although with future growth of Tucson it is
expected that the major usage will be for
municipal water supply.
The purpose of this paper is to illustrate some
of the traditional monitoring-tracer programs used
in the study area and to describe a potential new
tracer, trichlorofluoromethane (trade name Freon
11), for tracing effluent movement and for use in
conjunction with predictive models.
HYDROGEOLOGY
The Tucson Basin lies within the Basin and
Range Province of the southwestern United States
where downfaulted or rifted blocks, bounded by
steeply-dipping normal faults, form the valleys.
The material filling the intermontane basin
consists of locally derived clastic material or
alluvium deposited since the Oligocene (about
40 million years ago). A secondary process of
erosion has formed three terrace levels in the
Tucson Basin (Davidson, 1973). These are from
oldest to youngest the University, Cemetery, and
the Jaynes terraces.
Davidson (1973) divided the sediments
comprising the aquifer in the study area into
three major units as shown in Figure 2. The
Tinaja Formation, Ts, consists of 200 to 500 feet
(60 to 150 meters) of sandy gravel with 10 to 50
percent silt which unconformably overlies the
Pantano, Tos. The Fort Lowell Formation, Qf,
unconformably overlying the Tinaja, is the most
productive aquifer in the Basin (Davidson, 1973).
In the study area this formation consists of 50 to
150 feet (15 to 45 meters) of loosely packed sands
and gravels, the thinnest zones underlying the
Santa Cruz River channel. The most common
lithology is granite, gneiss, and locally derived
volcanic clasts in a sand and montmorillonite
matrix. Overlying the Fort Lowell along the Santa
Cruz River are the Recent stream-channel deposits,
Qs, comprised of clean, reworked sand and sandy
gravels which grade laterally to a thin veneer of
sheet flow and alluvial fan sediments. The
thickness of the Recent deposits below the River
channel is 40 to 100 feet (12 to 30 meters),
averaging about 80 feet (24 meters).
Because most wells in the area are screened
throughout the entire saturated thickness and
penetrate into the Tinaja, reliable values of
individual formation transmissivities do not exist.
However, Davidson (1973) and others estimated the
transmissivities in the study area to be the following:
2500 to 7000 ft2 day"1 (230 to 650 m2 day"1) for
the Tinaja and Fort Lowell, and 10,000 to 40,000
108
-------
ft2 day"1 (930 to 3800 m2 day'1) for the Recent
channel-alluvium.
The Recent deposits, though very permeable,
are dry in most of the Basin and have only about
10 to 20 feet (3 to 6 meters) of saturated thickness
below the Santa Cruz. Their major hydrologic
significance is as a recharge conduit for flood and
sewage-effluent flows and excess irrigation waters.
The local ground-water flow direction is to the
northwest with a gradient of about 20 feet per mile
(3.8 meters per kilometer) as shown in Figure 3.
The ground-water outflow of the study area is
constricted by the surface and subsurface expression
of the northern end of the Tucson Mountains. The
low permeability of the volcanic rocks forming the
constriction tends to force the ground water closer
to the land surface.
TRADITIONAL MONITORING AND
TRACING TECHNIQUES
Historically, effluent from the City of Tucson
treatment plant has been utilized for irrigation
and/or discharged to the Santa Cruz River.
However, since 1971 the entire flow has been
discharged to the River. Concern over the
possibility of ground-water contamination, as a
by-product of these disposal practices, generated
several monitoring programs. Davis and Stafford
(1966) reported the results of a sampling program
in 1965-1966 characterizing the chemical quality,
temperature, microbiological and virological levels
in 105 wells. Organics were evaluated via the
carbon chloroform extract (CCE) method. Three
regions downstream of the plant manifested the
presence of ABS in well waters—a definite
indication of effluent recharge. Nitrate levels were
also greater than 45 mg/1 in a large number of
samples. Only 10 wells showed the presence of
coliform or fecal streptococci.
A study by Cluff et al. (1972) involved
sampling irrigation wells in the Spring of 1966 and
1970. In 1966, 23 wells were found with nitrate
levels between 4 and 122 mg/1. Individual concen-
trations in the same wells in 1970 ranged between
5 and 51 mg/1. Matlock et al. (1972) reported on a
sampling program using 400 wells in the area. Their
nitrate concentration maps were similar to those
presented by Davis and Stafford (1966), with high
values (up to 45 mg/1) in the region irrigated by
effluent. As with the distribution of ABS observed
by Davis and Stafford (1966), the levels of boron
from LAS type detergents were found to be in
excess of background in the vicinity of irrigated
areas and the Santa Cruz River.
'350 -
I mile
Fig. 3. Computer plotted potentiometric map of the Santa
Cruz River study area in feet above mean sea level.
109
-------
Schmidt (1972, 1973) reported on a
sampling program in which 20 large-capacity
irrigation wells were sampled monthly in the
1971-72 irrigation season. Schmidt found levels
of nitrate in excess of 55 mg/1 in the western
portion of the sampling area. He attributed the
distribution of nitrate in ground waters to three
main factors: a history of effluent irrigation in a
portion of the area; the regional movement of
ground water in the area; and well construction.
In an attempt to characterize quality changes
in effluent during river recharge, Wilson et al.
(1975) obtained water samples from a number of
shallow wells both within the vadose zone (i.e., in
perched water tables) and at the water table
(70 feet, 21 meters). Calcium and magnesium levels
increased in samples during recharge but sulfate
levels decreased. A loss of about 90 percent of the
total-N in river samples occurred during flow
through the vadose zone apparently because of the
combined effects of filtration of organics, ammonia
volatilization, ammonia sorption on clays,
assimilatory reduction, and denitrification.
Virtually complete removal of coliform organisms
was observed.
Kasper (1976) examined samples of well water
near the River to characterize the movement of
organics in the ground-water system. Total organic
content of recharged effluent was measured by the
miniaturized carbon chloroform extract method
(CCE-m). Effluent contained 3.8 mg/1 CCE-m, but
percolation through about 100 feet (30 meters) of
the vadose zone reduced CCE-m values to less than
0.5 mg/1.
Because of the importance of the Santa Cruz
River as a recharge source, studies have been
undertaken to characterize intake rates of effluent
together with quality transformations during
channel flow. Matlock (1966) found an average
channel loss of 6 ac-ft/acre/day (18300 m3/ha/day)
in a 6.3-mile (10-kilometer) reach downstream of
the treatment plant. Later, Sebenik et al. (1972)
measured an average loss of about 7.0 ac-ft/acre/
day (21300 m3/ha/day) in an 11-mile <17.5-
kilometer) downstream reach following a flood
flow.
Sebenik et al. (1972) and Sebenik (1975a)
sampled effluent in an 11.7-mile (19-kilometer)
reach of the River during high and low flow periods
in the Summer of 1971. Particular emphasis was
placed on characterizing nitrogen transformations in
samples. In general, ammonia-N decreased with
distance, with a corresponding increase in nitrate-N
during both low and high flows. Sebenik (1975b)
examined the relationships of dissolved oxygen
(D.O.) and biological oxygen demand (B.O.D.) in
effluent within the channel during high and low
flows of effluent. He found that D.O. levels were
higher at low effluent flows and conversely B.O.D.
values were higher during high flows. He related
these results to waste loadings, flow conditions,
phytoplankton growth and nitrification.
Water quality data (Courtesy Department of
Soils, Water and Engineering, University of Arizona,
Tucson, and Cortaro Water Users Association),
collected during 1971 from 280 wells near the
Santa Cruz River, were mapped by the authors to
indicate the spatial movement of the recharging
sewage effluent. The water quality parameters
chosen were calcium, magnesium, sodium, chloride,
bicarbonate, nitrate, electrical conductivity, and
total dissolved solids. The data were plotted and
contoured using an advanced computer graphics
package at the University of Arizona. The resulting
maps compared favorably with those of Matlock
et al. (1972). The aerial distribution of chloride
and nitrate, chosen to be representative, are shown
in Figures 4 and 5.
Shallow water quality patterns in the area
strongly reflect local hydrogeological controls. In
particular, the channel deposits have a hydraulic
conductivity about one order of magnitude greater
than the underlying Fort Lowell and Tinaja
formations. This causes recharging water to mound
near the interface with the Fort Lowell and to
spread laterally more rapidly than downward.
The spreading is reflected in the ground-water
quality of the selected wells (Table 1) plotted in
the Piper diagram (Figure 6). The shallow wells,
which only penetrate the upper zones of the Fort
Lowell (SF1A1, RRCY1, and IRA21), have a water
type similar to the effluent water in the Santa Cruz
(SCRR1). Deeper wells deriving most of their water
from the Fort Lowell and Tinaja have a similar
water type, but different from the shallow wells,
manifesting long-term vertical mixing of effluent
and indigenous ground water.
The areas of high chloride and nitrate con-
centrations, showntin Figures 4 and 5, tend to
correspond to areas that have been irrigated with
sewage effluent. Availability of an inexpensive,
abundant supply led to over irrigation with two
to three feet of effluent per year in excess of
consumptive use requirements (Schmidt, 1973).
Because the channel deposits underlying the
agricultural areas are highly permeable, over-
irrigation caused a large surface area of the aquifer
to receive effluent recharge. The problem was
110
-------
Fig. 4. Computer plotted isograms ot nitrate in mg/l for
the Santa Cruz River study area.
Fig. 5. Computer plotted isograms of chloride in mg/l for
the Santa Cruz River study area.
Ill
-------
Table 1. Representative Water Quality Analyses of Wells in the Santa Cruz River Study Area
Location
(Tn. Rn.)
Section
(13-12)2add
(13-13)5bba
(13-13)6cdd
(13-13)7cdd
(13-13)27cdd
(13-13)28cd
(13-13)34dbb
(13-13)17cdd
(SCRR1)
(13-12)lbdd
(IRPW1)
(13-12)lbdd
(IRA21)
(13-13)17cda
(RRCY1)
(13-13)17cca
(SF1A1)
(13-13)28bdc
(SF2A1)
Depth
feet/
meters
300/91
294/90
-
250/76
-
-
-
0
280/82
100/30
100/30
100/30
100/30
Ca
mg/l
83
57
119
60
124
115
82
51
125
41
51
48
125
Mg
mg/l
16
12
22
17
30
16
14
8
20
13
6
9
20
Na
mg/l
108
24
92
163
127
183
144
142
120
128
122
116
186
K
mg/l
-
-
-
13
13
-
-
0
6
13
14
15
—
HC03
mg/l
270
144
293
307
305
264
225
266
381
259
288
251
371
C03
mg/l
0
0
0
0
0
0
0
0
0
0
0
0
0
SO4
mg/l
185
49
298
130
257
310
240
149
25
140
84
123
252
a
mg/l
54
24
91
84
109
145
98
104
105
94
111
85
190
N03
mg/l
18
4
-
40
23
-
5
3
4
1
3
1
2
compounded by cascading water within irrigation
wells.
In contrast to the irrigated areas, effluent
recharge from the River is not as evident from
quality data. River seepage is controlled primarily
by the hydraulic conductivity of the river bed.
Several researchers (e.g., Sebenik et al., 1972) have
IRPKl
(13-13)Sbh
CA CL
Fig. 6. Piper diagram for selected wells in the Santa Cruz
River study area.
reported that intake rates are restricted through the
channel bottom during low flows because of
plugging by fine inorganic and organic sediments.
Occasional flooding scours these deposits
permitting substantial recharge until clogging
again retards intake rates.
The chloride ion distribution, a traditional
tracing method, is further complicated due to the
high natural background in the study area. The
two anomalous areas to the west of the River have
been correlated to lacustrian mud and siltstone
deposits which are frequently associated with
evaporites in the Basin and Range Province.
FREON TRACING APPLICATIONS
A new tracer, trichlorofluoromethane (C13CF),
has been evaluated for use in hydrological studies
(Thompson et al., 1974). Applications of this tracer
are similar to tracing applications of environmental
tritium. Trichlorofluoromethane is a commercially
produced compound manufactured since the
1930's, primarily by E. I. duPont de Nemours and
Company under the trade name of Freon 11.
C13CF has been widely used as an aerosol propellant,
working fluid in refrigeration systems, blowing
agent in plastic foams, and as an industrial solvent.
This compound and others like it have received
much attention recently due to their postulated
112
-------
role in the stratospheric destruction of ozone.
Tropospheric introduction of C13CF is a
function of its intended usage. The residence time
between manufacture and use of aerosol propellants
is short. Rowland and Molina (1975) estimate that
90 percent of the propellants are released within
6 months. The residence time for blowing agents
and solvents has been estimated to be near that
of aerosol propellants. Tropospheric introduction
of C13CF is not uniform, being greatest around
urban areas in industrial nations (Hester et al.,
1974). However, global tropospheric mixing is
rapid (on the order of months) and uniform. The
distribution of artificially produced radioactive
krypton-85 as a C13CF analog (Telegradas and
Ferber, 1975) shows relatively uniform mixing
within both hemispheres with only a 20 percent
reduction in concentration in the southern
hemisphere.
The introduction of C13CF into the hydrologic
cycle occurs when it is partitioned between the
gas and liquid phases during precipitation. There
are two possible mechanisms of incorporation in
precipitation, rain-out and wash-out. Rain-out
involves solution of the gas during drop formation
and wash-out involves solution as the raindrop
falls. Based on other trace gases (SOX and NOX) in
precipitation, rain-out is predominant (Moyers,
1976). If rain-out is the predominant mechanism,
then the C13CF concentration in precipitation will
be in equilibrium with the global atmospheric
distribution rather than near surface anomalies in
urban areas. The concentration of C13CF in
recharging water is therefore proportional to the
tropospheric concentration at that location and
time.
C13CF in water samples is separated and
quantitatively measured by a gas chromatograph
with pulsed electron-capture detector. The
chromatographic system developed at the
University of Arizona is a modification of an
Indiana University design (Thompson
1975). The system has several features distinguish-
ing it from the analytical quantification of
carbon-14 or tritium: it is fast and easily
operated; allows sample concentration; and most
importantly, is field operable.
The major demonstrated advantages of C13CF
as an environmental or artificially injected tracer
can be summarized as follows: no quality degrada-
tion (a problem with large concentrations of
chloride tracers); extremely low toxicity, thereby
minimizing health hazards; low sorptivity on clays
due to its low surface energy (a problem with
organic dyes); quick, economical field-operable
detection with equipment less expensive than for
most other tracers; a detection limit of 10~13 grams
per second; a steady known build-up in the atmos-
phere (atmospheric variability is a major problem
with tritium); and no loss through decay as with
radioactive isotopes.
To develop a predictive tool more quantitative
than the contouring described earlier requires a
computer model. Mass flow of water can now be
handled by numerous computer modeling
algorithms (e.g., Prickett and Lonnquist, 1971).
Modeling chemical mass flow is more difficult and
is still in its infancy. A finite-state mixing cell
model (FSM) has recently been developed at the
University of Arizona (Campana, 1975) which
accounts for chemical and water mass-flows
without requiring assumptions about dispersion
coefficients. The FSM has been calibrated and
verified for the Tucson Basin using carbon-14 as a
tracer (Campana, 1976) and for the Edwards
Limestone in Texas using tritium. A more detailed
FSM model is being calibrated for the study area
shown in Figure 1 using the collected historical
chemical data. Once calibrated, C13CF data will be
used in the FSM to demonstrate the tracer's
usefulness as an age dating tool in sewage effluent
tracing (Randall and Schultz, 1976).
C13CF concentrations, from the preliminary
analyses shown in Table 2, were mapped and
correlated with the corresponding nitrate values.
The resultant map and correlation analysis
indicate that there is no apparent correlation
between the two. This is not surprising because
mixing has been neglected and more importantly,
the historically increasing C13CF concentration in
the recharging water has not been accounted for.
However, because C13CF is an anthropogenic
(man-made) compound, its very presence in the
ground water near the Santa Cruz River indicates
that recharge has taken place since the 1930's.
This kind of qualitative information can be useful
in delineation of regional or local recharge areas
where other methods have proved impractical.
Furthermore, it is expected that C13CF when
utilized in a finite-state model will provide
quantitative information suitable for predictive
modeling.
ACKNOWLEDGMENTS
The work upon which this paper is based was
supported in part by funds provided by the United
States Department of the Interior, Office of Water
and Research and Technology, and E.I. duPont de
113
-------
Table 2. Trichlorofluoromethane (CI3CF) Concentrations
in Southern Arizona
Location
(Tn.Rn.)
Section
(12-12)16bbb
(12-12)16bbb (duplicate)
(12-12)17ddd
(12-12)21bab
(12-12)21bdd
(12-12)25cac
(12-12)26dba
(12-12)26dbd (First)
(12-12)26dbd (Second)
(12-12)36bab (First)
(12-12)36bab (Second)
(13-13)35adb
(A-35)
(13-13)16ddd
(Z-2)
(13-13)16bca
(Z-5)
(12-13)31ccc
(Z-13)
(12-13)31dcd
(Z-14)
St. David, Arizona
Santa Cruz River
©Silverbell Rd.
Tucson
Sample
Type
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Well
Surface
Water
Air(1976)
Concentration
(picograms/ml)
0.28
0.13
6.0*
0.62
12.*
0.19
0.13
5.6*
14.*
3.9*
0.36
3.4*
2.8*
18.*
17.*
2.1*
<0.02
0.08
0.70
* Possible sample contamination.
Nemours and Company. We gratefully acknowledge
the assistance of Glenn Thompson of the
Department of Geology, Indiana University, in
making the preliminary C13CF analyses.
REFERENCES
Arizona Water Commission. 1975. Phase I, Arizona state
water plan, inventory of resource and uses, Phoenix,
Arizona.
Campana, M. 1975. Finite-state models of transport
phenomena in hydrologic systems. Unpublished
doctoral dissertation, Dept. of Hydro, and Water
Res., Univ. of Ariz., Tucson.
Campana, M. 1976. Application of carbon-14 ground water
ages in calibrating a flow model of the Tucson Basin
aquifer, Arizona. Hydro, and Water Res. in Ariz.
and the Southwest, v. 6.
Cluff, C. B., K. T. DeCook, and W. G. Matlock. 1972.
Technical economic and legal aspects involved in the
exchange of sewage effluent for irrigation water for
municipal use, case study—city of Tucson. Tech.
Completion Report, Dept. of Interior, Off. of
Water Res. and Tech., project A-022-Ariz.
Davidson, E. S. 1973. Geohydrology and water resources
of the Tucson Basin, Arizona. U.S. Geol. Survey Water-
Supply Paper 1939-E.
Davis, G. E. and J. F. Stafford. 1966. First annual report,
Tucson wastewater reclamation project. Planning and
Research Section, Water and Sewers Dept., City of
Tucson, Ariz.
Dye, E. O. 1974. Annual report, fiscal year 1973-74. Dept.
of Water and Sewers, Wastewater Div., City of
Tucson, Ariz.
Hester, N. E., E. R. Stephens, and O. O. Taylor. 1974.
Fluorocarbons in the Los Angeles Basin. Jour. Air
Pollution Control Assoc. v. 24, no. 6.
Kasper, D. R. 1976. Verbal communication. Univ. of Ariz.,
Dept. of Civil Eng. and Eng. Mech.
Matlock, W. G. 1966. Sewage effluent recharge in an
ephemeral channel. Water and Sewage Works, v. 113,
no. 6, pp. 224-229-
Matlock, W. G., P. R. Davis, and R. L. Roth. 1972. Sewage
effluent pollution of a groundwater aquifer. 1972
Winter meeting, Amer. Soc. Agri. Eng. paper 72-709.
Moyers, J. L. 1976. Verbal communication. Univ. of Ariz.,
Analytical Center.
Prickett, T. A., and C. G. Lonnquist. 1971. Selected digital
computer techniques for groundwater resource
evaluation. 111. State Water Survey, Bull. 55.
Randall, J. H. and T. R. Schultz. 1976. Chlorofluoro-
carbons as hydrologic tracers, a new technology.
Hydro, and Water Res. in Ariz, and the Southwest, v. 6.
Rowland, F. S. and M. J. Molina. 1975. Chlorofluoro-
methanes in the environment. Rev. of Geophys. and
Space Phys. v. 13, no. 1.
Schmidt, K. D. 1972. Groundwater contamination in the
Cortaro area, Pima County, Arizona. Hydro, and
Water Res. in Ariz, and the Southwest, v. 2.
Schmidt, K. D. 1973. Groundwater quality in the Cortaro
area northwest of Tucson, Arizona. Water Res. Bull.
v. 9, no. 3.
Sebenik, P. G., C. B. Cluff and K. J. DeCook. 1972.
Nitrogen species transformations of sewage effluent
releases in a desert stream channel. Hydro, and Water
Res. in Ariz, and the Southwest, v. 2.
Sebenik, P. G. 1975a. Physicochemical transformations of
sewage effluent releases in an ephemeral stream
channel. Unpublished master's thesis, Dept. of Hydro.
and Water Res., Univ. of Ariz., Tucson, Ariz.
Sebenik, P. G. 1975b. Relationships of dissolved
oxygen and biochemical oxygen demand in sewage
effluent releases. Unpublished Master's thesis, Dept.
of Civil Eng. and Eng. Mech., Univ. of Ariz., Tucson,
Ariz.
Telegradas, K. and G. J. Ferber. 1975. Atmospheric
concentrations and inventory of krypton-85 in 1973.
Science, v. 190.
Thompson, G. M. 1975. Verbal communication. Indiana
Univ., Dept. of Geology.
Thompson, G. M., J. M. Hayes, and S. N. Davis. 1974.
Fluorocarbon tracers in hydrology. Geophys. Res.
Letters, v. 1, no. 4.
Wilson, L. G., R. A. Herbert and C. R. Ramsey. 1975.
Transformations in quality of recharging effluent
in the Santa Cruz River. Hydro, and Water Res. in
Ariz, and the Southwest, v. 5.
114
-------
DISCUSSION
The following questions were answered by Stanley N.
Davis after delivering his talk entitled 'Tracing Sewage
Effluent Recharge — Tucson, Arizona."
Q. If there is no ion, normally analyzed for, to pose a
problem in the usability of the ground water, why would
you want to stir up the public by using a tracer to prove
that the water contains sewage?
A. This question is a very good general question, related to
social and political justifications of ground-water studies.
As a purely scientific answer, if general knowledge can be
gained by hydrochemical studies, this is justification
enough. In the case of the Santa Cruz River north of
Tucson, it is common public knowledge that effluent is
entering the subsurface. Numerous studies have been
published and are also available to the public. Our interest
is in testing new techniques in an area which is already
quite well known. We do not anticipate that our studies
will add to public anxiety.
Q. What are typical freon concentrations in
domestic sewage effluents?
A. Insufficient measurements have been made to date. No
typical values can be given. The next-to-the-last analysis
given in Table 2 is of effluent that has entered the Santa
Cruz River. The value given is lower than would be
expected if it were in equilibrium with the atmosphere.
Another analysis made at the City of Tucson sewage plant
in April of 1976 also showed that the effluent was under-
saturated with respect to atmospheric Freon-11.
Q. Do the Tucson area waste treatment plants chlorinate
their effluent prior to discharge into the receiving stream?
If so, could you also look for chlorinated organics
present in the underlying ground^water body?
A. The Pima County Ina Road plant which discharges
approximately 10 mgpd does not chlorinate its effluent.
Effluent from the City of Tucson facility receives some
chlorine. Chlorination has been sporadic in the past. At
present, no one is analyzing for chlorinated organics in
ground water; however, it would be feasible using
analytical procedures similar to ours that incorporate the
electron-capture detection system.
Q. Many papers state that "a few feet of aerated zone
thickness beneath a disposal system will remove large
quantities of waste constituents indefinitely." Does your
study and/or experience support or refute this frequent
conclusion?
A. The present study does not contribute information to
answer this question. However, past studies cited in our
paper do give some information. Some waste
constituents such as chloride would not be removed; others
such as very small traces of certain heavy metals might be
removed for periods of several hundred years. Of course,
biological and suspended solids can be mechanically
filtered by underlying sediments.
Q. You mentioned that in arid environments, potassium
tends to travel great distances. Why is this so?
A. The possible use of potassium was mentioned in the oral
presentation but is not discussed in the written paper.
Potassium levels in the Tucson effluent most commonly
range from 15 to 20 mg/1, or somewhat higher than the
native ground water. The native ground water in arid
regions, however, is also high in potassium, commonly from
5 to 15 mg/1. This suggests, in turn, that many adsorption
sites are already occupied by potassium ions and that the
distribution coefficient would be relatively low with the
transport correspondingly more efficient.
Q. How much time and what is the expense to run a single
fluorocarbon analysis?
A. Once the sample has been collected in the proper
container, the process of stripping the Freon-11 and
passing it through the gas chromatograph takes less than
10 minutes. On-site duplicate samples should be run to
check for consistency of data. A typical field visit to a site,
including analyses, could take from 20 minutes to an hour
provided the water is readily sampled and all components of
the instrument are functioning. This is based on the
experience of Glenn Thompson of the Department of
Geology, Indiana University, who is presently making
routine analysis of surface-water samples. He can be
contacted for further details.
The cost of field analyses has not been estimated;
however, it should be limited to operator time and
transportation. There are very minimal expendable costs.
Based on our experience at the University of Arizona, the
cost of the electronics package and the chromatographic
and detector system, including parts and labor, is on the
order of $2500. A strip chart recorder is also convenient
to supply a backup record and visual aid.
115
-------
Monitoring Cyclic Fluctuations
o
in Ground-Water Quality
by Wayne A. Pettyjohn
ABSTRACT
The chemical quality of water in many shallow and
surficial aquifers exhibits cyclic fluctuations. These
fluctuations are caused by the intermittent flushing of
contaminants into the ground during recharge events. The
contaminants may be natural or reflect man's activities,
particularly waste disposal schemes.
Over the past 12 years an oil-field brine contaminated
aquifer in central Ohio has been monitored. Data from
three closely-spaced wells tapping selected parts of the
aquifer indicate that brine is flushed into the ground during
recharge events and that each contaminated mass maintains
much of its integrity as it sinks to the bottom of the
aquifer and then migrates laterally to the adjacent river.
The most concentrated mass that covers the largest area
infiltrates during the spring recharge period, but less
concentrated and smaller masses may occur any time
rainfall is sufficient to overcome the soil-moisture
deficiency.
Because of the cyclic nature of recontamination
events, care and common sense must be exercised in the
extrapolation of quality data, particularly in regard to
estimation of contaminant flushing rates.
INTRODUCTION
Within the past 10 years water pollution
investigations, reports, and legal actions have far
exceeded those of previous decades. In the
literature, ground-water pollution case histories
have become commonplace where before 1965
they were rare. Despite our having become much
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
^Department of Geology and Mineralogy, The Ohio
State University, 125 South Oval Drive, Columbus, Ohio
43210.
more aware of ground-water pollution, however, a
great deal still remains to be learned and
documented. A particular need is the development
of adequate ground-water quality monitoring
techniques that can provide us with information
on the temporal and areal migration of contaminants
so that we can develop accurate predictive models.
Samples of well water obtained for quality
analyses may be collected only once, now and then,
or perhaps on a regular schedule. The literature
indicates that even during the monitoring of
contaminated aquifers, samples may be collected
only during a set time period or until the concentra-
tion of the contaminant begins to decrease.
Furthermore, such important data as depth to
water, well depth, and construction details
commonly are not recorded.
We may not realize the pollution potential or
that which is occurring in an area merely on the
basis of a one-time collection period. We must
consider the contaminant both in time and space.
The lack of many of the basic details
mentioned above does not permit an investigator
to accurately monitor a contaminated aquifer,
determine when it may be flushed, or establish
legal responsibilities. In most instances, we
complain of a lack of information and stringent
time constraints on data collection, monitoring,
and perhaps well construction. Only in rare
circumstances will these problems be overcome.
Generally, we better perceive what should have
been done only long after the project has been
completed. Certainly there is no substitute for
sound planning.
The purpose of this Symposium is to
examine various methods that have been used to
clean up ground-water pollution. Determination
116
-------
of the effectiveness of these techniques, to a large
extent, depends on adequate monitoring, which
must be accomplished in such a manner that the
distribution of contaminants can be detected and
predicted in both space and time.
It is generally believed that the quality of
ground water changes very slowly, if at all. This
may be true in the majority of cases, at least when
we deal with deep aquifers, those not contaminated,
or those not greatly stressed by pumping or
construction. On the other hand, many surficial or
shallow aquifers are prone to rapid changes in
quality, the effect being measureable within hours
or days. Because rapid and perhaps repeated
changes in quality can occur, we must be prepared
to predict these events and understand the
principles involved so that we can deal with them
from legal, technical and practical viewpoints.
The chemical quality of water in many shallow
aquifers exhibits cyclic fluctuations that have
seldom been noticed, rarely reported, and apparently
never described in detail. The purpose of this report
is to shed some light on this phenomenon and to
describe the vertical, horizontal, and temporal
migration of contaminants introduced into a
shallow aquifer by certain waste-disposal methods,
agricultural techniques, and natural phenomena so
that we can develop adequate ground-water quality
monitoring techniques. A few examples to illustrate
how some traditional methods can lead to erroneous
conclusions will also be presented.
The major cause of cyclic quality fluctuations
is intermittent infiltration of water-soluble
substances that lie mainly on the land surface or in
the ground above the water table. These substances
include waste materials, such as brine and chemicals,
nitrogenous materials of natural or man-made origin,
and mining wastes, to mention only a few of an
almost limitless number.
Soluble materials that infiltrate during periods
of recharge retain much of their integrity as the
contaminated mass migrates to points of discharge,
such as wells, streams, springs, seeps, and drainage
tile. Following another period of recharge, a
second mass of contaminated water descends and
moves laterally. Several distinct masses may exist in
an aquifer at any instant in time, extending down a
flow line and thus resembling a string of beads or a
dripping faucet. A contaminated mass maintains a
relatively uniform shape as it migrates through an
aquifer because of laminar flow conditions and
perhaps differences in specific gravity, temperature
and viscosity between it and the native ground
water.
PREVIOUS WORK
Although commonly alluded to, cyclic
fluctuations of ground-water quality are rarely
described in the literature. In a number of reports,
water-quality graphs actually indicate these
fluctuations, but their significance is not described.
Pettyjohn (197la) reported that nitrate concentra-
tions in a surficial aquifer in North Dakota
increased substantially following periods of
precipitation. He assumed that the nitrate was
leached from decaying organic matter in the scores
of small sloughs and potholes that dot the area. In a
later report describing cyclic chloride fluctuations
in a streamside aquifer contaminated by oil-field
brine, Pettyjohn (1971b) suggested that salt is
washed downward into the aquifer during each
period of natural recharge. Pettyjohn (1975) also
described chloride fluctuations in Alum Creek in
central Ohio, a stream contaminated by the
discharge of brine-rich ground water.
Walker (1973a, b) pointed out that nitrate in
the soil is transported to surficial aquifers during
late fall and early spring recharge periods. He also
recognized that these highly mineralized masses of
water migrate through aquifers as bulb-like slugs
with relatively little mixing or dilution.
Toler and Pollock (1974) described the
accumulation of deicing salt in the unsaturated
zone along a major highway in northeastern
Massachusetts. They pointed out that some of the
salt was flushed deeper into the ground during the
spring recharge period, reaching the water table
during midsummer. This is clearly another example
of cyclic change in ground-water quality.
Nightingale and Bianchi (1973) discussed the
Leaky Acres ground-water recharge system at
Fresno, California. Their graphs show rapid increases
in specific conductance as well as nitrate and
chloride concentrations at the beginning of each
recharge period. Schmidt (in press) discussed
water-quality variations in pumping wells and
presented a rather thorough review of the
literature. The reader is referred to Schmidt's
report for additional information.
CYCLIC FLUCTUATIONS
This report describes the cyclic fluctuation of
chloride in a streamside alluvial aquifer in a humid
environment. Chloride-rich wastes were both
spilled on the ground and placed in pits for
disposal. Similar water pollution situations might
occur due to waste disposal by land spreading,
burial, use of disposal or holding ponds and
lagoons, and perhaps even septic tanks, among
117
-------
Fig. 1. Water-level surface and monitoring wells at the
contaminated site.
several others. The example described herein may
not be applicable to arid regions because of the
commonly great thickness of the unsaturated zone
and the limited amount of precipitation, soil
moisture and ground-water recharge.
The major part of this report is based on data
collected intermittently over the last 12 years,
largely by The Ohio State University students,
from an aquifer in central Ohio contaminated by
oil-field brine (Shaw, 1966; Boster, 1967; Hulman,
1969; Lehr, 1969; Pettyjohn, 1971b, 1975;
Kerzner, 1973). Similar phenomena, however, have
been observed in many other areas throughout the
country, particularly in humid and semiarid regions.
The contaminated site 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. The land surface rises to
the north (Figure 1). The alluvial deposits,
consisting of sand, gravel, silt and clay, range from
15 to about 35 feet in thickness. The underlying
bedrock is shale.
Since 1965, 23 test holes have been drilled in
this area, ranging from 5 to 25 feet in depth. 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.
The site became contaminated during
1964-65 following the excavation of three brine
holding ponds. One pit (Skiles) was used from
July 1, 1964 to June 30, 1965; about 126,000
barrels of brine were placed in it. Two other pits
(Slatzer number 1), also used at about the
same time, received 110,000 barrels of brine. The
impoundments, used only about a year, received
at least 236,000 barrels of brine containing about
35,000 mg/1 (milligrams per liter) of chloride.
The brine was spilled on the ground, but
largely stored in the holding ponds. Fluids leaked
from the ponds and soon contaminated the ground
water. When samples were first collected from
monitoring wells in July 1965, the aquifer, at least
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. During 1965-66
chloride concentrations in several wells, although
variable with time, were in excess of 15,000 mg/1.
By 1969 chloride concentrations in most of these
wells were less than 2,000 mg/1, in 1972 were less
than 1,000, and during a single sampling period in
1975, concentrations were less than 400 mg/1.
An interesting relationship becomes apparent
when examining the areal extent of ground-water
contamination with time. For example, the
position of the 1,000-mg/l isochlor as it existed
during selected months of 1965 and 1966, and the
500-mg/l isochlor during 1969, are shown in
Figure 2. Note that the enclosed area during
1965-66 changes monthly but that the changes,
even those from one month to the next, do not
necessarily encompass smaller areas. This suggests
Fig. 2. Areal extent of the contaminated area enclosed by
the 1,000 mg/l (1965, 1966) and the 500 mg/1 (1969)
isochlors during selected months. Contours based on data
from monitoring wells shown in Figure 1.
118
-------
=r-Land surface
^—:.W_ater ieyej4
Gravel packx'-'.
creen/.
Fig. 3. Completion details of three closely-spaced monitoring
wells.
that a linear relationship between chloride
concentration and time (flushing rate) does not
exist. For example, the area in September 1965 is
larger than the January 1966 area indicating a
natural flushing effect on the reservoir during the
four-month period. But when the size of the
January 1966 area is compared with the area that
existed in February, and the March and April 1966
areas are compared, it is apparent that the
contaminated zone has increased in size.
Significantly, both February and April were wetter
than previous months.
An examination of maps for 1969 shows a
similar phenomenon (Figure 2). Monthly decreases
in the size of the contaminated area are readily
explainable as resulting from less rainfall and
natural cleansing of the aquifer. It can be seen then,
that in this area, both the size of the contaminated
area and the chloride concentrations are functions
of precipitation.
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
and receives water from the entire aquifer (Figure 3).
The location of these three wells are shown in
Figure 1 at the position marked A. It is assumed that
the first two wells represent the quality that exists
at depths of 9 feet and 23 feet, respectively, and
that the well gravel-packed throughout the entire
thickness of the aquifer provides a composite
sample of the reservoir. It is also assumed that when
the composite well has a higher concentration than
both the deeper and shallower wells, the most highly
JASOKDJFMAKOJASOKD
1965 1966
Fig. 4. Fluctuation of chloride content in closely-spaced
wells of different depths during 1965,1966 and 1969.
mineralized water lies between 9 and 23 feet.
The chloride fluctuations in the three wells
during 1965, 1966, and 1969 are shown in Figure 4.
Notice that at certain times the highest concentra-
tions occur at the shallowest depth, at other times
at the greatest depth, and at still others the
greatest concentration must lie somewhere in
the middle of the aquifer. The only means for
accounting for this phenomenon is intermittent
reintroduction of the contaminant, which is
puzzling in view of the fact that oil-field activities
ceased before any of the samples were collected.
Another technique illustrating the temporal-
vertical distribution of chloride in the aquifer is
shown in Figures 5 and 6. These illustrations are
based on monthly data obtained from the three
adjacent wells. Concentrations at depths of 9 and
23 feet were measured. Interpretation 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 a shallow
Fig. 5. Vertical distribution of chloride at the contaminated
site from October 1965 to April 1966.
119
-------
Fig. 6. Vertical distribution of chloride at the contaminated
site during 1969.
depth, but during December 1965 through April
1966 the highest concentrations were near the
bottom of the aquifer (Figure 5). Furthermore,
in November and December, water in the middle
of the aquifer was less mineralized than that
above or below.
Although greatly reduced, in January 1969 the
largest concentration of chloride was at the
shallowest depth, but the situation was reversed
during April and May (Figure 6). 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 concentra-
tions again appeared in the central part of the
aquifer.
The chloride fluctuations that occurred during
1965-66 and 1969 are shown schematically in
Figure 7. The October 1965 samples apparently
were collected 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 mass (1) had descended and was
migrating along the bottom of the aquifer, when
another recharge event occurred (2). By December,
mass (2) had reached the bottom of the aquifer and
was moving toward the river. Recharge events also
occurred in January 1966 (3) and in February (4).
The February event, however, represented the
spring runoff when evapotranspiration was
minimum 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 grossly contaminated throughout
a wide area. This mass eventually discharged into
the river.
In spite of the fact that brine disposal ceased
in mid-1965, the aquifer was recontaminated
several times during 1969 (Figure 7). Following an
established pattern, small recharge events took
place in January, February, and March 1969.
During the spring recharge period in April, there
was another massive influx of brine as salt was
flushed from the soil. This mass slowly diminished
in size and concentration as it flowed through the
aquifer towards and into the river. By August
chloride concentrations, although increasing with
depth, were far below previous months.
The water-quality data for 1965-66 and 1969
show that (1) the chloride concentrations
decreased substantially over the four-year period
but, (2) following each recharge event the aquifer
was recontaminated. Similar cyclic events have
been recorded from 1970 to 1975.
This study indicates that water-soluble
substances that lie on the land surface or at a
shallow depth 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 soil and on the
October 1965
November 1965
January 1969
February 1969
Fig. 7. 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.
120
-------
frequency of recharge events, which are controlled
by evapotranspiration, by the rate, duration, and
intensity of precipitation and by soil-moisture
conditions.
Throughout most of the year in humid and
semi-arid 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,
great quantities of contaminants 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.
When considering the possibility of cyclic
fluctuations of ground-water quality, several points
other than those alluded to here, must be
considered:
(1) Even if there is a constant source of
contamination, such as a septic tank, leaking
transmission line, etc., more highly mineralized
and cyclic masses could be generated through
flushing during recharge events.
(2) In arid or semiarid regions where there is
a thick unsaturated zone, the contaminated mass
may never, or only rarely, reach the water table.
(3) In areas of high potential evapotranspira-
tion, large concentrations of salts may lie near land
surface. Since all of the available moisture is used
by vegetation or evaporated, cyclic events would
not be likely to occur. If, however, a rain of
sutficient intensity and duration occurred, some of
these salts could be washed downward.
(4) Even during steady-state conditions, such
as induced infiltration from a contaminated river,
the quality of the water may show cyclic fluctua-
tions due to recharge between the river and the
well or increased head in a holding pond due to the
inflow of runoff, rain, or more waste.
(5) One must also consider aquifer stratigraphy,
paying particular attention to horizontal and
vertical permeability. Confining beds could cause a
perched mass to migrate laterally, eventually to
discharge through springs, seeps, or drainage tile.
From the discharge point, it might again infiltrate
and continue downgradient as an isolated mass. For
example, Walker (197 3a) described a case where
mink farm wastes containing nitrate and bacteria
flowed overland on glacial till more than a quarter
mile before entering a creviced dolomite aquifer.
(6) Most commonly, cyclic fluctuations occur
where water-soluble materials lie on the ground or
in the ground above the water table and where
the system is affected by direct infiltration. Cyclic
effects may be masked where the contaminant must
flow through confining beds.
(7) Movement of the contaminant in the
saturated zone will be influenced by the difference
in density between the contaminant and the
receiving water. If the density of the contaminated
mass is greater than the receiving water, and if the
stratigraphy permits, the contaminated mass will
move vertically downward and then horizontally.
(8) Wastes are not likely to infiltrate when the
ground is frozen or when evapotranspiration is at
its maximum. In a temperate or colder climate the
greatest influx should occur during the nongrowing
season, particularly in the fall and spring.
(9) Any soluble material may be transported
deeper into the ground; enough rainfall events or
prolonged rain may cause it to reach the water
table.
CAVEAT EMPTOR
Oftentimes two or three water samples,
collected over a period of a few years, are used to
predict a flushing rate. If samples are collected at
just the right time, we might assume a predictable
rate, but we could be wrong. For example, an
interesting relationship can be observed if we plot
the chloride concentration of samples collected
from a single well during the same month but in
different years. Water from a well in the Olentangy
River area discussed in the preceding pages had a
chloride content of 35,750 mg/1 in August 1965,
12,750 in August 1966 and only 380 mg/1 in
August 1969 (Figure 8). A straight line neatly fits
all three points when plotted on semilog paper. If
we assume the flushing trend will continue, the
concentration should return to background (about
15 mg/1) by mid-1972. We might also plot chloride
concentrations as they existed in the same well in
April 1966 and 1969. Since it is easy to fit a
straight line between only two points, the graph
readily shows us that the area should be flushed
clear by early 1973. Sampling and analysis showed,
however, that the chloride content in April 1972
was not 15 mg/1 but 1,900 mg/1; a recharge event
had occurred just before the sample was
collected.
Although seemingly ridiculous, we might use
the same logic and work backward (both literally
and figuratively) and fit a line through the con-
centrations of April 1969 (890 mg/1) and April
121
-------
1 .000
Fig. 8. Graph for predicting flushing rates.
1972 (1900 mg/1). In this case the graph indicates
that the chloride concentration in the reservoir was
at background in 1952, which it most certainly
was, and that by 1987 the aquifer would contain
a pure unadulterated brine, in spite of the fact that
brine disposal ceased entirely in 1965!
A hypothetical cross section of an aquifer as
it might have existed after two recharge events is
shown in Figure 9a, along with 10 observation
wells of different depths. A plan view of the area
using only the "A" wells to develop isochlors is
shown in Figure 9b. A similar map using only the
"B" wells appears in Figure 9c, while Figure 9d
shows isochlors determined from all the wells.
Actual conditions, however, are shown in Figure
9e. If the monitoring wells were of different
depths and closely spaced, the actual situation
would have been evident.
The previous two illustrations show that we
can make significant errors by using traditional
methods. Only by employing knowledge of the
aquifer system, determination of recharge events,
and closely-spaced monitoring wells of different
depths, coupled with common sense, can we
adequately understand the movement of con-
taminants in ground water and thus design
programs for monitoring ground-water quality in
contaminated areas.
SUMMARY
Existing data indicate that in many situations,
cyclic fluctuations of ground-water quality can
occur and in fact may be common. These fluctua-
tions are greatly influenced by the characteristics
of the wastes, recharge events, and aquifer
stratigraphy. Cyclic events can best be monitored
by using a series of closely-spaced wells, each of
which is screened opposite a small part of the
aquifer and withdraws water from only that
limited section. Moreover, samples should be
collected from these wells at closely-spaced,
regular intervals until the hydrologic nature of the
site is recognized. Furthermore, we must not
<700<700 850 700 1030
ABA B A
112 2 3
750 1025^700^700 800
B ABBA
3 4455
Wells
700
"A" and "B" Wells
700
Fig. 9. Cross section (A) of an aquifer containing two
contaminated masses and plan view interpretations of
isochlors based on data from "A" wells (B), "B" wells (C),
"A" and "B" wells (D), and actual conditions (E).
122
-------
blithely pass over or ignore quality data that appear
to be anomalous for they may tell us far more than
the expected analysis.
REFERENCES CITED
Boster, R. S. 1967. A study of ground-water contamination
due to oil-field brines in Morrow and Delaware
Counties, Ohio, with emphasis on detection utilizing
electrical resistivity techniques. Unpubl. M.Sc. thesis,
The Ohio State Univ., 191 pp.
Hulman, B. D. 1969. A ground-water contamination study
in the Delaware area, Ohio. Unpubl. senior thesis,
The Ohio State Univ., 34 pp.
Kerzner, Stuart. 1973. A study of chloride contamination
in Delaware County, Ohio. Unpubl. senior thesis,
The Ohio State Univ., 24 pp.
Lehr, J. H. 1969. A study of ground-water contamination
due to saline water disposal in the Morrow County
oil fields. Ohio State University Water Resources
Center, 81 pp.
Nightingale, H. I. and W. C. Bianchi. 1973. Ground-water
recharge for urban use: Leaky Acres project. Ground
Water, v. 11, no. 6, pp. 36-43.
Pettyjohn, Wayne A. 1971a. Water resources of Renville
and Ward counties. North Dakota State Water
Comm., County Ground-Water Studies 11, Part III,
100pp.
Pettyjohn, Wayne A. 1971b. Water pollution by oil-field
brines and related industrial wastes in Ohio. Ohio
Jour. Sci. v. 71, no. 5, pp. 257-269.
Pettyjohn, Wayne A. 1975. Chloride contamination in Alum
Creek, central Ohio. Ground Water, v. 13, no. 4,
pp. 332-339.
Shaw, J. E. 1966. An investigation of ground-water contami-
nation by oil-field brine disposal in Morrow and
Delaware counties, Ohio. Unpubl. M.Sc. thesis, The
Ohio State Univ., 127 pp.
Schmidt, K. D. (in press). Water quality variations for
pumping wells. Ground Water.
Toler, L. G. and S. J. Pollock. 1974. Retention of chloride
in the unsaturated zone. Jour. Research, U.S. Geol.
Survey, v. 2, no. 1, pp. 119-123.
Walker, W. H. 1973a. Ground-water nitrate pollution in
rural areas. Ground Water, v. 11, no. 5, pp. 19-22.
Walker, W. H. 1973b. Where have all the toxic chemicals
gone? Ground Water, v. 11, no. 2, pp. 11-20.
DISCUSSION
The following questions were answered by Wayne A.
Pettyjohn after delivering his talk entitled "Monitoring
Cyclic Fluctuations in Ground-Water Quality."
Q. Isn't there a possibility that you have more than one
source of chloride and that migration occurs under
conditions of differential head under anisotropic
conditions?
A. Water from wells tapping till or alluvium in adjacent
areas contains less than 15 mg/1 of chloride. Likewise,
chloride concentrations in the underlying shale and
limestone are less than 25 mg/1. There are no septic tanks or
any other waste disposal sites nearby except for the
abandoned brine ponds. Road salting is not carried out on
the country lane that crosses the site. Furthermore, the
head in the alluvium is higher than that in the underlying
shale or limestone. Therefore, brine disposal during
1964-65 provided the only source of chloride in the
studied site.
Q. Does the rate of flushing in your example have any
correlation with the rate of ground-water movement or is
it strictly related to recharge rates or other factors?
A. The rate of flushing of this system depends largely on
recharge events. Once the contaminants reach the aquifer,
they move fairly quickly. If the system were not con-
tinually recontaminated, the aquifer would probably have
been flushed clear of chloride years ago.
Q. How long will it take for the brines to be completely
flushed out of (I) the pit, (2) the zone of aeration, and
(3) the aquifer?
A. Until the brine is completely flushed from the zone of
aeration, the aquifer will be intermittently recontami-
nated. The rate of flushing of the unsaturated zone is
related to the frequency and magnitude of recharge,
which, in turn, depends on the rate, duration, and
intensity of rainfall, the season, soil-moisture conditions,
grain size, and land use, among others. It is evident then
that it is impossible to predict when the area will be
flushed clear of the contaminant.
Q. Was there a direct correlation between rainfall and
re contamination?
A. In general, there is a close correlation between rainfall
and reintroduction of the contaminant. However, the
amount of rain required to trigger an event depends on
the antecedent soil-moisture conditions. For example,
during the growing season more rain is required than during
the early spring or fall. A rain that might cause a mass to
form in the fall might have no effect during winter when
the soil is frozen.
Q. Are the rates of movement, both vertical and horizontal,
reasonable to allow the slugs to move as postulated
considering vertical and horizontal permeabilities and heads?
A. Analysis of quality data permitted a rough estimate of
123
-------
the rates of contaminant movement. Vertical migration
appears to occur at an average rate of about .5 feet per
day, while horizontal movement towards the Olentangy
River is about 7 feet per day. Therefore, a mass would
reach the bottom of the aquifer in 20 to 30 days and
move from there to the river in about 25 days.
Q. If changes are due to recharge events, what is the best
method to determine the background level?
A. Background concentrations could be determined when
evapotranspiration rates are the highest, when the soil is
frozen, or several days after a recharge event. The time of
frame, of course, is dependent on the vertical permeability
and thickness of the unsaturated zone and the aquifer. One
should not attempt to determine background concentra-
tions in contaminated areas, but should limit sample collec-
tion to sites some distance upgradient or on the opposite
side of a drain or divide.
Q. Why aren 't the high concentrations of chloride initially
flushed out? What mechanism caused the salt accumulation?
A. At least 236,000 barrels of brine were disposed of at the
site. As a result, the unsaturated zone, which consists
largely of silt and clay, serves as a huge sink for sodium
chloride, particularly in the vicinity of the pits. The
alluvium contains an abundance of near vertical openings,
such as dessication cracks, and animal and root borings,
many of which are quite large. It is postulated that salty
soil moisture diffuses from the surrounding contaminated
but less permeable material into these larger openings,
remaining there either as a salt crust (following evaporation)
or as salty water, until it is flushed deeper into the ground
during a recharge event.
In the unsaturated zone large pores drain quickly,
while small pores drain much more slowly. Perhaps 90
percent of the water moves through 2 percent of the void
space. However, gravity drainage of salty water from the
small pores slowly moves downward, forming a highly
mineralized zone near the water table. Much of this
solution also is flushed into the aquifer during recharge.
The zone is then slowly remineralized by descending
gravity drainage from the small pores while the larger
openings are reloaded by diffusion. When this stage is
reached, a recharge event will cause another slug to form.
If the alluvium had consisted of coarse material,
such as sand or gravel, the salt would probably have been
flushed from the system fairly rapidly. Very likely the
flushing rate could be increased by irrigation.
124
-------
The Lycoming County, Pennsylvania,
Sanitary Landfill: State-of-the-Art
in Ground-Water Protectiona
by M. Todd Giddings, Jr.
ABSTRACT
A 120-acre sanitary landfill site is being developed in
north-central Pennsylvania to serve a six-county area. The
design of the site will prevent ground-water contamination
by leachate from the refuse through the use of a 20-mil
(.5 mm) PVC plastic liner which will lie between two
protective layers of sand containing drainpipe networks.
Sampling of the composite ground-water flow from
the underdrainpipe network in the sand blanket under the
liner will provide sensitive monitoring of the liner
performance. In the unlikely event of a leak, the pipe
draining that area would be diverted to the leachate treat-
ment facility, thus providing a backup leachate collection
system. A thick (15 to 75 feet, 4.6 to 22.9 metres)
glacial till deposit is present at the site and confines
ground-water flow within the underlying shale bedrock.
The low permeability of the till and the artesian head
within the bedrock flow system provide additional
protection against ground-water contamination.
Operation of the site will be by the area-fill method;
the refuse will be deposited in 8-foot (2.4-metre) lifts up to
a maximum height of 120 feet (36.6 metres). Based on an
initial refuse deposit rate of 400 tons (363 metric tons)
per day, the site is expected to have a 20-year life at a
disposal cost of approximately $5.00 per ton.
Local residents who felt this project posed a severe
threat to their wells strenuously opposed the project and
appealed the State permit. The Pennsylvania Environmental
Hearing Board upheld the permit; the decision was appealed
to Commonwealth Court and was upheld. Residents
remain unconvinced the design and site conditions will
provide adequate protection and have petitioned the
Pennsylvania Supreme Court to consider another appeal.
A baseline water-quality monitoring program has been
undertaken, to establish on-site conditions and a private-well
sampling program will begin before the landfill is in
operation. Till and bedrock monitoring wells at the site
will be used to evaluate the performance of this landfill;
analysis of the nearby private wells will be provided to the
owners to demonstrate confidence in the design and the
satisfactory operation of the ground-water protection
measures.
aPresented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
^Principal, Todd Giddings and Associates, Consulting
Hydrogeologists, 140 West Fairmount Avenue, State
College, Pennsylvania 16801.
INTRODUCTION
This 120-acre (48.6-hectare) landfill site is
located on the Federal Allenwood Prison Camp
property, approximately ten miles south of
Williamsport, Pennsylvania. The Lycoming County
Commissioners are developing this site to serve
Lycoming County and the adjacent counties of
Clinton, Columbia, Montour, Union, and part of
Northumberland. All of these counties are located
in the north-central part of the State where it has
been necessary to develop a multi-county landfill
operation in order for the sparsely populated areas
to reach the minimum necessary solid waste
volumes to make an environmentally-sound solid
waste disposal site economically feasible. The site
will serve an area of approximately 3000 square
miles (7770 square kilometers) with a total
population approaching 300,000 persons and, thus,
will allow for the future development of a regional
resource-recovery program.
