WATER POLLUTION CONTROL RESEARCH SERIES • 16060 GRB 08/71
     THE NATIONAL
GROUND WATER QUALITY
      SYMPOSIUM
 U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to V7ater Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, B.C. 20^60.

-------
                           PROCEEDINGS
                                of the
         NATIONAL GROUND WATER  QUALITY SYMPOSIUM
                          Cosponsored by the
                    Environmental Protection  Agency
                               and the
                    National Water Well Association
                          August 25-27, 1971
                           Denver, Colorado
                 Dedicated to C. E. Jacob (1914-1970)
                        Contract No. 68-01-0004
                        Project # 16060 GRB
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, B.C. 20402 - Price $1.78

-------
                  EPA Review Notice
This report has been reviewed by the Environmental Protection
Agency and approved for publication.  Approval does not
signify that the contents necessarily reflect the views  and
policies of the Environmental Protection Agency nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
                         ii

-------
                    FOREWORD
THE OBJECTIVE
by Jay H. Lehra
     During the past decade, our nation's water  problems have
been front page news across the country. Everyone is aware of our
tremendous surface water  pollution problems, yet the threat to
our vast underground water resources remains remote to most of
the American  public as well as to  much of  the  scientific com-
munity.
     With surface water resources decreasing in both quantity and
quality, ground water, as the primary alternate supply, must be
protected where it is pure,  and improved where it is not. The
technology  to accomplish most of these  tasks should be made
available  to  the  engineers,  the architects,  the  planners, the
managers, and, of course, the ultimate users.
     In an effort  to focus public and scientific attention on the
broad  and basic ground-water pollution problems, either present
or developing, the first National Ground Water Quality Symposium
was planned.  Its objective was to
     "bring together the nucleus of  men, methods and ideas
     capable  of yielding  solutions  to problems  which will
     insure the protection and  restoration of the quality of
     our vast ground-water resources, whose development is
     destined  to double  and perhaps triple  in the coming
     decade.  It is intended that the  published transactions of
     the Symposium will serve as a  guide to the State-of-the-
     Art on ground-water quality."
     In the following 211  pages, the reader should find that these
 objectives  were met in a most comprehensive and  useful  manner.
 It is hoped that this manuscript will serve as a guide to potential
 problems  and  workable  solutions  which must  be intelligently
 considered if  our valuable resource is to be protected.
      Executive Director, National Water Well Association, 88 East Broad
 Street, Columbus, Ohio 43215.

-------
           PROCEEDINGS OF THE NATIONAL GROUND WATER QUALITY SYMPOSIUM
Cosponsored  by the Environmental Protection Agency and the National Water Well Association, August 25-27,
                   1971, Denver, Colorado.  Dedicated to C. E. Jacob (1914-1970).
                                 Contract No. 68-01-0004
                        TABLE OF CONTENTS

    TITLE PAGE
 iii FOREWORD-THE OBJECTIVE	Jay H. Lehr
  2 EPA'S ROLE IN GROUND-WATER PROTECTION 	Stanley M. Greenfield
  7 THE CONTRIBUTIONS OF C.  E. JACOB TO SCIENTIFIC
    HYDROLOGY AND ENGINEERING WORKS	Zane Spiegel
 10 SUBSURFACE DISPOSAL OF  LIQUID INDUSTRIAL WASTES
    IN ALABAMA-A CURRENT STATUS REPORT	William E. Tucker
 20 SUBSURFACE STORAGE AND DISPOSAL IN ILLINOIS	H. F. Smith
 29 FEASIBILITY OF RECHARGING TREATED SEWAGE
    EFFLUENT INTO A DEEP SANDSTONE AQUIFER	 Richard J. Schicht
 36 BULL SESSION 1-GROUND-WATER WASTE DISPOSAL RECHARGE AND REUSE
 45 PESTICIDE CONTAMINATION OF A SHALLOW BORED WELL
    IN THE SOUTHEASTERN COASTAL PLAINS	M. J. Lewallen
 50 GASOLINE POLLUTION OF AGROUND-WATER
    RESERVOIR-A CASE HISTORY	  D. E. Williams & D. G. Wilder
 57 PETROLEUM CONTAMINATION OF GROUND WATER IN MARYLAND	John R. Matis
 62 BULL SESSION 2-CHEMICAL CONTAMINATION OF GROUND WATER
 76 GROUND-WATER POLLUTION POTENTIAL OF A LANDFILL
    ABOVE THE WATER TABLE	M. A. Apgar & D. Langmuir
 97 GROUND-WATER POLLUTION AND SANITARY
    LANDFILLS-A CRITICAL REVIEW	A. E. Zanoni
111 EFFECT OF EARLY DAY MINING OPERATIONS
    ON PRESENT DAY WATER QUALITY	L. L. Mink, R. E. Williams,
                                                                         & A. T. Wallace
121 BULL SESSION 3-SOLID WASTE-ITS GROUND-WATER POLLUTION POTENTIAL
136 METHODS OF GEOLOGIC EVALUATION OF POLLUTION
    POTENTIAL AT MOUNTAIN HOMESITES	James P. Waltz
144 NITRATE IN GROUND WATER OF THE FRESNO-CLOVIS
    METROPOLITAN AREA, CALIFORNIA	Kenneth  D.Schmidt
159 THE USE, ABUSE AND RECOVERY OF A GLACIAL AQUIFER	Edward M. Burt
167 BULL SESSION 4-AQUIFER PROTECTION AND REHABILITATION
182 A SYSTEMS APPROACH TO MANAGEMENT OF
    THE HANFORD GROUND-WATER BASIN	D. B. Cearlock
193 SALTY GROUND WATER AND METEORIC FLUSHING OF
    CONTAMINATED AQUIFERS  IN WEST VIRGINIA	Benton M. Wilmoth
201 PROBABLE IMPACT OF NTA ON GROUND WATER	W. J. Dunlap, R. L. Cosby,
                                                               J. F. McNabb, B.  E. Bledsoe,
                                                                          & M. R. Scalf

-------
ERA'S ROLE IN GROUND-WATER PROTECTION3
by Stanley M. Greenfieldb
     The Environmental Protection Agency's mission with respect
to the nation's ground-water resources is not explicitly mentioned
in the laws under which our predecessor agencies operated, nor is
it mentioned in President Nixon's reorganization plan and message
to Congress  in July 1970 announcing the establishment of this
new Agency.
     It did not need to be mentioned specifically. It is implicit in
all the water quality laws carried out by the  Federal Water Pollu-
tion  Control  Administration  and  the Federal  Water  Quality
Administration. It  is implicit  in the  various public health  laws
under which the old Bureau of Water Hygiene established recom-
mended standards for drinking water and gave technical  assistance
to municipal water-supply systems throughout the country.
     Here is what the President said in his message to Congress in
1970 proposing the establishment of EPA:  "Our national govern-
ment today is not structured to make a coordinated attack on the
pollutants which  debase the air we breathe, the water  we drink,
and  the land that grows our food.  . .  For pollution control pur-
poses, the environment must be perceived as a single, interrelated
system. . .  The sources of air, water, and land pollution are inter-
related and often interchangeable. . ."
     "This reorganization," the President continued, "would per-
mit  response to the environmental  problems  in a manner beyond
the  previous capability of  our pollution control  programs. The
EPA  would  have  the capacity to do research on important
pollutants irrespective of the media in which they appear, and on
the impact of these pollutants on the total environment. . .  This
consolidation of  authorities would help assure that we do not
create new environmental  problems in the process of controlling
existing ones."
     I think that  is clear enough for anyone to understand. Water
is water—a basic  natural resource,  whether  it is on  or  under the
      Presented at the National Ground Water Quality Symposium in Denver,
Colorado, August  25-27, 1971 (sponsored jointly by the Environmental
Protection Agency and the National Water Well Association).
     "Assistant Administrator for Research and Monitoring, Environmental
Protection Agency, Washington, D. C. 20460.

-------
surface of the earth. The nation's waterways, used for transport,
industrial cooling and processing, recreation, and a host of other
purposes, are surface waters; and a large part of our environmental
protection  efforts  are  to  restore and  maintain surface-water
quality. But ve are  not  limited to surface water. This would be
absurd. A substantial portion of our total water is underground; in
many parts of the nation the only water resource  is underground.
     And  the line of demarcation between ground and surface
water is indistinct. It's not a line but  a continuum, variable from
place to  place and time  to time. Ground water becomes surface
water and  vice  versa.  Both  are  parts of the same  system,  the
hydrosphere.
     There's an analogy here with the atmosphere. Ground water,
especially the least-known portions that are deeply buried  and
move very slowly, can be likened to the stratosphere,  which mixes
very slowly with the troposphere below it, and in which pollutant
gases or particles have very long residence times and slow fall-out.
Yet  in the  recent  public debate  and  discussion over possible
pollution of the stratosphere by the  supersonic transport plane,
neither side  argued that the thin upper regions of the atmosphere
were too  high up to  be of environmental concern!
     Ground-water quality is a legitimate concern of  EPA, as this
Conference demonstrates. It  is probably  true that lack of specific
mention of ground-water resources perse  in the laws and executive
proclamations on which  EPA  is based has  caused us to  give
ground-water problems  less  attention  than other aspects of the
environment during the first nine months of our official existence.
This is an  oversight—if you  could  call it that, perhaps "lack of
emphasis" would be a better term—which  we intend to remedy.
     I endorse heartily the statement  issued  in October 1970 by
the then Commissioner of the FWQA, on  the disposal  of wastes by
subsurface  injection—deep-well disposal. That policy order  was
clear and brief:  two typed pages. After noting the increasing use
of deep-well disposal of hazardous wastes, the risk of contamina-
tion pathways,  the  order  put the FWQA on record  as opposing
such waste disposal "without  strict controls and a clear demonstra-
tion" that such disposals will not harm "present or potential sub-
surface water supplies ... or otherwise damage the environment."

-------
     The order then went on to list seven criteria to be "critically
evaluated" for all proposals for deep-well injection:

     1.  Alternative disposal  methods  have been  explored and
found less satisfactory in environmental risk.
     2.  Appropriate  preinjection tests  have been made to allow
prediction of what will happen to the wastes.
     3.  There is "adequate  evidence"  to show that the injected
wastes will not  interfere with present  or potential  resources and
will not produce "other environmental hazards."
     4.  The  best practical measures of waste pretreatment have
been applied.
     5.  The  injection system  has been designed and built in the
best possible manner.
     6.  Provision  has  been  made for  "adequate and continuous
monitoring"  of both the injection operation and its environmental
effects.
     7.  The  well  will be plugged  "below present or potential"
water resources whenever its use is discontinued.

     These  seven  caveats appear to give perfect  protection. But
they contain some circularities, some "weasel" words.
     What constitutes "adequate evidence" that there will be no
pollution?
      Preinjection  tests  must  "allow" prediction  of what will
happen to the wastes, but how reliable must these predictions be?
And how long a time must they cover?
      Pretreatment is obviously not as good as complete treatment,
so why call it "best"?
      How can alternative measures of  disposal ever be considered
 "less satisfactory" in terms  of  environmental protection, when
 complete  detoxification and  neutralization of the pollutant is
 included as  an  alternative? How can we neglect to consider such
 complete treatment as an alternative?
      What  constitutes "adequate  and continuous monitoring"?
 And when a disposal is sealed off,  does the monitoring continue?
 How long?
      The  FWQA  policy statement  concludes  with  a  ringing
 declaration  that  we  recognize  all  subsurface  injection as a
 temporary   means of  disposal, to be  abandoned when  better
 methods are developed, giving greater environmental protection.
      There appears to be a kind of schizophrenia here.
      Waste materials that we dare not release to that part of the
 environment we know the most about-the air, surface waters, and
 the  top layers of the land-can be safely contained by pushing
 them farther down, where our knowledge of fluid movements and
 physical and chemical and  biological changes is very scant indeed.
 It is just such out-of-sight, out-of-mind attitudes and actions that
 have brought about our present pollution problems.
      To say "this is only a temporary expedient till we find a
 better disposal method" just glosses over the unpleasant truth that
 we are doing things we have condemned in others.
      There is a parallel in the current concern over dumping wastes
 at sea.  It has long been assumed that if you take the stuff out far
 enough  and sink it deep  enough, no harm will  result-that the
 oceans are  so  voluminous they can dilute anything to innocuous
 levels. The Irish  poet, Oliver Gogarty, 75 years ago expressed it
 thus: "And up the  back garden,  the sound comes to me of the
 lapsing, unsellable, whispering sea."
      We know that  he was wrong; the sea is soilable. The earth's
 crust is soilable too, and vulnerable to damage by  man's activities.

-------
not only in ways that we  may  predict and make allowance for,
but also in unexpected ways.
     For example, the problems of liquid waste injection are not
simply of a geological and hydrodynamic nature. The problem of
the thermal and chemical fate of the waste materials should have a
major  role in  determining the areas and  amounts of  allowed
disposal.
     The result of not  considering  the  chemical-heat problem
was  demonstrated  when  the Hammermill  Paper Co., in Erie,
Pennsylvania, began operating  a sixteen hundred  foot well in a
dolomite formation.  Approximately  fifty-five thousand  barrels
per day of spent sulfite liquor containing a fiber, titanium dioxide,
clay and lignin-like compounds were pumped into  the well  at
pressures of eleven to thirteen hundred  pounds per square inch
between 1964  and 1968. When the well blew up, one hundred
fifty thousand  gallons  per  day of waste were spewed into Lake
Erie. Apparently the well blew when the injection tubing corroded
and the underground pressure was released to the casing.
     The most dramatic  instance  of unexpected  problems,  of
course,  was the apparent stimulation of earthquakes  in 1962 and
1963 and again in 1964-66 by the  injection of liquid wastes two
miles below the Rocky Mountain Arsenal near Denver, Colorado.
Less dramatic—but probably more important in the long run—are
the instances of gradual but massive changes in ground-water levels
and  ground-water quality  by mining, by oil  and  gas drilling,  by
saline water injection, and also by pumping up ground water faster
than its natural  recharge rate.
     Just as the cure of schizophrenia  begins with recognizing that
we have the disease, so there is great therapeutic value in candidly
acknowledging  that protecting  the nation's ground water, and
deep-well waste injection are likely to collide.
     They  are  not mutually exclusive, however,  any more than
septic tank sewage systems are  mutually exclusive with  surface-
water quality.  Both can co-exist with the proper controls. If the
soil  is  of  the  right type  and  is not overloaded, the physical,
chemical, and bacterial processes will completely purify the sewage
and maintain the quality of the surface water.
     Our specific problems with deep-well injection are: first, how
to identify and classify areas of safe injection so that fresh water
aquifers are  not encroached;  second,  how  to determine what
volumes of wastes can  be injected  safely; third, how to establish
chemical standards on wastes to minimize the dangers of injection;
and  fourth, how to monitor and record deep-well injections  to
insure the engineering standards of their operation.
     We have built up a considerable  body of knowledge  about
ground-water supplies, recharge times, flow rates, and the fluctua-
tions that can be  expected  over  long periods  of time. We need to
know a  lot more,  of  course, but  we should not denigrate the
knowledge we have, which has been  vital to our existence.
     We have  long experience  with deep-well  injection—tens  of
thousands of injection wells in  oil  and gas fields, for instance-
both for the disposal of unwanted brine and for the stimulation of
flow from wells that have been depleted.
     This long  experience, however, has not been  properly  moni-
tored in the sense expressed in the  FWQA criteria  previously
mentioned. Oil  men, in general, have been concerned only with
getting  rid  of the  stuff. When they put it back in  the same zone
from which they extracted the oil  and gas, there  is  little change
in the fluid gradient;  the natural dynamic conditions are at least
partially restored.  But  I  am  informed  that the  current and
increasing practice  is  to return  the brine to some other zone—

-------
usually the shallowest available salt-water aquifer—from which the
chance of migration into fresh-water aquifers or surface waters is
great. This has already happened in  certain parts of Texas and
Oklahoma. It doesn't happen immediately but over a considerable
period,  maybe  years after the injections start, so it  is hard to
trace and to establish the cause.
     Arthur Piper of the  U. S.  Geological Survey has suggested
the possibility that the noxious chemical wastes whose injection
triggered  the earthquakes at Denver might eventually migrate
eastward to outcrops and  ground-water source areas in Nebraska,
hundreds  of  miles  away.  Piper  said  that might take  several
thousand  years. He expressed the hope that "over that space and
during that time  the waste will have been rendered harmless by
sorption,  dispersion, and degradation. . .  Data for assessing the
hazard definitely are neither at hand nor  readily  obtainable. .  .
For  the  most persistently noxious wastes,  a responsible  society
cannot knowingly create even a remote hazard."
     It seems to me that the Denver earthquakes—and similar, less
intensively studied  cases in oil fields  in western Colorado, Texas,
and  Utah—have served a very good purpose in  alerting us to this
kind of long-term danger. There's nothing like an earthquake to
arouse public interest! And we  need public interest if we are to
get support for the research needed to find out what happens
down there in the dark and the heat and the pressure.
     We in EPA  have  a clear call to study, test, and evaluate all
aspects of the  ground-water  environment.  This includes, but  is
not  limited to, the  use of that  environment as a  depository for
wastes.
     Our  research  and monitoring efforts will be designed to
complement and  not conflict with those of the  Geological Survey,
NOAA, and other governmental agencies, both  State and Federal,
that are concerned with land and water use.
     Above all we  will need basic research in what happens to
water and other fluids down there in the dark and the heat and the
pressure.
     To accomplish this task, we must take full advantage of the
myriad of talents and resources that  are available at the universi-
ties, in the States, in  other  Federal  agencies, and in the private
sector.  This Symposium  has  resulted from our association with
Dr.  Lehr and the  National Water Well  Association. It has proven to
be one with many worthwhile aspects. It has collected those in the
field of ground-water resources who in their infrequent idle hours,
I am sure, have resolved many problems and found new approaches
to those yet unresolved. I think this gathering will go a long way to
show collectively the  importance  of sound water resources de-
velopment and more importantly the need for protecting our vast
underground supplies of fresh water. Perhaps  by our association
with those of you working with the day-to-day aspects of ground-
water management, and our association with the diverse member-
ship of the National Water Well Association, we can achieve that
most important goal in research—the goal of transferring technol-
ogy from the laboratory  to those  who  must ultimately make it
work.  !f it has made a  start in that direction, then this Symposium
has most  assuredly been successful.
     A while back  I quoted an Irish poet on the sea. Now I'd like
to quote  another Celtic bard, the Welshman, Dylan Thomas:  "The
force that drives the water through the rock, drives my green age."
     We need to  know a lot more than we do about the force that
drives the water through the rock.  If we are to  have a "green age"
for ourselves and our posterity, we need pure and abundant water
in the ground.

-------
The  Contributions  of C.  E. Jacob  to
Scientific   Hydrology  and  Engineering  Works
by Zane Spiegel
                  ABSTRACT
     C. E. Jacob was the first of a small group of hydrol-
ogists who followed C. V. Theis' lead in application  of the
theory of nonsteady boundary value problems to both well
hydraulics and regional ground-water movement. Some of
his scientific  contributions  were  the explanation of the
storage and elastic properties of artesian aquifers, the semi-
log method of aquifer test analysis, methods of analysis of
tests on unconfined leaky aquifers, application of doublet
theory to paired  polarized heat-pump wells, the theory of
step and constant-head aquifer tests by wells,  salt-water
encroachment, the tutelage of hydrology students at several
universities, solutions to hundreds of field problems en-
countered in  his extensive  consulting activities in many
parts of the world, and his method of determining aquifer
recharge  and  transmissivity  from the response of water
levels in  wells to varying natural recharge. One  of his last
works was a study of the natural recharge to Southampton
on Long Island, which showed that the recharge was much
less than the 21 inches per year assumed previously.
     I  would like to acknowledge the cooperation
of several of C. E. Jacob's clients and close associ-
ates in  the preparation of this paper, among them
C. V. Theis, Stan Lohman, John Ferris of the U.S.
G.S., and Bill  Guyton,  formerly of the U.S.G.S.,
and  especially  to  Bettie  Morgando of the firm of
C. E. Jacob and  Assoc., who is still working very
hard for the  firm to take care of all the details
involved in settling his business affairs.
     The purpose  of this Symposium, as stated on
the  cover  of  our Program  Guide,  is "to  bring
together a  nucleus  of   men, methods and  ideas
capable  of  yielding solutions to problems which
must be solved  to insure  the  protection and
restoration of the quality of our vast ground-water
resources, .  . . "  C.  E. Jacob  has already  provided
the means to solve  many of our problems,  if we
could look  at his many reports, particularly  in his
last twenty years of consulting.
     C. E. Jacob began his career in ground  water
with  field experience with  the  U.  S. Geological
Survey. Assisting R. M. Leggette, he made detailed
studies  of water  level changes in the complicated
aquifer systems of Croton Valley and Long Island,
New York. This work gave him the practical experi-
ence  needed  to  arouse  his  scientific interest in
learning and  then  explaining  the  physical and
     Presented  at the National Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
     bGround-Water Hydrologist, P. O. Box 1541, Santa
Fe, New Mexico 87501.
mathematical basis of observed phenomena. He was
the first of a small number of hydrologists who
followed the lead of C.  V.  Theis  in applying the
theory of nonsteady boundary value problems to
both  well  hydraulics  and regional  ground-water
movement, in the English language.
     In  his  first  five years  of experience, several
important papers  were published  on  the storage
and  elastic properties  of aquifers  in coastal  plain
sediments.  These properties  had been recognized
qualitatively by Theis previously, but Jacob  pro-
vided the first quantitative explanation of artesian
storage,  related  this explanation to  the  tidal and
barometric effects noted by others, first recognized
the effects of railway  trains on water levels, and
later coined the term  "storativity" which unified
storage  concepts.  As a result  of  his  detailed ac-
quaintance with  highly permeable areas of  Long
Island and Texas, Jacob was  the first  American
hydrologist to connect the fact that natural fluctu-
ations in water levels  of wells are  the  best indi-
cators of rates of recharge  to aquifers with the
theory of nonsteady boundary-value problems, to
give a direct method for determining both the true
recharge rate and the average effective transmissivi-
ty. This  method requires the selection of only one
assumed or independently determined value—the
specific yield. Perhaps due to man's natural inclina-
tion to be more concerned with disturbances caused
by himself than  with those  caused  by  natural
phenomena, Jacob is best remembered for his con-
tributions  to  understanding  of  storativity, the
interpretation  of the  extremely valuable  semi-
logarithmic plots of pumping tests, and the general-
ization of Theis' model for aquifer test analysis to
tests on unconfined and leaky artesian aquifers, and
to  step  tests  and determination of effective well
radius. Interchange of  ideas with more experienced
Survey hydrologists  such as Theis, Lohman, Ferris,
and Guyton contributed greatly to  the  success of
his early research. An  example of  the value of this
kind  of interaction  was the 1952 publication by
Jacob and Lohman  of a solution  to the constant-
head well problem  which  is  useful  in areas of
flowing wells. His ability in mathematics  permitted
coordination with the ideas of Muskat and Hubbert
developed in the petroleum industry.
     Twenty-five years before a logical terminology7
and system of consistent units were agreed to by
the  profession,  Jacob  was  appointed  by   Stan

-------
Lohman to chair a subcommittee of the American
Geophysical Union to improve usage of the term
permeability  and  its  units. Many of  the sub-
committee's recommendations in 1946 were finally
adopted by another committee chaired by Lohman
after many thousands of students and self-taught
hydrologists had labored with the ridiculous system
Jacob and others had tried so long to improve.
    The end of his first fifteen years of remarkable
original contributions to  hydrology was marked in
1950 by publication of the well-known chapter on
"Flow of  Ground Water" in the treatise "Engineer-
ing Hydraulics." This work summarized the results
of most of his original work to that date. A large
proportion of this important work was his original
work.
     Some of the important topics of that work
were (1) a  brief exposition of Darcy's Law correctly
stated  in terms of hydraulic head, not pressure, in
a form similar to  that of Hubbert's correction of
one of the many unfortunate fallacies in hydrology;
(2) a discussion of the range of validity of Darcy's
law, (3) the difference between the terms permea-
bility  and transmission constant; (4) the velocity
potential; (5) derivation of differential  equations
and their relation to boundary-condition equations;
(6) the  range  of validity  of steady  radial-flow
equations; (7) the effect of a well on a uniform
flowfield, (8)  recirculating pairs of wells (doublets)
and well-stream systems; (9) the general theory of
images; (10)  specific capacity; (11) the superposi-
tion theorem; (12)  tidal fluctuations;  (13)  well
characteristics; (14) effects  of a well between two
parallel streams; and (15) adjustments  for uncon-
fined flow.
     The  next  few years  (the  early  fifties)  are
notable for the large number of important studies
that were  made, not so much by himself, but by his
colleagues and students, on various types  of well
models and   on  more  detailed  aspects  of  the
correlation of precipitation and ground-water levels
on Long  Island. These  years were especially  im-
portant for  the beginning  of publication of an
astounding number  of papers on well hydraulics
and stream-connected flow by Prof. Mahdi Hantush,
Jacob's most renowned former student.
     After 1947,  as ground-water consulting  be-
came Jacob's dominant activity, formal publication
of important  discoveries waned, but his character-
istic custom of attacking a  practical problem with
new methods continued. Engineers, like Cipolletti,
have been honored  by having structures they de-
signed named after them, but the great engineering
works that benefited  from Jacob's analyses in his
last twenty  years have  neither his name nor  a
surface expression,  and for many  of the  works,
even his reports are only obscure sections of large
8
tomes  by the  engineering  firms to  which  he
rendered able assistance.
     Much of Jacob's work was done as a "con-
sultant's  consultant," wherein he provided a broad
knowledge of ground-water principles and a phe-
nomenal  ability  to  solve problems.  Most of  his
work can be classified in categories such as (a) the
effects of a given water-supply development; (b) the
mechanism and effects of mine  drainage; (c) the
effectiveness of drainage by wells; (d) well injection
procedures; and (e) the control of uplift pressures
near dams.
     The scientific contributions contained in many
of these reports may some day be considered as im-
portant as those of his first fifteen years of hydro-
logic contributions. Jacob developed a form for his
major  consulting reports that might well be taken
as a model by consultants in many fields. Typically,
these reports contained a synopsis giving his impor-
tant conclusions; a concise statement of the prob-
lems and the circumstances which  led  to  his
involvement in them; a section  on the important
physical   principles  on which  his analyses  were
based,  with detailed  references, many  of  which
were his own; definitions of special technical terms
used in  the  report; a concise  summary  of the
geographic, geologic and hydrologic aspects  of the
problem; and finally  the hydrologic core of the
report, which frequently contained significant orig-
inal contributions to scientific hydrology or engi-
neering practice.  In  addition, it was his practice to
add detailed appendices of the basic data on which
he based his conclusions.
     The detail and originality of Jacob's treatment
of all the sections of his consulting reports can be
fully appreciated only by reading these reports.  I
hope that some day a complete catalog of his works
can  be made available to the hydrologic profession,
and that the full reports will eventually be made
available  for  reference.  Many  of the consulting
reports have  already been released for public use,
and efforts are being made to obtain the release of
the  others. New Mexico Institute of Mining and
Technology in  Socorro,  New Mexico,  has been
designated custodian of these important contribu-
tions.  The task  of  cataloguing  Jacob's works  for
general library use is an enormous effort, far  be-
yond the present resources of New Mexico Tech.
If, during these  three days, an effort  to  obtain
additional support for this task can be initiated, the
dedication of this Symposium to the memory of
C. E. Jacob will take on a new meaning.
     One of the most interesting characteristics of
Jacob's  wide  range of consulting reports was  the
emphasis which he put on a careful description of
the  long-range effects of each  project with which
he was associated. This emphasis was undoubtedly

-------
due to his recognition that neither the public nor
his clients were generally aware of the inevitability
of diversion of streamflow by wells, the  far-reaching
effects of wells in  slightly leaky aquifer  systems,
and the effects of aquifer nonhomogeneity on the
rate of movement of contaminants. His  understand-
ing of hydraulic effects in complexly layered aqui-
fers led him to conclude in several reports on both
agricultural  and  mining  problems  that  drainage
works were more effective if installed in deeper
semiconfined beds than in the  shallow aquifers that
ultimately were to be dewatered. This concept was
applied in the Yuma area to reduce the number of
wells  needed  to  accomplish  a given depth  of
drainage.
     The great magnitude of effects of many of the
projects which Jacob studied inevitably has brought
about considerable controversy, or at least concern,
as to the public welfare. For example, the economy
of the entire country of Israel is dependent on the
validity of  computations  made on the  basis of
Jacob's theories; enormous costs are  at  stake in
the drainage of the Indus Valley in West Pakistan;
international relations  may  be affected  by the
success or failure of the Yuma drainage project in
southern Arizona, along the Mexican border.
     Through many years of activity in  problems of
production of fresh water  in areas underlain by
salt water, Jacob recognized that future generations
would be faced with enormously serious shortages
of fresh water. As a result of this foresight, in 1965
he patented two improved methods for producing
fresh water from wells. Negotiations are in progress
by  Jacob's estate   to  develop   and  test  these
patented methods.
     In recent  years  possibly  one of the  most
significant contributions  to society made by the
hydrology profession was a report on  "Evaluation
of  Potential  Impact   of  Phosphate  Mining  on
Ground-Water  Resources of Eastern North  Caro-
lina," prepared by a Board of Consultants of which
Professor  Jacob was Chairman. The report,  made
for the North  Carolina Department of Water Re-
sources in 1967, is typical of Jacob's reports in
form and  content.  While  the  report contains a
standard  application  of the  well-accepted  tech-
niques of basin-wide engineering analysis,  based on
his  own leaky-aquifer  theory, it also contains one
special feature that distinguishes it from most other
reports of this  type—a review  of certain aspects of
water law, with a recommendation for the enact-
ment  of  State legislation  establishing  a permit
system for all new large uses of ground water.
     As a  direct result  of this report by Jacob's
Board of  Consultants, the North Carolina General
Assembly enacted the  "Water Use Act of 1967,"
which requires a permit from the State for large
diversions of surface or ground  water in any area
previously declared to be  a "capacity-use area."
The  legislation follows  Jacob's  recommendations
closely. The  rapid  action of the  North Carolina
General  Assembly  in  adopting  Jacob's  recom-
mendations is evidence of the clarity and forceful-
ness of the report.
     One of Jacob's last consulting reports (Decem-
ber,  1968) was an extension  of  earlier work by
himself and Boris Bermes, a former student at the
University of Utah, in which he applied variations
of his important nonsteady mathematical model of
water-level fluctuations  to  Southampton, a penin-
sular  portion  of Long Island.  Using a value  of
specific yield that had been determined by detailed
U.S.G.S.  studies in a nearby  area, he computed
values of average annual recharge ranging from 8 to
14 inches per  year, substantially lower than  the
previously accepted value of 21 inches per year that
had been based on very rough estimates. Although
engineering consultants working on similar prob-
lems in central and eastern Long Island had rejected
Jacob's method and  conclusions, the  basic  data
available  for similar areas of Long Island, southern
New Jersey,  and Cape Cod confirm Jacob's con-
clusions.  Detailed analysis of all  the factors affect-
ing recharge  have revealed  valid  reasons for  the
lower values he computed. Re-evaluation of  Jacob's
1945 recharge computation for central  Long Island
on the basis of a revised specific yield of 0.25 gives
a recharge value of 16 inches per year, correlative
with those determined  from Jacob's  later  South-
ampton work. Revised ground-water outflow com-
putations,  based on  transmissivity values deter-
mined by correct applications of Jacob's semilog
and  leaky aquifer  methods of analysis, and by
refinements of Jacob's methods by Hantush, cor-
roborate  the conclusion that Long Island's recharge
(and hence  natural discharge)  is  much closer to
Jacob's value of 16 inches per unit area than to 21
inches per unit area per year,  based  on the con-
venient roundness of the ratio  21 to 42 inches of
precipitation. The implications of a twenty percent
overestimate   of  the  natural  water  supply  are
enormous, when it  is recalled  that water-supply
projects  do not generally  contain any  factor of
safety, and millions of people may depend on the
consequences of a rough estimate.
     In the final analysis, it may come to pass that
Jacob's real contribution will be recognized to be
the foundation which he  laid for a greater social
awareness of the importance of water in our future
life,  by  the combined effects of  all his scientific
and  engineering studies. No matter what you or  I
do  to attempt to  solve these  problems, no  real
solution will  be attained without this social aware-
ness by ourselves and the public.
                                              9

-------
Subsurface Disposal  of Liquid  Industrial  Wastes
in  Alabama—A  Current  Status  Report"
by William E. Tuckerb
                  ABSTRACT
     Four subsurface disposal wells have been drilled and
completed in Alabama. These are: Stauffer Chemical Com-
pany, Mobile County; Ciba-Geigy, Inc., Washington County;
U. S. Steel Corp., Jefferson County; and Reichhold Chemi-
cals, Inc., Tuscaloosa County. The Geological Survey  of
Alabama has been directly  involved in all four projects.
The Survey served as  a consultant to the Alabama Water
Improvement Commission, the State agency responsible for
protection of surface and ground water in Alabama, on the
Stauffer and Ciba-Geigy projects,  and as consultant and
supervisor on the U. S. Steel Corporation and Reichhold
Chemicals, Inc., projects. These projects were undertaken
as a research  effort  to insure that the responsible State
agencies are fully cognizant of all aspects of this method of
waste disposal. It is  a policy in Alabama that subsurface
disposal is permissible for some wastes if the well is properly
designed and completed in an appropriate geologic environ-
ment and if conventional methods of waste treatment have
been evaluated and proved to be inadequate.
     The Stauffer well, operating at  75 gallons per minute
and 500 psi, is the only subsurface disposal system, other
than oilfield brine disposal wells, that is currently in opera-
tion. The Stauffer and Ciba-Geigy wells are in the Coastal
Plains geological province and the U. S. Steel and Reichhold
Chemicals, Inc., wells are in Paleozoic sediments of the
Warrior Basin. A general discussion of the geology, drilling,
completion, and  testing techniques  is presented for the
two geologic provinces involved.

               INTRODUCTION
    The transition of Alabama in the last ten years
from  a raw material producing  State to a manu-
facturing State has  placed an increasing burden on
our water resources. In  1965, the State legislature
amended the Water Improvement Commission Act
creating  an  agency to  protect  the surface-  and
ground-water supplies of our State and  removing
the grandfather industries from exemption.  The
Geological Survey of Alabama is authorized by the
legislature to  conduct  studies on  the  occurrence
and availability of surface and ground water. This
responsibility has been expanded to consider water
      Presented at the  National Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
     "Geological Survey of Alabama, P. O. Drawer O,
University, Alabama 35486.
usage and the environmental impact of industrial
and  population  evolution. The Survey  serves as a
consultant to the Water Improvement Commission
in the fields of geology,  hydrology, and geologic
engineering.
     As subsurface disposal of industrial waste be-
came popular in  the  middle  1960's, the Survey
undertook research projects to determine any and
all aspects of this method of waste disposal. To date
the Survey has  served as regulatory supervisor on
three wells and directed the total program on two
wells.  In January,  1971,  a  general policy was
adopted  by the Alabama Water Improvement Com-
mission regarding  subsurface disposal in Alabama.
The  general  policy approved of this method of
disposal provided that all other methods had been
examined and exhausted; and geologic and hydro-
logic conditions exist such that the environment is
protected to the maximum extent possible.
     Application for a subsurface disposal system
in the form of  a feasibility study is made  to the
Alabama  Water Improvement Commission.  The
Geological Survey  reviews the material  presented
and  makes its own recommendation to  the  Com-
mission.  If the plan is approved, the applicant  is
allowed to drill and  test the  well.  The applicant
then applies for permission to utilize the well and
information  gathered while drilling  and testing  is
reviewed by the Survey and further recommenda-
tions regarding use, requirements, and monitoring
procedures are  prescribed. So far, five  subsurface
disposal wells  have been drilled and completed in
accordance with this policy. These are: (1) Stauffer
Chemical  Company at Mobile;  (2)  Ciba-Geigy at
Mclntosh—2 wells; (3) U. S. Steel in Birmingham;
and (4) Reichhold Chemical in Tuscaloosa (Figure
1).

            GENERAL GEOLOGY
     In order to form a geologic frame of reference,
the State can be divided  into three major geologic
provinces. These are:  (1)  the Crystalline area in
east-central  Alabama,  (2)  the  Warrior Basin in
northwest Alabama, and (3) the  Coastal Plains in
10

-------
                           O   OILFIELD

                             .  INDUSTRIAL
                                                    Fig. 2, Generalized cross section of north central Alabama.
Fig. 1. Disposal wells in Alabama.
Fig. 3. Generalized cross section of south Alabama.
south  and southwest  Alabama. A cross  section
through  the  Warrior  Basin  exhibits sedimentary
formations ranging from Pennsylvania!!  to Cambri-
an in age. Disposal horizons are chiefly available in
the Ordovician and Cambrian Ordovician by virtue
of secondary porosity  in the form of fractures and
solution  cavities.  A  cross  section through  the
Coastal Plain area exhibits sedimentary rock from
Miocene to Jurassic age with disposal horizons
available in all formations by virtue ot intergranular
porosity in sandstones and  secondary  porosity in
limestones and dolomites.
     The Stauffer and Ciba-Geigy wells are in  the
Central Gulf Coastal Plain geological province and
the easternmost extension of the Mississippi Interi-
or Salt Dome Basin. Most o! the geologic structures
found in Lower Cretaceous or younger sediments in
this basin are the result of movement of the under-
lying Louann Salt.  Salt at depth,  responds as  a
plastic medium and  will  move into zones of weak-
ness in response to sediment  loading.  Structures
formed as positive features by  salt swells or domes
and as collapse-type  features such as grabens where
salt  was removed  occur  in  southwest  Alabama
(Figure 4).
     The most prominent structural features within
the  salt  basin  in  southwest  Alabama arc  the
Hatchetigbee  anticline,  the  Jackson  rault-Klepac
dome, the  Mobile graben, and the  Giihertown-
Coffeeville-West  Bend.  Pollard and  Bethel fault
zones. Other  important structures are the domal
anticlines at Citronelle, South Carlton and Chntom
and the piercement salt dome at Mclntosh.
     A complex, north-south oriented fault  system
known as the Mobile graben extends fn>ni Jackson
Alabama, south  to Mobile Bay.  The Jackson  tauli  is
the northernmost  fault on the east flank  of  the
graben system.  It is down thrown to the wev  and
in the subsurface exhibits approximateK  5,(>uo  teet
of throw at a  depth  of 2,450 reel  where it  was
penetrated by the Champion  Klepac No.  l  well
drilled by  Humble Oil and Refining fomuanv in
1955. The large amount ot displacement probably
resulted from  subsidence as salt was being removed
to supply domal growth of the Klepac dome on the
upthrown site.

-------
Fig. 4. Major structural features in Alabama.
     The major taulr representing the west flank of
the grabcn opposite  the  Jackson  fault has never
been  penetrated. It  is possible  that the northern
part of the Mobile  graben  is only  a  half-graben
where the Jackson fault comprises the east flank
of a structural low.
     Farther to the south, faults representing both
east and  west flanks of the graben  have been
mapped. The graben, as determined  from existing
well control,  turns westward north of Mobile Bay
and can be projected into Mobile  County west of
the Bay.
     The  Mclnrosh   salt  dome  in  Washington
Count}' is the shallowest  piercement salt dome in
the Mississippi Interior Salt Dome Basin. The dome
pierces Miocene sediments within  410 feet of the
surface of the ground and is approximately 1 mile
in diameter. There is little or no well  control for
structurally mapping radial and peripheral faulting
commonly associated with   the piercement  salt
dome. The location  and  conliguration of the  rim
synciine immediately adjacent to  or surrounding
the Mclntosh dome is interpretive.  It is believed to
have  been  formed by subsidence  above the area
where salt flowed toward the center  of the  salt
dome.
     The deposits  of Late Cretaceous  age in Ala-
bama crop out in a crescent-shaped belt across the
12
central part of the State. In western Alabama these
beds dip  southwestward and, in ascending  order,
include the Tuscaloosa Group, which comprises the
Coker  and  Gordo  Formations; the McShan  and
Eutaw Formations;  and the  Selma Group,  which
comprises the Mooreville and  Demopolis Chalks.
the Ripley Formation, and the  Prairie Bluff Chalk.
In eastern Alabama the beds dip southward and the
chalk formations of the Selma Group merge lateral-
ly into formations consisting chiefly of sand  and
clay. Along the Chattahoochee River deposits of
Late Cretaceous age, in  ascending order, are repre-
sented by the Tuscaloosa Group (undifferentiated),
the Eutaw Formation, the Bluff town and Ripley
Formations, and the Providence Sand.
     The  Upper Cretaceous  Series  includes two
important faunal zones, the Exogyra ponderosa and
the Exogyra costata,  that can  be traced throughout
the Atlantic and Gulf Coastal Plains. The Exogyra
ponderosa zone  extends stratigraphically from the
Tombigbee Sand Member of  the Eutaw Formation
almost to the bottom of the Ripley Formation, and
the Exogym costata zone extends from the  top of
the Demopolis Chalk to the Prairie Bluff Chalk.
     The  Tertiary formations in Alabama  consist
predominantly of marine elastics, and are  transi-
tional  in character between the clastic and largely
nonmarine formations of Mississippi and the  car-
bonate rocks of the  Florida peninsula. In Alabama
the Paleocene, upper Eocene,  and Oligocene forma-
tions grade into limestone; one of the lower Eocene
formations grades into  limestone downdip  in the
subsurface;  and  the middle Eocene formations
become increasingly  calcareous  eastward. Alabama
and  northwestern  Florida  are included  in  the
eastern part of  the North Gulf Coast sedimentary
province  characterized by clastic deposits chiefly,
in contrast to the carbonate anhydrite fades in the
Florida peninsula sedimentary province (Toulmin,
1955).
     The U. S. Steel and Reichhold Chemicals, Inc.,
wells are in Paleozoic formations. Geologically, the
Reichhold plant site  is located in the Black Warrior
sedimentary basin. The Black Warrior is a wedge-
shaped basin extending across Alabama and Missis-
sippi and contains considerable  thickness of clastic
and  carbonate sediments. The  basin is a negative
feature  bounded to  the  north by  the Nashville
dome,  to the east  and south  by the  ridges  and
thrust faults of the Appalachian front, and extends
to the west into  the State of Mississippi.
     The geologic structure in  most of the Black
Warrior  basin  is fairly  simple, with formations
generally  dipping to  the  south or southwest. In
Alabama, the salient structure existing in the Black
Warrior basin is  the Sequatchie  anticline. This anti-

-------
cline, a.' subsurface extension of  one of the Ap-
palachian  foreland ridges,  plunges  beneath the
surface in Blount County, Alabama, and probably
extends far to the southwest, perhaps as far as the
Mississippi  State line. The Sequatchie anticline in
the subsurface has been referred to as the Blounts-
ville or Browns Valley anticline. The Sequatchie
anticline passes  beneath Tuscaloosa  County, but
the axial trend fades out and cannot be traced on
the surface beyond the  eastern half of the county.
Semmes (1929) shows  the  anticlinal  axis turning
south in Tuscaloosa County, but detailed structural
mapping supporting this trend is not included in
his report.  The  existence of an  unconformable
Cretaceous veneer  over the western  part of the
county prohibits the use of Pennsylvanian outcrops
in tracing the trend of the anticlinal axis. Moreover,
thickening   of  Pennsylvanian  sediments  to  the
southwest  along  the  direction of axial plunge  of
the anticline coupled with the unconformity exist-
ing at the Mississippian  contact may tend to mask
the trend of the Sequatchie anticline in the older
and  deeper  sediments, from  reflection in  the
Pennsylvanian.
      Semmes (1929) shows  the  Warrior syncline,
the complimenting structure to  the  north  of the
Sequatchie  anticline,  trending northeast to south-
west with  the  axis passing a few miles east of the
city of Tuscaloosa. For the  reasons given previous-
ly, it is possible that the main structural axis of
both the  Sequatchie  anticline and the companion
Warrior syncline extend to the southwest as far as
the Mississippi State line with the dominant struc-
tural expression existing in pre-Pennsylvanian sedi-
ments.  If this  is the case, the anticline may pass
north of  the city  of  Tuscaloosa, and not be  re-
flected strongly in Pennsylvanian age rock.
      Detailed mapping of  the  coal  beds  of the
Pennsylvanian sequence indicates that minor fold-
ing and faulting is widespread in sediments of this
age in the basin. Anticlinal trends are local and tend
to disappear rapidly.  Faults show small displace-
ment and are of limited areal extent. It is possible
that many faults are confined to the Pennsylvanian
and fade out before reaching the Mississippian. In
general, the  magnitude of folding and faulting in
the Pennsylvanian decreases  to the northwest away
from the major thrust faults marking the boundary
between the Black Warrior basin and the Appa-
lachian front.
      The  Birmingham anticline is the major struc-
tural feature in the Fairfield area. It is a breached,
asymmetrical anticline, trending northeast, under-
lain by limestones, shales, sandstones, chert,  dolo-
mite  and   some  iron  ore   seams.   Rocks  from
Cretaceous to Cambrian outcrop  in  the Fairfield
area,  with Jurassic, Triassic, and  Permian rocks
not being represented. The Mesozoic and Paleozoic
sediments generally dip to the southeast except
where folded or faulted.
     Faults  associated with the Birmingham anti-
cline include thrust faults, strike-slip faults, reverse
and normal  faults, and relaxation faults. The geol-
ogy of the area is  often complicated by the pres-
ence of so many folds and faults.
     Near Vance, Alabama, the Birmingham anti-
cline plunges to the southwest and is overlain by
Mesozoic sediments of the Tuscaloosa Group.

            NATURAL RESOURCES
     The Coastal  Plain  of Alabama is the most
important ground-water producing area in Alabama.
The consistent occurrence  of good water-bearing
sand  and gravel beds make the development  of
large-capacity wells  quite   common.  Most  out-
cropping sands in this area are considered good to
excellent aquifers. Water  well depths vary from a
few feet to  1,700 feet and yields range up to over
1,000 gpm (gallons per minute). Well yields in the
Paleozoic vary considerably and, although topogra-
phy is important in locating a successful well, the
contributing factor for a high yield well is  the
penetration of solution channels and cavities within
limestone reservoirs. Yields range from a few gal-
lons to several thousand gallons per minute (350 to
3,500 gpm).
     The Citronelle Field, the South Carlton Field
and the  Tensaw Lake  Field are located in the area
of the Ciba-Geigy wells and the Stauffer well. Salt,
limestone, and lignite are other mineral resources
that are being developed or have potential as miner-
al deposits in the Coastal Plain.
     Mineral resources in the area of the U. S. Steel
and Reichhold, Inc., wells  are primarily iron ore,
coal, and limestone.

        DISPOSAL WELLS IN ALABAMA
Stauffer Chemical Company Well
     The Stauffer Chemical Company disposal well
was the first industrial subsurface  disposal system
drilled and completed in Alabama. This well  was
drilled to a total depth of 4300 feet, and eventually
completed through perforations at 3400 feet with a
tested injection rate of 420 gpm at a surface pres-
sure of 400 psi. This well has been  in operation for
21A years without interruption, except for a remedi-
al acid treatment.