The design of this site will prevent ground-
water contamination by leachate from the refuse
through the utilization of a tough, plastic membrane
liner which will isolate the landfill from the ground-
water flow systems. The leachate flow will be
controlled and collected by the liner, subjected to
biological and chemical treatment as required,
and then will be recirculated back through the
landfill to accelerate the stabilization of the refuse
mass. The importance of this new design in
managing the leachate problem is underscored by
the Federal Appalachian Regional Commission's
1.2-million-dollar grant for construction of this
sanitary landfill facility; this project will be
administered by the U.S. Environmental Protection
Agency (EPA).
HYDROGEOLOGIC FRAMEWORK
The site is located on the south slope of a hill
within the Allenwood Prison Camp property in
Brady Township, Lycoming County, and lies
immediately west of U.S. Route 15. Topography at
125
-------
SECTION I
SCALED
0 250 500 FEET
SECTION 2
0 250 500 FEET
PROPOSED SANITARY LANDFILL
LOCATION PLAN
IOOO 2000 FEET
Fig. 1. Location plans and profiles.
the site is very gentle; near the upper or north edge
of the site, slopes are approximately 13 percent
but over much of the site the slopes are 5 percent
or less (Figure 1). There is approximately 140 feet
(42.7 metres) of relief on the site between the north
and south boundaries (near the top and bottom of
the hill). Within the site, there are several small
gullies which drain the slope area and are up to
several feet deep, carrying surface-water flow during
heavy rainfall or snow-melt periods.
A detailed reconnaissance study of the site
was conducted to map all pertinent surface features
and to lay out the test drilling program. The site
was flown to obtain vertical aerial photographs,
and a photogrammetric topographic map was
produced at a scale of 1 inch (2.54 centimeters)
equals 100 feet (30.5 metres) and at a contour
interval of 2 feet (.6 metre). This topographic map
and aerial photographs taken at several different
times were used to map features during the field
study; this map was also used to present hydro-
geologic data obtained during the subsurface
exploration program.
A drilling program was conducted to determine
the characteristics of the glacial till and the bedrock
beneath the site. Eleven test borings were drilled
through the compact, Illinoian glacial till, and
cores were taken in the upper 5 feet (1.5 metres)
of the underlying bedrock. Split-spoon samples
were taken at 5-foot (1.5-metre) intervals within
the till, and water levels were measured in all holes
during and after drilling. The glacial till ranges in
thickness from 15 to 75 feet (4.6 to 22.9 metres),
with the area of thickest till located at the base of
the hill along the south edge of the site. The non-
sorted and non-stratified till material contains a
wide range of sediment sizes including clay, silt,
sand, pebbles, cobbles, and boulders, and is
composed primarily of silt and clay material with
scattered cobbles and boulders throughout. A
detailed investigation of the soil series which
have developed on this till was not undertaken
because the construction of this site will involve
excavation of all of the topsoil and several feet of
subsoil in order to prepare a surface upon which
to install the membrane liner.
The glacial till lies on bedrock which is
dominantly a dark-gray shale with minor inter-
bedded limestone and calcareous shale beds, with
no evidence of solution openings. Eight of the
11 cores drilled at the bottom of each test boring
encountered only dark-gray-brown shale. The other
3 holes encountered shale with some limestone
interbeds. Strata beneath the site are quite flat,
dipping only 10 degrees to the north.
Four ground-water quality monitoring wells,
126
-------
each 150 feet (45.7 metres) deep, were drilled
around the perimeter of the site. All of these wells
penetrated at least 100 feet (30.5 metres) of
bedrock composed almost entirely of dark-gray
shale. Cuttings from these air-rotary test holes
indicated that some calcareous material was present
in Wells 2, 3, and 4 (Figure 2); but because the
bedrock is so nearly horizontal throughout the
site area, and because the spacing of these wells
around the perimeter of the site provides good
control on the continuity and thickness of the
predominantly shale bedrock beneath the
glacial till cover, it was concluded that the
limestone and calcareous shale material were
present as discontinuous interbeds within the
predominantly shale bedrock.
Bedrock crops out near the site, along the
north edge where there is an abandoned shale
quarry exposing flat-lying, interbedded sandstone
and shale strata. No local folds or faults were
observed in road cuts or shale quarries adjacent
to the site.
A refraction seismograph was used to supple-
ment the soil boring program by running traverses
where the drill holes were furthest apart on the site.
These data, in addition to the bedrock elevations
from the test boring program, indicate that the
bedrock surface beneath the proposed site has
moderate relief apparently due to scouring and
plucking by the Pleistocene-age ice sheet which
overrode the hill immediately to the north. A
bedrock-surface contour map was constructed from
the test boring, water well, and seismographic
survey data.
Drainage gullies are well developed throughout
the site area and carry storm runoff and the
discharges from several small springs within the
glacial till. The watershed located above the north
edge of the site is only a few acres in area and,
thus, there is only a small amount of overland
flow which will have to be diverted around the site.
Soils throughout the site area contain a fairly
consistent fragipan layer which causes shallow
perching of water within the soil zone on this
denser, clay-rich strata. Thus, there is a shallow
perched ground-water flow system developed within
the till, and lateral flow to small springs is common
throughout the site.
Ground water within the shale bedrock
beneath the till is confined by the overlying till
blanket; all 4 of the air rotary test holes encountered
ground water under artesian conditions, and the
water levels in these wells all rose to within a few
feet of the ground surface. The artesian heads
ranged from approximately 9 feet (2.7 metres) at
the north (uphill) edge of the site to more than
60 feet (18.3 metres) at the south (downhill) edge
of the site. Water was first encountered in the
bedrock at depths of 80 feet (24.4 metres) or
more beneath the ground surface. The yields of the
two wells along the north edge of the site, which
are to be background water-quality monitoring
points, were approximately 1 and 3 gallons per
minute (.06 and .19 liters per second); the yields
of the 2 downgradient monitoring wells along the
southern edge of the site were approximately 20
and 30 gallons per minute (1.26 and 1.89 liters
per second). Each of the 4 monitoring wells was
cased and grouted at least 10 feet (3.1 metres) into
solid bedrock so that no shallow ground-water flow
contributed to the water levels or water-quality
samples. Initially, 3-hour pumping tests were
conducted on each of the monitoring wells to
develop their yield and to obtain base-line ground-
water quality samples. The ground-water quality
samples were analyzed for a selected list of chemical
parameters designed to establish base-line conditions
with regard to those chemical parameters generally
indicative of leachate contamination.
DESIGN FEATURES
Development of the site will involve the
sequential placement of the 20-mil (.5. -mm) PVC
liner on 10 sections of the site (Figure 2). These
sections, or "fields," will allow for the staged
development of the site in order to match the
current need for refuse disposal capacity. The first
field to be excavated will provide a gently sloping
floor which will then slope more steeply up to the
terraced level of the second field to be developed.
Eventually, the third and fourth fields will be
developed along the west edge of the site to
complete development of the entire western
section of the landfill. A similar sequential
development of terraced fields will allow develop-
ment of the central section of the site, and finally
the eastern section of the site will be filled.
The landfill will be operated by the area-fill
method with the refuse being emplaced in 8-foot
(2.4-metre) lifts and covered daily. The soil
material excavated during preparation of each field
will be stockpiled for use as daily, intermediate,
and final cover. The maximum thickness of the
refuse will be 120 feet (36.6 metres) where the
gently sloping plateau reaches its maximum height
above the slope of the hill. Thus, the configuration
of the finished landfill will be a bench or a terrace
on the side of the existing hill (see Figure 1).
127
-------
Fig. 2. Site design plan.
-------
During construction of each of the fields,
appropriate sedimentation and erosion control
measures will be taken to control runoff from the
site. As new fields are developed, previously used
areas will receive final grading and seeding. The
configuration and slopes of the finished landfill
will be graded and seeded to provide pasture land
for grazing, which is the present use of the landfill
site.
The landfill is designed to receive municipal
solid waste; chemical or hazardous industrial
wastes will not be accepted at the site. The refuse
will be spread and compacted by a landfill
compactor machine in order to obtain an in-place
refuse density of approximately 1000 pounds per
cubic yard (600 kilograms per cubic metre).
The design of the ground-water protection
measures includes several positive features not
previously used in a full-scale landfill application.
The leachate control system is a continuous
multi-layer barrier, which separates the undisturbed
glacial till and ground water from the refuse and
leachate. From the bottom to the top, the barrier
consists of the following layers: 12 inches (30.5
centimeters) of highly permeable sand (emplaced
on the undisturbed till), a 20-mil (.5 millimeter)
thick continuous sheet of impermeable PVC
plastic, 6 inches (15.24 centimeters) of sand, and
12 inches (30.5 centimeters) of select clay soil
(glacial till) (see Figure 3). Sets of perforated,
horizontal drainpipes will be installed in gravel
envelopes within the 6-inch (15.24-centimeter)
sand layer overlying the plastic liner in order to
carry leachate to the treatment and recirculation
facility. There will be another set of horizontal
drainpipes, also located in gravel envelopes, within
the 12-inch (30.5-centimeter) sand layer beneath
the liner, to provide an exit for ground-water
seepage occurring beneath the liner. Much of the
site is a ground-water discharge area for the shallow
ground-water flow system located within the
glacial till. Periodic water-level measurements made
in plastic pipes installed in the 11 test borings
located throughout the site indicate that through-
out much of the site shallow water-table levels are
within a few feet of the ground surface during all
seasons of the year. This shallow ground-water flow
system will be encountered during excavation and
construction of each field, and the seepage from
this flow system will drain from beneath the liner
through the ground-water underdrain network (see
Figure 2).
This ground-water underdrain network will
provide a sensitive monitoring system to detect a
leak in the liner. Should leachate be detected in the
ground-water discharge from beneath the liner,
the drainpipe set containing the contaminated water
would be located by the sampling of each major,
trunk drainpipe. The drainpipe set containing the
contaminated water would be isolated, and that
water would be conducted to the leachate treatment
facility. Thus, this technique would provide a
backup leachate collection system which would
recapture the contaminated water should a leak
occur. The possibility of a leak occurring is
considered to be extremely unlikely, due to the
protection of the liner by the sand layers, and the
select soil cover. Also, the first lift (8 ft. or 2.4 m.
thick) of refuse emplaced on the liner will not
contain any coarse rubble or construction debris.
Daily monitoring of the conductivity of the
composite ground-water underdrain discharge will
be utilized as an early warning monitoring system
to evaluate the performance of the liner.
In addition to the engineered ground-water
protection measures discussed above, there are also
additional, naturally-occurring ground-water
protection features at this site. The site is located
within a shallow ground-water discharge zone
within the glacial till. This flow system will oppose
2
Fig. 3. Design environmental protection measures.
129
-------
GEOLOGIC SECTION
Fig. 4. Natural environmental protection features.
any leachate trying to enter the till beneath the site.
There will also exist a permeability contrast at the
interface between the 12-inch (30.5-centimeter)
sand underdrain blanket and the undisturbed glacial
till. The contrast in permeability between this sand
blanket and the undisturbed till will be approximate-
ly one million times; that is, the permeability of
the sand will average one million times greater
than permeability of the till.
Ground water within the shale bedrock under-
lying the till is under artesian pressure, with heads
ranging from 10 feet (3.1 metres) to more than
60 feet (18.3 metres). This artesian pressure would
prevent leachate from entering the bedrock flow
system, should the leachate ever manage to enter
the confining layer of glacial till above the bedrock.
Finally, should leachate somehow enter the shale
bedrock flow system and be detected in the down-
gradient monitoring wells, the low average
permeability of this bedrock unit would allow for
on-site extraction of the contaminated water
before there is any degradation of the ground water
in the region (see Figure 4).
Thus, there are 4 components of the multi-
layer barrier which are designed to separate the
leachate from the ground water, and there are also
4 components to the naturally-occurring environ-
mental protection system. The primary leachate
collection system is designed with a backup
collection system to recapture leachate within the
ground-water underdrains. Further, several natural
hydrogeologic features of the site are taken
advantage of to provide additional protection of
the ground-water resources of the region.
Finally, an early warning underdrain monitoring
program and a quarterly monitoring well
sampling program will be conducted to demonstrate
the satisfactory performance of this leachate
control system.
CITIZENS' OPPOSITION
Local residents who felt that the proposed
landfill site posed a severe threat to their domestic
water-supply wells opposed the investigation and
development of this site from the earliest phases of
A thick mantle of compact glacial
till overlying bedrock.
The site is a groundwater discharge
area for water moving in the
glacial till.
3) A low permeability bedrock.
The groundwater in the bedrock is
under artesian pressure.
the project. Meetings were held with citizens'
groups, and the site investigation procedure,
landfill design requirements, and ground-water
monitoring system were explained in detail. Most
of the citizens remained unconvinced that there
were adequate safeguards either within the
design of the site or within the hydrogeologic
features underlying the site, and the citizens' group
retained a hydrogeologic consultant to assist them
in evaluating the ground-water pollution potential
of this project. Reasoning that the citizens might
be fearful of the project out of ignorance of the
safeguards, they were invited to inspect all phases
of the hydrogeologic investigation of the site and
were provided copies of all preliminary and final
site investigation reports and landfill design plans.
During one of their tours of the site, the
citizens discovered a shallow depression which they
said was a sinkhole in the early stages of formation.
Cavernous limestone strata are present in the area
adjacent to the site, and unusual ground-water flow
conditions and sinkholes are prevalent in local
knowledge and folklore. A detailed evaluation of
this depression was undertaken; the backhoe
excavation revealed parallel buried stone walls on
the margin of the depression and pottery artifacts
at an old floor elevation within the depression.
Thus, it was concluded that the depression was an
old root cellar or building foundation associated
with a farmstead present on the site before it was
acquired for federal use.
After a thorough review of the hydrogeologic
conditions at the site and after evaluation of the
proposed design and development plan for the
landfill operation, and with consideration of the
questions and concerns raised by the citizens'
group, the Pennsylvania Department of Environ-
mental Resources issued a solid waste disposal
permit for this landfill site. This permit was
appealed to the State Environmental Hearing
Board by the opponents of the project who still
maintained that there were not sufficient environ-
mental protection measures at the site or in the
design to adequately safeguard their ground-water
supplies. It was also claimed that the Department
130
-------
of Environmental Resources had not evaluated the
design concepts for the site in enough detail in
order to provide assurances that the site would
function as designed. Additional questions were
raised concerning the strength and durability of
the 20-mil (.5 millimeter) PVC liner as the key
component of the leachate control system.
The Environmental Hearing Board held 10
days of hearings wherein numerous experts
testified as to each aspect of the site conditions
and landfill design. Because neither the citizens
nor their consultants raised any significant
questions which were not thoroughly addressed
or evaluated by the site study or landfill design,
the Environmental Hearing Board upheld the
issuance of this permit without reservation. The
decision of the Environmental Hearing Board set
forth findings of fact and findings of law which
stated that adequate subsurface investigations had
been conducted, adequate environmental protection
measures had been designed into the site, and the
review process carried out for this application had
been thorough and proper in the procedure of
evaluating these data and concepts.
In spite of this strong decision, the citizens'
group appealed the Environmental Hearing Board's
decision to the Commonwealth Court of
Pennsylvania. Because the county had anticipated
the possibility of this appeal, they had put into the
record of the Environmental Hearing Board
proceedings a considerable amount of detail
through the testimony of their technical
witnesses. Thus, the complete and thorough
documentation and record which had been
developed was then reviewed by the Commonwealth
Court. That court stated that their review of what
they termed a voluminous record compelled them
to conclude that the Environmental Hearing Board
decision was supported by substantial evidence and
that the citizens' group was not denied due process
of law nor was the permit issued arbitrarily or
capriciously. Therefore, 9 months after the Environ-
mental Hearing Board decision, the Commonwealth
Court affirmed that decision.
Based on the strength of these 2 court
decisions, Lycoming County is now proceeding
with the final design of this site and the preparation
of bid documents for construction. The citizens'
group has now filed a petition asking the Supreme
Court of Pennsylvania to hear an appeal to the
ruling of the Commonwealth Court. In view of the
strength of the previous decisions, it is considered
unlikely that the Supreme Court will consider that
appeal.
WATER QUALITY MONITORING PROGRAM
During the 2 years within which the construc-
tion of the landfill has been delayed by the permit
appeal and court action, a program has been
conducted which involved the periodic sampling of
the monitoring wells and periodic measurement of
water-table levels throughout the site. The 4
monitoring wells have been pumped and sampled
annually for a period of 4 years; this program has
provided base-line ground-water quality data for
evaluation of long-term quality trends. In addition
to the 4 bedrock flow system monitoring wells
which were constructed on-site during the sub-
surface investigation work, another bedrock flow
system monitoring well will be constructed down-
gradient of the site, and 2 shallower downgradient
monitoring wells will be completed within the
glacial till flow system at the south margin of the
site. Thus, in addition to the early warning monitor-
ing system provided by sampling the composite,
ground-water underdrain discharge, there will be
2 monitoring wells at the downgradient (south)
margin of the site within the glacial till, and 3
wells at the downgradient margin of the site
located within the shale bedrock. These 5 wells will
provide a sensitive, early-warning monitoring
system for use in evaluating the performance of
the leachate control system.
In addition to the continued ground-water
quality baseline monitoring program on-site and in
addition to the completion of 3 additional wells of
this monitoring program prior to operation of the
site, a private well monitoring program has been
initiated. There are approximately 40 homes with
private water-supply wells located within Vt mile
(.8 kilometer) of the landfill site and 6 other springs
or wells at a greater distance from the site which
represent well-known or unique features within
the regional flow system (which includes a
significant zone of carbonate strata). The private
well ground-water quality monitoring program (at
no cost to the homeowner) would be for the
purpose of documenting the existing ground-water
quality conditions at nearby homes prior to and
during operation of the landfill. This would protect
both the county and the homeowners in the event
of a change in water-quality conditions. Lycoming
County will benefit from this program through the
"insurance policy" effect which would result from
the documentation of the water-quality conditions
at each of the wells. These water-quality data
would allow a thorough evaluation of any claims
alleging damages from the landfill operation to a
domestic well. By providing each homeowner with
131
-------
a copy of the analytical results each time his
well is sampled, this program would clearly
demonstrate that the county cares that adjacent
residents worry about their water quality and that
the county has nothing to hide and has absolute
confidence in the ground-water quality protective
measures at their landfill site.
SUMMARY AND CONCLUSIONS
This particular project was selected from
among more than 75 sanitary landfill projects in
which the author has been involved because there
are several unique aspects of this project which set
it apart from the usual case-history example project.
1. The utilization of a plastic membrane liner
for leachate collection and control at this site as
part of a 4-component engineered barrier system
represents the state-of-the-art in lined sanitary
landfill design.
2. Backup leachate collection through diversion
of the ground-water underdrains provides an
additional level of protection in the unlikely event
of liner failure.
3. The naturally-occurring ground-water
protection features present at this site provide yet
another level of protection against the degradation
of ground water from leachate at the site.
4. The Federal Appalachian Regional Com-
mission grant of 1.2 million dollars for the construc-
tion of this sanitary landfill emphasizes the
importance of leachate control in this design.
5. This project has served as a test case
wherein the completed specifications for the
composition, installation, performance, and
guarantee of the membrane liner to be installed at
this site are presented in the EPA publication
"Liners for Land Disposal Sites; an Assessment."
6. Site problems are overcome through the use
of a membrane liner, and site benefits are utilized
in the design of this landfill to increase the degree of
ground-waier protection and to improve the
sensitivity of performance monitoring.
7. Questions and concerns raised by the
citizens' group opposed to the development of this
site were investigated and answered during the
hydrogeologic investigation of the site and during
the design of the landfill and development of the
operational plans.
8. The landfill site permit issued by the
Pennsylvania Department of Environmental
Resources was appealed by the citizens' group. Ten
days of technical testimony before the Environ-
mental Hearing Board examined all aspects of the
site conditions, landfill design, potential environ-
mental impacts, ground-water protection measures,
and monitoring systems. The permit issuance was
upheld by the Environmental Hearing Board
without reservation.
9. Anticipating another appeal of the Hearing
Board decision, a thorough presentation of all
aspects of the landfill site investigation and design
comments was presented before the Environmental
Hearing Board. The Commonwealth Court affirmed
the issuance of the permit due to the substantial
evidence supporting the decision on the initial
appeal.
10. This project, in spite of its state-of-the-art
design, primary and multiple backup ground-water
protection measures, and an extremely sensitive
ground-water quality monitoring system does not
satisfy some opponents. However, the regional
problem of solid waste disposal was too great for
this project to be dropped in the face of unfounded
claims that the ground-water resources in the area
would be contaminated by leachate from this
proposed landfill site.
11. At an initial refuse disposal volume of 400
tons (363 metric tons) per day, it will be necessary
to develop 4 fields of the site in order to have a
lifetime of approximately 10 years. The total unit
operating and capital cost for the first 10 years
operation is estimated to be $5.00 per ton, which
compares very favorably with fees at other landfill
sites which do not provide the same degree of
ground-water protection.
ACKNOWLEDGEMENT
The author wishes to thank the Leonard S.
Wegman Company for furnishing their site design
drawings to illustrate this report.
REFERENCES
Geswein, Alan J. 1975. Liners for land disposal sites: an
assessment. U.S. Environmental Protection Agency
Report, EPA-530/SW/137, 66 pages.
Giddings, Todd and Associates. 1973. Ground-water module
phase I and supplemental hydrogeologic report,
Lycoming County landfill site at the Allenwood prison
camp. Report submitted to the Pennsylvania Depart-
ment of Environmental Resources, 42 pages, 5 maps.
Wegman, Leonard S. Co., Inc. 1974. Status report and
application for permit, Allenwood control sanitary
landfill. Site design report submitted to the Pennsyl-
vania Department of Environmental Resources, 60
pages, 8 drawing sheets.
132
-------
DISCUSSION
The following questions were answered by M. Todd
Giddings, Jr. after delivering his talk entitled 'The Lycoming
County, Pennsylvania, Sanitary Landfill: State-of-the-Art in
Ground-Water Protection."
Q. by Bruce Brott. Were background monitoring wells
installed in both the bedrock and till aquifers?
A. Background monitoring wells were constructed only
in the bedrock aquifer. Two downgradient monitoring wells
will be constructed in the till aquifer, and one additional
downgradient monitoring well will be constructed in the
bedrock aquifer to provide a total of three downgradient
monitoring wells in the bedrock flow system.
Q. by Jeffrey A. Gilman. Was the regional ground-water
flow direction taken into consideration in the design and
placement of the monitoring wells?
A. The regional ground-water flow directions in both the
till and bedrock flow systems, as determined from the
hydrogeologic site investigation, were used to determine
the location for placing additional downgradient monitoring
wells in both of these flow systems.
Q. by James Braithwaite. In general, do you feel it would
be prudent to delineate aquifer and soils characteristics
outside and adjacent to the site in order to evaluate the
contingency ofleachate escape from the collection
measures?
A. Existing geologic mapping in the area of the site was
confirmed in detail by the hydrogeologic investigation
conducted within the site boundary. A reconnaissance
study was made of the surrounding area up to a distance of
between '/4 and '/2 mile from the site in order to confirm
the consistency of aquifer and soils characteristics outside
of the area of detailed study.
Q. by James Gibb. Is there a monitoring network around
the lagoon?
A. The downgradient monitoring wells for the landfill site
are also located downgradient from the leachate-holding
lagoon and, thus, will serve to monitor its performance also.
Q. by John Quagliotti. What type of lining does the leachate-
holding lagoon have?
A. This lagoon will be lined with a 30-mil PVC liner placed
on a prepared sand layer.
Q. by Herman Bouwer. What will be the cost of the 20-mil
PVC material to be used to line the landfill site area?
A. The installed cost of this material is estimated to be
$2.50 per square yard.
Q. by Charles L. Kleeman. Did you investigate the
possibility of degradation of the PVC liner by industrial
wastes or chemical sludges if they are to be allowed in
the landfill?
A. Disposal of industrial wastes and chemical sludges will
not be permitted at this landfill site. Since there is a
possibility that a small quantity of an industrial waste or
solvent could be disposed of in the landfill without the
knowledge of the operator, a one-foot-thick clayey soil
barrier has been provided on top of the six-inch sand layer
which overlies the liner. When this soil becomes
saturated, it will provide a capillary barrier and perch the
flow of leachate immediately above this layer. The
permeability of the six-inch sand layer and drainpipe
network overlying the liner will prevent saturated
conditions from existing and prevent the prolonged
contact of concentrated leachate flow with the liner.
Q. by Alexander Zaporozec. Why was the expensive
leachate collection system designed when the natural
conditions are favorable for the attenuation ofleachate?
A. The shallow water-table condition within the ground-
water discharge zone at this site and the large quantity of
refuse to be disposed of at the site would not allow for
the adequate attenuation of large quantities of leachate
within the glacial till underlying the site. The liner not only
will serve to collect any leachate which is generated, but it
will also prevent ground-water seepage from entering the
refuse and generating large quantities of leachate. By
minimizing the quantities of leachate which are generated,
the over-all leachate treatment and site operation costs will
be lower.
Q. by Robert D. Mutch. What provisions have been made
for leachate treatment and for leachate generation
minimization?
A. Leachate will be collected from the pipe network over-
lying the liner and flow to the leachate-holding lagoon where
it will be treated by aeration from floating-fan aerators. The
leachate quality may also require additional chemical
treatment within the lagoon. Final cover on the landfill
will be at least three to four feet of the native soil material
from the site, which is a high-clay-content glacial till. This
soil will be compacted, graded, and seeded in order to
promote runoff and the evapotranspiration loss of water by
plants.
Q. by James C. Warman. Have you considered transpiration
drying of the landfill as described by Dr. Fred Molz at the
Second National Ground Water Quality Symposium?
A. Leachate quantities and rates of generation were calcu-
lated by utilizing a water-balance method which took into
account evapotranspiration of water from the surface of the
landfill.
Q. by David W. Miller. What other types of treatment will
be u.sed in addition to the lagoon for the treatment of
leachate?
A. When the quantity of leachate generated from the landfill
exceeds the amount which can be recirculated back into
the landfill, a treatment plant will be constructed
on-site. The leachate will be recirculated to accelerate
stabilization of the refuse mass and to cause the decrease
in leachate potency so that treatment costs in this plant in
order to meet stream discharge criteria will not be prohibitive.
Q. by G. J. Thabaraj. What, if any, are the state and federal
requirements for discharge ofleachate to surface waters?
A. Discharge criteria covering a range of chemical parameters
similar to those established for sewage treatment plants have
133
-------
been set forth by the Pennsylvania Department of Environ-
mental Resources for the receiving stream adjacent to this
site.
Q. by George Clark. What mechanisms will you utilize to
detect and prevent migration of methane gas?
A. Gas vents will be constructed at a spacing of 200 feet
throughout the refuse deposit area and will penetrate the
entire depth of the fill. These vents will be constructed of
rock-filled, perforated concrete pipe sections and will allow
leachate perched within the fill to drain to the leachate pipe
network in addition to allowing the escape of gas from
within the refuse mass.
Q. by Kenneth Lustig. What was the cost of the site study,
landfill design, and well construction to date?
A. The cost of the site study including test drilling and the
monitoring wells, the design of the landfill site and
preparation of construction documents, and the coordination
time of the county staff and legal fees to defend the permit
through two appeal proceedings is approximately $300,000.
Q. by Kenneth Lustig. What will the annual operating cost
be of the completed landfill?
A. The annual operating cost will be approximately
$600,000 per year or about $5 per ton of refuse deposited
at the site.
Q. by Edgar Meiser. How much money has the Environmental
Protection Agency allocated in grant funds for the study,
design, and development of this site, and how does this
grant affect the anticipated $5 per ton disposal cost?
A. A grant of $1,299,379 is being administered by the
Environmental Protection Agency for this project. If these
funds had not been available, the disposal cost at this site
would probably be in excess of $6 per ton.
Q. by K. E. Childs. What is the life span in years of the PVC
liner; will the liner last longer than the site will produce
leachate; and is there a fund set aside to correct a failure if
it occurs?
A. The 20-mil PVC liner is expected to last far in excess of
20 years at which time it is expected that the potency of
any leachate which is produced will be minimal due to
recirculation and treatment of the leachate. There will be
an operating fund set aside for contingencies, which could
be used to correct problems with the leachate control
system.
Q. by K. E. Childs. 7s there any possibility of sinkhole
development beneath the site which could cause a major
liner failure?
A. There is no possibility of sinkhole development beneath
the site because there is more than 1800 feet of shale strata
underlying the site. Cavernous limestone strata are present
beneath the shale strata, but the formation of solution
openings at that depth below the site would have no effect
on the structural stability.
Q. by Walter Steingraber and Jack E. Sceva. What course of
action will be followed in the event that the leachate
collection pipes become clogged with iron oxyhydroxide
precipitates, as experienced in the collection of leachate from
the Army Creek Landfill site in Delaware?
A. Iron will be removed from the treated leachate which
will be recirculated into the landfill, and the soil layer
overlying the leachate drain system will also serve as a
filter to prevent precipitates generated within the fill from
reaching the leachate collection pipes. Should the pipes
become clogged, there is adequate permeability within the
gravel envelopes and sand layer adjacent to the pipes to
allow for free drainage of leachate to the collection sumps
at the lower end of the landfill.
Q. by Orest Tokarsky. Did the consultants hired by the
citizens' action group agree with most of your conclusions?
A. Yes. During the hearings held following the appeal of
this permit, a geologist and an engineer hired by the
citizens stated that they felt the site investigation was
comprehensive and entirely adequate in scope, and that the
over-all design of the landfill represented the current state-
of-the-art.
The paper by Michael Apgar, and the paper by James Mang, Robert
Stearns, and Dallas Weaver were not available for publication.
134
-------
Ground-Water Chemical Quality Management
by Artificial Rechargea
by Harry I. Nightingale and William C. Bianchi
:b
ABSTRACT
The effectiveness of basin ground-water recharge at
the Leaky Acres Recharge Facility in Fresno, California
for improving the regional ground-water quality was studied
as 65,815,000 m3 of high-quality surface water was
recharged from 1971 through 1975. Observation wells at
the facility showed some variability in chemical parameters
associated with each recharge period. The long-term
decreases in salinity could be described by a power law
decay curve fitted by regression analysis.
Without a special network of observation wells
outside the facility, scientific evaluation of the enclave of
recharged water is not possible. A practical evaluation of
water-quality changes is possible from producing water wells
around the facility. However, the pumping well discharge-
time variations, well depth, aquifer sequence, and prior use
of surrounding land must be considered, since all of these
factors affect the pumped-water quality and its seasonal
variability. Recharge at Leaky Acres had noticeably
decreased the ground-water salinity for a distance of up to
1.6 km in the direction of the regional ground-water
movement.
INTRODUCTION
Ground-water chemical quality has frequently
been degraded as a result of man's activities and
use of this resource (Kaufman, 1974). The causes
Water Management Research, Western Region, Agri-
cultural Research Service, U.S. Department of Agriculture,
Fresno, California 93726. In cooperation with the City of
Fresno Department of Public Works—Water Division.
Presented at The Third National Ground Water Quality
Symposium, Las Vegas, Nevada, September 15-17, 1976.
t>Soil Scientists, Western Region, Agricultural Research
Service, U.S. Department of Agriculture, Fresno, California
93726.
of degradation and alternative projects to prevent it
or perhaps improve the chemical quality are unique
for each ground-water basin (Gofer and Owen, 1975;
Schmidt, 1975a).
Artificial ground-water recharge is an alterna-
tive procedure for consideration in ground-water
quality management, especially when low-salt
surface water is economically available and the
geology is favorable. Even wastewater is currently
being considered for recharge, but more research is
needed before its utilization for recharge to aquifers
for unlimited use becomes a reality (Baffa, 1975;
Bouwer, 1974; Gofer and Owen, 1975).
In 1970, the City of Fresno, California, which
is completely dependent on ground water for its
domestic and industrial supply, initiated a basin-
type artificial recharge project. Surface water from
the Kings River is delivered at a cost of $8.107 per
1000 m3 ($10 per ac-ft). This water is available
because urbanization is replacing irrigated agri-
culture. The recharge facility, Leaky Acres, is
located about 7.2 km upgradient from the main
pumping depression which is apparent in the
regional ground-water gradient due to the urban
well field. During the first 5 years (1971 through
1975) 65,815,000 m3 (17.388 billion gal) of
low-salt water has been recharged at Leaky Acres.
Various research investigations associated with the
large basin-type recharge facility were conducted
during this time (Nightingale and Bianchi, 1973;
Bianchi and Lang, 1974;McCormick, 1975;
Nightingale, 1975; Bianchi and Nightingale, 1975).
Evaluating the effectiveness of a basin
recharge facility based on the regional ground-water
quality is analogous in many respects to evaluating
135
-------
contamination enclaves from surface waste disposal
sites (LeGrand, 1965; Kimmel and Braids, 1974;
Palmquist and Sendlein, 1975).
We could not economically justify determining
the detailed geological and ground-water hydro-
logical factors needed to develop a theoretical
shape of an enclave of high-quality recharged water
for this specific site. We expected the ground-water
chemical quality to improve without bacteriological
changes in the ground water. Therefore, like at
Leaky Acres, evaluating most ground-water quality
changes, associated with large recharge facilities,
will depend upon sampling the existing wells and
the variations accompanying: multiaquifer extrac-
tion, well depths, different usage, and non-ideal
areal distribution. This report considers five years
of ground-water quality data associated with
recharge at Leaky Acres as determined under the
above conditions.
STUDY AREA AND METHODS
Geologically, the Leaky Acres Recharge
Facility is located on a compound alluvial fan
(Page and LeBlanc, 1969) composed of vertical
and lateral heterogeneous layers of clay, silt, and
sand. Water movement in the fan sediments is
vertically impeded by the less permeable clay and
silt horizons, as indicated by perched water, while
lateral transmission is semiconfined specifically to
the sandy horizons. The geology and stratigraphy
in the upper 30 m at Leaky Acres has been well
defined (Nightingale and Bianchi, 1973; Bianchi
and Nightingale, 1975). Generally, there are three
low permeability layers at 4, 11, and 20 m depths;
each is about 1 to 3 m thick. In 1975, these layers
limited the facility's average infiltration rate to
14.2 cm/day. The ground-water mound dissipates
rapidly after each recharge period and is indicative
of productive aquifers, as indicated by Schmidt
(1975b). Well yields usually exceeded 126 I/sec
(2,000 gpm); specific capacities averaged about
25 1/sec/m (120 gpm/ft) of drawdown and trans-
missivities averaged more than 2,480 nWday
(200,000 gpd/ft).
Figure 1 presents the location of Leaky Acres
and the location and use classification of peripheral
wells used to evaluate the influence of recharge on
the regional ground-water quality. Behnke and
Haskell (1968) conducted a ground-water nitrate
study in this area. The urban development south
of E. Dakota Avenue (Figure 1; except for Fresno
Air Terminal) consists of lot sizes of about 557 m2
(6,000 sq ft), each with an individual septic tank
disposal system so the nitrate contamination in the
area was about 25 mg/1. Since north of E. Dakota
Avenue is mostly abandoned farm land with few
home septic tanks because of the clear-zone for
the airport, nitrate levels were usually less than
10 mg/1.
The initial areal distribution of ground-water
specific electrical conductivity (SEC) in the study
area (Figure 2) was established in 1971 from data
collected in the summer (May through August)
before recharge. The rural area north of E. Dakota
Avenue usually had SEC < 200 ^mhos/cm.
The surface water being recharged at Leaky
Acres is a low salinity, Ca-MgHCO3 type water,
delivered from the Kings River (29 km east of
Fresno) to the north side of Leaky Acres by the
Fresno Irrigation District's Gould Canal. Samples
for analyses were collected at inlets to the upper-
most basins on the same schedule as the ground-
water sampling at Leaky Acres.
Changes in ground-water quality beneath
Leaky Acres were observed by sampling ten 20.32
cm I.D. PVC plastic-cased water-quality wells
midway between top and bottom of slotted
section below the third slowly permeable layer.
The frequency of sampling of the wells at Leaky
Acres had decreased from twice a week in 1971, to
once a week in 1972, and semimonthly in 1973
through 1975. Ground water from 25 to 30
producing peripheral wells of various use-classifica-
tion was usually sampled semimonthly, unless
they were temporarily out of service, (i.e. irrigation
well during the winter), or monthly if they were
E DAKOTA
*
E SHIELDS
E CORNELL
E PRINCETON
E CLINTON
E McKINLEY
?
OI3
52
22*
e*
I
Z
z
• 62
-------
N
»t" E.GETTYSBURG ."230 \
E.ASHLAN
cc
lij
o
z
»240
• 243
I00«
119*
•164
176 •
'•201
E.DAKO
•170
•121
• 108
• 67 84*
•200
• 119
• 110.131 L86
•157
•H2|32/ «I50
*74I62/
Table 1. Selected Canal Water Quality Parameters.
Monthly Mean for 1971 Through 1975
Cl~ NO3 Temp. D.O.*
Month mg/l mg/l FTU* C mg/l
Jan. 2.51 2.77 13.0 8.8 8.00
Feb. 3.34 6.00 13.7 11.0 7.88
Mar. 1.58 1.21 8.7 13.0 9.80
Apr. 1.94 1.50 4.8 14.9 9.60
May 1.08 1.98 4.2 16.9 9.77
June 0.87 1.45 4.1 20.1 8.70
July 0.79 1.25 4.6 20.3 9.38
Aug. 0.72 0.95 2.8 19.8 9.00
Sep. 1.25 1.81 3.0 18.8 9.70
TA I301 . .158 Oct. 0.78 1.28 3.6 16.6 9.34
E. SHIELDS .407
• 282
10 ' Nov. - - -No Recharge- -
Dec. 2.65 1.78 4.5 10.0 8.00
. [ QQ O 1 C
* Turbidity in Formazin Turbidity Units.
Dissolved oxygen.
.346
• 338 *243 electrical conductivity (SEC, /imhos/cm) when in
£ the 20 to 650 nmhos/cm range, by the regression
.462 ^ equation: TDS = (SEC + 22.79)71.62.
2 36 • _
•251
f
z
RESULTS AND DISCUSSION
Fig. 2. Initial (May through August 1971) mean ground-
water specific electrical conductivity, /imhos/cm, for study
area.
fairly constant in quality. Water samples were
immediately analyzed for their nitrate, chloride,
and specific electrical conductivity (U.S. EPA,
1974). Ground-water turbidity and temperature
changes were also measured. Major cations and
anions were analyzed once yearly. For this study
area, the total dissolved solids (TDS, mg/l) can
be estimated accurately from the specific
ZI40
"|30
^100
o 90
UJ
to 80
of 70
H 60
£ 50
tu 40
£ 30
2 20
uj 10
i — i — i — i
JanFMAMJT AS ON Dec
MONTH
Fig. 3. Monthly mean specific electrical conductivity of the
recharge water for 1971 through 1975 recharge periods. No
recharge in November.
Canal Water Quality
The salinity of the canal water as evaluated
by the SEC was usually between 30 and 70
/xmhos/cm (Figure 3) during the recharge periods.
The higher values resulted from runoff of some
winter and spring storms. Table 1 presents the
monthly mean chloride, nitrate, turbidity,
temperature, and dissolved oxygen for 1971
through 1975. These data establish the excellent
quality of the water before recharge.
Ground-Water Quality Trends — Beneath
Leaky Acres
Figure 4 summarizes the ground-water SEC
trends during five years of recharging 65,815,000 m3
of high-quality water. Shortly after the beginning of
each annual recharge period, there was some
increase in ground-water SEC, but a new, low, SEC
was attained by the end of the recharge period.
The fluctuations in the Spring of 1973, shows the
impact of turning the water on and off because of
turbid canal water or shut-down periods in the
western basins because of insect problems. The
jump in SEC during August and September 1974,
is in response to a gypsum treatment over tile
lines used for water collection for a recharge well
study.
Because of the variability (Figure 4) caused
by irregular recharge periods, a power law decay
curve was fitted to the monthly mean SEC data by
regression analysis. The fit is satisfactory with the
137
-------
1260 -
Fig. 4. Response in ground-water SEC beneath Leaky Acres
as related to annual recharge periods. Each data point is the
mean of all observation wells and sampling times for a
month. In the regression decay curve equation, X is the
number beginning with X= 1, for January 1971.
coefficient of determination, r2 equal to 0.651. The
regression equation would predict a ground-water
SEC of 44 jumhos/cm by the December of 1980,
120 months after recharge started. This is
reasonable if there is no recharge during December
through February, when only water that is
slightly higher in salinity is available. Essentially,
the quality of the ground water beneath Leaky
Acres is nearly in a steady-state equilibrium with
the recharge-water quality.
Monthly mean ground-water nitrate levels
beneath Leaky Acres are presented in Table 2. The
nitrates in the soil profile before recharge were
essentially leached out during the first recharge
period. The slightly higher nitrate levels in the
spring resulted partly from mineralization of
organic matter during the winter dry period and
partly from higher nitrate levels (5-15 mg
in the canal water when it contains surface runoff
water from local storms. Fluctuations in the
ground-water chloride levels were similar to the
nitrate fluctuations. Before recharge the chloride
level averaged 2.8 ± 0.3 mg/1 for all wells at
Leaky Acres, by 1975 the high spring level was
1.7 ± 0.1 mg/1 and season's low was 0.7 ± 0.1 mg/1.
Periodically, the concentration of the major
cations and anions in the ground water were
determined. Table 3 presents the results only for
the June 1971 and October 1975 samplings.
Between these times the concentrations are
between those shown. Fluctuations in these
chemical parameters within a given recharge period
are similar to the fluctuations in SEC shown in
Figure 4. The milliequivalents per liter ratio of
Ca2+ to Mg2+ has increased from 0.61 to 1.16 during
the five-year period. The Ca2+ to Mg2+ ratio for the
canal water for the 1975 data was 1.13, about the
same as that for the ground water. The soluble
silica concentration was within the 30 to 50 mg/1
which is typical of the regional ground water. The
pH has remained between 6.7 and 7.0. The dissolved
oxygen content of the ground water averaged 8 to 9
mg/1 at the start of a recharge period and decreased
to 5 to 7 mg/1 by the end of a recharge period.
Ground-Water Quality Trends — Beyond
Leaky Acres
In evaluating the chemical data to determine
the significance of recharge at Leaky Acres on the
surrounding regional ground-water chemical
quality, we must remember several factors: (a) The
domestic wells are shallow, usually 30 to 40 m
deep, open-bottom, and the volume of pumped
water for house and yard use is small (as compared
with public supply and irrigation wells), and the
Table 2. Ground-Water Nitrate, mg/1. Trends Beneath Leaky Acres. Monthly Mean, and
for Some Months the Standard Deviation of Mean
Month
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1971
-
29.2 + 8.3*
—
—
—
15. 3 ±4.9
12.6
5.3
3.6
3.2 ± 1.0
2.3
3.5
1972
3.5
2.4
6.6
3.2
3.5
2.4 + 0.5
1.8
3.9
2.7
2.4 + 0.8
2.7
8.1
1973
6.2 ±0.9
4.5
3.2
3.2
2.8
2.3 ±0.4
3.2
2.5
2.0
1.7 ±0.2
1.1
1.2
1974
4.9
5.0
5.1
4.2
2.5
1.6 + 0.1
__
1.0
-
1975
_
4.0 + 0.5
1.7 + 0.1
_
0.7 ±0.1
1.2
Before recharge.
138
-------
Table 3. Change in the Mean and Standard Deviation of
the Mean of Selected Water-Quality Parameters for All
Wells at Leaky Acres Between 1971 and 1975
Quality
Parameter
mg/l
Na+
K+
Ca2+
Mg2+
HC03"
Cl"
N03
S042~
Si02
June 2, 1971
6.95 ± 1.32
1.98 ±0.20
5. 97 ±1.04
5. 90 ±0.60
69.67 ±5.23
1.85 ±0.21
12.6 ±4.52
38.6 ±2.4
48.0 ± 3.9
Oct. 28, 1975
2. 77 ±0.14
0.82 ± 0.02
4.89 ± 0.26
2. 55 ±0.09
36.12+ 1.68
0.70 ± 0.04
0.73 ± 0.09
7.6 ± 1.1
37.7 ± 3.4
Percent
Decrease
in Mean
60.1
58.6
18.1
56.8
48.2
62.2
94.2
80.3
21.4
observed annual variability in water quality is low.
For example, in 1974 periphery well PW-10 had an
SEC coefficient of variability (Cv) of 2.9%. The
domestic wells penetrate only the upper aquifers
and can be polluted if they are in dense, septic tank
areas, like PW-17, with a mean NO3~ level of 45.3
mg/l with Cv of 7.5% in 1974, as compared with
domestic well PW-11 (outside the dense septic tank
area), with a mean 7.6 mg NO371 in 1974. (b) The
urban area, which had only septic tanks, was
converting to the Fresno Sewer System, and most
connections were completed by the end of 1974.
Thus, this pollution source will eventually be
eliminated, (c) The public-supply wells are much
deeper (75 to 100 m) and usually not perforated
above 36 to 45 m. Fluctuations in pumped-water
quality were observed in the public-supply wells in
this area in response to seasonal use demand and
drawdown. For example, PW-21 had a 1974 mean
NO3~ level of 32.3 mg/l with Cv of 27.1% and PW-7
had Cv of 80.3% variability.
The mean ground-water chloride concentration
for the PW-7 (Figure 5), shows that during the
summer months of maximum pumping, lower
1971 ' 1972
1975
1973 ' 1974
Time,Years
Fig. 5. Ground-water mean chloride concentration for each
third of a year for peripheral well (PW) No. 7.
chloride levels (or SEC) were observed. This
variability, which was not the same for all public-
water wells complicates the long-term evaluation
of effect of Leaky Acres' recharge. Regression
analysis is probably the best practical way to make
long-term evaluations. The seasonal cyclic
variability is not great and for long-term effects
could be disregarded.
Table 4 presents the results of power curve
regression analysis for ground-water SEC for wells
at various distances from Leaky Acres, as a function
of time. The coefficient of determination for PW-6,
7, 14, and 21, in a south to southwest direction
decreased with distance from Leaky Acres. This
statistical method of describing trends in ground-
water quality associated with basin-type recharge
seems acceptable, considering expected life of the
recharge facility, which is probably less than 30
years, and the unjustified costs of a scientific
ground-water hydrological and quality study.
In the past ground-water nitrate levels in the
public-supply wells of this area of high septic tank
Table 4. Ground-Water SEC Time-Trends Predictions for Peripheral Wells at Various Distances and Directions
from Leaky Acres Based on Power Law Curve Regression Analysis of 1971 Through 1975 Data
PW
NO.
2
6
7
14
21
11
Approx.
Distance
m
200
200
400
800
1200
800
Approx.
Direction
SE
S
SSW
ssw
SSW
w
No. of
Obs.
93
68
75
85
104
83
Coef.
of
Deter., r2
0.835
0.693
0.592
0.217
0.106
0.404
Estimated SEC, y, jimbos/cm for Dec.
Regression Equation* 1971
y = axb x=12
A -> r / "0. 36 -i A y-
y = 356 x 146
y = 389 x"°-29 189
y = 346 x~°-24 190
y = 387x-°-064 330
y = 619x~°-13 448
y = 246x"°-12 183
1972
x=24
113
155
161
316
409
168
1973
x=36
98
138
146
308
388
160
1974
x=48
88
126
137
302
374
154
1975
x=60
82
118
130
298
363
150
1980
x=120
64
97
110
285
332
138
y = estimated specific electrical conductivity, jumhos/cm, after x-months from January 1971 when x = 1.
139
-------
Table 5. Mean Ground-Water Nitrates, mg/l, for May through August, During Maximum Pumping
Peripheral
Well
No.
2
16
8
6
4
7
10
20
14
17
11
19
21
18
Use
Domestic
Domestic
Domestic
Public
Irrigation
Public
Domestic
Public
Public
Domestic
Domestic
Public
Public
Public
Approximate
Distance
from Basins, m
200
200
200
200
400
400
600
800
800
900
800
1200
1200
1600
Nitrates, mg/l
1971
7.4
4.3
8.2
14.2
34.1
14.4
9.9
18.4
30.0
43.8
6.0
18.7
35.1
23.4
1972
8.2
2.3
12.2
15.1
OFF
14.7
12.1
21.4
30.1
44.9
7.1
23.5
36.9
25.1
1973
3.4
2.5
5.9
9.1
30.9
8.9
11.0
16.7
31.5
39.6
5.0
19.9
31.2
23.3
1974
3.0
4.1
2.8
7.4
OFF
6.1
10.9
22.1
22.8
44.7
7.4
28.8
33.7
27.6
1975
2.0
2.6
2.1
6.3
32.8
4.9
7.5
26.2
35.8
44.5
6.6
28.4
31.9
24.5
density have received considerable attention.
Table 5 presents summer (May through August)
trends in mean ground-water nitrate levels, ranked
according to approximate distance from the
recharge basins. In general, only wells less than
about 600 m have shown a nitrate decrease, and
initially these were low in nitrates, because they
were upgradient from the dense septic tank area.
The nitrate levels of pumped water in the septic
tank area have not greatly changed during this
observation period. Unsaturated flow of poor
quality water from the soil around the abandoned
septic tanks probably will continue for several
years because of necessary watering of lawns and
gardens.
Figure 6 shows the areal distribution of the
1971-1975 percent decrease in ground-water
salinity as evaluated by SEC from data for May
through August (maximum pumping period).
The decrease in SEC south of E. Shields Avenue
is evident, but as shown in Table 5 the nitrate
level apparently has not yet been noticeably
reduced by recharge, probably because of the high
septic tank density.
The Leaky Acres Ground-Water Recharge
Facility demonstrates that ground-water quality
can be improved or managed by basin-type
recharge using high-quality water. Scientific
evaluation of the enclave and movement of the
recharged water is difficult without a specially
built network of expensive observation wells and
monitoring equipment. All recharge facilities
should be monitored, which is possible if there are
producing water wells around the facility.
However, consideration must be given to their
pumping capacity, delivery time schedule, depth,
and zones of extraction. The historic and present
land use around the recharge facility, and the
stratification will affect the chemical quality of
the pumped water and its seasonal variability.
REFERENCES
Baffa, J. J. 1975. Artificial ground-water recharge and
wastewater reclamation. Jour. Am. Water Works
E.
t..
E.Sh
• E.Gettysburg .
O.Smi
0.80 Km
Ashlon c
o
5
m
S z
•o
at
O —
z . —
Jakofo —
Q 0-10% Decrease i
gll -30
mi31"50
Bfflsi-ioo *
• — ^*-
~ — j"i
' 1 •
^^= — iU
a^si
elds ^^^ =
— . -ZT" ~f
Pnrlpherol — —
well-*. —
J
nSEC
•
III
\^*
•*-!
•
— >>
w
a>
. .E
$
z
Fig. 6. Percent decrease in ground-water specific electrical
conductivity (SEC) between 1971 and 1975, based on May
through August data for these years.
140
-------
Assoc. v. 67, no. 9, pp. 471-476.
Behnke, J. J. and E. E. Haskell, Jr. 1968. Ground-water
nitrates distribution beneath Fresno, Calif. Jour.
Am. Water Works Assoc. v. 60, no. 4, pp. 477-480.
Bianchi, W. C. and G. J. Lang. 1974. The city of Fresno's
Leaky Acres ground-water recharge project-
construction and performance. Jour. Am. Water Works
Assoc. v. 66, no. 3, pp. 176-180.
Bianchi, W. C. and H. I. Nightingale. 1975. Hammer seismic
timing as a tool for artificial recharge site selection.
Soil Sci. Soc. Am. Proc. v. 39, no. 4, pp. 747-751.
Bouwer, H. 1974. Renovating municipal wastewater by
high-rate infiltration for ground-water recharge. Jour.
Am. Waterworks Assoc. v. 66, no. 3, pp. 159-162.
Gofer, J. R. and L. W. Owen. 1975. Solving the adverse salt
balance in the Orange County ground-water basin.
Jour. Am. Water Works Assoc. v. 67, no. 9, pp.
481-486.
Kaufman, W. J. 1974. Chemical pollution of ground waters.
Jour. Am. Water Works Assoc. v. 66, no. 3, pp.
152-159.
Kimmel, G. E. and O. C. Braids. 1974. Leachate plumes in a
highly permeable aquifer. Ground Water, v. 12, no. 6,
pp. 388-392.
LeGrand, H. E. 1965. Patterns of contaminated zones of
water in the ground. Water Resources Research, v. 1,
no. 1, pp. 83-95.
McCormick, R. L. 1975. Filter-pack installation and
redevelopment techniques for shallow recharge shafts.
Ground Water, v. 13, no. 5, pp. 400-405.
Nightingale, H. I. 1975. Ground-water recharge rates from
thermometry. Ground Water, v. 13, no. 4, pp.
340-344.
Nightingale, H. I. and W. C. Bianchi. 1973. Ground-water
recharge for urban use: Leaky Acres Project. Ground
Water, v. 11, no. 6, pp. 36-43.
Page, R. W. and R. A. LeBlanc. 1969. Geology, hydrology,
and water quality in the Fresno area, California. U.S.
Geol. Survey Open-File Report. Menlo Park, Calif.
Palmquist, R. and L.V.A. Sendlein. 1975. The configuration
of contamination enclaves from refuse disposal sites
on floodplains. Ground Water, v. 13, no. 2, pp. 167-
181.
Schmidt, K. D. 1975a. Salt balance in groundwater of the
Tulare Lake Basin, California. Proceeding of the 1975
meetings of the Arizona Section — Amer. Water Res.
Assn. and the Hydrology Section — Arizona Academy
of Science, v. 5, Hydrology of Water Resources in
Arizona and Southwest, pp. 177-184.
Schmidt, K. D. 1975b. Regional Sewering and groundwater
quality in the southern San Joaquin Valley. Water
Resources Bulletin, v. 11, no. 3, pp. 514-525.
U.S. Environmental Protection Agency. 1974. Methods for
chemical analysis of water and wastes, EPA-625-/6-
74-003. Office of Technology Transfer, Washington,
D.C.
DISCUSSION
The following questions were answered by Harry I.
Nightingale after delivering his talk entitled "Ground-Water
Chemical Quality Management by Artificial Recharge."
Q. by Abe Kreitman. Has the recharge capacity of the basins
changed with time? If so, what is the nature of the change
and what is the nature of the rehabilitative procedures
employed?
A. During the first five recharge periods the measured
average infiltration rate for all basins has not statistically
changed. The basin soils have been disked once during the
winter drying period to mix the accumulation of organic
material into the soil. We feel that disking should not be
done unless necessary to prevent surface sealing due to
accumulation of organic or inorganic sediments.
Q. by J. Brown. Will the amount of recharge water available
for the Leaky Acres project be affected by continued dry
conditions in California or are quality criteria the only
reasons for discontinuing recharge?
A. During the dry year of 1976, water was made available
for recharge through October even after the Fresno
Irrigation District had to shut off surface-water deliveries
to farmers on July 15. Perhaps this would indicate a
higher-use priority, especially when the farmers also have
ground water available for irrigation in water-short years.
For the second part of the question, the historical chemical
quality of the surface water available to Fresno for recharge
would indicate that its quality would probably never be
a criteria for discontinuing recharge.
Q. by Leonard Konikow. Would you recommend the use of
a deterministic model for studying artificial recharge
problems?
A. Yes, but only from the point of view that you have a
real need and can justify the cost. There is no question that
each recharge site is unique, and quantity and quality
modeling is beneficial to our understanding recharge
problems, but the problem of cost is difficult to justify.
Q. by K. E. Childs. (a) Specifically what was the source of
the recharge water? (b) Were the actual zones receiving
recharge water monitored or were the monitoring wells in a
different horizon?
A. (a) The source of recharge water is the Kings River about
18 miles (29 km) to the east of Leaky Acres and is delivered
by the Fresno Irrigation District's Gould Canal, (b) The
zones receiving recharge water were actually monitored at
Leaky Acres by observation wells perforated in the lower
soil zones beneath the basins. Outside the bounds of
Leaky Acres, domestic, public supply, and irrigation wells
were monitored and they extracted water from many
141
-------
different sandy zones, which for wells further away from
Leaky Acres, probably are not continuous with the zones
monitored beneath the basins.
Q. by L. A. Swain. Do yon have any evidence that shows
whether the better quality water recharged is vertically
mixed throughout aquifer depth or floats on top of older
ground water?
A. The construction of the water wells monitored around
Leaky Acres precludes the determination of vertical mixing
based on chemical data of pumped water. The density
difference between the native ground water and the
recharged water is small.
Q. by Mike Kaczmarek. In areas of short water supply where
ground water is a major source of drinking water and
municipal supply, where will you obtain recharge water?
A. The management of the ground-water quality and
quantity in a given area does demand the input of water of
equal or better quality to maintain the quantity of water
in storage, as well as an output of excess salt, be it to the
ocean or a salt lake. The need for input by recharge should
be first minimized by area-wide water conservation,
especially keeping precipitation within the area, and
maximum re-use of waste waters, then high
quality surface water should be imported. If additional
water is not economically available, then the people of the
urban area should determine how they can continue to
exist within their limited water supply and management
capabilities.
Q. What was the problem (need for recharge), and what does
it cost?
A. The problem in Fresno, California, is typical of urban
areas in arid climates in that the water table has been
falling as use has increased and septic tanks have resulted in
a slow deterioration in ground-water quality in some areas.
For the Leaky Acres Recharge Facility, and assuming
12,000 acre-feet per year of recharge, the total users costs
are $16.98 per ac.-ft., which is composed of $4.56 per
ac.-ft. for recharge operations, $10.00 per ac.-ft. for
purchase of water (Bureau of Reclamation), and $2.42 per
ac.-ft. for energy costs only to recover the water. For Leaky
Acres there is a secondary benefit to the facility, namely
benefits associated with Federal Airport clear zone land
purchase contributions. If this benefit is considered, then
the recharge costs are $3.53 instead of $4.56 per ac.-ft.,
which would reduce the total users' costs to $15.95 per
ac.-ft.
142
-------
The Dashte-Naz Ground-Water Barrier
3,
and Recharge Project
by Dennis E. Williams
ABSTRACT
The danger from salination poses a significant threat
to development of ground-water resources for domestic or
irrigation use in near-coastal environments throughout the
world. The Dashte-Naz farm area near the Caspian Sea in
northern Iran is no exception. Large-scale ground-water
development in this region has been limited to wells south
of the Caspian S.ea near the base of the Alborz mountains
where sediments are coarse and natural recharge high. In
the northern regions, however, exploitation of the ground-
water resources for agriculture and livestock purposes has
proven fruitless with most of the high-yielding wells turning
saline after only several irrigation pumping seasons. The
source of the salination comes not from encroachment of
Caspian Sea water but rather from sediments saturated
with connate or fossil waters. A wide zone of dispersion
separates the salinated aquifers from the fresh-water
aquifers nearer the mountains.
The Dashte-Naz ground-water barrier and recharge
project was conceived for purposes of overcoming the
problem of exploitation of fresh ground-water reserves in
the vicinity of saline-water fronts. The project presently
under construction is a pilot research project on the
practicality of using an injection well barrier for storing
water underground in areas immediately adjacent to saline
ground-water reservoirs. The results from this project will be
used to design other water-supply projects in Iran not only
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
b Agro-Water Consulting Engineers, Tehran, Iran.
along the Caspian Sea coast, but along the Persian Gulf
as well.
INTRODUCTION
Salination of Aquifers
The word "contamination" has been used in
the past to describe the condition of ground water
which has been degraded in quality by the
invasion of saline waters. However, contamination
is generally associated with harmful chemicals or
sewage, and is hazardous to public health. For
technical continuity, therefore, it seems necessary
to use a term explicit in meaning but not limited
by definition as to source and concentration. The
word "salinated" has been used throughout the
literature for the past seven or eight years to
describe this condition of aquifer degradation.
Salination of aquifers can result from several
conditions depending upon the geologic and
hydrologic factors active in the particular area. One
of the most common occurrences resulting in
aquifer degradation is from intrusion of saline water
in near-coastal alluvial deposits. For example, along
the coast of southern California, intrusion of salt
water was first noticed in the early 1900's along
the bay margin of Mission Valley in San Diego
County. By the 1930's, the encroachment was
detected as far north as Los Angeles, Ventura and
Orange Counties. Since the 1940's, the natural
seaward hydraulic gradient in many of the near-
coastal communities has been reversed and control
of salt-water intrusion is now the object of several
143
-------
major projects along the coast of southern
California. The problem, however, is not unique to
southern California and is a common occurrence
throughout all coastal areas in the world (the
aquifers adjacent to the Caspian Sea or Iran being
no exception).
The extent and degree of salination depends
not only upon the geohydrologic factors affecting
the vertical and lateral movement of the saline-
water front, but also upon the extraction and
recharge which is active in the area. The extraction
or recharge may be natural or man-made, but
whichever the case, the dynamic movement and
shape of the saline-water front is greatly affected.
To properly manage and control ground-water
resources in close proximity to potential saline-
water encroachment, one must first understand
the ground-water hydraulics of flow governing
this encroachment with a degree of confidence
enabling proper prevention and control of all
undesirable factors.
Definite progress has been and is being made
regarding conservation and protection of ground-
water resources near salinated zones in the
northern coastal area of Iran near the Caspian Sea.
In this area a project is presently underway in
which an injection well barrier is being used to
protect fresh ground-water supplies from
salinated aquifers lying seaward to the north.
The results from this study show that the well-
known principle of a ground-water pressure ridge
(as practiced extensively throughout the world
for the past fifteen years), can and is being adopted
for use in coastal areas of Iran (Figure 1).
History of Ground-Water Development in Iran
Since ancient times, Iran has been an agri-
cultural country Limited surface-water resources
Fig. 1. Control of salt-water encroachment using an
injection well barrier.
in most areas reflect the low annual average
precipitation forcing farmers to develop ground-
water resources to meet irrigation demands.
Evidence of thousands of years of ground-water
development can be seen by the numerous
hand-dug shallow wells and ghanats scattered
throughout the country.
Ghanats represent a very old system of
ground-water exploitation dating back thousands
of years. Basically, they consist of subterranean
tunnels dug mainly in alluvial fans with the upper
portion of the tunnel contacting the water table.
These tunnels act as drains on the phreatic aquifer
providing a continuous flow of water, the amount
and duration of flow dependent upon tunnel
seepage losses and natural recharge. In recent years,
decreasing phreatic water levels and ghanat
maintenance problems have forced farmers to seek
the untapped ground-water reserves lying in the
deeper confined and semi-confined aquifers.
The advent of modern day drilling equipment
has thus produced an explosion in the number of
wells tapping the deeper aquifer zones with the
result being an overdraft or mining condition
occurring in many areas. Recent control by the
government of Iran now requires conservation of
all ground-water resources by application of
modern planning methods on ground-water
exploitation and use. The Dashte-Naz ground-
water barrier and artificial recharge project reflects
such planning.
THE DASHTE-NAZ AREA
Location
The area known as Dashte-Naz is located near
the Caspian Sea in the northern part of Iran (see
Figure 2). The total project area covers about
34,000 hectares with the Dashte-Naz farm area
itself occupying about 3,000 hectares. The project
area contains only one distinct topographic
feature; that is, a plain with relatively smooth
topographic highs and elevations ranging from -28
to +40 meters above the mean sea level of the
Persian Gulf. In the southern portion of the
project area, two alluvial fans exist formed by the
accumulation of sediments from two large rivers
(Tajan and Neka). A topographic depression exists
between these alluvial fans as the result of the lack
of sedimentation between the rivers.
The Need for Protection
The Dashte-Naz ground-water barrier initiated
out of the necessity to supply irrigation water to
144
-------
MIDDLE EAST AREA
Fig. 2. General project location map.
the Dashte-Naz farm area. Originally, the project
was conceived for the purpose of importing water
from outside sources as it was thought that local
ground-water resources existing in the aquifer
systems within the farm boundaries were not fit
for either irrigation or domestic purposes (early
deep exploratory wells yielded electrical con-
ductivities of 6000-7000 micromhos/cm). This
latter theory was quickly disproved once results
from precise exploratory cores were made available.
Geohydrologic and geochemical test results show
that fresh-water aquifers do exist in certain areas
of the Dashte-Naz farm, namely in the southwest
and northeast portions. However, exploitation of
these fresh-water supplies is severely hindered by
the ever-present threat of saline-water encroach-
ment from the north.
Ground-Water Salinity Problems at Dashte-Naz
The two large alluvial fans in the project area
are the result of deposition of sediments carried
by the Taj an and Neka Rivers. The Taj an alluvial
fan extends northward more than 20 km from the
base of the Alborz mountains while the Neka
alluvial fan, considerably smaller in areal extent,
continues only 10 to 15 km northward from the
base of the mountains. A definite topographic low
point dividing these alluvial fans occurs in the
vicinity of the Dashte-Naz farm. This low point
coincides approximately with the mid-point
between the two river systems (see Figure 3). The
transgression and regression action of the ancestral
Caspian Sea in Plio-Pleistocene times formed many
interlayered alluvial and marine deposits in these
alluvial fans (undisturbed core samples taken
during the exploration phase show alluvial sediments
interbedded with marine sands, silts and marls).
The salinity of ground water in the Dashte-Naz
area generally increases from south to north
reflecting the direction of alluviation. In the alluvial
fans where the amount of natural recharge is high,
finger-like lenses of fresh water protrude northward
into areas of brackish-water aquifers. Specifically,
fresh ground water occurs in abundant quantity
near the Taj an and Neka Rivers where coarse
sediments exist and transmissivities are high.
Continuous circulation of fresh water in these areas
has resulted in a flushing action over the past several
million years cleansing the aquifers of the original
connate waters. Away from the rivers, however,
the magnitude of the natural flux decreases
significantly resulting in brackish-water aquifers
(Figure 4).
Geohydrologic and Geochemical Investigations
During the investigation phase of the project,
over eighty deep exploratory boreholes were drilled
for the purpose of differentiating fresh and
salinated aquifers. Precise coring using the double-
core barrel method combined with a complete
suite of geophysical borehole logs produced the
most concrete results regarding the geologic
formations underlying the farm area. It soon
became evident that fresh-water aquifers did in fact
exist in the western and easterri portions of the
farm near the alluvial fans of the Taj an and Neka
Rivers. These fresh-water aquifers, however, are
limited to the upper 100 meters or so of depth
Fig. 3. Dashte-Naz farm area.
145
-------
Fig. 4. Relation between fresh and salinated aquifers in
Dashte-Naz.
ER AQUIFERS
Fig. 5. Geologic cross section through upper fresh-water
aquifers in the southwest farm area.
and are separated from deeper high-pressure
salinated aquifers by thick silt and clay confining
layers (see Figure 5).
A geologic "type" section was developed
describing the fresh-water aquifers in the Dashte-
Naz area and consisting of three zones:
1. The A zone is composed of alluvial sands
and gravels typically found between 30 and 40
meters depth.
2. The B zone consists of a clean marine sand
deposit lying between 50 and 60 meters depth.
3. The C zone is composed of layers of fine
beach sands interbedded with some silts and
generally found between 75 to 90 meters depth.
The three fresh-water aquifers are separated
hydraulically from each other by aquicludes of
plastic clay and silt. The confining layers are
"tight" as evidenced by separate piezometric
levels measured in piezometers completed in the
individual aquifers.
A thick silt and clay aquiclude underlies the
fresh-water aquifers at about 90 or 100 meters
depth and separates the fresh upper layers from
deeper connate-water sands. The pressure in these
lower salinated zones ranges from two to three
atmospheres with a chloride iron content of
12 grams/liter (twice that of the Caspian Sea
water).
These lower aquifers formed when the over-
burden began to increase during the sedimentation
process, forcing the salty waters to migrate from
the silt and clay layers into adjoining sands. As the
sands were compressed much less than the clays,
they received more and more saline waters with
internal pressures increasing as the overburden
became thicker. The end result is a formation
containing "pockets" of high-pressure saline-water
sands hydraulically isolated from the fresh-
circulating ground water.
A general geohydrological model of the area
was developed after results from the exploratory
drilling and testing became apparent. A two-
layered model was adopted: the upper layer (above
100 meters) consisting of the fresh circulating
ground water; and the lower layer (below 130
1SOCHLORS I MG / L )
FRESH -WATER AQUIFERS
Fig. 6. Interface between fresh water and salinated
aquifers in Dashte-Naz.
146
-------
meters) consisting of the high-pressure connate-
water sands.
Identification of the areal variation of salinity
in the upper fresh-water zones was the major
purpose of the exploration works. Defining the
zone of dispersion between the fresh-water aquifers
lying near the alluvial fans and the brackish-water
zones existing in the low point between the fans
took many months of careful drilling, testing and
evaluation. Coring, vertical electrical conductivity
profiles and electrical resistivity, gamma ray and
gamma ray neutron logs were used extensively to
locate this zone (see Figure 6).
Isochlor maps combined with aquifer test
analyses showed that the most promising area for
construction of a ground-water barrier and
artificial recharge project would be in the southwest
portion of the farm. Although the northeast
portion of the farm showed a similar fresh-
saline water dispersion zone, it was discarded as the
first choice for an artificial recharge and recovery
scheme due to the fact that the aquifers were less
distinct and significantly lower yielding.
MANAGEMENT AND PROTECTION OF THE
GROUND-WATER RESOURCES IN THE
DASHTE-NAZ AREA
The Pumped-Storage Concept as Applied to
Ground-Water Reservoirs
The principle of pumped-storage hydroelectric
plants is common practice in many surface-water
reservoirs. These plants maintain enough storage
during off-peak hours for use during peak hours of
the same day. The pumped-storage plant generates
high-value peak power from the headwater to the
tailwater pool when the demand is high. Low-value
power is then used to pump water from the
tailwater pool to the headwater pool during periods
of low demand. The value of the pumped-storage
plant lies in its efficiency to convert low-value off-
peak power to high-value peak power.
The Dashte-Naz ground-water barrier and
artificial recharge project operates on a plan similar
to the pumped-storage system of a hydroelectric
plant. During the non-irrigation season (off-peak
period) when the demand for water is low, water
from outlying well fields is piped into the project
area and stored underground for later use when
water demand is high. The water is physically stored
through a series of injection wells tapping the
upper three fresh-water aquifers. The artificial
recharge cycle lasts about nine months of the
year (September through May). During the
three-month irrigation season (June, July, August),
the imported water is then used to help meet the
field irrigation-water requirements and the
previously artificially recharged water, recovered
through extraction wells makes up the balance of
the irrigation demand.
Field Irrigation-Water Requirements
As the first real solution to the problem of
meeting the field irrigation-water requirements at
Dashte-Naz, supply wells were drilled in the
vicinity of the Tajan and Neka Rivers. These wells
were out of the geohydrologic influence of the
farm area due to high natural recharge near the
rivers and thus posed no danger of inducing
saline-water encroachment. The pumped ground
water from these outside source wells is conveyed
into the farm area irrigation canals through a
system of pipelines.
The water presently imported is able to
provide the necessary irrigation water for most of
the farm area. The eastern portion of the farm is
supplied by imported ground water pumped from
ten wells drilled near the Neka River. The western
farm area is irrigated by ten wells drilled near the
Tajan River and conveyed to the farm by two
pipelines. The central portion of the farm has a
deficiency in water supply which will be met by
utilization of the ground-water resources available
within the farm boundaries itself.
Injection Well Barrier
To meet the water requirements in the central
farm area, the concept of a ground-water barrier
was conceived. The design provides irrigation water
by storing imported water, as obtained from the
outlying well fields, underground into the three
fresh-water aquifers of the southwest area and
pumping it out later when needed for irrigation. A
total of five production wells are needed in the
southwest area to recover the artificially recharged
water. These recovery wells will then export the
water to the central farm area through a pipeline
and pumping station (see Figure 7). In order to
protect the southwest well field from the threat of
saline-water encroachment during the pumping
season, a ground-water barrier of approximately
2,000 m in length has been constructed. The
barrier consists of nine injection wells spaced
approximately 250 meters apart. The alignment
of the barrier, as seen on Figure 4, parallels the
isochlor lines and is normal to the direction of
encroachment.
The spacing, injection rates and total length
of the barrier were designed based on results
147
-------
Fig. 7. Dashte-Naz area showing injection wells, recovery
wells and imported water supplies.
from interference testing and mathematical
simulation models of the area.
The injection wells are constructed so that
selective injection can be made into individual or
various combinations of the three aquifer zones in
the event that salination occurs in any one or all
of these zones. The three aquifers are cemented off
from each other and pneumatic packers installed
to prevent cross flow. At the present time, inter-
mixing of the aquifer zones is permitted through
the screen sections as the ground-water quality in
all three zones is excellent. Figure 8 shows a
typical cross section of an injection well.
Source of Water for Irrigation and
Artificial Recharge
During the early planning stage, several
sources of water for artificial recharge were
investigated. Surface-water diversions were
discarded due to irregular flows and sedimentation
problems. Ground water imported from areas
outside of the influence of the Dashte-Naz farm
area proved to be the most practical and reliable
source of injection water.
Three main well fields were thus developed
and connected to the Dashte-Naz farm area
through a pipeline conveyance system. The ten
wells developed near the Neka River on the eastern
side of the farm, provide an instantaneous flow of
2,400 m3/hr.
Two separate pipelines from the western side
import water from ten wells tapping aquifers near
the Tajan River, producing 2,500 m3/hr.
The Artificial Recharge and Recovery Scheme
at Dashte-Naz
Operation of the ground-water barrier consists
of injecting water imported from the Tajan River
Fig. 8. Cross section of typical injection well.
supply wells during the non-irrigation season
(September through May). At the end of this
9-month injection cycle, a rise of approximately 8
meters will have occurred in the piezometric
surface in the vicinity of the southwest barrier
wells. At this time, the water previously piped from
the Tajan River wells will discharge into the
irrigation canal system to meet the field irrigation-
water requirements of the southwest area.
Also at this time, the five production wells
located in the southwest farm area will be turned
on to recover the previously artificially recharged
water. The collective discharge from the recovery
wells will be piped into the central farm area to
meet the irrigation-water requirements there. The
ten supply wells near the Neka River will also be
148
-------
B
BARRIER
i
INJECTION
IRRGATION
-
"
3
-i
TIME
1
1
PUMP NG
-
•i
MONTHS ^
Fig. 9. Artificial recharge and recovery cycles.
started and their water used to meet the field
irrigation-water requirements in the eastern farm
area (see Figure 9).
Supervisory Control System
To properly regulate, control and monitor the
complex system of injection and production wells
throughout the year, a Supervisory Control
System is presently being installed whereby pump
motors, pipeline valves, and water-quality sensing
devices can be interrogated and remotely con-
trolled through a central computer housed in a
new farm headquarters building. From this central
control room, continuous records will be kept on
all well flows, pressures and water-quality changes,
so that immediate regulation can be made should
adverse conditions arise.
The Supervisory Control System basically
consists of five remote stations which relay
digital information to a Supervisory Control Master
Unit which in turn conveys the information to
the computer (see Figure 10). The computer
converts these digital data into pressure readings,
valve set points, flow measurement and electrical
Fig. 10. Supervisory control system layout.
conductivity and chloride ion concentrations and
stores the data on memory disks generating a
historical record of operation.
The real-time data are continuously checked
against minimum and maximum allowable ranges,
and alarm messages printed on the operator console
if the ranges are exceeded. Background programs
consisting of ground-water simulation models and
pipeline hydraulic flow models are automatically
loaded into the computer if conditions warrant
new operating set points.
The operator of the computer in the central
control center can, at any time, selectively
interrogate remote stations to obtain information
on any specific measurand. Predictive models can
also be run simultaneously during real-time data
acquisition in order to anticipate operating regimen
changes due to variations in injection well pressures,
flow rates, or pipeline flow variations.
The Supervisory Control System at Dashte-
Naz represents application of modern-day-computer
technology to the field of water-supply management
and allows a degree of flexibility and accuracy of
operation virtually unattainable by conventional
water-management techniques.
SUMMARY
Protection of fresh-water supplies from
saline-water encroachment has been practiced for
many years throughout the world. However, the
results from the Dashte-Naz project will be the first
such in Iran to illustrate conservation and mana; <--
ment of ground-water reservoirs in salinated areas.
Results from the Dashte-Naz ground-water barrier
and recharge project will be used to design ground-
water barriers in other Caspian Sea coastal areas
and along the Persian Gulf. Operation of the
project requires extensive and careful engineering
and management if the project is to become a
smooth and efficient operation. One of the most
important tools to ensure this efficiency is the
remote control monitoring of all key elements in
the project. A Supervisory Control System which
monitors and regulates the artificial recharge and
recovery scheme is therefore indispensable. In
addition to regulating water flows, hydrometeoro-
logical and soil moisture information will also be
relayed to the Supervisory Control System
computer to aid decision-making during times of
irrigation and harvesting.
In summary, the Dashte-Naz ground-water
barrier and recharge project is a practical example
of applying new methods and technology to the
field of ground-water resources management.
149
-------
DISCUSSION
The following questions were answered by Dennis E.
Williams after delivering his talk entitled "The Dashte-IMaz
Ground-Water Barrier and Recharge Project."
Q. by K. Childs. What was the total cost of the project, and
•who paid? What percent of injected water was recovered?
A. The total cost of the project including the exploration,
final design, and supervision of construction was approxi-
mately $6,000,000. The Iranian Government funded the
project and the Ministry of Energy administered the project.
The percent of injected water recovered will not be
'accurately known until after several years of Project opera-
tion but from the initial tests, it is predicted that 85 to 90
percent of the injected water will be recovered by inland
well fields.
Q. by Bill Wilmitts. Was a barrier well line also considered
for the east well field?
A. Yes, an injection well barrier was considered to protect
the northeastern portion of the Dashte-Naz ground-water
reservoir; however, as the first stage it was felt the southwest
reservoirs were more geologically suited for ground-water
storage due to better definition of the aquifer systems. In
the future, an injection well barrier will most probably be
constructed to protect the northeast aquifers as well.
Q. Are the sediments you referred to called "the Kuwait
group"as the oil companies named it?
A. No, the sediments underlying the Dashte-Naz farm area
are mainly interbedded Quaternary alluvial and marine
deposits and, as far as I know, the N.I.O.C. (National
Iranian Oil Company) has not given any name to these
formations.
Q. by C. Roberts. Is there a functional relationship between
injection rate and assimilation rate of the fresh water into
the salt-water wedge?
A. The quantitative relationship between injection rate and
loss of fresh water to the dispersion zone was determined
analytically through the use of a distributed parameter
digital computer simulation model and verified somewhat
experimentally with field tests.
Q. How many acres are mechanically irrigated today?
A. Approximately 3,000 hectares or 7,400 acres are irrigated
by a combination of moveable sprinklers, center-pivot
sprinklers and conventional ditch and furrow irrigation.
Q. by R. P. Chagnon. Did you have any problems with cross-
flow between the three aquifers; are they hydro logically
connected and how long will the well last?
A. The three aquifer systems are hydraulically separated
from each other by thick aquicludes of silt and clay. This
has been verified by different piezometric surfaces measured
in the individual aquifers. At present, cross-flow is taking
place between the aquifer systems through the screen
sections but this is not felt undesirable as all three
aquifers presently contain fresh water and the pressure
gradients are small. However, should one of the aquifers
become salinated it will be mechanically isolated by a
packer to prevent an undesirable cross-flow condition.
The life of an injection well depends mainly upon the
clogging rate of the screen sections. It is foreseen that
redevelopment at least once a year will be needed; however,
some injection wells in the West Coast Basin Barrier
Project of Southern California have not been redeveloped
in ten years.
Q. What about vertical upward intrusion underneath the
pumping well field?
A. It is hoped that upconing of saline waters from the under-
lying connate-water aquifers will not take place due to the
careful attention paid to completion of the injection and
recovery wells. A thick aquiclude separating the fresh and
connate zone exists between 100 meters to approximately
130 meters depth and unless this aquiclude is penetrated
no upconing of undesirable waters should take place.
Q. by J. Brown. Is there any danger of areal subsidence
caused by the release of artesian pressure of the connate
water?
A. Since there will be no pumping of the connate-water
aquifers there will be no release of artesian pressure in these
zones. As previously mentioned, only the upper 100 meters
of fresh-water aquifers will be used for the artificial
recharge and recovery scheme.
Q. by Don Runnells. Do the native people have the training
to maintain this complex system or is it the plan to keep on
consultants indefinitely?
A. All Iranian government contracts call for training of local
personnel to properly operate and maintain the project
once the foreign consultants terminate their work. For the
Dashte-Naz Project the consultants have foreseen training
at least eight to ten qualified Iranian personnel to operate
the system. Normally, a lesser number of people would be
trained for a project of this size but, due to the sophistica-
tion and complexity of the Dashte-Naz Project, it is felt
advisable to train as many as possible.
Q. by Jean Schmidt. By what mechanism does salt water
migrate out of the clays into the sands at depths of 100
meters or so ?
A. During the accumulation of sediments in the region, water
of deposition was trapped in the sediments due to this
process. Progressively increasing pressures caused the by-then
salty waters to be "squeezed out" of their clays and silts
migrating into the adjoining sand layers which, owing to
their properties, were compressed much less than the
finer-grained silts and clays.
As this mechanism continued, the sands received more
and more saline waters with internal pressures progressively
increasing as the overburden became thicker. The end result
is that these lower zones now contain fossil or connate
water with the internal aquifer pressures quite high, often
reaching several atmospheres in some areas.
Q. by Thomas R. Schultz. Is this system economical with
such costly technology?
A. No, this system is not economical as judged by benefit-
cost analysis techniques. However, this system was never
intended to be economical as it is a research project on the
150
-------
practical development of ground-water resources in the
vicinity of salinated zones. The results from this project
will be used to design other recharge and recovery schemes
not only along the Caspian Sea coast but the Persian Gulf
as well. In areas where this method of ground-water
conservation and management is the only alternative left
for water supply, the benefit of just having the water
available cannot be judged by standard economic
indicators.
Q. by W. B. Wilkinson. What is the suspended solid
content of the recharged water? Is there any danger of
clogging of the wells and, if so, are back washing facilities
available?
A. The recharged water will be imported ground water
and the suspended solid content (after initial start-up of
the wells) will on the average be below 1.0 mg per liter. At
least once a year, or when the injection pressures rise and
injection rates decline, backwashing facilities will be
accomplished by mechanical surging and airlift pumping.
Chemical treatment will also be performed.
Q. by Dale Ralston. What is the approximate water cost per
acre irrigated?
A. The specific construction cost would be approximately
$810 per irrigated acre. The annual operating cost of the
supply-well system is approximately $50,000. This amounts
to approximately $7 per acre irrigated. However, these
figures do not reflect the cost of building the existing
irrigation and drainage canal and road system, but only
reflect the civil and appurtenant works connected with the
ground-water barrier and recharge portion of the project.
151
-------
Injection/Extraction Well System—A Unique
3,
Seawater Intrusion Barrier
by N. Thomas Sheahan
ABSTRACT
A multiple-aquifer system in the bayfront area of
Palo Alto, California is being intruded with seawater from
San Francisco Bay. In order to combat this potential
degradation of the ground-water supplies in the area, a
sea-water intrusion barrier is being constructed consisting
of a series of injection wells used to inject 2.0 million
gallons per day (7.6 X 106 1/d) of reclaimed wastewater into
a shallow aquifer. The injected water is subsequently
removed by a similar system of extraction wells to avoid
any possible degradation of the water-supply aquifers from
this source, and to allow reuse of the reclaimed wastewater.
The investigation phase included test drilling, aquifer
testing and injection testing to determine the feasibility
of the injection/extraction (I/E) concept. The number,
spacing and location of I/E doublets were optimized using a
digital computer model. The double-cased, double-screened
wells were constructed using corrosion-resistant materials
and were designed for ease of routine maintenance. In
operation, injection and extraction will be computer
controlled by sensing piezometric levels in a series of monitor
wells. Water pumped from the extraction wells will be sold
for industrial and agricultural purposes. The I/E well
system has been approved for 87J/2 percent Federal and
State grant funding.
INTRODUCTION
Among the objectives of the Santa Clara Valley
Water District is the prevention of degradation of
ground-water quality in the aquifers supplying
various municipalities and industries in their district.
Presented at The Third National Ground Water
Quality Symposium, September 15-17, 1976, Las Vegas,
Nevada.
DChief Geologist, Brown and Caldwell, Consulting
Engineers, 150 South Arroyo Parkway, Bin 83, Arroyo
Annex, Pasadena, California 91109.
The quality of ground water for these purposes is
generally good. In a few locations, however, aquifers
once used to produce good water have become
intruded with salt water. One such area is the
shallow aquifer, above 150 feet (45.7 m) depth,
around South San Francisco Bay. The aquifers
deeper than 150 feet (45.7 m) still produce good
water and are apparently protected by thick, clay
aquicludes which have prevented salt water from
the bay or from shallower aquifers from migrating
to these deeper producing zones. Heavy pumping
in the deep aquifers has produced a hydraulic
gradient in the area away from the bay toward the
producing well zones. Continued heavy pumping
from the deeper aquifers may, in time, produce a
degradation of these aquifers due to salt-water
migration from the shallower zone or from the bay.
Recognizing the need for control measures to reduce
the salt-water intrusion into the shallow aquifers, a
system has been designed which will incorporate
injection of water into the shallow aquifers by
wells in order to control the seawater intrusion.
The proposed salinity intrusion barrier
contemplates the use of a series of injection wells
and extraction wells located in pairs, termed
doublets. These doublets form two lines of injection
and extraction wells located essentially parallel to
the bay front. This arrangement will allow injection
of reclaimed water into the injection wells and the
complete removal of this water through the
extraction wells.
An investigation of ground-water conditions in
the area was prefaced by a review of the physical
environmental characteristics of the area in order
to provide a basis for evaluation of changes in the
ground-water regime.
152
-------
MOUNTAIN VIEW
Fig. 1. Study area.
Study Area
The study area is located in the northwesterly
corner of Santa Clara County, in the City of Palo
Alto, along the south shore of San Francisco Bay.
As shown in Figure 1, the area is bounded approxi-
mately by San Francisco Bay on the east, El
Camino Real on the west, the City of Mountain
View on the south and the City of Menlo Park on
the north.
Topography
The study area is composed principally of flat
lowlands and marshy baylands adjacent to San
Francisco Bay. Generally the topography
slopes northeasterly towards San Francisco Bay,
and elevations in the area range from mean sea
level at the bay front to a maximum of approxi-
mately 25 feet (7.6 m) above mean sea level in the
western portion of the study area. The portion of
the study area between Bayshore Freeway and the
San Francisco Bay coast has been largely set aside
by the City of Palo Alto for Baylands Park Preserve.
There are, however, several areas of commercial
and professional development near Embarcadero
Road and along the Freeway.
Climate
The area exhibits normal ocean-moderated
conditions. Temperatures in the area range from a
mean of 47°F (8.30°C) in January to 66° F (18.90°
C) in July with a mean annual temperature of 57° F
(13.90°C). Average annual precipitation in the study
area is approximately 15 inches (38.1 cm) although
in the higher ground to the west of the study area,
the average annual precipitation amounts to as
much as 40 inches (101.6 cm). The higher rainfall
to the west of the study area accounts for the
extensive drainage patterns present and is the source
of recharge water for the ground-water aquifers in
the area.
153
-------
Geology
The geology of the area has been studied and
described extensively in reports of the Department
of Water Resources (1967) and the reader is referred
to these publications for more detailed information
on the subject. For the purposes of this paper, the
discussion of geology is limited to the immediate
local geology in the study area and particularly in
the vicinity of the test well site, as it applies to this
investigation.
The principal geological deposits of interest to
this study are the alluvial sediments in the area.
The alluvium consists of a series of essentially
flat-lying sand and gravel aquifers separated by
extensive clay aquicludes and aquitards. The
alluvium is underlain by the Santa Clara Formation
which consists of loosely consolidated to compact,
poorly-sorted sands, gravels, silts and clays. The
total thickness of alluvium and Santa Clara Forma-
tion in the study area is in excess of 1,000 feet
(305 m) as shown by logs of wells in the vicinity.
The alluvium is overlain, in portions of the
study area, by soft, low-density clay deposits
commonly referred to as bay mud. The thickness of
the bay mud deposits is quite variable and ranges
from 0 to 20 feet (6.1 m).
Ground-Water Occurrence
The principal waterbearing deposits, and the
ones of most significance to this investigation, are
the unconsolidated sands and gravels occurring in
the alluvium. Based on the grouping of aquifers
within the alluvium, the south bay area has been
divided into various subareas which group areas of
similar waterbearing formations (DWR, 1967). The
study area includes portions of the Niles subarea
and the San Francisquito subareas. Although the
alluvium extends to approximately 1,000 feet
(305 m) depth in the area, the aquifers of particular
importance to this study occur in the upper 200
feet (61 m) of these deposits. Salt-water intrusion
has occurred in the shallow deposits but the deeper
deposits in the study area appear to contain fresh
water. For this reason, only the aquifers occurring
between ground surface and approximately 200 feet
(61 m) depth are considered in detail here.
FIELD INVESTIGATION
In order to produce detailed information con-
cerning the ground-water conditions and geology
in the area of proposed injection of water for the
seawater barrier, a program of test drilling and
aquifer test analysis was developed. The site for
initial test drilling was selected in order to provide
the most representative information for the area.
In addition, one of the goals of the drilling program
was to result in the installation of a permanent
injection well which could be utilized for initial
testing, and later for continued testing and monitor-
ing when incorporated as one of the operational
wells.
Selection of Drilling Site
The site selected for the initial test drilling
program is a portion of the City of Palo Alto
property north of and adjacent to Embarcadero
Road and east of Bayshore Freeway. The site was
convenient to a water-supply source from the City
of Palo Alto and for water disposal along
Embarcadero Road. In addition, the site provided
sufficient area in the vicinity for the installation of
the second injection/extraction well required to
complete the doublet, and the site is reasonably
close to the proposed water reclamation plant.
Examination of drillers' logs and geophysical logs
in the area indicated that this site would show
geological conditions reasonably similar to those
expected generally throughout the study area.
Test Drilling
The test drilling consisted of drilling
mud-rotary holes, 8-inch (20.3-cm) minimum
diameter, to a depth of approximately 200 feet
(61 m) and taking undisturbed samples of the soil
at various locations during drilling. In test hole-
observation well number 1 (TH/OW-1), undisturbed
samples were taken approximately every 5 feet
(1.5 m) in order to obtain the maximum amount of
information concerning the subsurface. Samples of
the materials encountered during drilling were
examined on site both visually and with a
mechanical sieve apparatus to determine grain-size
distributions, material types, consistency and other
soil characteristics. After each hole was drilled,
geophysical logging was performed in the well to
verify the written descriptive log and to obtain
additional information concerning material types
and water qualities. Locations of test holes are
shown on Figure 2.
Observation Well Construction
After each TH/OW had been drilled and
geophysically logged, and an analysis made of the
information obtained, various sand and gravel
formations were selected for installation of
piezometers for measuring water-level changes and
obtaining water-quality samples. Piezometers were
constructed of IVi-inch (3.175-cm) polyvinyl-
154
-------
RADIAL DISTANCE AND BEARING
OF TH/OW'S FROM I/E-I
INJECTION/
EXTRACTION
WELL I/E-I
Fig. 2. Test well site.
chloride (PVC) pipe with IH-inch (3.175-cm) PVC
slotted well screens. The well screens were set and
gravel-packed opposite to the particular aquifer
formations to be monitored and these zones were
sealed from other aquifer zones by cement seals.
A review of the geological information
obtained during the test drilling and geophysical
logging program indicated the presence of two
shallow aquifers, herein referred to as the "20-
foot" (6.2-m) aquifer and the "45-foot" (13.7-m)
aquifer both of whi'ch are invaded by sea water, as
evidenced by analyses of water samples. Below the
45-foot aquifer, there exists an extensive clay
aquiclude to a depth of approximately 180 feet
(54.9 m). Below the aquiclude, there occurs
another producing aquifer, herein called the
"185-foot" (56.4-m) aquifer. This aquifer contains
fresh water, and analyses of water samples showed
no evidence of seawater contamination. The
locations of the various aquifers found in the test
drilling and their lateral relationships are shown in
the geologic section in Figure 3. Slight variations
in the formations were observed between adjacent
borings; for example at TH/OW-1, there appeared
to be a merger of the 20-foot and 45-foot aquifers.
It was not determined whether this merger was due
to previous construction in the area or was a natural
occurrence. Comparison of additional subsurface
information available near the north edge of the
site indicated that the 20-foot aquifer is underlain
in all cases by a clay aquitard as shown on Figure 3.
However, south and east of the site area, additional
test borings previously performed for another
project did not disclose the presence of the 20-foot
zone. It was, therefore, inferred that the 20-foot
zone is an interrupted lense, although the fact that
it is intruded by seawater indicates that it is areally
extensive in other directions.
Some variation in the 45-foot aquifer was
demonstrated by the test drilling on site; however,
the variation is essentially in aquifer thickness and
not in material type. The aquifer is essentially a
sand and gravel zone in all three test holes. The
45-foot zone, or lenses similar to this formation,
occurs over a fairly widespread area.
The 20-foot aquifer and the 45-foot aquifer
are separated, in all three test holes, by a zone
composed of varying amounts of silt, clay and fine
sand. This zone, which is termed "aquitard" on
Figure 3, has a measurable permeability but the
155
-------
permeability is apparently much less than that of
the formations above and below it. Examination of
the areal geology confirmed that this relationship of
two aquifers separated by an aquitard occurs
extensively throughout the area.
A variation was found in the water-table
elevations among the three aquifers examined. The
185-foot aquifer exhibited flowing conditions in
the vicinity of TH/OW-1 and water levels at or
near ground surface at the other test holes. The
water levels in the shallower aquifers were 3 or 4
feet (0.9-1.2 m) below ground surface, generally,
and showed a difference between themselves of
approximately one-half foot (0.15 m).
Well Construction
After completion of the test holes, the design
data necessary for construction of the initial I/E
well was prepared. Among the items considered in
the design phase were size of the well casing to
allow ample room for anticipated pumping
equipment, selection of materials of construction
which would adequately resist normal corrosion as
well as chemicals used in future development tech-
niques, location, size and length of well screens
and slot size openings to insure negligible head
losses during injection and extraction, facility of
future maintenance, and adequate sealing of
aquifers.
Figure 4 shows the construction details of
the initial injection/extraction well, I/E-1. The
location and thickness of the cement seals used to
separate the adjacent aquifers and to seal the zone
near ground level are also shown on Figure 4, as
well as the screen lengths, screen locations and
gravel-pack dimensions.
There are various methods of well drilling
available for shallow wells of this nature. For this
particular installation, the reverse circulation
method was selected. This method of drilling
requires no drilling fluid additives, such as
bentonite, which may adversely seal the water-
producing formations. Excellent samples of the
materials encountered during drilling are also
available with this method which allow verification
of material types, and depths and thicknesses of
formations during well construction.
AQUIFER TESTING AND ANALYSES
In order to provide data necessary for the
ultimate design of an I/E system, a series of aquifer
tests were performed on each of the water-
producing aquifers at the test site. The tests were
designed to yield information concerning the
aquifer characteristics and the relationship
between aquifers in response to pumping and
injection. The following presents the methods of
testing, the analyses of the test data and a summary
of the geohydrology of the site as determined by
analyses of the aquifer-test data.
20-FOOT AQUIFER
AQUITARD
"45-FOOT" AQUIFER
Fig. 3. Geologic section A-A'
156
-------
°°° -
[fe££aj
0 Q o
°o°0
'jo^0 nc
c/cT o o o °
* . _ n n °
•ELEV. 2.86 msl
•ELEV. 2.34 msl
ELEV. 2.62 msl
/—GROUND SURFACE
6" PVC COUPLING (TYP)
6" SCH. 80 PVC
CASING (TYP)
GRAVEL-PACK, "LAPIS
NO. 4" (TYP)
6" WIRE-WOUND, PIPE
SIZE, STAINLESS STEEL
WELL SCREEN (TYP)
• FILL GRAVEL (TYP)
•CEMENT SEAL (TYP)
24" DIA. DRILLED HOLE
STEEL CENTRALIZER (TYP)
PVC PLUG-END BELL (TYP)
:^i-'—BOTTOM OF DRILLED HOLE
Fig. 4. Construction details, I/E-1.
Aquifer-Test Procedure
The aquifer testing of each screened aquifer
was performed by pumping at a constant rate
from I/E-1 for the period of approximately 24
hours, during which time changes in water level
responding to the pumping were monitored in the
pumping well, TH/OW's screened in that formation
and in formations above and below the formation
being tested. A series of water-level measurements
were made in all the wells prior to starting the
pumping portion of the test in order to establish a
trend of water levels, and after pumping stopped,
water-level monitoring was continued in all the
wells to obtain data on the recovery of water
levels after pumping.
In addition to the test wells and the pumping
well at the test well site, water levels were measured
in seven additional existing wells at various
distances and directions from the test site in order
to determine whether or not wells in these vicinities
would respond to pumping at the test site.
Aquifer Test, 185-foot Zone
During the aquifer test of the 185-foot zone,
the pumping rate in I/E-1 was held to a constant
flow of approximately 165 gallons per minute
(gpm) (10.4 1/s). Water-level changes were
monitored in the pumping well, and TH/OW-2.
In addition, water levels were monitored in the
shallow aquifers in all three test holes and the I/E
wells in the shallow aquifers. The water-level data
were reduced to values of drawdown, corrected
drawdown and recovery plotted against time since
pumping began or time since pumping ceased. The
drawdown values were corrected to account for
projected trends of water levels prior to the begin-
ning of the aquifer test. Pumping was continued
for 24 hours, after which recovery was monitored
for 24 hours. No response was noted in the shallow
aquifers to pumping in the 185-foot zone. The data
in these wells therefore were not considered in the
following analysis.
The initial examination of the test data from
the 185-foot zone indicated that the aquifer is an
artesian, anisotropic aquifer, bounded on at least
one side by a nearly impermeable boundary. The
lack of response in the shallow aquifers during
testing of this aquifer confirms the fact that it is a
confined, artesian aquifer with no hydraulic
communication to overlying zones.
Due to the confined nature of the aquifer in
the 185-foot zone, the Theis nonequilibrium
method of analysis was applied to the data (1935).
The nonequilibrium formula is shown in Equations
(1), (2) and (3).
s= 114.6 QW(u)/T
= (e"u/u)du
u
(1)
(2)
(3)
u = 2693 r2S/Tt
s = drawdown, in feet;
Q = discharge of well, in gpm;
T = transmissivity, in gpd/ft;
r = radial distance from pumping well, in feet;
157
-------
S = storativity, a decimal fraction;
t = time since pumping started, in minutes.
In applying this method of analysis to the
time-drawdown data for each observation well and
the pumping well, it was found that the calculated
coefficient of storativity increased with time. These
variations have been shown to be characteristic
of an aquifer which is bounded by one or more
impermeable boundaries (Sheahan, 1967). In order
to determine the distance to and direction of the
apparent impermeable boundary, the theory of
images (Ferris, et al, 1962) was applied to the test
data. This method, which is essentially a graphical
superposition method, and a nongraphical method
(Sheahan, 1967) of analysis for boundary conditions
were utilized in determining the boundary charac-
teristics of the aquifer.
The results of this portion of the analysis
indicated that the 185-foot aquifer has an average
transmissivity of 6,300 gallons per day per foot
(gpd/ft) (78.12 m2/d) and that there is apparently
only one impermeable boundary reflected in the
test data. The boundary was determined to be to
the east of the injection/extraction well, a distance
of approximately 900 feet (274 m).
In making calculations for the coefficient of
storativity from the time-drawdown data for
each of the wells, based on that portion of the data
which did not reflect the boundary effects, a large
variation in the magnitude of storativity was
observed. This variation in apparent storativity is
normally indicative of anisotropic conditions in
the aquifer (Papadopulos, 1965). In order to
calculate the directions of the major and minor
axes of transmissivity and the values of the major
and minor transmissivities in an anisotropic aquifer,
a minimum of three observation points are required
in addition to the pumping well. In this particular
case, only two observation wells were available and
therefore a complete analysis for anisotropy cannot
be made. However, by assuming that TH/OW-1
was on a principal axis, a calculation of an approxi-
mate range of minimum-to-maximum transmissivity
and a value for storativity can be made. Analysis
under these assumptions indicated that the trans-
missivity may range from as low as 3,000 gpd/ft
(37.2 m2/d) to as high as 12,500 gpd/ft (155 m2/d)
and that the storativity is approximately 0.002
under these conditions. Based on the nature of the
materials comprising this aquifer as shown on the
drilling logs, together with the apparent thickness
of the zone penetrated, the range of values of
transmissivity, the average transmissivity and the
apparent storativity values appear reasonable. It is
noted that, if a value of storativity S, is assumed,
calculation can be made of the other variables.
Aquifer-Test Analysis, 45-foot Zone
The test procedure employed in performing
the aquifer test of the 4 5-foot zone was similar to
that of the deeper, 185-foot zone, in that it
consisted of pumping the 45-foot aquifer of the
I/E well at a constant rate of 80 gpm (5 1/s) for a
period of 24 hours, preceded by a rest period.
During the rest period, the 24-hour pumping period
and the 24-hour recovery period following pumping,
water levels in the observation wells and pumping
wells were observed and recorded. Water levels
were monitored in both the 20-foot zone and the
45-foot zone during this test; however, the 185-foot
zone was not continuously monitored since the
aquifer test of that zone indicated it to be a
confined aquifer and not hydraulically connected
with the shallower aquifers. The water-level data
obtained during the aquifer testing was reduced to
values of drawdown and recovery which were
plotted against the time since pumping began and
time since pumping ceased, respectively. A graph
of the aquifer-test data on a full logarithmic graph
is shown in Figure 5. Due to the fact that the
recovery readings were taken for a full 24 hours,
corrections were necessary to these data to account
for projected trends of drawdown due to the pump-
ing portion of the test. These corrected recovery
readings are also shown on Figure 5.
During the pumping and recovery portions of
the test of the 45-foot aquifer, the water levels in
the 20-foot aquifer were observed to respond to
pumping in the 45-foot zone. The response of the
20-foot aquifer indicated conclusively that the
45-foot zone was obtaining recharge during the
test from the shallower, 20-foot zone, and thereby
must be considered a leaky aquifer, or more
explicitly, a semiconfined artesian aquifer.
Examination of the test data also indicated that the
aquifer had not reached steady-state conditions
and therefore must be considered in a nonequilibri-
um state for analysis.
In order to take into consideration the leaky
artesian, nonequilibrium characteristics of the test
data from this aquifer, the modified Hantush
method of analysis (1955) was employed. The
modified Hantush method utilizes a family of
type-curves plotted according to Equations (4),
(5) and (6).
s= 114.6QH(u,0)/T
(4)
158
-------
*
/
o.
~_ 1 1 1 — Mill
LEGEND
O DRAWDOWN
X RECOVERY
1 1 1 — Mill
MATPH PfMIU
1 1 1 1 — Mill
r
/ (VIM 1 ^H rWiro p
/ I/E- 1
/ u* 1.0X10-'
/ H(u.3) = IO.O
+ CORRECTED RECOVERY / t'lZ.Omin.
° 9 x 9
XX X
_
-
X
x 4^
\g x "*
X
X \
X \
x \
- X \
r \
- ° \
e MATCH POINT -^
TH/OW-Za
u = 0.01
H(u,(3) = 3.0
s -3.2ft.
t = 2.1 mln.
P- 0.003
[ 1 1 I 1 1
/ p» 0.0007
/ I/E-I (45') rw*I.O'
X0xxxxx Oxx o x x oxx
£
1 TH/OW-2a
XXX X X X t X X XXXXXXXX-
LM5'' rj^6'x xx xx x xxxx*
vXXXXXX X X|«x»
.xxxxxxxx x x x x |
TH/OW-3 (45') r'1311 ., n + x°O x x1
x° XXXX X X X * *°'
I
{Q
Z' TH/OW-I (45'
.._.CH POINT
TH/OW-3 Q
u = O.OI
H(u,|3) = I.O x
s = l.05ft. x
t = 19.5 mln.
p = 0.09 x x
„ o
x
1 1 1 1 1 1 1 1
„ x 0 XO X * "
XX. xo
+o
+ O*" x
xxx xx
{~f& X
0^ X X
) r = 496' o' x
X
" BEST FIT
x TO ALL DATA
0 80 gpm
u 1.0
H(u,p) 1.0
s 1.05ft.
t/r2 I.I XlO-'mln./ft.2
P/r 7.0X10-''
T 8700gpd/ft.
S 3.6X10-'
1 I 1 1 1 1 1 1
1 1 1 1 1 1 !_
—
_
x+ + x
^-x03 t x -
+ '-
Ox x x
-
*+# + +
X X
xx x x
1
<)
/r
/
/
t— MATCH POINT
TH/OW-I
u = 0.001
H(u,p) = I.O
8=1.05 ft.
t =2700 mln.
[3 = 0.35
1 1 1 1 1 1
Fig. 5. Aquifer test, 45-foot zone.
100
TIME I, MINUTES
I,OOO
10,000
H(u,0) = /
u
erfc [0 Vu/Vy(y-u)] dy (5)
S' r
S"
where
K',K" = hydraulic conductivities of semipervious
confining layers, in gpd/sq ft;
b', b" = thickness of the semipervious confining
beds, in feet;
S', S" = storativities of the semipervious confining
beds, as decimal fractions;
y = constant of integration;
and s, Q, T, r, S, and t are as previously defined.
The analysis of data by this method consists of
attempting to match one of the family of type-
curves to the data as shown on Figure 5 using a
superposition procedure. When a position of
best-fit is obtained, values of the parameters are
read from the superimposed graph at a convenient
match-point and the storativity, transmissivity,
and leakance characteristics of the aquifer are
calculated. In the particular case under analysis, it
is obvious that there is leakage occurring only from
above the 45-foot aquifer. This is substantiated by
the fact that the 185-foot aquifer exhibited no
hydraulic connection to the shallower aquifers,
and the fact that there are no water-producing
zones of any significant continuity occurring
between the 45-foot aquifer and the 185-foot
aquifer.