Ciba-Geigy Wells
      The Ciba-Geigy  Corporation  has  drilled  two
subsurface disposal wells. The first well was drilled
to a total depth of 7500 feet and final completion
                                              13

-------
has been effected at 3800 feet through a screen and
gravel pack. An injection rate of 420 gpm at 800
psi has been established for this well. The number
2 well was drilled to 2500 feet and completed at
2000 feet through a screen  and gravel pack. Fresh
waters occur in this area to a depth of 600 feet.
Surface casing has been set on each of these wells
at approximately 1100 feet and cement circulated
to the surface. The number 2 well has an injection
capacity of 800 gpm at 300 psi.

     Two wells have been drilled and completed in
the older sedimentary rock of north-central  Ala-
bama. These were for U. S. Steel  Corporation in
Birmingham and Reichhold Chemical Company in
Tuscaloosa.  Both wells  are  of the open-hole com-
pletion type and penetrated  the same rock units.
U. S. Steel Weil
     The U. S. Steel well has an injection capacity
of 300 gpm  at 600 psi surface pressure. The well
was drilled and completed to a total depth of 6,072
feet. All  operations were under  the  direct super-
vision  of the Geological  Survey  of Alabama as a
research project in the field of deep-well disposal
of aqueous wastes. The project was financed by a
research grant from U. S. Steel Corporation to the
Alabama Geological Survey. All steps taken in the
drilling and completion  of the well were designed
for the protection of surface and subsurface water
resources and other natural resources of the area.
     Electric logs,  cores, drill stem tests, and in-
jectivity tests run  during the drilling of the well
indicated the feasibility of completing the well as a
deep-well disposal system. The disposal horizon was
completed "open-hole" from  4,415 feet to 6,072
feet. The well was cased and cemented as follows-.

16" casing set at 64 feet     —Cemented to surface
103/4" casing set at 1,100 feet—Cemented to surface
75/8" casing set at 4,415 feet -Cemented to 2,720
                            feet
27/8" injection tubing set at 4,434  feet

     The  results of injection tests indicate that
either  of two zones have a sufficient capacity to
handle the present volume of pickling  liquor waste.
The  uppermost zone has the capacity of accepting
a greater volume-rate at a lower pressure  than the
lower  zone.  The well is  so  constructed  that the
lower zone would commence to  take fluid in the
event the upper zone fails to function  properly.
     Injection tests run during and after the com-
pletion indicated the well's ability to  receive fluid
without pressure build-up. The absence of pressure
build-up after extended injection indicates a sal-
aquifer of considerable areal extent and capacity.
     The  primary injection  zone,  4,415  feet to
to 4,630  feet, is  a  fine grained, very calcareous
sandstone  with  shale laminations.  Porosity  and
permeability is present in the rock unit by virtue of
fractures.   The  zone  of  secondary importance,
5,550  feet to 5,770 feet, is a dolomite with frac-
tured porosity and permeability.
     Due  to the high percentage  of carbonate ma-
terial in the sandstone and almost total carbonate
composition of the dolomite, consideration must be
given the reaction  between the waste FeCl2 + HCL
or FeSO4 + H2 SO4 and the formation constituents.
Injection  of waste HC1 effluent  into a carbonate
formation will result in a dissolution of the soluble
parts of the rock and will  cause  an increase in
permeability, thereby increasing the injection rate
and/or reducing the pressure required. The reaction
products will be water, carbon dioxide, and calcium
chloride—the latter  two being  soluble  in water.
Adversely,  the pH increase resulting from  these
reactions may cause Fe(OH)3 precipitate and pos-
sible plugging.

    Fe+" + 3(OH)~ = Fe(OH)3 (gel precipitate)
    Fe"  + 2(OH)~ = Fe(OH)2 (precipitate)
     In a  sulfuric acid reaction with a carbonate
formation,  the calcium released will be  reprecipi-
tated  as CaSO4  (anhydrite) or CaSO4  •  2H2O
(gypsum). The high iron content of the waste efflu-
ent (1.2 moles/liter) coupled with the pH increase
from mineral reactions will cause precipitation of
Fe(OH)3 gel.
     There  is considerable  controversy among ex-
perts as to whether the precipitation resulting from
acid waste injection into a  carbonate reservoir will
plug  the   pore  space and  reduce  permeability.
Fortunately,  experience   has  shown  that  low
strength acid waste may be injected into carbonate
strata.   Hammermill  Paper  Company  at  Erie,
Pennsylvania,  and Dupont at  New Johnsonville,
Tennessee, have successfully injected acid waste in-
to carbonate strata.
     From  all  indications,  the  porosity  of the
injection  zone in  the subject  well is  fractured
rather  than intergranular. This  should prove more
favorable to acid injection because less surface area
is available for reaction, and the "zone of reaction"
should extend to a greater distance from the  well
bore. Considering  the  two extraneous precipitates
caused by the reaction of H2 SO4 with a carbonate
compared with the HC1 reaction, it must be stated
that the HC1 reaction would be  more compatible
with the salaquifer water and rock than the H2 SO4
14

-------
pickle  liquor reaction.  However,  if  the H2 SO4
pickle liquor was reduced in acid strength by dilu-
tion  of 1 part acid to 1 part water or 1 part acid to
2 parts water, there  is no reason why  the well and
receptor  formation should not function equally as
well.
Reichhold Well
     The Reichhold well  has an injection capacity
of 600 gpm at 700 psi. A generalized surface treat-
ment  facility  includes  pretreatment  filtration,
storage, further filtration and injection.
     Initial feasibility and geological studies for the
Reichhold well were made by the Geological Survey
of Alabama and Deep Well Pollution Control Cor-
poration. The results of this study were presented
in a  report entitled "Study of the Potential of Sub-
surface  Disposal  in the Tuscaloosa  Area," dated
September 30, 1968. All aspects of the project were
under the direct supervision of the Geological Sur-
vey of Alabama as a research project in the field of
deep-well disposal  of  aqueous wastes. The project
was  financed by a  research grant from the Federal
Water   Quality  Administration  and  Reichhold
Chemicals,  Incorporated.  The objectives  of  this
phase of the project were to design and complete a
well that would assure protection of surface and
subsurface water resources and natural resources of
the  area,  and to locate a horizon in the deep sub-
surface formations that would be receptive to the
plant's waste effluent.
     As in the case of the U.  S. Steel well, electric
logs, cores, drill stem  tests  and injectivity  tests run
during the drilling of the well  indicated the feasibil-
ity of completing the well as a deep-well subsurface
disposal system. Permeability exists in two main
horizons in the well. The upper zone is from 6620
to 6670 feet and  the lower zone is from 7036 to
7230 feet.  The upper zone is sandstone, pale red
to light gray, very fine to fine grained, hematitic
in part, calcareous, and quartzose. The  lower zone
is limestone, yellowish gray to very light brownish
gray, dense to crystalline dolomitic and sucrosic in
part. Permeability  and porosity exist in both hori-
zons by virtue of  natural fractures and cavernous
vugs.
     Electric  logs  and drilling characteristics indi-
cate a highly fractured zone in the Floyd Shale in
the  interval between 5070-5600 feet. A completion
in this  zone should be considered as an alternative
should  the lower two  zones cease to  function
properly.  This interval  is presently  cased  off and
cemented.  A  completion  in  this  interval would
necessitate a secondary  cement job and  perfora-
tions throughout the injection horizon.
            SPECIAL TECHNIQUES
     During the drilling of a disposal well, a geolo-
gist constantly examines  cuttings and maintains a
sample log. This information is used to determine
zones of importance in which open-hole drill stem
tests should be  made. The drill stem tests provide
formation  pressures, producing rates and  a sample
of the reservoir fluid. Electric log techniques are
used extensively to  evaluate the formations pene-
trated  and to  determine potential  injection  hori-
zons. The  data collected while drilling the wells is
utilized to determine  the casing and cement pro-
gram. Preliminary injection tests are conducted to
establish rates and pressures  that will be utilized in
the design of surface  equipment. A composite of
this  information is made for a graphical  presenta-
tion  and exhibits the results of drilling and testing a
well. The most advanced  logging techniques availa-
ble were utilized in Alabama.
     These included the dual induction lateral log,
the borehole compensated sonic log, the formation
density log,  sidewall neutron porosity log, radio-
active tracer log, and temperature log. Each of these
logs  is  also recorded  on  magnetic tape and  com-
posited by computer  in the form of a synergetic
log which  is a continuous record of all formations
penetrated showing porosity, permeability, water
saturation, and lithology. The cement bond  log is
useful in determining the amount ot cement fill-up,
and  the degree of bonding between the casing and
the  cement, and  the cement  and the formation.

                WELL MONITORING SYSTEMS
BSffi
                      PRESSURE
                           ]. Injection samples
                           2- injection tubing - icng string annuius


                      INJECTION RATE AND CUMULATIVE VOLUME
                      DRIP SAMPLES
                      TEMPERATURE
                      DENSITY
p
1

I
g

i
j
'
IS 1
— v. — '
5
^Tfff,
r5*
rFi
?
*3
•j
•i
i
i'

J
 Fig. 5. Well monitoring systems.


 Temperature logs and radioactive tracer logs are run
 to determine the zone of entry of the injected fluid.
 When the well is placed in operation, the well must
 be  monitored  by  continuous clock recorders, re-
 cording pressure at the injection tubing, the annuius
                                               )5

-------
between the injection tubing and long string and the
annulus between the long string and surface casing
(Figure 5). Injection rates and cumulative volumes
are recorded on continuous clock recorders.

          MONITORING TECHNIQUES
     Drip  samples  of the  wastes are taken  and
analyzed periodically. Temperature and density are
recorded on  a continuous basis  and corrosion  tags
are analyzed  periodically. Sample  stations are set
up on surface streams and  water-supply wells are
sampled periodically. Through an  agreement with
the Air  National Guard, aerial  photos (color  and
infrared) are made quarterly. An additional seismic
monitoring system has been added to the  Reich-
hold  well to  study any possible effects that injec-
tion  may  have on the tectonic activities of  that
area  (Figure  6).  The  system consists  of ten  geo-
phones placed around the well.  The geophones are
connected by a telephone service  to  a seismic re-
corder at the Survey office. Continuous films (16
mm)  are made of seismic  activities  in  the area.
These are transmitted daily to the  U. S. Geological
Survey Earthquake Center at Menlo Park, Califor-
               TKLEPHONE LINE
                   SEISMIC TELEMETRY BY FREQUENCY DIVISION-

                               MULTIPLEXING
 Fig. 6. Seismic monitoring system.
nia, for analysis. Two monitor wells will be drilled
between the two injection wells at Ciba-Geigy and
completed in  the  lowermost fresh-water  aquifer.
Samples from these wells will be analyzed peri-
odically for changes in chemical quality.

           SELECTED  REFERENCES
Bergstrom, R.  E. 1968. Feasibility of subsurface disposal of
     industrial waste in Illinois. Illinois Geol. Survey Circ.
     426, 18 pp.
Craft, B. C, and M. F. Hawkins, Jr. 1959. Applied petrole-
     um reservoir engineering. Prentice-Hall, Englewood
     Cliffs, New Jersey, 437 pp.
Hartman, C. D. 1966. Deep well disposal at Midwest Steel.
     Iron and Steel Engr. no. 12, pp. 118-121.
Healy, J. H., W. W. Rubey, D. T. Griggs, and C. B. Raleigh.
     1968. The Denver earthquakes. Science, v. 161, no.
     3848, pp. 1301-1310.
Henkel,  H. O. 1955. Deep well disposal of chemical wastes.
     Chem. Engr. Prog. v. 51, no. 12, pp. 551-554.
Louis Koenig Research. 1964. Ultimate  disposal of ad-
     vanced-treatment waste. U.  S. Dept. Health, Educ.
     and Welfare  Environmental Health Ser. 999-WP-10,
     146 pp.
Paradise, S. J. 1956. Disposal of fine  chemical  wastes. In
     Industrial Waste Conf., 10th, 1955 Proc. Purdue Univ.
     Ext. Ser. 89, pp. 49-60.
Semmes, D.  R. 1929. Oil and  gas in Alabama. Alabama
     Geol. Survey Spec. Rept. 15, 408  pp.
Sheldrick, M. G. 1969. Deep well disposal: are  safeguards
     being ignored? Chem. Engr. April 7, pp. 74-78.
Talbot, J. S.,  and P. Beardon. 1964. The deep well method
      of industrial  waste disposal. Chem. Engr. Prog. v. 60,
      no. 1, pp. 49-52.
Toulmin,  L. D. 1955a. Cenozoic geology of southeastern
      Alabama, Florida, and Georgia. Am. Assoc.  Petroleum
     Geologists Bull. v. 39, no. 2, pp. 207-235.
Toulmin, L. D. 1955b. Tertiary formations of west-central
      Alabama. In Guides to southeastern geology. Geol.
     Soc. America  Guidebook,  1955 Ann. Mtg., New
     Orleans, pp. 465-489.
Warner, D. L. 1965. Deep-well injection of liquid waste. U.
     S. Dept. Health, Educ. and  Welfare Environmental
     Health Ser. 999-WP-21, 55 pp.
Warner, D. L. 1967. Deep wells for industrial waste injection
      in the United States.  U. S. Dept. Health,  Educ. and
     Welfare Water  Pollution Control Research Ser. WP-
      20-10, 45 pp.
                                              DISCUSSION
 The following questions were answered by William
 E. Tucker after delivering his talk entitled "Sub-
 surface Disposal  of Liquid  Industrial Wastes in
 Alabama—A Current Status Report."
 Q.  We'd like  to  bear more discussion on whether
 conventional  methods  of waste  treatment  have
 been  evaluated and proven inadequate. Example:
 Brines can almost  always be disposed  of in more
 conventional  ways. Aren't  we really speaking of
 economics?
 A. This  particular brine is tremendously concen-
 trated and, of course, I think economics definitely
 fit into  any of these systems.  The  present con-
 ventional system,  in this particular plant, has been
 to store the brine  in a lagoon for evaporation. And,
 in an area where you have 53 inches of rain a year,
 it  doesn't evaporate  very much, and, of course, is
 inadequate. I'm not familiar with  all of the chemi-
 cal process  techniques  of separating brines—but
 most consume  a  tremendous quantity  of energy
 creating  other  forms of pollution and, of course,
 introducing the economic question again.
 16

-------
Q. Are you monitoring the effects of ingestion of
acid  wastes on  the carbonate  rock reservoir: and
how  do you determine the compatibility  of the
injection fluid with the reservoir fluid?
A. Well, this causes no end of  concern in the U. S.
Steel well which will be  a sulphuric acid injected
into  the carbonate reservoir. You can duplicate it
in the laboratory with samples of the waste. When
you  lower the pH, you start getting precipitation of
the sulphates. This is  a problem  that's  in  debate
throughout  the  country. In  subsurface disposal
wells, will  these  precipitants  harm  the  injection
well? There are some active sulphuric acid injection
wells that function perfectly well. But, when  you
see those precipitants  come out of solution in the
laboratory,  it really scares you. We think that one
advantage at the U. S. Steel well is the  fractured
system of porosity rather than intergranular poros-
ity.  Of course, the rock material will be in contact
with the fluid,  but  it will eventually  neutralize
itself at a greater distance from the well. An option
for that well, if precipitants clog the fractures, is to
change  the pickling process to a hydrochloric sys-
tem and dispose of the WPL in  the disposal well.
Q. How does  the Alabama Geological Survey de-
termine and/or define the expression, ''appropriate
geologic environment, "and also how do you deter-
mine an inadequate treatment method which would
then indicate that you could—or should—consider
deep well disposal?
A. Well, the  appropriate geologic environment is
determined by the  knowledge of the  subsurface
 gained  from data while drilling the  well, reservoir
 parameters, porosity, permeability,  compatibility,
 fractured gradients  in the area,  structural move-
 ments—all these factors that I've tried to  present in
the  talk, are used to determine the appropriate geo-
 logic environment. At the same time, the well has
 to  be  mechanically  constructed  to conduct  the
 waste to the disposal horizon and protect the fresh
 water resources.  Each well is judged by  its  own
 merits in an evaluation-type decision. We do make
 laboratory  compatibility  studies of the waste efflu-
 ent with the rock and with the formation water.
 Q.  Does the Alabama Geological Survey expect to
 bear the cost  of all  monitoring of waste disposal
 wells drilled in the future? And will detailed moni-
 toring be done on all waste disposal sites?
 A.  Yes, we will bear  the cost. This has been set up
 in our  operational budget, allocated by the State
 Legislature. There are many areas, of course, where
 matching funds and research funds are available to
 help supplement these efforts.  For instance, the
 seismic study  and monitoring on the Reichhold  is
 being financed by the USGS. The Environmental
Protection Agency paid 23.3% of the toral cost of
the well  and funds were budgeted for 5 years of
monitoring  the  well. Detailed monitoring will be
utilized on all waste disposal wells.
Q. Comment  on  the  need for  rehabilitation of
waste disposal wells—how often might you have to
do this and what methods might you use7
A. Typically, we think our monitoring procedures
will  turn up  times that  remedial  work  will be
necessary. The Stauffer well was  in operation for
about two  years before it required a remedial acid
job to clean up the perforations. This is typical for
a brine disposal  well in an  oil field or a producing
well in an oil field.
Q. Regarding  the  chemical reaction between acid
waste and carbonate rock,  how about the produc-
tion of CO2 and its effect on formation pressures?
A.  Since we  are  operating  at   3,000 or 4,000
pounds pressure, most of the CO:  will be in  solu-
tion of  formation fluids  and spent acid water.
Sometimes this reduces viscosity. CO:  is not gener-
ally  a problem unless the  pressure is such that it
comes out as solution.
Q,  Will  there  be publications describiiig  -;;:Lse dis-
posal holes, or disposal wells, or general informa-
tion on the disposal program.?
A;  Yes,  we have in preparation a State-wide report
on subsurface disposal which the Survey hopes to
be able to release this fall.
Q.  How do injection rates compare  between the
Stauffer well, which was  perforated—presumably
with a  jet perforation  system—and the gravel
packed Ciba-Geigy well? Do you see any problems
involved with jet perforation for any injection well,
especially in larger diameter wells in the 16 to 20
inch size?
A.  The  Ciba-Geigy  gravel-packed  well  presents
higher injection rates at lower pressures. Of course,
you get pressure drops, cross perforations because
cf  friction  loss. If the conditions are suitable. I
think perhaps the enlarged  hole, under-reamed prior
to  gravel packing, presents mechanically the  best
systems. In some cases, this is not practical, and
incompetent  formations will continually cave in
and slough  off into the well. The perforating tech-
nique doesn't lend itself very readily to, say, an acid
 injection.  You  would  have acid in contact with
the  metal causing undue corrosion unless you used
 some type of  exotic alloy  steel.
 Q. Have you  noted any decline in injection capaci-
 ty  on the  one  well that's been in operation for a
 couple of years?
 A.  It did decline up to the remedial acid job, and
 the acid job restored the well to its former capacity.
                                              17

-------
Q. For  wells in which you are disposing of acid,
and using a cement grout  to seal the  casing, this
might readily dissolve—that is, some cement grout
might readily dissolve if leaks develop. Why not use
inert,  or acid-resistant grout?
A. Correct. In  some cases, we use a resin-type of
cement—that is, acid-resistant. Most often it is best
to complete the well  in such  a  manner  that the
waste will never be in contact with the cement.
You may  have noticed  on the  gravel-packs and
open-hole completion that  the waste is going down
below the casing and cement and this prevents it
from coming in contact with the cement. Some of
the cement used is high density of about  15 or 16
pounds  per gallon and it has  a lot of nonreactive
material in the cement. But, there are resin cements
available,  plastics that  can  be used if the situation
develops.
Q. The  sense of this question  asks whether hydro-
chloric  acid is,  in fact, a waste that could not be
handled at the surface with conventional treatment
techniques and, if  that is   true,  then  why  is the
decision made to use a disposal well for handling of
bvdrocbhiric acid?
A. We  would  greatly  encourage  all industries to
recover  whatever products  they can off of a waste
stream  I  think this  is an area  that  will be the
eventual solution to most  of our waste problems.
We have one steel industry in Alabama that has a
hydrochloric acid pickling  system and they reclaim
all of the acid. This is, of course, the ideal solution—
to restore and  renew your waste. The U. S. Steel
well will  dispose of  125  gallons per minute of
sulphuric  acid,  which  is about 10 percent active.
This is  a tremendous amount of potentially useful
material, and while these wells are in operation, we
hope  to   continue  to encourage companies to
restore  their waste. But, I think it is just something
that will  happen down the road quite some time
off.
Q.  In  the case of a  State agency directed  well
disposal system for an  industry,  who  hears the
responsibility for possible contamination, the State
or the industry?
A.  I  imagine  that  would  get into a pretty  hairy
legal situation.  We were wearing the unusual hat of
a consultant or prime contractor on these two jobs
and I assume that the usual liability the contractor
has for  a completed project would exist.

Q.  What  do you hope to  see from air photo and
infrared methods with regard to monitoring and
injection  systems and, an  allied question, why do
you want to fly side looking radar for monitoring?
A.  South Alabama, the Mobile area, of course has
quite a few fault systems. We don't know for sure,
but we think these faults are sealed as evidenced by
the nature of the  sediments, but we found these
particular photographs very useful  in our part of
the country for picking out fracture traces, fault
and joint systems and lineation on the surface. We
have flown a  side looking radar  on hydrologic
problems and were able to locate fault systems. If a
subsurface disposal system  has any possibility of
leaking along fault planes we  would like to know
where. We feel that the infrared will show up the
areas of tree kill or vegetation destruction. Because
they are flown on a periodic basis we have  a con-
tinuous historical record.
Q.  Two questions related to economics: first, what
was the approximate cost to  the industrial clients of
deep disposal well planning and completion, and
the other question—what are some typical costs of
the disposal wells mentioned?
A.  In the  case of the U. S. Steel well, the Survey
received  $30,000 for its participation. The cost of
the well, including that $30,000 was approximately
$240,000. On the Reichhold well, the total cost of
the well was approximately $300,000 and the Sur-
vey  received $50,000 for their participation. These
figure out to approximately  $30 a foot. This would
compare closely with oil and gas drilling and com-
pletion in the  same area. Oil  and gas wells  in this
area usually run $10 to $15 a foot for a completed
well, but they  don't take quite the pains in  testing
and coring and sampling  and  logging. We used
practically every tool available to the oil industry
in our work. So, that's one factor that increases the
cost.
Q.  Another one related to  a former subject: have
you had any complaints from  consulting engineer-
ing firms about the Geological Survey of Alabama
acting  as a consultant  in  this way? Would some
type of consulting by the State  Survey be permis-
sible outside Alabama  and, lastly, did Alabama's
Attorney General render an  opinion on this?
A.  So far, we haven't  had  any  complaints.  I
noticed Jack Talbot was in the auditorium  and he
is in this  type of business, but  he hasn't com-
plained.  Actually we  have worked  closely with
consultants to  the mutual benefit of both parties. 1
don't think  we'd be interested  in  doing  outside
work,  except contributing  our  experience  to  the
state of the art.  In fact,  I doubt if it  would be
permissible for us to do otherwise. We did receive
an Attorney General's opinion on this subject. Our
legislative  act  enables the Survey  to  study all
phenomenon  associated with geology so we are
given a pretty broad area of operation. The philoso-
phy behind our being  involved  in  this is that the
 18

-------
wells  we were involved in were  considerably re-
moved from  an area of active oil  and gas explora-
tion. It was in an area where the nearest explora-
tory well was 75 miles away, and we felt that we
were more suited, knowing outcrops and knowing
the stratigraphy, than  perhaps  a consulting  firm
from  out  of State.  There are two more areas
within the  State  where  we  would like  to  drill
subsurface  disposal wells. Again they are areas that
we don't have much subsurface information about.
Very  likely  we will  be  involved  in those  two
projects, but I think it will end after that.

Q. Do you  have  a monitoring system  that will
identify leaks early in the stage by monitoring over
the injection zone, rather than waiting until a leak
might be serious  enough to  come  up to  hit  an
aquifer which you are monitoring?
A. None  of the wells  are presently set up to  do
that.  We  feel that we can use the pressures  and
injection rates,  pressure fall-off curves and things
like that to determine if there is a leak anywhere
and these are recorded on time  charts which are
analyzed continually and irregularities will  show
up in the pressure recordings.

Q. This, incidentally,  was  done by  a geologist
working in  Colorado and he  raises a good point.
They  have had to fuss with  this for  awhile. Has
there  been any noticeable seismic activity related
to any of your disposal wells?
A. No, there has not. I think I might have misled
you  on  the seismic monitoring of the Reichhold
well. That well is not in operation at this time. We
are recording historical data prior  to the actual
injection of waste into that well. This should occur
about  the first  of 1972. As  far as seismic monitor-
ing is  concerned, the Reichhold well will be our
main  area of study—the seismic activity caused by
the injection into the Reichhold well. We do not
have  seismic monitoring capabilities  around the
other wells in the State.
                                                                                                   19

-------
Subsurface  Storage  and  Disposal  in  Illinois
byH. F. Smith b
                   ABSTRACT
     The storage of both liquids and gases in underground
strata has become rather common in Illinois.
     The problem of disposal of fluid industrial wastes has
caused greatest concern, especially for the possible effects
on ground-water quality.  Necessary precautions have been
established in the requirements of the Illinois  Environ-
mental Protection Agency (IEPA) which has authority to
control,  prevent, and abate pollution of streams, lakes,
ponds, and other surface and underground waters in the
State. Before any construction can begin on storage in
subsurface  strata, a permit must be secured from the IEPA.
Eight basic design policies have been adopted that have to
be met before a construction permit will be issued.
     Sandstone,  limestone, and dolomite are the  basic
lithologies  most commonly considered as potential disposal
reservoirs in  Illinois. From well cores  of such formations,
essential laboratory studies are conducted defining the rate
of fluid movement, porosity, and permeability of the rock
and the pressure distribution within the aquifer.
     Plugging of the injection horizon is the most serious
cause of damage to a fluid injection system and results from
the  forming of an impermeable deposit on the well bore or
plugging in the formation itself. In either case the plugging
may result  from a number of listed causes.
     To date there have been four industrial waste disposal
wells in  Illinois.  Variation in conditions is illustrated by
these cases of disposal wells that have been authorized by
the  State.
     The  first practical use of underground gas storage in
Illinois was at Waterloo in 1950. Since then, the number of
projects and  their capacities have grown continuously. At
the  present  time there  are  24 underground gas storage
projects. Gas injection pressures must be kept below the
fracturing  pressure of the  caprock.  In underground gas
storage reservoirs in Illinois, injection pressures of approxi-
mately 0.55 psi per foot are often used.
     Secondary recovery by water flooding accounted for
73.4 percent of the total oil production in Illinois during
1968. In that year there were 880 active projects in Illinois
with 13,107 water injection wells that injected 2581 million
gallons of water.

                INTRODUCTION
     The  storage of both liquids and gases in sub-
surface  strata has  become rather common in  Illi-
nois. These  practices  include:  (1) storm water
runoff with storage for  only a few days,  (2) arti-
ficial recharge for  several  months, (3) natural gas
for a season, (4) brines for an indefinite period, and
(5) industrial wastes for permanent disposal.
     Presented at the National Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
     ''Head, Hydrology Section, Illinois State Water Sur-
vey, Box 232, Urbana, Illinois 61801.
     The problem of  disposal  of fluid  industrial
wastes has  caused greatest concern, especially for
the possible effects on ground-water quality. Neces-
sary  precautions  have  been  established  in  the
requirements of the Illinois Environmental Protec-
tion  Agency,  which  has authority to  control,
prevent,  and  abate  pollution  of streams, lakes,
ponds, and other surface and underground waters
in the State.
     As  the  State Legislature  and  the  Federal
Government, encouraged by conservation agencies,
tighten controls on  surface  disposal and  enforce
existing  legislation, the problem  of fluid waste
disposal  will  become  considerably greater in the
future.
     The first major industry that was faced  with
high  volumes  of  fluid waste  disposal  was  the
petroleum  industry, which produced large quanti-
ties of salt water with their crude oil.
     The  petroleum   industry's  solution  to  the
restriction  of surface disposal was to initiate injec-
tion of brines into suitable subsurface strata.
     Pollution  of streams and potable  water re-
sources  by industrial wastes including brines pre-
sents a great hazard. Every year there is an increase
in industrial water consumption, and a correspond-
ing rise in fluid waste that must be disposed of.
     Liquids, gases, and liquid wastes in reasonable
quantities probably can be safely injected into the
underground strata, if such operations are responsi-
bly  managed  with adherence to  rigid  standards.
Deep injection  waste   disposal is no more  than
storage in  underground space  of  which  little is
attainable in some areas and which is exhaustible in
most areas (Piper, 1969). These and other problems
are reflected in the following descriptions of the
policies in effect and the existing installations and
proposed projects in Illinois.

       SUBSURFACE WASTE  DISPOSAL
     Before any construction  can begin  on storage
in subsurface strata, a permit must be secured from
the Illinois Environmental Protection Agency. Be-
fore the issuance of a permit for a particular facili-
ty, the Board requires the submission of such plans,
specifications, and other information as it deems
relevant   and  necessary.  The  general  procedure
followed is that the interested industry  prepares a
comprehensive  feasibility report and submits  it to
20

-------
the Board's  Technical  Fecretary for review and
comment,  and  if satisfactory,  receives  tentative
approval subject to the submission  of final con-
struction documents for the proposed test wells
prior to drilling. The following  basic policies have
been adopted:
     (1) All  zones to be  considered for disposal
must contain brine waters having over 10,000 mg/1
total dissolved solids. This  equates to a  1 percent
solution of total dissolved solids. Fresh water  is
now defined as water of 10,000 mg/1 or less total
dissolved solids. A figure of 5000 mg/1 or less had
been used until recently to define fresh water, but
this was revised upward  to  preserve ground waters
that may be  usable  if  and when current experi-
mental methods of desalination are made practical
for the production of potable water  from brackish
waters.
     (2) There must be an effective and adequate
impermeable barrier overlying the disposal zone  to
prevent upward migration  of either the wastes  or
displaced brine waters into fresh water zones.
     (3) The well bore must be double cased and
the annular space grouted to a point 200  feet below
the lowest fresh water zone. The outer casing is to
be set into the next lower suitable rock barrier.
     (4) The industry must  indicate the character
 and volume of waters to be injected into the deep
 well.  Compatibility of the wastes with the  forma-
 tion fluids must be indicated.
      (5) Operational injection pressures to be used
 must  be indicated in the plan documents. Injection
 pressures  that  cause fracturing of the rock forma-
 tions  in the injection zone have not been permitted.
      (6)  Detailed information  about the proposed
 surface injection equipment  must be provided as a
 part of the plan documents.
      (7)  The  industry is required as a condition of
 the permit to submit daily injection report  records
 each  month  to the Illinois Environmental  Protec-
 tion Agency. These operating reports must  include
 the character and volume of the waste and show
 the injection pressures.
       (8)  Currently the Illinois Environmental Pro-
 tection Agency is reviewing  the  necessity   for
 requiring  the  installation  of observation wells to
 detect the escape of any  fluids from the disposal
 zone.
       The  disposal of industrial wastes in deep wells
 is considered relatively  new to  Illinois. It is realized
 that pollution of any underground potable or fresh
 waters will undoubtedly represent  long-term dam-
 age.  It is believed  and hoped  that  the  procedures
governing the installation and operation of deep
disposal wells are conservative. If there is error in
engineering  and  administrative  judgment, it is
hoped that  it will be toward the conservative side
of the scale so that the critical natural resource of
underground  water  is protected for  present and
future use.
     The primary desirable characteristic  of a sub-
surface disposal  reservoir is that it be deep enough
to  preclude any  possibility  of contamination of
potable ground-water supplies. Second, the disposal
reservoir  must  be deep  enough to preclude  any
adverse effect on  oil  or gas production in the area,
and third, it must be a suitable geologic formation.
To  be suitable, the  subsurface  strata must  have
sufficient porosity and permeability to accept fluid
wastes at reasonable  pressures and contain them
indefinitely. The  selected strata must  have a cap-
rock that is sufficiently  impermeable vertically to
prevent upward migration of injected  fluid. Sand-
stone, limestone, and   dolomite   are  the  basic
lithologies most commonly considered as potential
disposal reservoirs in  Illinois.
      The most desirable horizon  is  a sandstone,
 containing  salt  water (minimum of  10,000 ppm of
 dissolved solids), with high  porosity  and permea-
 bility. Also there are several limestone and dolomite
 reservoirs in  Illinois  that contain great volumes of
 salt water  and  can  be  considered  for storage or
 disposal  areas.  The  usefulness of these types of
 sedimentary formations  varies since their porosity
 and permeability depend mainly on fractured and
 developed solution channels.
      From well cores of such formations, essential
 laboratory  studies are conducted defining  the rate
 of  fluid movement,  porosity, and permeability of
 the rock and the pressure distribution within the
 saltwater aquifer.
      Plugging of the injection horizon is the most
 serious cause of damage to a fluid injection system
 and results from the forming  of an  impermeable
 deposit  on the  well bore or plugging  in the forma-
 tion itself. In either  case, the plugging may  result
 from a number of causes:
      (1)  Injection  of fluid containing  suspended
 solids which are often chemical precipitates.
      (2) Accumulation of corrosion products.
      (3) Precipitation within the  formation as  a
 result of incompatibility of the injected fluid with
 the connate water of the formation.
      (4) Injection  of insufficiently  treated  water
 resulting in bacteriological plugging.
       (5) Swelling of mineral  constituents in the
 formation.
                                                21

-------
Fig. 1. Waste disposal well locations.
     In  most  cases  the waste product, prior to
injection, will require treatment such as filtration to
remove any suspended solids that might deposit on
the face of the formation. It may require treatment
by additives to stabilize the solution or to prevent
excessive equipment corrosion.
     Laboratory  tests  on  core  samples  ot  the
formation will  give information on the maximum
particle  size  of the  suspended  solids that  can be
passed in the filtration process.
     The  first  deep  disposal well in Illinois  was
placed in operation in  November 1965. Since then
three other facilities have been placed in operation.
     To date there have been four industrial waste
disposal wells  in Illinois, one  in Clark County, two
in Douglas County,  and one in Putnam County
(Figure   1).  Variation  in conditions  in  Illinois is
illustrated by these cases of disposal wells that have
been authorized by  the  Sanitary  Water  Board
(Bergstrom, 1968).
     The  well  in Clark County  disposes 55 to 65
gallons per minute (gpm) of alkaline solution with
about 65,000 to 100,000  mg/1  dissolved minerals
into creviced Devonian age limestone at a depth of
about 2500 feet. Potable water occurs in the glacial
drift and the upper part of the Pennsylvanian rocks,
but below a  depth of a few hundred feet, miner-
alization  of  water increases rather  sharply  with
depth. Upward migration of alkaline waste water
would probably  be prevented  by  shales in the
Mississippian  and Pennsylvanian rocks.  The water
at 2500 feet had total dissolved minerals of  more
than 14,000 mg/1.
     A somewhat deeper disposal well now in  oper-
ation in Douglas Count}" was designed tor injection
of about 35.000 gallons  per day (gpd) of hydro-
chloric  acid   into  creviced  Potosi   Dolomite ot
Cambrian age at  a depth  of  about 5000  teet.
Originally designed for injection of wastes into the
St.  Peter Sandstone, the well  was  drilled deeper
when the permeability  of the St.  Peter Sandstone
was  found to be too low. Substantial  shale sections
occur above  the disposal  zone,  and  the  more per-
meable  formations contain  water  with  more than
5000 mg/1  dissolved  minerals  below  the lower
Pennsylvanian (Figure 2). The  water at  5000 feet
had  total dissolved solids of more than 20,000 mg/1.
A second waste  disposal well  having about the
same depth and construction features as the first
Douglas County well, Figure 3, has been drilled and
placed  in operation in the vicinity of the  first
Douglas County well.
      The well in Putnam County is for the disposal
of about 100,000 gpd of hydrochloric and chromic
acid from a steel mill. It is in the area where the St.
Peter  Sandstone  contains  usable water,  but  the
Ironton-Galesville and Mt. Simon Sandstones con-
tain brackish to saline water, injection of wastes  is
into a zone of the Mt. Simon Sandstone at a depth
of about 4000 feet. The low permeability of shah-
beds in the  Eau  Claire Formation  and the large
reservoir capacity  of the Mt. Simon  Sandstone will
prevent the  acid waste from moving upward into
the  zone of potable water.
      Conditions in the  northeastern part of Illinois
where potable ground water extends into the  upper
part of the Mt.  Simon  Sandstone  are illustrated by
a test hole recently drilled in DuPage County. The
well penetrated the entire Mt.  Simon  Sandstone
(2200  feet  thick)  and reached  the Precambrian
crystalline basement rocks  at  a  depth  of  about
4000 feet. Potable  water extended  to a depth of
about 2300  feet,  or 500 feet into the  Mt. Simon.
 Below  this depth the mineralization ot  the water
gradually increased to more than 90,000mg/'l at the
base of the Mt. Simon. The  1700 feet of Mt. Simon
 that contain nonpotable water consist mainly of
 sandstone, with only  a few thin  shaly zones. It is
 questionable whether there are impermeable shah-
 zones  in the lower Mt.  Simon  Sandstone with

-------
DEPTH
(feet)
         Velstcol Chem. Corp.
          NE SW  12-11N-I2W
            CLARK COUNTY
           Elev. &32f  K.B.
Wolcr
Quality
        \
          f t5OO~
          E 70OO~
                             Cabot Corp.
                            NW SE 3I-16N-8E
                             DOUGLAS COUNTY
                             Elev. 698'  K.B.
Jones £ Laughlin  Corp.
   SW SW 3-32N-2W
   PUTNAM COUNTY
   Elev.  527' K.B.
        \
              EXPLANATION
             WATER QUALITY

        22 Less than 1000 ppm TDM

        2 '°00 to 5000  ppm TDM

        ^ More than 5000 ppm TDM
      BOOO~ Water  quality in ppm  TDM
Am. Potash & Chem Corp. Test
      SE SE 9-39N-3E
      DU PAGE COUNTY
      Elev. lk\ ' K.B.
Fig. 2. Construction details and formations penetrated by disposal wells.

Fig. 3. Douglas County disposal well no. 2.
                                             sufficient  thickness  to prevent  any wastes  from
                                             moving upward  into the zone  of potable water.
                                             This test hole has been abandoned.
                                                  For all new disposal wells, after completion of
                                             drilling and  evaluation  is made of all  available
                                             subsurface information (sample descriptions, core
                                             descriptions, logs, porosity, and permeability anal-
                                             yses), injection  tests are made for a period of up to
                                             30 days. The injection tests are  made with filtered
                                             and treated water to preserve the injection charac-
                                             teristics of the formation. Data  from the injection
                                             test will be used to determine the pressure volume
                                             relationship  of  the  formation for future  injection.
                                             In the case of the Putnam County well, calculation
                                             of the rate of advance of the interface between the
                                             water and the pickle liquor showed that the radius
                                             of interface would be about 460 feet in  20 years.

                                                CHICAGO AREA DEEP TUNNEL PROJECT
                                                  The  early  sewer systems in the Chicago area
                                             were combined  sewers  intended to handle  both
                                             storm water  and raw sewage.  This practice has con-
                                             tinued, for the most part, until the present. The
                                             present combined sewer system  serves  300 square
                                             miles of heavily populated area.
                                                  In time of storm, the capacity of the inter-
                                             ceptor sewer and treatment system is too small to
                                             handle  both sewage  and storm  water. Therefore,
                                             during such  periods relief is obtained by  discharge
                                                                                            23

-------
of the mixture of storm water and raw domestic
and industrial sewage to the waterways. The over-
flows from the combined sewer system enter the
Illinois Waterway at some 400 locations (Sheaffer
and Zeizel, 1969).
     When the overflows are too  large for the
waterway system to accommodate, it is necessary,
on  rare  occasions, to discharge  this mixture  of
storm water and sewage to Lake Michigan. Such a
discharge  occurred on August 16, 1968. Although
such occasions have been rare (only four times in
the last 25 years), they are detrimental to the water
supply  and  recreational  activities  of  the area.
Associated with this on many occasions has been
flood damage along the waterway.
     An  economical solution to  the problem has
been  offered — it  is known  as  the deep  tunnel
project.  It  combines  the (1) City  of Chicago
underflow concept  with  the (2)  Metropolitan
Sanitary District deep tunnel project.

(1) Chicago Underflow Concept
     It is  proposed to construct a large  tunnel
under the rivers and  canals into  which all of the
combined sewers in Chicago would discharge. Thus
spillages   from the combined sewers  would not
pollute the surface streams but would be directed
to  the underground tunnel  referred to  as the
"underflow mainstream." The plan  calls for about
54 miles of mainstream tunnel, plus approximately
40 miles of branch tunnels. The mainstream tunnel
would function as a huge surge tank, taking the
overflow which the treatment plants cannot handle
during periods of rainfall.  After  the flow had
peaked,  the  underflow tunnel would  be pumped
out and sent to Chicago's sewage  treatment plants.
To connect the combined sewer to the mainstream,
tunnel shafts would be dropped down to it at each
outflow sewer's point of discharge.
     Chicago has started construction on the first
underflow sewer  which  runs from the  Chicago
River  westward paralleling  Lawrence  Avenue for
about 22,100 feet. The tunnel will have two sec-
tions (Pikarsky and Keifer, 1967). The first section
will be 9300 feet long and 17 feet in diameter, and
the second section will be 12,800 feet long and 12
feet in diameter. A branch tunnel 4000 feet long
and 12 feet  in diameter extending along  Harding
Avenue and south of Lawrence Avenue is proposed.
     Construction is being done by a tunneling
mole that is  used to bore through the Niagaran
Limestone stratum at a  depth of about 200 feet
below the ground surface.
     Ground water is to be protected by lining or
otherwise sealing the tunnel in any  sections where
there is leakage. The tunnel will be about  125 feet
24
below  the present water  level in the Niagaran
Limestone. There is a possibility that there will be
some leakage from the tunnel toward the aquifer
during  storm  periods that might  surcharge  the
system. This  maximum  period  of surcharge is
estimated  to be less  than  three  days and would
permit a small flow toward the aquifer.
    Monitoring wells, equipped  with water-level
measuring  devices  and provisions  for  collecting
water samples, are to be placed along the tunnel.
This monitoring system will indicate any change in
aquifer pressure or in water quality in the vicinity
of the tunnel that would indicate leakage.

(2) Metropolitan Sanitary District Deep Tunnel
    The  general  concept  of the  Metropolitan
Sanitary District deep tunnel system, as presently
proposed, combines the City of Chicago underflow
concept  with  an underground  storage  concept
(Sheaffer  and Zeizel, 1969). The tunnels designed
for flow under pressure,  to utilize  the large head
available,  will  conduct the overflow water to  a
central mined storage reservoir, some  830 feet
below  the land surface. The  mined storage reservoir,
made up of large unlined chambers in the Galena-
Platteville dolomite, will consist of two sections, a
settling chamber and the main storage reservoir.
     Sewage will first flow into a settling chamber,
which  will be large enough to contain runoff from
small and medium storms,  and  will retain much of
the solid loads of large storm runoffs. The partially
treated sewage will then flow into the main storage
reservoir,  from where it will be  pumped  through
reversible pump turbines  to a diked reservoir on
the surface using off-peak power.  Storm  water
stored in the surface reservoir will further improve
in quality by sedimentation and oxidation, and will
then be fully treated and released gradually  to the
waterways.  It  is estimated that  the deep tunnel
system will eliminate 99.5 percent of the pollution
load presently  reaching the  waterways  through
storm  overflows from the sewers. Both surface and
lower  reservoirs  will serve the dual function of
storm  water retention and storage for hydroelectric
generation.
     The  tunnels  and mined  storage area of the
deep tunnel system will be excavated in the Galena-
Platteville dolomite stratum that underlies the area
at  a depth of about  800  feet below the ground
surface.
     The  ground water in this aquifer is presently
somewhat above  the elevation of  the settling
chambers and the mined  storage area. However,
because  of heavy usage,  the natural  recharge  is
inadequate  to maintain the ground-water level in
the aquifer, which is falling generally at an average

-------
rate of about 13 feet per year throughout the area,
and soon will be below the storage area elevation.
    To adequately protect the water levels in the
aquifer they must be maintained above the tunnels
and mined storage area. This  can be done in one of
two ways, (1) by control of ground-water pumping,
or (2) by  artificial  recharge of the aquifer. It  is
believed more conservative at present to recharge
the aquifer.
    The feasibility  of aquifer recharge has been
based on  extensive  ground-water investigations,
including  an  electric analog  computer study and
field  drilling, seismic  surveys,  well  geophysical
logging, and aquifer pumping  tests.
    The tunnels and mined  storage area will not
be  allowed  to   become  pressurized   above the
ground-water pressure.
       ARTIFICIAL RECHARGE OF THE
     CAMBRIAN-ORDOVICIAN  AQUIFER
    WITH TREATED SEWAGE EFFLUENT
     Artificial  recharge  has  been  suggested as a
remedial  measure for some of the  water-supply
problems of the Chicago metropolitan area (Sheaffer
and Zeizel, 1966; McDonald and Sasman, 1967; and
Jones and Heil, 1967). Since the aquifer is deeply
buried it will be necessary to recharge the aquifer
through wells.
     The source  of water  considered  for artificial
recharge is sewage effluent (reclaimed water) from
tertiary treatment plants. Gulp (1969) states that
reclaimed water represents a  large potential source
of water.
     A major problem in well artificial recharge
operations is maintaining injection capacity.  Clog-
ging of injection  wells has been reported by many
investigators. Although the recharge water may be
classified as of drinking quality, further treatment
may be necessary or methods of well rehabilitation
may be  needed to  maintain  injection capacity. In
addition, the effectiveness of the filtering action
of consolidated sandstones needs to be established.
     The State Water Survey has planned a study to
reproduce the processes occurring in injection wells
under controlled  conditions in the laboratory. The
object of this research is to study  the feasibility of
recharging effluent from  a tertiary treatment plant
into the  deep sandstone aquifer  in  northeastern
Illinois,  and  to  investigate  methods of further
treatment of tertiary treated  sewage effluent so
that it  can  be used  successfully  to  recharge the
aquifer system.
     Upon completion of this initial research it is
proposed that a field  research project consisting of
a deep sandstone recharge well with four water
level and sampling  wells  be constructed in the
Chicago  metropolitan  area and operated  together
with the necessary treatment facilities.
       UNDERGROUND GAS STORAGE
     Space heating in homes and other buildings
consumes large  amounts of gas.  Because of the
seasonal  fluctuations in the demand  for gas for
space heating, the total gas demand generally varies
considerably from summer to winter.
     To make better use of the gas pipelines from
the gas fields to the users through  the year, the gas
distributors have acquired more heating customers
than  the  pipelines can  supply in the middle of
winter. Then,  any deficiency in gas supply is made
up by using gas stored underground during summer
months.
     The  first practical use of underground gas
storage  in Illinois was  by Mississippi River Fuel
Corporation at  Waterloo in 1950 (Bushbach and
Bond, 1967).  In 1952,  Natural Gas Pipeline and
Panhandle Eastern Pipeline Companies started their
large projects at  Herscher and Waverly, respectively.
Since then,  the number of projects and  their
capacities have grown continuously. At the present
time there are 24 underground gas  storage projects,
and  Illinois ranks fifth  in total reservoir capacity
among States that  have underground  gas storage.
Pennsylvania,  Michigan,  Ohio, and  West Virginia
each  have \l/2 to 2 times the capacity of Illinois
(Martinson, et al., 1966). Illinois has more potential
aquifer storage capacity than any other State.
     To store natural gas underground the follow-
ing conditions are  needed:  (1) rock  layers with
sufficient permeability and porosity to accept and
hold gas, (2) an impermeable caprock overlying the
storage rock to  prevent  upward migration of gas,
and (3) a geologic trap to keep the gas from moving
in a horizontal direction. This trap may be a dome
or closed anticline caused by gentle upward arching
of the strata,  a  stratigraphic trap caused by updip
gradation of the reservoir rock from sandstone to
shale, or a trap caused  by  faulting  that seals the
updip side of the  reservoir  by replacement of an
impermeable bed adjacent to the reservoir.
     The  porous storage rock in  a geologic  trap
under the  caprock  is  called a  reservoir. This
reservoir may  have been  filled originally with oil or
gas and thus may be a depleted oil or gas reservoir.
On the other hand, the reservoir may have been
filled originally with water;  in this case, it is called
a natural aquifer. The water in an  aquifer could be
fresh or salty. In Illinois fresh water aquifers are not
used  for gas storage because they  are too valuable
as sources of water for human consumption.
                                             25

-------
     When gas is injected into an aquifer, water is
displaced  from the pores of the  rock around the
injection  well.  The water  and rock  around the
storage  gas bubble is  compressed. In  a typical
storage  aquifer roughly half of the space for the
injected gas is created by the compression of the
solid rock matrix and half is created by the com-
pression of the water in the pores  of the rock. Near
the storage bubble the fluid pressure in the pores of
the rock is the same as it is within the bubble.
     The  higher the injection pressure, the greater
the rate at which the storage bubble is developed.
If gas is injected at too high a pressure, the caprock
may be fractured. Experience with hydraulic frac-
turing  of oil and gas producing  wells to increase
production  of these  wells  shows that sand-face
pressures  from about  0.7 to 1.0 psi  per foot of
depth  are required  to cause fracturing. In gas
storage  aquifers, injection  pressures of  approxi-
mately 0.55 psi per foot are often used.
     Besides pressure  per  foot of depth, the differ-
ence between the injection pressure and the initial
fluid pressure  in the aquifer  must also be con-
sidered. In the early stages  of the development of
the bubble, this difference is usually held at about
100 psi.  If experience shows that this causes no
leakage problems,  the  pressure difference is then
increased to 200 psi or more, but should not exceed
400 psi except in special  cases where extra control
precautions can be  taken.
     Pipeline companies and other companies sub-
ject to the regulations of the Federal Power Com-
mission  must  satisfy  the  requirements  of  that
Commission with respect to any  proposed under-
ground gas storage project. Public  utilities operating
in Illinois are subject to the "Public Utilities Act"
(Illinois Revised  Statutes, 1965, chapter 111 2/3).
Under Section 55 of this act, the Illinois Commerce
Commission  is directed to  issue  an order author-
izing a new facility (such as an underground gas
storage project)  after it  has found that the new
facility is necessary.
     The Natural Gas Storage Act (Illinois Revised
Statutes,  1965,  chapter  104, sections  104-112)
gives storage  companies  the right to use private
property  for  gas storage purposes in the manner
provided  for by the  law of eminent domain. Ac-
cording to the act, before the right of condemna-
tion can be exercised, the corporation must receive
an  order from the Illinois Commerce Commission
approving the proposed storage project. The Com-
merce  Commission cannot issue such  an  order
unless it finds that the proposed storage (1) will
be  confined to  strata  lying more than  500 feet
below the surface, (2) will not injure  any water
resources, and (3)  will involve no condemnation of
26
any interest in any geological stratum containing
oil, gas, or coal in commercial quantities within the
area of the proposed storage.
    In addition, gas storage companies that operate
in Illinois are  subject to certain rules  and regula-
tions  of  the  Illinois Department  of  Mines  and
Minerals and the Illinois Environmental Protection
Agency. Each  well that is drilled requires a permit
from  the  Department of Mines and Minerals. Fur-
thermore, the  storage company must  furnish evi-
dence  to the  Illinois Environmental Protection
Agency that the proposed storage project will not
result in the pollution of potable water.
     Observation  wells are required by the Illinois
Commerce  Commission  for measuring  pressures
within the gas bubble and for measuring changes in
water pressures in all directions from the bubble in
order to detect any possible gas leakage. Records of
these daily observations are  submitted  monthly to
the Commerce Commission and the Illinois State
Water Survey for study.

         WATERFLOOD OPERATIONS
     Waterflooding has become  an effective and
economical  means of stimulating  oil  production
and is essential to the oil industry of the Illinois
Basin.  Secondary recovery by  waterflooding  ac-
counted for 73.4 percent of the total oil produc-
tion in  Illinois during 1968 (Lawry,  1969). In 1968
there  were 880  active  projects in Illinois with
13,107   water  injection  wells  that  injected
518,5 81,000 barrels of water (2581 million gallons).
     Most of the water used in waterflooding is
brine that originates from the same formations that
produce oil. Thus, waterflooding presents very few
problems in contaminating the ground-water aqui-
fer as it replaces oil and  water that has been with-
drawn.
     Good  well  construction is controlled by  a
permit  and  inspection system administered  by  the
Illinois  Department of Mines and Minerals.

                   SUMMARY
     Great  caution must be exercised in the selec-
tion of formations to receive liquids and gases  and
liquid wastes for stotage underground.  Confining
formations  to prevent migration  of  wastes  and
stored material to the ground-water aquifer must be
assured before injection of liquids  into  permeable
and  porous formations  is  permitted.  Adequate
protection of the potable ground-water reservoirs
must be established and maintained at,all times.
     This is,done in Illinois through an adequate
permit  and inspection  system  requiring  a  high
standard  of facility design. These designs for  any
given project must satisfy the requirements of the
Federal Power Commission, the Illinois Department

-------
of Mines  and Minerals,  the Illinois Environmental
Protection Agency, or the Illinois Commerce Com-
mission, depending on the circumstances.
                 REFERENCES
Bergstrom, Robert E. 1968. Feasibility of subsurface dis-
     posal of industrial wastes in Illinois. Illinois Geological
     Survey Circular 426.
Bushbach, T. C., and D. C.  Bond. 1967. Underground stor-
     age of natural gas in Illinois, 1967. Illinois Geological
     Survey, Illinois Petroleum 86.
Gulp, Russell.  1969.  Uses  of reclaimed water. Water and
     Wastes Engineering, v. 6, no. 7.
Jones, David W., and Richard W. Heil. 1967. Beneficial uses
     for  northwest  area reclaimed water.  Metropolitan
     Sanitary District of Greater Chicago Report.
Lawry, T.  F. 1969. Petroleum industry  in Illinois 1968,
     Part  II, Water flood  operations. Illinois Geological
     Survey, Illinois Petroleum 92.
Martinson, E. V., et al.  1966. The underground storage of
     gas in the United  States and Canada, December 31,
     1965, 15th Annual Report on Statistics, Committee
     on Underground Storage, American Gas Association,
     Inc.
McDonald,  C. K.,  and  R. T. Sasman. 1967.  Artificial
     ground-water recharge investigation in northeastern
     Illinois. Ground Water, v. 5, no. 2, April.
Pikarsky,  Milton, and Cling  J. Keifer. 1967. Underflow
     sewers for Chicago. Civil Engineering—ASCE, May.
Piper, Arthur M. 1969. Disposal of liquid wastes by injection
     underground—neither myth  nor  millennium.  U. S.
     Geological Smvey Circular 631.
Sheaffer,1 J. R., and A. J.  Zeizel. 1966. The water resource in
     northeastern  Illinois: Planning its use. Northeastern
     Illinois Planning Commission Technical Report No. 4,
     June.
Sheaffer, J. R., and A. J.  Zeizel. 1969. The impact of the
     deep tunnel  plan on water resources of  northeast
     Illinois. Harza Engineering Company and Bauer Engi-
     neering, Inc., February.
                                             DISCUSSION
The following questions were answered  by H.  F.
Smith after delivering his talk entitled "Subsurface
Storage and Disposal in Illinois."
          ;   t
Q.  Who is going to pay for the  artificial recharge
program—the public, the beneficiaries, or  others?
A.  The artificial recharge program will be paid for
by  the public—in other words, the people who are
going to use  the water. In  this case the  public
would be the beneficiaries.
Q.  What is the storage capacity in days of storm
water runoff systems?
A.  The  storage design was  for 2 inches of runoff.
Q.  Assuming  that  fluids   break  through  to \ the
surface,  or to aquifers,  what do you do for an
encore—apologize, pump into the stream,  rapidly
build a treatment plant, change the permit form, or
what?
A.  This problem  was  encountered in  the early
beginnings at the Herscher Field not long after it
was put in operation. They had too few injection
wells so it was decided  to  increase  the  pressures,
and they probably got a fracturing pressure with
some  cement that wasn't as good as it could have
been. They  had  some  migration of gas  from the
lower zones up into an upper aquifer. In fact, two
or  three farmers had free gas for awhile which they
didn't appreciate. They were paid something for
their  inconvenience  and  I  think  everyone  was
satisfied with the final results of more wells, lower
injection pressures and  better well construction.
The same could apply  to  wastes, as long  as they
aren't too toxic. The wastes being disposed of at
present in Illinois are the kind that can be diluted
with water.
Q.  This one's a little allied.  Michigan  has had a
couple  of bad gas leaks in private wells. Has Illinois
had any  problems and how did they correct the
leaks? How  much damage did brine leakage do to
fresh-water aquifers, and was it corrected?
A.  Well,  I just mentioned a little about  the gas
leaks. We have had two shallow aquifers go bad by
brine intrusion. The  correction was made by mov-
ing five miles to another well field. I don't  think it
has actually  been decided who is going to pay the
cost of the five miles of pipeline.

Q.  You described four disposal systems in  Illinois.
Have you had  other applications in Illinois—bow
many, and how many of those have been accepted
to date?
A.  The  IEPA  has  requests  for  two  additional
permits for  disposal  wells. To date, permits have
not been issued for either well.

Q.  Regarding the DuPage  County Company that
was unwilling to give a definite analysis of their
effluents for underground  waste  disposal, who
permitted their discharge to surface water  in their
present system?  Is it now being  monitored, and
what  standards  must  they  comply  with—what
penalties must they face?
A.  The Company is now using an  oxidation pond,
and there are suspicions that it has leaked. They
have monitored wells as much as  2500 feet from
the pond.  Where contamination  was noted,  the
Company ran a pipeline from  the city supply to the
affected wells.
Q. Your printed summary noted plugging problems
related' to both the well bore and the formation.
Bill Tucker  noted a  similar problem and possible
                                                27

-------
solution  by remedial acidizing. Could you expand
on both the problem and possible solutions?
A.  I did not get into this because of time, but we
are having a discussion in Illinois  on what  is a
maximum  size particulate  that can be put  into
these aquifers. Some of  the  consultants want to
use  a  filter  that  will  filter  down  to about 15
microns. We  think perhaps a 5 micron filter will be
necessary. This is being substantiated by field data.
The effluent  in the well in Putnam County which is
from a steel  mill's waste,  is filtered through diato-
maceous earth filters.  Their particulate is reduced
to between 10 and  15  microns. I understand that
their injection pressures have continued to rise, and
they believe  that within the next six months they
are going to  have to rehabilitate the well. I am not
familiar  with how they intend to do it, but most
rehabilitation of the deep sandstone in our area is
done by some type of explosive such as primacord,
to  clean off  the well  bore. They are injecting a
pickling  acid which has a  pH of about IVa so they
have a real strong acid in the well. I am not familiar
with an  acid they could use that would help them
more.  They  may depend  on shooting  and then
cleaning out  for this particular rehabilitation.
Q.  How can an  injection well be justified under
Section 12 of the Illinois Environmental Protection
Act?
A.  The  IEPA tells  me that under  this particular
Section they are  not able, at this time, to prohibit
the construction  of  an injection well if it is proper-
ly constructed and the effluent is acceptable.
Q.  What is  the  rationale  of thinking  regarding
changing criteria  for salinity of injection horizons
from 5,000  at present to  20,000 milligrams per
liter in  the  future? Will  desalting perhaps  make
water usable  and needed in the future?
A.  Yes,  we think that desalting can make brackish
water  usable economically.  Five  years ago we
thought  that 5,000 ppm  of  salinity would be a
maximum  for conversion but we are  now of the
opinion  that with certain  types of new procedures
and processes, water with 10,000 ppm and perhaps
20,000 ppm of salinity can be demineralized in an
economical fashion. We  are raising our sights to
that figure.
Q.  What is your opinion of a bill recently intro-
duced in Washington  proposing that the federal
government regulate all subsurface injections?
A.  I believe I'll pass that for the time being.
Q.  Has the storage of gas in sandstone domes in
Illinois had any effect on local deep aquifer systems
regarding quality, pressure, flow, etc. ?
A.  None that we know of. We have had observation
wells from about 2,000 feet up to 20 miles from
these domes, and we are unable to see that the gas
pressure has  had any  great effect on the aquifer
system. In fact, it has had less effect than we really
thought it would have.

Q.  Regarding increasing pressure, how much dif-
ferential sheer will you permit between formation
and the casing?
A.  I don't have that answer right at hand.

Q.  Are periodic, complete chemical analyses, in-
cluding  base  metals,  required for  all injection
fluids?
A.  Yes. We originally  asked for daily samples, but
at  the present they are  collecting composite  sam-
ples over a period of two or three days,

Q.  What specific records are required from indus-
trial disposal well operators?
A.  The records  required are: an analysis of the
fluid injected, the rate, the injection  pressure, and
annulus pressure.

Q.  Have there been any failures or problems with
any of the storage or disposal systems, beyond the
ones you mentioned in your talk?
A.  I do not  know of any  failures that  have oc-
curred other  than those mentioned. The systems
have been  designed, I  think, rather well. The IEPA
has insisted and, of course, the former Department
of  Health who  formally  issued permits,  insisted
that the companies use good design  and follow  a
prescribed method of monitoring, and they  have
done real  well.  One or two gas wells went bad in
the early beginnings.  Since they have limited the
pressures on all  wells  so there  is no fracturing of
the formation or any  possible  disturbance of the
cement bond, the wells have worked real well.
 28

-------
Feasibility of Recharging  Treated  Sewage  Effluent
into  a  Deep Sandstone  Aquifer"
by Richard J. Schichtb

                 ABSTRACT
     Artificial  recharge with tertiary treated sewage efflu-
ent  has  been  suggested as one remedial  measure  for
projected ground-water deficits in the Chicago region, A
deep  sandstone aquifer, an important  source of ground
water in the region, offers the best opportunity for artificial
recharge. Recharge will be through wells since the aquifer
is deeply buried.  Expected problems in maintaining well
injection capacity were  studied  by recharging  treated
effluent through formation cores of  the sandstone. Some
success was had in maintaining recharge rates at constant
heads for several days.

               INTRODUCTION
     The Illinois State Water Survey is engaged in a
comprehensive  water-supply  study  of the metro-
politan Chicago region (northeastern Illinois) (Fig-
ure 1), an area  of about 4000 square miles with a
population of about 7.0 million.  It is expected that
results of this  study will  aid in determining the
State's strategy  for meeting the demands for water
for the region to the year 2020. The Water Survey
will  investigate  alternatives and possibilities from
all sources and the economics  of  each. It is ex-
pected that the study will include use of desalted
water, artificial  recharge, reuse, diversion of surface
water,  and  importation  of ground  water  from
outside of the region.
     The importance of ground water in the area is
emphasized by  the population dependent  upon
ground water as a source of water. Census figures
for the 6-county area are given below.
                            U. S. Census Count
                                 for 1970
6-County Region                 6,978,947
Cook County                     5,492,369
     Suburban Cook              2,125,412
     Chicago                     3,366,957
Du Page County                    491,882
Kane County                       251,005
Lake County                       382,638
McHenry County                   111,555
Will County                        249,498
    n
    Presented  at the  National Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
    "Engineer,  Illinois State Water Survey, Box 232,
Urbana, Illinois 61801.
Fig. 1. Study area.

It is estimated that about 25 percent of the region's
total  population—this  includes the population of
most of Du Page, Kane, McHenry, and Will Counties,
over one-half of Lake County and about one-fourth
of Suburban Cook County—are presently dependent
upon ground water as a source of water. Population
and percent of population in the region dependent
upon  ground water from  work by Schicht and
Moench (1971)  for 1980,  2000,  and  2020 are
given below.
                        Population and Percent
                            of Population
                           Dependent Upon
                            Ground Water
                                      Percent
                                        28.9
                                        40.1
                                        51.4
Year
1980
2000
2020
Projected
Population
8,258,390
10,867,805
14,641,950
Population
2,389,160
4,363,620
7,526,290
                                                                                            29

-------
     The  report  by  Schicht and Moench (1971)
stated  that  the total potable ground water in the
6-county  area was greater than the  total projected
demand  by the  year 2020. To meet demand in
some areas, however, importation of ground water
or  surface  water or  development of  alternative
sources locally may be necessary. Artificial recharge
is strongly being considered as a remedial measure
for the projected water-supply problems.
     It may  seem  strange  that Chicago and her
suburbs do  not have an unlimited supply of water
from Lake Michigan. But the  Lake  Michigan water
supply is limited  by the U. S. Supreme Court decree
of  1966 to 3200 cubic feet per second (cfs). The
Supreme  Court decree states: "The  State of Illinois
may make  application for  a  modification  of this
decree so as to permit the  diversion of additional
water  from Lake Michigan for domestic use when
and if it  appears that  the reasonable needs of the
Northeastern  Illinois Metropolitan Region (com-
prising Cook, Du Page, Kane,  Lake, McHenry, and
Will Counties) for water for such use cannot be met
from  the water  resources available to  the  region,
including both ground and surface water and the
water  authorized by act  of Congress and permitted
by this decree to be diverted from Lake Michigan,
and if it further  appears that all feasible means
reasonably  available to the  State of Illinois and its
municipalities, political  subdivisions, agencies and
              instrumentalities have been  employed to improve
              the water quality of the Sanitary and Ship Canal
              and to conserve and manage the water resources of
              the region and the use of water therein in accord-
              ance with the best modern scientific knowledge and
              engineering practice."

                   PREVIOUS ARTIFICIAL  RECHARGE
                             INVESTIGATIONS
                   Artificial  recharge into two aquifer systems,
              referred  to as the shallow  aquifers and the  deep
              sandstone  aquifer, have been considered  (gener-
              alized  aquifer descriptions  are given in Table  1).
              Recharge into the shallow aquifers (shallow sands
              and  gravels and shallow dolomites) has ,been dis-
              cussed in  the "Water  Resources in  Northeastern
              Illinois:  Planning  Its  Use," Northeastern  Illinois
              Metropolitan  Area Planning Commission Technical
              Report No. 4. The sources of water for artifkial
              recharge considered in the report were high flow in
              surface streams,  water that has been used for cool-
              ing,  certain industrial waste waters, and  treated
              domestic sewage. The pit method  was considered
              mainly because of the experience and success with
              the pit recharge operation at Peoria, Illinois (Suter
              and Harmeson, I960)., Areas where the pit method
              of artificial recharge  is potentially favorable were
              determined for the  planning commission report
              based on data from the Illinois State Geological and
                                   Table 1. Generalized Aquifer Descriptions
        Geologic Units
Thickness
  (feet)
          Water Yielding Properties.
 Glacial Drift1

 Silurian Dolomite1

 Maquoketa Shale

 Galena-Platteville Dolomite2
 Glenwood-St. Peter Sandstone2

 Prairie Du Chien, Trempealeau
 Dolomite, and Franconia
 Formations'
 Ironton-Galesville Sandstone2

 Eau Claire Shale

 Eau Claire and
 Mr. Simon Sandstones
   0-400+

   0-400+

   0-250

 150-350
  75-650

  45-790


 103-275

 235-450

2000±
Yields of wells variable, some well yields greater
than 1000 gpm.
Yields of wells variable, some well yields greater
than 1000 gpm.
Generally not water yielding, acts as barrier be-
tween shallow and deep aquifers.
Water yielding where not capped by shales.
Estimated transmissivity 15 percent that of Cam-
brian-Ordovician aquifer.
Estimated transmissivity 35 percent that of Cam-
brian-Ordovician aquifer.

Estimated transmissivity 50 percent that of Cam-
brian-Ordovician aquifer.
Generally not water yielding, acts as barrier be-
tween Ironton-Galesville and Mt. Simon.
Moderate amounts of water, permeability between
that of Glenwood-St. Peter and Ironton-Galesville,
water quality problem with depth.
 1 Glacial drift and shallow dolomite are referred to as the shallow aquifers.
 2 Collectively referred to as Cambrian-Ordovician or deep sandstone aquifer.
 30

-------
Water Surveys. These areas were limited to reaches
of major  streams in the region.  It was estimated
that a potential of 121 million gallons  per  day
(mgd) could be recharged. The aquifers outlined by
the Planning Commission as favorable for artificial
recharge  are in areas for the  most  part  where
ground-water deficits are not expected to occur
until after 2020.
     McDonald and Sasman  (1967) described a
study made for Park Forest-Chicago Heights  that
indicated  that artificial recharge to the Silurian
dolomite aquifer, the source of water supply in the
area, is feasible.
     They  also described  five   existing  artificial
recharge  operations utilizing recharge wells. At all
five sites water circulating through  closed cooling
systems is recharged. One installation withdraws
water from the  deep  sandstone aquifer  and re-
charges a shallower sandstone at  an average rate of
335,000  gallons  per day  (gpd). The other  four
installations withdraw  water from  and  recharge
water to  the  shallow  dolomite at average rates
ranging from 25,000 to  395,000 gpd.

      THE DEEP SANDSTONE AQUIFER
     The  heavily pumped Cambrian-Ordovician or
deep  sandstone aquifer, described by Suter et al.
(1959) is presently being considered  as offering the
best  opportunity  for  artificial  recharge in  the
Chicago region.
     The deep sandstone  aquifer consists in de-
scending order of  the Galena-Platteville dolomite,
Glenwood-St. Peter sandstone, and Prairie du Chien
Series of  Ordovician  age;  and  the Trempealeau
dolomite, Franconia Formation, and  Ironton-Gales-
ville  sandstone of Cambrian age. The  sequence,
structure, and general characteristics of these rocks
are shown in Figure 2.  The deep sandstone aquifer
is separated from  the Mt. Simon aquifer by shale
beds of the Eau Claire  Formation. The Maquoketa
Formation  above   the  Galena-Platteville dolomite
acts as a barrier between the shallow dolomite and
deeper  aquifers,  and  confines  the  water  in the
deeper aquifers under artesian pressure. Available
data  indicate that  on  a regional basis the entire
sequence  of strata, from  the top of the Galena-
Platteville  to the top of the shale beds of the Eau
Claire Formation,  behaves  hydraulically as  one
aquifer.
     The  Ironton-Galesville sandstone is the most
productive formation of the deep sandstone aquifer.
The  Galena-Platteville  dolomite and  Prairie du
Chien Series generally  are  not well creviced; the
Trempealeau dolomite is locally  well creviced. The
Glenwood-St. Peter sandstone and Franconia  For-
mation yield small to moderate amounts of water.
     The  deep sandstone  aquifer  receives water
from overlying glacial  deposits mostly in areas of
Kane,  McHenry,  Kendall,  Boone, and  DeKalb
Counties where the Galena-Platteville dolomite is
the uppermost bedrock formation below the glacial
deposits. This is west of the border of the Maquo-
keta Formation.  Recharge of the glacial deposits
occurs from precipitation that falls locally. Vertical
leakage of water through the Maquoketa Formation
into  the  deep  sandstone  aquifer is appreciable
under the  influence of large differentials  in head
between  shallow  deposits and the deep sandstone
aquifer.
     There are advantages  in recharging the deep
sandstone aquifer. The  aquifer is uniform in perme-
ability  and thickness  throughout  the  region, so
artificial  recharge would benefit  many  users. The
piezometric surface is well defined so that recharge
installations can be located to serve specific areas.
Continued withdrawals from the aquifer in excess
of the practical sustained yield, 46 mgd as estimated
by  Suter et al.  (1959), will  create considerable
storage volume. The practical sustained yield has
been exceeded each year since 1958. It is estimated
that water from  the deep  sandstone will be with-
drawn at a rate of 622  mgd by 2020 (the estimated
usable water in storage is large, estimated to be 1.6
X  1013  gallons).
     Since the deep sandstone aquifer  is deeply
buried  it will be necessary to recharge the aquifer
through wells.  It has been suggested that  effluent
that meets drinking water  quality  standards from
tertiary treatment  plants be used as a  source of
recharge water. According to Jones and Heil (1967)
one of the potential markets for the effluent is for
ground-water recharge. They report the interest of
Water  Commissions in northeastern  Illinois  in
utilizing effluent from  tertiary treatment plants for
artificial ground-water recharge.
     Because clogging  of injection wells has been
reported by many investigators, including  Rebhun
and Schwarz (1968), Bruington and Scares (1965),
and  Sniegocki  (1963), it was  decided to conduct
laboratory studies  to  acquaint  the Water Survey
with some  of  the  problems that may  arise from
recharging sewage effluent  into the deep sandstone
aquifer. The Water Survey recently completed  a
pilot project where sewage effluent from a second-
ary  treatment plant was  treated and  recharged
through  formation  cores  of  the deep  sandstone
aquifer. The objectives of the project were to study
what water treatment  is needed to maintain injec-
tion capacity,  to study methods of rehabilitation,
and to investigate methods of further treatment of
secondary or tertiary  treated  sewage effluent  so
that it  can be successfully used to recharge  the
                                              31

-------
aquifer system and meet water quality standards.
     The pilot project  was  described by Smith,
Schicht,  and Humphreys (1970) and is summarized
below.
     The new  Urbana-Champaign Sanitary District
Plant was  selected as  the initial laboratory  site.
Since the plant provides only secondary treatment
(activated sludge), additional treatment,  filtration
and chlorination,  was provided. It was later found
necessary to modify the  pH of the effluent.
     Three  types  of filters were used:  micro-
strainers, diatomaceous earth, and rapid sand. All
three filters were successful in  markedly filtering
out suspended solids when operating properly. Two
microstrainers  were  constructed,  one  with  26
micron cloth,  the other with 10 micron cloth. The
diatomaceous  earth filter was a type used for small
swimming pools. A chlorinator was installed ahead
of the filters.
     The  filtered and  chlorinated effluent  was
                                                          AQUIFER	~—~_^	I	_„	
                                                                                      "
            f -s-  -^r-^u^/I
                                                  Prairie d u C h i e n
                                                    Tre^peaIeau,
                                                    and f >~an con i a
Fig. 2. Cross sections of the structure and stratigraphy of the bedrock and piezometnc profiles of the deep sandstone aquifer
in the Chicago region.

-------
injected through a sandstone core used to simulate
a well in the deep sandstone aquifer. The Ironton-
Galesville  sandstone, since it is the most productive
formation  in the  deep  sandstone  aquifer, was
selected for  investigation. The cores, 3 inches in
diameter,  were provided  by the Northern  Illinois
Gas Company, which collected them during a site
investigation  for  underground  gas  storage. The
cores were  cut  into 3  inch lengths  and a 1 inch
diameter hole was drilled through their centers.
     The  cores were placed in  a specially designed
permeameter. Deionized  water  was injected  to
eliminate  plugging caused by  drilling and  cutting
and to determine head-discharge relationships. Hor-
izontal  permeabilities  determined by the Thiem
equation  indicated that permeabilities of the sand-
stone cores are comparable to permeabilities of the
sandstone determined  from aquifer  and pumping
tests.
     The  permeameter  is designed so that effluent
can  be injected into  the  center  so  that  flow is
outward  through the core similar to an injection
well. This  will  be  referred to in subsequent dis-
cussions as recharge. Water can also  be  injected so
 that flow is from the outside toward the center of
 the core simulating pumping and will be referred to
 as backwashing. The rate of recharge was controlled
 from  a constant head tank so  that initially  the
 recharge  rate into  an  actual  deep sandstone well
 would  be simulated. At first the normal procedure
 was to recharge for several hours and  then back-
 wash with  deionized water for a period  ot 15 to 60
 minutes. The purpose  of backwashing was to simu-
 late pumping which under  field conditions  is one
 method that can be used for well rehabilitation.
      Analyses of the filtered effluent were made to
 determine ammonium, nitrate,  chemical oxygen de-
 mand, biochemical oxygen demand, and bacterial
 counts. Ranges of the above are given below.
    NH4
NO,
                               COD
                          BOD
  0.0 to 5.5   19.8 to 57.3    16.8 to 83.3   0 to 8.0

       Bacteria plate counts  range from  0 to too
  numerous to count.
       Tests for chlorine residual were made several
  times daily.  A typical chlorine residual  would be
  7.8  mg/1 after filtration,  6.6 mg/1 before recharge,
  and 4.50 mg/1 after recharge.
       Recharge rates through a core with recharge
  water filtered by  a 10 micron microstrainer and
  through a core with  recharge  water filtered by a
  diatomaceous earth filter are shown in  Figures 3
  and 4. It is apparent from inspection of the graphs
  that little success was had in maintaining recharge
                                               A*
                                                                              s    —,
                                    Fig. 3. Discharge  rates  through core 6 with 10 micron
                                    microstrainer effluent.
                                           13     16
                                                             *'AKLH
                                     Fig. 4. Discharge rates through core 10 with diatomaceous
                                     earth effluent.
                                     rates. Some success was had, however, in rehabili-
                                     tating the cores by backwashing.
                                          Clogging was attributed  to material precipi-
                                     tating out of the recharge water after chlorination.
                                     Mineral analysis  of  the recharge  water at several
                                     sampling  points  indicated  the  precipitate was
                                     calcium phosphate and calcium silicate. The addi-
                                     tion of chlorine (Sodium Hypochlorate) as a disin-
                                     fectant was  raising the pH,  thus decreasing  the
                                     solubility of the effluent.  The  pH of the effluent
                                     before chlorination was 7.6. The pH after was 8.4.
                                     It was decided  to maintain a pH no  greater than
                                     7.0  by  adding  sufficient  quantities  of  diluted
                                     hydrochloric acid immediately after chlorination.
                                                                                   33

-------
     Discharge  rates through a core with filtered,
chlorinated, and pH modified effluent is shown in
Figure 5. During this trial it was possible to main-
tain constant discharge rates. Considerable decline
was noted,  however,  between  August  14 and  17
when recharge was discontinued over the weekend.
This decline may be attributed to biological activity
or a precipitate.
Fig. 5. Discharge rates through core 7 with rapid sand filter
effluent.
     It is planned to continue this pilot study using
effluent  from  a treatment plant in  the  Chicago
region. One objective of further study will be to
maintain constant discharge rates for longer periods
of time.
                  REFERENCES
Bruington, Arthur E. and Frederick D. Scares. 1965. Oper-
     ating a sea  water barrier project. Proceedings of the
     American Society  of Civil  Engineers, Journal of the
     Irrigation and Drainage Division, v. 91 (IR1), March.
Jones, David W. and Richard W. Heil. 1967. Beneficial uses
     for northwest  area  reclaimed  water. Metropolitan
     Sanitary District of Greater Chicago Report.
McDonald, C. K. and R. T. Sasman. 1967. Artificial ground-
     water recharge investigations in northeastern Illinois.
     Illinois State Water Survey Reprint No. 74.
Rebhun, M. and J. Schwarz. 1968. Clogging and contamina-
     tion  processes  in  recharge  wells.  Water  Resources
     Research, v. 4 (6), December.
Schicht, R. J. and Allen Moench. 1971. Projected ground-
     water deficiencies in northeastern Illinois, 1980-2020.
     Illinois State Water Survey Circular 101.
Smith,  H.  F., R.  J. Schicht, and H. W. Humphreys.  1970.
     Pilot scale investigations of well recharge using cored
     samples. Artificial Ground-water Recharge Conference
     at University of Reading, England, Paper 11.
Sniegocki,  R.  T.  1963. Problems  in artificial recharge
     through wells in the Grand Prairie Region, Arkansas.
     U. S. Geological Survey Water-Supply Paper 1615-F.
Surer,  Max  and  Robert H. Harmeson.  1960.  Artificial
     ground-water recharge at Peoria, Illinois. Illinois State
     Water Survey Bulletin 48.
Suter, Max, R. E.  Bergstrom, H. F. Smith, G. H. Emrich, W.
     C. Walton, and T. E. Larson. 1959. Preliminary report
     on ground-water resources  of the Chicago Region,
     Illinois. Illinois  State Water Survey and Geological
     Survey Cooperative Ground-Water Report 1.
The Water Resources  in Northeastern Illinois: Planning Its
     Use.  1966. Northeastern Illinois Planning Commission
     Technical Report No. 4. Chicago.
                                              DISCUSSION
The following questions were answered by Richard
J. Schicht after delivering his talk entitled "Feasi-
bility of Recharging Treated Sewage Effluent into a
Deep Sandstone Aquifer."
Q.  What are  the legal and moral ramifications of
diversion of streamflow by well pumping?
A.  In  Illinois  I'm  not  familiar  with  any  legal
ramifications  of  diversion of streamflow  by well
pumping. Most of  the  diversion of streamflow by
induced  infiltration, to  any appreciable  extent,
would be along our major streams and has not been
great enough  to reduce the streamflow any appre-
ciable amount.  Now the moral  ramification I'm
sure is a topic for another session.

Q.  Do  you have any cost, projections, say a cost
per  WOO gallons,  to  treat and then inject the
treated municipal sewage effluent?
A.  The only data I've seen on cost per 1000 gallons
to treat sewage is for secondary treatment. As far
as  I understand,  tertiary treatment of sewage
effluent is  still in the experimental stage,  and the
cost would not be realistic. However, we plan to
make  cost estimates of treating and  injecting at
some stage in the future, but I have no  information
on this.
Q. Have you yet attempted to scale your results to
actual field conditions? What is the expected injec-
tion rate per well?
A. We try to maintain a realistic injection rate that
would be comparable to actual field injection. You
can  compare  about  300  milliliters  per minute
through  the core with several  hundred gallons per
minute in an actual injection well.
Q. How do you propose to  backwash  the aquifer?
A. We're going  to  have two strings of pipe in the
well, one  for injection and one for pumping. The
pumping is intended to serve as backwashing.
Q. Do you know of anyone m the  United States
that is now recharging a confined aquifer?
A. Yes,  for example there is  injection of cooling
water  in the Chicago region into the deep sandstone
aquifer on a small scale.
34

-------
Q. Have you compared, or do you plan to compare
the direct use of polished effluent versus injecting
it into an aquifer?
A. Yes, we plan to; however,  there are probably
social reasons for injecting the sewage effluent into
the aquifer rather than direct  use. There's also a
very slow  time of  travel in  this aquifer. We're
talking about movement of less than a foot per day,
and this in itself would give a final polishing to the
effluent.
Q. Shouldn't the quality of water used to recharge
the aquifer system  be comparable or compatible
with the existing water quality in the aquifer? Will
the State require tertiary  treatment, such as nutri-
ent and heavy metal removal from the water to be
recharged?
A. I think the quality of water will be comparable,
at least within  certain limits, to  the  water  in the
aquifer. The effluent will have to meet drinking wa-
ter quality standards.
Q. How do you remove nutrients (phosphates and
nitrates) from your tertiary treatment plant?
A. We have  no plans at the present time for re-
 moving nitrates or phosphates. Nitrates were one of
 the chemical determinations made.  The nitrate
 content was usually  below 45  parts  per million.
 Analysis for phosphates were not made.
 Q. Do you foresee any problems with the forma-
 tion  of algae due to the  temperature of recharged
 water?
 A. We had problems with algae in the project and
 foresee problems during actual field injection.
 Q. 7s it feasible to  recharge storm water instead of
 treated sewage effluents? Are there any  studies
 being undertaken on the kind of treatment required
 for storm water tunoff to be used as recharge?
 A. Yes, it would be feasible. In relation to injecting
 into the deep sandstone  aquifer, we would have to
 consider the storm water runoff as sewage effluent
 and  treat it accordingly. We have  made  studies
 where the storm water runoff has been considered
 for  use  in  pit  or  trench  recharge in  surficial
 formations.
  Q. // common law governs withdrawals  of water,
how will the recharged stored water be kept for the
"recbarger" instead of being withdrawn by another
party?
A. The "users" would have to be the "rechargers."
In the Chicago region, in the deep sandstone aquifer
with a  well-defined piezometric surface, limiting
flow lines could be drawn from the point of injec-
tion,  or points  of injection, to determine  bene-
ficiaries of the recharge water and charge the "user"
based on the benefits from reduced pumping lifts.
It is not proposed that the State of Illinois be the
"recharger." We would  operate a demonstration
project and eventually a group of industries and/or
municipalities would undertake  the  operation  of
the recharge installations.

Q. Have you yet experienced  the formation  of
organic clogging mass at the liquid-solid interface,
as related  to aquifer injection now, similar  to the
masses that are characteristic in a surface spreading
operation for the field seepage of sewage effluents?
 The loss of infiltration  capacity  begins to show up
 in about 50-100 days as a  result  of these  layers.
How long have you run your system continuously?
 A. No, we didn't experience the build-up of these
 organic clogging mats.  I don't believe we actually
 ran the experiments  long enough.  In answer to the
 question  as to  how  long we ran the tests, this is
 what we  hope to do in subsequent studies—be able
 to recharge for  a longer  period of time. The longest
 that we ran continuously was five days.
 Q.  //  the  chloride ion is  not removed from  the
 sewage water,  will a long-term recharge program
 ultimately result in brackish formation water?
 A.  Well, I think you're talking about a period of
 time here. What I should point out is that since the
 aquifer will be  depleted at some time in the future
 you will  have  a storage reservoir underneath  the
 metropolitan  area  which   can  be  used to great
 advantage. Its use, by not  going into some sort of
 recharge program is greatly limited. We have only
  considered  the recharging  of  sewage  effluent in
  this particular  study. There is also a possibility of
  recharging Lake  Michigan  water directly during
  periods  when  the levels in the lake are critically
  high.
                                                                                                   35

-------
  Bull  Session  1—Ground-Water Waste  Disposal  Recharge  and  Reuse
Session  Chairman:  James  C. Warman,  Director,
  Water Resources Research Institute. Auburn Uni-
  versity, Auburn, Alabama 36830.

Leonardo Alvarez, Association of Engineering Geol-
ogists, Palo Alto, California:
     I'd like to bring up a year-old question as to
the direction of research that we should undertake
in developing some criteria for disposal of wastes
to wells. We notice that some  of the projects we
have  in Illinois haven't run into  any trouble and
they don't expect to, but there are  some ambigui-
ties here  and there which  bring up the question,
"Why is it so-and-so?" There is no answer. Why is
the pressure head  so-and-so, and there is not a
logical answer. I'm bringing up this question—this
topic—what do you think is the logical way to go
about setting up some sort of criteria? I'm sure
industry has done  this for a long time, but does
their  work  apply to what we are doing and, if it
does, should we modify it, and so on and so forth.
James C. Warman, Lead Bull:
     I'm  sure you  don't mean to say there is no
answer—you mean to say we don't yet understand
the answer.
John Rold, Colorado State Geol. Survey, Denver:
      I think that most of the answers we are look-
ing for about disposal wells are  in  the petroleum
engineers' heads and in the research reports that
petroleum  engineers  have read.  When we started
working in Colorado on our rules and regulations,
we found that the petroleum engineers with major
oil companies and  research organizations had  a
fantastic  amount of answers to questions that we,
as geologists, hadn't even begun to comprehend
when we started squirting something underground.
We think of this as a new field. In Illinois, they
talked about having four waste disposal sites and
yet  there are  13,000 underground brine disposal
sites  in the State. You have difference in viscosity
of fluids  and difference in compatibility, but they
have been fighting this problem  for 100 years. It
may  be  that there is research  yet  to  be done on
compatibility of fluids  and the  like,  and the  be-
havior of fluids in the ground. This research would
take  a lot of monitoring-I feel that we need more
 monitoring  than  we  are  getting.  Many of  these
 wells, of course, are  7,000 feet deep and  we're
 talking about a $40,000 or $50,000 hole just for
 monitoring. Then  you  have the  possibility  that
 your monitoring hole may be a leak in the injection
 aquifer. For most people concerned about injection
 wells, just going through the literature of the petro-
 36
leum engineering profession, or getting with a good
petroleum engineer, somebody like Hank Van Fulin,
will give them most of the answers that we have.
Jack Talbot, Brazos Oil & Gas, Houston, Texas:
     The preceding comment is very much in order
and, certainly,  Dr. Van  Fulin  and people of his
stature, have contributed a great deal to the basic
and fundamental information that has to do with
this  subject. But  there  are  still differences.  For
example, some of the injected fluids are corrosive
and difficult to clarify—much more so than oil field
salt waters. Peculiar, subtle metallurgical and  ma-
terials-engineering  problems are  involved, and some
additional and new investigations really need to be
made.  One of the  things that has surprised me  and
my colleagues is that there is so little information
as regards the tolerance of permeable formations to
suspended solids. We know we can pump things like
cottonseed hulls into fractured limestones, and yet
formations  exhibiting intergranular porosity  are
very sensitive to suspended matter. But the degree
of clarification required is not known. The question
came up yesterday—can you inject 10-micron parti-
cles or is it  only  0.3-micron particles? There  is a
great deal of work still to be done in this area.
John  Fryberger, Consulting Ground-Water Geolo-
gist, Engineering Enterprises, Norman, Oklahoma:
     The  prime  area  where we have worked in
Oklahoma is in salt water disposal wells, rather than
industrial  wastes. Our experience has shown  that
the oil companies are not applying the expertise
that I'm  sure they have, at least in salt water dis-
posal. I can cite, as an example, salt water disposal
wells  in the Glorieta Formation  where they had
perforated 10 feet of casing in formation consisting
of about 100 feet of fine-grain sandstone. Conse-
quently they were using pressures up to about 500
psi ground surface for  their salt water disposal in
a formation capable of taking the salt water with
no surface pressure applied. It's a matter of  well
design that made  them use the pressure.  I think we
 need  a uniform requirement that a  disposal  well
 application  should  include,  in the  test drilling
 procedure, a determination of the transmissivity of
 the injection zone to be used. Once you know the
 transmissivity and use the projected injection rates,
 you can  calculate the formation pressure build-up
 which may be  far  different  from  the required
 injection pressures because of skin losses and plug-
 ging of the injection well. If  required, injection
 pressures might exceed  what you would normally
 allow for formation pressure, but would still be a
 safe  formation pressure to use. If you know the

-------
formation transmissivity, you can calculate what
the formation pressure will be with time. This is a
safeguard  that I have not  seen  in  my  study of
State regulations required now for injection wells.
I'd like to see this adopted.
Speaker not identified:
     Would  these regulations be within the oil and
gas industry? Or, within some other agency?

John Fryberger:
     1 would  think  that  the regulatory  agency
charged with the responsibility of industrial waste
wells should have this requirement. The regulatory
agency  responsible for salt water  disposal wells,
quite often is a separate agency.
Speaker not identified:
     Do  you think that there should  be a  dual
responsibility?
John Fryberger:
     No, but quite  often, that's the  way it is with
the State agencies.

John Rold:
      On the subject of regulations, Jay Lehr asked
me  to talk  about Colorado's problems.  We're the
most recent State to come up with rules and regula-
tions for disposal wells.  I didn't get  around to
putting together a paper for the formal program, so
he asked me  to discuss this tonight. Colorado had
the unhappy experience  of having a deep waste
disposal  well  which  became quite famous,  the
Rocky  Mountain Arsenal Well,  without any rules
and  regulations. Then the State passed a law on
water pollution and the Water Pollution  Control
Commission went through about a year of hearings
on an application  for a radioactive waste disposal
well in northeast Colorado. They still didn't  have
any rules and regulations or guidelines. The appli-
cant didn't know what was expected of him and
the Water  Pollution Control Commission didn't
know what was expected of them, so they turned
down the application after about a year's hearings.
      The Commission then put together  a Com-
mittee to come up with a set of rules and regula-
tions. There  were  about 25  people on  the Com-
mittee  — geologists,  lawyers, sanitary  engineers,
petroleum  engineers, people from Dowell   and
Halburton  Service Companies and many  of our
State agencies. We had a Geological Committee, a
Reservoir Committee, a Sanitation or Health Com-
mittee and a Legal Committee. Everybody sat down
and first  of all wrote to all the States that had any
experience with waste disposal wells and any laws—
they picked all their brains. Then we made a basic
policy decision before we started—that we couldn't
go with hard and fast rules such as Illinois has used
about  5,000 ppm and  10,000 ppm. One of the
reasons is that some of our  wastes  in  Colorado
actually are good enough that they wouldn't hurt
water of 2,000 ppm. We've been in the situation of
having to inject oil field fluids that are  1000 ppm
down a hole rather than let them run into a river
of 1500 ppm.
     We  came up with  a series  of factors which
should be considered  individually for the situation
at hand.  There are 25 or 30 factors that have to be
looked at—for example, the  transmissivity of the
reservoir and coefficient of storage. We realize that
it may cost the company wanting a  disposal well
probably  several  hundred thousand dollars to get
all of the data together that would be needed for
an application. So, we have a preliminary applica-
tion situation whereby the applicant can get  to-
gether all the data he has, or can put together at a
reasonable cost, and bring the proposition into the
Water  Pollution Control Commission and see if it
has any chance of flying at all. So he has some idea
as to whether or not they're going  to  let him go
ahead with the proposal.
     The application has to  include a legal descrip-
tion and a map of the area with  a two-mile radius
of the proposed  well, showing all the surface and
mineral owners.  It has  to show  all test holes and
penetrations,  description and depth of  all the
formations penetrated by any well or mine on the
map, local topography, industry, fish and wild life,
description  of the mineral resources believed to be
present in the area, and the probable effect of the
system on the mineral resources. They have to have
a geologic  structure  map  showing stratigraphic
sections,  plus  a geologic  cross  section  drawn
through  the area and the proposed well, a descrip-
tion of all water resources, surface and subsurface,
within the  zone  of influence  or possible influence
of the system, a classification of those waters, and
a map showing vertical and lateral limits of surface
and subsurface supplies. There must be a descrip-
tion  of  the chemical, physical, radiological and
biological properties and characteristics  of waste to
be disposed of  in the  system, and treatment pro-
posed for such  waste including copies of all plans
and specifications of the system and its appurte-
nances. A statement  of all sources relied upon for
information must be set forth in  the application. If
the  system is to be an injection well, he has to have
potentiometric surface  maps of the disposal forma-
tions and those formations immediately above and
below. Copies of all drill stem tests, installations
 and data used in making the maps must be includ-
 ed.  We ran into  that on a previous application be-
 cause there were a lot  of numbers on the map that
                                              37

-------
couldn't be traced to basic data when we were
trying to check them. You must state the location
and nature of present and potential  use of fluids
from  the disposal  or affected formation and the
volume, rate and injection pressure of the fluid to
be injected. Then include the following geological
and physical characteristics of the injection interval
and the overlying and underlying permeable barrier:
thickness,  areal  extent,  lithology,  that's  grade.
mineralogy, type mineralogy, matrix, amount and
type of cement, clay content and clay mineralogy.
You must  include  the effective porosity and how
determined, permeability—vertical and horizontal—
and how determined, mechanical logs, and  forma-
tion tests. And you must furnish the coefficient of
storage  of  the aquifer, the amount and extent of
natural  fracturing,  location,  extent and effects of
known or suspected faulting, extent and effects of
natural  solution  channels, the fluid saturation in
the formation, the  formation fluid chemistry with
indicated local and regional variation, temperature
of the formation and how determined, formation
and fluid pressures, fracturing ingredients, osmotic
characteristics of the rocks and fluids comprising
and contiguous  to  the reservoir and an indication
of the effect of the injected wastes on the contigu-
ous formation in the event of leakage, diffusion and
dispersion characteristics of wastes in the formation
fluid  and  the  effect of  gravity segregation, the
compatibility  of  injected wastes  with physical,
chemical and biological characteristics of the reser-
voir. And, of course, you must have the engineering
data on the well which is normal and  as detailed
for the geology.
    Now,  you can see why we give them a chance
for a  preliminary go to see if their project applies.
We didn't want any disposal  wells unless we knew
they were  safe.  Since these  regulations went into
effect on July 1, 1971, we have had an application
and it was granted for a waste  disposal well by
Shell  Oil Company, over  in  the Peonce Basin for
an oil shale experiment. The regulations are feasi-
ble. Shell said it was  a real pain in the neck and a
lot  of a problem but, by the time they made out
their  application,  I think they knew more  about
the formation and their project than they did when
they started.