Concerning other physical limitations in this
aquifer such as impermeable boundaries or
anisotropy, the fact that excellent match was
possible with the type-curves to each of the data
plots indicates that there are no boundary
conditions of any significance affecting the test
data. An examination of the relationship between
drawdown and distance from the pumping well
shows that the drawdown varies essentially linearly
with the logarithm of distance indicating that the
aquifer is isotropic.
Although the data from each TH/OW may be
analyzed individually to determine a unique value
for each of the aquifer coefficients, it is of greater
159
-------
importance to evaluate all of the test data from all
the observation points together, in order to obtain
those values of the coefficients which are most
representative of the entire aquifer. In order to
perform this type of analysis, the data shown on
Figure 5 were replotted as a graph of drawdown, s,
and recovery, s', versus a value of time divided by
the square of the radius from the pumping well,
t/r2. The superposition method of curve-matching
was applied to this graph using a family of type-
curves based on Equations(4), (5) and (6),and a
position of best-fit was determined. From this
analysis, the match-points corresponding to each
set of test data were determined and are shown
on Figure 5.
The composite aquifer-test data analysis
yielded the value of 8,700 gpd/ft (108 m2/d) for
transmissivity, and a coefficient of storativity of
0.000036. These values appear to be reasonable
based on the general geology and the relative
geometry of the aquifer. Using an average thickness
of 13 feet (4 m) for the overlying semipervious
confining layer, as determined from the logs of the
test holes and a value of /3/r equal to 0.0007, as
determined from the analysis, the storativity and
vertical hydraulic conductivity of the semipervious
confining layer were calculated. These calculations
yielded a storativity of 0.001 and a value of vertical
hydraulic conductivity of 0.032 gallons per day per
square foot (gpd/sq ft) (0.013 m/d).
During the analysis of this set of test data, it
was observed that the rate of change in water level
in the 20-foot aquifer at I/E-1, TH/OW-1, and
TH/OW-2a, with respect to the logarithm of time,
were all essentially equal. The rate of change, or
slope per log cycle time, of the data from TH/OW-3,
however, was approximately twice the slope of the
data plot from the other wells. Examination of
borings made immediately south and east of TH/
OW-3 indicate that the shallow, 20-foot aquifer,
appears to be discontinuous in that direction.
Therefore the increase in slope observed from
TH/OW-3 may be a response to an impermeable
boundary in the 20-foot aquifer. The leakance
characteristics in this vicinity may also be different.
This aspect of that aquifer is examined further in
the analysis of test data from the 20-foot aquifer.
Aquifer-Test Analysis, 20-foot Zone
Due to the limited depth available for
drawdown in the shallow aquifer, a very low
constant pumping rate was used during this test.
The I/E well in the 20-foot zone was pumped at
an average pumping rate of 8.6 gpm (0.54 1/s) for
a period of 24 hours after which the pump was
shut down and recovery readings were taken for
17 hours following pumping. During the test,
changes in water level were monitored in both the
20-foot aquifer and the 45-foot aquifer in all three
TH/OW's and I/E-1. The small magnitude of
drawdown in the wells produced a data graph
which was considered less than satisfactory for
proper analysis. However, the recovery data from
the test was able to be monitored more
exactly than the drawdown data and did produce
usable data. Consequently, the water-level
readings taken during the recovery period were
reduced to a plot of recovery versus time.
The 20-foot aquifer essentially reached
equilibrium prior to the end of the test period.
The aquifer reached equilibrium due to the leakage
occurring from the 45-foot aquifer below, and
probably additional leakage from the overburden
materials above the 20-foot zone. The short time
period required for this aquifer to reach steady state
is due partly to the low value of pumping rate
used during the test, and also due to the apparent
additional leakage from the overlying beds.
Since the aquifer reached equilibrium during the
test period, the leaky artesian formula of Hantush
and Jacob (1955), which considers an aquifer
approaching equilibrium was used in the data
analysis. The leaky artesian formula is shown in
Equations (7), (8) and (9).
s = (114.6 Q/T)W(u,r/B)
(7)
W(u, r/B) = J (1/u) exp (- u - r2/4 B2u) du (8)
u
B = VTb'/K' (9)
where all variables shown are as previously defined.
For this analysis, the data were replotted as
recovery, s', versus a value of time divided by the
square of the distance from the pumping well, t/r2,
and a family of type-curves, based on Equations
(7), (8) and (9), was superimposed to a position of
best-fit to the data. Utilizing the portion of the
data prior to the point at which steady state was
reached, values of transmissivity and storativity
were determined. The transmissivity was found to
be 3,500 gpd/ft (43.4 m2/d) and the storativity
value determined was 0.002.
Since the leakage to the 20-foot aquifer
apparently occurs from both above and below that
aquifer, and since the analysis method used
assumes leakage from only one bed, it is not
possible to determine exact values of the vertical
160
-------
hydraulic conductivity for the adjacent beds.
However, the purpose of obtaining information
concerning the aquifer characteristics is to allow
prediction of future response to pumping or
injection stresses on these aquifers, and a more
usable parameter for this purpose is the range in
values of B. From type-curve analysis, the value
of B was determined to be in the range of 3 3 to
131 and averaged about 105.
Since, as indicated above, the latter part of
the test data represents essentially steady-state
conditions, it is possible to verify the value of
transmissivity using a different method of analysis.
The steady-state leaky artesian formula (Jacob,
1946) is shown in Equation (10).
s = (229 Q/T) K0 (r/B)
(10)
where
K0 (r/B) = modified Bessel function of the
second kind and zero order;
and s, Q, T, r and B are as previously defined.
The values of r/B obtained from the nonsteady-
state analysis were plotted versus the steady-state,
equilibrium recovery for each of the observation
wells. To this data was matched a type-curve of
the steady-state leaky artesian aquifer conditions.
Analysis by this method yielded a value trans-
missivity of 3,650 gpd/ft (45.3 m2/d) which
compares favorably with the value of 3,500 gpd/ft
(43.3 nWd) obtained from the nonsteady-state
method of analysis.
Summary of Geohydrology of Test Site
The subsurface consists of essentially three
separate aquifers within the depth of investigation.
The deepest aquifer, the 185-foot aquifer, is a fully
confined, artesian aquifer with an average trans-
missivity of 6,300 gpd/ft (78.1 m2/d) and a
storativity of 0.002, but is anisotropic in character
and may have directional transmissivities ranging as
low as 3,000 gpd/ft (37.2 m2/d) to as high as
12,500 gpd/ft (155 m2/d). The aquifer appears to be
bounded by an impermeable hydraulic boundary
located at a distance of approximately 900 feet
(274 m) east of the I/E well. There was no hydraulic
continuity observed between the 185-foot aquifer
and the two shallower aquifers.
The shallow aquifers, consisting of a 20-foot
aquifer and a 45-foot aquifer, exhibit a certain
degree of hydraulic communication. Pumping of
the 45-foot aquifer created a water-level response
in the 20-foot aquifer. The 45-foot aquifer showed
characteristics of a leaky artesian aquifer upon
analysis. Although the pumping rate used in the
20-foot aquifer test was too small to allow any
observable correlative response in the 45-foot
aquifer, the fact that there is hydraulic communica-
tion between the two was established during the
test of the 45-foot aquifer and verifies the apparent
correlative response to pumping of the 20-foot
aquifer.
Considering the fact that the depth to water
level in both shallow aquifers is about the same,
roughly 4 feet (1.2 m) below ground surface, it
was anticipated that injection into the 45-foot
aquifer at pressures greater than about 5 feet
(1.5 m) of head, might eventually cause water-
logging of the surface of the ground due to leakage
from the 45-foot aquifer, through the 20-foot
aquifer, and up into the overburden material
above this zone. The 20-foot aquifer, however,
provides a means for controlling the waterlogging
effect that might possibly occur. While injecting
into the 45-foot aquifer at a fairly high rate, the
20-foot aquifer can be pumped at a much smaller
rate, but a rate sufficient to control waterlogging
in the area.
Injection Testing
After the completion of aquifer testing of
the three aquifers encountered in the test well site,
actual injection testing was performed on the 45-
foot aquifer. The purposes for which the injection
testing was performed were actually threefold.
First, although the analysis of the aquifer-test data
provided the information necessary to predict the
results of injection, it is valuable to be able to
confirm the accuracy of these predictive tech-
niques. Second, injection testing produces addi-
tional, usable information not obtained during the
aquifer testing which was of value in design and
operation of the injection system. Third, and
possibly most important, it is valuable to
demonstrate that the proposed injection procedure
will perform as expected under actual field
conditions, and with complete control of the
operation.
In designing the injection test, the first
criteria of importance was a value of maximum
injection pressure to be used during
injection. In order to arrive at this maximum figure,
the criterion was established that the injection
pressure should be no greater than the net loading
of soil at the point of injection. In order to calcu-
late this, the static loading of soil down to the top
of the aquifer was determined to be 35.3 pounds
per square inch (psi) (2.48 kgs/cm2). Due to the
161
-------
artesian nature of the 45-foot aquifer, the hydro-
static pressure in the formation is 15.6 psi (1.10
kgs/cm2). The net static loading, therefore, is the
difference between these two values, or approxi-
mately 20 psi (1.4 kgs/cm2), at the design point.
Based on the aquifer-test coefficients available,
this value corresponds to an injection pumping
rate of approximately 200 gpm (12.6 1/s).
In order to obtain additional data concerning
the possibilities of waterlogging in the area,
additional shallow borings were made in the vicinity
of the test holes on site and were equipped with
piezometers to a depth of approximately 3 to 5 feet
(0.9 to 1.5 m). These piezometers were also
monitored during the injection testing. In addition,
concrete monuments were set in the soil at various
locations around the test well site, and accurate
elevations were determined at these monuments.
The elevations of the monuments were monitored
during the injection testing to determine whether
or not heaving of the upper layer might be occur-
ring due to the injection pressures.
Due to limitations in the availability of
injection pressures at the site, the injection test was
actually performed at an injection pressure of
approximately 12 feet (3.66 m) and at a pumping
rate between 90 and 100 gpm (5.7 and 6.3 1/s).
The test was conducted using a constant-head,
variable-capacity method which varied slightly from
the aquifer test in that the aquifer tests were
performed with a constant pumping rate and a
variable head. The water used was fresh water from
the City of Palo Alto municipal water supply, and
was injected into the 45-foot aquifer at I/E-1. The
test was continued for 48 hours under these condi-
tions, and changes in water levels were monitored
in the I/E well and the TH/OW wells in both the
20-foot and 45-foot aquifers. At the end of the
48-hour period, both the injection and the pumping
operations ceased and recovery readings were
obtained.
Examination of the test data at the end of
24 hours of injection showed that the water level in
the 20-foot aquifer at TH/OW-3 had risen to a point
approximately 6 inches below ground surface. By
contrast, the water level in the 20-foot zone at
TH/OW-2a, 85 feet (26 m) closer to the injection
well, was still nearly 2 feet (0.61 m) below ground
surface. This rapid rise at TH/OW-3, southeast of
the injection well, correlated with what was
expected in that area based on the aquifer-test
analysis of the 45-foot aquifer which indicated the
possibility of an impermeable barrier in the 20-foot
aquifer to the southeast of the site. At that point
in the test, the shallow, 3 to 5-foot (0.9 to 1.5-m)
piezometers still did not show any waterlogging of
the overburden materials.
After 24 hours of injection alone, and at the
point at which the water level in the 20-foot aquifer
in the vicinity of TH/OW-3 was 6 inches (15.2 cm)
below ground surface, pumping was started in the
I/E well at the 20-foot depth at a rate of approxi-
mately 8 gpm (0.5 1/s). This pumping continued
for the next 24 hours, during which time no
additional rise in the water level in the shallow
aquifer at TH/OW-3 was observed. In fact, the
water level actually receded approximately 0.15
feet (0.05 m) during that time. The shallow
piezometers showed no response to injection during
the entire 48 hours of injection, and there was no
change in elevations of the concrete monuments
observed during testing.
The injection test therefore demonstrated
that fresh water can be injected into the 45-foot
aquifer with no detrimental effects and that
possibilities of waterlogging can be adequately
controlled by pumping the shallow aquifer,
thereby providing complete operational control of
the injection system.
Prediction of Injection Response
As a result of the injection testing, the actual
response of the 20-foot and 45-foot aquifers to
injection of fresh water under the test conditions
was known. The next step was to verify that this
response could be accurately predicted based on
the knowledge of the aquifer characteristics
determined by the aquifer testing.
In order to calculate the expected response of
injecting at a constant head, a constant pumping
rate was selected based on the range of injection
rates used during the injection test and a variable-
head, constant-rate method, similar to that used in
analyzing the aquifer-test data for the 45-foot
aquifer, was employed, since under the test condi-
tions at I/E-1, very little change in pumping rate
was observed. The predicted response and the
actual response of injection for two of the test
holes in the 45-foot aquifer are shown on Figure 6.
Similar data for the 20-foot aquifer for the same
two test holes are shown plotted on Figure 7. As
can be seen from the Figures, there is an excellent
correlation of predicted response with actual
response for the 45-foot aquifer, and for the
nearest test hole in the 20-foot aquifer. The low
magnitude of water-level rise in the 20-foot aquifer
at TH/OW-1, located 496 feet (151m) from the
I/E well, may account for the slight deviation of
162
-------
TIME f, MINUTES
0
1.0
2.0
SO
60
7.0
'0 IOO I,OOO 4f
^~~~~~~:
o
0
^\ o
X
^
PREDICTION OF RESPC
OF 45' AQUIFIER TO \t
• AT 97gpm TO ACTUAL
FROM TEST.
~^\S TH/OW-I (45')
x,o
o~o
LEGEND
O INJECTION Tl
X PREDICTED F
^0
"^x^ TH/OW-3 (4!
NSE °°0
ilJECTION °
. RESPONSE
°~o.
1ST DATA
ESPONSE
)')
£""0*^^
i
TIME f, MINUTES
0
I.O
\~,
\
g 20
Uj
-j
30
10 IOO I.OOO 4.C
X
\o
0
o
v 0
PREDICTION OF RESPON
OF 20' AQUIFER TO INJ
AT97gpm TO ACTUAL
FROM TEST.
LEGEND
0 IN.IRVrinN TF
X PREDICTED R
C\ u
V
>E X\p
ACTION \°
RESPONSE \Q
°\
TH/OW-3 C
TH/OW-I (2d)
ST DATA
:SPONSE
°o
^:
i i
Fig. 6. Response of 45-foot zone to injection into 45-foot
zone.
Fig. 7. Response of 20-foot zone to injection into 45-foot
zone.
actual response to predicted response observed for
that well. It is noted, however, that the predicted
response is conservative in that it shows a greater
rise of water level than actually occurred.
In summary, therefore, it can be said that the
aquifer characteristics as determined by aquifer
testing are suitable for accurate prediction of
injection response both in the injecting aquifer
and in the aquifer into which leakage is occurring.
Even the effects of boundaries in the aquifer being
recharged can be taken into account in making these
predictions.
Water Quality
Throughout the course of the field investiga-
tion, water samples were obtained from the various
monitoring wells for determination of chlorides
and specific conductance. A summary of the
water-quality data is shown graphically on Figure 8.
As shown by the Figure, there is a general
correlation between specific conductance of the
water samples obtained and the amount of
chlorides present in the samples. There is also an
apparent grouping of samples which correlates
with the location of the wells and the depth of the
aquifers from which the samples were taken. The
lowest chloride contents were found in the 185-foot
aquifer at the test site. Chlorides in these samples
averaged 28 milligrams per liter (mg/1). By contrast,
the amount of chlorides in the shallow domestic
wells, located approximately one-half to one mile
farther away from the Bay than the test well site,
showed an average chloride content of approximate-
ly 110 mg/1.
COMPUTER SIMULATION
The proposed operation of the injection
system consists of injecting reclaimed water into
a series of wells in order to hold back seawater
intrusion, and to extract this water from a series
of extraction wells in order to keep the reclaimed
water from migrating out of the area of the
injection wells and into the vicinity of producing
water-supply wells. Given an injection/extraction
doublet located in an aquifer whose characteristics
are known, under a regional hydraulic gradient
greater than zero, there is a limiting streamline,
the path of which can be determined, which will
: 1 1 — i — i i i IT
! /-
I-'
/
\
(9\
>*^
^-185' ZONE
l I I 1 L 1 1 1
1— i — I — n I I I
-SHALLOW WELLS
WEST OF BAYSHORE
FREEWAY
\
1
i
1 ; — TT-I 1 1 L
45' ZONE-}
^20' ZONE
A= BEFORE
INJECTION
B- DURING
INJECTION
, i i i I i i
IOO IfOO
CHLORIDE ION, ma/1
Fig. 8. Relationship of chloride to specific conductance.
163
-------
EXTRACTION
WELL
INJECTION
WELL
RECLAIMED WATER
Fig. 9. Diagrammatic illustration of operation of injection/extraction doublet.
flow from the injection well to the extraction well,
encompassing an area surrounding the doublet.
Under the given circumstances, no particle of
water can move outside of the perimeter of the area
enscribed by the path of the limiting streamline. In
addition, all other particles of water from the
injection well will follow streamlines within that
perimeter between the injection well and the
extraction wells (D'Acosta and Bennett, 1960). A
diagrammatic section illustrating this condition is
shown in Figure 9.
Under simple, straightforward conditions, the
path of the limiting streamline can be determined
analytically. However, with conditions of multiple
I/E doublets, complex aquifer conditions, and
interaction between aquifers, the calculation of the
path of the limiting streamline becomes a task more
suitable to the use of a digital computer. For this
reason, a computer simulation approach (Ramey,
et al., 1973) was used to determine the path of
the limiting streamline under the conditions
particularly applicable to the study area.
Aquifer Characteristics
The computer model was designed to include
the aquifer characteristics from the 45-foot zone.
These aquifer characteristics include the following
values:
T= 8,700 gpd/ft (108 mVd);
S = 0.000036;
j3/r - 0.0007;
Q = varies;
t = very large, to approach equilibrium;
r = varies.
The digital computer model essentially is a mathe-
164
matical model applying Equations (4), (5) and (6),
incorporating the above aquifer characteristics as
determined from the aquifer testing.
In order to determine a value for the hydraulic
gradient through the area which would be repre-
sentative of the actual field conditions, the hydraulic
gradients in various portions of the study area were
calculated. These calculations were based on
water-level elevations in various wells and the
gradients varied from nearly zero to as much as
0.002 ft/ft (feet per foot).
Preliminary computer simulations indicated
that the area encompassed by the limiting
streamline was quite sensitive to variations in
hydraulic gradients. It was determined that as the
hydraulic gradient became greater, the area
encompassed by the limiting streamline became
smaller. Therefore, in order to present a conservative
approach, the largest gradient measured was
utilized in preparing the simulation model.
Injection/Extraction Doublet
The digital computer simulation model was
first applied to the problem of determining the
effects of spacing between the injection well and
the extraction well, and variations in pumping
rates at the wells. Since it was anticipated that the
injection rate and extraction rate will be equal in
actual operation, this relationship was also used
in the simulation. Figure 10 shows the path of a
particle of water along the limiting streamline in a
single doublet oriented as shown with respect to
the normal hydraulic gradient. In this instance, the
limiting streamline traces the approximate shape of
an ellipse. By applying the simulation model to
various well spacings and pumping rates, the
relative relationships among doublet spacing,
pumping rate and area encompassed by the limiting
-------
streamline were determined. It was found that at a
constant spacing, the area is proportional to the
rate of pumping, such that if the pumping rate is
doubled, the area encompassed by the limiting
streamline is also essentially doubled. It was
further determined that the area is roughly
proportional to the product of the rate times the
spacing. For example, the area encompassed by
the limiting streamline for the case of injection and
extraction wells spaced 1,000 feet (304.8 m) apart,
pumping at 50 gpm (3.15 1/s), is the same as the
area resulting from wells spaced 500 feet (152.4 m)
apart pumping at 100 gpm (6.31 1/s). It was also
shown, as mentioned previously, that the area is
inversely proportional to the regional ground-water
hydraulic gradient through the area.
The path of the limiting streamline becomes
somewhat more complicated when multiple
injection/extraction (I/E) doublets are considered.
Obviously, if three doublets are spaced side-by-side,
they will have the same effect on the limiting
streamline as would a single doublet pumping at
three times the pumping rate. On the other hand,
if the doublets are spaced great distances apart,
each doublet will operate as if the other doublets
do not exist. However, in the intermediate zone
between these two extremes, there is a particular
spacing of doublets at which the path of the
limiting streamline from the center doublet changes
from a path completely encompassing all three
doublets to a path that encompasses only the central
doublet. This effect of doublet spacing on the path
of the limiting streamline is illustrated in Figure 11.
In Example A on this Figure, the doublets are
relatively close and the path of the limiting stream-
line encompassing all three doublets follows
essentially the same path as if the system con-
EXISTING HYDRAULIC
GRADIENT 0.003 ft./ft.
sisted of only a single doublet pumping at three
times the rate of each of the doublets shown.
Example B shows three doublets spaced somewhat
farther apart. The difference in shape of the area
encompassed by the path of the limiting streamline
in this example, as compared with Example A,
shows the effects of the increased spacing. The
effect of the spreading out of the doublets, or
EXISTING HYDRAULIC
GRADIENT 0.002 ft./ft.
(TYPICAL)
EXAMPLE A
SPACING 500 FEET
EXAMPLE B
SPACING 2000 FEET
EXAMPLE C
SPACING 2500 FEET
Fig. 10. Limiting streamline, one doublet.
Fig. 11. Effect of doublet spacing on limiting streamline.
165
-------
increase in side-to-side spacing, causes the limiting
streamline path to approach the point at which it
will change direction and pass between adjacent
doublets. Example C shows the doublets located at
a spacing just greater than the critical spacing at
which the streamlines have, in fact, broken through
or changed direction, and each streamline path is
encompassing an area around only its own doublet.
It is noted that in the case of Example C in
Figure 11, there is, in addition to the limiting
streamline shown, another streamline resulting from
the pre-existing hydraulic gradient which is also
passing between the doublets. If this existing
hydraulic gradient is bringing salt water from a
source such as the Bay toward the line of I/E
doublets, the doublets spaced as shown in
Example C are not effectively holding back
seawater intrusion. There is, obviously, for a given
pumping rate and well spacing, a critical doublet
spacing, somewhere between that shown in
Example B and Example C, such that if the
doublets are spaced less than that critical spacing,
the line of I/E doublets will do an effective job as
a barrier to seawater intrusion. It was therefore
necessary to determine the critical spacing require-
ments for the location of I/E doublets, and the
relationship of the critical spacing to either the I/E
pumping rates, the number of I/E doublets, or both.
As stated above, the area encompassed by the
path of the limiting streamline is directly propor-
tional to the pumping rate and to the spacing
between wells. Therefore, it is certain that the
critical spacing between the doublets, depending
as it does on the path of the streamline, is also
dependent upon the spacing between the injection
and the extraction well in each doublet. From an
evaluation of the physical constraints of the study
area and the geohydrologic parameters, the range
of separation between the injection well and the
extraction well in a given doublet was determined
to be between 500 and 1,500 feet (152.4 and
457.2 m). For the purposes of this analysis, this
dimension was assumed to be 1,000 feet (304.8
m). In addition, the aquifer characteristics and
maximum injection pressure indicated that the
pumping rates for both injection and extraction
wells are in the range of 100 gpm (6.31 1/s) to
200 gpm (12.6 1/s). The pumping rate value of
150 gpm (9.5 1/s) was assumed as a reasonable
value for this analysis.
To determine the relationship of the critical
spacing between I/E doublets to both the pumping
rate and the number of doublets, the computer
simulation technique was first applied to a three-
doublet system to calculate the critical doublet
spacing for various pumping rates and various well
spacings within each doublet. The results of this
analysis indicated that, for a particular well
spacing in a three-doublet system, as the pumping
rate in both the injection and the extraction well
increased, the critical spacing between doublets
approached a linear relationship with the pumping
rate for any given spacing between wells in the
doublet. With this information in hand, it was
therefore possible to predict the critical doublet
spacing and therefore the maximum design spacing
for a three-doublet system.
Since the physical constraints of aquifer
characteristics and site parameters limited the
assumed distance between the injection well and
the extraction well in a doublet to approximately
1,000 feet (304.8 m) and to pumping rates of
approximately 150 gpm (9.5 1/s), a three-doublet
system was not sufficiently broad to prevent
seawater intrusion through the study area.
Therefore, it was necessary to go to a greater
number of doublets. The additional I/E doublets
raised the question of whether or not the critical
doublet spacing will change with the addition of
doublets. The computer simulation model was
utilized to determine the relationship between
critical doublet spacing and number of doublets at
any particular pumping rate. The analysis indicated
that there is a reduction in critical doublet spacing
with the increased number of doublets.
Evaluation of the computer simulation data
showed that the critical spacing between doublets
is inversely proportional to an exponential of the
number of doublets. For the particular case
assumed, that of wells separated a distance of
1,000 feet (304.8 m) in each doublet, pumping at
150 gpm (9.5 1/s), it was determined that, for
nine doublets, the critical spacing between doublets
is approximately 1,800 feet (585 m) and for seven
doublets, the critical spacing is approximately
2,000 feet (610 m). Since the actual design spacing
between doublets must be less than the critical
spacing, and in order that the system be
conservatively designed to accommodate any
variation which may exist in aquifer characteristics
or in the existing hydraulic gradient, a design
spacing between doublets was established at 1,000
feet (305 m).
Recommended Location of Injection/Extraction
Doublets
Taking into account the physical and cultural
characteristics of the study area, together with the
166
-------
design criteria established above for location of the
doublet system, the recommended locations of
I/E doublets were determined. The system requires
a total of nine I/E doublets, the total capacity of
which would be approximately 2.0 million gallons
per day (mgd) (7.6 X 106 1/d). Because of the
conservative approach taken in design of the
system, a certain amount of latitude was available
for use in final location and construction of the
doublets.
DESIGN AND CONSTRUCTION
Following the analysis of the computer
simulation, the seawater intrusion barrier system
was designed. The system includes a total of nine
doublets (18 wells), as shown on Figure 12, and 37
monitor wells. Eight existing monitor wells were
incorporated into the system together with the
initial injection/extraction well (test well).
Monitor Wells
Three types of monitor wells were designed
for the injection/extraction well system; shallow
wells ("S" series), mid-depth wells ("M" series)
and deep wells ("D" series). Monitor wells were
designed and located to serve as monitoring points
for ultimate operation of the system and as test
holes to obtain pertinent information on formation
thicknesses and material types for design of the
injection/extraction wells.
The "D" series wells were planned to monitor
water levels and water quality in the 185-foot
aquifer and to provide a supply of fresh water for
future possible injection purposes. Each well is
designed with 8-inch (20.32 cm) schedule 80
polyvinyl chloride (PVC) casing, type 304 stainless
steel wire-wound type well screen, gravel-packed in
an 18-inch (45.77 cm) drilled hole. The annulus
from ground surface to a point 5 feet (1.5 m) above
the top of the screen is sealed with neat cement
grout to protect the lower aquifer from possible
contamination by water from the shallow zone.
The "S" series wells were designed to
penetrate the 20-foot and 45-foot aquifers. The
six "M" series wells were designed similarly to the
"S" wells with the exception that the initial drilled
hole was planned for an approximate depth of
100 feet (30.5 m) for the "M" wells, while the
"S" wells were scheduled to be only about 60 feet
(18.3 m) total depth.
Each well in the "S" and "M" series is
double-cased and double-screened with a neat
cement grout seal separating the 20-foot and
45-foot aquifers. Well design includes 3-inch
(7.62 cm) PVC casing and PVC wire-wound-type
well screen, gravel-packed in a 12-inch (30.5-cm)
drilled hole.
Both cuttings samples and split-spoon samples
were specified to be obtained during drilling and
each hole was to be geophysically logged and
calipered to provide additional information for
correlation with the geology of the area and for
design of the injection/extraction wells.
Injection/Extraction Wells
In order to provide the facility for future use
of each of the injection/extraction (I/E) wells for
either injection or extraction, each of these wells
were designed identically. These wells also are
double-cased, double-screened, with neat cement
grout separating the 20-foot and 45-foot aquifers.
The casing and screen for the 45-foot zone were
designed to be 8-inch (20.32 cm) nominal size to
allow installation of pumping or injection equip-
ment. Since the 20-foot zone will be pumped only
to control waterlogging, the casing and screen for
this zone is 6-inch (15.24 cm).
All casing is Schedule 80 PVC and the screens
are type 304 stainless steel, wire-wound, keystone-
type screens. Each screen is gravel-packed in a
24-inch (61 cm) drilled hole.
Funding
Since the injection/extraction well system is
part of the final treatment and disposal process of
the regional wastewater treatment plant supplying
the reclaimed wastewater for injection, application
was made to the Environmental Protection Agency
(EPA) and the California State Water Resources
Control Board (SWRCB) for grant funding under
P.L. 92-500. After review of the predesign report
prepared by the Santa Clara Valley Water District
and the report of Geohydrologic Investigation,
both EPA and SWRCB approved the project for
grant funding. Under this program, EPA provides
75% funding of grant eligible costs while SWRCB
provides an additional 12Vi%.
The bid price of the injection/extraction well
system was $400,000. Contracts were signed in
late 1975 and construction began in January 1976.
Construction
The project was scheduled for completion by
the end of August 1976. However, a three-month
delay due to right-of-entry restrictions occurred in
late spring and early summer. Consequently, at the
time of preparation of this paper, only the monitor
wells have been constructed.
167
-------
EGIONAL WASTEWATER
TREATMENT WORKS AND
WATER RECLAMATION PLANT
EXTRACTION WE L TYP
S.7T
TOM'S WELL"
NJECTION WEL
Fig. 12. Injection/extraction well locations for seawater intrusion barrier.
168
-------
It is anticipated that after completion of
construction, a supplemental paper will be prepared
describing the construction and operation
characteristics of the system.
ACKNOWLEDGMENTS
Assistance during the course of this study was
provided by Mr. William M. Roman, Water Resources
Engineer, Santa Clara Valley Water District; Mr.
Thomas I. Iwamura, Geologist, Santa Clara Valley
Water District; Mr. Ray Remmel, Chief Engineer,
Utilities, Water Gas and Sewer, City of Palo Alto;
Mr. Perry Wood, Hydrologist, U.S. Geological
Survey, Menlo Park, California; Dr. Mohinder S.
Gulati, Union Oil Company, Santa Ana, California;
and Dr. Henry J. Ramey, Jr., Professor and Acting
Chairman, Department of Petroleum Engineering,
Stanford University. Mr. Mohamed Soliman and
Mr. Syed Tariq, both students in the Graduate
Program, Department of Petroleum Engineering,
Stanford University, assisted in the test drilling
and aquifer-test portions of the field investigation.
REFERENCES
D'Acosta, J. A., and R. R. Bennett. 1960. The pattern of
flow in the vicinity of a recharging and discharging
pair of wells in an aquifer having areal parallel flow.
International Assoc. of Hydrologists, Pub. No. 52,
pp. 524-536.
Department of Water Resources, State of California. 1967.
Evaluation of groundwater resources, South Bay,
Appendix A: geology. August.
Ferris, J. G., D. B. Knowles, R. H. Brown, and R. W.
Stallman. 1962. Theory of aquifer tests. U.S.
Geological Survey Water-Supply Paper 1536-E,
pp. 144-166.
Hantush, M. S., and C. E. Jacob. 1955. Non-steady radial
flow in an infinite leaky aquifer. Am. Geophys. Union
Trans, v. 36(1).
Hantush, M. S. 1960. Modification of the theory of leaky
aquifers. Jour. Geophys. Research, v. 65, no. 11,
November.
Jacob, C. E. 1946. Radial flow in a leaky artesian aquifer.
Am. Geophys. Union Trans, v. 27, no. 2, pp. 198-
205.
Papadopulos, I. S. 1965. Nonsteady flow to a well in an
infinite anisotropic aquifer. Symposium on Hydrology
of Fractured Rocks, UNESCO, Dubrovnik, Yugo-
slavia. Paper no. 25, August.
Ramey, H. J., Jr., A. Kumar, and M. S. Gulati. 1973.
Gas well test analysis under water drive conditions.
Monograph Project 61-51, Pipeline Research Com-
mittee, American Gas Association, Arlington,
Virginia.
Sheahan, N. T. 1967. A non-graphical method of
determining u and W(u). Ground Water, v. 5, no. 2,
pp. 31-35.
Theis, C. V. 1935. Relation between the lowering of the
piezometric surface and the rate and duration of
discharge of a well using groundwater storage. Am.
Geophys. Union Trans. Part 2, pp. 519-524.
DISCUSSION
The following questions were answered by N. Thomas
Sheahan after delivering his talk entitled "Injection/
Extraction Well System—A Unique Seawater Intrusion
Barrier."
Q. by Peter Perez. Would you mention the R & D program
to be conducted by Stanford University for transmission of
viruses and organics between the injection and extraction
wells?
A. As a long-term goal, the proposed injection/extraction
facility will be used for research to determine the feasibility
of such a system for reclaiming water for potable uses. A
research program has been established by Stanford
University, supported by U.S. Environmental Protection
Agency research grant EPA-R-804431, in order to answer
some of the significant questions as necessary to realize
this long-term goal. The major objectives of this research
study are as follows:
1. To determine the effects the injected wastewater
will have on the chemical, physical and biological quality
of the basin and injected waters.
2. To determine the effect injected wastewater will
have on the hydrologic characteristics of the aquifer.
3. To seek the optimum quality for injected water
which will result in a high-quality basin water and minimum
damage to the hydrologic characteristics of the aquifer.
4. To develop generalized mathematical models for
describing the movement of water, the changes in hydrologic
characteristics, and resulting changes in water quality from
wastewater injection in order to make the results of most
value for application in other similar projects.
Along these lines, technical report No. 206 dated
April 1976 entitled "Preproject Water Quality Evaluation
for the Palo Alto Water Reclamation Facility" and the
first quarterly progress report dated August 5, 1976
entitled "Groundwater Injection of Reclaimed Water in
Palo Alto" have been prepared by the Department of
Civil Engineering, Stanford University.
Q. by Jon O. Nowlin. Was consideration given to the effect
on the hydrologic characteristics of the aquitard of using
lower-quality effluent as the recharge fluid?
A. The aquitard allows leakage from the 45-foot zone, into
which injection is to take place, to the shallower 20-foot
zone. The use of reclaimed wastewater or treatment plant
effluent as the recharge fluid may have one of two effects
on the aquitard; it may either decrease its permeability or
increase its permeability. A third alternative would be no
effect at all. Since the project is designed to allow leakage
through the aquitard and to control such leakage by
pumpage from the 20-foot aquifer, an increase in
169
-------
permeability of the aquitard would have negligible effect on
the operation of the project. If, on the other hand, use of
reclaimed wastewater caused a decrease in permeability in
the aquitard, a reduction in leakage would occur requiring
less pumpage from the 20-foot aquifer.
Q. by Stanley N. Davis. I question the postulated mechanism
of sea-water intrusion. Are you aware of the large number of
abandoned wells including the old Spring Valley Water
Company test wells? These are near the salt marsh edge of
the bay. They were never adequately filled and subsequently
rusted out thus feeding salt water vertically into the under-
lying aquifers.
A. Not only the Spring Valley Water Company test wells,
but many similar abandoned wells may exist in the study
area. All of these wells are potential sources of degradation
to the lower aquifers by leakage through the wells from
the upper, salinated aquifers. In addition, further westward
there is a greater degree of vertical hydraulic communication
between the shallow aquifers and the deeper, water-
producing zones, which occurs naturally due to the modes
of deposition of the sediments. The site for construction of
the injection/extraction system was selected so as to preclude
any problems such as abandoned wells in its vicinity. There
are no abandoned wells located between the injection/
extraction barrier and the Bayfront area, the source of
saline intrusion into the shallow aquifers in the project area.
Q. by Stanley N. Davis. Upper zones have a total dissolved
solids of about 5,000 ppm. Is this "highly saline"? (See
City records of Palo Alto Gold Course test well of about
1964.)
A. The upper zones, in fact, have total dissolved solids well
above 5,000 ppm. For example, the test hole drilled by the
Santa Clara Valley Water District in 1972, located immedi-
ately east of the injection/extraction system, shows a total
dissolved solids of 17,400 mg/1 in 1972 and a chloride
content of greater than 10,000 mg/1. The Palo Alto Golf
Course test well penetrates both the shallow zones and
the deeper, fresh-water zones. Perhaps the water-quality
analyses from this well reflect a mixture of both salinated
ground water and fresh ground water.
Q. by Russel E. Darr. Do I understand correctly that the
injection is in the 40-foot aquifer and the withdrawal is
in the 20-foot aquifer?
A. That is not a correct understanding. Both the injection
and extraction occur in the same 45-foot aquifer. Injection
is planned for the line of wells on the bayward side and
extraction is to occur through the line of wells on the
landward side. The purpose of the 20-foot aquifer wells is
only to allow pumpage to control waterlogging'. Injection
into the 45-foot aquifer may produce leakage through the
aquitard separating the two shallow aquifers and by pumping
the 20-foot aquifer, such leakage can be controlled to
eliminate any possibilities of waterlogging of the ground
surface.
Q. by Dennis Goldman. Were longer aquifer tests run? Do
you feel that a 24-hour test should be extrapolated to
imply zero leakage?
A. This question concerns the zone between the base of
the 45-foot aquifer and the lower 185-foot aquifer, herein
termed an aquiclude. The total time of pumping for
testing of zones on either side of this aquiclude, including
both the pumping tests and the injection tests, amounted
to over 96 hours of pumping. Concerning extrapolation of
aquifer-test data to imply zero leakage, it is important to
consider the scope and purposes of testing as well as the
physical geology and other limitations of the test site.
Examination of the regional geology of the area shows that
the aquiclude changes in character and lithology to the
west of the test site. In this area westerly, the aquiclude
provides vertical intercommunication between the shallow
and deep aquifer due to a pinching out of the clay zones,
an increase in permeability due to changes in lithology, or
a combination of the two. In addition, the potential for
abandoned wells located westerly of the test site which
provide vertical intercommunication from the shallow
to the deep aquifer through the aquiclude is also present.
With continued pumping in a confined aquifer, the area of
influence or zone of depression due to such pumping
eventually reaches far beyond the area of immediate
interest at the test site. Such continued pumpage in the
project area would eventually show evidence of leakage from
the shallow to the deeper zones. The purpose of the
aquifer test, however, was to determine the characteristics
of the aquiclude in the vicinity of the test site. Therefore,
for purposes of this investigation, the pumping tests
performed were adequate to determine the characteristics
of the aquiclude in view of both the project requirements
and the physical geology of the site.
Q. You stated that injection of wastewater would increase
head and prevent intrusion. What was the head in the upper
aquifer before injection? How much can the original head be
increased?
A. The head in the 45-foot aquifer at the injection/extraction
well prior to injection was at an elevation of —3.23 mean
sea level. It was determined that the original head can be
increased by a maximum of 20 psi, or approximately 46
feet, without creating any potential for heaving or leakage
around seals in the well.
Q. Recharge and subsequent extraction has no benefit over
salt-water intrusion if the water cannot be used. Do you
agree?
A. I do not agree. Salt-water intrusion has the potential of
degradation for the entire ground-water aquifer under
heavy pumping conditions in the Palo Alto area, and is
therefore detrimental. Recharge and subsequent extraction
have the benefit of controlling seawater intrusion as well as
providing utilization of the recharged and extracted water
for industrial and agricultural purposes downstream of the
project area.
Q. Would you save pumping costs if you stopped seawater
from coming into the bay if the Corps of Engineers were to
build a locks and dam under Golden Gate Bridge?
A. This may be a tongue-in-cheek question. Even if the
Corps of Engineers controlled the water quality within the
bay and maintained fresh water in San Francisco Bay, the
existence of salt evaporation ponds in the vicinity of the
project site present the potential for further salination of
shallow aquifers through leaching from these salt ponds.
Therefore, such a project would still be required to prevent
salination of the shallow aquifers and to provide protection
of the deeper water supplies.
170
-------
A Nonstructural Approach to Control
Salt Accumulation in Ground Water"
by Otto J. Helweg
ABSTRACT
One of the more subtle and dangerous areas of
pollution is that occurring in ground water from normal
irrigation practices. The pollution from salt buildup is
presently one of the unsolved problems in managing
stream-aquifer systems. This paper presents several
strategies that offer a possibility of controlling this salt
buildup. For example, instead of applying irrigation
ground water near the site of the well, the ASTRAN
Method transfers it downstream to be applied on land where
the ground water is a lower quality thereby controlling the
increase in salt concentration. Instead of preventing seepage
loss in delivery canals, the percolating water is used to
maintain ground-water quality. Finally, timed releases of
return flow remove salts without exceeding surface-water
quality constraints.
INTRODUCTION
As is often the case, the most obvious need
receives attention first even though it may not be
the most important; consequently, the visible
problem of surface-water pollution has been
emphasized. When the efforts of the NWWA pointed
to the equal, if not greater, need to address ground-
water pollution, the emphasis has been on dramatic
changes such as oil spills, wastes from feedlots, etc.,
instead of more subtle sources of ground-water
degradation. For example, normal irrigation pro-
cedures slowly but steadily increase the concentra-
tion of total dissolved solids (TDS or salt) in
0.5 m pure water evapotranspired
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
b Assistant Professor, Department of Civil Engineering,
University of California, Davis, California 95616.
TDS = 0 L0 m
irrigat
f 1 HM i
i « i & i —
/x/TTrv~
0.5 m3 water percolated TDS = 1000 mg 1
V
Groundwater TDS = 500 mg 1 \
UJ
3 ot
on water
500 mg 1
^
1 1 1 1 III
I i i r
Fig. 1. The degradation of ground water from irrigation.
ground water. The process, as shown in Figure 1,
works like this. An amount of irrigation ground
water (say 1m3) with a TDS of 500 mg/1, is
applied to the land. One-half of the water (but
hardly any of the dissolved solids) is evapotrans-
pired (used by the plants). This means that one-half
of a cubic meter of water with a TDS of 1000 mg/1
will percolate down to the water table and increase
the salinity of the ground water. The natural
recharge of the aquifer and lateral ground-water
movement is not rapid enough to move the more
saline water out so the concentration of salts
increases until the ground water and/or land
becomes unusable. Before the land was farmed,
nature had reached an equilibrium (or nearly so)
whereby the discharge of salts via lateral movement
171
-------
equaled the increase in concentration from natural
vegetation; moreover, the applied water was rain
which had a very low salt content. Though
measurements have confirmed this process, as yet
there is no economic procedure to control it,
notwithstanding, many proposals which may
retard degradation or delay it for a period of years.
STANDARD METHODS OF SALT CONTROL
A few forward-looking organizations are
addressing the problem of ground-water degradation.
Among these is the Colorado River Water Quality
Improvement Program. The Program Officer has
listed the various possibilities in water-quality
planning and management (Maletic, 1974). These
are given in Table 1.
The management techniques presented herein
would appear to fall under Category III 3, Ground-
Water Management, since they involve selective
pumping, application, and release of ground water.
They might also be classified under Category IV,
River System Management. These techniques are
not presented as the ultimate solution to ground-
water salinity problems, but a potentially viable
alternative for consideration among the many other
alternatives. The challenge to water planners and
managers is to find the optimal mix of alternatives
that will meet water-quality goals in the most
beneficial way.
ACCELERATED SALT TRANSPORT
(ASTRAN) METHOD
Many stream-aquifer systems may be
conducive to application of a potentially cost-
effective salinity management technique referred
to as the Accelerated Salt Transport (ASTRAN)
Method (Helweg and Labadie, 1976). The usual
irrigation practice is to apply ground water on fields
near the well supplying the water. The ASTRAN
Method, however, encourages the application of
ground water on downstream fields (via canals or
pipes) instead of on nearby fields. That is, fields
proximate to a pumping well should be irrigated,
to at least some extent, by water from upstream
wells. Salts in the pumped water can therefore be
transported downstream at a faster rate than would
occur naturally through flow in the saturated zone;
hence, the name Accelerated Salt Transport
(ASTRAN) Method. The slow movement of ground
water tends to cause an accumulation of salts from
normal irrigation practice, since drainage water adds
salt to the aquifer at a faster rate than it can be
naturally transported downstream. The idea behind
this technique is to simply augment the natural
process by transporting pumped water downstream
via a surface distribution system of some kind.
The management technique is illustrated in
Figure 2. This diagram depicts an increasing salt
concentration downstream, since many stream-
aquifer systems have this characteristic. Such a
condition would not seem to be necessary for
successful application of the ASTRAN Method. It
is, however, required for application of the manage-
ment algorithm presented in this report. Appropriate
modification of the algorithm could be carried out,
in case this condition does not exist.
Table 1. Major Categories of Salinity Control
I. POINT SOURCES
1. Desalt
2. Divert/Evaporate
3. Divert/Special Use
4. Plug Wells
5. Deep Injection
II. NATURAL DIFFUSE SOURCES
1. Collect/Desalt
2. Collect/Evaporate
3. Collect/Special Use
4. Watershed Management
a. Vegetative conversions
b. Forest management
c. Structural measures
d. Water harvesting
e. Reduced sediment production
5. Phreatophyte Control
a. Control of spread
b. Replacement vegetation
c. Antitranspirants
III. IRRIGATION SOURCES
1. Improved On-Farm Irrigation Use
a. Irrigation scheduling
b. Improved farm irrigation systems
1) Pipes and lining
2) Automation
3) Advanced systems
2. Improved Water Conveyance Systems
a. Piles, lining
3. Ground-Water Management
a. Water-table control (drainage)
b. Selective pumping
c. Ground-water recharge
4. Return Flow Management
a. Collect/desalt
b. Collect/special use
5. Evaporation Suppression
IV. RIVER SYSTEM MANAGEMENT
1. Alteration of Time Pattern of Streamflow
2. Alteration of Time Pattern of Saline Discharges
V. DILUTION
1. Augmentation
a. Weather modification
b. Geothermal resources
c. Desalting
d. Wastewater reclamation
e. Conservation practices
2. Importation
172
-------
-Grwndwater Being Transported Downstream
(Accelerated Salt Flow)
Irritation water
(Same Concentration
as Pumping Well)
Drainage Water
With Increased
Concentration
-^ Increasing Salt Concentration
Fig. 2. A schematic diagram of the Accelerated Salt
Transport (ASTRAN) Method.
Quantifying the ASTRAN Method entails
finding out how much water should be taken from
each source and where it should be applied. It is
certainly possible to pump more than needed and
maintain a hydrologic balance by artificial
recharge if the constraints and economics so
demand. If water quality were not a consideration,
one could just apply a subset of Linear Program-
ming (LP) called the Transportation Problem (see
Wagner, 1975).
The Transportation Problem is designed for a
situation where there are supply points (sources),
demand points, and costs to transport the items
between each source and demand. Figure 3
illustrates this situation. There is a least cost
solution to this' problem which can lead to a
solution to the ASTRAN Method. Figure 4 shows
the Bonsall subbasin in California, where the
ASTRAN Method was tested by simulation models.
The subbasin was divided into four sections, each
one being both a source and demand. Additionally,
two other sources were available; ground water
from upstream basins (USGW), and imported State
Project water (SPW). There was also an additional
demand and that was for water needed downstream
or exported out of the subbasin (EXP).
The amount of water demanded is set by the
crops and irrigated land. The amount of water
available is determined by the aquifer size and
QUANTITY TRANSPORTED FROM
SUPPLY i TO DEMAND j
"ii
S
u
p
p
L
I
E
S
Fig. 3. The transportation problem.
characteristics. The costs for transporting water
between sources and demands is calculated by
water costs (as for imported water), pumping
costs (for ground water), and the distance it needs
to be transported (costs of canals, etc.). Mathe-
matically, the problem is formulated as follows:
n m
j=l, . . . ,m
subject to:
m
2 qij<
0, i=l, .
. m
1=1, . , m
(1)
Fig. 4. Supply and demand points for Bonsall subbasin.
173
-------
in which:
c;: = the cost of transporting water from
source section i to demand section j
($/AF);
q;; = the amount of water transferred from
itoj (AF/yr);
s; = the amount of water available at source
section i (AF/yr);
d: = the amount of water needed at demand
section j (AF/yr).
As mentioned above, the Transportation
Algorithm is only the basis of a solution strategy
because, as yet, water quality has not been con-
sidered. The TDS of each supply is different and in
order to control ground-water degradation the
applied irrigation water must be of sufficiently
higher quality to maintain mass balance of salts.
Consequently, the algorithm needs an additional
water-quality constraint. The problem is to find the
quality of the applied irrigation water needed in
order to maintain ground-water quality. An initial
approximation is to assume a static situation and
use the leaching formula:
— ECV
X ET
(2)
in which:
Dw
ET
= the depth of the supplied irrigation
water for leaching (cm);
= the evapotranspiration or consumptive
use (cm);
= the electroconductivity of the drainage
water percolating past the root zone
(micromhos/cm2);
= the electroconductivity of the applied
irrigation water (micromhos/cm2).
EC is easily found from TDS by the relation,
TDS = -2 + 0.683 EC.
Rearranging terms, assuming six sources and
four demands of irrigation water as in Figure 4
yields:
EC
W
6
Cmj 2
i=l
in which:
CmjET:, j=l, . , 4 (3)
J J
= Concentration of Applied Irrigation Water
from source i(i = 1, ,, 5)
1
I I I I
Ground Surface
Root pS^iS^^^^SSf^^
Zone
Unsaturated
Zone
= Concentration of Drainage Water
v Groundwater Table
C - = Assumed Concentration
of Groundwater
Fig. 5. Water-quality schematic of demand node j.
trations of the drainage water at section
j (mg/1);
C; = average TDS concentration of irrigation
water from source section i (mg/1);
qj; = amount of water transferred from
source section i to demand j (AF/yr);
ET; = evapotranspiration at section j (AF/yr).
Figure 5 shows the locations of the various waters
for formula (3).
The situation is a little more complex than
Figure 5 because the aquifer is not static as assumed
by the leaching formula. Ground water is flowing in
and out of the aquifer from the boundaries plus
water is entering or leaving the aquifer from rain or
via the river. For the Bonsall basin, it was found that
the water-quality constraint could be up to 300
mg/1 greater than the actual ground water and
maintain a near stable salt balance. Figure 6 shows
the long-term solution for Bonsall basin and Figure
7 shows the improvement of the downstream
portion of the basin the ASTRAN Method would
have had if it had been applied back in 1966.
Cmj = upper bound on average TDS concen-
Fig. 6. Long-term solution for Bonsall subbasin (numbers
are water transfers in acre-feet/year).
174
-------
2500 r
2000
u>
E
v>
a
1500
1000
Mean Historic Values
DCON=100-
T
1960
1962
1964
YEAR
1966
1968
Fig. 7. Ground-water quality, comparing the historic values
with the ASTRAN Method (DCON is the difference between
ground-water quality and drainage-water quality).
CONTROLLED SEEPAGE RECHARGE
Depending on the situation, the ASTRAN
Method may call for artificial recharge to meet the
water-quality and quantity constraints. Often this
can be accomplished by Controlled Seepage
Recharge (GSR) that is allowing seepage in delivery
canals purposely by so designing the water delivery
system that the seeping water not only helps to
maintain a hydrologic balance, but a water-quality
balance. Figure 8 shows how this might be done.
Most irrigation systems' designers consider
seepage a loss, and line delivery canals to prevent
this. Though this may often be desirable, lining
canals is not always called for. For example,
Konikow and Bredehoeft (1974) showed that
unlined canals were, in fact, helping to maintain
the ground-water quality of the Arkansas River
basin in Colorado. Basically, the canals can be
designed for two purposes, water delivery and
artificial recharge. It may be that such a multi-
purpose use of the canal would make the construc-
tion of artificial recharge facilities unnecessary.
TIMED GROUND-WATER RELEASE
It is not yet clear what the impact of the
Federal Water Pollution Control Act (PL 92-500)
will have on irrigation return flow; but if a river is
to maintain a certain water quality such as a
controlled TDS, the irrigation return flow from
the drains (which cause the river to exceed this
during part of the year) may be prohibited.
Normally, irrigated areas are in climates where the
Summer river flows are low; consequently, the
relatively saline water returning to the river from
irrigated lands have a much higher effect on the
surface-water quality than they would during the
Winter or Spring months. An example of Timed
Ground-water Release (TGR) is operating tiled
places so that the outfalls into the rivers can be
shut during the Summer months and then opened
during months of high flow. Moreover, if pumping
is called for to export saline water, this can be
accomplished during these same periods of high
flows.
For example, if the TDS of a river were
constant, and the maximum allowable TDS were
800 mg/1, a timed ground-water release policy as
shown in Figure 9 during the months of January
through March would not exceed the surface-water-
quality standards. Often, however, the TDS during
-1500
Groundwater Release
(TDS =1200 me/1)
- 300 -
- 200
- 100
Fig. 8. Controlled seepage recharge from an unlined ditch.
J F M A M J J A S ' 0 ' N ' D
MONTHS
Fig. 9. The change in river water quality from timed
ground-water releases.
175
-------
the periods of high flow is lower than the low
flows, so this would show an even more favorable
picture than the one illustrated. Of course the rate
of ground-water release need not be constant but
could be matched to the quantity of the natural
flow. Some work along these lines has been done
in Australia in the Murray River basin (Williams,
1972).