Jack Talbot:
    Do the oil  field salt water disposers have to
comply with these same requirements?

John Rold:
    No. In Colorado the disposal of oil field brine
is handled  by the Oil and Gas Commission. They
don't  have to use these regulations; however, the
Water Pollution Control Commission has essentially
38
said, "Yes, the Oil and Gas Commission has author-
ity over that, but if they goof up once, we'll take it
over." So,  it's putting a little bit more surveillance
on the brine disposal.

Unidentified Speaker:
     It is  fortunate for the prospective injectors
that  Colorado  is not a large industrial State, be-
cause the  information that's requested  here, if it
must be supplied, is simply not available.
John Rold:
     We are aware of that and aware of the proba-
bility that to get some of this data you'd actually
have to drill one or more test holes  down through
the disposal aquifer. We weren't looking to become
a haven for subsurface waste disposal wells. A lot of
people said we'd never get one,  but we've  already
got  one  after  the  regulations  came into effect.
Incidentally,  the applicant for the unsuccessful
operation that was turned down earlier and had so
much trouble was on the Committee and helped
write these rules.

Zane Spiegel, Ground-Water Hydrologist, Santa Fe,
New Mexico:
     I just wanted to ask if this  well you speak of
was  in an  area with existing knowledge from  oil
exploration.
John Rold:
     There was an  amount  of existing regional in-
formation  in  that  there were two or three holes
that had been drilled within a few hundred feet, as
well  as the hole proposed  for injection. So, they
did have  the required data.
James C. Warman:
     Would anyone else like to comment on these
regulations, their complexity or adequacy, or would
you like to make other comments about regulations
at the moment?
John Fryberger:
     I'd like to ask a question of everyone here. Do
you  think  it would  be reasonable to require waste
disposal well applicants to furnish the transmissiv-
ity .  . . well,  now, let me back up ... in his appli-
cation to furnish all  of the background data that
you  were talking about—the existing wells in the
area, the geological structure of the horizon that he
would intend to inject into, information  that he
could obtain  without actually drilling a hole down
there,  and have  his application be  either turned
down on the basis of that information or tentative-
ly accepted on the  basis of that information. If it
was tentatively accepted then he  would be required
to construct a well—if he wanted  to do it this way—
go right ahead and construct his  disposal well as he
would intend to use it,  and  perform tests to deter-

-------
mine the transmissivity of the formation. Based on
these tests, the regulatory agency could then deter-
mine that rate of injection at which the potentio-
metric surface  in the injection zone would not be
above the water table or potentiometric surface in
the lowest fresh-water aquifer. And, at that rate of
injection, even if there would be a leak, there'd be
no way for the injected fluid to get into the fresh-
water aquifer. So it would be, in a sense, a fail-safe
type of operation. Do you follow me?

James C. Warman:
     Are there any comments on that? That  sounds
like a practical approach to me.

Jack Talbot:
     Well, it  is a  practical  approach,  and  this  is
what almost all regulatory agencies do, regardless
of whether the discharge is to a surface stream or to
the subsurface, or whatever method is used. Any
scheme  that's used to  control, or to  alleviate a
pollution problem is an interim thing. The  per-
mission to  use it  is an interim thing.  I know of
only one State that deviates from the policy-that
requires a permit  to drill a well for this purpose—
 and that's Alabama. Now you have to have there a
 permit to drill an  exploratory oil or gas well. You
 have to have a permit to drill a salt water disposal
 well, but for  some reason in Alabama you can go
 out with some understanding with the State people
 to make a penetration and get the data. Now they
 may, or may not, let you use the well, and this
 policy is the same as that followed by other States.
 This has the  same policy  as  Louisiana. They can
 withdraw the permit at any time to discharge into
 the  subsurface.  As regards your limitations on
 pressure, there can be  some real problems  because
 of the nature of the fresh  ground waters, the pres-
 sures under which they exist, and the pressures re-
 quired to inject the effluent to the deep subsurface,
 these have to be  examined on an individual basis
 and considered on an individual basis.

  H. F.  Smith, Illinois State Water Survey,  Urbana,
  Illinois:
      In Illinois, they have to have  a permit to drill
  an exploratory hole for waste disposal. They need
  to furnish the information as best they know simi-
  lar to what was outlined here in Colorado—maybe
  not so detailed. But then they  get a permit to drill
  the exploratory  hole  and  that  is tested for its
  transmissivity and compatibility of the effluents
  and so forth. That was the case of the pilot hole
  that was drilled in a northern Illinois county. From
  all indications, the pilot hole looked like  it would
  make a well. Following drilling, permeability tests,
  and  running  cores,  they did  try injection and
couldn't get any injection—the well was abandoned.
But, for all intents and purposes, it looked like a
well until it got to that point. They had a permit to
do it. One of the reasons I think they have a permit
is  because  you have to  specify the fluids you're
going to inject. You can't deviate from that fluid
without going back and having  the  permit cor-
rected. The regulatory agency turns down a lot of
applications because they just won't permit certain
types of fluid to be injected. The permit to drill the
well itself is a more or less temporary affair, but it's
a permit. If they have demonstrated that it's a good
well, then  they do  get a permit for operation. But,
until that  time, it's just a permit to  test and the
permit is required to even start off with one hole.

James C. Warman:
     As John Fryberger pointed  out, most of the
regulatory agencies are approaching these requests
for permits to construct a disposal  well on the
same basis that they consider a request to dispose
of a liquid effluent to a surface stream. The request
for a permit to dispose an effluent to  a stream
 carries  with  it  a  description—some  characteriza-
 tion-of the fluid,  the quantities,  and also the kind
 of  treatment  that you  are giving  this  material
 before you  are  turning  it loose, or  proposing to
 turn it loose. These same kinds of things are  con-
 sidered by the regulatory agencies on this disposal
 proposition. You're not just going in to request to
 punch  an exploratory hole, but you're really going
 in with a reasonably complete design package of
 the kind of hole you're going to  drill, how  you're
 going to  construct it, the kind of tests that you're
 going to  run  on  this well that will provide the data
 that will  give assurance to the regulatory  agency
 that you  have considered these several parameters
 that we need be  concerned with to satisfy ourselves
 that we're  not  going to  violate the  fresh-water
 aquifers. I don't know of any State that isn't really
 looking toward  this approach. Some of  them are
 doing it more or less thoroughly than others and I
 suspect most States that are getting into this will
 move toward tightening up these requirements for
 even this preliminary or feasibility approach, rather
 than making it  easier as we go  along. Now,  if we
 have explored this facet to your satisfaction-if no
 one objects-I'd like to  revert back to the original
 question, which was, "What kinds of research might
 we need to get  into to provide some of the answers
  that we do not now have?"

  Zane Spiegel:
       I'd  like to  make  one comment  about Mr.
  Fryberger's  question or suggestion  in relation to
  the relative heads in the  injection zone, or proposed
  injection zone, and the heads  of overlying fresh-
                                               39

-------
water aquifers. I think we have two general circum-
stances that we  might encounter in this problem.
One is a normal circumstance in which the deep
aquifers, or salt water zones, already have a higher
head  than  the  overlying fresh zone. And this is
actually  the  normal circumstance. In  much of
Oklahoma, Texas and southeastern New Mexico,
oil and gas production has been going on since the
1920's or 1930's and early  1940's.  The pressures
in these lower zones—the salt water bearing zones-
have been lowered to the extent that they are much
lower than  the  overlying  zones.  Therefore,  you
could adhere to  his proposal very easily, and inject
significant  quantities of water  under substantial
new heads—differential heads—without exceeding
these fresh-water heads. But I think in virgin areas
that  have  not had their potentials lowered by
previous production of either oil or salt water, it
might not be possible to do what you suggest.

James C. Warman:
     What we're saying again is that each case is an
individual case and must be considered only on that
basis. Now, let me try again to revert back to the
first  question  which  got  such  short  treatment.
Where are the areas of knowledge in which we are
quite lacking? Where are the research areas? Where
do we need some help?

John Rold:
     I'd  like to throw one  out that's really  been
bugging us, and that's the seismic effects of injec-
tion. We  have one in our  backyard, here, which
has been blamed for several thousand earthquakes
and the people  that  know more  about seismic
activity than  I—some  of the real  experts  in this
area—don't believe it. Most people feel it probably
is so, probably  because the proponents of the
theory have gotten better coverage than the people
that are  against it. You've probably even seen in
such journals as  True Magazine about the seismic
effects of the arsenal well. We, however, also have
an  injection—which is salt water this time, just an
ordinary salt water injection pressure maintenance
over  in an oil field—and they are monitoring a lot
of earthquakes in that oil field. They're all  small,
Richter magnitudes of 1,2, and every two or three
years you  get one in that area  that you can feel.
But,  of course,  everybody's watching and waiting
for  them  now.  This possible  seismic  effect  of
injection wells is one  avenue that really ought to
be studied in considerable detail and very thorough-
ly,  or we'll get slapped.
      You may have heard that the GSA Meeting in
Milwaukee had  an excellent paper on the Seismic
Effect of  Dams,  i.e., the  loading of  a dam with
water causing seismic effects. The speaker had five
40
charts and the statistical correlation was beautiful
to show  that they  built the dam and  then the
earthquakes started.  When he was about done he
said, "Oh, by the way, that number four, they
never built the dam." So, if you start monitoring
any area with a good enough seismograph, you're
going to pick up a lot of earthquakes. If I were an
oil company, I'd be monitoring any area—especially
an area  of population—the day after I got the idea
that I might start injecting fluid into a reservoir.

Frank Kresse, W. A. Wahler & Assoc., Palo Alto,
California:
     The  U.  S.  Geological Survey  is  trying to
trigger  earthquakes  with the idea of controlling
major earthquakes. Of course, this is  a very exotic
thing, but they're really talking about this seriously.
The idea  is to construct  an injection well along a
fault  that has a  potential  for major earthquakes
and, by injecting  fluids,  cause small earthquakes
that are undamaging to the point that the strain is
taken off the fault. This is wrought with all kinds
of problems as we  can all imagine.  That's really
the kind  of thing that they're imagining with this
experiment and it  should provoke quite a bit of
discussion and thought, anyhow.  I'd like to regress
just a little bit. Most of you are  familiar with the
Baldwin Reservoir failure in the Los Angeles area
in  1963.  The failure of this dam was caused by a
fault offset under the reservoir and under the dam,
and the classic explanation for this was that it was
caused  by subsidence from oil withdrawal in adja-
cent  oil fields.  There was a $30  million lawsuit
which was settled out of court. After the settlement
some of  the  things that  probably  occurred came
out and  there's  been  a  couple  of  papers—Doug
Hamilton's and others on this  particular thing.
They've given a pretty good case for the idea that
this offset was caused not by the subsidence, but
by injection  of  water flooding  of this  same oil
field, and that the offset was really a jacking effect
of the fault itself which caused the failure. They've
got a lot of time and space correlation and so on,
and I feel that this surely was the thing that hap-
pened here.
     So this  is an example of what can happen,
maybe  an unusual one, but one of the other things
that can happen from injection wells. I think there
are at least three or four cases that have enough
documentation  where  there  can be little doubt
that this  has happened. When I was involved in the
 Baldwin  Reservoir  lawsuit, I went through  the
literature and tried to find all of the cases where
 there was at  least  some documentation and I feel
 that there are at least 12 cases, world-wide, where
 there is at least some evidence of increasing earth-

-------
quake  activity  due  to the  construction  of  a
reservoir.
Jack Talbot:
     I wonder if an initial bit of research couldn't
be simply to  find out reasonably how many types
of injection wells there are  in the  country, and
where they are. No one seems to know this. How
many wells inject salt water into formations other
than that of origin?  What have been the seismic
effects of this? And what have been the effects of
withdrawals,  not  as  regards  subsidence, but  as
regards earth tremors? I think someone could do a
good study of  the literature and the records that
we have  in this country and come up with some
useful  information.   It may allay  some of  the
apprehensions that go with injection of fluids and
building dams and the like.

James C. Warman:
     One  of  the things in which  we can stand a
little research is in the  pure economics  of what it
really is going  to cost to go through the kinds of
testing that the  regulatory agencies are going to
have to require—the construction and the testing of
the well itself, and the operation and monitoring of
this  well. Then we get into  the philosophical bit
about who's  going to pay for the  monitoring. I'm
not  sure where that  cost needs to fall.  We've had
some research  going  on in Alabama by Dr. David
M.  Grubbs  at  the   University  of Alabama.  Dr.
Grubbs has  been working with well logs from Oil
and  Gas  Board records and other existing informa-
tion. He  sought  to determine the most probable
areas, especially in the Coastal Plain part of Ala-
bama, where you might likely be successful in deep
well disposal.  He used available data to put to-
gether ideas on how much liquid of what kind you
could  inject  in these  zones in these  various
feasible sites that he  has identified. And then he's
tied  this thing  into your need as an industry  with a
given effluent.  If you  will characterize the effluent,
the quantity  and type of material you want to get
rid of,  he will take  your disposal  requirements,
consider the  geographic, hydrologic, and geologic
setting in which  you want to operate, and run all
this  through  the computer. The computer will—on
the  machine—drill a  well and attempt to operate
that well to meet your requirements.  If it can't
accept your waste in  the quantities that you want
to get rid of and the  pressures that you're going to
have to handle, the computer will drill another well
and  carry cost data with this.
     This approach  needs  refinement, but  it's
interesting and  promising. If the computer indicates
that you  have  to drill five wells,  you look  at the
economics of this and say, "Yes,  I can handle that
kind of expense in the construction and operation
costs to dispose of my waste that way." Or you
may say, "No, this is out of the ball park and we
have to find a better way to treat this waste at the
surface and get rid of it with conventional or new-
technology."
     Other kinds of work going on at the Universi-
ty of Alabama and elsewhere are concerned with
compatibility of some of the wastes that we know
people are going to try to dispose of through deep
wells—the compatibility of this  waste  with  the
liquid in the reservoir and with the reservoir rock
itself. What other  kinds of problems are we going
to  have  to solve  before we can move on with
safety?
Richard  J.  Schicht, Illinois  State Water Survey,
Urbana, Illinois:
     Has anyone here had any experience using a
buffer fluid injected before the waste effluent?
Jack Talbot:
     Yes, I've been involved in this and it is one of
the less  exact sciences.  Sometimes  I  think the
reason this is done is to make everyone feel better.
I'm not really  sure what  purpose injection  of the
buffer fluid, usually fresh water, serves. But it  is
convenient,  besides making everyone feel better, in
that when you start injecting  fluid into a well it's
prudent, normally, to test the surface equipment
and it's nice to do this with  fresh water,  so you
pump a calculated volume. We use a formula that
Don Warner developed—or his colleagues  did—at
the Taft Sanitary Engineering Center, which, so far,
has worked like a charm. But I'm not really sure
that the buffer fluids are  necessary. I'd like to say
something  about  compatibility.  A great deal of
research can  be done in this area, but I'm not sure
that this is one of the problems of injecting strange
fluids into a formation. We've had the experience
of taking a core—a limestone core—of intergranular
porosity and  injecting dilute sulfuric  acid into  it
and plugging it immediately. Yet, in a real life
situation in the field, it works fine. Now these are
very dilute acid sulfates,  and I can't suggest what
the result would  be if you  injected  a more con-
centrated solution. The compatibility is something
that one should be concerned about; but the fact
that core lab work says that you're going  to plug
the formation  need not stop you from making an
attempt.
James C. Warman:
     Any more than when core  lab work says  it
won't  plug the formation is this a guarantee that
injection will be a success. Laboratory results do
offer some  mighty good cautions. If you were
considering an effluent that immediately plugged a
                                             41

-------
porous limestone core that came from your host
rock, you'd at least better have your eyes open.

Jack Talbot:
    Exactly. I'm not denying that  caution should
be used in this area. I'm just saying that if a person
or a company has a difficult injection problem and
there  appears to  be  no other way  to  solve it, he
should not drill  the  well  based  solely upon the
laboratory data. Incidentally, the tendency around
the country is becoming more and more to  insist
that the applicant examine all reasonable alterna-
tive means—sometimes unreasonable alternatives—
to  the solution of his pollution  problem.  This is
something I, personally, agree with.  The subsurface
is limited in its volume, its capacity, and its ability
to  accept  these  fluids,  and  things that  can be
treated  on the surface really  ought to be  treated
there.  But, at the same time, there  are streams
generated by manufacturing plants that have so far
defied human ingenuity  in  solving  the  surface
treatment.  And there are those who say there will
never be a solution; they may be overly pessimistic.
The chemists tell me that the chloride effluent is
one very tough problem.

Zane  Spiegel:
     I have comments on three  points and I'll go
backwards. One is that I agree  completely with that
last statement about the alternatives,  because one
of  the alternatives I  think that is  being proposed
more and more is a recycling  type  of treatment in
which things  are recovered. Many cases have been
mentioned in the last two days where an industry
has been forced to  do something like that and
afterwards they've discovered that recovery was
really the most profitable thing  to do in the first
place. I guess we haven't looked  at  these things
carefully  enough and  the future  of  recycling is
going to become more  and more  attractive. The
second  point is that I think an obvious explanation
for the difference between laboratory results and
occasional  better results from actual tests on forma-
tions is that there may be denser  porosity in the
core  because the core  obviously  is a piece that
hasn't been fractured. The third point is to go back
to  the kind of basic research that is needed. 1 don't
know whether you'd want to call  it  research or not.
It's just a collection of the kind  of basic data that
we need, the actual  formation potentials that exist
and sometimes can be obtained from existing files
of data,  particularly  in  the oil  industry,  and
development of better methods of collecting these
data. I  think the oil industry  has been working on
that for the  last few years. They were rather lax in
doing that in the earlier years. Much  of the data
that we do have is either of no value  now because
conditions have changed  or  it was never of any
value because the data was just no good to start
with. One of the basic things we need is develop-
ment and application of methods of collecting good
data on existing wells.

John Fryberger:
     I'd like to expand on what Mr. Spiegel was
just  saying.  You're referring to the problem of
gathering information on existing wells in the area
where the proposed injection well is to be drilled,
and  especially those wells for .which there is no
record. That's a tough one. In some work that I've
done recently, I've tried to gather data on problems
that have been created due to salt water disposal or
water flood operations in contamination of ground
water or surface water supply.
     One example that really sticks in my mind is
where they had an injection  well about 3,000 feet
deep for water flood and  about a quarter of a mile
away there was a hole about  500 feet deep.  I don't
even know why the hole  was drilled. Some time
after injection was started, salt water poured from
this 500 feet deep hole and it  obviously was coming
from this 3,000 feet deep  injection well. It  flowed
out at the rate of several hundred barrels  a day,
flowed into a stream and contaminated the stream
for  15 or 20 miles before it was finally observed.
That  illustrates  how water might move  under-
ground from one zone up to other zones,  maybe
through  several wells,  interconnecting  different
zones.
      This, to me, illustrates the futility in a  way of
having observation wells—monitoring wells—in the
vicinity of an injection well. I really don't think
the  information that these safeguards obtain would
be worth  the  expense,  because water can move  a
considerable distance away from the injection well
before it  comes up through a fracture zone  and
never be detected in any  monitoring system. So, I
haven't been able  to envision in my own  mind  a
monitoring system that would really do the job at
a reasonable cost.

Jack Talbot:
      If you're talking about  intergranular porosity
rocks,  I  just can't agree with that because in an
injection system  situation your greatest pressure
difference between the natural head and the injec-
tion formation and the injection pressure exists at
the well bore. For what you're talking about, some-
thing could happen where fractured porosity exists.
 But we worry more about the well itself in an inter-
granular porosity situation,  which is what  we like
 to stay with,  than we  do about what is going to
 happen out away from the well bore a half mile or
 so,  or  even less, because at that distance where the
 42

-------
porosity is intergranular and possesses some reason-
able degree of homogeneity, the difference between
your natural pressure and your injection profile is
going to  be minimal. In the case you described, I
would  suspect that  you've  just recited  a bad
cement job.
John Fryberger:
     This was  in  SE Kansas where the  water was
coming up the 500 feet deep well from a 3,000 feet
deep injection  well—I didn't  get  much more in-
formation other than the specific depth. I got the
information from the Kansas Health Department.
     You would propose then putting monitoring
wells, say within ten feet—or some very close prox-
imity to the injection well—is that right?

Jack Talbot:
     I probably wouldn't want to inject in an area
like southeastern Kansas in the first place. If you're
talking about somewhere else where the rocks pos-
sess reasonable permeability and porosity and the
geology is fairly well known—yes, I would want the
monitor well near the injection well. Or, perhaps
the injection  well itself  could  be engineered  to
serve as a monitor well.

John Fryberger:
     Yes, I like that idea, but if you had a separate
monitoring well you would probably want to put
it  to the bottom of the lowest  fresh water zone.
However,  it might be likely  that there would  be
other permeable zones below the lowest fresh water
that could take the injection  fluid that might  be
coming up alongside the casing because of a poor
cement job, and the fluid would move laterally and
might  find  some other avenue  up. There are so
many geologic conditions that,  in my opinion,
make monitoring not too feasible. In some cases it
would  be. As  a general thing, I don't think I'd be
for it.

John Rold:
     One of the things I  think we're going to be
faced with in  the near future is the highly toxic
fluids—the  real problem  fluids.  As soon as  the
Corps of Engineers puts the clamps down on waste
going into navigable streams or their tributaries we
are going to have a lot of people thinking, "Well,
maybe we ought to put it down a hole." Many of
these fluids are not too bad. There's no real prob-
lem with some of the fluids and it could be fairly
economical, in many instances, to put in an injec-
tion well and get rid of these rather innocuous or
fairly safe fluids that way. Of course, once you say
underground waste disposal well  in most circles,
you've got a fantastic emotional problem. Every-
body, including the Sierra Club, is very down on
underground waste disposal wells.
     But I think that  we should start thinking in
terms of our being  a lot better off to put a lot of
these fluids in a salt water formation at 3,000 feet
than running them down the river. There's an awful
lot of that going on. We're just now beginning to
find out how much liquid is being surreptitiously
or otherwise run into  a storm sewer or a sanitary
sewer or into a creek. I think that we're going to
be facing a lot of small-time innocuous fluids that
could well be put underground. There again, each
case should be looked at on its own merit.

Jack Woodard, Florida Dept. of Natural Resources,
Tallahassee, Florida:
     We've  already  hit  the spots you were just
referring to  about  getting some  of the rather
innocuous materials out of  surface streams,  espe-
cially in the southeastern section of  Florida. In
Florida  half of our deep injection  wells are dis-
posing of treated sewage effluent.  We're  afraid we
might want  to salvage some  of this stuff at a later
date. Also, we don't have much room to build  dams
for storage reservoirs in south Florida.  We're  look-
ing at disposal wells  seriously  with the idea of
injecting storm water and recovering it at a later
date. There is a little bit in the  literature about
this, but I would like to hear, particularly,  from
any  of  the State  regulatory  agencies who  have
criteria on what it  would take to inject with the
idea of recovering   this particular water.  And I
might add  that we  are putting storm water into a
salt  water  formation.  Everything from  200 feet
down to a 3000 feet injection horizon is a poorer
quality  than  what   we're putting into the  3000
horizon, so I am not too worried about  leakage.
Richard J. Schicht:
     I don't know if you're familiar with the USGS
project in Norfolk, Virginia. I think they're doing
something similar.

Jack Talbot:
     I'd like to ask  Mr.  Woodard  if there are any
plans to  do  some experimentation for testing to
determine or  estimate what the quality of  fresh
water will be  if it's injected into  a  saline aquifer
and  then  recovered. It seems to  me  that this is
something that needs to  be known before there's a
great deal of money and effort spent  in this sort
of thing.
Jack Woodard:
     Israel  didn't have very good  luck with  their
project  and recovery rates. It  has proven to be
economically feasible to dispose of storm water and
                                             43

-------
particularly  sewage  effluents into deep storage
wells, but both of these procedures are fairly new.
Right now the entire amount of the monitoring
program  in Florida is handled by the private indus-
try which injects  the waste. I'm trying to get some
program  set up for research in this area, but I was
looking particularly  for anybody who might  be
doing that. I'm going to pick their brain.

Zane Spiegel:
     I know of a couple of cases where water was
injected into sandstone aquifers and then recovered
at a fairly reasonable rate of recovery—something
near 80% as I recall.

Jack Woodard:
     Yes. Everything I have found has been  in the
sandstone aquifers and we don't have any.
James C. Warman:
     All  right. Anybody want to shift gears?
John Farbiter:
     I would like to ask a question of people who
might know.  What  fluid is normally used in the
annulus  between the long string and the injection
tubing? The ideal fluid would be noncorrosive and
nonelectrolyte. I've  heard that diesel oil can be
used.
Richard J. Schicht:
     That  is  what  they are using in Illinois. I
haven't heard any complaints about it up to now
so it seems to  be working all right.
John Farbiter:
     That's the only one that I've heard of and I
was just wondering if there have been others.
Unidentified:
     Most commonly, fresh water is used. In my
experience you can add an oxygen scavenger if you
want to  be especially  particular about it, but nor-
mally  this  isn't  done. Mud is  normally avoided
because  of the problems  of fishing the injection
tube out after some period of operation.

Dick Brown, U. S. Geol. Survey, Lubbock, Texas:
     I was waiting to hear someone start on injec-
tion of surface water into aquifers, especially aqui-
fers as a recharge operation. We're involved  in the
research  in the high plains on artificial recharge. 1
thought  that  when  this time came I would com-
ment  on your question  as to what size particles
you could  get into an  aquifer system.  We don't
really know this, of course, but we have done a
little work which we  thought was interesting.
     We have sand columns in the laboratory and
these  consist  of fine  and medium sand. We have
injected through these columns at a range of from
20 to 50 psi at a constant rate. We get a reduction
in permeability  of  70% on  some of these. We've
taken cores out of these columns and sent them to
Sperry Rand and had scanning electron micrographs
made of the materials taken from the columns after
they are plugged to this extent. What has surprised
us greatly  is  that  in  the  scanning micrographs,
sediments look as if they're  an  open boulder field,
and the surface deposit is kind of a layer of sea-
weed over the sand grains and you don't even see it
until you get to 10,000 magnification, and yet this
is  enough to cut permeability  70%. We  have not
done  much with different kinds  of clays yet. In
preliminary  experiments  we  ran a  mixture of
kaolinite and  water through millipor filters; then
we ran pure water and then a mixture  of mont-
morillonite and water and we found that we got a
couple of orders of magnitude  more water through
in a given amount of time with the kaolinite than
we did with  the montmorillonite. In fact we  got
more water through with kaolinite than water by
itself. This was 500 mg.  per liter of kaolinite, and
then we ran 500 mg. per liter  of kaolinite through
first and ran the montmorillonite through later and
got  more montmorillonite through than we had
through the filter without any kaolinite on it.
     So it makes a big difference as to what you
run through. In much of the literature on artificial
recharge you see the sediment content expressed as
a  Jackson Turbidity Unit and  here again the JTU
readings for a kaolinite-water mixture are  totally
different than for a montmorillonite-water mixture.
There is a magnitude of difference in JTU units for
the same concentrations, so there is a great deal of
work to be done.

James C. Warman:
     Having solved "all"  the problems of deep well
disposal, I think we may have run down for  this
evening. Thank you all for coming in and partici-
pating in the bull session. I  think it's been a lot of
fun to sit dov/n together and go at it this way. We
learned a lot by listening to  the formal speakers
and having a chance to write questions to them, but
this bull session is another kind of dialogue that is
also quite helpful. We probably need to do more of
this,  and less of that,  with  all  deference. More
auspicious groups, for example  the  Gordon  Re-
search Conferences, have found that this type of
dialogue can  be highly productive. We hope that
this bull session and the  other three underway this
evening will have turned out some thoughts that we
can  all  pick  up  from  the  Proceedings of  this
Symposium.  Let's get on  with  the business of
solving problems. Thank you.
 44

-------
  Pesticide Contamination  of a  Shallow  Bored  Well
  in the  Southeastern  Coastal Plains"
  by M. J. Lewallenb

                   ABSTRACT
       A comprehensive examination was made of a shallow
  farm well which was contaminated with persistent pesticides
  when contaminated soil was used as backfill material
  around the well casing. The well location was less than 25
  feet from a site previously used for flushing an insecticide
  sprayer.
       Pesticide level in  the water has been monitored for
  more  than 4  years, during which a gradual decline in
  concentration has occurred. Soil core samples taken in the
  area surrounding the well indicate relatively high surface
  contamination but very little downward movement. Sedi-
  ment  samples from the bottom  of  the well  exhibited
  highest concentration of all samples.

                 INTRODUCTION
      Pesticide contamination of water supplies is a
  matter  of  concern to all people  associated with
  water quality. Because of this interest, many in-
  vestigations have been  conducted concerning  the
  occurrence, cause, and solution of problems associ-
  ated with  pesticide contamination. However, most
  research has  been directed  toward surface water
  and municipal supplies.
      Large amounts of pesticides are handled and
  possibly even stored in close  proximity to many
  individual  water supplies. Quite often these materi-
  als are in their most concentrated form when they
  are closest to the water supply. Such situations can
 occur during filling and flushing of sprayers. Re-
 gardless of  these  seemingly  great hazards,  few
 documented cases exist of pesticide contamination
 of individual water supplies. This report documents
 the occurrence and describes the conditions which
 allowed toxaphene, DDT [1,1, l-trichloro-2, 2-bis-
 (p-chlorophenyl)  ethane]  and its  metabolites  to
 enter  a  private water supply in the southeastern
 Coastal Plains.
 Well Construction
     The well was constructed  in late 1966 and is
 typical of many  installations  in  the  Southeast,
 where  rather  small water yields are required and
      Contribution from the Livestock Engineering and
Farm Structures Branch, Agricultural Engineering Research
Division, Agricultural Research Service, USDA. Presented at
the National Ground Water Quality Symposium, Denver,
Colorado, August 25-27, 1971.
     bAgricultural Engineer,  U. S. Department of Agri-
culture, Southern Piedmont Conservation Research Center,
Watkinsville, Georgia 30677.
       sprayer wash area—^
                                       HOUSE
                       I    '
             ,fence  x_|
0 10 20 30
scale, feet
 Fig. 1. Layout of the farm yard.
 geologic formations permit the boring of a large
 diameter hole.  Typically, the well is  30 inches in
 diameter and cased with 24 inches inside-diameter
 concrete pipes which are 48 inches long. The pipes
 are fitted  together by  a tongue-and-groove type
 joint without sealing joints in the casing. Depth of
 the well is 40  feet, and the water table averages
 approximately 18 feet below the surface.
      Contrary to established recommendations  (U.
 S. Department of Health, Education, and Welfare,
 1962), soil was backfilled around the well casing
 without using either concrete or clay grout. Typi-
 cally, after a few days the inadequately compacted
 backfill settled, producing  a maze of  channels
 through  the  backfill material. The owner then
 removed soil from  an adjoining area which had
 been used  to fill and flush the farm sprayer and
 placed the contaminated soil around the well casing.
 Layout  of  the  farm yard is  shown in  Figure  1.
 Within a few days pesticide contamination was
 detectable organoleptically in the water.

         EXPERIMENTAL PROCEDURE
 Water Samples
     Collection of water samples began in January
 1967 and has continued periodically since that
 time. Typically, 4-liter samples were collected  at
 the  tap adjacent  to  the pump. Samples were
 returned immediately  to the  laboratory  and ex-
 tracted  in accordance with FWPCA methods (U.  S.
 Department of the Interior, 1969).
 Sediment Samples
     Sediment samples have been taken from the
well bottom both to reduce contamination as well
                                                                                               45

-------
 as for analysis. The first sample was obtained by
 bucket in August '1968. Additional sediment was
 pumped from the well in September 1969 and again
 in February 1970. Fractionation by gravity separa-
 tion was attempted .on the latter sample to obtain
 estimates  of the  pesticide load carried  by each
 particle size range. However, complete separation
 of the silt-clay particles was not obtainable proba-
 bly  because  a dispersant was not  added to the
 mixture  to  aid  separation. The dispersant was
 avoided to prevent the possibility of biasing  pesti-
 cide determination.            '

 Soil Samples
      Surface  soil samples  were collected in June
 1968 from  the immediate area surrounding the
 well. In July 1968, soil core samples were taken to
 a depth of 6 feet using  a 3-inch diameter bucket
 auger. Soil backfill around the well casing was also
 sampled to  a depth of 4 feet. The soil core was
 separated  into  1-foot  increments except the top
 foot which  was  divided into  a 0-3-inch surface
 sample and  a 3-12-inch  subsurface  sample.  Core
 samples were taken again in June 1969 in the area
 of highest contamination. The sampling device used
 was  an impact driven sampling  tube.  These soil
 cores were separated into 1-fopt increments except
 the top foot which was  divided into three 4-inch
 samples. Soil samples were air dried in the labora-
 tory and  extracted using the method  of Guenzi
 and Beard (1967).
      All samples were analyzed for the presence of
 toxaphene, DDT, ODD  [1,  i-dichloro-2,  2-bis-
 (p-chlorophenyl) ethane], and DDE [1, 1-dichloro-
 2, 2-bis-(p-chlorophenyl) ethylene]  using electron
 capture equipped gas-liquid chromatographs. Quan-
 tification methods used in this study are described
 in FDA Pesticide Analytical Manual (U. S. Depart-
 ment of Health, Education, and Welfare, 1968).

          RESULTS AND DISCUSSION
 Water Samples
    .  Analysis of water samples taken in  January
 1967 showed 20.1, 1.5 and 0.5 ppb (jug/liter) of
 toxaphene, DDT,  and DDE, respectively,  while
 March  levels  declined  to 9'.3,  0.5  and  0.3  ppb,
 respectively. Contaminant levels'continued to de-
 cline, and in June 1968 levels were 8.4, 0.1 and 0.1
 ppb  of toxaphene,  DDT, and DDE, respectively.
 The June 1968 sample was the last one to exhibit
 contamination levels greater than the criteria estab-
 lished for public water supplies  (U. S. Department
 of  the Interior, 1968). Figures 2, 3, and 4  show
 that contaminant levels have oscillated at relatively
'low levels from  summer  1967 "until the  present.
 Presumably an equilibrium exists in the pesticide
 content  of the sediment and water. This is con-
 46
  0.3


  0.2
 Fig. 2. DDT residue in the water as drawn from the tap.
Fig. 3. DDE residue in the water as drawn from the tap.
Fig. 4. Toxaphene residue in the water as drawn from the
tap.
trolled by the absorptive capacity of the sediment
and the solubility of the  pesticides  in water.
Solubility of DDT is reported  to  be 1.2 ppb at
25  degrees.|C (Bowman, Acree, and Corbett, 1960).
Sediment Samples
     Pesticide levels of sediment samples are shown
in Table 1. Generally, levels in the sediment are
approximately 10" greater than levels in the water.
An exception to  this is ODD  which appears in
sediment and does not appear in the water. Physical
characteristics of the sediment such as color and
particle size appear  identical to that of surface
soil. Hence,r both appearance and results of pesti-
cide analysis indicate that  contaminated surface
soil has been transferred to the well bottom.
     Incomplete  fractionation   of  the February
1970 sediment  samples  prevent the drawing of
definite conclusions regarding the relative pesticide
load of each particle size range. Reasons are not
completely clear as to why the combined silt-clay

-------
 Table 1. Pesticide Residue in Well Sediment, ppm (/ug/g)
Time of
sample
Aug. 1968
Sept. 1969
Feb. 1970


Complete or
fraction
Complete
Complete
Sand
Silt and clay
Clav
Organic matter
DDT
8.75
3.93
0.13
10.1
0.48
23.8
DDD
11.61
8.65
0.35
7.51
2.30
145.0
DDE
1.25
0.24
0.05
0.66
0.17
13.3
Toxa-
phene
34.6
6.85
0.88
10.2
2.55
235.0
    Table 3. Pesticide Residue in Soil Core Samples,
              July 1968, ppm 
-------
that contamination exists in the soil backfill for the
complete depth of the well.
     Table 4. Pesticide Residue in Soil Core Samples,
                June 1969, ppm
Distance in feet
and direction Sample
from well depth
Backfill









5-East








10-East








1 5-East








20- East








25-East








0-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
7-8 feet
0-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
O-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
0-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
0-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
0-4 inches
4-8 inches
8-12 inches
1-2 feet
2-3 feet
3-4 feet
4-5 feet
5-6 feet
6-7 feet
DDT ODD DDE
.019
—
.013
.080
.161
.187
1.516
.733
.654
.573
.049
.091
.024
.014
T
T
-
-
-
.198
.024
.012
-
-
-
-
-
-
1.196
T
T
-
-
T
-
T
-
.188
.020
T
-
-
-
-
-
-
.086
—
-
-
-
-
-
-
—
.012
— —
T*
- .030
.017
.051
.245
.112
.093
.091
.015
.032
T
T
T
T
- -
- -
— —
.055
- .018
.015
- -
- -
- -
- -
- -
— —
.283
T
T
- -
- -
	 *j*
- -
	 -T*
— —
.045 .055
T
T
- -
- -
- -
- -
- -
- —
.025 .025
T
T
- -
- -
- -
- -
- -
— —
Toxa-
phene
-
—
-
.208
.543
.543
5.208
3.400
3.000
2.253
-
.417
-
-
-
-
-
-
—
.700
-
-
—
—
—
—
—
—
7.128
—
-
-
—
—
—
—
—
.625
—
-
—
—
—
—
—
—
.217
—
—
—
—
—
—
—
—
  * Values less than .010 ppm are shown as trace (T).
  48
     The occurrence of DDD in sediment samples
and  its absence in soil backfill  samples, as well as
water  samples, raised questions beyond the scope
of this study. However, Miskus, Blair, and Casida
(1965) studied the conversion of DDT to DDD in
lake water  and concluded that  oxygen content of
the water helped control the conversion rate. Also,
they concluded  that DDD  appears  to  be  more
stable than  DDT under certain conditions.

                   SUMMARY
     This study of a private water supply demon-
strates the susceptibility of improperly constructed
wells to contamination with pesticides. Results of
monitoring water and soil samples for more than 4
years  point to the persistence of both DDT  and
toxaphene. Probably  because of very low solubili-
ties  of these insecticides in water, contamination in
the  water  remained low.  Contamination  levels in
the  sediment were much  higher. Soil sampling in
the  area around  the well indicates  the pesticides
have  moved  only slightly  downward,  and  con-
tamination of the  well occurred through the move-
ment to the water table of surface soil containing
adsorbed pesticides.
      Results of these studies emphasize the impor-
tance  of good design  and proper installation for
 shallow wells on farms.
      The author gratefully acknowledges the assist-
 ance  of Dr.  A.  W.  White, Soil Scientist, in the
 analytical determinations made in this study.

                  REFERENCES
 Bowman, M. C., F. Acree, Jr., and M. K. Corbett.  1960.
      Solubility of C14 DDT in water. Jour, of Agricultural
      and FoodChem. 8:406.
 Guenzi, W.  D., and W. E. Beard.  1967. Movement and
      persistence of DDT and lindane in soil columns. Soil
      Sci. Soc. Amer. Proc. 31(5):644-647.
 Miskus, R. P., D. P. Blair, and J. E. Casida. 1965. Conversion
      of DDT to DDD by bovine rumen fluid, lake water,
      and reduced porphyrins. Jour,  of  Agricultural and
      FoodChem. 13-.481-483.
 Trautman, W. L., G. Chesters, and H.  B. Pionke. 1968.
      Organochlorine insecticide composition of randomly
      selected soils from nine states-1967. Pesticides Moni-
      toring Jour. 2(2):93-96.
 U. S. Department of Health, Education, and Welfare.  Food
      and Drug Administration. 1968. Methods which detect
      multiple residues. In Pesticide Analytical Manual, Vol.
      I. Looseleaf.
 U. S. Department of Health, Education, and Welfare. Public
      Health Service. 1962. Manual of individual water-sup-
      ply systems. 121 pp.
 U. S. Department of the Interior. Federal Water Pollution
      Control Administration. 1968. Water quality criteria.
       234 pp.
 U. S. Department of the Interior. Federal Water Pollution
      Control Administration. 1969.  FWPCA method  for
       chlorinated  hydrocarbon  pesticides in  water and
       wastewater. 29 pp.

-------
                 DISCUSSION
The  following questions were  answered by M. J.
Lewallen after delivering his talk entitled "Pesticide
Contamination of a  Shallow  Bored Well in the
Southeastern Coastal Plains."
Q. Was this  a routine sampling?  Did the owner
suspect contamination and request specific analy-
sis? Did this result from a special survey or study or
were there other reasons for sampling and analyses?
A. Contamination was detected by taste in the
water in this instance. I think,  in general, most of
these materials we're speaking  of have taste thres-
holds in the low  parts per  billion  range. In this
particular instance, the owner detected a strange
taste  and  odor  in the  water. At that point he
realized  what had happened and asked  us  what
steps could be  taken.
Q. Has pesticide contamination been measured in
nearby streams?
A. No, they haven't. Pesticide determinations have
been  made in a well which  is approximately 100
feet  from the present well and no pesticides have
been detected in the second well.
Q. What are levels of DDT,  DDE, toxaphene, and
so forth, in  soils  at various depths under natural
conditions in agricultural areas not grossly polluted?
A. This will vary  with soils. In general, the down-
ward movement of these  materials is very  slight.
Probably 6 or 8 inches is about the greatest depth
you'd expect to find these pesticides in these soil
 conditions.
 Q. Do you know of any adverse  effects upon the
 owner or the animals in the area?
 A. No, I know  of no effects. Even though this is a
 small farming operation, water has been used for
 farm  animals continually  since installation. I say
 that without further  comment, because I'm of the
 opinion that no  analysis  was made of any meat
 animal  which left the farm,  so there could have
 been unknown contamination.
 Q.  What method was used to  separate the organic
 components from the soil?
 A.  The method used was wet sieving.
 Q.  Could you discuss that a little more?
 A.  I might point out that problems encountered in
 separating soil sediment for pesticide analysis were
 much greater than I anticipated at the beginning. In
 fact, I know of no completely fool-proof method of
 doing  this. Possibly  a high-speed centrifuge could
 be used for this separation.
      The  sediment was placed  in  a container of
 water,  and  sieved with appropriate size sieves to
 remove sand and progressively smaller fractions as
 far  as sieving is possible. Because it is not possible
 to  sieve down  into  the clay fraction, the final
separation was attempted by a gravity  separation
method. The silt and clay fraction was mixed in a
container of water, stirred thoroughly, and allowed
to set an appropriate time for the settling of the
larger particles, after  which  clay  particles were
siphoned off in the water fraction. This procedure
was  repeated  six or seven times. Water was then
allowed to evaporate from each mixture. I feel that
the determination for  the silt fraction is probably
biased upwards because of. an unknown amount of
clay which was not dispersed adequately and ad-
hered to the silt particles.
Q. Max,  what was the soil texture, and can you
speculate why the silt had more pesticides than the
clay?
A.  I think the soil here is a  sandy loam. I believe
the  question  as to why  the silt  contained more
pesticide was answered  in the  discussion  on the
previous question.
Q.  The data showing  vertical movement of pesti-
cide to the well is not very clear. Would you repeat
how you know pesticide moved to the well through
the ground?
A.  I'm sorry  I was not clear on that point. In fact,
I did not intend to indicate that pesticide moved
vertically through the soil to the well. The pesticide
contaminated  soil  was  transferred to  the water
table and the bottom  of the well. This was shown
 by  the soil profile chart which  showed  backfill
 material of  the  same color, texture,  and  particle
 size as top soil, and also from pesticide concentra-
 tions in top soil as compared to pesticide  concen-
 trations in the well sediment. Concerning the down-
 ward  movement  of these materials  in  the  area
 where  the  flushing  and  spraying  operation had
 taken place,  I intended to show that only a slight
 downward movement occurred.
 Q.  Max,  did you  investigate any ionic  change
 which might have caused an increase in chloride in
 the ground water? Also,  have the total dissolved
 solids been affected?
 A.  No, I have not looked at that.
 Q.  We have  about five questions  here concerning
 the term organoleptically. What is the term organo-
 leptically?
 A.  That would be when you have a strange taste or
 odor present.
 Q.  What is the symbol p,p' that you had before the
 DDT in the abstract? That was probably an error, I
 imagine.
 A.  No, p,p' is the para para isomer of DDT. DDT
 is  composed of the para para, p,p', and the ortho
 para, o,p',  isomers. Commercial grade DDT is com-
 posed  mainly of the  p,p' isomer with  a small
 portion of the o,p' isomer.
                                              49

-------
Gasoline  Pollution   of a  Ground-Water Reservoir
A  Case  History
by Dennis E. Williams'3 and Dale G. Wilder0
                   ABSTRACT
     A leak in a product gasoline pipeline near the City of
 Los Angeles has caused contaminarion of a valuable ground-
 water supply.  Since  1968, it is estimated that 250,000
 gallons of gasoline have seeped  into  the underground
 reservoir,  limiting the value of a well field  adjacent to the
 contaminated area.
     Remedial measures include extensive analytical studies
 of the two-fluid flow system as well as an all out effort in
 the field  to try and clean up  the gasoline  and restore the
 aquifer to service.  The field methods involve an elaborate
 system of "skimming"  wells designed to  produce a high
 gasoline/water ratio and on-site treatment facilities at many
 locations  throughout the area. To  date, 50,000 gallons of
 free gasoline  have  been removed from the aquifer in this
 manner.
                 THE PROBLEM
      In  September 1968, Forest Lawn Memorial
 Park (located  in the  City of Glendale, California)
 reported pumping free  gasoline  from one of their
 large irrigation wells.  Upon removal of the produc-
 tion pump  they proceeded to  pump  and bail over
 1,000 gallons  of free gasoline from the well. A
 drilling contractor retained by Forest Lawn drilled
 some 20 small diameter wells  in the area in order
 to determine the quantity and extent of the gaso-
 line. From the results of these first 20 wells it was
 estimated that approximately 160 thousand square
 feet of area were influenced with the average depth
 of free gasoline ranging from 6 to 18 inches. Using
 an average porosity of  30 percent the volume  of
 free gasoline existing in  the reservoir was estimated
 to be about 250,000 gallons (see Figure 1).
      Shortly after, the Western Oil and Gas Associ-
 ation  (referred to as VVOGA) was contacted and
 proceeded with  pressure tests of a buried 8-inch
 gasoline  pipeline located  in the area in  hopes  ol
      ^Presented at  the  National Ground Water Quality
 Symposium, Denver, Colorado, August 25-27, 1971.
      ^Ground-Water Hydrologist, Los Angeles Department
 of Water and Power, 111 N. Hope St., Room 1411, Los
 Angeles, California 90054. (Now Senior Hydrologist, Louis
 Berger, Inc., Tehran, Iran.)
      cHydrologic Engineer, Los Angeles Department of
 Water and Power,  111  N. Hope St., Room 1411, Los
 Angeles, California 90054.
             STRONG TASTE
             AND ODOR AREA
Fig. 1. Contaminated area and ground-water contours, as of
1968.

determining the source and quantity of any leaks.
Although initial tests were  negative, later  test  re-
sults showed possible evidence of a leak, and flow-
in the pipeline was immediately terminated. WOGA
then initiated an abatement program and proceeded
to hire experts with the hope  of attaining  rapid
solutions  to  the contamination. In  November of
1969,  the  Los Angeles  Department  of Water and
Power turned off its nearby Pollock Well Field to
prevent the contamination from migrating into the
influence cone of the wells.
     Extensive drilling and testing continued in the
area under the direction of WOGA to further define
the problem. In April, 1969, it was evident that
some  immediate action  had to be taken to prevent
contamination of all producing wells in the area. It
was decided  to concentrate on an intensive drilling
and pumping program to  control the lateral spread
of gasoline.
 50

-------
              GEOHYDROLOGY
     The area involved  is located  in the southeast
section of the San Fernando Valley where the basin
narrows  and  contacts  the coastal  plain  of Los
Angeles.  In the Forest  Lawn  area, the valley is
approximately two miles wide and  composed main-
ly  of  alluvial  outwash  from the  San  Gabriel
Mountains lying to  the north. Sands and gravels
interbedded between lenses of  silt and clay consti-
tute an important ground-water reservoir.
     The  ground-water  reservoir  is  bounded  by
Plutonic  basement rocks on the east and northeast,
by  nonwater-bearing Elysian  hill formations to the
west and  southwest, and nonwater-bearing sedi-
ments on the south and southeast. In the immediate
Forest Lawn area, the alluvium overlies a basement
complex seen outcropping in the  hills to the east.
Immediately overlying  the  basement  rocks, and
underlying the alluvium, is a layer of grey organic
clay.
     The random nature of the depositional cycles
has caused the  ground-water flow to be predom-
inantly anisotropic with higher hydraulic conduc-
tivities in the direction of alluviation. Interbedded
with the  sand  and gravel deposits  are extensive
lenses  of brown clayey-silt which forms localized
barriers  (aquitards)  to  vertical ground-water flow
and have  created  multiple  aquifer systems  (see
 Figure 2). The depth of the unconsolidated material
ranges from 150 feet to 250 feet  deep in the reser-
voir area.
     Aquifer tests conducted in some of the recent-
 ly   drilled  wells  show  hydraulic  conductivities
 ranging  from lows of 500 gpd/ft:  and less in the
 silty zones, to over 2,000 gpd/ft2 in the clean sands
 and gravels. Storativities range from 0.10  to 0.005
 suggesting semiconfined aquifers with the confining
 layers being  somewhat leaky. These  aquifer test
 results have been  supported  by  geologic  samples
 and electric logs.
      The  only  significant structural feature in the
 area is the east-west trending  Raymond Fault. The
       A
       z
                             «
            - BROWN CLAYEY SILT

               GREY ORGANIC CLAY	
fault displaces bedrock (Figure 2) but as of yet, no
evidence of displacement (geologic or hydrologic)
in the recent alluvium has been found.