CONCLUSIONS
The main point of this paper is that often
water-quality objectives can be approached by
including nonstructural measures in the over-all
management program. These measures have the
advantage of requiring low initial investment
costs. For example, one estimate for decreasing the
rate of ground-water degradation, but not halting
it, in the Bonsall area was $3,900,000 (Santa
Margarita-San Luis Rey Watershed Planning Agency,
1973) while the ASTRAN Method could com-
pletely control degradation at roughly 10% of that
figure (Helweg and Labadie, 1976). Nonstructural
measures provide greater future flexibility in that
future options have not been decreased by
constructing fixed facilities.
However, several words of caution are needed.
As with many water problems, the technical
solution is much easier than the political. For the
CSR or the TGR methods to work effectively, the
whole river basin should be coordinated and this is
difficult when there are many ditch companies or
irrigation districts. Nevertheless, the potential gains
of nonstructural measures to control salt accumula-
tion seem to warrant their implementations.
ACKNOWLEDGEMENTS
The research for this paper was partially
supported by the Office of Water Research and
Technology and the State of California.
REFERENCES
Helweg, O. J. and J. W. Labadie. 1976. Accelerated salt
transport method for managing groundwater quality.
Water Resources Bulletin, v. 12, no. 4, August.
Konikow, L. F. and J. D. Bredehoeft. 1974. Modeling flow
and chemical quality changes in an irrigated
stream-aquifer system. Water Resources Research.
v. 10, no. 3, June, pp. 546-562.
Maletic, J. T. 1974. Current approaches and alternatives to
salinity management in the Colorado River basin.
Salinity in Water Resources. J. E. Flack and C. H.
Howe, eds. Marriman Pub. Co., Boulder, pp. 11-29.
Santa Margarita-San Luis Rey Watershed Planning Agency.
1973. Joint administration committee of the
comprehensive water qualtiy management study.
2 vols., December.
Wagner, H. M. 1975. Principles of operations research.
Prentice-Hall, Englewood Cliffs.
Williams, A. F. 1972. A summary of progress reports
covering investigations of groundwater hydrology
adjacent to the River Murray in South Australia
with some additional notes. Department of Mines,
South Australia, 27 September.
DISCUSSION
The following questions were answered by Otto J. Helweg
after delivering his talk entitled "A Nonstructural Approach
to Control Salt Accumulation in Ground Water."
Q. by Don Lundy. The ground-water flow equation presented
in one slide is written for one-dimensional flow in the
x-direction. Subsequent maps and sections indicate
component flow in at least one more direction (horizontal).
How do you justify flow in a y-direction?
A. The flow equation as shown on the slide was:
— (Tij — ) = S- + W(x,y,t)
3x;
It should have been:
ox;
ot
in which:
S
h
W
a ah
9x; 9x;
3h
' = S — +W(xj,t)
= the spatial cartesian coordinates (i=l,2)(L);
4
= the second order transmissivity tensor (L2/T);
= the storage coefficient (dimensionless);
= the piezometric surface above some datum
(L);
= the volume flux per unit area (L/t).
So the model was two-dimensional; however, because of the
long, narrow geometry of the river basin, we could have
modeled it with a one-dimensional model. The reason we
didn't was because the USGS had already calibrated a
176
-------
two-dimensional quantity model and it seemed reasonable
to utilize the work they had already completed.
Q. by Leonard Konikow. What value of dispersivity did you
use and how sensitive was the model to this parameter?
A. Our final values for BETA (the characteristic length)
was 100 and the value we used for DLTRAT (the ratio of
transverse to longitudinal dispersivity) was 0.3. We did an
informal sensitivity analysis with these two numbers and the
model was not sensitive to them. In fact, with BETA equal
to zero (i.e., dispersivity equal to zero), we still got good
results.
Q. by Robert 0. Thomas. 7s there any work being
accomplished on the accelerated transmittal of a contami-
nant—at least transmittal evidenced by odor, etc. ? I refer
specifically to phenol movement of over 10 miles in a week
in a ground-water body moving in the low hundreds of
feet per year. This occurred in the lower San Gedonel
Valley, California in the early thirties.
A. Not to my knowledge.
Q. by Robert O. Thomas. Does not properly managed
conjunctive use, using surface water for both recharge and
supplemental applied water, have a large positive effect on
maintenance of ground^water quality?
A. How do you "properly manage conjunctive use"? We
are proposing the ASTRAN Method as a "properly managed
conjunctive use" program. However, the ASTRAN Method
can be applied where there is no surface water available to
be imported into the river basin. The problem is that there
is often not enough good quality surface water to go around.
Q. by G. H. Hendricks. Would you briefly discuss rainfall
amounts relative to leaching? What are the limitations?
A. The precipitation average at the upper end of the basin
is about 13 inches. We did not consider the leaching effect
of this rain in relation to the root zone because we were
investigating the water quality on a larger scale, i.e. the
ground-water table. The precipitation was included in the
simulation model, however. There have been studies on
leaching from precipitation, but the information was not
pertinent to our study.
Q. Who in USGS developed the model that incorporates
water quality? What division of USGS?
A. The simulation model was developed by Bredehoeft and
Konikow of the Water Resources Research section of the
USGS at Denver. Their work was based on research done
at Colorado State University by Redell and Sunada. This
model solves both the flow equation and the transport
equation and is one of the few working models that does so.
Q. by John E. Mann, Jr. Do you know the magnitude of
TDS pick-up shown by the Imperial Valley drains?
A. The average electroconductivity (EC) of the irrigation
water from the Colorado River is 1300 micromhos/cm and
the drainage water from the tiles ranges from 2300 to
3900 micromhos/cm. If you are talking just about surface
runoff, the EC increases from 5 to 10%.
Q. by W. B. Wilkinson. Were sensitivity tests applied to
the simulation model and if so, which of the aquifer's
properties proved to be the most critical with respect to
decision making? How were the dispersivity values used in
the model determined?
A. Concerning the flow portion of the simulation model,
initial water levels were the most sensitive factor, followed
by transmissivity and storage coefficient. Informal
sensitivity tests were used, but the parameters, such as
transmissivity, were not formally optimized.
The dispersivity values were taken from USGS values
measured in similar systems. Since they proved to have very
little effect on the results, there was no need to spend money
testing the aquifers in the basin.
Q. by L. A. Swain. You assume (in model) that recharged
water mixes uniformly throughout entire vertical thickness
of aquifer. Do you have any monitoring data to support
this idea rather than the recharge water being stratified atop
the aquifer water?
A. We do not have any monitoring data because there is no
artificial recharge presently being done. If artificial
recharge would be used, the design of the pits (or wells)
would have a great effect on the mixing. Because of the
geometry of the aquifer though (i.e., shallow with not
much layering), we would not expect stratification to be a
major problem. Remember, though, the ASTRAN Method
does not necessarily need artificial recharge; artificial
recharge would only be used if the decision maker wanted to
so improve the ground-water quality that the normal
transfer of irrigation water was insufficient.
Q. by Donald Runnells. Does the ASTRAN model consider
any chemical reactions such as precipitation of salts, ion
exchange, etc.?
A. No. Investigators at the University of Arizona and
others have modeled the chemical reactions in the root zone
and vadose zone, but doing this would have greatly
complicated the simulation model and yielded little if any
benefit. Remember that we are managing water quality on a
large scale so the models are appropriately (we trust)
simplified.
Q. by K. Childs. Does the import of water address the
problem or simply the symptom? Is this not using dilution
as a solution to pollution?
How does your suggested methodology address the
total quantity of salts in the hydrologic regime?
A. To answer the first part, we feel that the importation of
water can be merely diluting saline water, but using the
ASTRAN Method would prevent this. What the ASTRAN
Method does is to apply the imported water at the best
place in order to maintain the appropriate water-quality
gradient between drainage water and applied water. The
point of application would also assist in moving the salts
downstream. Moreover, as stated above, imported water is
not necessary to implement the ASTRAN Method.
The simulation model calculated the mass
balance of salts, though at the time we used it there were
some inaccuracies that have subsequently been corrected
by Knokow at the USGS. Consequently, the ASTRAN
Method not only controlled water quality (as measured
by total dissolved solids), but caused a salt balance in the
system also. Obviously water quality and mass balance of
salts are connected, but the decision maker can look at
both in deciding to what degree he wants to control the
ground-water quality.
177
-------
Quantifying the Natural Flushout
of Alluvial Aquifers"
by John S. Fryberger and William H. Bellisb
ABSTRACT
A plume of contaminated water may extend down-
stream for several miles in an alluvial aquifer. Hydrologic
forces acting on the aquifer generally tend to dilute the
contaminant downstream from the source. After the
pollution has been stopped, how rapidly will these hydrologic
forces flush out the contaminant? The answer to this
question may be calculated by using the mass-balance
equation herein developed.
This mass-balance mathematical model incorporates
all the hydrologic forces acting on the alluvial aquifer that
affect the concentration of the contaminant. Methods are
presented to define and quantify each of the hydrologic
forces. These forces include (1) present quantity and
quality (with respect to contaminant) of water in the
alluvium and the change in the quality downstream,
(2) quantity and quality of ground-water inflow,
(3) quantity and quality of flood inflow, (4) quantity and
quality of base flow, (5) quantity of recharge from
precipitation, and (6) quantity of loss from
evapotranspiration.
The equation is first balanced to agree with past
conditions and observed field data and is then used to
predict future quality changes after the pollution is stopped.
The method is applicable primarily to situations in
which the pollution has been taking place over a long
period of time and the water quality in the alluvium has
reached equilibrium at any given point. Modifications may
be possible to permit use of the general approach to
short-term or slug-type pollution events.
INTRODUCTION
Along the drainage systems of the Red River
of Oklahoma and northern Texas and the Arkansas
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
byice President and Hydrogeologist; and Hydro-
geologist, respectively, Engineering Enterprises, Inc.,
1225 W. Main, Norman, OK 73069.
River of Oklahoma and Kansas are significant
natural salt-water springs and seeps which discharge
approximately 20,000 tons of salt (NaCl) per day
into these river systems.
This salt-water inflow has a significant adverse
effect on the quality of the surface water as well
as contaminating the river alluvium for hundreds
of miles downstream from these natural springs.
The Corps of Engineers has the responsibility for
developing methods of control, which in most cases
will involve the collection and disposal of brine in
off-stream evaporation reservoirs, thus essentially
stopping the salt from entering the river systems.
However, the flushout of the salt from the
alluvium downstream from the source areas and
the effect of such flushout on surface-water
quality are of major concern.
As a result of this interest a mathematical
model was developed for the Corps of Engineers
which describes the hydrologic forces that control
the flushout process. This model was applied to the
Red River and several tributaries in Oklahoma and
Texas with apparent realistic success. The concepts
and procedures developed during the course of this
study work may have general application for the
prediction of the long-term impact and fate of
pollution plumes in alluvium.
The application of the mass-balance model is
presently limited to those areas where the source
of pollution has been active for sufficient time to
allow a state of equilibrium to be established for
the decay of pollutant concentration downstream
from the source. Modification of the model will be
required to define the future flushout of plumes
not yet at equilibrium with the flushout forces.
BASIC CONCEPT
The fundamental principles of a hydrologic
mass-balance equation can be expressed by defining
178
-------
the volume of water and the chemical concentra-
tion for each parameter as follows:
VC0 + 1C; - LQ = VCr
(1)
where:
V = original and ending volume of water,
C0 = original concentration of the
contamination in V,
I = volume of water being added
(inflow) to the original volume,
Q = concentration of the water added,
L = volume of water removed after
mixing (equal to volume added),
Cr = concentration of water after mixing
(applied both to the water removed
and water remaining), and
VCr = result in which the volume is the
same as the original volume but the
concentration has changed.
The inflow water, I, with concentration, Q, is
mixed with the initial volume of water with a
concentration, C0 , to form a new mixture with a
concentration of Cr. Then the quantity, L, which
is equal to the quantity, I, but having a concentra-
tion of Cr, is removed, leaving the original volume,
V, with a concentration of Cr .
Because there are two components to each
factor, it is possible to write two equations and,
therefore, solve for two unknowns.
VC0
I-L = V,
- LCr = VCr.
(2)
(3)
This mass-balance principle is herein expanded to
define the hydrologic forces acting on an alluvial
aquifer. These forces are depicted in Figure 1.
THE MODEL
General
Expansion of the mass-balance concept to
describe the hydrologic forces acting on an alluvial
aquifer results in the following equation:
QCa + GCg + FCf - BQ + P - E + RCr (4)
M &
where:
Q = original quantity of water stored in
the alluvium per unit distance,
G = quantity of ground-water inflow
from bedrock into the alluvium,
QCq + GCg + FCf - BCr + P - E = RCr
Where 0 = Onginoi quonlily ol water stored in tne alluvium per uml distance
G " Ground water inllow into the alluvium (ram bedrock
F - Flood inllow into trie I
8 a Base How - primarily
Fig. 1. Schematic of alluvial flushout equation.
F = quantity of flood inflow that mixes
with the alluvial water,
B = quantity of base flow from the
alluvium into the surface stream,
P = quantity of precipitation that mixes
with the alluvial water,
E - quantity of water lost from the
alluvium by evapotranspiration,
R = quantity of water remaining in the
alluvium, which over a long term is
equal to Q, and
C = contaminant concentration of each
respective volume of water in
milligrams per liter (mg/1).
Concentrations of P and E are assumed to
equal 1 and are not shown in the
equation.
In general, the model is applied to field situa-
tions by calculating and adjusting the values of
each component to match observed field conditions.
Future predictions may then be made by changing
the value of the key input factor to its future value.
Detailed discussions of the mechanics of apply-
ing the model follow later. The basic steps are:
a. Determine from field observations the
present decay rate of the contaminant downstream
from the source, Cq - Cr, per unit distance. This
change in the alluvial water concentration, Cq - Cr,
takes place over a calculable period of time, t.
b. Over the time period, t, calculate the values
of GCa.Pand E.
D
c. The values for F and B may be calculated
using simultaneous equations.
179
-------
d. After all values of the factors have been
defined so that the equation balances using the
observed values of Cq and Cr, adjustments can be
made if the values of any parameters appear
unreasonable. An independent check is available
for the value for F to determine its reasonableness.
e. Future values for Cr can then be made by
changing one of the input parameters such as Cf to
an estimated future value.
CALCULATING EACH COMPONENT
As a given volume of water moves slowly
downstream in the alluvium, the concentration of
the contaminant decreases as a result of the
hydrologic forces acting on that volume. It is
assumed that although water is added and removed,
the volume of water remains constant over the long
term. However, the quality changes with distance
downstream from the source and can be determined
by field testing. By calculating the time for water in
the alluvium to move a unit distance, and because
the change in the quality over that distance is
known, Cq - Qi, the change in quality can be
related to time. The factors that cause the change
in quality over that same time period can then be
calculated and the equation established.
QCq and RCr
The original quantity, Q, and resultant
quantity, R, are equal and represent the volume of
water in an average one-mile stretch of the river
along the study reach. As this volume of water
moves downstream the concentration of the
pollutant decreases. The average decrease, Cq - Cr,
per mile, or some other unit distance, is
established by the existing decay gradient deter-
mined from test drilling, as illustrated in Figure 2.
Note that, in the example shown in Figure 2, the
river has been divided into three flushout reaches
on the basis of changing hydrogeologic conditions
along the river.
The time required for the alluvial water to
travel one mile is then calculated. The velocity of
the ground water is calculated using the following
equation:
KI
v =
where:
7.48 Pe
v = velocity in feet per day,
K= permeability in gpd/ft2,
I = hydraulic gradient,
(5)
lOpOO
r '
1
? 1000-
S
ONE-MILE
. USED IN F
CALCULATI
F
STRETCH J-
NS
L U S'
^
-**l [ : A IWFRAGf CONTAM NflWT rONrFfJTBBT.m,
\ , ; pun , wfi nara
20 40 6
RIVER MILES D
\
X
\
\
: V^
, , , i >s-*<; —
i
1
0 80 100 IZO 140 160
Fig. 2. An example of natural present decay of contaminant
downstream from source.
Pe = effective porosity, a value less than
total porosity, Pt, due to friction and
viscosity—Pe = Pt/2 is suggested,
7.48 = constant.
The time of travel, t, for one mile is then
determined by:
t in years =
5280 ft/mi
(365d/yr)(vft/d)
(6)
GCg
The quantity of ground-water inflow, G, into
the alluvium from the adjacent bedrock can be
determined by standard methods. The average
ground-water gradient toward the river along the
reach being studied is obtained from water-level-
contour maps. The permeability is averaged from
pumping tests or estimated based on a knowledge
of the formation types. The quantity, G, is estimated
using the permeability, gradient, area of inflow into
the alluvium and the time of travel determined from
equation (6). This represents the quantity of ground
water that mixes with the original quantity, Q, in
the alluvium during the time that Q water travels
one mile (the unit distance chosen for this example).
365 KIA (t)
325,829
(7)
where:
G = quantity of ground-water inflow in
acre-feet in t, time,
K = permeability in gpd,
I = ground-water gradient in the bedrock,
180
-------
A = area of inflow = (5280 ft) (depth of
alluvium) (2),
t = time in years from equation (6),
365 = days per year,
325,829= gallons per acre-foot.
The concentration of the pollutant in the
ground water in the bedrock, Cg, is determined
from chemical analyses of bedrock-water samples.
If no pollutant is present, Cg is assumed to equal 1.
The quantity of precipitation, P, that
infiltrates into the alluvium along a one-mile stretch
of river is a fraction of the total precipitation, Pt,
and is dependent on several factors such as soil
type, vegetation, slope, rainfall intensity and
duration, and temperature. It is generally not
practical to define this quantity precisely. An
estimate ranging from 5 to 30 percent of the total
precipitation can be made by judging the effects
of the controlling factors. P is calculated as follows:
P =
JPtAt
12
(8)
where:
P = quantity of precipitation mixing
with alluvial water in acre-feet,
J = judgment factor, ranging from .05
to .3,
Pt = total average annual precipitation
in inches,
A = area of the alluvium in one mile in
acres,
t = time in years from equation (6),
12 = correction factor to obtain acre-feet.
The quantity of water lost from the alluvium by
evapotranspiration, E, is an important factor in the
mass-balance equation but is not easily determined.
Methods for calculating evaporation are found in
Sleight (1917), Hughes and McDonald (1966),
Sorey and Matlock (1969), Hellwig (1973), and
Gaetwood and others (1950).
Gruff and Thompson (1967) concluded, in a
comparative study of different methods for calcu-
lating transpiration, that the Blaney and Griddle
(1950) method is generally the most practical. The
reader is referred to the above references for detailed
instructions on the calculation of evapotranspiration
losses.
Factors that influence transpiration losses
include:
Depth of water table.
Type of vegetation.
Density of vegetation.
Percentage of area covered by vegetation.
Area per mile length of river underlain by
alluvium.
Climatological factors.
Factors that influence evaporation losses
include:
Temperature.
Wind.
Depth of water table.
Soil types and grain size.
Potential evaporation and transpiration losses
can be calculated for each subarea in the one-mile
stretch as follows:
Percentage of total area for open running
water (river surface).
Percentage of total area where the water
table is less than two feet deep.
Percentage of total area where the water
table is more than two feet deep.
Percentage of total area vegetated, and
corrected for density of vegetation.
The sum of the preceding evaporation and
transpiration losses constitutes the total potential
evapotranspiration, Et. The amount of evapo-
transpiration that affects the quantity of ground
water in the alluvium, E, is the total potential
evapotranspiration, Et, less that part satisfied by
soil water above the water table. The amount of
soil water may be estimated as that precipitation
that does not run off or does not percolate to
the water table. The percent of total precipitation
that runs off is governed by the amount of
precipitation, time distribution and intensity of
storms, soil type, surface slope, and density of
vegetation. Judgement also must be used in
estimating the amount of slope wash or other input
to the alluvium from runoff from the bedrock-
drainage area.
Although the subarea fractions of Et are calcu-
lated in terms of acre-feet, the value of Et should be
expressed in terms of inches over the entire area of
the alluvium in the one-mile stretch.
E is then calculated as follows:
E =
[Et-(Pt-P-Pr)] At
12
(9)
181
-------
where:
E = quantity of evapotranspiration
affecting the alluvial water in acre-
feet over time, t,
Et = total potential annual evapotranspira-
tion in inches,
Pt = total annual precipitation in inches,
Pr = estimated amount of total precipita-
tion that leaves the system as direct
runoff, in inches,
P = quantity of precipitation that
percolates to the water table from
equation (8) but expressed in inches,
A = area underlain by alluvium in acres
in one-mile stretch,
t = time in years from equation (6).
FCf
During periods of flooding, considerable
amounts of flood water, F, enter the alluvium and
mix with the alluvial water. The quality of this
water, Cf, is determined from flow-concentration-
duration curves or other direct measurements of
the concentration of the flood waters.
The quantity of flood inflow, F, over time, t,
can be directly estimated from gage height, flood
frequency curves, unsaturated thickness, surface
profiles across the river, and porosity data.
However, this calculation is even more difficult
than calculating the other factors and is done
primarily as an order-of-magnitude check on the
value of F determined using the equation. The
quantity of flood inflow can be more easily
determined using the mass-balance equation. At
this point in the discussion all the factors have
been calculated except F, flood inflow, and B,
base outflow. Using the two basic equations, these
unknowns may be determined as follows:
The quantity-only equation is:
- (13)
Q + G + F-B + P- E='R,
and because Q = R, then
(10)
G + F + P = B + E or inflow = outflow, (11)
and B = F + G + P-E. Then by substitution for B
in the full equation,
QC + GC + FCf -
Cr- Cf
BCr
The quantity of base outflow, B, from the
alluvium into the surface flow is extremely
difficult to measure. Therefore, this value is
determined from the quantity-only part of the
equation after other values have been calculated as
follows:
B = F + G + P- E.
(14)
P-E)Cr + P-E = RCr , (12)
and solving for F
The concentration of the base outflow, Cr, is
assumed to average the same as the resulting water
concentration, Cr, as developed in the basic
concept under equation (1). It should be noted
that the actual concentration of the base outflow
varies with time and that a more accurate repre-
sentation would be to express the pollutant removal
in terms of tons per year rather than as a concentra-
tion. This value represents the quantity of pollutant
or base load, BL, that moves from the alluvium into
the surface flow as the alluvial water flows
downstream a unit distance in time, t.
After all the factors in the mass-balance
equation have been initially estimated, it may be
necessary to adjust some of the parameters to get
the equation to balance and to obtain reasonable
values for each factor. Once this is done, then future
predictions can be made based on an assumed or
known change of one of the factors.
FUTURE PREDICTIONS
The objective of the mass-balance equation
is to predict the rate of future flushout after the
contamination source has been stopped. The present
application of the equation is predicated on the
assumption that the hydrologic forces and the
contaminant concentration are in equilibrium, and
that a future change in a major factor is required
to bring about a reduction in alluvial contaminant
concentration leading to a new equilibrium.
On the project for which the equation was
developed, the future change will be the value for
the chloride concentration of the flood water, Cf.
In the vicinity of the salt-spring areas, and for miles
downstream, large salt flats are present. During each
flood the salt from the flats is redissolved and adds
considerably to the salt load. Elimination of the salt
flats will result in a significant decrease in the
chloride concentration of the flood water. There-
fore, to predict the future chloride concentration of
alluvial water, the value of Cf is reduced to reflect
its future concentration, Q,, and the equation is
182
-------
rearranged and solved for a new value of Cr as
follows:
QCq + GCg + FCfl + P-E
ri ~ R + B
(15)
This new value for Qi represents the future
concentration of the contaminant in the alluvium at
the downstream end of the one-mile stretch of river
at the end of time, t, after the change has been
initiated.
Also, at the end of time, t, a new value for
Cq will be established at the upstream end of the
one-mile stretch. This new value, Cq1( can be
determined from a plot of the field data, Cr vs.
Cq - Cr, as demonstrated by Figure 3. Information
for this curve is from the same field data used to
construct Figure 2. Cq - Cr is the contaminant
reduction per mile, and this difference decreases as
the concentration Cr decreases. Therefore, for each
new value of Cr such as Cri, Cr2 . Crn calculated
at the end of times (tt, t2 . . tn) using the equation,
a new value for Cqi, Cq2 . . . Cqn can be determined
by adding the appropriate value of Cq - Cr from
Figure 3 to Cr to find the new Cq.
Cr , mg/l CONTAMINANT ]
/
/
/
/
/
/
/ °
/
/
/
/
/
/
/
/
/
/
4AN6E IN CONTAMINA
PER RIVER MILE AS
CONTAMINANT CC
»JT CONCENTRATION
A FUNCTION OF
NCENTRATION
0 50 100 150 200 2!
Cq - C, ,mg/l CONTAMINANT
0
Fig. 3.An example of the rate of decay of contaminant
versus contaminant concentration. Used to determine
Cqn when Crn is known.
CONTAMINANT CONCENTRATION ^ IN mg / 1 £ |
00
\
\
\
\
\
\
\
N
\
R
V
'N^.
"^-^J
ESIDUAL
ENTRATIOf
FLUSH -OUT
j
CONTAMU
i IN AL
REACH I
> O c
ANT
5 O <
0 50 100 150 200 250 300
YEARS AFTER STOPPING CONTAMINANT INFLOW
Fig. 4. An example of a future decay curve calculated
from repeated iterations of the equation.
Then for the second time interval, t2, the
equation is solved as follows:
Cr,=
QCQ1 +GCg + FCf, +P- E
i o
R + B
(16)
The future values for Cn, Cr2 Crn calcu-
lated by repeated runs of the equation can be
plotted to show the future decay of the contami-
nant concentration in the alluvium and the new
equilibrium level as illustrated in Figure 4.
In addition, the future contaminant load
entering the surface flow from the alluvium can be
determined by plotting the values of base load,
BL, which is the average of BCr over a given time
period expressed in tons per year. This value may
be determined using the following equation:
BL =
1.359X 10'3 B(Crn +Crn-i)
2t
(17)
where:
BL = load from base flow in tons per year,
B = quantity of base flow over t years
from equation (14),
Crn +C
= average contaminant concentration
of base flow within the time interval
of tn and tn-i ,
183
-------
TONS OF CONTAMINANT PER YEAR /RIVER MILE _ |
«
\
\
s
\
\
\
CONT/
FROM
FLUSI-
VMINANT
BASE
-OUT REA
LOAD
FLOW
CH I
0 50 100 ISO 200 250 300
YEARS AFTER STOPPING CONTAMINANT INFLOW
Fig. 5. An example of base load, BL, in tons of contaminant
(chlorides) per year per river mile.
1.359 X 1CT3 = constant to obtain tons of chloride
from concentration and volume. A
different constant is required for
other contaminants.
Figure 5 is a plot of future contaminant load
from base flow. Note that the points connected by
the curve are halfway between the points on Figure
4 on the time scale because the average load between
each time interval, tn to tn-!, was calculated.
If the affected river is contaminated for a long
distance downstream from the source, then it
should be divided into several reaches. Each factor
for the equation should be calculated separately
for each flushout reach. Some of the factors, such
as GCg, FCf and even E, may change considerably
between flushout reaches. The one-mile stretch used
in the equation should represent the midpoint of
each flushout reach. After calculating the future
decay curve or residual concentration at the
midpoint of each reach, new profiles of contami-
nant concentration vs. distance downstream from
the source can be plotted at selected future times.
Such a plot is demonstrated in Figure 6.
CONCLUSIONS AND RECOMMENDATIONS
The mass-balance equation herein developed
has been applied to five rivers in Oklahoma and
Texas. The predicted future salt flushout from the
alluvium is quite different from river to river,
reflecting differences in the hydrologic conditions
along the rivers.
Although the equation is believed to represent
a valid approximation of future flushout, the
accuracy of the future predictions is dependent on
the accuracy with which the various factors can be
calculated. The difficulty in accurately determining
the values of the various factors is fully recognized;
however, the initial equation-balancing process has
proved valuable in establishing reasonable values
for the factors and at the same time allowing the
equation to reflect observed field conditions.
It should be stressed again that the equation is
presently intended for use in situations where an
equilibrium exists between contaminant concentra-
tion and the hydrologic forces. It is recommended
that potential users of this equation make modifica-
tions before applying the concept to contaminant
plumes not yet in equilibrium with the hydrologic
forces. For instance, in many such cases the value
of the concentration of flood-inflow water, Cf,
would not be the cause of future changes. Instead,
the field-measured existing concentration gradient,
Cq ~ Cr, per unit distance could be related to time,
and the basic concept could still be applied.
Modifications required for those applications are
left to the reader.
It is further recommended that users of this
mass-balance equation use a computer for the main
equation and for calculating the more complicated
input factors. Although the calculations may be
done by hand calculator, considerable time is
required both for balancing the equation and for
the predictive iterations.
100 120
RIVER MILES DOWNSTREAM FROM CONTAMINANT SOURCE
Fig. 6. An example of predicted future contaminant con-
centration downstream from the source based on calcu-
lated decay rates for three flushout reaches.
184
-------
REFERENCES
Blaney, H. F., and W. D. Griddle. 1950. Determining water
requirements in irrigated areas of climatological and
irrigation data. U.S. Department of Agriculture, Soil
Conservation Service, Technician Paper 96, 48 pp.
Gruff, R- W., and T. H. Thompson. 1967. A comparison of
methods of estimating potential evapotranspiration
from climatological data in arid and subhumid
environments. U.S. Geological Survey Water-Supply
Paper 1839-M, 28 pp.
Gatewood, J. S., T. W. Robinson, B. R. Colby, J. P. Hein,
and L. C. Halpenny. 1950. Use of water by bottom-
land vegetation in lower Safford Valley, Arizona. U.S.
Geological Survey Water-Supply Paper 1103, 210 pp.
Hellwig, D.H.R. 1973. Evaporation of water from sand, 4—
the influence of the depth of the water table and
particle size distribution of the sand. Journal of
Hydrology, v. 18, pp. 317-327.
Hughes, Gilbert H., and Charles C. McDonald. 1966.
Determination of water use by phreatophytes and
hydrophytes. Journal of Hydraulic Division,
American Society of Civil Engineers, v. 92, pp. 63-81.
Sleight, R. B. 1917. Evaporation from the surfaces of
water and river-bed materials. Journal of Agricultural
Research, v. 10.
Sorey, Michael L., and William G. Matlock. 1969. Evapora-
tion from an ephemeral streambed. Journal of
Hydraulics Division, American Society of Civil
Engineers, v. 95, pp. 423-438.
DISCUSSION
The following questions were answered by John S.
Fryberger after delivering his talk entitled "Quantifying
the Natural Flushout of Alluvial Aquifers."
Ct.How can the equation represent a true mass-balance
without concentration values for P and E?
A. The concentration values, C, for P (precipitation) and E
(evapotranspiration) are assumed to equal 1 and therefore
are not shown. In a strict sense, P and E should have
corresponding concentration values of Cp and Ce. From a
practical viewpoint they were omitted because they are
assumed to equal 1.
Also, it was pointed out during the discussion that P
and E could be omitted from the main equation and used
only in the water-only equation (11). This approach may be
workable but has not yet been fully evaluated.
Q. Would it be possible for chloride-contaminated alluvium
(assuming no further addition of chloride) to be flushed by
a single precipitation event?
A. I cannot invision that a single precipitation event could
have much effect where significant contamination is present.
Q. Have you considered the application ofthermonics
equations to concentration parameters?
A. No.
Q.. Assuming you have determined flood discharge values
from a statistical analysis of discharge records, what data
would you look for to aid in estimating that portion of
flood discharge that enters the aquifer?
A. The most straight-forward method would be to install
water-level recorders in observation wells to measure changes
in ground-water levels during flood events. However, such
historical records are extremely rare and would require years
to accumulate for an ongoing project. The approach used in
our work is as follows: (1) From cross sections across the
river, determine that water depth above river bed required
to flood a significant part of the alluvial flood plain. (2) From
gage-height versus flow records determine that flow that
occurs at the gage height or water depth found by step 1.
(3) From flood records construct partial-duration flow
frequency curves at each gaging station to establish the
frequency of the critical flood determined by steps 1 and
2. (4) Determine the volume of otherwise unsaturated
alluvium that would be saturated by the floods determined
by previous steps. (5) Multiply the flood frequency (for
instance 0.2 floods per year) times t years from equation
(6) times the volume of newly saturated alluvium from step
4 above. (6) For our work we have then assumed a storage
coefficient of 0.2 and also that only 0.4 of the total flood
inflow percolated down to mix with the normal alluvial
water.
In this way we estimated a value for F independent
of the F value calculated by the equations. If necessary, the
least reliable independent variables, G and Cg, and lesser
reliable independent variables, P and E, were adjusted so
that the two methods for determining F were in reasonable
agreement.
Q. How do the precipitation and flood waters mix with
the deep alluvial water, and does density stratification, play
a role?
A. Assuming that the alluvium is not highly and consistently
stratified and that the ratio of vertical to horizontal
permeability is relatively low, then at least some of the
flood water and precipitation inflow will percolate deep,
especially in rivers where the flood plain width is great
compared to the depth of alluvium. Stratification due to
density was most pronounced in those rivers where the
alluvium was relatively deep and the flood plain narrow.
In such cases both the present day and future calculated
flushout rates are low.
The paper by William H. Walker was not available for publication.
185
-------
Improving the Sanitary Protection of
Ground Water in Severely Folded,
Fractured, and Creviced Limestone"
by Elmer E. Jones, Jr. and Charles M. Murray0
ABSTRACT
The construction of wells yielding safe, sanitary
water in areas of severely folded, fractured, and creviced
limestone depends on the retention of enough sediment in
the transmission paths to restrict the movement of con-
taminants. The described research has been directed
towards two problems—the identification of terrain and
hydrologic features that indicate a lack of adequate
sanitary protection of ground water in the formation, and
means of improving this protection. Intermittent spring-
sinks along streams or in low-lying areas indicate a
reversal of ground-water flow in the area. When the water
level rises above a critical level sediments that have
accumulated in the formation over many years may be
discharged in a few days, resulting in a loss of filtering
ability.
Ground-water levels and bacteriological contamination
have been monitored on four wells in Washington County,
Maryland, since 1973. Ground-water quality deteriorated
after tropical storm Agnes washed sediments from the
formation. Techniques for preventing surface contaminant
entry, improving filtering ability of the formation, and
providing controlled relief of hydrostatic pressure have
been studied as ways of protecting ground water in the
formation.
INTRODUCTION
Many local studies have been made on
bacteriological quality of individual well-water
supplies. Normally, about 40% of the wells yield
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
bResearch Agricultural Engineer, USDA-ARS-CFBPE,
Bldg. 228, BARC-East, Beltsville, Maryland 20705.
cResearch Technician, Maryland Environmental
Service, Annapolis, Maryland 21400.
water of unacceptable bacteriological quality
(Whitsell, 1975). In one Tennessee county, 85% of
the drilled wells were contaminated with 4 or more
coliform per 100 ml and 54% contained fecal
coliform (USEPA, 1971). The report stated,
"Generally speaking, quality of well construction in
all three counties is so poor, that one or more
deficiencies threatening the safety of the source
could be found in nearly every well. " Persons
responsible for public health protection, ground-
water development and management must con-
sider: (1) How many wells would yield safe sanitary
water if the best available technology were applied?
(2) Under what circumstances is the best available
technology inadequate?
The 1969 Community Water Supply Survey
(USHEW, 1970) provided a partial answer to the
first question. Of 621 municipal wells studied, 9%
contained coliform but only 2% contained
confirmed fecal coliform. This report stated, "Most
of the wells having total coliform or fecal coliform
densities greater than 4/100 ml were constructed, or
so located as to make contamination of the water a
not-too-unexpected result." Application nationally
of best available technology should result in about
95% of wells yielding water of acceptable
bacteriological quality.
In some areas, currently accepted best
available technology is difficult to define or apply.
Areas with severely folded, fractured, and creviced
limestone have a higher percentage of contami-
nated wells than areas with sandstone or alluvial
aquifers. Many limestones have low primary
permeability, and ground-water movement of
importance is through fractures, bedding planes,
fissures, and solution channels. The sediment
186
-------
Direction of ground-voter flow in wel seoson
Direction of ground- water (low in dry season
High stoge of water table
Low stage of water table
Fig. 1. Profile of a karstified carbonate terrane showing
fluctuation of the water table, the uneven profile of the
water table resulting from differences in permeability, and
the direction of ground-water flow in wet and dry seasons
(LeGrand, 1971).
contained in many of these openings provides
adequate sanitary protection to the ground water
entering the wells. Many of these wells, however,
can be pumped at rates that will remove sediment
from the formation. If enough sediment is
removed, a path for transmitting contaminants
from the surface to the production zone may be
created. Although not commonly done, the produc-
tion zone of wells constructed in severely folded,
fractured, and creviced limestone can be screened
and packed with sand or gravel to prevent loss of
sediment during pumping.
In this paper, we will describe some hydro-
geologic situations where extreme natural hydro-
logic events can remove large amounts of
sediment from a formation and procedures that
may be used to restore or improve the sanitary
protection of ground water.
TECHNICAL BRIEF
Normally, water-table contours are approxi-
mately parallel with land surface contours, but they
change as the water table rises and falls. Generally,
water flows toward and parallel with streams and
rivers. Karst aquifers are less uniform in contour
because of wide differences in recharge and
permeability, but the same conditions are normally
anticipated. The situation illustrated on the left-
hand side of the valleys in Figure 1 is considered
abnormal; in dry weather, the slope of the water
table is opposite the slope of the land. When the
water table is high, however, ground water flows
to the nearest stream. This reversal of ground-water
flow has been documented by LeGrand and
Stringfield (1971), but the deleterious effects on
ground-water protection have not been described.
A severely folded, fractured, and creviced
limestone presents an extremely complex three-
dimensional flow network. The network is difficult
to characterize because of the wide variability in
the width, spacing, and orientation of fractures
and solution channels, as well as in the size,
distribution, consolidation, density and bridging of
the filling sediment. But regardless of its complexity,
any given part of the formation will have a path of
least resistance. If the overburden is an expansive
clay soil that when saturated serves as an aquiclude,
the hydraulic gradient to discharge points, or
openings, may be quite high as a result of artesian
pressure. The openings may be old wells; spring-
sinks; worm, insect larvae, or crayfish holes; or
recent borings by man.
The development of a spring-sink contaminant
transmission path in a karst formation initially
having adequate filtration involves a series of
complex events. When an opening is made through
the aquiclude, discharge begins. The capacity of
the upward flow of water to transport sediment
is not necessarily limited by Stokes Law because
the sediment is nonhomogeneous. Silt and clay in
suspension increase the apparent fluid density,
making it easier to suspend sand which in turn
makes it easier to suspend gravel. Suspended
particles too large to be transported will move to
the side walls. As sediment is removed, hydraulic
resistance is reduced and sediment can be trans-
ported from deeper or farther in the formation.
The discharging flow path must develop in the
existing fractures, bedding planes and crevices, and
must obey the laws governing sediment transport
and hydraulic resistance. The mass of the largest
particle that a flowing fluid can move is propor-
tional to the sixth power of the velocity of the
fluid near the particle. The hydraulic resistance of a
pipe or closed channel is inversely proportional to
the fifth power of the diameter or thickness of the
pipe or channel. For horizontal flow sections, a
channel will develop along the roof of the
section and its capacity will be limited by the
sediment transport capacity of the vertical segments.
Thus, in a sediment-filled solution channel that is
several square feet in cross section, the active pipe
flow area may be less than one square inch. The
flow path, to the extent permitted by the forma-
tion, will develop toward the area of highest head.
An open well bore is a low resistance path through
the formation that may interconnect many cracks
and crevices, thus becoming a point of high head.
When a well is in a discharge path, with water
entering the bore at one or more points and
leaving the bore through one or more paths, the
water may be extremely turbid but free of
bacterial indicators.
As water levels decline, discharge from spring-
sinks will cease and the flow path will be able to
187
-------
accept recharge. How serious the hazard for ground-
water contamination might be depends on the
nature and availability of nearby contaminant
sources, the availability of a transporting medium,
and, the capacity of the contaminant transmission
path. If we assume the spring discharge point was
in or near a clipped pasture, runoff would include
manure with fecal organisms, grass clippings,
bacteria and soil particles from soil erosion and
atmospheric contaminants associated with
precipitation. The recharge influent could contain
10s to 109 coliform per 100 ml.
Organic debris and soil particles may lodge in
the contaminant transmission path and develop a
sufficient filter to reduce or eliminate bacterial
contamination for a time. As the organic matter
decomposes, however, the carbon dioxide produced
accelerates the dissolution of adjacent limestone.
Also as the water level drops, sediment in cracks
and crevices may dry, shrink, and compact, once
more opening the contaminant transmission paths.
Thus, the gradient from any recharge point, sink or
doline to the water level in a well increases. This
permits increased flow velocity and sediment
transport capacity. Sediment may accumulate in
the well, move out into other flow paths, or be
pumped from the well.
The rate of recharge and rise in the water
table is important in determining the nature of
the discharge from an intermittent spring-sink.
A rapid rise of 10 feet or more that results in
discharge will remove much more sediment from
the formation than will a gradual rise over
several weeks that gives clays an opportunity to
expand and sediments a chance to settle before
discharge begins.
Intermittent spring-sinks will not normally
persist in stream bottoms. Bed load sediments
having a fall velocity greater than the discharge
velocity of the spring will enter the outlet. As
discharge declines, progressively finer sediments
can enter. When discharge ceases and recharge
begins, the entire bed load contained in the
recharge flow may enter. When the stream bottom
is in a gaining mode, bottom areas that have
been through this process may be identified by the
presence of coarser bottom sediments. When the
stream bed is dry, such areas may be noted by the
tendency of the thin clay-silt sediment layer to
dry, crack and curl over the coarser sediments
which function as a capillary barrier.
When a site is identified where reversal of the
ground-water gradient is removing filtering
sediment and allowing contaminants to move
rapidly through the formation, the ground water
may need improved protection. Identifying and
closing the spring-sinks normally will not produce
permanent improvement; pressure will increase and
discharge will probably develop at another site. A
better solution is to provide an engineered path of
least resistance that can discharge ground water to
reduce the hydraulic gradient on the spring-sink
area and prevent the entry of contaminants.
If the proper sand-water ratio and fill rate are
maintained, large quantities of sand can be placed
in the formation through sinks. When discharge of
sediment is halted, nature will fill the flow paths. If
contaminant transmission paths end in or pass
through a well, a filter and screen may greatly
improve water quality.
FIELD STUDY
The field study reported was conducted on
farm WA10. Well WA10A was one of 24 wells
being monitored for pesticides in the Hagerstown
Valley, Washington County, Maryland. Farm WA10
was selected for detailed study because it appeared
that pesticides might be entering Well A below
the water table. Details of early field work have
been published (Jones, 1973-74).
The Hagerstown Valley is part of a broad
synclinal structure known as the Massanutten
synclinorium. It includes the area between South
Mountain on the east and Fairview Mountain on
the west. The rocks in the valley are highly
contorted, fractured, and contain many faults
and folds.
Farm WA10 is located on the Beekmantown
Group, which is made of three formations:
Stonehenge limestone, Rockdale Run formation,
and Pinesburg Station dolomite.
The basal formation, the Stonehenge lime-
stone, outcrops in narrow belts over a wide area of
the Hagerstown Valley. It is composed of a large
lower algal limestone member, and a thin-bedded
upper mechanical limestone member. The Stone-
henge formation is about 750 feet thick in the
Valley (Slaughter and Darling, 1961; Sando, 1957).
An east-west geologic cross section of the Valley
is shown in Figure 2. A partial map of the farm
area is shown in Figure 3. Many small spring-sinks
are around the edges of area F and along the
stream.
Major operations or occurrences in the field
study are listed in Table 1. Table 2 provides a brief
description of the wells. Wells A and B were
grouted and cased by the sinking shoe technique,
using Type III A cement with a 60% water/cement
188
-------
SOUTH MOUNTAIN
Fig. 2. East-west geologic cross section of Hagerstown
Valley. Farm WA-10 is east of the Conococheague Creek
area.
ratio. Well A had a contaminant entry point 79
feet below grade, which leaked air at about 2 gpm
(Jones, 1973). After open bore grouting, the bore
tested airtight to 100 feet at 35 psi. From initial
testing, it appeared that all contaminant entry had
been stopped.
Before reconstruction of Well A, chlorine
demand fluctuated widely and sometimes exceeded
chlorinator capacity when using standard laundry
bleach. Since reconstruction, chlorine demand has
not varied more than 1 ppm, the demand equivalent
to a change in ammonia concentration of 0.1 ppm.
Coliform were detected in Well A on March
23, 1973 with 12/100 ml at 10 minutes pumping
time. Subsequent counts were 80 on April 2, 990
on April 25, and 700 on April 27. Because nine
months had passed since tropical storm Agnes
Table 1. Major Operations or Events in Field Study,
1969 - 1976
May 1969 Well B constructed, cased and grouted to
100 feet.
May 1970 Evidence of major surface infiltration
noticed at Well A.
Feb. 1971 Well A reconstructed, cased and grouted to
60 feet.
Jan. 1972 Well A open bore grouted to 100 feet.
June 1972 Tropical storm Agnes, 8.1 inches rain.
May 1973 Dye test of septic tank drainage to Well A
negative. Gradient from Well D to Well A
observed.
July 1973 Dye test Well D to Well A.
Aug. 1973 Large sink hole filled 500 feet northeast of
WellD.
Sept. 1973 Well D casing set, 2.25 yards cement grout
and 10 tons sand placed.
Nov. 1973 10 tons sand placed Well D. Sand recovered
from Wells A and C. Well A returned to
service.
Jan. 1974 Major ground-water discharge at site known
as Old Well, northeast of Well D.
Feb. 1974 Sand fill of Well D completed.
July 1974 Dye test Old Well site to Well D. Relief well
installed at Old Well site.
April 1975 Well C reconstructed.
Sept. 1975 Tropical storm Eloise, 8.7 inches rain.
Dec. 1975 Relief drain line installed. All samples Well C
negative for coliform.
April 1976 Well A negative for coliform, 2 consecutive
weeks. First negatives since March 1973.
occurred, it was difficult to initially associate these
results with that storm.
Water levels in the wells from June 1973 to
January 1976 are shown in Figure 4. Coliform
concentration for Wells A and D and weekly
precipitation are shown for the same period in
Figure 5. Water levels began their normal summer
decline in June 1973, and coliform counts
increased. That a major contaminant transmission
path had developed, was obvious. Results of a dye
test through the septic tank in May 1973 were
negative. However, because water level recordings
Table 2. Description of Farm WA10 Wells A-D
Fig. 3. Partial map of Farm WA-10 showing Wells A,
original drilled well; B, new drilled well; C, dug well; D, old
dug well; E, cistern; F, intermittent spring-sinks; G, old
well site.
Surface Elevation1
Depth, feet
Production Level,
feet
Diameter, inches
Specific Capacity,
gpm/ft
A
99.6
131
127+
6V4
6.3
B
101.8
198
173
5
0.07
C
100.5
32
p
42
>10?
D
94.6
32
?
48
>30
1 Elevations are from an arbitrary datum point of 100.00
feet.
189
-------
N 0 ' I F M A M I I A S 0 N D r I
N D [ I F M A M ) I A S
Fig. 4. Well water levels. Farm WA10, June 1973 to January 1976.
showed a gradient from Well D to Well A, a dye
test was performed. The rapidity with which the
dye traveled 160 feet from Well D and penetrated
about 90 feet deeper to enter the bottom of Well
A and the sharpness of the peak obtained
(Figure 6) were surprising.
The stream north of the farmstead has a
nearly level bottom from the ford northwest of
the garage for 300 feet east and northeast. In July
1973, all water in the stream was entering a
spring-sink hole north of the stream channel
about 500 feet northeast of Well D. The sink
hole was filled with 1,500 pounds of sand and
3 bags of cement, which caused water to flow
down the stream and back up into the spring-sink
area F This apparently caused the coliform count
to be temporarily higher in Well D than in Well A
in late August.
The age and history of Well D are uncertain,
but it is probably more than 100 years old. Much
of the area between Well D and the spring-sink
area F probably is stone fill, since many stones
are removed from the fields after each plowing.
The well, approximately 48 inches in diameter
with stone curbing over 2 feet thick in places, is
32 feet deep and terminates in a predominantly
Fob Mar Apr May June July Aug Sapl Ocl Nov Dec
Fig. 5. Bacteriological data for Wells A and D with associated precipitation.
190
-------
calcareous sediment-filled solution channel or
crevice about 5 feet deep. In 1936, the well was
nearly dry and a black walnut crib about 2X3 feet
was sunk to the bottom of the sediment to obtain
water.
The sediment consists primarily of fine-to-
medium sand-size calcite particles. Also present
are larger pieces of extremely porous (rotten)
limestone, with sand and soil particles and some
large organic debris. The upper few inches are
predominantly organic. This organic layer may
have been responsible for the wide variation in
coliform counts obtained with suction samples—
500/100 ml 1 foot below the water table and
5,000,000/100 ml 1 foot above the sediment.
In reconstructing Well D, we first tried to
remove the bottom sediment, but sediment seemed
to be entering the well as fast as it was removed.
Therefore, we mixed pea gravel in the calcareous
sediment by using air from a 100-cfm compressor.
A slotted 5-inch casing was jetted and driven to the
bottom of the sediment. A 2-inch tell tale was set
alongside the casing and 2.25 yards of ready-mix
grout was placed over the sediment by tremie.
After the well was producing an acceptable
yield, we began to backfill the upper bore with
sand dumped directly into the bore from a dump
truck. About 30% of the first 10-ton load was lost
into the formation. Figure 7 shows the condition of
the bore 4 days after placement. A small stream of
£
u°
< 4
o
o>
£ 1
«
X
a
fc? 1
A
0 60 120 180 240 300 360
Minutes After Injection Began
Fig. 6. Results of dye test Well D to Well A. Cj dye con-
centration in Well D, Co dye concentration in water
samples from Well A.
Fig. 7. Sand in bore of Well D, 4 days after placement.
water was flowing over the sand from the north
and exiting through the small hole at the south.
Normally, this flow would have reached the water
table through the curbing without being noticed.
We washed as much sand as possible into the
formation to obtain relatively stable conditions
before dumping in more sand. After the second
load of sand was placed in the well, sand was
recovered from Wells A and C.
Membrane filter techniques were used for the
bacteriological examination of well water. For
colony counts of more than 20 per 100 ml, efforts
were made to obtain counts between 20 and 50
colonies with a reproducibility of ±2 colonies.
In counting colonies, the sediment also was
examined with a low-power microscope. Occasion-
ally, water samples were filtered solely for
sediment examination.
From the time sand was placed in Well D in
November 1973 to the end of February 1974,
coliform counts declined. The ground water had
reversed flow and was flowing from the well sites
to the spring-sink and stream area. A site about
50 feet east of the spring-sink area, known as the
Old Well began discharging water, perhaps as much
as 200 gallons per minute. When discharge ceased
at the Old Well and area F, coliform counts soon
increased in Wells A and D. Heavy rainfall in early
April caused a sharp rise in water tables and a
decrease in coliform counts. When water levels
declined in May, coliform counts again increased.
When water levels in Wells D and A were about
1 foot above stream level (91.25 feet arbitrary
191
-------
datum), spring-sinks were discharging and coliform
counts were low. When well-water levels were lower
than this, coliform counts increased after heavy
precipitation, as shown in Figures 4 and 5.
Backfilling Well D with sand increased the
hydraulic resistance in and near its bore, thereby
increasing discharge at other points. A relief well
was installed at the Old Well site G to permit
ground-water discharge and prevent surface-water
entry, but it did not function as desired. The 2 to 3
feet of sand and mud heaved into the casing by
rapid water-table rises had very low permeability
after settling. Hydrostatic heads 2 feet above
ground surface were measured through the sediment
in the relief well.
When water levels rose rapidly, as in April
1974, water in Well C became extremely turbid.
Water entered the bottom of the bore, rose, and
exited through the porous curbing, probably
discharging to the stream northwest of the house.
Although extremely turbid, the water was free of
coliform and fecal strep organisms. The appearance
of sand from Well D in both Wells A and C indicates
a common flow path much of the way. It was
suspected that Well C would respond rapidly to
rainfall, with its water level rising sufficiently to
cause water to flow to Well A. Discharge of
sediment from Well C was regarded as especially
undesirable because Wells A, B, and D normally
had positive gradients to it.
In reconstructing Well C, 1 foot of 5-inch PVC
screen with 0.040-inch slots and an open area of
about 19 square inches per foot, was used with
schedule 40 PVC casing. This was set in the well by
hand. After setting, the casing and screen were
weighted internally with 150 pounds of steel pipe
to stabilize them while gravel and sand were
placed. The 3,800 pounds of bagged No. 3 well
gravel, 1.7 to 2.0-mm size, were placed around and
above the screen. The bucket of the front-end
loader used to transport sand carried about
400 pounds. After every 3 to 4 buckets of sand
were placed, the fill level in the well was measured
at several points. The average fill depth is shown in
HECONSTRUCTION OF WEIL WAIOC
10 20 30 40 50 60 70 80
THOUSANDS OF POUNDS OF FILL
Fig. 8. Fill material use in reconstructing Well C. Fill
requirements varied from 1,000 pounds per foot to over
3,000 pounds per foot.
Figure 8. The sand hitting the water, surged the
water moving sand into the formation. After the
sand fill was above the water table the sand was
washed and jetted into the curbing and formation.
The water level was about 21 feet above the well
bottom.
Coliform count for Well C has been negative
since December 1975, except for 2 samples having
1 coliform per 100 ml (Figure 9). Tropical storm
Eloise had little effect on bacteriological quality of
Well C.
In 1975, ground-water levels were extremely
high. Spring-sinks fluctuated between discharge
and recharge several times. When coliform counts in
Well A would exceed those in Well D could be
predicted from water levels in Well D. The critical
elevation was 92.3 feet. This was confirmed by dye
tests from Well D to a seep in area F. Because
reversal of the ground-water gradient was
responsible for opening the contaminant trans-
mission paths, ways to reduce or eliminate the
hydrostatic pressure on the spring-sink area were
> *• ^
5 o • 2
o>
o
1
^
\ L A.. /
V^7
L ,
July Aug Sept Oct Nov
k
\
\
k/\
^ -,
WA
OCf
— \
V
lee
onstrui
ted
V/
s
^
Zero |
\/
terlOO
A
m\-sr
Dec Jan Fob Mar Apr Mav'Juna Jul
-------
considered. The yard southeast of Well D often
had standing water at high-water periods. This, in
effect, was a reservoir area exerting a hydrostatic
head of 1.7 to 2.7 feet of water on the spring-sink
area.
In December 1975, a 200-foot-long, 4-inch
drain line was installed. The first 100 feet from
the creek to the fence west of the tool shed was
installed in "solid" rock. In this rock, 4 to 6 inches
of sand fill were placed under the fiberglass-
jacketed, corrugated polyethylene tubing. In soil,
the tubing was placed on 12 to 18 inches of sand
fill and covered with 8 inches of sand. The sand fill
provided over 400 square feet of surface area below
the drain line in soil. The drain line has discharged
for only a short time since installation. However,
well points near Well D and the drain line indicate
that the drain has reduced water levels in the sand
fill at Well D.
DISCUSSION
Wells constructed in severely folded, fractured,
and creviced limestone depend on sediment in
cracks and crevices to protect water quality. In
areas where the ground-water gradient is subject
to reversal, major storms can remove in a few days
the sediment accumulated over many years. Effects
of this loss may not be apparent until ground-water
flow returns to normal.
Differences in water level between the wells
studied cannot be equated with hydraulic resistance.
Differences in recharge, discharge, and hydraulic
control sections may have major effects. However,
when the spring-sink areas are discharging, the
smaller differences in water level between Well D
and Wells A, B, and C are attributed to decreased
hydraulic resistance due to the removal of sediment
from the formation. Water levels from December
1974 to March 1975 demonstrate this trend.
For sediment to be transported from the
formation, the sixth power law of sediment
transport and the fifth power law of hydraulic
resistance must be satisfied. This results in the
development of flow paths that have uniform
characteristics, as shown by the dye test in
Figure 6.
In reconstructing Well D, we made several
mistakes. The well should have had a screen and
sand pack. Placement of the grout over the
calcareous sediments was a serious mistake. Once
set, the grout could not move and prevented the
sand fill from settling towards the production
zone but served as a roof over unstable sediments.
Use of clean concrete sand may be a mistake in
some cases, because larger particles may bridge
and retain finer particles that could move further
into the formation.
Based on the results obtained in back-filling
Wells C and D with sand, placing sand in the well
in smaller quantities with a front-end loader is
recommended over dumping sand directly into the
well from a truck. It is believed that two factors
are involved: the available energy per pound of
sand to transport it through the curbing and into
the formation, and the difficulty of working
through 6 to 8 feet of sand. A pipe water jet could
be used to settle sand in Well D more than a year
after placement. Some of the difference in success
in backfilling the two wells was undoubtedly due
t-o differences in formations.
Major spring-sinks should be prevented from
recharging grossly contaminated water. If it is
desired to retain them as discharge points, extreme
care must be taken to avoid disturbing their natural
development. Several small water boils developed
within 6 feet of the relief well. The burrowing by
earthworms, crayfish, and other soil organisms
reduces the ability of heavy soils, which function
as aquicludes and are periodically subjected to
positive hydrostatic heads, to adequately protect
ground water from contamination.
The results obtained from Well C indicate
that it is most economical to provide filtration
immediately adjacent to the well. Its depth
compared to Wells A and B raises questions about
the trend of regulations to require deeper casing
and grouting as the primary means of providing
adequate sanitary protection.
SUMMARY
In the field study described, reversal of
ground-water flow in an area of severely folded,
fractured, and creviced limestone overlain with a
heavy soil capable of serving as an aquiclude
removed large amounts of sediment from the
formation. The loss of filtration and reduced
hydraulic resistance due to loss of sediment
permitted rapid movement of surface contaminants
when recharge began. Three remedial measures
were investigated: improving the filtering ability
of the formation, providing sanitary protection to
major discharge points, and relieving the hydro-
static pressure on spring-sink areas. Best results
were obtained by improving the filtration in and
around Well C. The relief drain installed in
December 1975 cannot yet be properly evaluated.
If successful, it will reduce sediment loss from
and contaminant entry into the formation.
193
-------
ACKNOWLEDGMENTS
We thank the many people who assisted us
with the planning and implementation of field work
and testing and evaluation of results. We especially
thank the farm family, without whose permission
and cooperation the field studies could not have
been made.
REFERENCES
Jones, E. E. Jr. 1973. Well construction helps determine
water quality. J. Env. Health. 35(5):443-449.
Jones, E. E. Jr. 1974. Evaluating well construction. J. Env.
Health. 36:6.
LeGrand, H. E. and V. T. Stringfield. 1971. Water levels in
carbonate rock terranes. Ground Water. 9(3): 4-10.
Sando, Wm. J. 1957. Beekmantown Group, (Lower
Ordovician) of Maryland. Geological Society of
America Memoirs 68.
Slaughter, T. H. and J. M. Darling. 1961. Allegeny and
Washington Counties Water Resources Bulletin 24.
Dept. of Geology, Mines, and Water Resources,
State of Maryland.
U.S. Environmental Protection Agency. 1971. Evaluation
of the Tennessee Water Supply Program. Bureau of
Water Hygiene Region IV, Atlanta, Georgia.
U.S. Department of Health, Education, and Welfare,
Public Health Service. 1970. Community water
supply study, analysis of national survey findings.
Bureau of Water Hygiene.
Whitsell, W. J. 1975. Rural water supply: how bad a
problem. Chapter 15, Water Pollution Control in
Low Density Areas. Edited by W. J. Jewell and R.
Swan. University Press of New England.
Suggested Reading on Sedimentation and Sediment
Transport
American Water Works Association. 1955. Mixing and
sedimentation basins. R401, pp. 15-36. Revision of
American Water Works Association Water Quality
and Treatment, Chapter 8.
Vanoni, V. A. 1966. Sediment transportation mechanics:
initiation of motion. Progress report of the Task
Committee on preparation of sedimentation manual.
V. A. Vanoni, Chairman. J. Hy. Div. ASCE. v. 92.
(Hy 2) Proc. Paper 4738, pp. 291-314.
Beverage, J. P. and J. K. Culbertson. 1964. Hyper concentra-
tions of suspended sediment. J. Hy. Div. ASCE. v. 90
(Hy 6):117-128.
DISCUSSION
The following questions were answered by Elmer E. Jones,
Jr. after delivering his talk entitled "Improving the Sanitary
Protection of Ground Water in Severely Folded, Fractured,
and Creviced Limestone."
Q. When examining for coliforms in waters having high
turbidity, isn't it recommended that the multiple tube
method be used instead of the membrane filtration
procedure?
A. We use the Multiple Tube MPN method when necessary
and as a quality control check. As stated in the paper, we
regard the examination of sediment particles when counting
colonies an advantage. The dilution required for counts over
1000/100 ml greatly reduces or eliminates the effects of
turbidity. For waters containing high coliform counts and
microscopic organic debris, interference from nuisance
organisms (Standard Plate Count) may be a greater problem
than the sediment. I hope my paper will raise the question
of the long-term protection of ground water when sediments
interfere with the examination of well waters having low
coliform counts, 0-10/100 ml.
Q. Not knowing the geology of the area and depth to non-
bearing water horizons, I'd like to ask, if possible why did
you not plug the old wells, drill a new one, case and grout
below the depth of the affected wells?
A. It is hoped that our research will ultimately be of value
over large areas. This research would not be practical for a
single individual. Well B was constructed during the study.
It also was affected by the extreme high-water levels
associated with tropical storm Agnes. After April 1973,
the coliform counts were less than 5% of those in Well A.
At this time, I cannot recommend either a safe separation
distance or depth of casing and grouting in areas where the
sanitary protection of the formation is reduced by the
removal of sediment from the formation by reversal of
ground-water flow.
When the specific capacity of the wells are compared,
Table 2, it appears that the cost of reconstructing an
existing well of proven capacity may be less than the risk
of drilling new wells that may be also contaminated, low
yielding, or dry. In reconstructing old dug wells such as
Well C and obtaining water of acceptable bacteriological
quality, we are also proving techniques that can be used to
close old wells to protect ground water.
In much of Maryland, 200 feet is regarded as the
practical depth for obtaining wells with adequate yield. In
the Hagerstown Valley, most of the wells deeper than
200 feet have specific capacities less than 0.1 gallon per
minute per foot.
194
-------
Land Disposal of Hazardous Wastes: An Example
from Hopewell, Virginia"
by D. H. Walzb and K. T. Chestnut, Jr.c
ABSTRACT
In early September 1975 a Technical Committee was
established by the Virginia State Department of Health to
clean up and dispose of a manufacturing plant which had
produced the pesticide Kepone, in Hopewell, Virginia. The
Committee recommended that the hazardous wastes be
buried in a virgin section of the new City of Hopewell
landfill. The disposal pit was designed to encapsulate the
wastes in clay and plastic. The project was completed in
early March 1976. The disposal site is monitored by the
Virginia State Water Control Board via an observation well
and an underdrain system.
INTRODUCTION
In late July 1975 the Virginia State Department
of Health was informed that a blood sample taken
from an employee of a pesticide plant in Hopewell,
Virginia was found to contain a significant quantity
of Kepone. A subsequent visit to the plant by
State officials resulted in the immediate voluntary
closing of the facility. The analyses of blood from
other workers and from local residents showed that
the manner in which Kepone had been manufactured
had created a serious health hazard for the entire
community. The independent City of Hopewell,
which has a population of approximately 25,000,
lies at the confluence of the Appomattox and James
Rivers (Figure 1).
Presented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
b Assistant Professor, Ground-Water Resources
Program, J. Sargeant Reynolds Community College, P.O.
Box 12084, Richmond, VA 23241.
cConsultant, Bureau of Solid Waste and Vector
Control, Virginia State Department of Health, 35 Fillmore
Street, Petersburg, VA 23803.
Kepone, a chlorinated hydrocarbon, had been
manufactured at this plant since early 1974. The
Hopewell product then was shipped to Baltimore,
Maryland where bait and/or other inert materials
were added to form the marketable pesticide.
Kepone was used in the United States to control
ants and roaches and in Latin America as a
pesticide for banana pests.
As part of an area-wide environmental and
health assessment of the Kepone contamination, the
Virginia State Department of Health established an
Fig. 1. Map of City of Hopewell, Virginia area.
195
-------
Fig. 2. Physiographic setting of disposal pit. "A" is under-
drain standpipe; "B" is observation well. Elevations in
feet above mean sea level.
Ad Hoc Technical Committee in early September
1975 to formulate a program for and supervise the
operations of the cleanup and disposal of the
Kepone manufacturing plant. The Committee was
composed of the following professionals: (1) two
industrial hygienists (Bureau of Industrial Hygiene,
Virginia State Department of Health), (2) a ground-
water geologist (Virginia State Water Control Board),
(3) a landfill consultant (Bureau of Solid Waste and
Vector Control, Virginia State Department of
Health), (3) a civil engineer (City of Hopewell),
and (5) an epidemiologist (Bureau of Epidemiology,
Virginia State Department of Health). Representa-
tives of the Virginia Air Pollution Control Board
also attended several meetings of the Technical
Committee.
After reviewing several alternatives, the Tech-
nical Committee concluded that the only feasible
way to dispose of the plant and its related contami-
nated materials was to encapsulate the hazardous
wastes in a lined disposal pit. A virgin section of
the City of Hopewell's new landfill site was
selected as a potential disposal area because of its
remote location and good physiographic setting.
DISPOSAL-PIT EVALUATION
The disposal site is located on a small knoll
in the southwest corner of the new landfill area
(Figure 2). The site is drained on the north and
south by intermittent streams. Maximum elevation
of the knoll is 39 feet (11.9 m) above mean sea
level (msl). The natural setting of the site isolates
the hazardous wastes from the municipal and
industrial waste disposal sections of the City
landfill.
Three deep borings at the site indicated the
presence of 28 to 30 feet (8.5 to 9.1 m) of hard
clays containing laterally discontinuous lenses of
clayey silts, clayey sands, and silty sands. Medium-
grained sands lie beneath this upper zone of clays.
These sediments are part of the Coastal Plain of
Virginia.
Water was encountered in the borings at an
average elevation of 5 feet (1.5 m) above msl.
Observation wells drilled in the old City landfill
in March 1976 suggest that the water in the borings
at the disposal site probably represents a perched
water table. Ground-water flow is to the southeast.
The disposal site initially was accessible from
paved streets only via a power transmission line
dirt road. Prior to excavation of the disposal pit,
this road was widened and improved. At present,
the road serves as the entrance to the new City
landfill.
There are no residences within 1000 feet
(305 m) of the site; the area is served by public
sewer and water and is patrolled regularly.
Upon the recommendations of the Technical
Committee, the Hopewell City Council donated this
tract of land for the burial site in late September
1975. The State Department of Health issued a
"Permit to Operate a Landfill" shortly thereafter.
DISPOSAL-PIT DESIGN
Initially, the contaminated materials which
were to be landfilled consisted of the following
(Figure 3):
1. a concrete pad which covered the entire
outdoor working area of the plant;
2. approximately two feet (0.6 m) of soil
beneath the pad;
3. process reactors, tanks, filters, and piping,
chemical storage tanks, and other miscellaneous
material in the pad area; and,
4. internal process apparatuses in the plant's
manufacturing building.
The size of the disposal pit was determined by
estimating the total amount of materials to be
removed from the plant site and multiplying by a
safety factor of 2 (100 percent). Estimates were
received from four sources; the representative
average was 2500 cubic yards (1911 m3). When the
excavation was completed, the pit measured 130
feet by 95 feet (39.6 m by 29.0 m) along the upper
perimeter; 95 feet by 40 feet (29.0 m by 12.2 m)
along the floor. The depth of the pit was 16 feet
(4.9 m) in the center. The calculated total
196
-------
MANUFACTURING BUIIDING
0.
r- O
Z I
s „
DRYERS —
-c.
N
1
0
O
PROCE
TANK
1
oo
55
*l
\
\
-* \
\
s
STO
\
D
TANKS
WELDING
SHOP
O O O
PARKING AREA
SCALE
Fig. 3. Layout of Kepone manufacturing plant. Section
between manufacturing building and office is outdoor
working area.
available volume was approximately 4600 cu yds
(3517m3).
The Technical Committee recommended that
the disposal pit be engineered with the following
safeguards on the floor of the pit (Figure 4):
1. a one-foot (0.3-m) thick layer of compacted
clay in the bottom of the pit;
2. an underdrain system in the top of the clay;
3. a layer of clean sand on top of the under-
drain system;
4. a 30-mil (0.76-mm) thick, nylon-reinforced,
plastic liner (seamed at the factory); and,
5. a protective cushion of three feet (0.9 m)
and one foot (0.3 m) of clean clayey sand on the
liner on the floor and sides of the pit, respectively.
The underdrain system consists of: (1) a
section of 4-inch (10.2-cm) perforated plastic pipe
in a 6-inch (15.2-cm) square gravel-filled trench
graded to the northern end of the pit; (2) two
6-inch (15.2-cm) square gravel-filled trenches
located approximately 25 feet (7.6 m) from each
end of the pit and graded to the main trench; and
(3) a standpipe at the northern end of the pit. If
water is found in the standpipe, samples will be
collected and analyzed. The water then will be
removed and, if contaminants have been found in
the samples, will be treated prior to final discharge.
Clean sand and clean clayey sands were
imported from a nearby quarry. Sediments which
were free from sharp and/or irregularly shaped
objects were needed to protect the plastic liners
from rupture.
The bottom plastic liner was entrenched along
the top edge of the disposal pit; the top plastic
liner (discussed below) was entrenched 2 feet
(0.6 m) to the exterior of the bottom liner. The
liners were left unsealed to allow for the escape of
any gases which might be generated within the fill.
PRE-DISPOSAL CLEANUP AT PLANT
In mid-November 1975 specially trained
workmen began the cleanup operation at the plant.
The office building was cleaned, painted, and
outfitted with showers, dressing rooms, and storage
space for safety equipment needed by the workmen.
Dismantling of the plant began one week later.
Under the supervision of State and private industrial
hygienists, the workers dismantled the plant and
discarded the scrap materials in dumpsters. When a
dumpster was full, it was covered with a tarpaulin
and moved to the Hopewell Sewage Treatment
Plant for safe storage (Figure 1).
This phase of the cleanup was completed in
early February 1976; all that remained of the plant
at this date were: (1) the shells of the office and
manufacturing buildings, (2) the concrete pad,
(3) the wastewater holding tanks, and (4) approxi-
mately 350 55-gallon (280-1) drums which
£
(N -I
0J
Clayey sand
Clayey sand
Clay
Contaminated
Plastic liner
Plastic li ner
Con laminated
" waste Undisturbed clay -
Clayey
sand
Clean sand
-
Clay
Top Seat
Bo!torn Seal
Fig. 4. Cross section through top and bottom seals of
disposal pit.
197
-------
contained miscellaneous Kepone-contaminated
materials (these drums presently are in storage
awaiting final disposal).
All rainwater and cleanup wastewaters at the
plant site were drained to the holding tanks; as
these tanks filled, the liquids were pumped to rail
tank cars and later treated.
DISPOSAL OPERATION
On February 24, 1976, the disposal pit was
ready to receive the contaminated wastes. The
environmental safeguards had been placed in the pit
and the liner had been covered with its clayey-sand
blanket.
All personnel involved with the operation
were instructed by State industrial hygienists on
the use of safety equipment, which included paper
or plastic suits, shoe coverings or rubber boots,
safety glasses, hard hats, and respirators.
In addition to the air samplers at the plant
site, Hi Vol filters were set up by the Virginia Air
Pollution Control Board midway along the
predetermined transfer route and at the disposal
pit (Figure 1). A weather vane also was installed
at the site to aid the personnel in staying upwind
from the contaminated wastes.
At the plant site the concrete pad which
served as the foundation for the outdoor working
area was ripped up. Materials were transferred to
the plastic-lined beds of tandem dump trucks and
covered with tarpaulins. At the disposal site the
loads were dumped along the edges of the pit; the
trucks were swept out and returned to the plant
site.
The equipment at the disposal site included a
small bulldozer and a crane with a 1-cu yd (0.76-
m3) clamshell bucket. The first Kepone-
contaminated materials placed in the pit consisted
of sections of the concrete pad and the soil from
underneath the pad. Concrete slabs were laid flat in
the center of the pit; concrete footings and curbings
were placed along the edges of the pit floor.
Foundation soil was released from the bucket when
still elevated over the pit to take advantage of the
added compaction from freefalling.
After three feet (0.9 m) of the above materials
had been put in the disposal pit, the dumpsters
were brought from storage to the pit area and
unloaded. Filters, small tanks, and other fiber or
metal containers were crushed as well as possible
outside of the pit and then placed in the hole. All
pipes, worm screws, angle iron, aluminum, copper,
and iron sheets were placed horizontally in the pit.
Clay, which was stockpiled during the excavation
of the pit, routinely was mixed with the debris in
the hole to fill voids created by the irregularly
shaped materials.
During the disposal operation, officials from
the Commonwealth of Virginia and from the City
of Hopewell, the owners of the plant property,
and the contractors of the cleanup operations
condemned the cinder-block manufacturing building
at the plant. The building then was demolished
and transported along with two feet (0.6 m) of
foundation soil to the disposal site.
The final stage of the cleanup at the plant
site included: (1) the grading of the area behind
the office building, and (2) the puncturing and
filling of the wastewater storage tanks.
DISPOSAL-SITE COMPLETION
After all the contaminated materials had been
placed in the hole, the pile of rubble was spread to
the edges of the pit with a drag line.
To prevent the infiltration of surface water,
the disposal pit was sealed across the upper surface
(Figure 4). The impermeable cap consisted of the
following materials (in order of installation):
1. four feet (1.2 m) of compacted clay;
2. 8 inches (20.3 cm) of clean clayey sand;
3. a 15-mil (0.38-mm) thick plastic liner;
4. one foot (0.3 m) of clean clayey sand; and,
5. three feet (0.9 m) of compacted clay.
The site was graded and sloped to prevent the
ponding of surface water and seeded to prevent
erosion. Signs were erected at the four corners of
the pit to warn against trespassing and excavation in
the area; a bronze plaque affixed to a granite slab
was installed as a permanent identifying marker at
the center of the west face of the pit; and a
statement which identifies the contents of the
disposal pit was attached to the property deed. The
disposal project was completed in early March 1976.
The disposal pit was uncovered from the
beginning of the disposal operations on February
24, 1976 until the top plastic liner was installed
on March 6, 1976. During this period, the weather
conditions were excellent; the average maximum
temperature was 76°F (24.5°C); minimum
averaged 48° F (8.9°C). Thirteen-hundredths of an
inch (0.3 3 cm) of rain fell the night before the
plastic cover was installed; the top clay seal,
however, already had been emplaced at this
time. The hazardous-waste materials, therefore,
were essentially dry when landfilled.
198
-------
MONITORING PROGRAM
In September 1975 a 28-foot (8.5-m) deep
observation well was installed downslope from the
site (Figure 2). A number of background water-
quality samples were collected from the well
between October 1975 and January 1976. The
underdrain and the observation well are part of an
area-wide monitoring program by the Virginia
State Water Control Board. During the initial
phase of this program, the collection of water
samples from the well and the inspection of the
underdrain have been made on a monthly basis.
SUMMARY
1. The manner in which the pesticide Kepone
had been manufactured had created a serious health
hazard for the Hopewell, Virginia area.
2. An Ad Hoc Technical Committee, which
was established by the Virginia State Department of
Health, concluded that the only feasible way to rid
the community of the source of Kepone was to
bury the plant and its contaminated contents.
3. A disposal pit at the new City of Hopewell
landfill site was designed to contain the wastes in an
environmentally safe manner.
4. The hazardous wastes were landfilled from
February 24, 1976 to March 6, 1976. During this
period, weather conditions were optimal for this
type of project.
5. The disposal site presently is monitored on a
monthly basis by the Virginia State Water Control
Board via an observation well and underdrain
standpipe.
ACKNOWLEDGEMENTS
The authors wish to extend their appreciation
to the other members of the Technical Committee:
C. R. Fortney and J. A. Saunders, Industrial
Hygienists, F. E. Hughes, Jr., P.E., and R. S.
Jackson, M.D.
DISCUSSION
The following questions were answered by D. H. Walz after A. The liner has a guarantee of 20 years; realistically, the
"'""ring his talk entitled "Land Disposal of Hazardous Committee felt that the liner would serve as a short-term
s: An Example from Hopewell, Virginia." barrier until the materials in the fill stabilized.
Q. Who carried out the disposal operation and who paid for
A. The cleanup and disposal operation was performed by
Private industry at an estimated cost of $350,000 under
the supervision of officials from the Commonwealth of
Virginia.
Q. Do you expect the material to degrade in the course of
time? If not: is this the only feasible way of disposing of
the material without creating environmental hazards?
A. The greatest problem in dealing with Kepone has been
the lack of information available on the response of Kepone
to the environment. We do not know if Kepone will
degrade with time. The major desire of the Committee was
to rid the community of this environmental hazard as
quickly as possible; the Committee investigated several
alternatives but found encapsulation to be the most viable
in terms of time and money.
Q. What type of plastic in the liner? Will organics used in
Kepone manufacture react with the plastic liner?
A. The plastic used was polyvinylchloride (PVC). Prior
to approving the use of PVC, the Committee consulted
with EPA (Washington) and a California firm which was
preparing a report on liners for EPA. Both sources stated
that PVC was an acceptable material for use with Kepone.
Q. What about the cement or bonding agent used to join
the panels of the liner; will the seams be a potential zone
of weakness?
Q. What provisions were made to monitor the quality of
any gases escaping the site?
A. During the disposal operation, hexachlorocyclopenta-
diene (HCP), a raw material in the manufacture of Kepone,
did volatilize. There is evidence from samples collected at
the plant site that HCP bonds to particles because it is an
unsaturated compound and, therefore, is subject to
degradation.
A Hi Vol air sampler was in continuous operation at
the disposal site during the operation. There are no plans,
however, to sample the air quality at the site in the
future unless the situation warrants it.
Q. Is there a risk of spontaneous combustion within the
landfill?
A. No, the landfill is a heat sink and does not contain a
large amount of oxidizable materials.
Q. Was the possibility of reacting the insecticide with some
other chemical compound, one that would result in a less
hazardous product, considered?
A. Yes, but the Committee could not find such a chemical
compound within the time constraints of the project.
Q. What effects does Kepone have on people who came
in contact with it?
A. The acute chronic effects on people consist of neuro-
logical problems (tremors, aberrations of eye movements,
anxiety, and nervousness), sterility, and changes in liver
functions. In tests conducted by the National Cancer
199
-------
Institute, there was evidence of cancer formation in
laboratory animals.
Q. The "60Minutes"presentation mentioned that
concentrations of Kepone were discharged into the James
River. What monitoring has been done in the ground-water
areas near the river and what are the results'?
A. In March 1976 the Commonwealth of Virginia installed
ten observation wells throughout the Hopewell area. In
addition numerous domestic- and public-supply water wells
were sampled in critical areas. Kepone has been found in
wells on the north side of the plant site but only in concen-
trations at or very near (the highest has been 0.08 ppb
[jug/1]) the detectable limit (0.02 ppb [Mg/H ) for Kepone.
Samples from wells near the Sewage Treatment Plant show
levels of Kepone at or below the detectable limit; samples
from wells along the river show no detectable Kepone.
Analyses of soil and core samples show that the
bulk of Kepone has not migrated below the top six inches
(15.2 cm) of soil. It must be noted, however, that these
analyses represent short-term results from a potentially
long-term reaction with the environment.
Q. What about rodent control?
A. Officials from the State Department of Health routinely
check the disposal site. If problems do arise from the burrow-
ing of animals, growth of seedlings, trespassing, or other
unpredictable events, they will be handled immediately.
Q. Has consideration been given to installing a second
observation well in the drainage north of the present one?
A. Ground-water flow in the area is to the southeast; any
possible contamination would show up in the present well
first. Other wells could be constructed if necessary.
Q. How long will it take liquids to migrate through the clay
barrier?
A. No permeability tests were made on the clays; field
evidence indicated that such tests were not necessary.
200
-------
Land and Water Use Impacts on Ground-Water
Quality in Las Vegas Valley"
by Robert F. Kaufmannt
ABSTRACT
Marked changes in the occurrence and quality of
near-surface ground water in Las Vegas Valley, Nevada
result from urban and industrial land and water use
practices. In-valley recharge has increased tenfold in the
period 1943 to 1973 and now amounts to about 40,000
acre-feet/year (49.3 million m3). Ground-water flows
leaving the Valley have increased from 250 acre-feet/year
to about 12,000 acre-feet/year.
Twenty to 400 tritium units (T.U.) in shallow ground
water confirm widespread addition of recent recharge.
Trend-surface analysis of recent water-quality data for
depth intervals or "slices" of 0 to 50, 51 to 100, and 101 to
300 feet (0 to 15.2m, 15.5 to 39.5 m, 30.8 to 91.4m)
revealed that natural trends below a depth of 50 feet are
explainable in terms of broad hydrogeologic conditions.
From 0 to 50 feet quality is highly irregular and markedly
more influenced by land and water use practices and waste
disposal in particular. Chloride, TDS, and nitrate are
particularly diagnostic of return flows as is spring develop-
ment and (or) a rising water table resulting from increased
recharge and low vertical permeability. Statistical tests on
water-quality data for the period 1912 to 1968 yielded
generally insignificant change with time. However, the
aPresented at The Third National Ground Water
Quality Symposium, Las Vegas, Nevada, September 15-17,
1976.
t>Hydrogeologist, Office of Radiation Programs, U.S.
Environmental Protection Agency, Las Vegas, Nevada
89114; work conducted while Associate Research Professor
of Hydrology, Desert Research Institute, University of
Nevada System, Las Vegas, Nevada 89109.
extreme paucity of the data base makes any conclusion
questionable.
More efficient irrigation practices could reduce the
present irrigation water demand by 15,000 acre-feet/year
and reduce return flows by 11,000 acre-feet/year. Return
flows by the year 2000 could easily amount to 75,000
acre-feet/year or about three times the total water budget
of the Valley prior to urbanization. Therefore, ground-
water problems are likely to worsen and, if present monitor-
ing practices prevail, go unnoticed.
INTRODUCTION
The study was designed to determine the
effects of past water use on ground-water quality in
Las Vegas Valley and provide a basis for projecting
both physical and hydrochemical changes which are
likely to result from increasing water utilization in
the area.
Las Vegas Valley is an intermontane arid basin
containing a thick sequence of alluvial sediments
arranged vertically and laterally in a complex
system of aquifers and aquitards. The most
permeable sediments are in the northwestern and
west-central parts of the Valley where the principal
volume of ground-water extraction has occurred
to date, primarily from confined aquifers at depths
of 250 to 1,000 feet (see Figure 1). Except in the
west-central area, these aquifers are overlain by
extensive layers of poorly permeable clay, silt, or
caliche. Separating these productive zones from
generally less permeable sediments to the east is a
series of faults and associated scarps which impede
eastward ground-water flow. The eastern part of the
201
-------
Valley is characterized by primarily fine-grained
sediments and a shallow water table.
Under natural conditions, recharge to the
Valley fill was a result of precipitation in the
surrounding mountains, primarily those to the
west and north. Flow from the recharge areas
followed an easterly or southeasterly path which
subsequently involved lateral and then upward
movement in the valley fill either as diffuse seepage
or as localized flow towards springs discharging
along fault planes. Both the seepage and reinfiltra-
tion of the spring discharges resulted in recharge of
the near-surface zone, discharge from which was by
evaporation and, more importantly, evapotranspira-
tion. There was no pronounced flow of ground
water from the Valley and the principal discharge
areas such as the Meadows (Las Vegas) were well
removed from the present locus of discharge, Las
Vegas Wash.
FACTORS AFFECTING RETURN FLOWS
The principal sources of recharge to the
near-surface system include irrigation return flows,
septic tank and sewage treatment plant effluents,
industrial effluent ditches and disposal ponds,
and upward leakage from deeper, artesian aquifers.
Discharge from the system occurs as direct
evaporation, evapotranspiration from phreatophytes,
and discharge to surface-water courses and to Las
Vegas Wash in particular (see Figure 1). Under
natural conditions prevailing until the early 1940's,
the only ground-water outflow from the Valley was
about 250 acre-feet/year of underflow in Las
Vegas Wash. By 1972 an additional 12,300 acre-feet/
year of underflow surfaced in the lower reaches
of Las Vegas Wash (Kaufmann, 1971; Westphal and
Nork, 1972). Extensive mesquite groves and stands
of saltbush that formerly discharged 25,000
acre-feet/year have largely been removed and
replaced with suburban sprawl and ubiquitous
green lawn. Exclusive of Henderson and the BMI
effluents, recharge to the shallow aquifer in 1973
amounted to 38,500 acre-feet/year (Patt, 1976)
and represented water which was (1) going into
storage, (2) leaking downward to recharge the
deeper aquifers, and (3) flowing laterally toward
Las Vegas Wash.
GROUND-WATER QUALITY
Despite the intense urban/suburban develop-
ment and resulting water use in the Valley, past
documentation of ground-water quality, regardless
of depth, is extremely deficient. Limited discussions
of water quality are presented in Mendenhall (1909),
Carpenter (1915), Hardman and Miller (1934),
Maxey and Jameson (1948), and Malmberg (1965).
The only water-quality map produced prior to the
present study depicted zones or regions of TDS as
indicated by specific conductance. Although an
extensive water sampling program of the District
Health Department resulted in several thousand
water analyses for the period 1968 to 1973, there
was no attempt to synthesize and interpret the
data. Municipal wells of the Las Vegas Valley Water
District and the City of North Las Vegas were first
monitored on a regular basis in 1969.
Trend-Surface Analysis of Water Quality
The trend-surface technique involves fitting
polynomial surfaces to map data by means of a
general linear model incorporating a least-squares
fit of a planar or curvilinear surface to the observed
data. In the past, trend-surface analysis has been
applied primarily to stratigraphic, structural and
sedimentation problems but the method has not
been adequately tested and applied to ground-
water quality studies, particularly those in which
there is a need to reduce and generalize a great
mass of chemical data, some of which are of
questionable veracity. It was reasoned that broad
over-all trends are analytically more useful in
describing variations present in Las Vegas Valley.
Available chemical analyses for ground water
were screened for depth of well, year sampled,
and accuracy of data. Analyses from January 1,
1968 to the present were selected for depth
intervals 0 to 50 feet, 51 to 100 feet, and 101 to
300 feet providing the anion: cation ratio was in
the range 0.9 to 1.1.
First order or regional trends in TDS for the
three depth intervals considered are shown in
Figure 2. At depths of 101 to 300 feet, the
progressive increase in mineralization along the
valley-wide flow path is primarily a function of
chemical quality of water in the recharge zone; the
type, distribution and absorptive capacity of the
geologic matrix; the porosity and permeability of
the rocks and sediments; and the course of water
along the flow path. Differences in absolute
concentrations and concentration gradients for the
three depth intervals are a result of natural and
man-induced factors.
The role of natural hydrogeologic controls on
ground-water quality is shown in more detail in
Figure 3 which is a hand-contoured map of TDS
at depths of 200 to 300 feet. In general, the
pattern for TDS for the zone from 101 to 300 feet
shows progressive increase along the flow path.
202
-------
Areas of major
ground-water
withdrawal
(circled number is
in 100 acre-feet
in 1973)
T
121
S
T
22
S
Water table above
50 feet within
boundary
Direction of ground-water
flow in unconfined aquifers
Fig. 1. Hydrogeologic and cultural features influencing ground water in Las Vegas Valley.
This is expected because the dominant sink is the
lowland area centered on Las Vegas Wash, the
point toward which all ground water flows. The
mineral content of deep ground water in the
southeasterly portion of T21S/R62E rapidly
increases from 1,000 to 3,000 mg/C in a distance
of about four miles. Comparing this with the
change in the distance from the prime recharge area
in the Spring Mountains to the 1,000 mg/S contour
line, (roughly 30 miles or 48.3 kilometers), gives
an indication of the role played by sediment
composition and permeability of fine-grained
valley fill.
Sulfate concentrations in ground water at
203
-------
Fig. 2. First-degree trend surfaces for TDS in ground water.
Fig. 4. Fourth-degree trend for sulfate in ground water
101 to 300 feet deep.
depths of 101 to 300 feet are shown in Figure 4
which is a fourth-degree trend surface showing a
similar pattern to the TDS plot discussed above.
Minimum sulfate concentrations are associated
^s
J
& •
^c •
HENDERSON
Fig. 3. Concentrations of TDS in ground water 200 to 300
feet deep.
with recharge entering the Valley from the north
and northwest, with minor inflow from the
northwest flank of Sunrise Mountain. Inflow from
the west, southwest, and south, particularly the
latter, is enriched in sulfate with the result that
most of the ground water in Paradise Valley does
not meet the U.S. Public Health Service (1962)
standard.
Nitrate and chloride trend surfaces for the
depth interval from 101 to 300 feet had the lowest
coefficients of correlation, indicating numerous
local variations in comparison to broad, Valley-
wide trends in the parameters previously discussed.
In the case of chloride at depths of 0 to 50 feet,
most of the variations are related to effluent
disposal and are superimposed on the first-order,
regional trends shown in Figure 5. The low
concentrations of chloride below 50 feet suggest
prime recharge to the alluvial fill comes from source
area(s) low in chloride and is characterized by short
residence time, or relatively short flow paths, or
both. This suggests recharge to the valley fill may
also be associated with movement in the carbonate
aquifers rather than only in the alluvial aprons
flanking the carbonates.
Rapid urban and suburban growth in Las Vegas
Valley for over 30 years has involved consistent
disruption of natural drainage paths and associated
salt flux processes of the native soils. A typical
204
-------
leveling operation for a subdivision in the central
part of the Valley is shown in Figure 6. For over-
land and open-channel flow, the net effect has been
to collect and channelize flow through urbanized
areas, impede or block sheet flow, and to
reduce the discharge capacity of lesser tributaries.
This had led to localization of runoff and
increased local recharge to the near-surf ace zone.
Similarly, removal of caliche layers and hardpans
combined with the introduction of wastewater
return flows from interfluvial areas, formerly
characterized by sheet runoff and little if any
natural recharge, provides greater opportunity for
increased infiltration and increased salinity in
shallow ground water.
Effects of Return Flows on Water Quality
Nitrate, chloride and TDS concentrations are
particularly diagnostic of return flows beneath
the (1) principal urbanized portions of the Valley,
(2) areas of sewage disposal, (3) areas irrigated
with sewage, and (4) the industrial area in
Henderson. The principal sources of nitrate con-
tamination are shown in Figure 7. Detailed
documentation of water use and return flows
presented in the study by Kaufmann (1976) and
by Malmberg (1965) reveal returns from sewage
effluent, industrial wastes, cooling water, and
septic tank systems have infiltrated the near-surface
Fig. 5. First-degree trends for chloride in ground water.
• • ..
Fig. 6. Deep soil excavation and land leveling associated
with subdivision construction.
aquifer since 1912. In 1955 alone, 7,000 acre-feet
of effluent infiltrated in the eastern part of the
Valley, whereas, total annual recharge to the near-
surface aquifer over the entire Valley was estimated
at 14,000 acre-feet (Malmberg, 1965). For 1973,
Patt (1976) estimated about 18,000 acre-feet. At
the present time, there are 5,591 irrigated acres in
the Valley which receive 47,000 acre-feet/year,
only 42 percent of which is consumptively used.
Clearly, recharge stresses on the shallow system
are pronounced and warrant analysis of their effects
on water quality.
Sewage effluent applied to the Paradise Valley
and Winterwood golf courses since 1960 and 1965,
respectively, has caused little change in TDS, but
nitrate in shallow ground water increased to as
much as 140 mg/C. Phosphate in one well was 8.8
mg/2 or approximately 88 times the normal
background concentration. Sewage effluent from
individual treatment plants associated with several
major hotels and individual septic tanks also
contributed to nitrate concentrations in ground
waters. By 1973 approximately 4,000 septic tank
installations in the Valley, principally distributed
as shown in Figure 7, contributed an estimated
1,750 acre-feet/year of wastes to the near-surface
aquifer (Patt, 1976). Other miscellaneous nitrate
sources include the LDS Church Farm and
adjacent farms, and the sewage treatment plant and
sewage-irrigated golf course at Nellis Air Force Base.
The trend surfaces for nitrate are revealing,
albeit in a negative fashion. Whereas, nitrate in the
zone from 0 to 50 feet is regularly distributed and
quite high in the eastern part of the Valley, at
depths of 101 to 300 feet it is irregularly distributed
and generally quite low in concentration. Therefore,
waste disposal, largely in the eastern part of the
205
-------
LAND DISroSAL OF SEWAGE FROM
MUNICIPAL OK PRIVATE TREATMENT PLANT
Fig. 7. Sources of nitrate in ground water.
Valley, dominates the shallow nitrate pattern.
Although the volume of return flow from such
nonpoint sources is relatively small, the water-
quality effects in terms of elevated levels of nitrate
are readily noticeable.
Temporal Changes in Water Quality
Ground-water quality data from previous
studies, supplemented with analyses from the
present study, were analyzed for temporal changes.
Excluding (1) recent (1969-present) analyses from
producing wells in the Las Vegas Valley Water
District and North Las Vegas well fields, (2) the
water-quality data collected by the Health Depart-
ment or (3) the present study, there are approxi-
mately 412 analyses extending over an area of
150 square miles. These were collected from wells
ranging from 8 to 1,700 feet deep, although 44
percent of the historical water-quality data prior
to 1968 pertain to wells of unspecified or unknown
depth, thereby reducing their scientific value.
Furthermore, only two percent of the available
analyses document very shallow water quality in
the depth interval from 0 to 100 feet, compared
to 22 percent from wells deeper than 400 feet.
The remaining 32 percent of the wells range in
depth from 101 to 400 feet. In township 21/62
and 22/62, deteriorating ground-water quality
from return flows and solubilization of evaporites
from a rising water table is difficult to document
because there are essentially no chemical data for
the unconfined aquifer. Similarly, townships 20/60,
21/60, and 22/60 are devoid of baseline water-
quality data; hence changes in the system, from a
multitude of causes, are and will continue to be
extremely difficult to document. This is unfortu-
nate, considering the rapid urbanization underway
in the eastern portions of the Valley. Although
the data for townships 20/61, 21/61, and 22/61
extend over a period of approximately 45 years,
the paucity of analyses makes it difficult to
define the chemical state of the system at given
points or certain periods. With the exception of
townships 20/61, 21/61, and 21/62, essentially
nothing is known about historical ground-water
quality prior to the period from 1955 to 1962,
despite the fact that since the early 1940's sewage
and industrial wastes were deliberately allowed to
infiltrate in townships 20/62, 21/62, and 22/62.
Wells in various parts of the Valley for which
historical chemical data were available were
grouped as shown in Figure 8. The groups were
selected using criteria of (1) generally similar
hydrogeologic and water-quality conditions within
each group, (2) sufficient number of wells to
constitute a minimal sample size for the statistical
tests, and (3) different time periods in which
Fig. 8. Locations of wells sampled between 1915 and 1972.
206
-------
changes in water quality could be compared.
Within each areal grouping of historical ground-
water quality data, the values for TDS and chloride
were analyzed for change through time by means
of two nonparametric tests, the Mann-Whitney
test and, in cases where sufficient data were
present, the Kruskal-Wallis test for one-way
analysis of variance (Siegel, 1956). The test results
indicated that from 1912 to 1972 essentially no
significant change in ground-water quality occurred
anywhere in the Valley, regardless of the individual
time periods considered.
Return Flows of Industrial Origin
In terms of both flow volume and chemical
concentration, the major industrial waste disposal
operation adversely affecting ground-water quality
is the industrial complex in Henderson. Now known
as Basic Management, Inc. (BMI), the complex
began operations in the early 1940's and initially
produced magnesium metal. Starting in 1945 to
1950, various chemical manufacturing and minerals'
processing industries have been operating the
complex. Throughout the periods of operation,
liquid effluents have been conveyed via open
ditches to tailings ponds where disposal has been
by evaporation and percolation.
The study involved initial assessment of
lithologic boundary conditions and water quality
(Kaufmann, 1971), followed by a combined paper
analog and digital analysis of the flow regime
(Westphal and Nork, 1972). Ground-water inflow
was first estimated by Kaufmann (1971) for
calendar year 1970 by taking into account the
total monthly surface-water flows in the Wash at
(1) the sewage treatment plants, (2) the midpoint
adjacent to the tailing ponds and (3) at the outlet.
Net average daily ground-water return flow in the
upper reach of the Wash is 10.3 acre-feet, much of
which exits the tailing ponds. Lesser amounts are
from commercial irrigation with sewage, and
"other" general return flow from sources such as
lawn watering or golf course irrigation. Daily
accretion in the reach below Pabco Road is
approximately 17.4 acre-feet for a total of 27.7
acre-feet. In contrast, ground-water returns prior
to extensive in-valley water use were about 0.75
acre-feet per day.
Variations in ground-water quality were
determined with respect to lateral and vertical
positions relative to Las Vegas Wash and
particularly in relation to the BMI tailings ponds
and lagoons. Peak concentrations of chloride,
nitrate, and TDS are located in areas extending
from the plant area to Las Vegas Wash. Figure 9
illustrates the pattern for nitrate which is similar
to that for chloride and TDS, although these have
maximum concentrations of 3,900 mg/C and 9,400
mg/£, respectively. It is apparent from these
patterns that pollutants have migrated extensively
from the northern portion of the plant area and
from the tailings ponds. This is confirmed by the
tritium data.
The effects of industrial waste loading were
compared to the mass flux already in the Wash
from other sources. At each of three points in the
Wash, each point was alternately considered as a
"basepoint" where the flux was compared to that
at points upstream or downstream. The differ-
ential (between the points in the Wash or between
the Wash and the tributaries) divided by the mass
flux at the basepoint indicates the contribution
from ground-water or surface-water sources tribu-
tary to the Wash in the reach considered. Using
TDS for example, stations LW048 and LW049
(springs discharging industrial effluent) contributed
107,306 and 126,281 pounds/day (48,717 to
57,332 kilograms/day) on March 11 and 30, 1971,
respectively. This represents 24 and 31 percent
additions of TDS to the Wash relative to the load
present upstream from the influence area of the
ponds. Considering the mass flux in the Wash above
the ponds equal to 100 percent, total contributions
SCALE IN Mll.£S
LEGEND
Nitrite (n NQ3
tsoconcBni ration contour. 100 mg/l
• Location of inalyjts
Fig. 9. Nitrate in shallow ground water in the lower Las
Vegas Wash area.
207
-------
from the tailings ponds are as follows:
Parameter Increase
Na 86%
Cl 118%
NO3 161%
SO4 65%
IDS 86%
TRITIUM AS AIM INDICATOR
OF RETURN FLOWS
Of interest to the present study is the use of
Colorado River water in Las Vegas Valley in the
period 1942 through the summer of 1971. During
this period there was rapid growth in population
and widespread introduction of tritiated River water
for municipal and industrial use. Although the use
of such water increased greatly in 1972, the effects
of this usage were not recognizable within the time
frame of the study.
The effects of tritium concentrations in
precipitation over southern Nevada are considered
negligible. Long-term precipitation records for Las
Vegas Valley indicate that there has been a long-
term deficit in rainfall from 1924 to 1929 and
from 1942 to 1971. Positive departures from the
average occurred only in 1949, 1952, 1955, and
1965. Long-term average precipitation is 4.4 inches/
year (111.8 mm/year) whereas pan evaporation is
about 90 inches. Except along wash channels and
in areas of ponded water, it is unlikely that rainfall
resulted in widespread, in-valley recharge.
Tritium in Colorado River water distributed
within the Valley since 1961 can be roughly
estimated as the straight-line average of the Cisco
and Imperial Dam values (T. A. Wyerman, U.S.
Geological Survey, written communication). It is
estimated that tritium in the Colorado River water
imported into Las Vegas Valley peaked at about
750 T.U. in 1963 and 1964, with a uniform decline
to approximately 200 T.U. in 1973. Concentrations
in the period 1953 to 1963 are little known but
probably are on the order of 10 T.U. or less in
1953 and early 1954. Peaks of several hundred T.U.
may have occurred in 1958 and 1959 followed by a
second period of decline before levels reached a
maximum in 1963.
The most dramatic evidence of tritiated
Colorado River water recharging shallow ground
water is in the area of the BMI complex in
Henderson. Here, tritium concentrations range
from 212 to 385 T.U. in shallow ground water.
This indicates that at least locally the shallow
ground-water reservoir contains the probable
upper limit for tritium concentrations,
considering original concentrations in the Colorado
River and decay of 5.5 percent per year. In other
words, dilution of return flows has been essentially
nil in some areas. In view of recharge from the
tailings ponds, estimated to be 28.5 acre-feet/day
in December 1971, the extremely high
concentrations of tritium indicate little or no
dilution of the return flows.
The tritium data indicate waste discharging
from the ponds to the Wash is isotopically similar
to that from the Colorado River whereas water
from deeper zones along the Wash in the reach
above the ponds contains 3.9 to 5.4 T.U. Thus,
water discharging from depth but upgradient.of
the ponds, contains essentially background tritium
concentration and is unaffected by industrial
wastes. The influx of industrial effluents entering
the Wash transforms the underflow of the Wash in
the downstream direction from older, largely
natural water discharging from the Valley, to that
which is predominantly young and from the
Colorado River.
Tritium concentrations in shallow ground
water beneath the urban and suburban portions of
Las Vegas Valley average about 50 T.U. and are
considered indicative of return flows in the near-
surface zone. Concentrations are approximately
two times higher in the eastern half of the developed
area where Colorado River water, in various degrees
of dilution with deep ground water, has been
delivered since 1955.
MANAGEMENT IMPLICATIONS
The shallow ground-water system is complex
in terms of existing character, past and present
impacts of water development, urban land use,
and various types of wastewater disposal and
return flow impacts. Considering projected
water-supply demands (350,000 acre-feet/year, or
over two times present water use) and the
hydraulic and water-quality responses documented,
there is every reason to believe that shallow ground
water will become increasingly relevant to basin-
wide water management. To date, wastewater
management in Las Vegas Valley has resulted in
the widespread contamination of shallow aquifers
and relatively rapid emergence of return flows
either in surface-water courses or as a contaminated
veneer at the water table. Past disposal of
industrial and sanitary effluent to unlined ponds
near Henderson has been particularly unsuccessful.
Although at least 230,000 acre-feet of water were
disposed of to date, storage calculations show
208
-------
retention of only 3 to 6 percent of the amount
infiltrated, which is conservatively estimated at
40 percent. Thus, true waste containment has
been minimal and the disposal scheme is an
environmental failure insofar as wastes have
migrated extensively into adjacent surface-water
and ground-water resources over an area of about
16 square miles. Noticeable reduction in flow
volume and salt flux is expected to take years, and
several decades are necessary before marked change
in the total salt flux from ground water occurs
after the disposal practice is terminated.
Recharge to the shallow aquifer in the Valley
proper has clearly increased with time. Annual
accretion in 1943 was 3,800 acre-feet compared to
12,700 acre-feet in 1958, 25,600 acre-feet in
1965, and 38,500 acre-feet in 1973. Recharge to the
shallow system now exceeds the total water
budget for the Valley prior to settlement. Ineffi-
cient lawn watering accounts for much of the
recharge to the near-surf ace zone. If present water
and land use patterns prevail, recharge associated
with a future population of 750,000 may approach
75,000 acre-feet/year, or roughly three times the
total water flux prior to development. More
efficient irrigation could reduce the present water
demand by 15,000 acre-feet/year and reduce return
flows by 11,000 acre-feet/year.
Several policies can be incorporated in
management of the shallow ground-water zone:
I. Wastewater disposal (this has been the
actual practice in many parts of the Valley for
industrial, municipal, and domestic wastewaters).
II. Municipal or domestic water supply (this
too, has been actual practice up to the present time
in extensive areas of the Valley).
III. Supplemental water supply, for irrigation
and industrial purposes (where quality permits and
economic development is feasible).
IV. Stabilize, reduce or eliminate saline ground-
water effluent to Las Vegas Wash.
The above policies or combinations thereof
illustrate possible management directions. Each of
these will require tailored monitoring programs as
an integral aspect of return flow management.
Optimally, the policy and management approach
should integrate both water supply and wastewater
disposal within the physical framework of the
resource, as outlined herein and in relation to the
objectives and constraints posed by Colorado
River-Lake Mead salinity/eutrophication problems.
ACKNOWLEDGMENTS
The author is most grateful for the assistance
received from many associates within the Water
Resources Center, Desert Research Institute.
Particular recognition is owed to Messrs. Nate
Cooper, Herbert N. Friesen, John Sanders, M. J.
Miles, Robert J. McDonald, Ralph Patt and James
Dinger for their assistance in data collection and
reduction. Critical advice and review was gener-
ously provided by Drs. George B. Maxey, Gilbert
F. Cochran, Martin D. Mifflin, and Mr. Dale
Schulke. The research was funded largely by a
grant (project R800946) from the U.S. Environ-
mental Protection Agency. Particular gratitude
is owed to the project officer, Mr. Fredric Hoffman.
REFERENCES CITED
Carpenter, E. 1915. Ground water in southeastern Nevada.
U.S. Geological Survey Water-Supply Paper 365.
pp. 31-41.
Hardman, G. and M. R. Miller. 1934. The quality of the
waters of southeastern Nevada. University of Nevada,
Agricultural Experiment Station, Bulletin 136. 62 pp.
Kaufmann, R. F. 1971. Effects of Basic Management, Inc.
effluent disposal on the hydrogeology and water
quality of the lower Las Vegas Wash area, Nevada.
Center for Water Resources Research, Desert Research
Institute, Interim Progress Report to U.S. Environ-
mental Protection Agency on Project No. 13030 BOB.
176pp.
Kaufmann, R. F. 1976. Land and water use effects on
ground-water quality in Las Vegas Valley. Final
project report to U.S. Environmental Protection
Agency under Project R800946. Water Resources
Center, Desert Research Institute. 240 pp.
Malmberg, G. T. 1965. Available water supply of the Las
Vegas ground water basin, Nevada. U.S. Geological
Survey Water-Supply Paper 1780. 116 pp.