                  SOLUTION
     Since the  original indication  of the problem,
dating back as far as April 1968, nearly 70 wells
have  been  drilled  in  an  attempt to  adequately
define the problem and expedite the removal of the
                                        FIELD
                                *
              NEWMAN FIELD—   *"
                                >
                                £-•$. .W-26  FIELD
                              W-58
                                  *'*
             • PUMPING WELL
             «fc OBSERVATION WELL
  Fig. 2. Geologic cross section — Forest Lawn area.
                         POLLOCK FIELD
 ,         **>   -    <-fs     ^    RIVER
 Fig. 3. Pumping and observation wells.

 gasoline  (see  Figure  3).  Principal removal  tech-
 niques have  included  skimming operations  using
 both dual and single completions. In the dual com-
 pletion technique the lower aquifer was packed off
 and  clean  water pumped from this  zone  while
 another  small  pump removed  free gasoline and
 water from the upper contaminated zone above the
 packer.  (The packer in the well provided  effective
 separation  for  a  while;  however,  the confining
 layers were either leaky or only locally continuous
 as gasoline was eventually detected in objectionable
 amounts from the lower zone.)
      Widespread use was  made of  the  single com-
 pletion, in which small jet pumps were placed very
 near the surface of the water to remove the free
 gasoline  layer from the top  of the water surface.
 The skimming  wells would then pump into separa-
 tor  tank assemblies located in the  area.   These
 separators were composed of density stratification
 tanks coupled to a larger baffled tank where floata-
 tion and skimming, along with aeration in the water
                                               51

-------
progressively separated the gas-water mixture. Free
gasoline  was then temporarily stored in a separate
tank prior to disposal by tank trucks.
     During  the  three years  of clean up,  over
50,000 gallons of gasoline have been removed from
the reservoir. It  is believed  that  most of the free
gasoline  has been removed, as most of the wells no
longer show any measurable amounts of gasoline
(see Figure 4).
                     -\
                              ^  s^ ^^. ^"-. ^v
                           .g/O^-x \XX
         STRONG TASTE   /?   A',   \\ \  '
         AND ODOR AREA/X*/!';^  '\^ \
                                          \   y  -,
                      /    /'/; •m^/ /  i   \  \
                      f      -   l(^
           f       /     f,  - i  i    ^
      «f ;
        /
      &%JnJ—^J	&.
 !".      0 PUMPING WELL
 ; x     « OBSERV*'
 1 «„.    <£^
                          •  /. *•
.TIOM WELL         / '     / / •  A> .
           POLLOCK FIEL(D J^"  /
Fig. 4. Contaminated area and ground-water contours, as of
1971.
     The possibility that residual gasoline  might
exist which  could not be flushed out  with  water
was recognized long before the clean-up operation
was started and the difference between  the original
free  gasoline  estimate  and  the actual  amount
verified  this possibility. It is possible that the lack
of data  resulted  in erroneous first estimates.  It will
be recalled  that in the original estimate a porosity
of  30 percent was  used in the calculations, while
later pumping test results showed effective porosi-
ties of 10 percent and less.
     The difference between  the original 250,000
gallon estimate  and the  50,000  gallons removed
could  also  be explained by a  reduction in the
relative  permeability of the gasoline. In immiscible
(two fluids) flow, the permeability1 of the forma-
      1 Permeability as used here is actually the Hydraulic
 Conductivity used in ground-water terminology. Due to the
 large  amount of  literature in the petroleum field on two-
 fluid  flow,  the  term permeability  instead of hydraulic
 conductivity is used here for consistency.
                                       tion to one of these fluids  is dependent on  the
                                       saturation  of the other  fluid.  A relative permea-
                                       bility  graph  (Figure  5)  shows  the  relationship
                                       between the  observed  permeability of each fluid
                                       for various  saturations  to that  of the observed
                                       permeability if the sample were 100 percent satu-
                                       rated  with that  fluid  (this  permeability  ratio is
                                       denoted as Kr). The curves shown  in Figure 5 are
                                       not based  on any actual laboratory work  done in
                                       the  Forest  Lawn area  and  cannot  be used  for
                                       quantitative evaluation of  the  problem. This rela-
                                       tive permeability concept is not new and has had
                                       widespread  use  by  the  oil  companies in their
                                       production operations.
                                        Fig. 5. Relative permeability graph.


                                             The three sections shown  in Figure  5  (I, II,
                                        III) represent different flow characteristics resulting
                                        from different water saturations. In Region I there
                                        is a high saturation of gasoline; nearly all the fluid
                                        flow  is  composed  of gasoline because  it is  a con-
                                        tinuous  phase while water is discontinuous. This
                                        results  in a  low  relative  permeability to  water.
                                        Since gasoline is the nonwetting fluid2, the smaller
                                        capillaries will be filled with water (see Figure 6-a).
                                             In  Region  II both  water  and gasoline are
                                        continuous,  although not  necessarily in the same
                                        pores. This continuity allows flow  of both  water
                                        and gasoline. As the saturation of water increases, a
                                        larger percentage of water will flow (see Figure 6-b).
                                             In  Region III (Figure 6-c) there is  flow of
                                        water with little or no flow of gasoline. The smaller
                                        capillaries are filled with gasoline only when the
                                        pressure drop overcomes  capillary forces.  Thus,
                                             2 If adhesive force > cohesive force then  6  < 90
                                        degrees and  the liquid is said to be the wetting phase (see
                                        Figure 7).
 52

-------
                                            (a)
    GASOLINE FLOW
               PRESSURE DECREASE-
              REGSON 1 - GASOLINE FLOW
             WATER FLOW
     GASOLINE FLOW
                                            (b)
               PRESSURE DECREASE-
                RESIONII - MIXED FLOW
                                            (c)
               PRESSURE  DECRE4SE-
               RE6ION HI - WATER FLO*
Fig. 6. Gasoline and water flow.

until the pressure drop (Pgas - Pw) is greater than
Pcap, it is  impossible to move  the "snapped off"
gasoline bubbles through  the throats. As a result,
the gasoline does not fill the "neck" or pore throat,
and  becomes extremely  difficult  to move. The
principle is illustrated in Figure 7 with the follow-
ing conditions pertaining:
     For equilibrium:
        rgas
         - P  = P
           r w   L c
                          2 a cosi
                  ap
   For movement:

Pgas ~ PW ^ Pcap
                            nonwetting  gasoline
                      will displace the water.
If Pgas ~ PW < pcap    tnen  gasoline  cannot  dis-
                      place  water  and  therefore
                      cannot  move through the
                      capillary but will be forced
                      back  until  equilibrium  is
                      reached.
Where:
                                        20 cos 6
  Pcap =  Pressure due to surface energy =
                                           'c

    Pw =  Pressure on the water side of interface.

   Pgas =  Pressure on the gasoline side of interface.
                                                                                      NECK  OR
                                                                                   PORE  THROAT
Fig. 7. Pressures on a capillary interface.

     a =  Surface energy (a function of the fluids).

     B =  Angle of contact between fluid and rock
          matrix (a function  of both rock matrix
          and fluid).

     rc =  Radius of the capillary.

     This  relative  permeability  can  be  both  a
hindrance to the removal process as well as an aid
to producing clean water.  Since there will always
be  some channels or necks which are bigger than
others, it is virtually impossible to flush out all the
remaining  gasoline. The water which is  used to
flush the  sand will tend to flow through unblocked
or  continuous  water-filled  channels  rather  than
through the gasoline-blocked channels. Conversely,
water which is produced from a contaminated zone
may  be essentially clean if the gasoline saturation
has been reduced to a low level.
     This idea has been investigated by Los Angeles
Department of Water and Power hydrologists as a
means of cleaning the gasoline contaminated water
after it has been pumped. In the actual experiment,
filter sand  was  saturated with water and then, a
gasoline-water emulsion forced through the filter.
The ppm of gasoline  in  the discharge was notice-
ably lower due to this process.
      In addition to the lack of data  and reduced
relative permeability, another reason for the differ-
ence between  the  estimated spillage and  the total
recovered gasoline is the mixture of  gasoline and
water which occurs in the capillary  fringe. While
the capillary fringe will be compressed somewhat
by the  head of  gasoline, there will still be a zone
which contains water held by capillary forces and
gasoline which is free to  drain. When  the  depth of
"free" gasoline is measured in the well, the thick-
ness of gasoline includes both the depth of  gasoline,
which is  above the fringe, plus the  depth of gaso-
                                              53

-------
line which  is within the capillary fringe. This could
make the original estimate significantly higher than
it should be (see Figure 8V
                        , /    CAPILLARY FRINGE
                        ,',MBOTH WATER 8 GASOLINE)
 (100% WATER SATURATION'
•"7
                                  AQUIFER
                      BEDROCK
 Fig. 8. Gasoline-water table relationships.

      Ihe major current problem facing hydrologists
 and engineers is how to  adequately clean the satu-
 rated  and  unsaturated portions of the  aquifer of
 residual gasoline. Of several techniques which have
 been considered, one  of the more promising is to
 encourage  growth of naturally  occurring but nor-
 mally dormant bacteria  (principally genus pseudo-
 monas) which have been detected in water samples
 from the area. These bacteria grow on gasoline and
 have been  noted in wells and separator tanks in the
 Forest Lawn area.
      Preliminary studies indicate that the bacteria
 convert two  thirds  of the gasoline  to carbon diox-
 ide and one third to new bacrcna cells. The process
 is limited by axaiiabiiitv of oxygen, mineral nutri-
 ents and gasoline-water interfaces.  It is hoped that
 the  bacteria can be utilized to  clean the aquiter of
 the  residual gasoline without plugging up the wells
 and formation with bacterial slime. Work is present-
 ly  underway by WOGA  and petroleum  companies
 to solve this slime problem. Unfortunately most of
the work which has been done is not readily availa-
ble,  and  the  applicability of  this process to the
Forest Lawn problem cannot be currently analyzed.

   SUMMARY AMD  PRESENT OPERATIONS
     A gasoline spillage of  significant  magnitude
(250,000 gallons)  occurred  in the San Fernando
Valley, prior to November 1968, which threatened
an important ground-water supply. The City ot Los
Angeles  was forced to terminate production from
its Pollock Well Field to prevent contamination ot
the water supply. An  intensive  drilling program u .<•••>
undertaken to define  the limits and  contain the
spread of gasoline.
     Efforts to drain off the gasoline have resulted
in a total recover}" of 50,000 gallons  of gasoline.
Most of the wells now have  only a taste or odor ot
gasoline with  only two wells  showing any  "free"
gasoline.  At present,  operations arc ar a standstill
with biweekly measurements and monthly  samples
taken  in order  to monitor  any  changes  in the
contamination.
     It is hoped that bacteria  will eventually break
down  the residual  gasoline so  that the  acjuiier may
be restored to service. Studies ol the  feasibility ot
this technique are currently  being made by WOGA,
the Department of Water and Power, and others.
     The  results  of  this  contamination  are very
 costly. The loss of the Pollock Well Field  alone  is
 significant, in addition to the cost of drilling and
 operating the 70 wells which are required to con-
 tain the spill, and drain off the gasoline.
      The costs of remedial measures  necessary to
 eliminate the contamination  in a problem ot this
 type  far outweigh   the  relative!}'  low  costs  o!
 preventive maintenance.


              REFERENCES USED
 Hubbcn.M. Kine. 1Q56. Darcv's  i-,m and iht  ncld  equation-
      ol the tlov. ol underground iluids. Transactions ot inc
      AIM!-', v.  207. p. 222.
 Kimkenberg, L. .1.  !«42. I'hc- permeability or porous nu\i:;i
      to  liquids and gases.  API  Drilling and Production
      Practices, p. 201'.
 Muskat, M. 1946. The flow of homogeneous fluid-; thnmgh
      porous media, j. \V   Edwards,  Inc.. Ann  Aroor,
      Michigan, p. 763.
 Kabe,  C. L. 1967.  Unpublished  lecture notes.  Shell  Oil
      Company. Technical Training Division.
 D-t-

-------
                                           DISCUSSION
The following questions were answered by Dale G.
Wilder after delivering his talk entitled "Gasoline
Pollution  of a Ground-Water  Reservoir—A  Case
History" (coauthored by Dennis E. Williams).
Q. Dale, we have a two  part question here-(l)
What is the estimated ground-water volume within
the contaminated area? (2) What is theguesstimated
volume for clearing the aquifer down to the thres-
hold point for gasoline, considering hydrodynamic
dispersion and flow conditions?
A. On the  first question, I'm going to have  to:
hedge a little bit. The title of the paper refers to the
problem  as "A Case History" which is somewhat
misleading because it implies past rather  than
current history. However, the outcome is not yet
known and some information, if it were made
public, might influence that outcome. Let me give
you just a little bit of  the history of this problem.
The utilities and companies  involved,  and there
were several, felt that a solution  to the problem
was urgent; therefore  they initiated cleanup and
monitoring programs, assuming that they would be
able to recover their expenses once the source of
pollution  was found.  It now appears that there
possibly will be some  sort of litigation or  out-of-
court bargaining coming up in the future to recover
these  expenses. For this reason I can't release the
estimated ground-water volume or the amounts of
money involved.
    As far  as the volume for clearing the  aquifer
down to the threshold point  for gasoline,  I think
that the current feeling is that it can't be done sole-
ly by  pumping of  ground water, and this is the
reason for stressing the bacteriological approach.
The lowest gasoline contamination from the separa-
tion  facility  effluent  is  about  1  to 2 parts per
million which is well above the threshold value, and
into the objectionable range.
Q. We have three questions here. We'll go one at a
time.  Do you feel that the tank and pipeline test
techniques used for routine testing are adequate?
A. This, once again, gets into one of these problem
areas.  I  think, from  what I  have read, the test
procedures are adequate. However, the interpreta-
tion  of  the  results  is subject to argument.  The
procedure used was to pressurize the tank or pipe-
line and then attempt  to inject diesel fuel or  some
other  fluid into it. Any diesel fuel accepted by the
tank or pipeline would either indicate leakage or
some  other  problem.  The petroleum industry felt
that  on  these particular tests, temperature,  com-
pression  of entrapped  air, and so forth, explained
the volumes of diesel  fuel used.  However, it was
felt  by many people  in the water industry that
these tests did show possible leakage.
Q.  Were you  satisfied  with  the  test  of these
facilities after the problem was identified?
A.  I  think  we  were  basically satisfied  with the
second set of tests that were run. The Department
of  Water and Power  sent representatives out to
observe the tests, and we were satisfied that this
particular test was an adequate test. It was on the
basis of this test that the pipeline was closed in.

Q.  What kinds  of pumps were used? Were there
any special modifications to prevent explosion?
A.  Most of the  pumps were air or water operated
jet pumps.  There were strong odors built up which
indicated a possible explosive condition, but to my
knowledge,  as  far as the pumps were concerned,
there were no special precautions taken other than
shielding of all the electrical wiring.
Q.  Are the pipeline pumping records so poor that
a 250,000  gallon loss could  go  unnoticed,  and
could the   company  records  be  subpoenaed to
determine the actual loss?
A.  Let me  answer the second part of that first.  I
think that the  records could  be subpoenaed. As
far as "Are  the  records this poor," when pinned
down, WOGA admitted a 250,000 gallon loss would
probably be picked  up  in  the accounting pro-
cedures. I don't think that they are willing to admit
that  there  were 250,000 gallons lost. I'm not sure
just  exactly how  much they can detect.  They
haven't released this figure to us.
Q.  Was any attempt made  to  identify what com-
pany produced the gasoline by chemical analyses?
A.  Yes. There were several tests run and attempts
made to  identify this gasoline.  Unfortunately,
contrary to  some of the advertisements that the
industry puts out, WOGA tells us  that they can't
pin down the  brand of the gasoline strictly by
chemical analysis. Originally,  they  thought that it
was  several different types  of  gasoline  mixed
together; then they felt maybe it was just one type
which had been altered in the underground, and so
forth. However, they have not been able to identify
it,  at least to the point that  they are  willing to
admit that they know whose gasoline it is.
Q.  Did soluble components pollute the water all
the  way to the bottom of the aquifer, or  was water
from the bottom still drinkable?
A.  It appears that water from the lower portions
of  the aquifer was clean and able  to  be used for
domestic uses. In the Forest Lawn No. 4 well,  I
mentioned that they had pumped from the second
                                             55

-------
or the lower part of the aquifer. However, after
pumping for a long period of time, it appeared that
there was some downward leakage of gasoline into
the area. This was somewhat expected based on a
viscous model study made by Dr. Williams which
assumed  that the clay formed a leakage layer and
not a solid barrier to fluid flow.
Q.  Has the owner of the product pipeline shared
in the cost of removing the gasoline from ground
water in the problem area? If not, why not?
A.  Well, I imagine they have,  but this information
is not officially available to us. WOGA has repre-
sented the oil industry, and originally it was fairly
common  knowledge that the  industry as a whole
was sharing in the cost. We understand, or at least
get the feeling from unofficial sources, that recently
one company, specifically Mobil, has been sharing
the majority of the cost. However, I know of no
official notice as to who is paying the cost other
than that monies are coming through WOGA  for
the industry  as a whole. This inability to  obtain
definite information is one of  the problems that
you run  into with representatives  speaking for an
industry as a whole.
 Q. Why did you not use activated carbon filters to
 remove the taste and odor problems after the pure
 gasoline was removed?
 A. I'm sorry  I didn't  mention  that  when  I was
 talking about the sand  filter. Activated carbon was
 tried and found to be much more effective than the
 filter  sand  that was talked about. As  a result, this
 was suggested as  the means for  getting rid of the
 gasoline in the effluent from the separator tanks.
 The  only  attempt  that  has been  made by  the
 Department of Power  and Water was where they
 hooked up a small system to eliminate the gasoline
 from  the separator tank itself. It looked hopeful
 and was working fine; however, it was a very small,
 more or less experimental, model. The size of the
 facility  needed would  have  to be built by WOGA
 and they are not, to my knowledge at least, willing
 to do that at this time.  They do not feel that it
 would be economical.
Q. Has the  total cost-to-date of the Forest Lawn
problem been estimated?
A. Part of these costs have been estimated. It is
fairly common knowledge that WOGA has spent in
the range of $700,000  in their drilling and abate-
ment program.  However,  the  costs incurred by
other parties are not available because of possible
litigation.
Q. Did you measure the lead content of the water
from the wells?
A. I'm not really sure  of  the lead  content of the
water  itself. The content of the  gasoline samples
was  taken and it was found to be high-test ethyl
gasoline, but I'm not sure on the water samples. I'm
sure that the lead content of the water is measured
by our Sanitary Department, but I'm not aware of
the results.
Q. Was consideration given to removal or disposal
of gasoline by burning?
A. I really don't know whether  or not they con-
sidered burning the gasoline. To  my knowledge it
wasn't seriously considered.
Q. Will you inject the  gasoline decomposing bac-
teria through wells, and how far do you expect the
bacteria to move?
A. These bacteria occur  naturally in  the water
 itself, so fortunately we won't have to try to inject
 the  bacteria.  The biggest problem  I think will be
 to try  to  get  the  oxygen and nutrients to the
 bacteria to let them work on the gasoline, because
 the bacteria creates a slime which blocks off the
 oxygen and nutrients.
 Q.  What kind of contamination is occurring in the
 aquifer from Forest Lawn?
 A.  Well, basically,  the  aquifer around the Forest
 Lawn  area is not able  to be used at all for water
 supply at this time. It  appears that the pollution
 at least is under  control—that  is, we have not
 picked up any new gasoline contamination. As far
 as contamination by Forest Lawn itself, they use
 various chemicals, and it was suggested that the
 reason the  gasoline could not be identified was
 that there were chemicals from the  cemetery mixed
 in it.  However, 1 think that  this  was discounted
 later   and,  to  my  knowledge,  gasoline  is  the
 principal pollutant in the area.
  56

-------
Petroleum   Contamination  of  Ground  Water
in  Maryland
by John R. Matis"

                  ABSTRACT
     Petroleum contamination of ground water is a wide-
spread problem that plagues many areas. Historical data
collected in Maryland indicates that most counties in the
State record cases of this contamination problem annually.
Cases in the "hard-rock areas" west of the Fall Zone have
the  highest frequency of occurrence,  in contrast  to the
Coastal Plain geologic province to the  east. In both areas,
the  problems have  been very localized.  It is difficult to
handle petroleum contamination cases,  and the legal impli-
cations are very complex. A disturbing  factor is that many
petroleum fuels do not deteriorate in the ground-water
system. Further, the identification of specific petroleum
products in  ground water  is generally not possible with
present techniques. An investigation of a particular com-
plaint can often be split into  a preliminary phase and a
detailed site investigation  phase.  Once a source of  con-
tamination is located, it must be stopped or removed. Since
it is virtually impossible to remove the contaminant from
the ground water, legal and regulatory problems continue on
for months or years after an original complaint.

                INTRODUCTION
     Petroleum contamination of ground water can
be  expensive and time  consuming for water re-
sources regulatory agencies.  Typical of after-the-
fact ground-water quality problems, it is reported
or  defined only after  the damage  has occurred.
County health departments were  surveyed in 1971
for information  on the subject, and the historical
 data collected indicate that many  homeowners have
 been troubled with petroleum fuel  in their wells in
 the past.
      The geology, geography, and climate of Mary-
 land is similar to other Eastern Seaboard States and
 as  a result, the hydrogeologic principles of petrole-
 um contamination defined here, should apply  to
 other areas.  Primarily  because of this reason, the
 American  Petroleum Institute was contacted  in
  1971  for assistance and advice. The API expressed
 an interest in the problem and indicated that con-
 siderable research has been done in Europe on the
 topic.  This  report presents  guidelines which are
 applicable for use by water resources management
 agencies.
      Presented  at the National Ground Water Quality
 Symposium, Denver, Colorado, August 25-27, 1971.
      ^Groundwater Management Division,  Maryland De-
 partment  of  Water  Resources, State Office  Building,
 Annapolis, Maryland 21401.
          THE GEOLOGIC SETTING
     The geologic framework of the State of Mary-
land defines the severity and location of this con-
tamination problem.  In general,  the  Fall Zone,
where  the Coastal Plain  sediments contact hard
rocks of the Piedmont, acts as a dividing line  in
dealing  with complaints. The "hard-rock"  areas
west of the Fall  Zone have the highest  frequency
of cases of petroleum contamination in contrast to
the Coastal Plain geologic province to the east (see
Figure 1).
     From examining the data it would appear that
the area around Annapolis has a very high incidence
of contamination. However, the presence of State
offices in Annapolis and  the active county health
department staff is reflected in the greater number
of cases recorded for Anne Arunde! County (12).
Also, these statistics include  only those cases re-
ported to State agencies and county health depart-
ments. Oftentimes, other cases are not reported and
individual problems encountered from leaking gaso-
line station tanks are handled by  the company  in-
volved. As communication improves between coun-
ty  and State agencies,  and the public, more cases
will be reported.
     In both the hard-rock and Coastal Plain areas,
the problems are very localized, and individual cases
are often confined to several acres. The difference
in frequency of occurrence results from hydrologic
factors.  Silts  and clays in the Coastal  Plain sedi-
ments, for example, tend to attenuate hydrocarbon
fuels that leak from buried tanks and  from spills.
Also, the high density surface drainage  may inter-
cept a great deal of the material before  it becomes
widely dispersed. And finally, the dilution factor of
the ground water in the Coastal Plain is very great.
     In  the hard-rock  areas, ground-water flow is
 confined to fractures in the rock,  and hence flow is
 channelized to  some extent.  This means  that the
 contaminant  remains  more  concentrated in  the
 ground  water. Secondly, clays and silts are gener-
 ally absent and  attenuation is less important as a
 controlling process. In general then, the petroleum
 accumulates in the fractures,  which are defined by
 local geologic structure, and  flows in well-defined
 patterns. In limestone regions, or in other similar
 hydrologic situations, the potential travel distance
 of the contaminant would be much greater than in
                                               57

-------
Fig. 1. Cases of petroleum contamination recorded in Maryland counties, 1969-1970.
the Coastal Plain. Likewise, the transit time of the
fuel from the  source to the well could be much
shorter.

      COMMENTS AND OBSERVATIONS
     Usually a local sanitarian will receive the initial
complaint and  in Maryland, he then refers it to the
Maryland Department  of  Water Resources.  This
procedure has  developed  for several  reasons. For
example,  anyone  who  has  ever tried  to  define
source areas for the petroleum products or who has
confronted the aggrieved individual with the  con-
taminated well can  testify to the  extreme  diffi-
culty of handling these cases. The distraught home-
owner generally does not understand the nature of
the problem  or the  nature of ground-water  flow.
As  a result, tempers flare  and few cases are  ever
settled to the satisfaction of the aggrieved party(s).
Also, the State entered the problem here partly
because of the complex legal nature of contamina-
tion cases. Expert opinions  are needed for adequate
settlement, and of course, the State wanted to keep
track  of any  deleterious  water quality  changes
within its domain. Our present position is to define
the  situation, eliminate the source of contamina-
tion, and where appropriate, try to locate an alter-
nate source of water (such as another well site) for
the aggrieved party.
     The  most disturbing  and pervasive factor in
58
these cases,  is that gasoline  and  kerosene (low
boiling  point hydrocarbons)  essentially  do  not
deteriorate  in  the ground-water  system  (James
Grutsch, American Oil, personal communication).
Once they reach the water table they remain there
indefinitely or until they move out of the area with
the ground water. These low boiling point hydro-
carbons actually float on the water table at higher
concentrations. As a result, any  open hole,  and
particularly a pumping well, tends  to act as a
hydraulic sink and draws in the material.
     Identification of specific petroleum products
at the very low concentrations typically found in
contaminated  wells is generally not possible with
present techniques. The infrared spectrophotometer
procedure requires at least 1-10 ml. of concentrate.
Oftentimes it is virtually impossible to remove this
required   amount.  The  gas  chromatograph may
alternately be used, but it can only identify "low
boiling  point" or  "high boiling point hydrocar-
bons"—not  specific  unknowns like gasoline  and
kerosene. The human  threshold  for detection of
gasoline, however, is O.OO5 mg/1 and approximately
0.01 mg/1  for fuel oil  (McKee and Wolf, 1963. p.
230). As  a  result, you  can  smell and taste hydro-
carbon fuels in water even though you can't detect
it on the gas chromatograph.
     Of  course,  high concentrations of particular
products resulting from gross contamination can be

-------
 worked with in the lab and specific products identi-
 fied. This may be complicated, however, by aging
 of the fuels in the ground and by mixing of differ-
 ent products. It appears that microbial breakdown
 and oxidation of some petroleum fuels does occur
 over long periods of time.
     Frequently a contaminated well is sampled in
 the field,  and the contaminant volatilizes before
 tests are run in the lab. It is very disturbing to the
 homeowner to receive a negative report in the mail
 from the  State  Department of Health or Water
 Resources when  he knows that the water is  con-
 taminated. This happens so often in Maryland that
 we now accept personal taste-odor confirmation by
 the field investigator at  the time of sampling. Any
 samples actually collected are taken in glass bottles
 with rubber stoppers. Later in the lab, a microliter
 syringe can be inserted  to draw off vapors for gas
 chromatograph analysis.
     Once a positive  result is obtained, the investi-
 gation  then  centers  on the  location of possible
 sources. Some cases  are.very obvious and require
 little knowledge of hydrology, while others seem
 unbelievably complex.  The ability of the investi-
 gator to understand (or  at least have a feeling for)
 ground-water  hydrology becomes very important.
 He must  first determine where  the contaminant
 originates  and secondly, eliminate  the  source(s).
 The following outline lists procedures that might
 be  followed  when investigating  petroleum con-
 tamination cases.

     PROCEDURES  FOR  INVESTIGATION
     1. Is  the  complaint still valid; can the investi-
gator taste: or smell contamination in  the com-
plainant's ground water; when did the  problem
start?
       a.  Does the complainant have any source of
petroleum stored on his property?
       b. Has any repair been made recently on his
water system  (possibility of contamination from
plumber's tools, etc.)?
       c. Has the complainant or anyone else re-
cently spilled petroleum products on his property?
     During rainy weather, a rise in the water table
may concentrate  more  petroleum product on the
water surface. Or,, heavy rains may  actually flush
more material downward  to the ground water.
Hence the relationship of rain to the contamination
problem should  be asked  of all  parties.  Another
major point is that ground water flows downgradi-
ent under water table  conditions. If a series of wells
on  a slight grade are all contaminated, it is most
likely that the source of contamination is upgrade.
     2. Go to all nearby houses, garages, etc.
       a. First check for the presence of petroleum
 in the" ground water, then  essentially repeat as
 above.
        b. Look for sources of petroleum products.
        c. Ask about and  look  for  evidence  of
 accidental spillage.
     3. Check  with all nearby (within several hun-
 dred yards) gasoline stations and repeat as above.
        a. Identify the full names, addresses and
 phone  numbers  of  owners  and  operators, and
 gasoline distributors for the station(s).
        b. Ask for information on storage tanks or
 pump leaks.
          (1) Identify the owner(s) or responsible
 parties for all buried tanks of any kind.
          (2) How long have the tanks been in and
 have they been checked for leaks recently?
          (3) Have the tanks  been filled recently?
 Does  this correlate  to  the  appearance of  con-
 taminant  in  the  complainant's  well?
        c. Look at the complete physical layout of
 the station.
          (1) Identify all drains and ditches, both
 inside and outside the station.
          (2) Look for places  where oil or gasoline
 is drained into the ground.
        d. Try  to  determine if any operating, pro-
 cedure contributes to the problem such as:
    .     , (1) Throwing waste oil or gasoline out on
 the ground.                            ,
          (2) Flushing petroleum products down
 drains or septic tanks.
         ^(3) Unnecessary, spillage of gasoline when
 filling trucks, filling cars,;etc.
          (4) Poor storage facilities for old  drums
 full of oil  or other petroleum fuels.
          (5) Possible spillage by  other  parties  at
 the station site.
        e.  Notify the owner or operator that all the
 tanks and lines should be checked for leaks. (We
 have observed that leaks from broken buried lines
 and  not from the tanks themselves have  caused
 many gasoline station problems in Maryland. The
 operator of the station will generally detect such a
 leak as a loss of gasoline,  or as water in his tanks.)
 Any operating procedures or physical characteristic
 of the station layout that might contribute  to the
 problem should.be changed, or.altered accordingly.
 In Maryland we issue a formal pollution complaint
 to force compliance with these recommendations.
     The two main aspects of the service  station
 investigation  .concern  the physical  setup  of the
gasoline station and the operating procedures used
at the station. The tanks and lines generally consti-
tute  the parent oil company's property, and are
their responsibility. Other physical aspects of the
station (i. e. drains, pits, storage areas, the building)
                                             59

-------
may or may not be owned by the parent company.
Any operational procedures are the responsibility
of the owner and/or operator of the station.
     After all the above have been considered and
no  leak or source is found, the investigator can
probably  conclude that the gasoline station is not
the source of contamination.
     4.  Recheck the  local  area, from bouse  to
house,  to church,  to  repair shop,  etc., and try to
find the source.
     5.  Request a more detailed  investigation if
this final check fails to uncover a source. Other
considerations,  such  as the  complexity  of the
situation might also -warrant a more detailed study.
     This procedure  may  seem unduly  complex,
but it is important to try  to  define the limits of
contamination.  Some people, particularly those
who are  old or very poor, may  be reluctant to
admit  that they have a contaminated well. They
fear that  the  health department  may  condemn
their water supply or that they will  become  in-
volved  in  a costly lawsuit. The field  investigator
must allay their fears and obtain accurate informa-
tion.
       DETAILED SITE INVESTIGATION
     In spite of these time consuming and detailed
procedures, a hydrogeologist will  frequently have
to  make  a field  inspection  to  determine  the
source(s) of contamination. In many cases, several
possible sources may be  identified,  although in
other  cases,  the initial overview may disclose no
possible sources. In one such instance in Maryland,
vandals apparently  dumped  oil  down a home-
owner's well. To  pinpoint an  exact source, how-
ever, sometimes a test drilling program is required.
For this purpose in  Maryland, we use an auger
drilling rig.
     The Groundwater  Management  Division  re-
ceives  the preliminary data and  conclusions  (or
questions) of the field investigator. Upon recom-
mendation from him, we then proceed  with a more
detailed  site inspection.  Initially we  attempt to
visually determine the ground-water flow at the
site. We may be able to quickly locate sources of
contamination or define the contaminated area. If
auger  drilling  is required,  we  drill shallow holes
through the weathered  rock  or sediments to the
water table. During  the drilling,  soil  samples  are
collected for analysis and inspected by the hydro-
geologist in the field. A plastic pipe or casing can be
inserted in the hole  to let the water  collect. The
water in  the pipe is sampled and analyzed. Eventu-
 ally, "positive" data points may  define a source,
 such as a buried tank. It is significant to note that
 the expense involved  in test drilling often precludes
 this technique as a routine  investigatory tool.
 60
     Once it is located, the source of contamina-
tion must be  stopped  or removed. Maryland law
allows  us to issue  a formal pollution order which
requires compliance with our findings and recom-
mendations. If the situation warrants it, we may
also advise the aggrieved party to seek legal counsel.

      OTHER CONTROLLING  FACTORS
     Since  the configuration  of the water table
generally reflects topography, most gasoline con-
tamination cases can  be  easily defined. In one
interesting case, however,  it appeared that a hard
clay layer in the weathered rock zone  prevented
downward movement  of gasoline from  a nearby
service station. Spilled gasoline probably moved on
the top of this hard red clay, and then was inter-
cepted by nearby wells  penetrating it.
     Disturbed ground may  be more permeable
than surrounding  native rock and sediments. Be-
cause of its higher permeability, it  may intercept
and channelize contaminants. Northeast of Balti-
more a buried  continental pipeline developed  a
severe  leak, but all of the gasoline remained in the
disturbed  fill around the line. Where the pipeline
crossed a small stream, however, the fuel  seeped
out and, of course, became obvious as it floated
downstream.
     In another case a  buried tile foundation drain
intercepted and concentrated sizable amounts  of
gasoline and  Number 2 diesel fuel. Both of these
fuels were present in  the ground at the building
site, probably the result of a gasoline station leak
at some time in the past. The flammable materials
and fumes in the sump pit created an explosion
hazard near the building, and only after modifying
the drain system considerably  was the problem
brought under control. The first concern in this
instance was to remove or lessen the hazard. Con-
currently, efforts  were made to  separate the gaso-
line from the  discharging water.  However, the dis-
charge  of gasoline to  the city  storm  drains was
avoided because it also created  a potential explo-
sion hazard. Pockets of fuel might have collected in
low spots in the storm drain system, and invited a
disaster.
     The integrity of the wells in a petroleum con-
taminated area also seems to have an influence  on
the occurrence  of well contamination.  A longer
length  of  casing  and  a good grouting job may
save  some homeowners  from  problems of this
sort. Of course, in  the Coastal Plain  a new well
could  be drilled to  another aquifer in an uncon-
taminated zone.
     Over a period of time pumping may remove
some  of the  contaminant. Although we have ad-
vised the use of this technique for  cases of minor
contamination, it has not definitely proven success-

-------
ful. An activated carbon water filtering system can
also remove some types of petroleum fuels at very
low concentrations. However, in Maryland we are
not aware  of any water  conditioning equipment
that  can  remove significant amounts of petroleum
contaminant.
     Occasionally the petroleum contaminant may
move out of the area or be sufficiently dispersed so
that  it can no  longer be  tasted in the water.  The
aggrieved parties have then been advised to use the
water. In general,  water that is only slightly con-
taminated by petroleum fuels is not toxic to hu-
mans.  By  the  time the  concentration reaches a
toxic level, no one  would be able to stand the
smell or be able to drink it (McKee and Wolf, 1963,
p. 230).
                CONCLUSIONS
     To  summarize, the impact of the problem on
ground-water management in Maryland is greater
than had been expected. Each case requires at least
one  day in the field  by  an investigator and fre-
quently by a hydrogeologist  as well. Figure 1 indi-
cates that in 1969-1970 over 60 cases were recorded
by county health departments and Maryland De-
partment of Water Resources personnel.  County
health department sanitarians also contacted many
of the people involved and collected water samples.
As a result, the  problem has consumed a large
number of man hours.
     Fortunately to date, all of the cases  we  have
handled  have been settled out of court. It is con-
ceivable, however, that  the investigating hydro-
geologist may eventually have to appear in court as
an expert  witness to discuss the findings of his
investigation. Hence  the  obligation to present  a
factual definition  of the hydrologic environment is
critical.  With regard to clean-up activities, it may
be (and often is)  impossible to  remove the con-
taminant from the ground  water.  This  presents
interesting, if not continuing legal implications.
      It  seems that the   more cases you handle
 successfully, the more you receive for action. The
 problem has not been given priority in the  past,
 but will be handled more directly in the future.
           SELECTED REFERENCES
Anon. 1970. Modern practices  in infrared spectroscopy,
      laboratory manual. Beckman Instruments, Inc., Fuller-
      ton, California.
 Levy, Eugene. 1969. Application of gas chromatography to
      the analysis of petroleum  fractions. General Course
      manual, gas chromatography. Hewlett-Packard Com-
      pany, Baltimore, Maryland.
 McKee, Jack E.  and Harold W. Wolf. 1963. Water quality
      criteria. California  Department of Water  Resources
      Publication No. 3—A.
 Yokes, Harold E. and Jonathan Edwards. 1968. Geography
      and geology of Maryland. Maryland Geological Survey
      Bulletin 19.
                 DISCUSSION
The following questions were answered by John R.
Matis after delivering his talk entitled "Petroleum
Contamination of Ground Water in Maryland."
Q.  In a group of service stations, what method do
you recommend to determine which of the gasoline
stations' tanks is leaking? In other words, how do
you differentiate one from a group of many?
A.  You have asked a significant question. There is
no  good way at the  present time, as the previous
speaker indicated, to  determine the gasoline from,
say a Mobil  Oil station  or a Texaco station or an
ESSO  station,  in a  typical contaminated water
supply. Additives are used to identify a particular
company's product. If the petroleum contaminant
occurs  in  the parts  per million range, and  the
additives occur at a  parts per million of that, the
problem is obvious. These additives would therefore
occur   at extremely  low  concentrations  in  con-
taminated  ground water. Hence it would be impos-
sible to detect them  and differentiate gasolines. As
far as locating a source, we would suggest a test
hole drilling program, and leak testing of all tanks
and lines.
Q. Undoubtedly, the use of various storage tanks
is increasing with the increase in auto population.
Is  there  any feasible  preventative  maintenance
program that  will enable the owner or regulatory
agency  to  predict   the  probable  onset  of  tank
leakage as  the tanks age?
 A. The idea of a preventative maintenance program
 deserves more attention by the petroleum compa-
 nies. There are many factors which determine why
 a tank leaks.  Some tanks, for example, have been
 dug up and  the tar coating on the outside is just as
 fresh as when it was buried. However,  it is becom-
 ing increasingly obvious—because of contamination
 and the dollar suits involved—that some oil distrib-
 utors are  going to Fiberglas lined tanks which do
 not deteriorate, and  I think that this is one  of the
 answers we  have at the present time. Perhaps some
 changes in the construction of buried lines would
 also be desirable.
 Q. Are your  efforts, in part, responsible for the
 popularity of Fiberglas TOL tanks and piping?
 A. I'm sure our efforts are not the  total effort
 involved.  Research people in the petroleum indus-
 try have  said  that they are limited because they
 operate under an economic  condition.  In  other
 words, changes in tank construction and develop-
 ment of new identification techniques cost money.
 These  changes cannot be made without sufficient
 documentation. So  our documentation probably
 helps their programs along.
                                              61

-------
        Bull  Session  2—Chemical Contamination of Ground Water
Session Chairman: Allen Agnew, Director, State of
  Washington Water  Research  Center, Pullman,
  Washington 99163,
Allen Agnew, Lead Bull:
     Gentlemen, \ve are going to operate with two
mikes  tonight. As you  know,  this is all  being
recorded,  so when you speak please bear this in
mind and identity yourself each time.  I am  Allen
Apnew.  not  Bill  Cawlev  who,  as  vou know, is
  u                                *
unable to be here. Yesterday afternoon,  Bob Crowe
filled in for him, and I am filling in for him tonight.
     You  remember our session consisted of four
speakers,  one who was unable  to speak, but had
presented an abstract earlier, for our use. In  dis-
cussing ground-water  pollution, we talked about
pesticides, gasoline, and petroleum  in general. We
had  some unanswered questions. I  would like to
open this discussion by asking Dale Wilder if he has
anything furthei  to say in elaboration  of what he
said  or  in response to  questions he received  this
afternoon.
Dale  Wilder,  Hydrologic Engineer, Los  Angeles
Department of Water  and Power:
     The  only  comment I had  was  that there had
been some  discussion  after  both my paper  and
John Matis's paper, in relation to whether or not
the petroleum products can be identified by labora-
tory analysis. It appears from comments of at least
two  or three individuals that there are some labora-
tories that feel they can  analyze petroleum prod-
ucts as to brand,  type and even possibly area that
they  were  produced  from,  depending upon the
type  of sample they are given and whether or not
they  have a good standard to compare with.
     Of course, there  was also one other comment
that was brought up when John and I were talking—
that is, we might be able to use radioactive tracers
to inject  in  underground storage' tank leaks  and
pipeline leaks to  monitor  them. The problem that
we could see with this particular technique  is the
public hazard and the  public pressure involved.
      lust as a way of comment, we  had a reservoir
in Owens Valley, California, that appeared  to be
leaking. We had  some drains about a quarter of a
mile away from the reservoir, and had been picking
up quite a bit of water flow in them. The Hydrol-
ogy  Section wanted to account for  that water and
they didn't want  to double-budget it, so they were
wondering how we could actually find out  if this
water was coming from the  reservoir or if  it was
coming through  a  basalt area  to the  east of the
reservoir.  We considered  running radioactive ma-
terial and finally gave up on  this particular ap-
62
proach, for the very same reason that I think the
same approach would be difficult in the business of
trying to find oil leakage.
    There is no build-up of population at all in this
area. The valley is owned jointly by several people,
but the  Los Angeles  Department of Water and
Power  owns most of the valley floor; they bought
it up several  years ago to obtain the water rights.
So, it is not a problem of who owns the land or a
problem  of people living there, but it is  a problem
because  the  water  eventually  gets down to Los
Angeles  and, even  though  the half-life may be
sufficient that the radioactivity would already be
dissipated before it  reached Lo^ Angeles, it was
feared  that the public in Los Angeles would be very
sensitive to this. We've tried dyes very  unsuccess-
fully.
Jay Lehr, Executive Director, NWWA, Columbus,
Ohio:
     I've been wondering—listening to John's paper
on gasoline contamination and yours on  petroleum
contamination—how serious  a  problem this is on a
nation-wide  scale. Are we just talking about more
of an odd-ball situation? I mean, John talks about
something like  60 cases  in the State  of  Maryland.
Well,   that's  not too  many,   and they affected
seemingly very few people.  Your situation in Los
Angeles, of course, was an enormous problem but,
again,  it was accident in the nature of an off-shore
oil spill. How really  significant is  this  kind  of
problem  in  ground-water  development, nation-
wide?  Also  I  can  bring in Lewallen's paper on
pesticides—he almost went so  far as to say that in
an  area where  the  pesticide  contamination was.
prevalent, it  wasn't that serious. I'm wondering if
things  are as bad as we thought they were.
Dale Wilder:
     In relation to a couple of questions that you
asked, I originally  felt that the gasoline spillage
was probably a rather  isolated case. But, in talking
to a few individuals, I  find that maybe  it is a little
more ' common  than I realized. As  far as  actual
numbers are concerned, I  can't give  you  a real
handle on this, but I have a request already from
one individual  (who  wasn't   able to  make  the
conference) for a copy of the paper; he feels that
he has the same problem in  Ohio. There are similar
problems mentioned throughout  the nation,  al-
though they are isolated cases, as far as I know.
     In  relation to  whether  or not this may  be
similar to the pesticide problem where it is not all
that critical, I think that this would depend quite a
bit on the individual case. In our case, it was a very

-------
great  problem. We did have to close down a well
field,  and  Forest Lawn has  had  to buy  water,
rather than use their own water. However, there is
some possibility, that after the initial flushing took
place, we may have suffered no damage other than
objectionable taste. That is,  the water may be con-
taminated,  but it may be possible to clear it  up
through the use of some sort of filtration system,
such as activated charcoal. Now, one of the reasons
I feel that WOGA (Western Oil and Gas Association)
did not  go  to  the  activated charcoal  in their
separator facilities  was basically  economics. The
estimates  were  that  it would  cost  somewhere
around  $90  per acre-foot  to  clean this up  by
activated charcoal. We can  buy  that  water  for
around $50-$60 per acre-foot so they felt it was
better just to replace that water.
    Now, this isn't getting into the safety aspect.
What  do you do  with the gasoline and water
mixture,  when from a purely economic standpoint
it may be less costly1 just to ignore that particular
water resource and try to use some other resource?
If we were  in  a really bad pinch,  like we may pos-
sibly get  in because of the recent California earth-
quake, and because'of other  losses of water supply,
then maybe this would be a much greater problem
than it is today. Right now, from a practical view-
point, we probably can work around it but, from a
standpoint of  contamination, I think  it is a serious
problem.
Dick Rhindress, Pennsylvania Department of  En-
vironmental Resources, Harrisburg, Pennsylvania:
    I, too, have a very similar problem to yours in
California, but mine is in Pennsylvania.  I  think I
should respond to a number of the questions that
have been asked  already. On identification of the
product as  free gasoline, we have a little different
situation  because we can just bail it up in a bailer.
We have  been  able to go right to  the brand of the
gasoline and, in fact, mix several grades and brands
of gasoline  together to get the precise combination
we  have  on the  ground. It was  slightly trial and
error, but on the gas chromatograph, infrared and
ultraviolet tests,  we can come up with precisely
the same  curve. It's fantastic  how close it is, in fact.
    In  cases  where we've  had  just extractable
amounts—something we've extracted from basically
a water sample, we have been able to say that this
is gasoline or  this is fuel oil. On  a couple  of oc-
casions we  have come close  to being able to say a
product name, but we haven't been able to really
pin down a specific product. We've been  able to
exclude some. For instance,  we've been able to say
in  one  case where there are two pipelines involved,
that it's not the Esso pipeline, but. it  is one of the
products that the Sun pipeline happens to be carry-
ing.  But,  exactly  which of the Sun  products  we
can't determine.
     On the subject of radioactive tracers, we have
tossed this around, as well. We've come up with the
same problem of public acceptance of radioactive
tracers.  .  .we just forgot the idea right there. We
have, however,  been thinking about using some of
the isotope  tracers which are not radioactive. We
gave this  thought to the oil companies—we have
two  pipelines and three companies involved in our
worst problem—but they refused absolutely saying,
"we don't want any contaminant in our products."
Now this, of  course, would be  in the  parts per
billion range, but it was a contaminant to their oil
and  they didn't want it because they had a market-
ing product  and they  said this would  foul up their
marketing if it ever got to the public. Because all
actions taken by our department are public, if we
recommended  and asked  for  this,  it  would  be
public knowledge that these oil companies  had
accepted it.
     Jay asked  about  the significance  nation-wide.
In Maryland there is a report of about 60  spills. I
can  put  my hands on  about nine major ones in
Pennsylvania, one of  which is to the point where
we've had explosions in two houses. Thank God, no
one has been hurt, yet. We've had people evacuated
from their homes because  of liquid,  free gasoline
that's flowing  into  their  cellars.  In  the latest
instance, the people were out for 4 or 5 months.
This was  because the gasoline pool was in a karst
aquifer in the Great Valley of Pennsylvania, and it
just rises up  every time we have a rainfall.
     I received  a call tonight that we  have  a hurri-
cane headed up the coast and I'm supposed to be
home in case something happens. That's where we
live.
     On  the charcoal filtration thought, we also
propose charcoal filtration in practically every case
around the State, where we have the individual gas
stations polluting an individual  well. Usually,  it
doesn't work. The economics are just too much at
the gas  station-distributor level. The  oil company
does  not  stand behind the continued use  of the
charcoal filter—they just won't go into a long-term
economic responsibility like that. The people who
own the home either drill a new well, try a charcoal
filter for a while, give up and then just start boiling
water to get  the hydrocarbons to volatilize, or bring
in water from a neighbor's for drinking and use the
oily  water for showers and so forth. Charcoal filters
just don't work because  of economics.
Mike Bell, North Carolina  State Board of  Health,
Greenville, North Carolina:
     I wanted to reply to Jay's question about the
seriousness of gasoline contamination. I don't know
                                             63

-------
about it  on a national scale. I had one experience
where a town of about 400 people had a well con-
taminated; we traced it down finally to a build-up
of cleaning fluid  and waste  oil that a garage had
been dumping into a ditch. It traveled some 225
feet  through the  ground and contaminated  the
town well. This town, if they  had not had some
back-up wells, would have been completely out of
water. As a result, they did have to rely on an old
well  which had not been  abandoned, but had  al-
most been discarded because of mineral  quality.
So, we would have a serious problem without these
back-up  wells and we were quite concerned about
the loss of this well. This is one instance which was
very  serious; I wouldn't be  surprised to  see this
happening in other places because  of the lack of
concern that's been shown so far.