Maxey, G. B. and C. H. Jameson. 1948. Geology and water
resources of Las Vegas, Pahrump and Indian Spring
Valleys, Clark and Nye Counties, Nevada. Nevada
Department of Conservation and Natural Resources.
Water Resources Bulletin No. 5. 121 pp.
Mendenhall, W. C. 1909. Some desert watering places in
southeastern California and southwestern Nevada.
U.S. Geological Survey, Water-Supply Paper 224.
Patt, R. O. 1976. Las Vegas Valley water budget: relation-
ship of distribution, consumptive use, and recharge
to shallow ground water. Final project report under
Project R800946. Water Resources Center, Desert
Research Institute.
Siegel, S. 1956. Nonparametric statistics for the behavioral
sciences. McGraw-Hill Book Company, Inc., New
York. 312pp.
U.S. Public Health Service. 1962. Drinking water standards.
Publication 956. 61 pp.
Westphal, J. A. and W. E. Nork. 1972. Reconnaissance
analysis of effects of wastewater discharge on the
shallow ground-water flow system, lower Las Vegas
Valley, Nevada. Center for Water Resources Research,
Desert Research Institute, Project Report 19. 36 pp.
209
-------
Bull Session on Predicting Physical and Chemical Alteration
of Land-Treated Wastewater, and Land Disposal of Sewage
Session Moderator: Jay H. Lehr, Executive Director, NWWA,
500 W. Wilson Bridge Rd., Worthington, Ohio 43085.
Jay H. Lehr, Moderator: Let's introduce the speakers on the
panel: Leland L. Mink with Boise State University and on
leave of absence with EPA here in Las Vegas at the present
time; Donald Runnels, University of Colorado, Boulder;
Herman Bouwer, U.S. Water Control Laboratory, Phoenix,
Arizona; C. W. Fetter, Jr., University of Wisconsin at
Oshkosh; John Mann from Los Angeles; Nils W. Johnson,
U.S. Forest Service, Hiawatha National Forest, Escanaba,
Michigan.
Rojer Clissold, Hydrogeological Consultants Ltd.,
Edmonton, Alberta, Canada: Just one question to Nils. I
was wondering if you had any indication about boron as a
limiting factor in the beautification of the lake?
Nils Johnson: No. We have not found boron to be a limiting
problem in our lake. Within the upper peninsula there have
been quite a few limnological limitations showing that
phosphorus is the limiting nutrient, and we have done
nitrogen and phosphorus ratios and have determined that
phosphorus is indeed limiting.
W. Roger Hail, W. A. Wahler and Associates, Palo Alto,
California: I guess my question is to John Mann. Has there
been any research on drainage rates of materials of low-
field capacity?
John Mann: There have been some models on unsaturated
flow, usually in the wetter climates. I think we have a big
communication problem here; at least I do with one
gentleman from Maryland who suggested that the tensions
could develop in zones below the soil zone to 30 and 40 cm
with only gravity operating, and I have a great deal of
difficulty seeing how you get beyond field capacity which
is generally considered about 10 cm without having some
other mechanism coming into play. This is an area where
there has not been much work done.
W. Roger Hail: That's been my experience.
Herman Bouwer: Maybe I can comment a little bit on this.
This is being researched and worked on really for many
years by the soil physicists, and what governs the rate of
dominant movement in an unsaturated flow is the hydraulic
activity, which is the relationship between hydraulic
activity and either tension or water content. You can
plot it either way.
Now, when you go to water tension below field
capacity, the hydraulic activity is usually several log cycles
smaller than its saturation, so for most field soils, say a
few cm or a few dm per day below field capacity, you
begin to talk about fractions of mm, and in the wetter
zone then, in the saturated zone, you can extend unit
gradient so that the Darcy velocity, the Darcy downward
velocity, is about equal to the unsaturated hydraulic
activity. That's a fraction of a mm a day. Your microscopic
velocity will be larger because the wetter cross section is less
than the total cross section, so maybe that's three or five
times larger than your saturated hydraulic activity, or you
talk about maybe on the order of a tenth of a mm per day
microscopic velocity; so let's say a tenth, that would be a
foot per year, it would still be moving at the lower rate.
W. Roger Hail: Relative or percentagewise, does movement
in the vapor state become important in relation to move-
ment in the liquid state?
Herman Bouwer: Eventually you will get only vapor move-
ment, but when you got to that state, I don't think you
have to worry about ground-water condition; but as long
as you are above the hydroscopic water, you will have water
movement in a liquid phase and you will have downward
movement. Now, in a few years you may not detect much.
In a few hundred years you might be in for a surprise.
John Mann: This to me is a very important thing, whether
you say that the water is able to move down and then
somehow completely stop. If you adopt the concept that
the movement only slows to a very small rate, but that, in
fact, movement goes on regardless of the rate for centuries,
then you nevertheless are faced with the problem of some
type of water pollution or some potential for it.
Now, this may be theoretically true within the time
frame of lab experiments, but as the water moves down and
as the amount of water available at the lower levels gets
less and less, I still think that there is such a thing as a
pellicular front when you put water on the dry desert soils,
and these are dry. I have been involved in the bucket-
augering of many places in the creosote bush areas, for
example, and this stuff is absolutely dry. I think the
important thing is that perhaps the movement Mr. Bouwer
suggests does go on, but that other processes begin to
take over. Now, for example, the work done at Hanford by
Isaacson and his group indicates that vapor transfer is not
an important mechanism. The temperature control from
season to season has more influence on the moving of water
up and down, primarily. Perhaps it is slow, but perhaps
mostly in the vapor phase, so I think that regardless of how
you approach it, the concept that the water continues to
move down through these dry vadose zones even at an
extremely small rate, is important.
Now, if you start thinking in terms of centuries you
don't have to have a very high rate to conclude that it gets
there, but I think what we are faced with here is the
interposition of a number of other mechanisms which will
not only control, but perhaps even reverse this very tiny
tendency for water to continue flow down in the partially
saturated zone.
Donald Runnels: Can I say something on that? Not being a
hydrologist I keep looking to hydrologists for wisdom on
this, and I am constantly amazed at how little is known
about it. Bill Guyton proposes piston flow. He proposes that
210
-------
the ram that falls on the surface mechanically shoves down
the water that-pellicular water, the water on the surface of
the minerals well below field capacity. No gravity flow. You
have to wait until the next rainstorm comes along and put
a gradient down on this thing, and rather than the new
water draining through the interstices between the grains,
it pushes down the old water that was there before in a
mechanical piston-flow fashion. Guyton is a man who has
worked in arid regions for a long time, and he is still
guessing. I am amazed at how little is known about it.
Herman Bouwer: I think the piston flow holds true only
where there is fairly high water contents.
Donald Runnels: An important factor would be, I suppose,
compressibility. The water is, you know, not easily com-
pressed, and I suggested that this thing is mechanically
transmitted down sort of as a wave.
Herman Bouwer: I think in the top material, you have a
process of infiltration and redistribution and diffusion.
Eventually it might come down to the deeper depth.
John Mann: There is some evidence of this following very
heavy rains, and with an increase in barometric pressure
there have been some evidences of the increased barometric
pressure pushing on the saturated water in the zone, com-
pressing the air underneath it and transmitting a small
amount of that differential to the water, only in the
water-table situation directly, and we are dealing with
relatively wet areas and relatively short, partially
saturated zones. Now, it's almost as if you were to have a
flexible membrane on top of the system that the air
pressure above increasing, pushing on the air underneath,
and that being transmitted to the water table. I think we
are talking about very small magnitudes of movement, and
also rather specialized types of conditions.
Donald Runnels: Do I draw the conclusion that we don't
know much more about water in the vadose zone?
Herman Bouwer: I think if you ask the physicist, he will
agree that they know quite a bit on the theory, but when
it comes to applying the theory to an actual situation it's
the difficulty of getting a proper input data that prevents
him from making intelligent judgments. Its relationship
to the saturated hydraulic activity or water content or
tension can be determined in the lab on samples, but to do
this on field soil, natural conditions and then for a large
profile, say the entire vadose zone, it would be very difficult.
I think the general principles are known.
W. Roger Hail: In applying this practically for waste
disposal, with the current state of technology, what kind of
a safety factor do you think you should apply for, say,
disposing of potential contaminants? In my case I am not
suggesting very hazardous wastes, but, say, high TDS water,
for example.
John Mann: I think on a high TDS water you can tolerate a
much lower safety factor than you could where there is
some particular toxic element. If you are dealing with the
normal, merely elevated TDS saline solution with normal
ions, it will depend somewhat on the rate at which this
material is redissolved, and assuming that many of the
reactions that Don Runnels has talked about do not
permanently tie up these things in a way that they are not
easily redissolved with just one wetting field capacity, you
may not be able to get the pore volume flow through. And
so I think the net effect of this is, if we returned to an ice
age and you had a reinstitution of deep percolation of
rainfall, the amount of resolution of this material would
be relatively slow.
Donald Runnels: I am more concerned, Roger, with highly
toxic contaminants, and our philosophy has been that
unless they are tied up, regardless of a very large amount of
water passing through, we have to worry about them, and
perhaps they shouldn't be discharged at all. Take arsenic,
for example. Let's test the soil. Let's test the alluvium or
its capacity to remove arsenic irreversibly in essence against
many events of precipitation, and then we can be safe.
TDS is a tough one because if any water gets through,
the TDS will probably get through, too; so I'm on the other
side of the spectrum. There are highly toxic contaminants
and they have to be removed and tied up.
Jay H. Lehr: I have heard two different things at the
Symposium with regards to desert hydrology. I clearly had
the feeling at our first session that while we don't know a
great deal about all the things that happen as the water
moves down from the surface to the ground water, we do
know that the desert seems to be a better place to dispose
of liquid waste than in a humid climate, because it is slow
and there are good things that are likely to happen, and
that we can use the soil to treat waste.
Then last night at the Las Vegas Valley meeting I
didn't get the idea that they felt the picture was very good
at all; that is to say, taking the theoretical approach that
we have taken at the first session, and the practical problems
they have here at Las Vegas Valley with the growing
population and poor planning with regard to their disposal
of waste, they didn't seem particularly buoyed up by the
fact that they are here in the desert rather than back east.
John Mann: Jay, I have done some work here in this area,
and this is one of the world's most horrible examples, and
I don't think we want to be too influenced by this
particular situation, because it is perhaps one of the worst
ones, not in terms of the toxicity of the material, but in
terms of just the huge volume of wastes, the high salinity
of the waste rather than the true toxicity. Industries were
developed in World War II which had very large flows of
water. This was done under wartime conditions where no
consideration was given to disposal of wastes. It was
gung-ho for the war effort in a situation geologically that
was not too well known or understood, and so we have now
the heritage of these conditions developed during a wartime
push. The slowness of movement of these materials once
underground is just continued for decades. I think if there
is one thing in common in this type of pollution it is that
you can get it in the ground relatively quickly, and it takes
a long time to get it back out.
In addition to that we have superimposed upon these
rather thoughtless industrial disposals, an enormous growth
in domestic sewage, particularly municipal sewage, and in a
hydrologic or geohydrologic environment where everything
is funneled down one little wash. We have 40 million gallons
211
-------
a day, high in nitrate, high in phosphate, dumped into a river
which shows a rather sharp increase in salinity. We have here
one of the horrible examples that we can use as a gauge
against what we should do next time.
Ronald L. Barto, James M. Montgomery Consulting
Engineers, Inc., Pasadena, California: In talking about this
vadose zone, it seems to me that evaporation is going to take
a lot of, shall we call it, short-circuiting of the water on its
trip downward, and evaporation will cause a lot of this water
to move back to the surface and never make it to the
water table. I would like John to comment on this, particu-
larly in arid regions where we know that the field capacity
is so low. Does this evaporation go all the way to the water
table, particularly if the water table is, say, a hundred or
two hundred feet deep?
John Mann: I think there must be some sort of vapor
transfer. We know just from the fluctuations of water levels
and with barometric changes that there must be quite a bit
of air flow through the soil, and certainly with the flow of
heat in and out this will also have an effect. I personally feel
that the evaporation from pellicular films in desert regions
is certainly one reason why moisture levels are extremely
low. Somewhere in the past there was a colluvial period.
There was a time of old lakes. Many of our valleys, Death
Valley, the hottest place in the U.S., had a few hundred
feet of water in it. We must have had some rain passing
directly through the soil all the way to the water table.
Since the end of this time, there has been a progressive
removal of moisture between those materials and the water
table. The mechanism by which this occurred is very difficult.
I think airflow evaporation is one of them. Probably heat
flow is another. Then when we speak in terms of thousands
of years we get to the range in which soil zones would be
developed. We have a lot of feldspar minerals in our desert
soils, and the feldspar minerals particularly will hydrate and
tie up a lot of water as they become clays. I have seen
materials where you can still see the shadow of the cobbles.
These cobbles, which probably were granitic to start with
are now turned completely to clay. They are imbedded in a
clay matrix, and they are all clay, and you can see this from
the top of the zone down to fairly fresh material in the
lower part, and that has to be associated with the tying up
of water. The main source of water in the weathering below
the surface in a dry environment, would be the peHiculaf
films.
Donald Runnels: I would like to make just one very brief
comment about the downward movement, and then loss by
evaporation. I think the presence of zones of caliche in
desert regions today are showing us about how deeply most
of the water penetrates before it is lost. Mass balances that
are done on the caliches say that most of that calcium
has to be coming from the surface, from windblown dust,
and is moving down, and if you find a zone of caliche
that is four feet below the surface, that suggests that the
water is stopping at about four feet below the surface
before it is lost again by evaporation; and so I take it as a
good sign for a place to dispose of wastes, a place where we
have good thick caliches forming today.
Ronald L. Barto: Then you are suggesting that maybe 25,
50, or 100 feet is much too deep for evaporation to take
place, is that right?
Donald Runnels: No. I am thinking of the recharge of the
occasional precipitation event then being evaporated back
up, but I can't imagine why there shouldn't be evaporation
in a water table in desert regions. If we have a chemical
gradient from pure water at the water table to 10 percent
relative humidity at the surface, it seems to me that over a
long period of time you simply have to have evaporation. I
was addressing the occasional rainstorm.
C. W. Fetter: It seems to me there must be regions where,
because of natural processes, the soils are unsuitable for
agricultural purposes and the ground water has such
mineralization that you can't do anything more to harm it,
it's beyond use. Is there a potential for looking for areas
like this and then saying, well, this is a good area for
underground waste disposal because even if we do have
movement of these dissolved substances, even if they are
toxic, the ground water is toxic, the soil is saline, we might
as well use it for a. disposal area?
John Mann: Yes. I have gotten myself at cross-purposes with
certain regulatory agencies in suggesting that waste disposal
is a very important part of our existence, and I have proposed
even in some instances where an alternative water supply
is available, that certain now-fresh bodies of ground water
be dedicated to waste disposal. This might be an area in a
desert basin, usually where the water cannot flow readily to
other areas. We do have in most of the closed basins, the
Mojave Desert, the Antelope Valley area very good examples
of this. Within this system there is a good deal of recharge
from the San Gabriel Mountains. There is flow through a
series of basins down to Koehn Dry Lake, which is the
sump for the whole system. The water there now is probably
10,000 to 20,000 ppm. This is the sump of the natural
system. All of the salts from this system eventually will
wind up there because there is an integrated ground-water
flow down to the bottom of this. There is an isolated fault
basin. All the salts build up there.
The costs of waste disposal are becoming so high
that we can't divorce the problem of, say, trying to save
maybe $1 million worth of fresh water when you can get a
$10 million benefit out of using this basin for waste
disposal, assuming, of course, that you have another source
of water.
Waste disposal should be viewed as one of the
important economic elements. Designation of Class I waste
disposal sites in California means we can put highly toxic
elements there. Finding a spot like this that will supply all
of the rigid criteria is very difficult. I drove way back in a
mountain range on a dirt road and found a little dry lake in a
closed ground-water basin with internal drainage. I wanted
to designate it as a Class I site, but it turns out that there is
a Bureau of Land Management grazing lease on the area. To
my knowledge there hadn't been any grazing or grass there
for 30 years, but in the early forties there were a
couple of wet years and grass grew. Now, there is an
example of a wasted resource. Class I waste disposal are
disposing of toxic substances, things like Kepone, where
you can set it out of the way so it isn't going to create a
problem to existing facilities.
212
-------
Jay H. Lehr: This Kepone situation strikes me as being an
example of what can be done, and it can be done easier out
here than in humid Virginia. That was almost a military
operation the way they dismantled and buried it. Out here
in the desert we have areas where we could be doing that.
Why don't we?
John Mann: Well, as the great champion of the desert
environment, I don't want Kepone here! I don't want any
problems like that here-let's use Ohio! (Laughter)
Charles P. Vanderlyn, U.S. Environmental Protection
Agency, Washington, D.C.: They mentioned on the field
trip on Sunday that they were thinking about using the test
site area for radioactive waste material. Do you have any
comments on that?
John Mann: Yes. I will refer you to Ike Winograd's paper.
He has suggested putting the waste in containers and putting
them in the bottom of holes within the vadose zone. He
has gravel sheaths around these so that hydraulically any
water that might come down gets diverted off to the side
away from the container. This is certainly worth looking
into, because one of the biggest waste problems of our age
certainly is the disposal of nuclear wastes because of the
long half-lives. I worry very little about the ordinary ions
that we find in drinking water. It is somewhat like Bill
Walker pointed out this morning, which I am sure brought
gasps of horror to certain people, when he said they should
have taken the cyanide and let it run into the creek on
down into the Mississippi River. I gasped a little, too, at
that, as a matter of fact; but I could see if the thing were
done intelligently and were released with a knowledge of
what the dilution factor was, it could be done safely.
We have three types of wastes: ones where the only
thing is sort of a high salinity of the normal ions, and this
poses perhaps over a period of years the least type of
problem; the second is where you have toxic substances, you
have to be more careful; and the third and most difficult
are the radioactive wastes because of the half-life.
Donald Runnells: In our proposed regulations in New
Mexico regarding the previous question, we proposed not to
protect water that has TDS gradient of 10,000. Perhaps the
best use for that is disposal. In talking to environmental
people I have brought up this possibility to them about
disposal in the vadose zone, and it really follows Winograd's
proposal about getting rid of radioactive wastes. I
presented the picture that perhaps he can dispose of metallic
elements today that are problem materials, lead, copper,
arsenic, and wouldn't they like it if 10,000 years in the
future that material was still contained in one small area
in solid form and could then be excavated, if necessary,
and reextracted. Ten thousand years from now that might
be a resource, and somehow this strikes a favorable chord
with my environmental friends; if you have to get it out of
the vadose zone, you can.
Jay H. Lehr: Waste has been described as just substance
without a use, you know. One can say there is no waste. It
is just matter that we haven't found a use for yet.
Donald Runnels: In the vadose zone in an arid region, we
wouldn't even have to encapsulate some of these wastes
because nature would do it for us chemically.
Jay H. Lehr: Could not a great industry develop in the
desert, in the area of solid waste disposal? I mean, could
we not take thousands of acres of useless land and begin
disposing of the nation's solid waste in some type of vault
or nonvault?
Richard H. Pearl, Colorado Geological Survey, Denver:
This is a very cost-intensive procedure, and companies can't
afford that. For example, we have one there at the Denver
Metro area where Shell Oil Company has been making for
years this pesticide at the Rocky Mountain Arsenal. They
were ordered to clean up the water problem they had there.
They wanted to put it out on the Lowry Bombing Range
right over one of the major aquifers of the Denver Basin. A
lot of people raised objections to that. Well, there was a
disposal industry in Houston, Texas, and they were willing
to take this stuff, but Shell Oil couldn't afford to ship it
to Houston. They proposed to put it in deep-well disposal,
5,000 feet deep. That was turned down by the State.
Today it's still being buried out there over that aquifer.
Herman Bouwer: It is the low rainfall and the aridity of the
air that makes the desert attractive for waste disposal. In
some of the papers that were presented these last few days
on the garbage dumps and the landfills, the main reason you
have the leachate production in the eastern States is
because of the high rainfall. In the desert, sanitary landfills
are no problem because there is no rain, there is no
infiltration, and no leachate production. Maybe we
should have indoor garbage disposal areas in the East.
Keep the rainfall out, don't produce any leachate. Put a
roof over your sanitary landfills. You can collect the rain
water on the roof and use that for potable purposes.
Richard H. Pearl: You can do that. A lot of sanitary
landfills are being constructed today with impermeable
barriers on all six sides of them, and even on top.
Hopefully it is going to work and keep the moisture out of it.
Any rainfall that falls on the landfill in the filling stage is
collected and moved out of the area.
Ronald L. Barto: To illustrate the fact that you need to
keep the disposal sites reasonably close to where the
material is being manufactured, we are working for a
client in California where a Class I disposal site was closed
due basically to political reasons, and they were unable to
find another site to reopen, and as yet they are still looking.
For the past three to four years most people feel that
probably where these Class I wastes are being deposited
are either along the roads in Southern California or in
the sewer lines, because none of the other waste disposal
sites in adjoining areas have been receiving this waste. So
if you make it too costly to transport this waste to
adjacent areas, the producers will find other means to get
rid of it.
Nils Johnson: In designated desert areas and disposal areas
you would require complete reconsideration of a number
of industries. Industries, I would think, would tend to
move to the area where they could most easily dispose of
their waste and yet be near a usable natural resource to
use in whatever products they manufacture
213
-------
Jon O. Nowlin, USGS-WRD, Carson City, Nevada: I
would like to make a couple of different points. One is the
transportation issue. Right now, one of the reasons that
steel is not recycled more often is that there is a trans-
portation rebate in transferring iron ore from the mine to
the mill. There is an incentive in terms of the tax structure
that doesn't exist for scrap, so if you can't take scrap
cars and within the same State, recycle them back to the
mill, you are not going to be able to take wastes from
Nevada economically and ship them back out east.
No. 2, as a recently immigrated Nevadan, I would
like to defend our State against the people who want to use
it as a dump for the nation, and show why people are
against it within the State, and that is because of
examples such as Las Vegas.
No. 3, I think that we will take a giant step towards
coming up with the proper solutions, both politically and
technically, if we talk more of waste storage instead of
waste disposal, and get to this idea of taking some of
these materials that are now hazardous and let's talk right
from the start about storing, not disposing of them.
Charles P. Vanderlyn: There is tremendous energy potential
in our solid waste. I know in Germany, 20 percent of some of
the cities' electrical and generating capacities is in recycling
of the cities' ordinary wastes.
Russell E. Darr, Wright Water Engineers, Denver, Colorado:
I can give you an example of just how people are afraid of
something new. There is a certain brewery in Colorado that
was looking for various municipalities in the Denver area to
take their garbage for nothing to generate such a plan as
you have proposed in energy production; and they finally
gave up when they could not find a municipality which
would donate their garbage, so you have got a political
problem to overcome before you can look at possible
scientific and practical aspects.
Leland L. Mink: We have just completed quite an exhaustive
study of the Boise-Caldwell-Nampa urban area in south-
western Idaho on this idea of taking collection of garbage,
and one of several alternatives, there was a proposed
coal-fired plant. The industry was interested in utilizing
all the combustion material in this, but we could not talk
local governing agencies into even considering collection
and transportation to that State. Political reasons again.
Charles P. Vanderlyn: How do we as technical people
surmount the political problems involved in coming up to
solutions?
Jay H. Lehr: I think the answer to that question is the
development of continuing communications between the
scientific communities and what they are doing, and the
public. And now our time is up.
Bull Session on Controlling Pollution from Sanitary Landfills,
and Reduction of Nitrate Contamination
Session Moderator: David Miller, Geraghty & Miller, Inc.,
44 Sintsink Drive, Port Washington, IM.Y. 11050.
David Miller, Moderator: Our panel speakers are: Joseph H.
Baier, Suffolk County Dept. of Environmental Control,
Hauppauge, New York; Lawrence A. Eccles, U.S. Geological
Survey, Laguna Niguel, California; Todd Giddings, Todd
Giddings and Associates, State College, Pennsylvania; James
L. Mang, SCS Engineers, Long Beach, California; Michael A.
Apgar, State of Delaware Div. of Environmental Control,
Dover; and W. B. Wilkinson, Water Research Centre,
Buckinghamshire, England.
Ken Childs, Suwannee River Water Management District:
Mr. Baier, the two years of data suggested that the nitrates
were highest beneath the sites that had lowest application.
And then you concluded by saying that, however, based on
six months of data that wasn't presented today, we feel
that the original conclusion was true. Can you explain why
the first two years of data showed the other trend and
support this?
Joseph H. Baier: It wasn't two years of data. It's only what
I showed in the overheads. There was only one year of the
data, from March 1975 through March 1976, which took
you through one full growing season and just up to the start
of the second growing season. The only background data we
had on the nitrogen was our first samples as soon as we
put the wells in, which was approximately the first piece of
data. We have no idea what the historical water quality
results were on each of the demonstration plots. Why one
was higher than the other I don't know. Now, the second
planting, which took place this year, 1976, and is just
being harvested now, we continued with the monitoring
with the same number of wells and we put in more wells
at each site. We only had one well, which is approximately
centrally located on each of the plots, with a grower
practice and the recommended practice. We then added
another well upgrade or just immediately adjacent, off-site,
to tell us what the background was coming on to the site
and we added another well on each half in the plots
themselves to verify that we were getting the same nitrate
data in two spots. This year's data is now showing that the
214
-------
reduced nitrogen application loadings is yielding better
ground-water quality in the recommended plots over the
grower plots. In other words, those graphs are now reversed.
The trend started high and then started back down on just
about all of them. They went up and stayed up there for
most of the growing season, but they were coming down,
and that downward trend for the recommended practice
in 1975 was statistically significant and it continued on into
1976.
Ken Childs: So you think this relationship is one of time
delay from application to results?
Joseph H. Baier: Yes, you've got certain mineralization.
You've got organic matter that's built up in the soil for
years in the root zone and that has to adjust. The amount
of nitrate, which is being mineralized and leaching out—there
has to be adjustment between the soil interface, the crop
interface, the fertilizer application and ground water, and
I think we saw that. While the crops responded pretty much
as we had hoped, ground water took a little bit longer.
Charles Kreitler, Bureau of Economic Geology, University
of Texas at Austin: What percentage of that nitrogen that
you feel is going into the ground water is the result of
mineralization of the organic nitrogen in the soils versus
direct leaching of fertilizer into it?
Joseph H. Baier: Part of the study included a detailed
review or nitrogen budget of fertilization as to what could
happen to all different forms, the plants themselves, the
plant itself which is plowed under, the organic material
that is still remaining in the soil, the possibility of
volatilization, the possibility of runoff, losing some of it
because of that running off the site, and many other
features throughout the entire nitrogen budget for a
potato farm. And the fellow who did it was Jack Meissenger
from Cornell University. He came down to three basic
materials as controlling the whole situation. And as the
nitrogen loss is equal to the nitrogen applied, minus the
nitrogen in the tuber which is harvested, very simple
balance for the whole system, everything else came out as
being negligible, too small a quantity to really quantify or
to really have an effect upon the losses that you were talking
about, because we're talking close to a hundred pounds of
nitrogen per acre lost. But he did go into details on mineral-
izations, the amount of organic material present. He
discussed that some of the growers also plant the rye crop,
cover crop for erosion purposes, just to keep the topsoil
there in the winter. And they also add fertilizer to that. It's
an additional loading at the end of the year to get the
grass to grow quickly, and this all adds into the input.
Charles Kreitler: I've done some of the nitrogen isotope
work on Long Island. Steve Ragone and I are sampling. We
see some interesting problem in that the nitrogen isotope
ratios for the ground water in Suffolk County are running
in a range of approximately plus three parts per million as
plus seven parts per million, and the fertilizer in general is
down around zero parts per million. Somewhere we're
getting an isotopic shift that is causing our ground waters
to get isotopically heavier than the nitrate fertilizers. Now,
we see this in several other aquifers, and the amount of shift
appears to be dependent upon the type of fertilizer we
are using. If you use a urea anhydrous ammonia, you get a
further shift. Somewhere we're changing the isotopes in
this system. Somehow I'm wary of the concept that the
nitrate going into the ground water is equivalent to the
nitrate which was put on minus the nitrate that went off in
the tuber.
Joseph H. Baier: You're saying, in effect, there would be a
direct interconnection, that amount that's applied goes
right through?
Charles Kreitler: If there is a direct flushing of the nitrate
from application directly into the ground water, we should
expect to see that take isotopic composition. Ground water
is going to be identical to the fertilizer and we don't see this.
And I don't know where the problem is.
David Miller: Maybe for some of the people who don't know
some of the work that you've been doing, you could
explain it—some of the work you did in Queens and Nassau.
Charles Kreitler: Right. The paper in Ground Water two
years ago, I guess it is, we looked at natural nitrogen
isotopic variation of N 14/15, and we are able to
differentiate different isotopic ratios for nitrate sources
that originated from animal sources versus cultivation
phenomenon. We applied this technique to some areas in
western Texas, where we had very bad contamination. We
could identify the major source of contamination of organic
nitrogen in the soils due to cultivation. Enough fertilizer
used, but we could separate out the animal waste. We did
not see the isotopic signature there. Since then we have
looked at a number of different aquifers and in areas where
we have minimal cultivation, but, say, heavy dairy farming.
We see that the nitrate in those ground waters is related to
the nitrates coming off those barnyards. We just finished a
study on nitrates on Long Island, around 20 samples. You
can see that when you look at Suffolk County you have
one isotopic range, and that's Pleistocene there, and you
move towards New York City and you have the deep
Magothy aquifer. That has a slightly heavier isotopic range.
The Nassau Pleistocene is an even heavier isotopic range,
probably indicative of septic tank contamination there.
And Queens County supposedly never had much in the way
of cultivation at all and was based on septic tanks and
leaking sewer lines, and it has a much heavier isotopic
value.
Unknown: Mr. Eccles, if you determine any interconnec-
tions as they do exist in Los Angeles between the upper
aquifers and lower aquifers and you came up with some
face rate, pumping from the lower aquifers so they would
not be contaminated from the affected aquifers on the top
either through the interconnections themselves or through
any abandoned wells in the area which you described
yesterday.
Lawrence Eccles: At this point we do assume that the upper
portion of the aquifer is connected to the lower portion,
but you don't necessarily get recharged to the different '
portions of the aquifer from identical spots. The lower part
of the aquifer may be recharged from an entirely different
source than the upper part of the aquifer. The well number
two right next to well number one obviously is an inter-
215
-------
connection. Some water's going to flow downward in well
two through the ingradient and just by well one flow
down, of course, into well one. We'll probably recommend
that they plug well two up.
Unknown: My immediate concern is that there will be
greater pumping from the lower one, if that would affect
the gradient and localized recharge in the area itself from
areas which are contaminated with the nitrates.
Lawrence Eccles: This isn't something we've thought about
just yet. That would be part of the greater basin manage-
ment plan, to recommend what the yield from various zones
in the aquifers would be, and one thing that might work
would be a pump test.
W. B. Wilkinson: Mr. Baier, what was the thickness of the
unsaturated zone and have you been able to measure any
nitrate profiles down through the unsaturated zone or any
tritium profiles? I'm a little concerned about the interpre-
tations you've made on the basis of the nitrate concentra-
tions in the saturated zone in the ground water. It could be
that the rate of movement down through the unsaturated
zone is very slow and that what in fact you're observing is a
land use or a fertilizer application that may have occurred
possibly several years ago.
Joseph H. Baier: Specifically with respect to the demonstra-
tion plots that we selected, these plots were regular
farmed areas and have been farmed areas for 20-some odd
years. The only change would be occasionally the farmer
would not plow the field and just leave it for a year just
with no crops whatsoever or once in a while he would
rotate a crop, just to change the over-all composition. The
depth of the unsaturated zone varied from 2-16 feet. We did
this specifically because we wanted to be able to pump
the wells themselves from suction, using suction pumps.
The wells were installed with skimming wells right at the
top. The type of infiltration on Long Island is coarse sand
and gravel. You're talking about infiltration rates in the
neighborhood of feet per day. We have very little or hardly
any runoff, regardless of size of storms. If you would
travel through Long Island in a two or three-inch rainfall,
you'd see puddles all over the place. Two hours after the
rainfall had ended, the streets would be dry and no puddles
on the side. Everything has seeped in, so we know that it
goes right through. We explored tritium, but the fact that we
were actually using farmers' fields and he was going to have
to harvest the crop, regardless of the fact that tritium is
harmless, I don't think we could have talked him into
using it.
W. B. Wilkinson: I wasn't talking about the application of
tritium. I was talking about the measuring of tritium that's in
the rainfall as a result of thermonuclear explosion from
1953 onwards.
Joseph H. Baier: We had explored the possibility of applying
some just to see if this infiltration was directly looking
at the background, then applying a known quantity and
seeing what the result and reaction would be on the ground
water, since we knew we had fairly direct connection
between the land surface and the ground water. But there
were a lot of logistical problems there.
Ken Childs: Now, I have listened to you talk, Mr. Giddings,
and what your conclusion was essentially, was that site
suitability may be somewhat lacking but the installation of
a "plastic liner" guarantees the integrity of the site. This is
really the bottom line in an awful lot of proposals for
landfills today. In addition, I'd like to say that there is a
recent E.P.A. document suggesting the questionability of
liners. Now, can you demonstrate with some existing study
or documentation what the lifespan of a liner is, or a
different kind of liner and what the failures are?
Todd Giddings: The suitability of geology at the site and
the example I gave is an important aspect of the liner
installation and performance at that site. I think you can
see where the four natural backup features in the site made
the danger of leachate contamination of ground water
minimal, if not virtually impossible in the event of a major
liner failure at this site. So I was pointing out the utilization
of a marginal site through the installation of a liner and
further designing the installation of that liner in the
operation of the site to take advantage of the natural site
characteristics. I think it has to be in harmony. I agree with
you, if you go into a very bad geologic site and say
because I've lined it I'm okay, you may find it impossible
physically or economically to correct a leak should it occur.
The P.V.C. liner was chosen only after an exhaustive review
of literature and research on many types of liners, such as
hypalon, E.P.D.M., monimer, butyl rubber, petroleum
drive, petromat types and asphalt, and was selected based
on the available state-of-the-art in researching liner
performance under attack from "leachate." And the reason
I put that last word "leachate," in quotes, is because if you
look at published chemical data on the constituents in
leachate, sometimes the ranges are two or three orders of
magnitude for concentration of critical parameters. So
there's leachate and there's leachate. As far as a leachate
attack, there have been studies and there's literature
documenting satisfactory performance under attack from
strong leachate. Since we're in a new technology, we
don't have 25 years of data. There hasn't been P.V.C.
under a landfill for 25 years anywhere, so, unfortunately, the
tests sometimes are accelerated under extreme lab conditions.
Other field tests have been as long as 2 to 5 years, and so
we are in an area of chemical extrapolation, but it has
been addressed in the literature and we feel that since
that's a very important question, the soil layer immediately
below the refuse will become saturated with water and
provide a capillary barrier to contaminants that would move
down out of the fill that aren't admissible in water and
hopefully would perch above the liner. And the sand layer
above the liner has a permeability based on the leachate we
expect because we want to maintain an unsaturated
situation above a liner. We don't want leachate sitting on
a liner to have the time to attack it; thus, with regards to
the point you raised (what about the durability of a liner)
we put those features into this particular multilayer system.
Because I agree, those points you've raised are important.
They are not fully known, and we had to approach them
in that way.
Joseph H. Baier: How do your various respective communi-
ties feel about resource recovery as, let's say, an alternative
to landfills?
216
-------
n Chi ' don>t think y°u're going to have any success
with the recovery of garbage, or whatever, until there's
some way to support it economically, and right now I
don t thmk you can recover these materials economically.
In north Florida there are programs of aluminum can
collection centers, but that's, you know, just a drop in the
bucket. I don't really see it until it becomes economical.
Joseph H. Baier: You can recover paper. You can recover the
glass. You can pick up other metals besides aluminum.
I'm talking about almost complete resource recovery so
that you're left with basically garbage, and then you
incinerate that.
Randy Sweet, Consultant, Kelso, Washington: I know of a
county in Washington where they were funded by EPA for a
grinder. It's operating today. The price of disposal of
garbage, which they now bury, went from $2.50 a ton to
$8 a ton and they're still burying the stuff. The market-
ability of cardboard became a very big market in the
northwest and immediately everybody wanted into it, and
a lot of stores put in their own baling equipment. The
market went well over $50 a ton. For a long period of
time everybody was making money and then the bottom
fell out of the market and the bailers sat around. The
main market that people think they can depend upon is
the fuel market, and right now private industry in Oregon
is putting together a grinder. They've developed a grinder
for tires and I think that's the way it will go. They'll pick
out particular fractions which are easy to separate and
hard to dispose of and high B.T.U. values and use as a
fuel.
Jim Mang: I have a comment on your resource recovery
considerations. Most of the statements have been directed
at direct resource recovery. Sanitary landfills can play an
important role in indirect resource recovery due to the
installation of wells and the extraction of methane. This is a
feasible alternative, depending on the size of the landfill
and another, a variety of other considerations. In small
landfills enough methane can be recovered, say, that a
golf course is built on top of it to run the air conditioning,
etc.
Ken Childs: Mr. Mang, to quote your paper, "expensive
and not entirely effective are plastic liners," and then you
make a pitch for clay liners. Is there documentation to back
up "the efficiency of clay liners," and can you address
clay liners simply on the basis of personal ability or do
you have to address mineralogy? Don't you feel that it's
necessary to address the length of time that it will take the
leachate to move through that liner and the chemical
reactions that will take place between the mineralogy of
the sediment, including the clays and the leachate itself?
Jim Mang: I didn't make a pitch for clay liners. I think the
state-of-the-art is an area of flux where there's not that
much specifically known about either clay liners or P.V.C.
liners and their longevity. I think it's been demonstrated
that P V.C. liners won't do the job and the potential exists
for clay liners, and I think this is the only extent that
anybody can speak of in terms of longevity. Clay liners
in my opinion offer the most promise because they are a
natural soil product. What you have to do is utilize the
clay liner to force the leachate to float at a given area,
then get it out of the landfill. Leachate or certain
contaminants in leachates, some of the heavier metals,
will migrate through the clay and it is a function of what is
required by the regulatory agencies on what you can and
can't do.
Todd Giddings: I have to respond to your statements where
you say that it's been demonstrated that P.V.C. won't do
the job. I can cite you instances where I could demonstrate
that steel bridges have failed and therefore I can say that
steel won't do the job and we should stop building bridges
out of steel. Well, that's patently absurd, as is, I maintain,
the generalization that because P.V.C. has failed in certain
applications of landfill liners we should cease lining
landfills with P.V.C.
Jim Mang: I think your permeability differentiation is
your most ardent safeguard, but I just cannot see spending
the money on a blanket basis for P.V.C. liners when nobody
can demonstrate that they'll last a long period of time. I
wish you could show me a situation where they have
lasted over more than 20 years. I'd be glad to look at it.
Todd Giddings: Wilbur and Orville Wright wished there had
been someone else do it first, I'm sure, too. I'm not
advocating the general widespread use of P.V.C. liners as a
panacea for landfill site design. In fact, I'm a proponent of
natural flow system manipulation and control as part of
collecting leachate in a landfill site. And we in fact have
sites in Pennsylvania and elsewhere that are operating
without any type of liner, using the natural flow system,
manipulating that flow system and collecting the leachate.
Jim Mang: In your talk I could not understand why you
had the P.V.C. liner until you explained to me that it was a
cost-effective situation, reducing the flow of ground water
and reducing the flow of leachate. In that sense I could
understand it, and only in that sense.
Todd Giddings: The site again is in a ground-water discharge
zone. To place the refuse in contact with the ground and
to allow that water to seep into the base of the fill would
greatly increase the quantity of leachate generated, and in
fact you can visualize them thinking back to that sandwich
situation, draining the ground water from under the liner
keeps it out of contact with the refuse. And by reducing the
quantity of leachate generated, we realize a great savings.
Randy Sweet: I was wondering how you put together your
plumbing system. I believe you said 140 feet of garbage
stacked up in there?
Todd Giddings: 120.
Randy Sweet: Okay. A large face. One of the biggest
problems that I've run into in landfills is what I refer to as
the "nuisance seep" all along the face, one or two gallons
a minute in a multitude of places. How are you separating
that dirty water from your clean water so you don't end
up treating all the runoff from the landfill?
Todd Giddings: The gas vents which will be constructed
throughout the landfill-several dozen of them-will consist
217
-------
of precast sections of perforated concrete pipe, culvert pipe
on end, packed with gravel with a plastic or perhaps steel
pipe within the center of the major gas escape avenue. These
pipes will be added to and built up as the fill depth
increases. They're going to be the short circuit drain.
They'll be connected directly to the leachate overdrain
network and they'll be the short circuit to bleed off the
perched water, because we've recognized exactly what
you've pointed out, that the nuisance seep of leachate
breaking out of the front face of the landfill due to
perching on our intermediate or daily cover is a problem.
Randy Sweet: Have you looked at orientation underlifts at
all as far as moisture routing system within the landfill?
Not an entire lift, but the daily lift in the compaction.
Todd Giddings: We haven't specifically looked at the
orientation of the working face, but I recognize what
you're saying and would thank you for a very good
suggestion with regard to creating an anisotropic to be
favorable to prevent the breakout that we're trying to
prevent.
Unknown: How are you going to treat the leachate?
Todd Giddings: Initially, the leachate which comes out of a
drain system will go into a holding lagoon of approximately
1,000,000-gallon capacity, in which there will be floating
fan aerators. After aeration, that treating leachate will be
piped back on the landfill and recirculated to accelerate
decomposition of the refuse and stabilization of the mass.
This is along the lines of work done by Poland and
Hammond and others that's been reported where recircula-
tion accelerates the leaching process, because we're aware
we face the problem of how long are we going to treat the
leachate. We've done water balance calculations and there
will be a point where we'll have reached field capacity
and have more water coming out than we can recirculate
back to the fill. At that point, we'll have to look at what
the quality of that is with regard to what type of treatment
we'll have to construct at the site and then we'll have a
discharge to surface water under a State permit. Money
will be set aside in an insurance fund for just that type of
endeavor, to keep that cost down. We're going to
recirculate initially.
Unknown: This process continues on for a hundred years.
It will be exhausted and there will be a problem for the
children.
Todd Giddings: I beg to differ. Based on research work
by Poland and Hammond, they show that the major
constituents in leachate in recirculation situation begin to
decrease to acceptably low levels for a low level of
treatment or no treatment.
Lawrence Eccles: If the integrity of the liner is in
question, why not just use an additional liner, use two or
three if you have to? They're relatively inexpensive, aren't
they?
Todd Giddings: And by using the probability of reasoning,
if you double it, you multiply the probability together,
thereby achieving a greater safety than just a factor or two?
Lawrence Eccles: No, as time goes on, the leachate is
going to become less potent, so maybe it will react with
the first liner over so many years. Well, have a layer of clay
beneath the first liner and then a second liner. At least by
that time we'll be digging the garbage up to reprocess it
anyhow.
Todd Giddings: In our system, we have the backup naturally
occurring features beneath the liner, the underdrain, the
ground-water.discharge, the permeability contrast, the
artesian head, even beyond that, the bedrock with the
monitoring wells giving us an extraction capability that we
felt it was not cost-effective to go to two liners. We had
the natural features at our site in lieu of a second liner, but
I could visualize a site perhaps on Long Island or a major
aquifer where that protection might be warranted because
I agree it's unknown as to when that liner will finally fail.
Ken Childs: There's something I don't understand. If
effectively you're developing a closed system in your
landfill and you recirculate the leachate, how do you end
up in 20 years or 40 years with no leachate?
Todd Giddings: We definitely end up with leachate. We
performed a water budget or water balance, a series of
calculations, taking into account precipitation, infiltration,
runoff and the moisture deficiency of the incoming
refuse, and generally I think it's been found that a foot of
refuse has in the neighborhood of an absorbent capacity of
two inches of water. And neglecting short circuiting
channelizing and immediate through-flow in a through-
lift, we don't expect to have major quantities of
leachate generated. That is, we don't expect to reach field
capacity in our fill within the first 10 years because the
refuse will be coming in as a blotter material faster than
the precipitation is soaking into it. We definitely will reach
a point that we're going to have an excess because we are
in the humid east and there's some 30 to 40 inches of
rainfall. It's at the planned point that we have to treat
quantities of leachate that we'll have by circulation lowered
the concentrations, that the treatment will be minimal in
requirement for discharge. We'll then have to treat and
discharge.
Ken Childs: Mr. Apgar, in your opinion, were there any
interactions between the leachate and the matrix that the
leachate moved through?
Michael Apgar: Yes.
Ken Childs: If so, what was it? The second question is
frequently we assume that contaminants move out as some
sort of a mass or plume of waste. Did your studies or
results suggest that was or was not the case, or did you
actually see bifurcating plumes of individual constituents?
And third, what sort of a monitor system or how would
you suggest a ground-water monitoring system be designed
for a new landfill, one that's intended to detect contami-
nants rather than to pacify or satisfy regulation?
Michael Apgar: There is considerable interaction between
the contaminants in the leachate and the earth materials
that this leachate has moved through; that is, the strength
of the leachate that we're collecting from the recovery
218
-------
wells in the aquifer is not nearly as bad as it is originally
in the landfill itself. The sample chemical oxygen demand
that we're pumping now at a rate of 2,000,000 gallons a
day is down to somewhere around 25 mg/1. Granted, that's
diluted with a lot of fresh water in the aquifer.
Ken Childs: Exclude dilution and talk if you can about
absorption processes.
Michael Apgar: As the leachate moves through the ground,
you filter some out and you have precipitation and
absorption reactions. You have a whole host of things going
that will decrease the concentration of the organics,
inorganics. In the way of metals, we don't see any
concentrations of heavy metals other than iron and
magnesium oxide. So other metals, although they're there,
and although the Redox conditions are appropriate or are
such that they could be mobile, if they were there, they
just don't move. I think that there is a lot of treatment or
renovation as the stuff moves through the ground. It's just
that we're not getting enough here before the leachate gets
into an aquifer system. And once it's flowing through
quartz sands in a confined system where there's no input or
very little input of oxygenated fresh water, not much is
happening and not enough is happening to decrease the
concentration of the organics and the iron and ammonia
before it gets to the wells and that's why we're involved in
a contaminant recovery system there. Now, secondly, as
far as whether the contaminants are moving in a plume
or a slug, the geology is very complex. There's an inter-
bedded sequence of sands, silts and clays, and in some
places immediately adjacent to the landfill you can drill a
well and the water will be fresh and uncontaminated as
leachate has never moved to those positions. In other
wells immediately adjacent to the landfill we find that the
water quality is very degraded at times and it's pristine
at others. This indicates that we are getting contaminants
moving through the ground in pulses, even though the
base of the landfill is continually saturated. Somehow we
have slugs moving through and along one part of the flow
path. We're seeing contamination at one point in time and
no contamination at others. And this can be observed very
strikingly by pumping a well. I have seen wells pumping
clear water all of a sudden become very discolored and
foamy, or wells that were producing foamy colored water
clear up instantly.
So we have both things going. We have plumes and we
have slugs here. We've gone to the philosophy that we
shouldn't introduce leachate into the ground-water system.
That's the concept behind all these liners. So as an effective
monitoring program, first of all we ought to be monitoring
what's going on beneath the base of the liner, if in fact
we've got a liner. I think that to adequately monitor
whether water-quality degradation is occurring in such a
complex site is virtually impossible on a practical basis
because we've got a hundred wells along Llangollen. We
had to throw in 20, I think, before we could even get
together a picture as to where contamination was occurring
and where it wasn't. The first 4 or 5 wells right next to the
landfill that should have shown contamination came up
with good quality water. I think that what we do when we
permit a landfill site to go in is to make the decision that
the aquifer materials or whatever geologic materials con-
taining water are near the site can be written off or degraded
to an extent. That is, we're not proposing that we locate a
landfill next to a major aquifer or on top of a major aquifer.
I think to do that would be irresponsible.
Ken Childs: Can I conclude from what you've said that
you could not predict with any reliability where a waste
plume would occur downgradient from your landfill? And
if that's the case, then a simple monitor screen of one well
in the contaminant direction would not be adequate.
Michael Apgar: Well, I don't think that one monitor well
downgradient from any waste source is really
going to be adequate. Even if it's just one
continuous sand layer 50 feet thick, you're going to pick
up different water-quality results. I guess what you really
want to know as far as monitoring is not how bad things
are getting but just are they getting worse, is there leachate
there. On a first cut monitor basis and if the leachate is
there, then you've got to make a decision as to whether you
have a more intensive monitoring or restrict water users
from using the water.
Ken Childs: I have been involved with contamination and
recovery studies and I know the number involving
petroleum products where there's a residual that remains
in the saturated zone indefinitely. Does this same sort of
residual remain in the saturated zone for leachate? In other
words, you can pump all the water out a hundred percent
and then as new water comes in, there's still a residual
that comes out and continually contaminates the new
waters that pass through.
Michael Apgar: Yes, I certainly think that's the case.
W. B. Wilkinson: It has been suggested in Europe that
perhaps domestic waste, domestic landfills, may be sited
on aquifers on a thick and saturated zone and that the
unsaturated zone can do the job of cleaning up the leachate.
Would anyone subscribe to this idea? And if so, are there
any simple guidelines they could suggest as to thickness of
the unsaturated zone?
John Moser, Pennsylvania Dept. of Environmental
Resources: We use a ratio of one foot of soil for every foot
of refuse that you're going to place there. That is, put all
that soil in before you put any refuse. We have to plan
ahead how much fill you're going to put in there. This
ratio was developed on the basis of some studies done at
Penn State and it's not really entirely defined as to
whether that's adequate in all cases or not, but we have used
this and we are running out of sites very rapidly where we
think this is suitable. It takes a fairly decent soil.
Randy Sweet: I think the most important thing is to be
able to take a close look at the flow system, at the geometry
of the flow system, and be able to predict where leachate is
going to migrate.
Unknown: I think Mr. Apgar has done an excellent job in
cleaning the landfill which was designed with some
consultants 20-30 years ago. But why can't you put a well
in the landfill itself somewhere, pump it out from the
landfill itself and treat this by osmosis?
219
-------
Michael Apgar: The option has been investigated. It's very
difficult to get water to flow to any central point in this
landfill. We've got 23 wells and five sumps that were dug or
bored down into the landfill and we tried to investigate the
rate at which we could get water to flow to any of these
central points. We decided that we'd have to have a drain
network with a spacing of somewhere between 100-200
feet in the trash to start collecting water out of the trash.
To install this through 30 feet of trash, 20 feet or 10 feet of
which would be saturated, would cost a couple million
dollars. And since we weren't sure that was the way we
wanted to go and the reverse osmosis studies show we have
a large reject volume of water, we have about a 50%
reject through the reverse osmisis system. 50% of the water
came out good and we have 50% left, and it was just twice
as bad as it was before and we had nothing to do with it. So
we didn't go to reverse osmosis and collecting from the
landfill.
David Miller: Thank you all for participating. We're
supposed to reconvene across the hall.
Bull Session on Monitoring the Flow of Polluted Ground Water,
and Artificial Recharge as a Solution to Pollution
Session Moderator: Wayne Pettyjohn, The Ohio State
University, Columbus, Ohio.
Wayne Pettyjohn, Moderator: Our panel consists of these
speakers: Bruce Labno, Pfeifer and Shultz/HDR, Inc.,
Minneapolis, Minnesota; N. Thomas Sheahan, Brown and
Caldwell, Pasadena, California; Dennis Williams, Agro-Water
Consulting Engineers, Tehran, Iran; Harry Nightingale,
USDA, Fresno, California; and Stanley N. Davis, University
of Arizona, Tucson. I'll now open the session for questions
and discussion.
Gerald Hendricks, Sieco, Inc., Columbus, Indiana: I realize
the value of a tracer, such as Freon 11, in both tracing
ground-water flow, and in analyzing possible pollution. I
have heard that the Freon 11 is, more or less, a common
part of our atmosphere, and that perhaps the minor
amounts that you might find in sewage effluent would not
be revealing. I would like us to discuss the pros and cons of
it or any other suggested better tracer for sewage effluents?
Stanley IM. Davis: We do not as yet have enough actual data
on Freon 11 concentrations in sewage effluent. We have
sampled rivers downstream from industrial complexes
where we assume that the rivers have received industrial
waste, and these rivers contain Freon concentrations far
above the concentrations that you would have in equilibri-
um with the atmosphere. Also, a well near a large landfill
area in Indianapolis was sampled and very large concentra-
tions of Freon were obtained from that well, suggesting that
Freon might be indicative of some types of industrial
effluent, plus landfill leachate. Theoretically, the problem
with normal sewage effluent might be the fact that there's a
tendency to aerate the sewage in some of the sewage plants
and also the sewage contains a moderate amount of
organic material both in suspension and in solution that
might tend to absorb or actually combine with the Freon
some way to take it out of solution so that it is not easily
detected in the normal analysis. So my thinking at present
is that Freon 11 may be useful in distinguishing between.
say, sewage effluent and some industrial effluent, as well
as landfill leachate, but it may not be possible to distinguish
between sewage effluent and normal recharge water from
the atmosphere. My coauthors are here. Perhaps they would
wish to add something.