Allen Agnew:
     That  is a good comment, and  it is one that I
had  hoped would  come up  tonight.  That is, the
deliberate  or careless washing down of the filling-
station areas into  the city  drainage  system  and,
where there's infiltration of recharge from this
system, we are getting pollutants into our ground
water. Where it isn't going into the ground water,
it's going into the surface water and so it is a prob-
lem  of the integrated hydrology of the two types
of water.
Ben  Wilmoth, EPA, Wheeling, West Virginia:
     During  the  past year,  I've talked  with State
sanitarians who have investigated  several fuel  oil
contamination problems in  consolidated rocks. In
some cases, the contamination traveled as much as
300  feet in sandstone or shale. Then there was one
of these fuel oil problems with a school, in which
the  fuel  oil tank probably contaminated  the well
at the  school—probably  caused  by overfilling. I
don't know how much was spilled, but apparently
it doesn't take too much to result in a problem.
Another case—a thousand gallons of gasoline leaked
from a corroded buried tank and contaminated a
school well  300 feet away;  this  was also in con-
solidated rock, sandstone. In the Ohio River Valley,
petroleum chemicals from a  tank-car siding con-
taminated a municipal water supply; they  were
able  to  alleviate this by  aeration, but not  com-
pletely eliminate it.

Verne Farmer,  Humble  Oil  and Refining Co.,
Houston, Texas:
     Several people here have observed specific
instances and suggestions  of instances of pollution
by oil and gas. For about the last year I've been
trying to make an inventory of many various types
of hydrocarbon  spills into the soils and ground-
water systems, but my efforts haven't really come
64
to much. Does anyone here know of any such in-
ventory, or systematic collection of incidents that
has been assembled—perhaps by the U.S.G.S. or by
a State agency?
     The  API  several months ago conducted  a
survey among its members, most of whom are the
middle size  and bigger  oil companies, asking for
case histories and  examples of  spills that these
people had experienced. They got some 110 replies
but  most  of these were the bigger  examples, the
more spectacular types of things, and in my opinion
don't even begin to  represent a true, systematic
inventory.  My  interest  in this,  really,  is  that
Humble—along with most other major companies-
is pretty sincerely interested in trying to do every-
thing we can to both avoid and clean up these spills
whenever  they do  occur.  Really what we need.
more than anything  else, are precedents—history
and  experience with a large variety of these things.
We have a lot of dope on the  big ones like Forest
Lawn, Mechanicsburg, and Savannah, but those
aren't typical. Ones where 8 or 10 or 15 individual
wells may get polluted, these are more common.
And most of these things don't get a lot of publici-
ty. It's a local thing and that's about the end of it.
If anyone has  any  thoughts  on approaches  or
methodology  for handling these things, I  would
certainly welcome them.
Allen Agnew:
     Mr. Farmer, in your survey, did you tackle the
organization  of State sanitary engineers, the State
Board of Health people? These are the people who
not only  have the  information within the  States,
but they get  together annually to compare notes
and complain about inadequacies in the system, so
I would think they would be a good source.

Verne Farmer:
     It could be.  This  really  wasn't  a systematic
survey in  the sense that I wrote to  representatives
in all 50 States asking them for cases of this kind.
This was  a hit-or-miss  sort of thing. I talked to
about everybody that I could find, various agencies
and various States. My  experience  with this sam-
pling—which was certainly not complete—was that
not very many people  knew anything about it at all.
And I just assumed that probably this was typical
of what you would get if you ran a 50-State survey,
which I confess I did not.
Allen Agnew:
     We have at least one  State agency man here,
maybe more—and I don't want to throw the heat
on him, but I think that maybe he is in a  spot to
respond to this.
 Dick Rhindress:
      I am in a spot to respond to this. Companies

-------
which we are working with on petroleum product
clean-up  do not even notify us of other spills and
ground-water contamination problems which they
are having elsewhere in the State.
     To the second part of the  point that was made
about contacting a State sanitary engineer, I'm not
sure  it would  do much  good.  The regional people
in our Health Department don't get the word to the
central office.  They try  to handle  it on their own.
They don't even call in  the  geologists that they
know they  have, to  help them with  it.  This is a
problem  we have and I'm sure other States have it.
The  regional sanitarian  who is a  county man, or
maybe even at the township level, just does not use
the facilities of his  professional  help at the top
level, and  the word  will never get to the  State
sanitary  engineer. There's a break in the  chain-of-
command in Pennsylvania—I know it exists in  other
States  as well. It's a nice idea to ask him,  but  I
know the man personally—I  work with  him—and
he just doesn't know about them except what I tell
him  on the way home from work—not through an
official channel.
Allen Agnew:
     It seems  to me that we  need to have  a  focus
for our  communication,  and what is the focus?
Well, to me it seems that it would have  to be the
man at the State level—the man who is responsible,
and  if there is a breakdown  in the internal com-
munication, we've got to rectify that.  Because until
that is done, we're going to be sitting on two sides
of the fence griping at  each other about each one
not doing his job. So I think we need  to do a lot of
homework  inside the State  among  the different
layers of government to  clean up this inadequate
communication pipeline.
 Elmer Jones, Agricultural Engineer, Department of
Agriculture, Beltsville, Maryland:
     This problem  is bigger than one can see, be-
 cause people can have petroleum  products in their
well  and be  totally unaware of it if the  proper
 conditions exist—if the  production zone is  at the
 lower part of the well and  the  pump  is  deeply
 submerged and there is no water entering the upper
 casing so that turbulent entrainment of the lighter
 petroleum fractions might carry them down  to the
 pump.
      In  pumping sediment from  the bottom of a
 well for pesticide analysis,  we  obtained a large
 amount of a jelly-like substance; the  owners of the
 well admitted to having a fuel oil leak several years
 prior  to this, about 120 feet from  the well.  But
 they very strongly maintained that at no time had
 they been aware of  any petroleum product  in the
 well by taste, odor, or by seeing it as  a thin film on
 standing water.
    I was concerned whether plastic pipe would be
a potential health hazard, in that something would
be  extractable from  the  petroleum—gasoline or
more  solvent  materials. Normal  drinking water
would not extract  such material  from the pipe,
but such  an  accidental exposure  should  be  con-
sidered in considering selection criteria for plastic
pipe.  Floyd Taylor, of  the Public Health Service,
serves  on one of the building research advisory
boards on plastic pipe. Floyd has,  in the Public
Health Service interstate-carrier drinking-water pro-
gram, investigated several gasoline spills; the feeling
of the PHS is that basically such a  spill does not
represent  a health  hazard  because normally it is
possible  to construct  wells  cased  and  grouted
through  the  surface  layer  so that  you can  still
obtain water of good quality. This is not to defend
the spill, but to indicate that where this is a serious
problem,  well  construction should be taken into
account.
     Now, you can turn it around and look at it the
other way—say, a fuel truck driver accidentally mis-
takes the upper casing terminal of the well for the
entrance to the fuel oil tank, and dumps fuel oil
down the well. Here  is a case where you would
really like to  have a fully constructed well that  is
not cased to below the water  table, because when
the well is cased  to below the water table and you
force fuel oil down it, the oil will  come out the
bottom of the casing, rise to the water surface, and
go on to some other point where it can continue to
be a problem. So, the point of injection can be
 cleaned up rapidly, but the well does not permit
 the removal of the oil at the point of injection.
 Dale Wilder:
     There  are a couple of things that I wanted to
 comment on. One was in relation to the problem of
 having spills that go  unrecorded  and repaired.
 Basically, my  comment is  that we need  to  en-
 courage corporate citizenship  — we must  make
 people understand that it is really important to be
 a good corporate citizen.  To  give  you  a small
 example, when I was working offshore, in the  old
 days they got a little bit calloused about what they
 did with some of their petroleum products during
 the testing  of some  of these wells and, if a barrel
 or two went overboard, they worried about it  but
 not enough to make them stop  operations because
 of the economics  involved  in getting out a barge
 with a tank on it. This isn't all the fault of tool
 pushers or of all the people in the field, but this did
 happen.
      After  the  environmental push  began and we
 were notified that such  things were to be stopped,
 the word went around that there  was to be no
                                             65

-------
dumping overboard of any empty barrels having a
little bit of crude oil left in them, or anything like
this. However, this didn't really solve the problem
until they told us that the tool pusher, any engi-
neer, etc. would be responsible and  subject to fine
and jail. Of course, I think you realize that if I was
out on the platform and saw a man about to dump
a barrel overboard, I  would run and get the tool
pusher and  tell him  that this man was about  to
dump some  oil. I  would do this because it means
something to me personally. I think that once the
company really emphasized it and this went down
the line, it would work. This is the thing we have to
do with the  State agency. We need to get the thing
started from the top and work it on down.
     One other comment in relation to radioactive
isotope tracing—1 wonder  if it might be possible to
somehow either heat up the petroleum product, or
cool it down, then  run a temperature survey.

Harry LeGrand, U. S. Geological Survey, Raleigh,
North Carolina:
     I  was impressed with the  contrast between
Dale Wilder's and John Matis's presentations. Dale
was concerned  with a fairly large, extensively used
aquifer, whereas John  was concerned with a number
of little problems in fairly impermeable situations—
perhaps you'd  still call  them aquifers. Maybe we
should analyze  this particular contrast. I think that
if a water-table aquifer  is fairly extensive, if it is
fairly permeable  and is used  extensively,  then
you  will have a single large problem, such as Dale
described. On the  other hand, the Maryland situa-
tion for the  most part was in the Piedmont Region
where  the permeability  is  quite low.  Wells  are
productive for  domestic  purposes and  also in the
valley and ridge area,  it's the same way. But there
are multiple  little problems  in those regions and
in  all probability the ground is more contaminated
than we think; we just  hear about specific prob-
lems. There's no question about  Dale's problem, it
will arise and we can see it. But, certainly, beneath
some of our big cities—Denver, Washington, and
others—the water-table system, whether we call it
an aquifer or not,  is certainly polluted from petro-
leum  products and gunk of all types  in many
different ways  and, in many of those places, we
don't even use the water anyway. I don't know
whether  we  should  say  that  we  dedicate  that
ground for uses other than drinking water, but I
think the number of problems we  see is  really a
question of the number of complaints we hear.
Allen Agnew:
     Do we  have any further thoughts  along this
line? I have another  question that  might take  us
away from this  theme.
66
Verne Farmer:
     I just want  to  make one observation about
this  effect of gasoline,  or  any hydrocarbon,  in
contaminating ground water and how far down it
will contaminate. As you people may know, there's
been really a considerable amount of research—lab-
type work and field work—done on this subject in
Europe, an awful lot more than has been done in
the United States. And,  actually,  a great  deal  of
this  stuff has never  been translated into English,
unfortunately. I've seen some translations of some
of these articles in the  last few months; most of it
is  in German. I don't  know what the mechanics
might be, but I think that it would be very useful
for translations to be made available to any associa-
tion  or  groups such as this  one,  that are vitally
concerned with ground  water.
     I think it has been demonstrated pretty con-
clusively  that one  of  the  biggest  problems  of
the light hydrocarbons, particularly gasoline, is the
solubility of a lot of the components. Generally,
the heavier  the product, the less solubles it con-
tains, but by the time you get up to the gasoline
range, it has  a lot of stuff that is soluble and, if
you  end  up  with  a  body  of gasoline  on the
surface it will float. However, over a relatively short
period of time, these solubles will disperse down-
ward and soon contaminate this whole system. And
that, in some of the  instances  I have encountered,
has  really been the greatest  problem.  We  know
where the surface of the  water is and, if money is
no object and the oil hasn't spread too far, we can
clean that off pretty well. But, if you have these
contaminants  dissolved and  diffused  completely
through  the whole system, there's no  way to get
that stuff out other than  to produce  the entire
body of water in area. Now, that type of thing is
quite easily  cleaned up  with  activated charcoal
filters, but, as Dick  said, sometimes getting those
things installed and getting somebody to maintain
them is  the biggest  problem.  However, it can  be
done—there  is no question  about that. But that
solution  of  certain components of gasoline is, in
my experience, one of the biggest problems of gaso-
line contamination that we have.

Dick Rhindress:
     I'd like to state at this time that the Pennsyl-
vania Department of Environmental Resources has
just  started  to plan a seminar. We hope to hold it
early next year, next March or April, for people
who have dealt specifically with the hydrocarbon
clean-up job. We want to get together the people
who have really worked in the field on clean-up of
gasoline and fuel oil  in the ground. We are not, at
this  point, considering  the prevention  aspects, but

-------
 are just talking about the clean-up aspects. I would
 like some input from anyone who is involved with
 this sort of thing, as to whether they feel such a
 program or seminar should involve also the pre-
 vention aspects.  This seminar will  be announced
 in journals like Ground Water. I'd be very happy to
 hear from  anyone who has this kind of problem or
 can give me some ideas on what we should discuss
 and how the program should be handled. The basic
 concept, though,  is to  keep it at a working man's
 level, not  a  formal paper and esoteric discussion
 level.

 Allen Agnew:
     We need  a lot more of that—a lot more of the
 hard exchange of  ideas  by working stiffs—those  of
 us who get the dirt under our fingernails. Gentle-
 men, we've been  bearing  down  pretty  hard on
 hydrocarbons  and, as you recognized,  we did have
 a little  bit of exposure to  pesticides  in the talks
 during the last two days. I wonder if we might have
 any reactions or any thoughts along that line.

 Max Lewallen, Agricultural Engineer. U. S. Depart-
 ment of Agriculture, Watkinsville, Georgia:
     I  don't  have anything in particular to say
 about  the  pesticide  problem. I am  available  to
 answer questions  from the floor if anyone has any-
 thing to add.
 Mike Bell :
     i do have one question, or it may be a series  of
 questions, dealing with the behavior or the move-
 ment of pesticides in a tight clay soil. I've heard
 different comments about the movement of DDT,
 the adsorption of DDT on clay, and the movement
of organophosphates. I'll ask Max  Lewallen  in
particular, since he was the speaker  on pesticides,
but I'd also like any information from anybody
else on the movement of pesticides in general, or on
a specific pesticide in a tight clay soil.
Max Lewallen:
     The literature is fairly full—quite a bit of work
has been done, there's quite a bit being done,  and
yet there's lots more to be done. I  think, as  you
pointed out in general, that this type of material
 is going to adhere tightly to the clay portion of the
soil.  One thing I might add—some of my associates
at work, who  have been involved for several years
with runoff from agricultural lands, laughingly say
 now that  runoff is a new  ball game because of
 past measurements  having been  made on  total
runoff quantities instead of this being fractionated
 in an attempt  to correlate the more active portions
of the soil, and the clay portion in particular, to the
 transport of these  materials. I don't  have anything
 in particular  to say about what you  asked. I do
 say  that in general these hydrocarbons will  be
 tightly bound to the clays. It's going to depend  by
 and large on what happens  to  the clay particle
 itself.
 Mike  Bell:
     Let  me just give you a specific instance  on
 which I was basing this question. As a result of a
 fire  of a storage area for both  fertilizers and  all
 types  of  agricultural chemicals, we  experienced a
 problem of what to do with both solid wastes and
 a liquid waste. And we ended up putting both the
 solid waste and the liquid waste  in a landfill. This
 landfill was above the water table and was enclosed
 in a natural clay area. What I was concerned with
 is  the  danger  to the ground water.  Should we  be
 concerned with movement in soils like this or are
 we pretty safe in what we did?
 Max Lewallen:
     I  think as far as the larger quantities of pesti-
 cides are  concerned, such as you refer to, this has
 not  been  described adequately in the literature. I
 know  of  several groups  that are interested in this
 aspect of pesticide disposal right now. Perhaps you
 should  talk with some  of the people like  John
 Matis here from Maryland who is wondering how  to
 get rid of many pounds and gallons of these source
 materials, along with lots  of other people. This is
 an  unanswered  question right  now, I  think.  In
 general, in a clay soil such as you refer to,  I still
 personally expect very little  movement  of  these
 materials—if we're talking about reasonable quanti-
 ties—and I don't know myself what I'm referring  to
 by "reasonable quantities." But as far as large  doses
are concerned, I don't know if anyone knows what
will eventually happen to these materials.
Dick Rhindress:
     I  can only cite one  instance of major disposal
of chlorinated  hydrocarbons—in the Province of
Ontario.  When Canada  outlawed them,  they col-
lected  all DDT and associated  things that they had
outlawed, and stuck them underground beneath a
large lagoon. The lagoon, to the best of my knowl-
edge is not impervious—rather, it drips out through
the bottom. These pesticides were not encased  in
anything special, and apparently these eventually
are going  to  flush  into the ground-water  flow
regime. The lagoon  was in glacial tills that they
felt were fairly tight. They didn't have a geologist
on the job—at least  I've been told  this by  some
Canadian  geologist friends of mine  who  are very
concerned about this situation. It sounds like one
of the worst possible things that can be done, ex-
cept that it  sits there in the middle of an industrial
complex  where nobody  is using ground water.
Boy, if the containers rot off and it all goes into the
                                             67

-------
ground water in one slug, it could really do a job of
pollution on whatever river  or stream is adjacent.
It sounds like the worst possible way of handling
them.
Max Lewallen:
     A point worth noting  here, I  believe, is the
term that is thrown around in the  literature—the
half-life of these materials. I don't believe that this
term is valid in most cases, but rather it is valid
only under certain prescribed conditions. When we
don't know what the conditions are, I think it's
almost ridiculous to  talk in terms  of half-life of
these materials—for  instance,  DDT. Some people
say it has a half-life of seven years or ten years or
more.

Allen Agnew:
     I'm certainly glad to  have somebody question
some of these numbers that we keep putting in the
books, in the guidelines,  in the rules and regula-
tions. You know, I  am a geologist  but I run into
very interesting textbooks on hydrology. One is by
Ray Kazmann at Louisiana State University, called
"Modern Hydrology." What he did was puncture
many of our hydrologic sacred cows, and he used
the term "dirty  data." He said that  our reports are
full of dirty data—data that have  been collected
over the years in various ways and we have come to
accept them as the Bible. Well, about a year ago the
U.S.G.S. went  through its old data and gathered
information on heavy  metals. Well, they  put a
disclaimer  in the very front of the report in the
very first paragraph. But how many people read
that disclaimer, or how many people would pay any
attention  to it? They would  look  down  through
the tables and  see  the statistics of these heavy
metal  concentrations, then  they would take off
from that point as they would begin to gather their
new data on mercury, or whatever else they happen
to be working on. I think we need to be particular-
ly careful of this sort of thing. Well, once  I've said
this, I  realize that we're talking to ourselves,  and
this group here probably doesn't need to be warned
of  this problem. We should make  sure that the
people who will use the information that we put
out realize the bounds that should be on it.

Dick Rhindress:
     Off the subject—but not  quite. In Pennsyl-
vania we  have had  several  incidents where drugs
have had to be disposed of in large  quantities. The
ships' stores for the whole east coast for the Navy
are  adjacent to  Harrisburg, and they have  tre-
mendous amounts of drugs which go out of date.
What to do with them? Some of them cannot be
burned;  some  of them are not soluble  so  they
cannot be run through a sewage treatment plant;
others will wipe out a treatment plant even in small
doses. What do you do with them? So far, we've put
some in a landfill, but only the ones that we  have
judiciously  felt would be decomposed in a landfill
environment. Some of the others we're still up in
the  air  about. They're sitting over  there  in the
Naval Depot waiting for somebody to say what to
do with them. The Navy itself is also working on
what to do with them. I haven't seen any data on
this  problem.  We've had railroad cars and  trucks
overturned  with drugs in them; we had a fire in a
major railway yard which destroyed, or partially
destroyed, some drug shipments; and we had to get
rid of this stuff. And  this stuff is probably as bad,
in its own way, as petroleum or pesticides. What do
you  do with it?

Allen Agnew:
     Thank  you. This leads  into my  question. I
believe in your paper you were talking about the
cost of  abatement,  and you  said something like—
the  cost of the abatement  wells and  other costs
make the cost of prevention  worth more than the
cost of clean-up. So what we need to do is push for
preventive programs. My question is—what kind of
preventive programs, and how do we do it?
John Matis, Maryland Department of Water Re-
sources, Annapolis, Maryland:
     I've participated in  a number of these inter-
agency  groups where we try to decide what we
would do in an emergency situation with hazardous
wastes—if you want to call them hazardous wastes.
That term covers a multitude of sins—drugs, wastes
from hospitals,  pesticides,  petroleum, etc.  You
really need to develop ahead of time, some type of
emergency  procedure for getting rid of these things.
The trend that I've seen is toward high-temperature
incineration as the answer to everything. However,
as we know, many of these waste products  have
heavy metals which are not going to high-tempera-
ture  incinerate,  so  you're   left with the heavy
metals.
     Perhaps legislation to reduce supplies is neces-
sary to  permit transport of the unused quantities
out  of the State, out of the area. This is another
approach; it's been suggested that you have legisla-
tion requiring yearly  inventories of whatever they
had  not used, whether it be drugs or pesticides or
some other hazardous waste product, and that their
return to the manufacturer be permitted.
     Now  as  far as your   specific   problem of
getting rid of these drugs, I would say that return to
the  manufacturer would  be one way of doing  it.
This is  one way we've gotten rid of pesticides in
Maryland—large  quantities  of pesticides, I might
68

-------
add, which have been condemned or made illegal
for usage, such as DDT.
     In  States like Massachusetts, they went on a
massive  collection program and collected tons and
tons and tons of DDT and other things, and they
really don't know what to do with them now. This
is  simply because they didn't have some type  of
procedure developed at  the time to handle this
problem. I think you really won't know until you
actually encounter the situation. Is this an evasive
answer?
Harry LeGrand:
     The question arises as to whether we should
have a special burial ground for these exotic wastes
that don't attenuate in  the  ground as easily as
would regular sewage. I'm thinking of pesticides
and many exotic chemicals. We know that many of
our sanitary landfills are not ideally located, as far
as the best hydrogeologic criteria are concerned,
and we  know that, if we have to, we could upgrade
a site location in each county to the point that we
could have a separate spot, you might say, to bury
this particular refuge. Would this be suitable?
Doug Olesen, Battelle-IMorthwest, Richland, Wash-
ington:
     Battelle's prime contract is with the Atomic
Energy  Commission at Hanford, and the AEC in
particular is  faced  with  the problem of the long-
term retention of hazardous  materials. Their phi-
losophy to date has been simple containment,  and
they're working on a number of long-term disposal
techniques.  But  they definitely have adopted a
philosophy  of national repositories, Hanford being
one, and there are  a number of others. The Kansas
 Salt Mines are also under consideration—also under
 attack—but this is one obvious alternative.
      Battelle-Northwest  is also looking at national
 disposal sites, and we're also working in the area of
 hazardous materials and information systems for
 EPA. Our work is more or less cataloging what is
 known on  hazardous materials and  producing an
 information-system format. EPA then is working
 toward having some sort  of a system on line so
 that when you get a  major hazardous  spill  in  a
 particular area where the man on the scene is not
 particularly  knowledgeable about what is spilled or
 what should be done about it, hopefully this system
 will be such that a simple telephone call will hook
 him into a computer system and he'll be able to get
 all of  the  specific facts, including a contingency
 plan and where the particular materials or equip-
 ment are located for handling the type of spill  that
 he has. No, it isn't liquid only. (Note: this is in
 answer to "Liquid  only?" from unknown in audi-
 ence.)  EPA's emphasis  obviously has been in the
water area,  because  they started as FWQA, and
Battelle's work has been with EPA's Water Quality
Office so far.

Jay Lehr:
     I'd like to comment on Harry LeGrand's com-
ment. Sometime ago, I espoused the idea that  some-
body could make money  by building  treatment
plants that could handle industrial wastes, and if
the government got tough enough on the industry,
the latter would be in a position where they would
have to ship their waste to some central treatment
plant that is really equipped to take care of it. I
think we really can take care of virtually any nor-
mal industrial waste.  It's  just  a matter of eco-
nomics.  Even  your  sulphur  liquors in  the  paper
industry can  be taken care  of,  but  it is an eco-
nomic problem.

Speaker not identified:
     The way  we're trying to do it now, in a shot-
gun approach, is ridiculous. We must look at  it on a
 national level. I think it  can be done if we can
 mobilize everybody.
 John Matis:
      A  couple of  comments on things that have
 been said. Virginia and Maryland jointly are looking
 at the  possibility of high-temperature incineration
 for handling of hazardous waste—anything that can
 be incinerated. This is an ongoing project. As far as
 Lehr's   remark about  industrial  people  making
 money out of this,  it is happening in Baltimore at
 the present time. Rob Tyler, Inc. is an example-
 their slogan  is "We Never Refuse Refuse." They
 have a site which is 80 acres or so where probably
 the most toxic materials you can ever imagine have
 been deposited. Well, this is really not acceptable,
 and they know this, so they  are trying to  find
 other  ways.  The  State  has forced some  of  the
 industries which  will bring in  cyanide waste, for
 example,  to  take  it to areas where  they could
 reprocess the cyanide.
      Secondly, waste oil is a big problem in Mary-
 land, as it is everywhere—45 million gallons a year,
 or something like  this, is not accounted for. We
 don't allow these in landfills for the obvious con-
 tamination  and hazard  contamination problems.
 This  outfit,   Rob Tyler,  Inc.,  has  built a high-
 temperature  incineration facility in Baltimore  and
 they can make money with  it. They  take these
 things and burn them up. Of course, they  comply
 with the air pollution controls,  but  it is still a
 profitable  thing at the present  time. If you have
 the facility, or if you have the capital to build one
 of these things, you can make money. Now, I think
 this will  probably be a short-term affair because
                                              69

-------
States and many of the industries realize that they
do need the regional approach—they do need these
facilities to handle all this multitude of sins called
hazardous wastes. So you  will see more of these
facilities being operated and, as I say, Virginia and
Maryland  are  talking about  a joint effort at the
present time.
Dick Rhindress:
     On the subject of hazardous wastes, we find
they really fall into two categories: one category is
typified by the instance I gave before, the Navy
depot  that  has  tremendous quantities of drugs,
those we can at least store for awhile until we come
up with a plan. When somebody brings in a truck of
X chemical  from Ohio on its way to New Jersey,
and it splits in half on the Turnpike and we have to
shovel  it up in an hour or  two and know what to
do with it so  that  we can get the Turnpike open
again, that's a  whole different ball of wax. Some of
these chemicals have been such that we couldn't
send men in  to  handle them because we  had to
have special uniforms and  some wastes were ab-
sorbable through the skin. One was a solid, but it
was so soluble in water that we had to get it out of
there almost  immediately  because it  was  raining
about ten miles west of it on  the pike and moving
toward us. These are  the types of things that are
real problems.
     You have to get it out of there, protect it, and
what do you do with it? We, sort of jokingly some-
times say, "Aha,  we got it out of here, we got it to
New Jersey"  or "back into Ohio."  But this is
begging the question. You  know, you can't always
just  send  it back to the manufacturer. That's beg-
ging the question too, because if the manufacturer
is in your State, you're going to just relocate  the
problem. If he's  out of State, you're  just passing
•:he problem on to somebody else. This plan that
EPA  and Battelle  are working  on,  of having a
national clearing house on what to do with these
chemicals is fantastic. This is what we must have,
and  we  must have instant  access to it, for the
emergency type  situation  which  happens  all too
often at  midnight-this is when they always hap-
pen; they  never  happen at 9:00 o'clock Monday
morning when you can work on them.
Mike Bell:
     Some things to be considered in developing a
regional approach are the economics and the condi-
tions that are necessary for transporting a  hazard-
ous  waste. How  far does it have  to go and  what
precautions need to be taken? We ran into this.
Elmer Jones:
     I would like to  return to  the  problem of
pesticides in  well water, and I use  this term in
70
place of ground water because in every case I have
Looked at carefully, I'm  personally convinced the
pesticides entered the well above the water  table.
Any rain that falls today is going to  contain some
DDT. Most of the water samples that we've ana-
lyzed have been below concentrations reported for
rain water. The pesticides—chlorinated hydrocar-
bons—are almost totally insoluble and very tightly
bound  to particulate matter. Our findings are that
they are  almost invariably associated with  gross
bacterial  contamination and variable turbidity with
rainfall. Neither of these characteristics are charac-
teristics of good well construction or good quality
ground water. And, as John  Mads pointed out in
his  talk,  the  initial approach  with the people who
have this problem is tremendously important.
    There is one  well  in Maryland  that had  a
$100,000 lawsuit  over  it  because  of  the  home
being treated for termites with Chlordane. We were
terribly afraid we'd end  up with this in court be-
cause  we had monitored this well for about three
years.  It was  cased  with open-jointed  clay-tile
casing  (the lowest point in the pit  adjacent to the
house was the top of the clay tile casing). The pit
received  drainage from a road and  off the house.
When the Health Department was first contacted
about an abnormal taste in the  water,  I feel this
problem  would have  been eliminated almost im-
mediately if the  undesirable  nature  of this well
construction  and location had been pointed out to
them, and they had been urged to have a properly
protected well  constructed.  As it was, awaiting
litigation, they were  told  by the lawyers to do
nothing  to the well,  so they ended up  carrying
water into the house for four years while awaiting a
solution in court.
     But the earth is a good filter and the question
with regard  to  sanitary  protection of wells is "At
what point does the natural filtering ability of the
earth  begin?" This  is something I was concerned
about  for quite some time—having  dug a lot of
fishing  worms  in  my  life—that pesticides were
always found in the upper soil  layers and it was
said they never moved down—and  yet we  know
that soil  insects go much deeper than  this. As  it
turns out, in the deeper soil  profile some of the
pesticides undergo  much more  rapid degradation
by action of organisms. Apparently this degradation
is sufficiently rapid so  that samples taken at the
level of  some of the  major insects show that the
degradation  rate is  as fast as the pesticides move
down to this level.
Allen Agnew:
     Elmer,  I think that you've put the finger on a
subject that the N.W.W.A. is  particularly happy to
see emphasized—that  is, improved construction of

-------
wells.  This  is  one  of the  cardinal points that
N.W.W.A. has stood for and, in working with the
State associations of water well drillers, this is one
of the points they've attempted to get across. You
recognize that the State associations unfortunately
don't represent all  the  drillers  in the State, or all
the well  diggers. So we've got a big job to do back
home and I  think that the N.W.W.A. will be pretty
happy to see the kind  of  emphasis that we are
putting on it at this session.
     Now,  we're approaching  our close—you re-
member  that when Lady Godiva came roaring down
the street naked on her horse, she said, "I approach
my  clothes."  She  spelled it differently from the
way I spelled it just now. It is  9:30—the time that
I had hoped would be our  close here, but we still
have some people who haven't talked and I'm sure
that  you've  got  some  things to say that  are so
pregnant that you just can't stand not to say them.

Dick  Pearl,  Colorado  State  Geological  Survey,
Denver, Colorado:
      I was just wondering if  any of you people
from  the  East  have  had any   contact   »vith
any research being done on degradation of DDT in
the soil  by any  means?  The reason I  ask  is that
research is being carried on up at Colorado State
University and,  if  my  memory serves me  right,
they're  finding that cow manure is a very useful
agent in this. It  destroys DDT  very rapidly in the
soil, and I was wondering  if there's anything else
that maybe you've found that does the job.
 Max Lewallen:
      Elmer  Jones  has  already  provided you with
 some of that information. There is quite a bit of
 that work being  done presently. One other thing I
 might mention—you  referred   to  cow  manure—I
 believe alfalfa meal has been in the literature also
 as one of the things that aid in degrading some of
 these materials. But, when we talk of the degrada-
 tion of these materials, it's a similar situation to the
 half-life that  we referred  to  awhile  ago.  You're
 talking  about  very  specific  conditions on the
 degradation of any of them which can be either a
 fast or slow degradation. For instance, I remember
 reading  over  a  report  where  in  one particular
 instance DDT was being used in a tropical climate
 for the spraying of mud huts and the problem there
 was too short lived, under those conditions.

 Verne Farmer:
      Is  DDT not all as bad as the newspapers say
 it is?
 Elmer Jones:
      I think this is a yes-or-no type answer. DDT at
 high pH can be  very rapidly broken  down  by
hydrolysis. This is  a condition you  will not find
in nature.  A large  number of very  common an-
aerobic organisms, E.  coli being one of them, can
break down DDT. The  problem with DDT in the
environment has been largely with what you might
call the "accidental fraction. "The pp DDT, and I'm
not going to attempt to define this because  I am
not an  organic chemist, is the desirable pesticide.
The op DDT is the accidental portion and depend-
ing upon  the manufacture and the date of manu-
facture, this will run from 5 to 20% of the active
agent. Unfortunately, in the  environment, this  is
so similar  to certain female hormones that it (not
the pp but the op) has had an effect on maturity
and sexual activity of wild life. The  thing that has
been  a real bugaboo, and I think that perhaps last
year we may have  seen a pesticide backlash  (now
this is a personal opinion, not an official Depart-
ment of Agriculture policy statement), because of
the publicity on  DDT, its persistence, a coming
ban,  and a lack of a good substitute in some cases—
perhaps some farmers have tried to put on enough
to last five or six years.
      It is impossible to describe normal conditions
 but perhaps normally 50% of the DDT applied  to
 an area will be lost to the atmosphere within 48 to
 72 hours. As you know, they talk about finding it
 in the Arctic for seals  and in places that they are
 sure  no pesticides were ever used. The English have
 done much more work on pesticides in the atmos-
 phere than we have. Their findings are that in the
 atmosphere, its  average life,  or half-life, is long
 enough for it to go around the world about three
 times. I talked about it in rain water because I'm
 concerned with  pesticides  in water, but  these
 Englishmen, Adkins  and  Eggleston, estimate that
 about  eight and a half times as much is removed
 from the atmosphere by deposition as by rainfall.
 So you  have  a  continual replenishment over the
 earth's surface  of DDT and, until you know this
 rate  of replenishment  on the surface, you cannot
 accurately calculate the survival of this in the soil.
 I think this has been a complicating factor, but in
 the last two years there has been more and more
 information on  the degradation of these, and the
 anaerobic environment which we don't find  in our
 agricultural  soils  is much more conducive to such
 degradation.
      A good example  of  this is when my project
 on pesticides  was started by Robert Egan of the
 Maryland State Health Department, hetrachloric
 peroxide was frequently found in water supplies in
 the Hagerstown Valley. When this was banned from
 use on alfalfa, even though it persists in the soil for
 a  long period  of time,  our detecting it in  water
 supplies  essentially cleared up within the year. The
                                              71

-------
question is not so much the persistence, but the
damage that  it can do in its lifetime. And this is
because of its mobility, especially with regard to
water; you cannot make DDT stay in pure water.
It  will be carried out by  a  process known as co-
distillation. The DDT has a very low volatility, but
it will be attached to  the water molecules going off.

John Mat is:
    I have a  question for Lewallen or  Elmer Jones.
If  DDT goes around the world three times before it
is  finally degraded and it comes down in the rain
water, it would seem to have some kind of impact
on photosynthesis in general, whether it be in the
oceans or in  lakes or standing water bodies. Is this
true, or not?

Stan Ursic, U.S. Forest Service,Oxford. Mississippi:
    We have had some people working on the hard
pesticides as applied to forest land, and they have
found that the persistence is very low, as measured
in surface water from measured water sheds.  They
don't detect  the hard pesticides after a matter of a
few days, due to the high biotic activity, especially
within the forest floor.
     If I may continue,  I  have a question  and a
comment.  It seems  that most people at this  Sym-
posium are geologists, hydrologists, and  engineers.
The one paper we didn't hear today was by a re-
search  chemist. For a neophyte  getting into the
field,  I'd like to hear comments on your experi-
ences  about what to do to get reliable chemical
analyses. I realize that you can go out and  hire  a
Ph.D.  chemist, but  perhaps this  isn't always the
best solution, unless you're a big outfit like  ARS.
What  are  your  experiences? Do you go to  State
agencies for  your analyses? Do you  go to private
firms?
     My other comment is, I  was a little  disap-
pointed in not hearing anything at all this evening
about  industrial wastes,  and this seems to  be  a
problem that is certainly as important as gasoline
spills.  For  example, I  know  that  International
Paper  Company,  between  1970 and  1974, will
plan  to spend $100 million on pollution control
alone.  This is just one paper company—it happens
 to be a big one—but there are others, quite a few.
 Are we neglecting this area? Are we skirting this
 area for any  particular reason?

 John Mat is:
      Industrial wastes include  a  large number of
 things, very complicated wastes. Those  of us who
 are  interested in petroleum know  that we have
 waste products. I'm familiar with waste oil. This is
 an industrial waste, and it is a very severe problem.
 I  do  know that many of the petroleum companies,
 72
at least in the Baltimore area, are spending very
large amounts of money on pollution control meas-
ures. There obviously have been occurrences in the
past which have created  ground-water problems—
these are  really what we're  concerned with. It's
well documented that many of these waste lagoons
have caused phenolic concentrations in the ground
water beneath. When you say "industrial wastes"
exactly what are you thinking of? It can include
many, many things.

Stan Ursic:
     Well, I was thinking of the whole gamut—the
paper  industry, the  people who make these hard
pesticides that  we're talking about.  You  are all
familiar with the Mississippi River kill, and this sort
of  thing.  I'm  speaking  of  industrial wastes in
general.

John Matis:
     One point which has become very obvious to
us  in Maryland; when you turn off all the spigots
and all the discharge points into the surface water,
the liquid waste has to go somewhere. I think Dr.
Zanoni made  this point—every waste has to go
somewhere. And at the present time, with the very
restrictive surface water laws,  most  of it is no longer
being  transported in water; it is going  on the
ground, and  this is really our concern, whether it is
pesticides, or washings from pesticide formulator
plants, or from petroleum plants.  It's going on the
ground. This is indeed a problem and, as far as I
know, in the northeast  corridor from Washington
up to Boston, with  which I am familiar, this prob-
lem has not been  evaluated.  It includes chromium,
steel plants, paint, and all kinds of chemical works.
It  has not been evaluated; people know it exists,
and it could very  well be the most severe problem
that we're facing in ground water.


Max Lewallen:
     I might contribute to his query concerning the
analysis of these  materials.  I frankly  know of no
easy route to analysis of pesticides. If we're talking
about  controlled  conditions, controlled   experi-
ments, known  quantities, known pesticides being
used,  it is a much  easier problem than  say, if we
 start with a completely unknown sample. As you
 might well know, we've had analytical problems; a
 lot of others in the  pesticide business with whom I
 have talked have analytical problems also. There
 are unresolved points. A lot of times in pesticide
 analysis if we stick  with common compounds and
 known compounds, it eliminates troubles  quite a
 bit. Others  might want to  make some comment
 with regard to pesticide analysis.

-------
DougOlesen:
     We had  a similar problem with Acrolein in a
study that we were doing. The first year's work was
based on analytical techniques that were supposed-
ly acceptable and, after a full season of sampling,
we discovered that it was a field problem. We dis-
covered that  the actual analytical techniques were
not successful and we had to  do  a fair amount of
basic research and go through and develop a com-
pletely new technique for Acrolein. I think a lot of
people face the  same problem. There's nothing to
draw on; you just have to go it on your own until
you  can develop something that is satisfactory. We
did document this and it is available now for other
people, but many pesticides you run up against are
likely  to  force  you to  go  through that  same
procedure.

Stan Ursic:
     Thank you.  I was  also  thinking of just the
routine analysis  of the common cations and anions
as well, rather than some of the more sophisticated
or exotic chemicals.

Allen Agnew:
     Mr. Ursic, it seems to me that EPA just recent-
ly reissued the second edition of analytical meth-
ods, or techniques—that big,  spiral-bound manual
that FWPCA put out 3  or 4 years ago, which con-
tains standard analytical methods and apparently a
re-evaluation of these methods. EPA has been doing
some soul-searching too, in  the last 3 or 4 years on
methodology, both at  the  Evansville Station and
the  Wheeling Station that I know of, on the Ohio
River. And the U. S. Geological Survey, of course,
has  Hem's Water-Supply Paper on Analytical Meth-
ods. So  there are two major  reference works, and
 I am sure that there are others; however, you still
have the problem of the bench chemist—getting
reproducibility of results, and the fact that some of
 them feel more  at home with a certain analytical
 method than with  others.  So you really have to
 identify the methodology and the  lab that's run-
 ning the analyses, as  you use the  information in
 your report.
 John Matis:
      I'd like to make a few  comments on analyzing
 petroleum. With the gas chromatograph  (I really
 didn't go into  this in  my  presentation), you  can
 analyze things in terms of the light fraction or a
 heavy fraction.  In other words, something which
 falls in the range of a C-4, C-6, C-8 compound and
 something which is up  around a C-12  compound.
 This  does not  identify  the specific  product. In
  other words, if  you have an unknown  petroleum
  contaminant in water,  you  cannot identify it in
minute quantities now. If you don't know what it is
(in other words, it might be a kerosene, gasoline, or
fuel oil), you are stuck because the kerosene, gaso-
line and this No.  2  diesel fuel (which is  really  a
kerosene)  fall within the same range of identifica-
tion by your gas chromatograph techniques. This is
a real problem. The infrared spectrophotometer
technique requires that you isolate this stuff with
some  type of organic solvent—chloroform has been
used.  A number of papers have  been  written on
this, many of them unpublished,  by this Kankawi
group in Europe, but more recently by  Crider with
Standard Oil of California.
     And  you  need a large volume of sample to
concentrate, let's say 0.1 ml;  I don't know if  this
came across in  my  talk. Now this assumes that
nothing volatilizes in the meantime, and that you
can keep  the solution; their technique  has been to
get so  many milliliters of chloroform or organic
solvent, collect  the water in the  same bottle,  and
then draw out the solvent and analyze this concen-
trate, boil it, reconcentrate the material and then
analyze it  with the gas chromatograph  infrared
spectrophotometer.
     There  is a real problem, though,  with identi-
fication and sampling. Now,  we have been faced
with the situation where we really never did resolve
a number of incidents where we suspected it was
kerosene  or gasoline, because of this problem. As a
result,  very little legal action could  be  taken. I
don't know if this clarifies our sampling techniques.
 Dale, you may  want to talk about your experience
 in California.