Thomas R. Schultz, University of Arizona, Tucson: One
thing that maybe wasn't brought out in the talk that we're
attempting to stress with this technique is the fact that it
is a field-operable system and can be operated by people
with very little training. You have the ability of taking more
data right in the field and as was alluded to, it's very similar
to the use of tritium sampling, which is quite expensive. So
if this technique should become valid, many small consulting
firms and public agencies could use it.
Gerald Hendricks: I infer from what you're saying that a
gas chromatograph would be used as an analysis and if so,
other people could assist in obtaining the data rather than a
specialized group of high-quality lab people?
Thomas R. Schultz: That's right. Many other environmental
tracers have many different origins, such as tritium and
chloride. It's difficult to really know what the sources are.
With the Freon, we know exactly how and when it was
first introduced into the atmosphere so it should be much
easier for people to interpret the data. In regard to your
question on errors, I think, although I have no data to
demonstrate this, neither ourselves nor Glen Thompson of
Indiana University has gotten to the point yet where we
could look at analytical errors in the field. One of the
projects that we're getting ready to undertake is to calibrate
or verify this technique in the Edward Limestone in Texas.
There's a model that exists for the Edward Limestone which
takes into account the flow and also mixing and dispersion
of tritium. We would like to verify the Freon technique
with this model, and in the process, we would be able to
get some hard data to answer a question such as yours.
Stanley N. Davis: To amplify the questions you brought up
220
-------
concerning errors-one advantage in field analysis, of course,
is that you can repeat your sampling on-site and make sure
that you're getting consistent results from the gas
chromatographic system. Furthermore, at the present time,
Glen Thompson is preparing and will, I assume, distribute
some standard samples so that calibration can be made
between various workers interested in low values of Freon
concentration, so we not only can field-check the sampling,
and hopefully reduce any errors that may be present in the
sampling technique, but we can also assure ourselves that
we're getting some reproducible results in terms of other labs.
Kenneth W. Lustig, Panhandle Health District, Coeur d'Alene,
Idaho: As a small health department, we historically use
rhodamine dye for tracing septic tank effluent into adjacent
lakes. We're presently involved in a 208 effort with a couple
of agencies on an aquifer study. It's been proposed, in order
to try to substantiate that the effluent from the septic
tanks is recharging the aquifer, that we put some kind of
tracer in it. It's been suggested that we use rhodamine.
Would someone care to comment on the advantages or
disadvantages, outside of tritium, of something that a small
health department could use?
Thomas R. Schultz: Rhodamine is quickly absorbed to clay
particles. If you have an aquifer containing a significant
amount of clay, rhodamine or any of the fluorescein dyes
would be an extremely poor choice. Chlorofluorocarbons
have the lowest surface energies of any organic, so they're
less likely to be absorbed by other organics or clays. You
need something that could be detected at extremely low
concentrations, could be demonstrated not to occur in the
environment and might be the sort of thing that would hold
up in a court of law. Also it's possible that if there were
more than one polluter, each polluter would be obligated to
inject a small quantity of a certain tracer into his effluent
and these can be separated easily. This would cost about
$2,00 to $3,000 in parts plus labor. The operation of the
equipment is much like operating a component stereo
system, almost anyone can do it, and the analytical
techniques are firmly established.
Stanley IM. Davis: One possibility of a cheap tracer might
be bromide. It's not very concentrated in the environment,
it is relatively nontoxic, at least most of the bromides, and
if one has a specific ion electrode to pick up bromides, you
can usually measure down in a few tenths of a ppm. It
depends on the amount of dye solution that you're going
to anticipate, whether or not this would be practical, but
bromides have been used on some occasions.
Kenneth W. Lustig: The situation we're in is where we've
got an area pumping 5,000 gpm with less than half a foot of
drawdown. The surrounding area is being subdivided. We
have a 20-acre tract dedicated to us by a small industry,
and we dug a test well to sample the ground water with a
tracer. We're involved in a lawsuit at this time with the
building industry. We're at a very crucial point and we want
to use something that is going to stand up in a court of law.
The way the whole area develops is going to depend on
what happens in the next year or two. We have about a year
or year and a half's worth of samples now and the only
thing that's showing up that we can really count on are the
nitrates, which peak under the subdivision. We've got less
than a ppm elsewhere, and 9V4 under the subdivisions.
Stanley N. Davis: As you describe the problem, I would
doubt whether bromide would be a good solution because
you would probably have to use very large quantities and
there might be some public health question.
Kenneth W. Lustig: What constitutes a large amount of clay?
We've got areas where we mine the gravel and never even
wash it, but the well drillers we talk to say, "We ran into
two or three inches mixed with clay." Would that be
enough to filter some tracers out?
Thomas R. Schultz: I don't recall exact data but I would
think that would be enough. It would at least reduce it to
the point where the hydrogeologists from the other side
could make the data questionable.
Gerald Hendricks: I would ask Mr. Schultz if the Freon
equipment setup is going to be published?
Thomas R. Schultz: I believe Glen Thompson will
probably be describing some of the analytical techniques in
a publication he's working on now. It should be within six
months or a year that a person with proper facilities could
build one of these devices himself.
George Hoffman: I have a question for Dennis Williams.
Did you consider increasing the number of pumping wells
and doing away with the well barrier?
Dennis Williams: One of the purposes of the project was to
illustrate the principle of conservation of ground water
through underground storage. Possibly we could have made
up the difference that we got by recovering previously
stored water or by importing additional water, but that
would have defeated the purpose of the project. There are
many areas in which we'll have to use this principle of
bringing in excess water during the winter season so that
we can store it underground either through injection wells
or artificial recharge basins and then recover it during the
irrigation season.
The hydrologic situation as it exists without any
extraction or recharge is in equilibrium; in other words,
there's fresh water in certain areas, and our project was
to superimpose a ground-water mound upon this and
then recover the mound. The salt-water ridge is more or less
stabilized now through natural conditions. We're only
pumping out the mound. In other words, we can't start the
project pumping first and then recharging. Wo built the
mound about eight meters high. By the time irrigation is to
begin, and then durring the irrigation season, our recovery
wells will degrade this mound down to the original level
before we started.
George Hoffman: How did the salt-water aquifer get such
great heads? They're way above sea level and way above
the head of fresh water?
Dennis Williams: The only explanation we have is this zone
of high pressure fossil waters, laying below 100 or 200
meters, was formed during the transgression and regression
of the ancestral Caspian Sea in Plio-Pleistocene times. The
overburden was building up, and as a result the pressures
221
-------
were increasing and a lot of water was squeezed out of the
clays which compacted more than sand, and the water
then migrated into the sand zones, with the result of
increasing pressures. Our exploration has shown that there
is no hydrologic continuity between this lower zone and
the fresh circulating ground water due to an aquiclude,
20 to 30 meters thick, consisting of very tight organic
clays with some silts. This zone, for some reason, is just
sitting down there and it's geologically sealed off, but when
you do tap it, the water comes out of it like sea water.
George Hoffman: Then you don't think it's connected to
the sea? It must be sealed off?
Dennis Williams: It may be connected to the sea, but our
project is located 15 or 20 km inwards so we really haven't
studied the continuity. The aquifer is below sea level but it
may outcrop somewhere.
George Hoffman: Are these heads typical in other parts of
the country or the world?
Dennis Williams: My experience in California is we don't
have any heads near that high. For example, in Owens
Valley, where I've done a lot of work, we've encountered
maybe one atmosphere a head. We've encountered these
tremendous heads in several areas in Iran, not only along
the Caspian Sea but also along the Persian Gulf area. But
I'm sure it occurs when you have buildup of sediment with
a lot of clays.
N. Thomas Sheahan: Is there much of a temperature
increase in this lower salt zone? Could high pressures be
geothermal?
Dennis Williams: No, we didn't observe any temperature
increase. The temperature of the lower water was the normal
ground-water temperature; however, some of the shallower
wells, which smelled of hydrogen sulfide gas had tempera-
tures close to 80 degrees F; however, the water quality was
700 or 800 micromhos/cm.
Gerald Hendricks: Would Mr. Sheahan describe in more
detail the type of wells that they used for their project and
why they used those wells?
N. Thomas Sheahan: We had some principal points of
interest in the design criteria for these wells. First of all,
knowing that we were going to be recharging reclaimed water
and the potential for suspended solids which also goes with
it, a potential for plugging of wells, we do expect to have a
maintenance problem on the injection wells, because of
suspended solids, and as a result the materials of construction
and the types of construction were designed with this in
mind. Also, there's a serious research program that is starting
and.will be going concurrently with operation of this
system for a number of years, looking at water quality
aspects of injection and extraction of the reclaimed water
into underground formations. Therefore construction
materials were selected which would have the least possible
effect on water quality, so that our water-quality variations
would be minimized and directly related to the geology and
hydrology. The third thing was the physical requirements of
injection wells, somewhat higher pressures and structural
problems. As a result of these three major areas, we
selected Schedule 80 polyvinylchloride casing. We selected
an 8-inch well size for the injection wells principally for the
requirements of pumping equipment. The well screens were
304 stainless steel wire-wound well screens which, in my
opinion, is probably the best type of well screen for the
most efficient operation. The stainless steel also was
selected because of its nearly inert character and the least
chance of affecting water quality in any way, yet providing
sufficient structural strength in the screen area. The
inside diameter of the screen is the same as the inside
diameter of the casing, so with this type of configuration
we can effectively use both mechanical and chemical
development methods in the future. We can use various
types of chemicals.
The wells are gravel-pack construction. The importance
of this is to provide one additional design buffer between the
well screen, the last point that you have the control of
injection water, and the formation phase, where you lose
complete control of the injection water. By using a gravel-
packed well, we can install a gravel pack that provides an
effective filter as well as a medium to hold back, and so by
designing the gravel pack in this way, sediment that might
be deposited in the well during injection would be filtered
out near the well face, we can more effectively remove these
with a maintenance program in the future. The gravel-pack
was a 10 or 20 type gravel, nearly 1.0 uniformity coefficient.
Each of the wells is double cased and double screened. The
deep well is 8 inches in diameter, and the smaller well is
6 inches.
Gerald Hendricks: What was the diameter of the gravel pack,
and did you vary the anticipated rate of application in any
one well depending on the formation and size?
N. Thomas Sheahan: We will be varying the rates because
there will be variations in permeability, aquifer thickness,
and geology in the area.
E. W. Ramsey, Virginia State Water Control Board,
Richmond: With respect to the quality of water and its
injection into an aquifer, have there appeared any problems
with respect to pressure and also with respect to compati-
bility of the injection water and the aquifer into which the
water is injected?
Dennis Williams: The injection pressures that we're using
are designed not to exceed two atmospheres, which is 28
or 30 p.s.i.
Harry Nightingale: While on this subject of recharged wells,
we have one at our Leaky Acres Recharge Facility. At this
facility we're using about the top five feet of soil to filter
the canal water. The water is collected and it is gravitied
down the recharge well. This recharge well is a cheap well.
Anytime you start a recharge operation, you are going to
get clogging of the interface between the well and the
aquifer. The system which we are working on is just about
the opposite of a good water well; that is, the well is a
sander when you pump it. The well is built with a four-foot
diameter conduit pipe down to the first 20 feet. An eight-
inch well goes down to about 250 feet, but the well is
gravel packed with l-ll/2 inches of nice round river rock.
This year we started out at about 500 gpm—this is a
222
-------
gravity-red well-down to about 350 gpm. We've had some
clogging on the surface of the aquifer interface. We don't
pump it very hard. We want to mine the sand that's in the
aquifer. This is where the colloids have been filtered out.
We mine this sand into the well and pump some of it out.
When you do this, the gravel around the casing starts moving
down and then out into the aquifer. We replace the gravel
at the top and this way we hope some of the colloids get
filtered out. We can measure the colloids and actually
recover the original recharge capacity. At this facility, the
ground-water mound beneath the basins is pretty well in a
chemical equilibrium with the canal water and the colloids
in movement are in equilibrium with the soil water as well
as the ground water. When these colloids are injected they
are not going from a low saline to a high saline environ-
ment and continue to precipitate; they are going to an
environment similar to where the colloids came from, so
they move a tremendous distance. We know they've moved
at least 400 meters out into the aquifer.
Wayne Pettyjohn: Would anyone attempt to answer the
problems that we all face in trying to get technical ideas
across to administrators, politicians and other people?
N. Thomas Sheahan: At the risk of maybe oversimplifying
things, we ran an injection test, and one of the most
important parts was a demonstration project. We did not,
at that time, deal with the water-quality aspects of it, but
perhaps a demonstration test over a period of time during
which sufficient water-quality data is made available to
draw some conclusions, however valid, and maybe this is
the easiest way. Administrators are much more receptive
to a comment like, "We have done this," rather than,
"We can do this."
Wayne Pettyjohn: We've had several questions on construc-
tion details of recharge wells, and I'd like to briefly explain
some that I constructed several years ago which were
described in Ground Water about ten years ago. This was a
situation where there was sand from the surface down to a
depth of about ten feet, about fifteen to twenty feet of
silt and clay below that, and a major aquifer underlying that.
Water levels were down about 70 feet. We constructed a Y-
shaped canal system attached to a basin and pumped
raw unfiltered water directly from a river, discharging it into
the basin. When the basin was full, the water flowed down
the Y-shaped canal system. There was a lot of recharge
from the basin itself. In the canal system we constructed
two types of wells. The first type were bored wells deep
enough to go about ten feet below the bottom of the
confining bed. They were 30 inches in diameter, filled with
sewer rock and covered with pea gravel. As the highly
turbid water flowed down the wells, they plugged quite
rapidly and it was almost impossible to clean them. Using
a dragline with a 3/4-yard clam attachment, we excavated
4 holes, which are, in reality, about 12 feet in diameter.
About 20 feet of corrugated culvert, 6 feet in diameter,
were installed and gravel-packed with pea gravel on the
inside and the outside. Obviously, the rate of recharge
through the large-diameter wells, which were very
inexpensive, was high. Of course, with time, the
rates decreased as the sediment began to accumulate in the
gravel pack. It was a relatively simple matter, however, to
remove the gravel pack with a clam, which was then stock
piled adjacent to the recharge site and clean gravel was
installed. After a year or so, much of the gravel that had
been removed was washed clean of the sediment that had
accumulated. It could then be reused. It was a very inex-
pensive method and rather primitive, but even so, we were
able to recharge about 4,000,000 gpd, although this
decreased with time. The original proposed solution to the
water shortage was a pipeline to the Missouri River; 43 miles
of pipeline at a 1959 estimated cost of $12 million. The city
had neither the money nor the time to construct a pipeline
and that's why the recharge system was built. The total
system, including land acquisition and the construction of
an $87,000 dam, was about $192,000. Furthermore, when
the system was put into operation, the water level rose
more than 22 feet within 6 months of operation. The point
I'm trying to make is that it is possible to develop
artificial recharge systems quickly and with limited funds.
I'd like to ask a question of Mr. Nightingale. I noticed
on one of the slides that at the beginning of each recharge
period the mineral content, whether it was chloride or
nitrate, increased and then it declined rapidly. This
appeared to be cyclic. Would you comment on this?
Harry Nightingale: Well, say we've been recharging all
summer long and we've reached a low value EC for that
year. When the water is turned off in October, the profile
drains rather rapidly but, when the profile is pretty well
drained, there are zones of soddy clay soils which are
micropores, and we feel there is diffusion of salt from the
micropores into the main flow channels. There appears to be
some downward percolation of this more saline water, but
most of the water moves out to the large openings so when
the next recharge period occurs, the water goes through
the bigger channels rather than the smaller ones since the
salt has diffused into these channels, and a new peak of salt
is leached into the ground water, and then, as the recharge
period continues, the salinity continues to drop down.
Wayne Pettyjohn: Might the increase represent water-
soluble products that are forming in the unsaturated zone
between recharge periods and you're just merely flushing
those out?
Harry Nightingale: Dissolution of the rocks does occur as
well as evaporation from the soil surface. We've looked at
the EC of saturation extract after recharge and found it is
very low, but after 3 months, in the wintertime, it will be
back up. Some water must be moving back up,
evaporating, and the rocks are slowly dissolving. I might
add, since you were talking about recharge shafts, that
we've done considerable work on recharge shafts, and one
we had set up this summer ran for 66 days. We put 6.4
million gallons of water down this one shaft. This is
irrigation canal water, but it does go through one screen
filter, about like a flyscreen, which is just enough to
filter out any moss, beer cans, etc. Again, the recharge
shafts are built different; they're 4 feet in diameter and 32
feet deep. We've got about 10 or 12 feet of good sand, but
again we go to gravel which we can add as we mine the
aquifer. The shaft is built so that we have the conduit pipe
which is a 10-inch pipe, and at the bottom of the shaft
there is an iron cage. On top of the cage there are some
rocks. In this case as the recharge water flows out into the
aquifer, it becomes plugged. On this shaft recharge is started
223
-------
at 475 gpm, but after 2 days, it was down to 100 gpm.
After 66 days it was down to about 50 gpm. To increase
the rate, the first thing we do is try to mine the sand.
The top of the shaft is sealed and when a submersible pump
is turned on, you get a shock connection and a high velocity.
The sand sometimes will pump out but a lot tends to go
down into the cage where we can get at it. Our idea is that
these shafts are very cheap to build, they can be connected
to an irrigation canal system and they only occupy a few
square feet. But our experiences show that the canal water
should be pretty clean. The water which we're using has a
turbidity of about 3 to 5. Sooner or later the shaft will
need to be abandoned. Only a few hundred dollars invested
in it and all you have to do is move over about 15 feet.
Wayne Pettyjohn: It seems to me there's no question but
that artificial recharge techniques will provide a solution
to many of the water problems that we have. Why they
aren't used more is certainly beyond me.
Stanley N. Davis: A short question for Mr. Nightingale.
For many years, storm water in Fresno was drained into
so-called dry wells. Are those still operating or have they
abandoned them?
Harry Nightingale: Basically they've been abandoned.
Ken Brewster, Illinois Division of Water Resources:
With a little background, I'd like to address this to Mr.
Sheahan. In the Chicago area there's been a proposal to
inject tertiary-treated sewage effluent that has also undergone
advance treatment, into the two primary rock aquifer
systems. Do you think you could have a doublet of the
type of design that you proposed that would adequately
remove a contaminant? What problems might be associated
with such a project? The initial project, as proposed by the
Federal EPA and the Metro Sanitary District of Greater
Chicago, would require an extraction system during the
test. If in fact it did work, the extraction well would be
put on standby for emergency purposes only.
IM. Thomas Sheahan: The key to the injection-extraction
doublet operation, within a limiting streamline, is a
regional gradient. Without a regional gradient, there is no
limiting the streamline, so it has to be there naturally or
created artificially by overpumping the extraction well. To
determine what the water quality would be after injection
and passage through the aquifer when it comes back out the
well, the wastes could be mixed with a certain amount of
native aquifer water because of the overpumping of the
extraction well required to produce the artificial gradient.
This may produce a problem. The only way around that
would be to have sampling points between the two wells
to use for basic data. The extraction wells could be used
simply as a tool to make the system work and completely
extract the water. On that basis, I would guess, without
knowing any more about the geology, that, in either
aquifer, you could construct such a system and adequately
control all the water injected by use of extraction wells.
When you talk about injecting into two different zones,
though, this may unnecessarily complicate the hydrology
and it may or may not be possible to extract all the
water from multiple-screened wells.
Wayne Pettyjohn: Ladies and gentlemen, I'm afraid that
our time is about up. We are supposed to reconvene across
the hall for a summary session. Thank you.
Bull Session on Managing the Movement of Contaminants,
and Protecting Mines, Wells, and Pits
Session Moderator: Tyler E. Gass, IMWWA, Worthington,
Ohio.
Tyler E. Gass, Moderator: Our panel consists of these
speakers: John Fryberger, Engineering Enterprises, Inc.,
Norman, Oklahoma; Otto Helweg, University of
California, Davis; Elmer E. Jones, Agricultural Research
Service, Beltsville, Maryland; and David Walz, Virginia State
Water Control Board, Richmond, Virginia.
Fred Hoffman, U.S.EPA, San Francisco, California: I
would like to address my questions to Otto Helweg. This
morning he discussed the use of ground water in an
agricultural community where the water was
becoming more saline downgradient. In many places, we
have similar situations. I'm thinking of an example where
we have a valley such as you described this morning with
increasing downgradient salinity. Artificial recharge has
been established upstream with water that has higher TDS
than the water that is in place now in the upper parts of
the valley. Have you thought how ASTRAN might work,
or how you might manage that type of situation?
Otto Helweg: Obviously to make the system work you've
got to have a water quality gradient someplace. Ideally if
you could start upstream with high-quality, usually surface
water that's higher quality than the ground water, then
you've got it made. If you don't have that, the only other
thing you could do would be to maybe set aside a section
of the upstream aquifer that is of high quality or the
highest you have and transfer that downstream and depending
on your setup, the imported water then could be recharged
at some point where the ground-water quality could be
perhaps just a little bit worse than the recharged water,
but I realize that you might need it upstream rather than
downstream, so it's a little difficult to comment.
Fred Hoffman: Okay, just in general then, your ASTRAN is
224
-------
only going to work if you have a perfect situation: good
water quality upgradient, increasing salinity downgradient?
Otto Helweg: I think it's more general than that. I have
never run into a river basin that you did not have this
water-quality degradation as you go downstream. If you
have that, the method can work, technically. The only
constraint might be the political, institutional or legal
constraints whereby if there was no way you could use the
highest quality upstream water downstream.
L. Jan Turk, University of Texas, Austin: I can show you a
lot of basins in Texas in which the opposite is true. Our
streams flow from the arid to the humid regions and the
water quality improves toward the coast so we do have
those situations there. These are major river basins. They
flow from the western arid region to the coast and the
rainfall dramatically increases. We have a lot of salt springs;
they start out as saline rivers, and there's a tendency for
those that flow across the northern and central part of the
State to improve in water quality. Some of them get more
polluted from bacterial sources.
Otto Helweg: I would say the ASTRAN method would be
of no value. On the upstream part where it is arid, do you
see a reversal in the gradient?
L. Jan Turk: Generally speaking there's a fairly steady
improvement downstream because of tributaries coming in,
and the increase in rainfall and increase in runoff results
in dilution of the water as you go. This is really unusual
because most streams flow from the humid areas to the dry
areas and this is just exactly the opposite.
Tyler Gass: I think the situation would become more
common as you approach more and more humid areas
even if the stream started in the humid areas. There are
many streams in the eastern part of the U.S. where there
are natural forms of contamination that do become diluted
out as the stream goes downgradient.
George Robinson, Peabody Coal Company: We have a
situation in the upper Colorado River Basin where they
have the salinity control plant. We wanted to dewater a
mine which had fairly high quality water to add to the
system, and applied for an MPDS permit and were originally
or initially declined because the fact is we are adding the
TDS to the system and regardless of how much water
we're adding to it they stated that the water would be
evaporated and the TDS would be still in the system so
they declined us the permit. Now we have no alternative
but to close the mine down even though we are releasing
higher quality water to the system. I would like to see
your management plan adopted by EPA and I think some
of the regulatory agencies should be aware of the systems
we're talking about because with these constraints placed
on no discharge, it makes it almost impossible for industry
to work within the guidelines of the law.
Elmer Jones: In Public Law 92.5 it basically states, I
believe, that a natural discharge is not a pollutant; if man
controls that discharge it then becomes a point discharge
and is a pollutant. Have they effected that practice of
trying to pump the salt water into storage ponds and
discharge it at high-water flows?
John Fryberger. To my knowledge the Corps of Engineers
has not chosen that as a potential solution and the reason
is during high-water flows the rivers carry a great load. The
concentration is lower then but the total load in terms of
tons per day is very, very high. This is primarily because
of the re-solution of the salt that has been deposited in the
salt plains. This salt is redissolved and carried out by the
flood waters so although the concentration is not as high,
still the total load is very high during the flood stage. The
Army Corps of Engineers have considered building a low-level
dam across the stream, just downstream from the source
of the salt and pump from the reservoir behind this dam
into the off-stream evapotranspiration sites. This technique
for control has been evaluated and in some cases it will
work all right but in many other cases it will not because
most of the load, in terms of tons per day, is carried down
by the flood waters, so to capture the water and then
release it into the flood waters is already being done under
the natural system. Salt is being captured by evapotranspira-
tion forces, being held in place and then released to the
flood waters.
Elmer Jones: Isn't the salinity in low base flow the greatest
concern?
John Fryberger: No, I would say it's the flood-water load
that is of greatest concern. For instance, the Red River is
dammed at Lake Texoma on the eastern side of the border
between Oklahoma and Texas. Lake Texoma originally was
intended to be used as a water-supply reservoir for some of
the towns in northern Texas. It can't be used because it's
too salty and getting saltier, and it's getting so primarily
because of the flood flow that picks up all that salt.
Lee Burton, Oklahoma State Department of Health,
Oklahoma City: On base flows, normally in those streams
in Oklahoma or in that part of the country, the base flow is
usually made up of contributions by return flow from the
terrace deposits and alluvial material in the local area and
usually it's of much better quality than that of flood flows.
The flood flows do pick up the salt, and therefore it's during
floodtime they are of particular importance as far as the
salt load. When they do have flows, in your equation you
presented today, does it take into consideration the period
of time when those streams sometimes will have a loss of
flow at one point and then later on downstream they will
have an increase in flow? I know you were developing
formulas so you'd have a general secretion of water
downstream but some of those streams, particularly through
the salt zones, do have a loss of flow.
John Fryberger: Yes, that's very true. In the upper reaches
in most of those rivers there are long periods of time when
there is very little if any surface flow and all of the water
that is in the alluvium and coming into the alluvium, say,
from the ground-water inflow into the alluvium is being
lost by evapotranspiration. In fact, the sum of the spring
inflow and the natural ground-water inflow below the spring
is being consumed and lost by evapotranspiration so the
only time when you have base flow, really, is following a
flood event. The flood saturates the alluvium above the
water table and it forces the water that's in the alluvium
225
-------
down and up into the river channel and comes out as base
flow. The base flow we're talking about in the equation in
some reaches of the river is only occurring following a flood
event, but it is the base flow that carries the salt out of
the alluvium. It's the only mechanism you can have to get
the salt out of the alluvium. Sometimes those streams are
dry or there might be little ponds here or there but they're
quite dry.
For instance, the amount of ground-water inflow per
year in Reach 1 of the Red River we calculated to be 59
acre-feet per year whereas the evapotranspiration losses for
that same one-mile stretch were calculated to be 844
acre-feet per year so, in some of these rivers, you have a lot
more evapotranspiration loss than you have ground-water
inflow. Of course, in addition to the ground-water inflow
we have precipitation which we calculated to be 428
acre-feet per year; and we have our flood inflow which was
449 acre-feet per year, so the sum of all the inflows is
greater than the evapotranspiration loss. The difference
between those two is the base flow.
Don Lundy, University of Wyoming, Laramie: When you
write QCa, did that mean Q. times the concentration or
does that mean just the volume?
John Fryberger: No, that is Q times the concentration. In
each factor you have a quantity of water times the
concentration of that water.
Don Lundy: I see, but when you get to precipitation in
evapotranspiration, you don't have a concentration follow-
ing the P or E and you have expressed those as acre-feet so —
John Fryberger: They could well have a concentration. It is
explained in the paper and I should have explained it today,
the concentration for the P and the E values is assumed to
be one. You could have Ce in all the concentrations of the
evapotranspiration losses and that is assumed to be one for
both P and E to have continuity.
Otto Helweg: Isn't there any possible way to come up with
a reservoir release policy downstream that could, say, pass
on the flood flows and maybe maintain the low flows? In
other words, try to store low chloride water and pass on
the degraded or salty water?
John Fryberger: The chloride concentration during a low-
flow period is higher than the concentration during the
flood periods but the total load in tons per day is lower
during the low-flow periods, so you would want to store
the water with the lowest concentration which would be
the flood water. But still, the flood waters are pretty high
in salt and they're carrying a high load in terms of tons
per day.
Otto Helweg: I don't understand why the reservoir is
degrading so badly; in other words, what was the water
quality when they first built it and how come it's getting
worse now?
John Fryberger: It's degrading because of the added evapo-
transpiration losses and concentrations by evapotranspira-
tion losses at the reservoir site itself. You're getting water
in all the time but you're losing water from the reservoir
all the time, too by evapotranspiration.
Elmer Jones: When you consider this on an annual basis,
the number of low-flow days exceeds the number of
high-flow days, so on a long-term basis what is the ratio of
salt transport?
John Fryberger: I can't answer that directly; I might
approach it though indirectly. Not for this flushout study
but for another part of the study, we did evaluate the
effect of a low-level dam. It was desired to prevent 80% of
the salt from going further downstream. The height of the
dam to do that and the amount of water that they would
have to pump out from behind the dam, especially in this
one river, was prohibitive because of the greater load
carried by the high flows. I don't have a ratio of the total
load carried by high flows versus total load carried by low
flows because it varies quite a bit from river to river. In
some of the rivers most of the load was carried during low
flow, and in those rivers a low-level dam would be effective.
In other rivers most of the load was carried by the high
flows, and in those cases the low-level dam would not be
effective because you'd have to have too big a dam.
Tyler Gass: Yesterday, Wayne Pettyjohn came up with his
drip theory in which ground-water contamination—in his
case, saline solutions seeped out of evaporation ponds for
oil-field brines. According to him, you would get a
massive drip or a contaminant pulse entering the ground-
water system and then moving out to the stream. In arid
areas, you have a situation where you seem to be getting
a pulse-type effect due to periodic high runoff or floods.
How would this philosophy of taking data affect some of
the projects that you gentlemen may have worked on?
John Fryberger: I can give one opinion, anyway. As far as
this drip theory would apply to the flushout equation that
I've been talking about, I really don't think it would apply.
The flushout theory is intended to be applied over long
periods of time. The time period for the one-mile travel in
the river that I had described is pretty long, 15 to 20 years
and his pulse idea wouldn't come into play. But we fully
recognize, for instance, following a flood event you would
have a slug of water that goes down into the alluvium and
mixes with the ground water. That slug of water is
moving through the ground water and finally is discharged
as base flow so if ground-water quality is measured at
different places in the alluvium, you'll be measuring the
effects of these slugs. If you were to measure the ground-
water quality at one place in the alluvium and say, "This is
characteristic of all the alluvium in this cross section," you
wouldn't have a very valid number. As far as determining
the water quality at one point downstream in the alluvium
you have to measure the water quality at several points in
order to negate the effect of the drip theory.
George Robinson: In Ohio I agree that the drip theory
applies as to what Wayne was mentioning yesterday, but in
Montana where we have a detailed reclamation program
going with moisture monitoring using neutron probes, the
drip theory, at least after two years, doesn't apply because
we haven't had a moisture front really pass to any great
depth. The moisture front from two years of data has only
226
-------
penetrated about a maximum of 10 feet. The rainfall here
is about 8-15 inches.
Donald Keech, Michigan Department of Public Health,
Lansing: Mr. Jones, was the air testing done below the
casing level in the lime rock?
Elmer Jones: Yes.
Donald Keech: Specifically what mechanism did you use
for the air test device? How could you pinpoint the exact
level of where either water was coming in or going out?
Elmer Jones: Your casing is set in rock such that you can
depress water 10 feet below the bottom of the casing with
air and maintain air pressure at that point; you have a sound
grout seal. In our tests we had four correlations for this but
only the pumping tests where we took samples under very
carefully timed conditions and the air pressure testing could
be directly related to the adequacy of the sanitary protection
of the formation. A caliper log at this 79-foot level revealed
a slight enlargement of the bore but you find in open bores,
many places where you'll have a slight enlargement. That
means nothing as far as sanitation protection is concerned.
The natural gamma log indicated a higher percentage of
clay at that depth. In dry weather, we would pump the well
and take a sample at one minute, two minutes, four
minutes and on a timed schedule like this. At the two-
minute pumping time we might get a jump of 100 times in
our coliform count, and then by the time you were
drawing water from the bottom of the bore, this would
have dropped down to less than 10% of what this peak was.
This happens to be the time it required water to move
from this depth up to the pump and into our sampling
point. The air pressure testing, to evaluate it quantitatively,
you have to convert your air from volume to pounds
because the density per cubic foot changes constantly as
you depress the water further in the well. But we find
that when we convert to pounds, which we can then convert
to volume at standard atmospheric conditions, a point like
this will have a fairly constant air loss with time. After
open bore grouting, this is to 100 feet, the well was air
tight at 35 pounds pressure to 100 feet which was 40 feet
below the bottom of the casing.
Donald Keech: Do you have charts on that, Elmer, that are
available?
Elmer Jones: No. I wanted to get a publication out on this
but at this time we don't have experience with certain
formations and only in wells where the casing supposedly
terminates in solid rock can we tell you how to have confi-
dence in it. We use this on testing wells in a monitoring area
and here the wells are screened with a fairly coarse sand. As
the water is pushed down into this screen it will drain out
and then be displaced by air, and your rate of air loss
will increase. However, we have been able to identify wells
that we decided obviously were not constructed to specifica-
tions. Where they were supposed to have 10 feet of
grout, they had about 10 inches. To go out and use this
method to evaluate a commercial producing well that
would be cased all the way down and screened into sand, I
would say that you could only use it with confidence
within the casing. You have an air leak in the casing or
near your upper casing installation if you cannot maintain
air pressure above the water table. We found in this
situation that if we cannot maintain 10 feet of air pressure
in the casing—if we take a suction sample from the top 6
inches of water in the well, our coliform count may be
several hundred times that of the pumped water sample,
which is a very clear indication that your contaminant is
entering above the water table and the concentration at the
water table in the well is the highest.
Lee Burton: You're speaking of this contamination that's
actually coming from the ground water, but is there any
possibility that when you're developing your well and the
use of your grout material that you have induced a
contamination there, and that you're just getting a return
from that rather than from the ground water itself?
Elmer Jones: No, we're talking about fecal strep, fecal
coliform, total coliform. The high pH of the grout will
disinfect the well for a period of time because the pH of
your water immediately after your grouting operation will
be around 12. After a rain the coliform count will go up
into the 10,000 or 100,000 area and then when the water
shifts to this discharging mode, as the water table continues
to rise, you drop off to zero or to a much lower count, so
the contaminant is not in our construction.
Donald Keech: Elmer, on the air test did you measure the
water level at the same time you were pressurizing the well?
Elmer Jones: We measured the water level before and after,
but not during. We use the initial water level plus the
pressure of the air converted to altitude, feet of water to
determine what the water level is in the well and in the
event that during the testing period there was a significant
change, we also measure this afterwards. Normally the
water level will recover to essentially the same elevation.
Ed Ross, Minnesota Department of Health: Mr. Jones, you
mentioned something about a cementing shoe. I would like
you to explain how you grout these wells, and how you
were able to find a zone that you didn't get a complete
grout. How did you locate that particular zone?
Elmer Jones: In the sinking shoe grouting technique you
open your bore to the depth necessary to obtain a
productive well and in doing this you should watch your
formation to find the desirable depth to terminate your
casing. Once you are sure you're going to have a productive
well, you then ream the bore to the depth that you're
going to set your casing. In the Johnson Well Manual you'll
find this sinking shoe technique described but they
recommend filling the well with clean plaster sand. We
found we couldn't use this because we cannot buy clean
enough plaster sand. So we back filled with No. 3 well
gravel up to within about two feet of where our casing
was going to terminate and then we put just about two
gallons of fine sand in to separate the grout from this. A
tremie is used to implace the grout. We placed approxi-
mately twice the theoretical volume of grout required to fill
the annular space. The sinking shoe is either a drive shoe
or a coupling in which you have set it in a funnel and made
this pointed nose with a quick setting cement. This is
screwed on the bottom of your casing and as you run
227
-------
this down the well the grout is displaced up the annular
space between the casing and the borehole. We try to hold
our sinking rate on the casing to about 2 feet per minute.
Then each time we make up another joint of casing we give
the casing 5 extra turns in the borehole so that we have a
high radial velocity between the casing, the grout and the
bore in addition to the high vertical velocity that we get
when the casing is sinking so there is no possibility of
having any bridging in this. This is why we have great
confidence in this grouting technique. The point at which
the contaminants were entering Well A was 19 feet below.
When Well A was reconstructed it was cased and grouted
to 60 feet and at 79 feet we had contaminants entering
and were losing air in our air test. This was 10 feet below the
bottom of the casing. When you have a sound casing and a
partially grouted annular space and you have a contaminant
entry point behind this, all your testing will tell you is that
your air is leaving the well at the bottom of the casing
and you can't tell at what point behind the casing it is
leaving the bore.
Ed Ross: That explained the sinking shoe but maybe I didn't
understand you about finding this one void space. How did you
determine where that was?
Elmer Jones: As the air pressure was brought up in the bore,
we keep a record of the time it takes for the pressure to
increase. If you go past a point at which you're losing air but
you're not losing air at the rate you're putting it in the
bore, the pressure will continue to increase but at a lower
rate and you may or may not reach a point at which you
lose all your air. The test plug that we had been using
was rated for 15 psi. We put C-clamps and angle irons on
this to hold this down and would go on up to 35 Ibs.
pressure. I would be extremely reluctant to advise people to
use these test plugs in excess of the manufacturer's
recommendation unless you have taken great care to see that
the test plug cannot blow out because it is heavy and should
it hit somebody you could almost be sure to have serious
injuries.
Donald Keech: Elmer, how did you grout the open bore hole?
Elmer Jones: Again we started with the same technique
that we used for the sinking shoe. We back filled the bore to
the depth at which we wanted to bring our grout down to.
We placed our tremie down there to where our grout went
to the bottom of this with the plastic barrier to separate the
sand and the grout and ran the grout down the tremie and
up, measuring, again, with this little hydrometer to see how
much grout it was taking to fill. Actually, in this case, it
was seeing how far 10 gallons of grout would rise in the
bore and from this we knew how fast the grout was rising
below 79 feet. It took more than 10 gallons to fill this at
79 feet. From there on up it took essentially the same
gallons of grout per foot of bore as it did before, and
then we brought the grout up into the casing about 10 feet,
so that if there were any settling we would have
continuity. However, had we stopped grouting at 70 feet
this would have been adequate because the air test had
previously shown that once air had bled out of the well
to the 79-foot level we could hold air for an hour at that
level with no discernible pressure change.
We redrilled the well. When the grout is at low
strength, I would not be hesitant to take the grout out of
the well and at that point I could do it with an air
compressor.
Tyler Gass: Mr. Walz, to what degree was there any public
opposition to putting the site where it was located?
David Walz: We went before the City Council and made a
presentation as far as the information we had gotten with
the site evaluation. The City really had no way of saying
"No" to us because they had accepted the problem that
had come with it. The City agreed to use part of their
landfill site to dispose of the Kepone. We had no reaction
from the residents.
Maxine S. Goad, New Mexico Environmental Improvement
Agency, Santa Fe: Mr. Walz, who actually did the waste
disposal operation?
David Walz: The actual disposal was carried out by a
private contractor engaged by Allied Chemical Corporation
but the disposal process and design was put together by
the State—by the Ad Hoc Technical Committee with a
representative of Allied Chemical in concert with us. At all
times during the disposal there was a State Supervisor at
both the plant and at the disposal site so it was done by
private contractors but under State supervision.
Maxine Goad: Dr. Fryberger, if I understand it correctly
in your flushing equation which had terms which were
water quantity multiplied by concentrations, what you had
were salt loading balances, is that correct?
John Fryberger: Yes.
Maxine Goad: May I assume that the salt lost through
evapotranspiration is extremely small?
John Fryberger: Yes, it would be very small. The concentra-
tion of salt is assumed to be one for purposes of the
equation. It's not even shown in the equation for either the
evapotranspiration or the precipitation. In the paper it is
explained that those values are assumed to be one.
Maxine Goad: Why did you choose to even include those
terms in your salt loading equation? Why did you not
have a separate, simultaneous equation for the water
balance—that is, two equations, one for salt and one for
water balance; why did you choose to combine them?
John Fryberger: I felt it was necessary to combine the two
because we are considering a mass balance. The evapo-
transpiration loss would remove water from the alluvium
and tend to concentrate the remaining water so it's
necessary to include that in the over-all equation. There is
a second equation that takes into consideration only the
water and not the salt. The second equation is a
simultaneous equation so you can solve for two unknowns
but that's the only way it's used.
Maxine Goad: I still don't understand why you would have
to have a term in there for the salt removed by evapo-
transpiration since it is salt we're balancing in the one
equation.
228
-------
John Fryberger: It's both salt and water in the main
equation so I felt both factors had to be considered: the
tons of salt and the amount of water because you have to
have both factors in order to have concentrations. Now,
the end result of the equation results in one factor being
concentration at a given time interval after the salt inflow
has been stopped so you're solving for concentration. The
second thing you're after are tons of salt per year coming
out of the alluvium that goes into the base flow. For the
first factor, the concentration, I think it is necessary to
include the evapotranspiration losses in the main equation
in order to arrive at that concentration.
Elmer Jones: I think the two terms are necessary. If you are
considering just the salt balance on a short-term basis I
think you could say the precipitation contains a zero
concentration; the evapotranspiration loss contains a zero
concentration but over a period of years the precipitation
falling on the outer edges of the Valley probably has
provided a driving force to drive the saline water in towards
your stream.
Maxine Goad: It's certainly true for concentration;
concentration is related both to the amount of water and
to the amount of salt.
Otto Helweg: So if this is true, you would have to include
ET even though it affects concentration, but, of course, it
doesn't affect the mass balance.
John Fryberger: Do I understand that you don't like the
idea of having a value of one assigned to the evapotrans-
piration losses when really and truly the value ought to be
zero?
Maxine Goad: Right.
John Fryberger: I agree with you; I don't like to assign a
value of one when you know it's probably not one and that
it is probably closer to zero, but I had to use one because
otherwise the factor would drop out— the E factor would
drop out.
Tyler Gass: We have to adjourn now. Thank you for your
participation.
229
-------
DISCUSSION
The following questions were answered by W. B. Wilkinson
after delivering his talk entitled "Prediction of Future
Nitrate Concentrations in Ground Water." (The paper
appears on pages 70-82.)
Q. by P. K. Saint. Where NO3-N levels exceed drinking
water standards in potable aquifers, is the information being
passed to health authorities, and what measures are being
taken?
A. Where the water authorities in England and Wales record
nitrate levels in public supplies in excess of the recommended
WHO's European drinking water standard of 50 mg/1 (11.3
mg/1 NOs-N), the health authorities are informed. They, in
turn, alert their staff, family practitioners and pediatricians
within the area. Environmental health departments will
also consider private supplies. The health authorities may
also advise mothers who are bottle feeding young infants
against excessive boiling of water for the infants' needs.
If the health authority considers it necessary, supplies of
low nitrate bottled water will be made available to infants.
This provision is made in all cases where the nitrate levels
exceed the 100 mg/1 (22.6 mg/1 NO3-N), this being the
WHO acceptable European Standards for drinking waters.
Health authorities are asked to immediately notify
the Department of the Environment of any confirmed
cases of methaemoglobinaemia so as to help the Department
and the water authorities in planning future policy.
Q. How were samples collected to prevent contamination
from overlying portions of the aquifer?
A. Preliminary samples from Chalk sites, and the majority
of samples from Bunter Sandstone sites were obtained by
power auger. Each sample was composed of cuttings from
the auger flights obtained during penetration of a 0.5-m or
1-m (I'/a- or 3-ft) interval. The augers were withdrawn from
the borehole between each sampling run and cleansed
thoroughly. Temporary lining tubes were inserted in
boreholes when unstable ground conditions threatened
collapse of material at high levels on to the auger flights.
Comparison of analyses on interstitial water from auger
samples with analyses of water from adjacent solid cored
holes suggested that contamination was minimal. Core
samples were obtained from the Chalk by driving a 450-mm
(18-in) long, 100-mm (4-in) diameter core barrel into the
rock. After withdrawal of the core barrel, the borehole
was enlarged to a diameter of 150 mm (6 in) to the base of
the cored hole, drilling rubble cleared out and another core
taken. Deep boreholes in the Bunter Sandstone were
drilled by rotary air flush methods to produce a 100-mm
(4-in) diameter core. Experience in other research
programmes had shown that this method limited the
invasion of the core by supernatant borehole liquids to the
outer 1 to 2 cm 0/2 to 1 in) of Bunter Sandstone cores.
Cores were extruded from the sample tubes on site. 5 cm
(2 in) from each end was discarded and the remainder
wrapped and sealed in two layers of polyethylene for
transport to the laboratory. Auger samples were similarly
packed for transport.
Q. by Don Runnells. How was the water extracted from
the unsaturated materials? Q. by C. Roberts. Could the
high speed centrifugation of the sample possibly produce
erroneously high (or low) values of nitrate and tritium
because of the non-"in situ " or unnatural conditions of
extraction?
A. Approximately 300 grams of material were removed from
the central parts of cores in the laboratory, excluding
material from the outer 2.5 cm (1 in) of the core. The
material was spun at 6,000 rpm in a refrigerated centrifuge
at a temperature of 12°C for one hour. Samples of auger
cuttings were processed in a similar manner. Chalk samples
generally yielded between 10 and 25 ml of interstitial
water and sandstone samples up to 10 ml. Replicate Chalk
samples have been extracted by centrifugation at 4,500
and 6,000 rpm and by leaching disaggregated material in
deionized water. Nitrate contents of the extracted waters
were found to be comparable within the limits of
experimental error. Electron micrographs of replicate
Chalk samples taken before and after centrifugation have
been examined for evidence of matrix distortion or
pressure solution of the carbonate minerals, but none has
been noted. It is concluded that the method of extraction
does not affect the value of the nitrate concentration.
Samples for tritium analysis were extracted by a standard
vacuum distillation procedure at the Harwell Laboratory
of the United Kingdom Atomic Energy Authority.
Q. by L. A. Swain. To what do you attribute the high
nitrate (105 mg/l NO3-N) at shallow depth in the Chalk on
the unfertilized grassland sites?
A. The high nitrate concentrations measured at shallow
depths in certain samples taken from unfertilized sites are
attributed to partial mineralization of the organic nitrogen
in the sample during the time between field extraction
and laboratory analysis of the sample. Further investigations
of this phenomenon, which appears to be limited to the
upper 1 to 2 metres (3 to 6 ft), at Chalk sites, are in hand.
Q. by M. Johnson. Is the tritium accumulation at 10 m a
moving front or an accumulation area? If it is an accumula-
tion area, why?
A. No direct evidence is yet available for movement of the
peak tritium accumulation in the Chalk in the United
Kingdom, but circumstantial evidence suggests that the
front is mobile. Smith et al. (1970) reported peak tritium
230
-------
concentrations at a depth of 4 metres (13 ft) in the Chalk,
and suggested a correlation with the 1963/64 thermonuclear
tritium maxima. Tritium profiles measured on samples
from boreholes drilled in the Chalk during 1974/75
exhibit peak concentrations at depths between 8 and 13
metres (26 and 43 ft) possibly indicating downward
migration rates of between 0.80 and 1.20 metres (2.6 and
3.9 ft) per year. Sukhija and Shah (1976) have
demonstrated the vertical movement of tritium fronts in
alluvial sediments in the Indian subcontinent.
Q. by J. M. Klein. Was any attempt made to determine if
there was a varying nitrogen isotope ratio (Nl*/Nl s) in the
pore waters collected from the varying land types and
land uses?
A. Measurement of nitrogen isotope ratios in pore waters
has not yet been undertaken but are being planned.
Q. How would you compare the nitrate in the Chalk with
the same in sand and gravel?
A. Studies have been concentrated predominantly on the
Chalk aquifer. No work has been undertaken in unconsoli-
dated sands or gravels. Sampling in the consolidated Bunter
Sandstone indicates that nitrate concentrations in
interstitial water are related to land usage in a manner
similar to that for the Chalk, but detailed interpretation of
the results have not yet been completed.
Q. How did the actual distribution of nitrates in the aquifer
correlate with the (1) land use pattern and (2) calculated
values (from the model)?
A. Detailed determination of the spatial distribution of
nitrate in the unsaturated and saturated zones of an aquifer
unit, and its relationship to land usage is the subject of
current work. The simplified catchment model reported in
the paper (Figure 20) shows good agreement between
calculated and measured concentrations in the ground water
and the actual land usage.
Q. by A. A. Baker. What is the effect of dispersion on the
predicted nitrate concentration?
A. Dispersion was included in both the vertical flow model
and in the catchment model. Its effect is to attenuate the
peak concentrations. The most significant contribution to
peak attenuation results from the variability of downward
migration rates through the unsaturated zone, which may
be considered as a macro-scale dispersion.
Q. by M. Apgar. What parameters are required for computa-
tions in your model for prediction of future nitrate
concentrations?
A. The vertical flow model required annual infiltration
and fertilizer application rates; land use records and the
depth of unsaturated chalk. Because of the slow rate of
solute movement through the unsaturated chalk, nitrate
input to the water table in any year may only be predicted
when data are available for earlier years. To run the
catchment model these data must be specified for each
zone of land use within the catchment. In addition the
ground-water level distribution, the aquifer properties,
and boundary conditions of the catchment are required.
Q by P. O. Seman. Did you also consider point-screens in
your study, like sanitary landfills, etc.?
A. Point sources were not considered in the investigations.
Their contribution to nitrate contamination of ground
water in the region studied is insignificant. The catchment
model assumed complete mixing of water and nitrate in
each zone which is acceptable for a distributed source, but
not for point sources.
Q. by L. Rowe. Did your predictive model allow for the
resaturation of the aquifer immediately above the saturated
zone?
A. No; the model calculations were based on yearly inputs
of rainfall and nitrate with the model years running from
April to March inclusive. It was assumed that there would
not be a significant change in water level from one year to
the next.
Q. by H. Bouwer. Would it be possible to reduce nitrate
input by better fertilizer use?
A. Fried et al. (1976) have recently summarized data
showing that excessive fertilizer applications may lead to
inefficient uptake by crops, leaving increasing residual
nitrate liable to leaching. Greene and Walker (1970) have
suggested that, under certain circumstances, restricting the
rate of fertilizer applications may be accompanied by a
reduction of nitrate concentration in Chalk ground water.
Efficient management of fertilizers may reduce the excess
nitrate available for leaching, and could include regulation
of timing and quantities of applications to take account of
individual crop requirements and residual amounts of
nitrate remaining in the rooting zone from previous crops
(Burns, 1975). However, Young, Hall and Oakes (1972) have
postulated that the increases in nitrate content of some
ground waters in the United Kingdom may be related to an
increase in arable acreage during the period 1939-45, as a
result of which organic nitrogen stored in the soil profiles
of long-term pasture became available for mineralization
and leaching. It is considered possible that changes from a
predominantly pastoral to an arable economy on other
areas may lead to similar increases in the rate of nitrate
leaching.
Q. Did you investigate any other chemical or biological
reactions that might interfere with the nitrogen movement
and form? Q. by M. Johnson. Have you ever detected
denitrification taking place at the saturated zone?
A. Investigation of potential biochemical nitrogen trans-
formations in the unsaturated and saturated zones is being
carried out and some of the preliminary results are
described by Young et al., 1976. No evidence has been
observed, to date, of in situ denitrification in the saturated
References
Fried, M., K. K. Tanji, and R. M. Van De Pol. 1976. Simpli-
fied long-term concept for evaluating leaching of
nitrogen from agricultural land. J. Environ. Qual. v. 5,
no. 2, pp. 197-200.
Greene, L. A. and P. Walker. 1970. See reference in text.
Burns, I. G. 1975. Simple methods of predicting the
leaching of nitrate from the root zone. In. Proc. Int.
Ass. Wat. Poll. Res. Specialised Conf. Nitrate as a
Water Pollutant, Copenhagen.
Young, C. P., E. S. Hall, and D. B. Oakes. 1976. See reference
in text of paper.
231
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/9-77-01^
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
PROCEEDINGS OF THE THIRD NATIONAL GROUND WATER
QUALITY SYMPOSIUM
5. REPORT DATE
June 1977 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Water Well Association
500 West Wilson Bridge Road
Worthington, Ohio 43085
10. PROGRAM ELEMENT NO.
1BA609
11. CONTRACT/GRANT NO.
68-03-2396
12. SPONSORING AGENCY NAME AND ADDRESS
Robert S. Kerr Environmental Research Lab.
Office of Research & Development
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
- Ada, OK
13. TYPE OF REPORT AND PERIOD COVERED
Final (3/31/76-1/31/77)
14. SPONSORING AGENCY CODE
EPA/600/15
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The Third National Ground Water Quality Symposium was held in Las Vegas,
September 15-17, 1976, in conjunction with the annual convention of the National
Water Well Association.
The Symposium was dedicated to the late Thomas P. Ahrens and the keynote
address was given by Charles C. Johnson, Jr., Chairman of the National Drinking
Water Advisory Council.
There were eight main sessions encompassing twenty-four technical papers.
These were concerned with the disposal of waste on the land, the movement of
pollutants in the subsurface, and artificial recharge. A special session was
dedicated to ground water in the Las Vegas Valley.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Water Quality
Water Resources
Water Pollution
Ground Water
Ground Water Movement
Artificial Recharge
13B
13. DISTRIBUTION STATEMENT
Release to public.
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
238
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
232
• U. 5. GOVERNMENT PRINTING OFFICE: 1977-757-056/6A58 Region No.5-11
------- |