 Dale Wilder:
      A couple  of things  have come up that I did
 want to talk about. One deals with sampling tech-
 niques-we found that, among other things, we had
 a very difficult time in estimating how much gaso-
 line was  in the well, and also which wells had! the
 taste and odor of gasoline. Part  of the problem is
 that when you're out in the field  sampling 70 wells,
 it is a difficult procedure cleaning up your M-scope,
 or whatever  you're using, to measure the levels
 with. Of course,  much of this is just developing a
 standard practice of carrying detergents  or what-
 ever  it is that will  cut the gasoline and get rid  of
 the odor.
      The other thing, mentioned earlier, is the cost
 of preventive measures in comparison to the cost of
 the remedial action. I mentioned in answer to one
 of the questions that was asked that $700,000 had
 been  spent by WOGA. This is not an overly large
 amount of money, as far as the petroleum industry
 is concerned, except that this is just a  small part of
 the total clean-up that was involved.  There were
                                              73

-------
several agencies who went ahead on their own and
did what they thought they could. Also the Depart-
ment of Water and Power has done this to insure
that  their wells will remain free from gasoline. In
addition to this, WOGA requested that we do some
monitoring for them. But, the point is that in the
past  it  has tended  to be  quite an economic con-
sideration, but now  it has become more economical
to make sure that it doesn't get  spilled in the first
place. It was a little hard to justify going ahead and
^pending this money. I think that we need to put
forth a  real effort, as an  organized group, to make
sure  thac economics isn't the only consideration
that  companies look at when they  are deciding
about preventive maintenance.
Doug Olesen:
     The business of tying economics to environ-
mental  considerations has just  become official
policy because of the recent Calvert Cliff decision—
that  was a lawsuit by interveners  against the Calvert
Cliff nuclear  plant,  where the  AEC and  other
government agencies actually lost the case and were
accused of absolute neglect in terms of the National
Environmental Policy Act. One  of the things that
came out of this  decision is that all environmental
impact statements now—in essentially all industry-
will  have to have  an economic consideration tied to
the environmental impact statement, so that it will
actually be  an  economic  trade-off  for various
alternative actions that industry can pursue.

Elmer Jones:
     I want to add my comments on the analytical
services. This is a problem because of the expense
of the equipment involved. I think there is a real
need for more environmental monitoring. If we are
going to protect our environment, we need to know
what is going on. Because of the shortage of quali-
fied  personnel and the cost of the exotic equipment
needed to test  in the parts per billion and parts per
trillion  range,  consideration should be given to
regional analytical  services where  you can have  a
well-equipped and  well-staffed lab with  people of
high caliber  to train the next  generation of ana-
lytical chemists for  environmental monitoring.

Mike Bell:
     I apologize  because  the time is running late,
but  1 do have an important question that I think
needs consideration, going back to contamination
of ground  water by gasoline. What  happens when
the  gasoline gets  into the ground? How far can we
expect  it to go; what conditions in the ground
determine how far  it goes: what happens once it is
there as far as its movement; does it degrade; it is
just  there-, what can we expect once it is  in the
ground?
74
Dale Wilder:
     In  our experience, it appeared that the gaso-
line  was very  mobile.  However, because  of  the
residual that was left behind, it tended to dissipate
itself. That is, a certain percentage of it would be
left behind in the pores that it traveled through,
and so the distance it would travel would depend
very much  on  the volume  that was originally
present. As  far as what happens to it after it is in
residual form, this is part of the problem that we're
having.  It appears that certain organisms can de-
grade it. Whether it degrades on its  own, I don't
know.

John Matis:
     In the case near Los Angeles which I explained
in my talk, the gasoline had been sitting there for
at least three years, or maybe for as much as ten
years. It did not move out of the area. It sat there
as a pocket of gasoline and did not degrade appre-
ciably. In another case where I personally examined
the well—it was a small area  perhaps one acre in
size—the gasoline had been sitting there for at least
six years, and the people's wells were still contami-
nated with strong odor and strong taste. So there
was a definite problem. The gasoline did not move,
as you might have imagined and as Dale Wilder has
indicated; the movement was limited. I don't know
why this is true. I think the people in Europe  are
working on that particular problem.
Doug Olesen:
     We face a similar problem with radioactive
waste, where it doesn't move with the water. Some
things move at something like I/1,000th the rate of
the water movement because of the soil-waste inter-
action (adsorption, precipitation—there are a lot of
things going on). In addition to biological degrada-
tion, we are dealing with an organic material. But
it always comes back  to the fact that the key to
the answer  has to be that you have to be able to
satisfactorily describe  a velocity pattern and  the
flow of the water system. If you can do that, then
you can deal with the soil-waste reaction that is
going on and try to make some prediction that the
contaminant itself  will move at  some fraction of
the  speed  of the water. It's always going to go
where the water goes, but it may not go as fast, and
that is the  crux of the matter; but the key is know-
ing where the water goes.
     On one other observation, it seems everybody
is talking about contingency plans for dealing with
emergencies  of  contamination  or  pollution of
ground water. This could be an excellent recom-
mendation from our group—a contingency plan to
deal with emergencies, be it in Washington or North
Carolina or California or Maryland or wherever.

-------
Harry LeGrand:
    Well, we  have' contingency  plans  when it
comes to a snowfall hazard,  and we have piles ot
salt and sand on  the ro'ad. I  can visualize piles of
clay, or something like that to help mop up some
exotic  chemical,  but  I  hope we don't have cow
manure piles to take care of the pesticides, as some-
one suggested earlier.
Allen Agnew:
     I move we adjourn with that one. I want to
thank everyone, for1 contributing your thoughts and
pointing out  the failings and the fallacies that we
are living with. We mentioned earlier the term
"corporate responsibility." I would like to remind
us that we also need  to have  individual "citizen
responsibility"—for example, our own municipal
waste-treatment plants, when we taxpayers won't
vote the  bond issues  necessary to clean  up  our
wastes. We are some  of the worst polluters that
exist  around  here. Gentlemen, let's all put  the
wheel  to  the  shoulder — or  some  other  mixed
metaphor — and get on with the job. Thank you.
                                                                                                 /5

-------
Ground-Water   Pollution  Potential  of a  Landfill
                                                o
Above  the  Water Table
                    h                       c
by Michael A. Apgar  and Donald Langmuir

                   ABSTRACT
     A study  of the character and movement of landfill
leachate through unsaturated soil was begun in 1967 at the
State  College  (Pennsylvania) Regional  Sanitary Landfill
which  has  operated  since  1962  employing the trench
method of  waste disposal. The landfill occupies a gently
sloping underdrained valley with a  water table more than
200 feet below land surface. Precipitation averages about
37 inches as rain per year. Residual sandy-clay  to sandy-
loam soils range from a few feet to  greater than  70 feet in
thickness on a sandy dolomite  bedrock.  Soil moisture
samples were extracted at different depths from beneath
two of the refuse  cells  using suction lysimeters. Water
samples were also  bailed  from these cells and pumped
from a water table well beneath  the landfill. Monthly or
less frequent analyses performed on  water samples included
Eh, pH, temperature, specific conductance, BOD, Cl, SC>4,
total alkalinity, NH3, NO2, NO3, PO4, Ca, Mg, Na, K, and
total Fe. Soil samples from  beneath the refuse  were sub-
jected  to particle size and x-ray analysis, and chemical
analysis of  soil pH, soluble salt content, exchangeable Ca,
Mg, Na, and K, cation exchange capacity, and extractable
P content.
     The study showed that the quality and quantity of
leachate  beneath a  landfill  varies  considerably with the
topographic setting  of landfill trenches  or cells.  Leachates
2  feet  under an upslope  cell which received only direct
precipitation, had  the following  maximum  values 3-13
months after  refuse  burial: specific  conductance  8445
jumhos, Cl  1890 mg/1, BOD 3300 mg/1, NH3-N  540 mg/1,
and total Fe 225 mg/1. Upon reaching a depth of 14.5 feet
after about  2V4 years or more, maximum values of these
species in the  leachate had been reduced  by 83%, 80%,
> 99%, > 99%, and 98%  respectively. In  contrast,  more
water,  including precontaminated surface and subsurface
runoff from adjacent upslope cells, infiltrated a downslope
cell, saturating  the refuse. Even after moving downward in
the soil to a depth of 36 feet in 7 years, the leachate beneath
this cell had a conductance  of 6600 /imhos, 600 mg/1 Cl,
over 9000 mg/1 BOD, 40 mg/1 NH3-N,  and  100 mg/1 total
Fe. A practically continuous depletion of inorganic species
in  the refuse as indicated  by the quality of leachate from
both cells has occurred with time. However, concentrations
of BOD and redox sensitive species such as Fe and NH3 in
the leachate have fluctuated in response to  changes in the
moisture content and temperature of the refuse.
     Presented  at the National  Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
     "Research Assistant, Department of Geosciences, The
Pennsylvania State University, University Park, Pennsylvania
16802 (presently Hydrogeologist, with Roy F. Weston, Inc.,
West Chester, Pennsylvania).
     ""Associate Professor of Geochemistry, Department of
Geosciences, The Pennsylvania State University, University
Park, Pennsylvania 16802.
     Leachate beneath instrumented cells is moving down-
ward in the subsoil  at the rate of 6-11 ft/yr.  Observed
mechanisms of leachate renovation during this downward
percolation, along with supporting evidence for each listed
parenthetically, include:  dilution and dispersion  (decrease
in Cl with  depth); oxidation (Eh  and  pH measurements,
decrease in BOD and Fe with depth); chemical precipitation
(decrease in soil extractable P after leachate percolation);
cation  exchange (increase in percent  base saturation  of
clays affected by  leachate, and depletion of NH3 under
reducing conditions—bacterial growth may  also  retard  or
remove NH3); and anion exchange (decrease in  SO4  with
depth). Removal of bacteria,  suspended ferric oxyhydrox-
ides and other particulates by soil filtration undoubtedly
also occurs. Although renovation takes place, it is  incapable
of preventing highly contaminated leachate from moving to
depths  of 50 feet or more in soils beneath downslope cells.
Ground-water contamination has occurred,  probably by
leachate channelled down fractures in locally shallow bed-
rock, or by leachate-contaminated runoff from heavy storms
which has entered  sinkholes or infiltrated along the valley
bottom. The study shows that improper design of landfills
emplaced above the water table in relatively permeable
soils and bedrock such as are found in many carbonate-rock
terranes, can result  in serious ground-water pollution.

                 INTRODUCTION
     Ninety percent of all  municipal  solid wastes
generated  in the United States are disposed of on
land. About half this  total goes into landfills, the
rest to open dumps (Hershaft, 1969).  In the near
future as  public outcry  against open  dumps be-
comes more universal, most of  such disposal  will
also give way to landfilling. Thus landfilling will be
the prevalent method of solid waste disposal in this
country for many years to come.
     A sanitary landfill is defined as an engineering
method  of land  waste disposal which creates  "no
nuisances  or hazards  to  public health or safety"
(American Society of Civil Engineers, 1959). How-
ever, in  practice any  operation  in  which refuse is
compacted and covered with soil on at least a daily
basis is considered a sanitary landfill. Landfills are
indeed more sanitary  than open dumps in terms of
controlling odor, fires, and vermin, and in permit-
ting site reuse. However,  leachate  produced in the
decomposition of refuse in dumps or landfills may
pollute surface  and ground water. The nature of
such  pollution will depend  on  whether  the de-
composition process takes place in the presence of
atmospheric oxygen or soil air (aerobic) or in the
76

-------
absence  of oxygen  (anaerobic).  Aerobic  decay
produces stable end  products such as CO2, NO3,
SO4, H2O,  and a relatively inert residue. Anaerobic
decay produces CO2, methane, ammonia, hydrogen
gas, alcohols and organic acids, and other partially
oxidized organic species which exert a high BOD
(biochemical oxygen demand). Metals such as iron
and manganese are mobilized and  may be concen-
trated in such reducing environments.
     Refuse cells usually become predominantly
anaerobic shortly  after emplacement (Merz and
Stone, 1962).  In  general, conditions are relatively
aerobic near the surface of a refuse cell and become
increasingly anaerobic  towards the bottom.  How-
ever,  within  refuse  cells unsaturated  by  water,
traces of oxygen  are often  present, so that zones
and pockets of anaerobic and aerobic decay will
exist side by side at any depth (Merz and  Stone,
1966).
     Leachate  quality is influenced by the compo-
sition and amount of refuse, its sorting and  degree
of compaction, the  amount of water  in contact
with it, and the temperature. Refuse constituents
which affect leachate quality include biodegradable
organic   matter   (chiefly  vegetable and  animal
wastes),  soluble inorganic  materials,  and  redox-
sensitive substances such as  most metals. The com-
position  of a  typical municipal refuse is given in
Table 1. In general, the thicker the  refuse, the more
polluted the leachate issuing from it (Quasim and
Burchinal,  1970).  The manner of sorting  of  the
refuse  is  important insofar as reactive materials
such as the biodegradable organics in garbage may
be isolated by relatively inert materials such  as
paper. Paper also absorbs and holds water which
would otherwise promote  the  decomposition  of
other organic materials.
    Table 1. Sample Municipal Refuse Composition —
         U. S. East Coast (after Kaiser, 1967)
Weight
Physical Percent
Cardboard
Newspaper
Miscellaneous paper
Plastic film
Leather, molded
plastics, rubber
Garbage
Grass and dirt
Textiles
\TJf\nA
wood
Glass, ceramics, stones
Metallics
7
14
25
2

2
12
10
3
10
8
Rough Chemical
Moisture
Carbon
Hydrogen
Oxygen
Nitrogen
Sulfur
Glass, ceramics, etc.
Metals
Ash, other inerts


Weight
Percent
28.0
25.0
3.3
21.1
0.5
0.1
9.3
7.2
5.5
100.0

                   100
     Water cannot readily enter refuse  which has
been  highly  compacted.  Such refuse settles less
after  burial and is  less susceptible to channelling
and  to changes in  permeability  than  less dense
material.  Compaction also  reduces  the rate of
atmospheric O2 diffusion into the cell and thus
favors anaerobic decay of the refuse.
     Water content determines the rate of refuse
decomposition, and whether  the  process is  pre-
dominantly aerobic  or anaerobic. Dobson (1964)
reported that  the  maximum  oxygen  uptake by
microbial decomposition  occurred in refuse with
40-80% moisture content. The maximum decompo-
sition rate was observed at  a moisture  content of
60%.
     Water content of the  refuse is a function of its
initial moisture content, water produced in decom-
position reactions, ground water  which  may enter
the refuse, infiltration of precipitation  and runoff
through soil cover, and lateral flow in the soil. The
amount of recharge to the refuse cell is determined
by the  permeability of the cover material and the
topographic setting, among  other factors. Vegeta-
tion,  or leaf mulch  cover disposed of on the site
will  prevent rapid  storm runoff,  and  thus  may
increase infiltration  to the buried  refuse. Vegeta-
tion,  however, will also reduce the water content
of the refuse through transpiration.
     Most water which infiltrates a refuse cell will
not  appear as leachate until all  the  refuse layers
have  reached  field capacity (Remson,  Fungaroli,
and  Lawrence, 1968). However,  some  leachate is
produced immediately after refuse burial by com-
paction  of initially  wet refuse  or channelling of
water through the refuse (Lane, 1969).
     Reduction in the permeability of the cell floor
may  cause water to pond in refuse buried in un-
saturated soils. Such permeability  losses  result from
compaction and smearing of the soil by  excavation
and  landfilling equipment,  from  precipitation of
gelatinous ferric oxyhydroxides, or formation of an
organic slime on the trench floor (McGauhey et al.,
1966; Lane, 1969).
     Temperature  exerts an influence on leachate
quality by controlling the rate of bacterial activity.
Microorganisms are able to grow  in refuse over the
temperature range 5-55° C,  but the largest number
grow between 15-30° C (Dobson, 1964). A drop of
5°C   results  in a 10-20% decrease in the rate of
bacterially - controlled oxidation   (Camp,  1963).
Thus, significantly less  BOD is produced  in a landfill
in a  temperate climate during winter  and spring
months than during the summer and fall  (Langmuir,
Apgar, and Parizek, 1971).
     Burial of refuse in direct or  intermittent con-
tact   with  the  zone of saturation has  repeatedly
                                                                                                 77

-------
been  shown to  cause  serious contamination ot
ground  water.  Unfortunately, natural depressions
and excavations  which  intercept the  water table
are frequently  employed as waste disposal sites.
Well documented cases  of ground-water pollution
resulting from such disposal have been reported in
New York (Carpenter and Setter, 1940), California
(California  State Water  Pollution Control  Board.
3954),  South  Dakota  (Andersen and Dornbush,
1968), and Illinois (Hughes et a!., 1969). Saturation
of the refuse by ground water not only permits the
ready solution  of inorganic material but also favors
exclusively  anaerobic decomposition of the organic
wastes present with the  formation of reduced and
highly undesirable organic and inorganic substances.
     In  contrast,  reported  instances ot  ground-
water  pollution  from  landfills situated above  the
water  table  are  virtually nonexistent. Thus, until
recently it has been widely assumed that a tew ieet
separation of refuse from underlying ground water
is sufficient to  prevent water quality  problems
(California  State Water Pollution Control Board,
 1954; Andersen and Dornbush,  1968; New York
State Department  of Health, 1969). This may be
true in scmiarid climates where significant amounts
of leachate are seldom  produced above the water
table.  However,  landfills located in humid environ-
ments can  be expected to produce leachate which
may percolate to the water table without sufficient
renovation  to avoid ground-water pollution (Emrich
and Landon, 1969).
     Ground water in carbonate terranes is particu-
larly  vulnerable  to  pollution by landfill leachates.
Carbonate  rocks may weather to produce residual
 soils  the thickness of which depends on the abun-
 dance of noncarbonate minerals in the parent rock
 and the extent of  subsequent erosion. When these
 sons are present above  the water table, they might
 be expected to minimize ground-water pollution by
 dispersing  and  retarding leachate flow and facili-
 tating its aerobic renovation. However, carbonate
 bedrock  may  be locally shallow or  exposed and
 often has  solution enlarged joints  and  fractures
 ranging to cavernous in size.  Such bedrock provides
 direct avenues for movement of raw leachate to the
 ground  water with little if  any  opportunity  lor
 improvement in leachate quality. Further, solution
 enlarged openings  in carbonate bedrock can permit
 Icachate-contaminated ground water to travel great
 distances rapidly and without filtration.

           PURPOSES OF THIS STUDY
      An investigation of the ground-water pollution
 potential of the State College Pennsylvania Region-
 al Sanitary Landfill has been pursued since 1967
 by the Pennsylvania State University in cooperation
with  the  Pennsylvania  Department  of Environ-
mental Resources (formerly the Pennsylvania De-
partment  of Health). Purposes of the study which
will be considered in this report include:
     1. Examination of the quantity and quality of
leachate produced in a humid-temperate climate by
a landfill in unsaturated soils.
     2. Determination  of the  rate of stabilization
of such a landfill as evident from changes in the
quality of its leachates with time.
     3. Study  of changes in leachate quality with
depth  in  underlying soils to  identify the  extent
and principal mechanisms ot natural leachate reno-
vation.
     4. Evaluation of the possible extent of ground-
water pollution  by  landfill leachate at a site with a
deep water table in  a carbonate rock terrane.
     The  study  was begun by R. R. Pari/.ek and B.
Lane. Results of the research  for the period Octo-
ber 1967 to July  1968 have been summarized by
Lane  and Pari/.ek' (1968), Lane (1969), Emrich and
Landon (1969), and Parizek and Lane (1970).


      DESCRIPTION OF THE STUDY AREA
     The State College Regional Sanitary Landfill is
located  1.8  miles  northwest  of State College  in
Centre County,  Pennsylvania (Figure 1 ). Also shown
 in the Figure are areas in the State directly under-
 lain by carbonate rocks which as noted are uniquely
vulnerable to ground-water pollution. The study
 area averages 37 inches  of precipitation as rain per
 year, distributed with a slight excess during summer
 months.  Of this 37 inches about 5  inches fall as
 snow, generally between the months ot November
 and April. Mean monthly temperatures range from
 a low of-2;' C in January to 22" C in July.
 Fig. 1. Location of  the State  College  Regional  Sanitary
 Landfill (see also Figure 23!, and major areas underlain by
 carbonate rocks in Pennsylvania.
 78

-------
     The  Landfill site, the  southwestern half of
which is shown in Figure 2,  occupies 108 acres, or
most of a  small valley on an upland  ridge in the
center  of a broad, carbonate vallev called Nittany
Valley.  The bottom of the small valley is dry and
underdrained  and contains  surface liows  only a
few  times  a year after the heaviest storms. Land-
filling  is done along  the  valley sides, the tops of
which parallel the dry stream bed.  The valley walls
are about 60 feet high and slope 7-10'" toward the
stream channel.
      Access Roods

  — — — Property bounaofy
 Fig. 2. Map of the southwest portion of the State College
 Regional Sanitary Landfill, showing  locations of holes for
 the 1962 and 1967 lysimeter networks, the 1962 cell tap,
 and the  deep well.

      Ground water exists under water tank condi-
 tions at a depth of about 200 feet in the  bedrock
 beneath the dry stream valley. Bedrock is the Upper
 Sandy  Member of  the  Gatesburg  Formation of
 Cambrian age,  an interbedded series of dolomites,
 sandy  dolomites  and  quartzites.  Residual  soils
 developed on this unit range from 0  to greater than
 70  feet in thickness. Although bedrock does not
 outcrop  on the  landfill property,  it immediately
 underlies the buried refuse in some places.
      Local soil textures range from  sandy ioam to
 loamy sand (Lane,  1969).  However a  typical soil
 profile is very  heterogeneous and may in  addition
 include lenses  of sand or clay, large quartzite or
dolomite  boulders,  and  occasional  i.!r,\\e-;ther?d
ledges of quart/ite or dolomire. These units control
the rate  and direction of soii moisture movement'
beneath the landfill. Sandy units allow mp.u move-
ment and channelling. wh:lc clavey kn^cs .'nd rock
ledges retard or deflect soil-water mo\t.;-iei:t  -'Lane
and Parizek. 1968).
     Before  1962, the Borough of  '-.n<  College
used the southern end of the present i mdhll  tract
as an open dump. In 1962. all exposed n. rnse was
covered, and the site has subsequently operated ,"•
a sanitary landfill. Five townships si: r'-o'inding State
College have  also disposed  of  their refuse at the
landfill since January, 1967. !r  197^ about 100
tons of refuse were brought to the site c.icii day tor
burial.
     The landfill is operated by the ••••_:,ch mcthoc;
(American Society of Civil  Engineers. I 9;)°). Indi-
vidual refuse cells arc about 30 ieef \\ <:ie. 12  to 16
feet deep,  and range from 200  to 1200  :eet ir
length. Inert, bulky  material including demolition
debris and tree  stumps are dumped  ;n  separate
areas  of  the  landfill to conserve  trench  MXICC.
Refuse eel's are covered with a fiin' :.;• er ot com-
 thickness. Leaves garnered each fall arc spread over
 the completed cells  as mulch cover. Final  restora-
 tion involves planting evergreen seedlings through-
 out  the  area,  although  nauir.ii p':'^  and  shrur.
 vegetation predominate about 5 years alter  burial.

    WATER AND SOIL SAMPLING METHODS
 Water Sampling Methods
      Soil moisture samples for this study '-vere ex-
 tracted  by means of  suction iysimeters from be-
 neath landfill  cells  which  had  been emphced m
 1962 and  1967.  The design  and operation of suc-
 tion Iysimeters ha^e been described in detail by
 Lane and  Pari/.ek (1968), and Parizck and  Lane
 (1970). Ihe cross section of a lysimeter  is shown in
 Figure  3.  The device, which can  hold .1  liter ol
 water, consists of a  plastic cylinder 25  inches long
 and 2 inches in  diameter, with a  2.5  inches long
 porous  ceramic tip  or cup. So:i moisture is  dwn
 through the  powdered quartz and the  ceramic tip
 into the lysimeter by  the application of a vpcuum
 greater  than  the  soil moisture tension through the
 pressure-vacuum  tube  A sample is then forced to
 the surface for collection by  exerting a gas  pressure
 on the pressure-vacuum tube.  As noted by Lane
 (1969) the  pore diameter  of  the ceramic  tip  is
 approximately 1  micron or  less.  Thus, most col-
 loidal-sized particles can enter the sampler; how-
 ever, many   particular.es  including  bacteria  and
 perhaps some BOD  active  substances are excluded
 from sampling.
                                               79

-------
   PRESSURE-
   VACUUM
      GROUT
BACKFILLED
   SOIL
                              SAMPLE OUT
                                                      TRENCH WALL
                                  /4 NYLON TUBE
                                 RUBBER CORK
                                  PLASTIC BODY
                                  POWDERED
                                  QUARTZ
                               __ POROUS
                             J    CERAMIC TIP
Fig. 3. Cross section of a suction lysimeter, modified after
Parizek and Lane (1970),
                                                      26'
                                                                      PL&ST1C  LINER

                                                                   OVER  TRENCH BOTTOM
                                                                           PLAN
                                FLOW OF LEAGHATE
                            <	UNCONTAMINfiTED
                                SOIL WATER
                CROSS-SECTION
Fig. 4. Plan and cross  section of the 1967 cell leachate
collection and sampling installation (modified after  Lane
and Parizek, 1968).
     The 1967 cell lysimeter network and leachate
collection  facility were designed and  installed by
Lane and Parizek (1968) (see also Lane, 1969). A
generalized plan  and  profiles of the installation are
shown in Figures 4 and 5. Prior to emplacement of
refuse a plastic  liner of double thickness  six-mil
polyethylene was laid along the bottom and 6 feet
up the sides  of  the trench to collect and conduct
leachate towards the  lysimeter  network.  A gal-
vanized  iron  spreader pipe  directly  above  the
network received and distributed the leachate even-
ly  into  a gravel-filled  trench  from  whence  it
filtered  downward  through  the  soil towards  the
sampling points. A  cross section of the lysimeter
sampling network beneath the 1967 cell is shown in
Figure  5.  Pelletized bentonite  was placed at inter-
vals between lysimeters during backfilling of each
hole with native soil to  prevent leachate channelling
to the sampling points.  Sampling tubes from  the
lysimeters were completed in a protective wooden
housing set on a concrete foundation on the floor
of the  landfill trench  (Figure  4). After the refuse
had been  emplaced to a depth of 10  feet in No-
vember  1967, access  to  the sampling tubes was
accomplished by raising the  hinged  roof  of the
housing.
SL-4


D-..,
y_ 9.
J-.S-
1
•
'
1

-2*
- 4'
-
L




!1
'-lO.S'



- 3'
-10'
-13.5'
--17'
       K
       *- 01
 LYSIMETER

DEPTH BELOW  REFUSE
                                            I.
 Fig. 5. Cross  section of the  1967 cell  lysimeter network
 (after Lane and Parizek, 1968).
 80

-------
     A second lysimeter network was installed in
November  1969  below a  landfill  trench  filled
during 1962. Shown in Figure  6  are  the relative
positions and depths of lysimeters installed in the
six  boreholes drilled through the  refuse.  Six-inch
steel casing installed in each hole prevented water
perched in the refuse from moving down the open
holes.  Either bentonite  pellets  or polyester resin
(Rohm and Haas  P-463) was emplaced at  intervals
as grout during backfilling  of the holes with native
soil to isolate the  individual lysimeters. Thermistors
were  also  positioned   in  the  holes  to  measure
changes in temperature with time  and depth. Sam-
pling and  pressure-vacuum  tubes for the lysimeters
and the thermistor leads were completed just above
the land  surface  in the steel casings which were
covered with padlocked, hinged steel caps.
     Water  ponded in the  1962 cell refuse was
sampled from a sump  or tap (Figure 2) drilled into
the cell, but not through its floor.
     The deep  monitoring  well located in Figure 2
was drilled near  the center  of the landfill area in
November,  1969.  Sandy  dolomite  bedrock was
encountered at 51  feet and the  water table at 229
feet below the land surface. The six-inch well was
drilled by the air rotary method to a depth of 350
feet, and cased  to  a  depth of  185 feet. Ground-
water samples for chemical analysis were obtained
by bailing or with a submersible pump.
           -2 5'
                    -6"STEEL-
                     CASING
    -59'
                               -16'
                               -28 5'
                                    -2.5'
                                    -105'
                                    -205'
                         I— DEPTH OF LYSIMETER
                           BELO* REFUSE
  Fig. 6. Cross section  of the 1962 cell lysimeter network.
Soil Sampling Methods
     Samples  for  soil  chemical  analysis were
augered from 3 holes beneath  the  1967 cell. Holes
SL-3 and SL-6 (Figure 4) were augered  5 feet apart
in November  1967,  before refuse emplacement.
Hole  SL-13 (Figure 4) was drive cored between
these holes to a depth of 9 feet, 2 years after burial
of the refuse. Drive cores for soil chemical analysis
were also collected in October  1969 from holes SL-
10 and SL-12, spaced 52 feet apart along the 1962
landfill trench (Figure 6).

  WATER AND SOIL CHEMICAL ANALYSIS
Water Chemical Analysis
     Chemical  analyses  were  performed  on  soil
moisture samples extracted from the two lysimeter
networks, on leachate bailed from the 1962 cell tap
and  1967  sampling house floor,  and  on  ground-
water samples from the deep well. Lysimeter sam-
ples  were  collected for analysis approximately
every  month.  The other sources were sampled less
often.  Routine  analyses  were  made  in the field
upon  sample collection of temperature, pH and Eh
(oxidation  potential) or DO  (dissolved oxygen).
The  pH  was measured with  a  combination  pH-
reference electrode and either a Coleman Medallion,
or  Sargent-Welch Model  PBX pH-millivolt meter.
 Eh  was  measured  using  a   platinum indicator
 electrode  and  saturated KC1  calomel  reference
 electrode, with either a Beckman Model G, or Orion
 Model 407  pH-millivolt meter. Detailed methods of
 Eh  and  pH  calibration  and  measurement  are
 described  elsewhere (Langmuir,  1971a).  DO was
 measured with a Yellow Springs Model 51  portable
 oxygen meter. Other analyses  run  routinely at Penn
 State included a qualitative  description of color,
 odor,  and  turbidity, and  laboratory analyses of
 specific conductance with a Beckman  conductivity
 bridge, Model  RC-1982,  and  dipping  glass  con-
 ductivity cell (cell  constant  near 1 cm"1 ), and Cl by
 titration with standard AgNO3 solution (Rainwater
 and  Thatcher, 1960). Total titrable alkalinity (end-
 point  near pH  4.5) was measured with  standard
 H2 SO4 within  48 hours  on  samples which  were
 refrigerated upon  collection  and brought  to room
 temperature  just prior  to  analysis. Samples  were
 also analyzed approximately  monthly within a day
 or two of collection for total  Fe, BOD  (biochemical
 oxygen demand),  NH.,, and  SO4  by  the  Pennsyl-
 vania  Department  of  Environmental  Resources
  Laboratories in  Harrisburg, Pennsylvania.  Total Fe
 was  determined on acidified  samples  by atomic
 absorption spectrometry, BOD by 5-day incubation
 of  diluted  samples, NH,   by  the  nesslerization
  method, and  SO4  by the turbidimetric  method.
  Analyses run bimonthly or less often  at Penn State
                                               81

-------
included  NO3  with  a  specific  ion  electrode
(Langmuir and Jacobson, 1970), and Na, K, and Mg
by atomic absorption spectrometry on field acidi-
fied  samples.  Occasional analyses run by  the  De-
partment of  Environmental  Resources  included
NO2  (sulfanilic  acid method), NO3  (phenoldisul-
fonic acid method), and PO4 (stannous chloride
method). All  analytical  techniques employed by
the Department  of Environmental Resources are
described in Standard Methods (1971).

Soil Chemical Analysis
     Soil pH, cation exchange capacity,  and ex-
changeable cation analyses  were run on soil  sam-
ples  cored from  beneath both the  1962  and 1967
refuse cells  during installation  of the  lysimeter
networks. Soil pH was recorded as the pH of a 1:1
paste of soil and distilled water. Determination of
the exchangeable cations on a soil involves displace-
ment of all adsorbed species with a  concentrated
extracting solution and application of a correction
factor for those cations which entered  solution
from  soluble salts present in the  sample.  This
correction is particularly  difficult \vhen  the  soil
moisture in the  samples is as high and variable in
dissolved solids  as landfill leachate. The extracting
solution used in. this study was IN NH4Ac buffered
at pH 7. Correction for soluble salts in  contami-
nated samples was made by washing 30 gram soil
samples  prior to extraction with either 100 ml of
40%  ethanol, or 150 ml of 95% ethanol (Jackson,
 1964). The 40% erhanol wash was used to remove
Ca, Mg,  and  K salts from soil samples. Because this
wash  leached most of  the  exchangeable Na  from
the soil,  soils to be analyzed for exchangeable Na
were instead washed with the 95% ethanol solution.
      Cation exchange capacities were obtained by
summing the exchangeable  cations  and  the ex-
changeable acidity. Exchangeable acidity was deter-
mined  by measuring  the  pH change of a pH 7
calcium  acetate-paranitrophenol extracting solution
 (Woodruff, 1947).

    VARIATIONS IN LEACHATE QUALITY
 The Significance of Redox Measurements
 in Leachate
      Water saturation or isolation of refuse materi-
 als can prevent contact between atmospheric oxy-
 gen  or oxygen-rich fresh recharge  and biodegrada-
 ble organics in  the refuse. Under  such conditions
 dissolved oxygen in the water and soil moisture is
 rapidly depleted by aerobic bacteria,  and anaerobic
 decomposition then predominates. In the anaerobic
 environment  iron can exist in its highly soluble and
 mobile  divalent form.  Although  only  total  iron
 analyses are  available for  leachate samples in this
study, previous experience indicates that the iron
is present chiefly as ferrous iron and also in part as
suspended ferric oxyhydroxides (Langmuir, 1969).
     Measurements of DO, or Eh when DO values
are less than 0.1 mg/1, and of pH, provide a useful
means  of  explaining  the relative  and  absolute
amounts of reduced and  oxidized species present
in leachate. Eh measurements must be interpreted
with  caution when made in  waters as  chemically
complex as leachate, however. A measured Eh is
the potential at which no net current flows between
the platinum indicator electrode and the reference
electrode. The net current in solution represents
the sum of all anodic and cathodic currents present
during measurement (Langmuir, 197la). Thus,  for
example,  if  both  dissolved   Fe2 +  and  Mn2 +  are
present  along  with  their higher  oxides in  sus-
pension, the Eh will represent the sum of positive
and negative currents for both oxidation-reduction
pairs and may thus  differ significantly from  the
measured Eh in a solution containing only the iron
species. Such an Eh is called  a mixed potential. An
additional  cause of mixed potentials is the  fact
that such reactions as the reduction of SO4 to H2 S,
or  NO3 to NH3  among others are too slow to
produce a significant electrode  current, whereas
rates of oxidation of H2 S or NH3 are fast enough
to produce such a current. Thus, if relatively large
concentrations of different oxidized and reduced
species are present in a leachate, mixed potential
effects may give a  measured  Eh  which has no
meaning in terms of the theoretical stabilities of the
different species present.
      Species which  enter  into rapid oxidation or
reduction  reactions that  can  thus control an Eh
measurement are called electroactive. The stability
and reproducibility of an Eh  measurement depends
not only on the presence of electroactive species,
but also on  their concentrations. Such species are
generally absent from  oxygenated  surface waters
(except iron-rich acid mine waters) or aerobic sub-
surface  waters,  so that  a   DO  measurement  is
usually the best indication of the oxidation state of
such waters.
      Redox  conditions in landfill leachates which
contain several milligrams per liter BOD or more,
are  probably  controlled  in  most cases by  the
presence  of ferric oxyhydroxide minerals in  the
soil, and the activity of anaerobic bacteria. Bacterial
activity leads  to  partial  reduction  of the oxyhy-
droxides to Fe2 + . Redox conditions  then reflect
equilibrium  between Fe2* and remaining solid or
 suspended oxyhydroxides as defined by the reversi-
 ble and electroactive reaction
Fe(OH)3 + 3H*
=  Fe2
                                    3H2O
(1)
 82

-------
where  Fe(OH)3  denotes the ferric  oxyhydroxides.
Because  dissolved ferrous iron and  suspended or
solid terric  oxyhydroxides are abundant and  are
the predominant electroaetive  species in contact
with most anaerobic  landfill leachates, meaningful
Kh measurements should often  be  possible in such
waters.
     The  theoretical  F.h-pH boundaries  for equi-
librium between the  important iron and  nitrogen
species are  shown  in Figure 7. The dashed lines
limit the tields of predominance of N()_,, NO, , and
N'Hj species.  1  hese  lines show Kh-pH conditions
tor equal  concentrations of nitrate and nitrite, and
tor nitrite and  ammonia, respectively. The solid
lines in  the  Figure  denote  Eh-pH conditions  for
equilibrium  between  suspended or solid ferric oxy-
hydroxides and  dissolved ferrous iron at 0.06 mg/1
and 56 mg/1. Also plotted are Eh-pH  measurements
on  a  set of  leaehate samples  collected  July  22,
1969   from  beneath  the  1967 landfill cell. Two
distinci sample groupings are shown  in the Figure.
These  include a  group of analyses which plot along
Fe;'-Fe(OH), boundaries. Electroaetive constitu-
 Fig. 7. Hh-pH diagram at 25JC  showing fields of predomi-
 nance of NO ,, NO-,  and NHJ, and boundaries for equilibri-
 um between Fe:* and Fe(OH)3 when Fe2* = 0.06 mg/l and
 58  mg/l.  A solubility product of 10"37-' for Fe(OH)3  is
 assumed.  Also  plotted  are analyses of samples collected
 July 22, 1969 from beneath the 1967 cell. Values beside
 points are total iron concentrations in milligrams per liter.
 Uncontaminatod samples contain BOD < 2 mg/l; contami-
 nated BOD > 2 mg/l.
ents including iron species are practical!}' absent in
the first sample group, so that sample Eh values are
only qualitatively  meaningful.  These samples are
relatively aerobic, and may contain DO in excess ot
0.1 mg/l. Although small amounts ol iron (probably
as suspended ferric oxyhydroxides) are present in 5
ot these  8 samples, the}' are  otherwise practically
free of anaerobic species including ammonia, and
contain less than  2 mg/l  BOD.  Four samples from
the  1967  cell are anaerobic with total iron values
between  80  and 225  nig/1. Because ol  their hiiih
Fe'* and  ferric  oxyhydroxide content this sample
group  plots  parallel to  Fe2"-F'e(OH), boundaries
on the diagram. That measured  iron concentrations
are not exactly  consistent with the FeJ~- Fc(()11) ,
boundaries in figure 7  (or Figure  8) is  to be ex-
pected for these boundaries are  based on the activi-
ty  ot  Fe:*  ion,  not the  total iron concentration.
Also, a particular solubility of l-'e(C)H), is assumed,
whereas  its solubility will vary widely (Langmuir
and Whittemore,  1971). The  four contaminated
samples  in  Figure  7  contain  a  substantial BOD
content. The  Figure shows why the aqueous nitro-
gen  species likely to occur in  anaerobic leachates is
ammonia, and why there is no point in  analyzing
waters  high  in  iron  for nitrate  or nitrite.  The
samples plotted  in Figure 7 were collected  only 7
months after standing  water was  last observed  in
fill materials of the 1967 cell.
     Figure 8 is a similar plot of Eh-pH measure-
ments on samples collected a  year later (July 22-
23,  1970) from  both the 1967 and  1962 cells. At
this time only two of the 1967 cell samples contain
significant total  iron,  and  the  iron (and BOD)
values are general!}' less than  in samples collected u
year previous. Ten of the twelve  samples from the
1967 installation have F.h's above  390 mv. These
samples  are nov.  aerobic although most  contain
substantial  amounts of  inorganic  pollutants.  All
samples  from beneath  the  1962  cell plot  along
Fe2*-Fe(OH)., boundaries. In  general, these  sam-
ples are anaerobic and  contaminated by  high total
iron,  BOD,  and  ammonia,  concentrations. Thus,
Eh-pH  measurements  provide  insights  regarding
both the  state of  renovation of  the leaehate, and
the  possible  presence  of particular oxidized and
reduced species.
Time Variations in Leaehate Quality Beneath the
1967 Ceil
     Because  samples were not collected from the
1962 instrumented cell until April, 1970, the study
of long-term  changes in leaehate  quality  is limited
to such changes  beneath the 1967 cell. The quality
of leachates  beneath  this  cell has  been  greatly
influenced bv the introduction into  the refuse of
                                                                                                    83

-------
   400
   300-
Eh
(mv)
   200-
    IOO-
Fig. 8. Eh-pH diagram showing total iron values in samples
from beneath the 1962 and 1967 refuse cells on July 22-23,
1970. See explanation for Figure 7.

about  8 inches of artificial recharge, and  about 6
inches  of natural precipitation as  rain,  between
October 30, 1967, and April 28, 1968 (Lane, 1969).
The artificial  recharge  was  taken from  a public
water-supply  well of the Pennsylvania State Uni-
versity. This well is  also in the Gatesburg Forma-
tion, and  its water  probably had a composition
similar to  that  from well  UN-17 (see  Table 4).
This 14-inch  total constituted perhaps more than
double the amount  of water which might be ex-
pected to infiltrate this refuse cell naturally during
a similar 6-month period. As a consequence, water
stood  in the bottom of the sampling house (Figure
4) from May through December, 1968.
     Time plots at four different depths of changes
in specific conductance, Cl, BOD, NH3-N, and total
Fe,  are shown in Figures  9-13 respectively.  Data
used to construct the plots tor dates before August
1968 are from Lane (1969). The plots show that
leachate moved to the shallowest lysimeters before
flooding of the cell, and almost immediately after
refuse   emplacement on  October  30, 1967. This
partly   reflects  compaction  during  emplacement,
and the high (over 55%) initial water content of the
refuse   (Lane and Parizek,  1968),  but  may also
have been caused by polluted surface runoff which
flooded the open lysimeter holes shortly before
refuse  emplacement (Lane, 1969).
84
                                                     Fig. 9. Change  of specific conductance with time in soil
                                                     moisture from  2 and 4 feet depths (SL-3) and  8 and 14.5
                                                     feet depths (SL-6) beneath the 1967 cell.
                                                      m,/
                                                      Fig. 10. Change in chloride concentration with time in soil
                                                      moisture  from 2 and 4 feet depths (SL-3) and 8 and 14.5
                                                      feet depths (SL-6) beneath the 1967 cell.
                                                      Fig. 11. Change in BOD with time in soil moisture from 2
                                                      and 4 feet depths (SL-3) and Sand 14.5 feet depths (SL-6)
                                                      beneath the 1967 cell.

-------
                         1969           1970
Fig. 12. Change in NH3-N with time in soil moisture from 2
and 4 feet depths (SL-3) and 8 and 14.5 feet depths (SL-6)
beneath the 1967 cell.
Fig. 13. Change in total iron with time in soil moisture from
2 and 4 feet depths (SL-3) and  8 and 14.5 feet depths
(SL-6! beneath the 1967 cell.

     The  concentration-time plots show that maxi-
mum specific  conductance and Cl values arrived
simultaneously at each of the 4 lysimeter depths.
These maxima were 8445  /umhos and  1890 mg/1
respectively, at the 2 feet depth after 4 months, and
 1470 jumhos and 375 mg/1 respectively, at the 14.5
feet depth after  31-32 months. Thus, by  the time
the  leachate  reached  14.5  feet,  maximum con-
ductance and Cl  values had decreased by  83% and
 80% respectively. The plots  show  that maximum
 contaminant values are moving downward  at the
 rate of approximately 6 ft/yr.
      Chloride  and  the  majority of other species
 (chiefly inorganics) which  contribute to  the con-
 ductance, are readily  leached from the refuse  into
 the  subsoil  after  the  refuse  has attained field
 capacity.  Because Cl is  practically unaffected by
 reactions in the  soil,  its decrease with depth must
 be  caused by  dispersion and dilution.  Dilution is
favored by the design of the 1967 sampling installa-
tion which permits fresh recharge  to infiltrate the
soil  adjacent to the sampling house where  it can
percolate downward and dilute the leachate (Figure
4).
     The behavior of redox-sensitive species includ-
ing BOD,  total iron, and ammonia (Figures  11-13)
is  more complex  and less systematic with depth
than that  of specific conductance or chloride. This
is  partly  because  conditions in uhsaturated soils
can  be both aerobic  and anaerobic within a short
distance.  Initial  conditions  in the  soil at  first
sampling  in  November 1967 were aerobic  as evi-
denced by  maximum concentrations  of 11 mg/1
NO3-N, 1.0  mg/1 NO,-N,  and  negligible ammonia
and iron concentrations (Lane, 1969). A BOD value
of  2500 mg/1 arrived  at the 2  feet depth  after  4
months, and a maximum  BOD of 1600 mg/1 at  4
feet after 8-9 months. With the earlier arrival of
leachate containing a few milligrams per liter BOD,
anaerobic  conditions were initiated,  and bacterial
reduction of nitrate and nitrite to  ammonia ensued
by reactions such as
                                                    NO; + CH,O + 5H+ + 4e = NH4OH + HCO3
                                                                                                 (2)
 where CH2 O represents the organic matter in BOD.
 Reduction reactions  such  as  (2) raise  pH  and
 probably account for the maximal pH values of 6.5
 and  7.5-8.5  observed in  the simulated  landfill
 studies  of  Quasim  and  Burchinal (1970),  and
 Fungaroli and Steiner (1971), respectively. Maxi-
 mum pH's between 7.3 and 8.6 were also observed
 in this study near the time when  conditions in the
 soil were changing from aerobic to anaerobic. The
 relatively  lower  pH  maximum  of Quasim  and
 Burchinal reflects the absence of soils and other
 materials in  their purely organic refuse which could
 adsorb or otherwise react with H ions.
      Although  small  amounts of  ammonia  were
 formed by reduction of  initially present NO3 and
 NO2, ammonia  has chiefly been  produced by the
 decomposition  of nitrogenous organic substances
 in the  refuse and leachate.  Maximum ammonia
 nitrogen values of 100 mg/1 or more arrived at the
 2  and  4 feet depths simultaneously with the BOD
 maxima. Insufficient sample was available for anal-
 ysis  from the 2 feet depth after January 1970. The
 gradual increase in NH3 -N at 4 feet after October
 1968 may be caused by its production from the
 breakdown of organic nitrogen in  the leachate.
      Because leachate which  contains nitrate  or
 nitrite is not sufficiently reduced to contain ferrous
 iron from the solution and reduction of ferric oxy-
 hydroxides  in the  soil,  the buildup of total iron
 (chiefly Fe(II))  by leaching of the soil could not
                                               85

-------
occur until the onset ot more reducing conditions,
As a result, maximum total iron values ot 100 mg/l
or more do not reach the 2 and 4 feet depths until
about  1  month  after  the  arrival  ot  maximum
ammonia  values  at these  depths. The behavior of
iron and ammonia at S and  14,5 feet depths is less
easily explained, although  clearly a major reduction
in these species has occurred with depth. Maximum
total iron values at 2 and  14.5 feet depths were  225
and 5.6nit£''l respectively, and represent a reduction
of 97.5°o.  Maximum ammonia values at 2 and 14.5
feet were 54s i  ,md 1.8 mg/l— equivalent to a reduc-
tion of more than 99%.
     In their studies of synthetic landfills Quasim
and Burchina! {197sii and  Fungaroli and  Steiner
(197i'» observed  maximal iron concentrations  at
the lowest observed pH's. A maximum total iron of
850 msi/1  was detected at pH  5.4 in Quasim and
Burchinal's study. 1-ungaroli and Steiner measured
total iron maxima of  1 TOO-1800 mg/l at a pH of
about  5.5.  1-xamination of  the  data plotted  in
 Hiiures  7 and 8 shows  that in an  actual  landfill,
total iron concentrations  also depend on the oxida-
tion state ot the s\stem, and in fact that maximum
 rota! iron values do not  occur at the  lowest  pH's
 but .it intermediate values.
      Probabl)  most of the water entering the refuse
 and subsoii before 197u was from the equivalent of
 14 inches of rain added  to the refuse prior to May,
 1968, Hcavv  ranis m the spring of 1970 increased
 the moisture  content  of the refuse to well abcwe
 field  capaciiv for  the first time since 1968, and
 re-established  strongly  anaerobic conditions.  1 hat
 no  increase in  specific  conductance or  chloride
 resulted  rv.rn this  development  shows  that the
 concentrations of soluble inorganic species in the
 re ruse had  been  significantly depleted. However,
      ases in BOD, ammonia, and iron  during  1970
                   icleased from the refuse when
Differences Between  the  1962 and  1967 Sampling
Installations
     Before a useful comparison can  be made ot
changes in the  1962 and 1967 cell leachates with
rime "and  depth, a discussion of differences  in the
two cell  installations  and  settings  is  necessary.
Features of the 1967 sampling system have proba-
bly  accelerated the  oxidation  and  dilution  of
leachates  in the subsoil  with  the result that more
leachate  renovation  is  occurring  than could  be
expected  under more typical  conditions  in the
landfill. As shown in  Figure 4, all leachate produced
in  the cell percolates through a  trench  down into
the soil and to  the lysimcters below. Oxidation of
                                                    leachate is possible  in both the trench, which can
                                                    receive oxygen through the sampling house, and in
                                                    soils  beneath  the trench which are adjacent to un-
                                                    saturated soils under the trench liner. Also, oxygen
                                                    in  fresh percolation waters can flow through the
                                                    inter-trench space next to the  sampling shed. This
                                                    fresh  soil water will also have  a diluting effect on
                                                    leachate quality. Soil moisture at the lysimeters  is
                                                    thus a mixture of leachate from the cell and fresh
                                                    percolation waters. Also, because  of its  upslope
                                                    position  (Figure 2) the  1967  cell  is well drained
                                                    and  cannot  intercept  appreciable  quantities  of
                                                    leachate from adjacent cells.
                                                         In contrast the  1962 cell is  emplaced in  a
                                                    relatively  low-lying area where concentration  of
                                                    precontammated  surface  runoft  from  adjacent
                                                    slopes frequently  occurs,  and repeatedly water-
                                                    saturates  the fill.  This  increases  the  volume  of
                                                    leachate  production  and  isolates  the  soil trom
                                                    fresh  sources of oxygen so that anaerobic  condi-
                                                    tions are  established  both  in the fill and at depth
                                                    beneath  it. The poor quality  and large volume  ot
                                                    leachate is further aggravated by downslope move-
                                                    ment  of leachates from adjacent landfill cells along
                                                    the  cell floors. Downslope movement and concen-
                                                    tration of leachate may also  follow bedrock sur-
                                                    faces  beneath the  soil which  in this case roughly
                                                    parallel the topography.
                                                          The above conditions have led to relatively
                                                    larger amounts  of BOD  in leachate  beneath the
                                                       iOOOh
                                                                       1962
                                                                       CELL
                                                                       SAMPLES
                                                          4   .
                                                           !00
                                                                 200
                                                                         500    1000   200C     5000   CCOC
                                                                       SPEC:F!C CONDUCTANCE Umhos)
                                                        Fig. 14. Comparison of  the BOD and  specific conductance
                                                        of leachates from beneath the 1962 and 1967 refuse cells
                                                        on July  22-23, 1970. Solid circles denote uncontaminated
                                                        samples  (< 2 mg/l BOD); open symbols denote contami-
                                                        nated samples (> 2 mg/l BOD).

-------
1962 cell than under the 1967 cell. This difference
is evident from  Figure  14 which shows  the  con-
tribution of BOD  to the  specific conductance  of
leachates collected from beneath the two cells on
July 22-23, 1970.

Leachates in Refuse of the 1962 and  1967 Cells
    A comparison of the quality  of standing water
in refuse from the 1962 cell tap, and standing water
from the floor  of the 1967 cell sampling house is
given  in Table  2. The data  show  that soluble
inorganic species have been relative!}' depleted from
the 1962-dated refuse during the 8  years since  its
burial. Thus, Na, K, Ca, and Cl concentrations  were
89 to 66% lower  in waters standing in the 1962
refuse than in waters within the 1967 refuse at the
times  of sampling. This reduction  in  soluble  in-
organic  species has not been accompanied by a drop
in BOD values  which are practically equal in the
two leachates.

Leachate Quality  Versus Depth  Under the 1962
and 1967 Cells
     Plots  of  the  chemical  quality of  leachates
beneath the  two cells  on July  22-23,  1970 are
shown in Figures 1 5-20. Also shown in each Figure
is a  schematic  profile  with the  general slope-
settings of the two cells. Specific conductance and
chloride profiles  beneath both  cells display the
same  general trends. However, because chloride is

 Table 2. Comparison of the Quality of Leachates Standing
  in Refuse in the 1962 and 1967 Cells (The 1967 cell pit
   bottom sample analysis is from Lane (1969).  Specific
  conductance (jj) is in micromhos at 25 C, Other analyses
    are in milligrams per liter unless otherwise noted)
T (° C)
Eh (mv)
pH

Cl
S04
N03-N
N02-N
NH3-N
Alkalinity as HCO3
(pH 4.5)
P04
BOD
Ca
Mg
Na
K
Total Fe
_
_
_
6,280
657
-
0.0
0. 1 8
60
_

0.8
1,820
1,040
_
350
160
128
15.1
+ 82
6.49
2,290
225
50
0.0
—
18
502

0.0
1,880
186
37
39
51
578
among the species most readily leached through the
soil, chloride values exaggerate trends shown by the
conductance. The Figures show that concentrations
of chloride and total ionic species (as conductance)
released from the  1967 refuse  have just begun to
drop, but have been  decreasing for several  years
under the 1962 cell. The leachate front beneath the
1962 refuse has moved about 50 feet in 8 years, or
at  an  average rate of  6 ft/yr. Under  the  1967
installation the leachate front is moving at an aver-
age rate of 11 ft/yr, and has reached about the 30
feet depth after 2 years and 9 months. This may be
compared to the 6 ft/yr vertical movement rate of
maximum conductance  and chloride values under
the 1967  cell. In the absence of long-term records
for the 1962 cell  it is not possible to examine the
effects  of dilution or dispersion on conductance
and Cl  values with depth.  Data already described
indicates   that  such   processes  have produced  a
significant reduction  in  conductance and Cl values
with depth beneath the 1967 cell.
                                                      Fig. 15. Specific conductance in leachates beneath the 1962
                                                      and 1967 cells, July 22-23, 1970.
                                                                 I0     fOO
                                                                CHLORIDEimq/^l
                                                      Fig. 16. Chloride concentrations in  leachate beneath the
                                                      1962 and 1967 cells, July 22-23, 1970.
                                                                                                    87

-------
                                          !OO    iOOO
                                      BOD tmgft}
             BOO 
Fig. 17. BOD concentrations in leachate beneath the 1962
and 1967 cells, July 22-23, 1970.
Fig. 18. The Eh  of leachate beneath the 1962 and 1967
cells, July 22-23, 1970.
Fig. 19. NH3-N in  leachate beneath the  1962 and  1967
cells, July 22-23, 1970.

     BOD concentrations beneath  the 1962 trench
(Figure 17) roughly parallel conductance values and
show that  largely anaerobic  conditions exist  to
depths near  50 feet.  This  is because most  infil-
trating water  in the vicinity of this cell has been
contaminated by  contact  with refuse. As yet no
obvious  reduction  in  the BOD released  from the
refuse is  apparent.  However, the  most anaerobic
leachate  is not directly under the  cell  but at a
depth of 10-15 feet based on the occurrence at this
depth of minimum  Eh (Figure  18) and maximum
NHj-N (Figure 19) and total Fe values (Figure 20).
The presence of relatively more aerobic leachates
above 10-15 feet indicates that significant depletion
of  biodegradable  organic  materials in  the 1962-
dated refuse is occurring.
 Fig. 20. Total iron in leachate beneath the 1962 and 1967
 cells. July 22-23,1970.

     BOD values issuing from  the  1967 refuse cell
 have  not yet  significantly decreased.  Because  of
 relatively  oxidizing  conditions  under  this  cell,
 BOD's become oxidized and  depleted  10-15 feet
 above the deepest penetration of the leachate. The
 lowest Eh's  are  directly  beneath the  1967 cell
 refuse. At greater depths conditions become more
 oxidizing. A lowering in the capacity ot the 1967
 refuse to  release  reduced species  has  not yet oc-
 curred. This  is further  evident from the plots  of
 NHj-N and total  Fe  which showr maxima for these
 species directly below the  1967 cell.
     The occurrence of the ammonia maximum at
 8 feet and iron maximum at 12 feet below the 1962
 site is consistent with the Eh drop over this depth
 interval. Iron values  in excess  of 10 mg/1 and Eh's
 less than 200 mv are present to depths of 45 feet
 below the  1962 cell. Under these conditions am-
 monia is  the stable nitrogen species and  in fact  no
 nitrate or nitrite has been detected at these depths.
 However,  Figure 19  shows a rapid decrease in am-
 monia below 25 feet. Explanation for this decrease
 may  be  retardation of the  ammonia caused  by
88

-------
 bacterial assimilation,  or ammonia adsorption by
 clays in the subsoil. The decrease in iron with depth
 may also in part reflect  cation exchange on clays,
 but probably is caused chiefly by oxidation of the
 leachate with depth.
      Under the 1967 cell, ammonia concentrations
 are present  at greater depths than are total  iron
 values.  The  absence of  nitrate or nitrite beneath
 this cell indicates that elimination  of the ammonia
 is  also  either by bacterial assimilation or cation
 exchange. The rapid increase in Eh and drop in  iron
 under this cell is most consistent with removal of
 the iron by oxidation to ferric oxyhydroxides.

     ION EXCHANGE  ON SOILS  BENEATH
                THE  LANDFILL
     Soils beneath  the State College Regional Sani-
 tary Landfill have clay contents  ranging from  2 to
 41% by weight (Lane,  1969). The  ability of these
 soils to  renovate or alter leachate  quality in large
 part depends on their ion  exchange  behavior, which
 in turn depends on the mineralogy of the clay-sized
 soil fraction. In Gatesburg-type soils this is chiefly
 kaolinite (75-80%), illite  (5-10%), montmorillonite
 (5%), vermiculite (0-5%), and chlorite (0-5%) (Lane,
 1969),  and ferric oxyhydroxides which  comprise
 from about  0.2  to 1.0% of the total soil  weight
 (Matelski et  al.,  1963).
     The role of cation exchange in leachate inter-
 action with uncontaminated soils depends  on  dif-
 ferences in the absolute and relative concentrations
 of  cations in  the leachate and in uncontaminated
 soil moisture. A comparison of typical cation con-
 centrations in soil  moisture before  and  after con-
 tamination is given in Table 3.
     Chemical analyses were made of  uncontam-
 inated soils in cores SL-3  and SL-6 taken at depths
 from  3 to 15-17 feet beneath the 1967 refuse  cell
 (Figure 5). The analyses  showed pH values to in-
 crease  with  depth from   about  4.7 to 5.7. The
 concentration of total  exchangeable bases  ranged
 from  0.7 to  1.6 meq/100 g  of soil;  the  cation
 exchange capacity from 1.7 to  5.0 meq/100 g of
 soil. The order of adsorption of cations was gener-
 ally H ?> Ca =» Mg > Na > K at all depths. With  the

      Table 3. Quality of Soil Moisture  Collected
     from Suction Lysimeters Beneath the Landfill
    (Concentrations are in milli-equivalents per liter)
Cation
               Contaminated,  Contaminated,
                 beneath        beneath
Uncontaminated    1967 cell       1962 cell
Ca
Mg
Na
K
0.6 -1.0
0.5 -1.0
0.05-0.2
0.05-0.2
0.2- 3
0.2-10
10 -40
5 -10
10 -25
7 -20
4 -10
0.1- 2.5
 exception of H ions, this is the same general order
 of cation concentrations  in  uncontaminated  soil
 moisture.
      Cation  concentrations  in contaminated  soil
 moisture under the 1967 cell are  in the decreasing
 order Na > K  > Mg > Ca, and leachate pH's gener-
 ally range from 6.1 to  7.1. The order of adsorbed
 cations  on soils from leachate-contaminated core
 SL-13,  at  depths from  2-7 feet was generally H >
 Na > Ca > Mg > K, with  soil pH  values decreasing
 with  depth  from 7.2 to  4.6. Total exchangeable
 bases ranged from 1.3 to 2.8 meq/100 g; the cation
 exchange  capacity  from  1.7  to  4.2  meq/100 g.
 Significantly, the adsorbed Na content increased
 from about 0.2 meq/100 g in uncontaminated soil
 to 0.6-0.8 meq/100 g in leachate-affected soil. The
 Na fraction  of total exchangeable bases increased
 from 12-24% in uncontaminated soil to 29-46% in
 contaminated soil.
     High  soil  pH's  (8.2) and  high Ca content (2.5
 meq/100 g) of soils directly beneath the 1967 cell
 were  evidently caused by the  complete solution of
 the overlying concrete  floor of the sampling shed
 by leachate. This resulted in precipitation of CaCO3
 in soil above the 2 feet depth as evidenced by soil
 effervescence in  HC1, and  calculated saturation of
 soil moisture with respect of CaCO3  (calcite). A
 further effect has been reduction of Ca values in
 the leachate which  has increased  the relative pre-
 dominance of Na in the leachate resulting in higher
 Na adsorption by the soil.
     Because  refuse in  the   1962  cell has been
 relatively depleted in the  most soluble inorganic
 species after 8 years of leaching, the order of cation
 concentrations  in leachate under  the cell was Ca
 *  Mg >  Na >  K. The pH of leachates ranged from
 5.1 to  6.5.  The order of adsorbed  cations  on
 contaminated soils from  cores SL-10  (4-42 feet)
 and SL-12 (4-56  feet) was  generally H > Ca *> Mg
 >  Na >  K, with soil pH values from 4.8 to 6.7.
 Total exchangeable  bases ranged from  1.0 to  2.9
 meq/100 g, the cation exchange capacity from  1.2
 to 4.8 meq/100 g.
     Unfortunately, exchangeable base  measure-
 ments were  not  made of  adsorbed  ammonia and
 ferrous  iron, although  these  species  may be  as
 important  as the species measured on  some clays
 contaminated by landfill leachate (Toth and Ott,
 1970).
    Probably  the  most significant  result of the
 cation exchange study was  the observation that the
percent base saturation at a given pH is significantly
higher in leachate-contaminated soils than in uncon-
taminated soils  (Figure 21). This could be expected
in  that  the amount of exchange  depends on the
concentrations  of metal cations relative to H ion
                                                                                                 89

-------
   7O(-
 co G.O'
    r
            UNCONTAMINATED
                 SAMPLES
  5.5
   S.Or-
                            CONTAMINATED
                               SAMPLES
    0       20       "SO       60      80      120
                  PERCENT BASE SATURATION
 Fig. 21. The  percent base saturation of leachate-contami-
 nated, and uncontaminated soils beneath the landfill.
 concentrations, and the former  are more concen-
 trated  in leachate  than in  uncontaminated  soil
 moisture at comparable pH's.
      The  pattern of  Cl  and SO4  concentrations
 below  the 1962  cell  suggests that SO4  is being
 removed from the leachate with depth. Thus, at the
 2 feet depth, SO4 and Cl values  are about 2 and 3
 meq/1 respectively, whereas SO4  decreases system-
 atically  to about  0.5  meq/1 and Cl increases to 7
 meq/1 at a depth of 36 feet beneath the cell. Re-
 moval of SO4 is  probably  not by bacterial reduc-
 tion  in that redox conditions are  not  sufficiently
 anaerobic, but may be caused by anion exchange
 on clays.  Kaolinite has an  exchange  capacity of
 about 6 meq/100 g for SO4 at  pH 6  (Wiklander,
 1964).  Therefore  a  soil  with  20-30% kaolinite
 would have an adsorption capacity for  SO4 of 1.2-
 2.0  meq/100 g.  Sulfate  might be adsorbed at
 approximately this concentration when  anions such
 as PO4 which are more preferentially adsorbed than
 SO4  are not  present  in the leachate.  However,
 clogging of exchange sites by organic  anions, and
 by ferric and manganese oxyhydroxides probably
 reduces  the  number of sites available  for sulfate
 exchange.
     PHOSPHORUS VARIATION  IN SOILS
          BENEATH THE  LANDFILL
     In uncontaminated  soils, phosphorus is  ad-
sorbed on clays and present as calcium, magnesium,
ferric and  aluminum phosphates. Phosphate con-
centrations in the leachate were consistently below
the limits of detection (0.05 mg/1  at  PO4 ). Ex-
tractable phosphorus  was determined in soil sam-
ples collected from  beneath the  1962 and  1967
refuse  cells  with a solution  of 0.025  N HC1 and
0.03 N NH4F (Bray and Kurtz. 1945). This method
removes  P  in   exchangeable positions  on  clay,
readily acid-soluble forms of P  (chiefly Ca phos-
phates), and a portion of the iron and aluminum
phosphates.  As shown in Figure  22,  at each depth
more extractable phosphorus was  present before
than after leachate had percolated through the soil.
The reduction in extractable phosphorus is proba-
bly caused by its precipitation as relatively insolu-
ble ferric iron and aluminum phosphates.
                 SO    .75  0     25    .50     75
                 EXTRACTABLE PHOSPHORUS(pemi
Fig. 22. The extractable phosphorus content of soils be-
neath the 1962 and 1967 cells. Solid lines denote leachate
contaminated soils; dashed lines  denote uncontaminated
soils.
 GROUND-WATER POLLUTION BY LEACHATE
     Ground-water movement in ca";>on.iu tuT.uics
typically occurs along fractures, iomts, fault planes
and other permeable zones often enlarged by solu-
tion of the rock. The Upper Sandy Member of the
Gatesburg  Formation is an interbedded sequence
of carbonate-cemented quartz sandstones and dolo-
mites which has a greater average primary permea-
bility than other more  pure dolomite  aquifers in
the study area. Because  of this high relative perme-
ability,  a water table map of the area surrounding
the landfill site  (Figure 23) shows a ground-water
trough  in the Upper  Sandy Member. The trough
90

-------
coincides with a topographic high, as the quartz in
the Upper Sandy  Member  causes  it to be more
resistant to erosion than  more pure  adjacent  car-
bonate rocks. Water  table  contours  in  Figure 23
show  that the  direction of  ground-water  flow
beneath the landfill is probably northwest  into the
ground-water  trough.  The  exact   direction  of
ground-water flow may not be at right angles to the
water level contours in that permeabilities in car-
bonate rocks are highly  anisotrophic (Davis  and
DeWiest, 1966).
                              STATE COLLEGE
                              REGIONAL SANITARY
                              LANDFi Ll
Fig. 23. Water table elevation in feet above mean sea level
in  the vicinity of the State  College  Regional Sanitary
Landfill, June 3-7, 1971.
    A deep  water  table  well  was  drilled on  a
fracture trace in November, 1969, in  the center of
the landfill property (Figure 2). Fracture traces are
natural linear features mappable on  aerial  photo-
graphs  which represent the surface expressions of
near-vertical fracture zones along which permeabili-
ty is usually much higher than in the surrounding
rock  (Lattman  and  Parizek,  1964V Water from
such a  permeable zone should be more representa-
tive of the local  ground-water depth and quality
than  water  obtained  from  tighter  interfracture
trace areas.
    During the first year  of observation, ground-
water levels in this well (Figure 24) showed gradual
seasonal trends independent of individual precipita-
tion events. In this respect the well was typical of
other wells located  on Gatesburg Ridge along this
                                                     Fig. 24. Fluctuations in the ground-water leve! in the deep
                                                     well,  and precipitation  recorded at the State  College
                                                     Regional  Sanitary Landfill from November, 1969 to June,
                                                     1971.
section of the valley (Giddini>s, 1971). Thus,  tin-
gradual  rise in water  level  beginning  in March.
1970 presumably corresponds to ^round-uater re-
charge which entered the  subsurface in the fall of
1969 or  of a previous year.
     Since early November. 1970, however, a sharp
continuous rise of the water  table has occurred in
response  to heavy rains.  This  shift in  recharge
characteristics may have been caused  by the wash-
ing away of a fine-grained detrital plus which had
partially  sealed  a  sinkhole  immediately  upvalley
from most of the landfill area ("sink area"  in Hgure
2). This  now  enables runoff  from  Park  Forest
Village, a housing development  500 teet southwest
of the landfill  property, ro  move rapidly  into  the
subsurface at this point. The specific  conductance
of storm runoff from Park Forest Village alonti this
gulley  ranged from less than  100  mnho1- dunn«
summer  thunderstorms in   1970  to  nearly  600
jumhos following the November 1 5th rainfall.  Al-
though some contamination  in  the present landfill
well water ma}'  be from this  source, such  recharge
could  be preventing severe  contamination oS  the
ground water by landfill leachate.
     Another possible  source of pollution  in  the
well water is a sewage oxidation lagoon about 1000
feet from the well  (Figure 2) which operated trom
1967 to  1969, and  serviced about 1000 people.
     The quality of water  from the landfill well is
given in Table  4.  With specific conductances  be-
tween  800 and 995 micromhos, and a DO content
of 2.3  mg/1 or less, measured in 1970, this ground
water was more  polluted  than  any other  pumped
from nearly 200 wells in carbonate  rocks within
ten  miles  of the landfill  site (Langmuir,  1971b).
Conductances of waters from 4 other wells which

-------
   Table 4. Background Ground-Water Quality in the Upper Sandy Member of the Gatesburg Formation (Well UN-17),
     and the Quality of Ground Water from the Well at the Landfill Site (Specific conductance (id) is in micromhos
                  at 25°C.  Other analyses are in milligrams per liter unless otherwise indicated)

Collection date:

Appearance & odor:

T(°C)
DO (% sat'n)
Eh (mv)
PH
M
BOD
Cl
SO4
Alkalinity as HCO3
(pH 4.5)
NO3-N
NH3-N
Fe (total)
Ca
Mg
Na
K
L'.V-; 7
8-27-70

clear,
odorless
9.5
93

8.04
215

1.7
8
128

1.5


22
15
0.5
0.7

4-23-70

clear.
odorless




995

41

300

1.8


13
92
22
4.0
LANDFILL
4-30-70 5-27-70 8-13-70
slight
clear, clear, turbidity,
odorless odorless odorless
10.0
21

6.89
860 800 890
1.1
23 41
17 22
278 450 485

2.2 1.5

0.4
14
84
46
9.7
WELL
9-23-70
slight
turbidity.
odorless
9.3

268
6.53
913
0.8
32
30
449

0.9
0.5
0.7





12-31-70 5-29-71
clear, musty
clear, leachate
odorless odor
10.0
< 1
46
7.50 7.31
810 950
7.5
26 50
42
520 463

0.8
0.8
2.1




 also tap the Upper Sandy Member range from 182
 to 345 jumbos. The closest of these wells has the
 best quality water and is about  % mile from  the
 landfill. A comparison of the quality of water from
 the landfill well with that from an uncontaminated
 well (UN-17) in the Upper Sandy Member  repre-
 senting background quality, is presented in Table 4.
 The Table also shows  that gradual deterioration of
 the quality of water in the landfill well is occurring
 with  time. The  analysis in  May 1971 shows the
 ground water to be anaerobic and to contain sig-
 nificant amounts of BOD, total iron, and ammonia.
 Other contaminant species  present  in abnormal
 amounts are Cl, SO4 , and HCO3. This latest deteri-
 oration in  quality is  evidently caused by mixing
 with leachate from the landfill.
     Results of a 21-hour pump  test  of the 6-inch
 diameter  landfill well at 10  gpm on May  28-29,
 1971 indicate that the transmissivity of the under-
 lying  formation is about 4500 gpd/ft. This is
 consistent with transmissivities for the Upper Sandy
 Member measured  by  Siddiqui (1969) of 260-620
 gpd/ft in  2 nonfracture trace wells, 370-68,400
 gpd/ft in  10 wells on single fracture  traces, and
 2310-423,000 gpd/ft in  11 wells on fracture trace
 intersections. These data demonstrate the extreme
 anisotropy of flows in carbonate bedrock and the
 difficulty of predicting ground-water flow volumes
through a given rock section.
92
     The residual soils at  the landfill site are too
heterogeneous to permit meaningful physical calcu-
lations  of  rates of  leachate flow through them
under  saturated conditions. The best  method of
following leachate progress through the soil is from
the chemical data which indicate that the leachate
front is moving downward at  6-11 ft/yr, but has
not  reached bedrock beneath  the two  lysimeter
networks.
     Because ground water under  the  landfill has
evidently become polluted by leachates  from the
buried refuse, channelling of leachate  down  frac-
tures  in  shallow  bedrock  must be  occurring.
Leachate may  also reach  the  ground-water table
when contaminated  landfill runoff infiltrates the
sandy stream bottom or enters sinkholes in the sink
area (Figure 2) at times of heavy storms.

  CONCLUSIONS  AND RECOMMENDATIONS
     Results of this and previous  studies support
the following generalizations and suggestions:
     1. In a landfill in unsaturated soils, topography
will be the chief control on the amount  and charac-
ter of moisture entering the refuse. Refuse cells on
hill slopes will generally receive less infiltration and
less precontaminated moisture  from adjacent cells,
so that their leachates will have lower BOD, iron
and  ammonia  concentrations than leachates from
cells located near valley bottoms. Cells near valley

-------
bottoms are likely to produce highly contaminated
anaerobic  leachate  which  is not effectively reno-
vated during movement through 40 or more feet of
soil.
    2. Marked decreases in the amounts of leach-
ate contaminants can occur with movement of the
leachate in unsaturated soils on hill slopes. In study
area subsoils the decrease in Cl and a similar reduc-
tion in other species is  caused by dilution and dis-
persion. The PO4  decrease results  chiefly from its
precipitation as ferric and aluminum phosphates.
Oxidation causes a reduction  in BOD and  ferrous
iron. Ion  exchange on  soil colloids decreases  SO4
and metal cation values generally. The reduction in
ammonia  with depth is also  caused  by cation ex-
change, or by bacterial assimilation. Physical filtra-
tion by soil removes bacteria, ferric oxyhydroxides,
and other suspended substances from the leachate.
     3. Inorganic  salts progressively decrease in
concentration in leachate  as  they are depleted in
the refuse. However, BOD and associated reduced
species including ferrous iron  and ammonia, fluctu-
ate in amount in leachate with changes in the water
content and temperature of the refuse.
     4. Eh  measurements  in leachate   are  often
meaningful, for leachate is typically poised by the
presence of abundant ferrous iron and suspended
or soil ferric oxyhydroxides. Along with pH, such
measurements are useful in predicting the presence
and  mobility  of redox   sensitive  species  in the
leachate.
     5. In unsaturated soils leachate quality can be
improved if undisturbed  strips of soil several feet
wide  are  left between landfill  cells so  that  fresh
recharge  can  enter the subsurface and dilute and
oxidize the leachate.
     6. Improvement  in leachate  quality will also
occur  if concrete demolition wastes are placed  in
the  bottom of landfill trenches. These lime-rich
wastes  tend to raise  leachate pH and precipitate
leachate Ca as CaCO,. The pH increase leads to an
increased base  exchange  capacity of the  soil and
thus to increased  adsorption  by the soil of remain-
ing cations in the leachate.
     7. This study and that of Hughes et al. (1971)
indicates  that location of landfills above the water
table in  permeable soils  and  rock may result  in
 more serious ground-water pollution than deposi-
 tion of the same wastes in an impervious zone be-
 low the water table.
     8. Finally, location of sanitary landfills in car-
 bonate terranes above or below the water table is
 likely to  cause ground-water pollution. Such pollu-
tion can be minimized by proper landfill design, or
by  provision  for  collection and treatment of the
leachate.

             ACKNOWLEDGEMENT
     Construction  of the leachate collection and
sampling installation in the 1967 dated refuse cell,
and sampling of leachates  from this cell prior to
August, 1968, was under the direction of B. E. Lane
and R. R. Parizek. Leachate sampling and partial
analysis  was conducted by E. T. Shuster between
July, 1968, and January, 1969. E. T.  Shuster also
played  a key roie in the field work after  this last
date. Dr. Parizek assisted in contracting and design-
ing the water table well, and the leachate sampling
network under  the 1962-dated  cell, and in coordi-
nating the research. Some of the chemical  analyses
were performed by the  Pennsylvania  Department
of  Environmental  Resources. Thanks  are also due
to  Dr.  Dale  Baker of the Penn State Agronomy
Department for advice on the soil-chemistry phase
of  this  study.  The  research also  benefitted from
assistance  and  suggestions offered  by graduate
students Roger Jacobson, Todd Giddings,  Leonard
Konikow, Darrel  Leap,  Donald Whittemore,  and
William  Sanner. The program of research of which
this is a part, has been  funded since  1966 by the
Mineral  Conservation Section of the Pennsylvania
State  University,  jointly  since  1967 with  the
Pennsylvania Department  of  Environmental  Re-
sources.

                  REFERENCES
American Society of Civil Engineers, Sanitary Engineering
      Division. 1959. Sanitary landfill. Manual of Engineer-
      ing Practice No. 39, 61 pp.
 Andersen, J. R., and J. N. Dornbush. 1968. Quality changes
      of shallow ground water resulting from refuse disposal
      at a gravel pit. Report of  Project SRI 3553, Water
      Resources Institute, South Dakota State University,
      Brookings, South Dakota.
 Bray, R. H., and  L. T. Kurtz. 1945. Determination of total,
      organic, and  available forms of  P  in soils. Soil Sci.
      v. 59, pp. 39-45.
 California Water Pollution Control Board. 1954.  Report on
      the investigation of leaching a sanitary landfill. State
      Water Pollution Control Board Pub. 10, 92 pp.
 Camp, T.  R.  1963. Water and its impurities.  Reinhold
      Publishing Corp., New York, pp. 245-247.
 Carpenter,  L. V., and  L. R. Setter. 1940.  Some notes on
      sanitary landfills. Am. Jour. Public Health, v. 30, pp.
      385-393.
 Davis, S. N., and R. J. M.  DeWiest. 1966. Hydrogeology.
      John Wiley and Sons, Inc., New York, 463 pp.
 Dobson, A. S. 1964. Microbial decomposition investigations
      in sanitary landfills. Ph.D. Thesis, Department of Civil
      Engineering, West Virginia University, 71 pp.
 Emrich,  G.  H., and R. A. Landon. 1969.  Generation of
      leachate from landfills and its subsurface movement.
      Proceedings, Second  Annual Northeastern Regional
                                                93

-------
     Anti-Pollution Conf.  July  22-25,  1969, College of
     Civil Engineering, University of Rhode Island.
Fungaroli. A. A., and R. L. Steiner. 1971. Laboratory study
     of the behavior of a sanitary  landfill.  Jour.  Water
     Pollution Control Fed. v. 43, pp. 252-267.
Giddings,  M.  T. 1971. Hydrologic budget of the Spring
     Creek drainage  basin,  central  Pennsylvania. Ph. D.
     Thesis, Department of Geosciences, The Pennsylvania
     State University (in  preparation).
Hershaft,  A.  1969. Solid  waste treatment.  Science and
     Technology. June, pp. 34-42.
Hughes. G. M.,  R. A. Landon, and R. N. Farvolden. 1969.
     Hydrogeologic data from four landfills in  northeastern
     Hiinois.  Illinois  State  Geological Survey  Environ-
     mental Geology Note  26. 42 pp.
Hughes. G. M..  R. A. Landon, and R. N. Farvolden. 1971.
     Summary  of findings of solid  waste  disposal sites in
     northeastern  Illinois.  Illinois State Geological Survey
     Environmental Geology  Note 45, 25 pp.
Jackson, M.  L.  1964.  Soil chemical analysis. Prentice-Hall.
     Englewood Cliffs, X. J..  498 pp.
Kaiser. E.  R. 1967. Refuse reduction processes. In Proceed-
     ings, Surgeon General's Conf.  Solid Waste  Manage-
     ment for Metropolitan Washington. U. S. Public Health
     Service Pub.  No. 1729, U. S.  Govt.  Printing Office,
     Washington, D. C.
Lane, B. E. 1969.  Sanitary  landfill  leachate  interactions
     with ,i carbonate-rock derived soil in  central Pennsyl-
     vania.  M.   S.  Thesis,  Department of Geology-Geo-
     physics. Pennsylvania  State  University, 197 pp.
Lane. B. E., and R. R. Parizek.  1968. Leachate movement
     in the subsoil beneath a  sanitary landfill trench traced
     by means of suction  lysimeters. Proceedings of 2nd
     Mid-Atlantic Industrial Waste Conf., pp. 261-277.
Langmuir. D. 1969. Geochemistry of iron in a  coastal-plain
     ground water of the Camden, New Jersey, area. U. S.
     Geol. Survey Prof. Paper 650-C, pp. 224-235.
Langmuir, D. 197la.  Eh-pH  determination. Chapt.  26, in
     Procedures in Sedimentary Petrology, R. E.  Carver,
     ed.  John Wiley  and  Sons,  Inc.,  New York,  pp.
     597-634.
Langmuir, D.  1971b. The geochemistry of some carbonate
     ground waters in central Pennsylvania. Geochim. et.
     Cosmochim. Acta. (in press).
Langmuir, D.,  M. A.  Apgar, and  R.  R. Parizek.  1971.
     Seasonal variations  in the production and quality of
     landfill leachates (in preparation).
Langmuir.  D.,  and  R.  L. Jacobson.  1970.  Specific-ion
     electrode  determination of  nitrate  in  some fresh
     waters and sewage effluents. Environ. Sci. and Tech-
     nol.  v. 4. pp. 834-838.
Langmuir. D.. and D. O. Whittemore.  1971. Variations in
     the  stability  of precipitated ferric oxyhydroxides. In
                                                                Nonequilibrium  Systems  and Processes  in  Natural
                                                                Water Chemistry, a Symposium,  J.  D.  Hem, ed.,
                                                                Advances in Chem. Series (in press).
                                                          Lattman,  L.  H., and R.  R.  Parizek.  1964. Relationship
                                                                between fracture traces and the occurrence of ground
                                                                water in carbonate rocks. Jour. Hydrol. v. 2, pp. 73-91.
                                                          Matelski, R. P. et al. 1963. Soil characterization laboratory
                                                                data. Soil characterization laboratory, The Pennsyl-
                                                                vania State University.
                                                          McGauhey, P. H., R. B. Krone, and J. H. Winneberger. 1966.
                                                                Soil mantle as a wastewater treatment system. San.
                                                                Eng.  Res. Lab. Rept. 66-7, Univ. of Calif., Berkeley,
                                                                120.
                                                          Merz, R. C. and R. Stone. 1962. Landfill settlement rates.
                                                                Public  Works, v.  93 (9), pp.  103-106,  210,  212.
                                                          Merz, R. C. and R. Stone. 1966. Sanitary landfill behavior
                                                                in an aerobic environment.  Public Works, v. 97 (1),
                                                                pp. 67-70.
                                                          New  York  State  Department  of  Health. 1969. Sanitary
                                                                landfill: planning design and operation. Albany, New
                                                                York, 40 pp.
                                                          Parizek, R. R., and B. E. Lane. 1970. Soil-water sampling
                                                                using pan and deep pressure-vacuum Sysimeters. Jour.
                                                                Hydrol. v. 11, pp.  1-21.
                                                          Quasim.  S. R., and  J. C. Burchinal.  1970.  Leaching  of
                                                                pollutants from  refuse beds. Jour. San, Eng. Div., Am.
                                                                Soc. Civil Eng. v. 42, pp. 371-379.
                                                          Rainwater, F. H.,  and L.  L. Thatcher. 1960. Methods for
                                                                collection and analysis of water samples.  U.  S. Geol.
                                                                Survey  Water-Supply Paper 1454, p. 141.
                                                          Remson, L, A. A. Fungaroli, and A. W. Lawrence. 1968.
                                                                Water movement in an  unsaturated sanitary landfill.
                                                                Jour. San. Eng., Proc. American  Soc. Civil Engineers.
                                                                v. 94, no. SA2, pp. 307-317.
                                                          Siddiqui,  S.  H. 1969.  Hydrogeologic  factors influencing
                                                                well yields and aquifer hydraulic properties of folded
                                                                and faulted carbonate rocks in central Pennsylvania.
                                                                Ph.  D.  Thesis,  Department of Geology-Geophysics,
                                                                The Pennsylvania State University, 502 pp.
                                                          Standard  Methods for the  Examination  of  Water and
                                                                Wastewater.  1971. 13th ed., American Public Health
                                                                Assoc. New York, New York, 874 pp.
                                                          Toth, S. J., and A. Ott. 1970. Characterization of bottom
                                                                sediments: cation exchange capacity and exchange-
                                                                able  cation status. Environ. Sci. and Technol. v. 14,
                                                                pp. 935-939.
                                                          Woodruff,  C. M. 1947. Determination  of the exchangeable
                                                                hydrogen and lime requirement of the soil by means
                                                                of the glass electrode and a buffer solution.  Soil Sci.
                                                                Soc. Am. Proc. v. 12, pp. 141-142.
                                                          Wiklander, L. 1964. Cation and anion exchange phenomena.
                                                                in Chemistry of the Soil.  F. E. Bear, ed.; Am. Chem.
                                                                Soc. Monograph Series No. 160, Reinhold, New York,
                                                                pp. 163-205.
                                                  DISCUSSION
The following questions were answered by Michael
Apgar  or  Dr.  Donald  Langmuir,  after delivering
their talk entitled "Ground-Water Pollution Poten-
tial of a Landfill Above the Water Table."
Q.  Hon'  did von measure the free  v
the refuse cell?
                                              table hi
A.  (Apgar) We had a sump drilled into the refuse
of the 1967-dated  cell that didn't go through the
floor of  the cell. The water level within  this cell
was generally 2-5  feet above  the  floor of the cell
depending on  the   season,  so  that  some of the
refuse material was always saturated. The ground-
water table was encountered 230 feet beneath the
valley bottom.
 94

-------
Q. // the leackate could be collected, what treat-
ment process would you recommend for it?
A. (Apgar)  It would have.ro go through some kind
of oxidation procedure. Some of the objectionable
constituents including  high BOD's, high iron, and
high ammonia values can be removed by raising pH
and dissolved  oxygen  levels. There's a landfill in
southeastern Pennsylvania  that has  a permit to
operate because they  agreed  to collect and treat
their leachate. In some cases one should be able to
collect leachate and dispose of it  in  a municipal
sewer system.  Dissolved  solids and BOD  contents
are much higher in leachate than  in  sewage, but
leachate would  generally constitute a small fraction
of the  total volume  of waste water entering a
sewage treatment plant.

Q. H/'-y not use  COD rather  than  BOD  measure-
ments to indicate the extent of organic pollution';'
Do you co:::
-------
 and from the university, are from a local television
 tube manufacturer. Possibly wastes from this latter
 source are contributing the rare earths.

 Q.  When it rains and surface runoff flows in the
 valley bottom, is the quality of this runoff affected
 by the landfill?
 A.  (Apgar)  Yes, as a matter of fact storm runoff
 from the landfill includes water which has  been in
 contact with the refuse. As the runoff in the stream
 channel flows through the site during wet weather,
 its quality progressively worsens.  Some contamina-
 tion of water in the deep well may be  occurring by
 movement of the runoff  into sinkholes, although
 the runoff is less contaminated than ground water
 from the well.

 Q.  You say  the deep well is contaminated by the
 landfill only. Can you exclude possible contamina-
 tion by the housing development nearby?
 A. (Apgar)  We didn't want to  overcomplicate the
 picture  during  this  presentation. We have been
 limited by budget restrictions  to  drilling one well.
 Obviously it is impossible to know anything about
 the  horizontal movement  of contaminated  ground
 water when  you only have one hole in the ground.
 Yes, there is a housing  development  2000 feet
 away, and a  sewage  lagoon  about 1500 feet from
 the  well failed in 1969, although the lagoon hasn't
 operated since then. Based on water levels in other
 wells in the area at least a quarter or half mile from
 the  landfill well, the  general direction of ground-
 water flow is probably not from the  area  of the
 lagoon toward the landfill well. Also, the well water
 is continuing to deteriorate in quality.  We consider
 this as further evidence that pollution of the well is
 by landfill leachate and not sewage from the lagoon.
      It  would certainly be most useful if we could
 obtain support to drill two or preferably three more
 wells close to the landfill so as to more exactly
 establish  the character and  direction  of ground-
 water flow and extent of leachate contamination of
 ground  water.
 Q. Are  your  suction  samplers similar  to  those
 devised by Cole? If not, how do they differ? Also,
 is there evidence  of differential filtration through
 the porous ceramic cup of the lysimeter?
 A. (Apgar) The lysimeters we used were obtained
 from the Soil Moisture Equipment Corporation in
 Santa Barbara, California.  I don't know if they are
 similar to equipment used by Cole. We consider
 samples collected in the lysimeters representative
 of the  quality of  the moisture on the outside. We
 are dealing with levels of contamination that are so
 highly different from the quality of the original soil
 moisture, that  whether or not the sampling devices
 exclude or absorb small amounts of material is not
 going  to  make much  difference. Naturally, we
 cannot examine contamination by bacteria or other
 suspended material  in  the  leachate  because these
 materials  are  likely to  be filtered out by the
 porous cup. However, such materials would likely
 be filtered out by the soil anyway.
 Q. Can you give any information on  the probable
 behavior of microbial contaminants and viruses  in
 the leachate?
 A. (Apgar) I  can't  do that with confidence be-
 cause, as I just mentioned, such suspended materials
 are probably filtered  out in sampling. We  did look
 for total  coliform  bacteria in selected  leachate
 samples from the lysimeters, but the tests all proved
 negative.

 Q. Some of your graphs indicate a marked increase
 of pollutants with  depth in  soils below the refuse.
 How far above the ground-water table should the
 base of a landfill be  to prevent pollution? Could a
standard be set?
 A. (Apgar)  Such a distance is going to vary from
one site to another.  You can't say that 30 feet or
even  230  feet as in this case, will be  sufficient to
prevent ground-water contamination  by  landfill
leachate. The safe distance will depend on the local
climate, topographic and geologic setting,  and  the
amount and  character  of soils  and  of refuse
materials.
96

-------
Ground-Water Pollution  and Sanitary  Landfills-
A  Critical  Reviewa
by A. E. Zanoni
              :b
                  ABSTRACT
     The principal studies concerned  with the ground-
water pollution potential from sanitary landfills and dump
grounds have been conducted in California, South Dakota,
Illinois and England. These studies have all demonstrated
that leachates are highly pollutional in characteristics, but
once they pass  into  the  surrounding  soil regime, the
attenuation mechanisms of dilution, adsorption and micro-
bial degradation tend to reduce the impact of this loading
on the underground-water supply.
     A  survey  of practice in twenty-one States in the
United States regarding ground-water pollution from land-
fill  operation showed  that not much  new  research was
underway; there  was much variation in the code and laws
dealing with ground-water pollution-, and suggested dis-
tances from  landfill to  water wells varied from  50 to 1000
feet.
     Finally, based on  the literature  findings  plus the
result of the State survey, a set of recommendations are
offered  to  minimize  ground-water pollution  problems
stemming from landfill  operations.

                INTRODUCTION
     Ground water  can be polluted in  numerous
ways in spite of the  protective  mantle which nature
has  provided.  Liquid  pollutants can originate, for
example, from waste water  stabilization  ponds,
sludge lagoons, barnyard runoff, septic tank leach-
ing fields or seepage pits, pit  privies and the  deep
well disposal  of certain industrial  wastes or treat-
ment plant effluents. Pollutants can also originate
from the leachates  of decomposing solid wastes as
in the case of  open dumps, sanitary landfills, solid
waste composting sites, industrial refuse, and treat-
 ment plant sludges. In the first case the  pollutants
 are  already dissolved  or  conveyed by  the  liquid
 stream, whereas in the second  case, sufficient water
 must pass  through the decomposing mass to "leach-
 out"  the  pollutants  and  convey them  to the
 ground-water source.
      There is no doubt that a large amount of the
 solid wastes generated in  the  coming years, or the
 residues and  by-products of solid waste treatment
 methods presently  known  or  to be developed, will
      Presented  at rhe National Ground Water Quality
 Symposium, Denver, Colorado, August 25-27, 1971.
      b Associate  Professor of Civil Engineering, Marquette
 University, Milwaukee, Wisconsin 53233.
be deposited on  the  land. Vaughan (1968), in his
interpretation of the  preliminary  findings of the
National Solid Wastes  Survey, stated that the aver-
age amount of solid waste collected in the U. S. is
over 5.3 pounds per person per day, or more than
190 million tons per year. It is predicted that these
values will increase to 8 pounds per person per day
and  340 million  tons by  the year  1980. The
amount of waste actually generated is considerably
more than noted above, amounting in  1967 to 10
pounds of household, commercial and industrial
wastes for every man, woman  and child  per day,
totalling over 360  million tons per year.  It is fur-
ther .estimated that the current annual expenditure
to  handle  and  dispose  of these wastes in this
country is $4.5 billion per year.
     This paper is concerned with the ground-water
pollution potential associated with the operation of
dumps, sanitary  landfills, and any other  practices
of the land disposal of solid wastes. The paper first
includes a critical review of the important literature
covering  the area  of the ground-water  pollution
potential from sanitary landfills  and dump grounds.
This is  followed  by a review of the practices in  21
States in the U. S.  related to this same topic. Based
on the information derived from these two sources,
a  series  of  recommendations  are suggested to a
"re
-------
lyzed the data obtained from the National Solid
Wastes Survey on  over 6000 land disposal sites.
Using the modest criteria that a sanitary landfill is
one in which there  is daily cover of refuse, no open
burning, and  no water pollution problem, it  was
estimated that only 6 percent of the 6000 can be
reasonably characterized as "sanitary landfills." In
view of the above  it was decided for the purposes
of this paper to use the term "sanitary landfill" or
simply "landfill" in  the very broadest  sense.  The
primary concern was simply what effect land dis-
posal of refuse or  solid wastes of any kind, either
in the form of a true sanitary landfill or an open
dump,  has on the  ground-water quality in  the
vicinity of the operation.


             LITERATURE REVIEW
Major Studies
     After examining some of the literature in this
area, it became readily apparent that certain groups
or States had conducted and are still conducting
most of the research. Without doubt since the early
 1950*s there has been more activity in the State of
California in this regard than any other State in the
country. One of the first studies was an in depth
investigation  on the  leaching of soluble  salts and
 alkalies from  incinerator ash dumps (State of Calif.,
 1952). Following  this study, the City of Riverside
 sanitary landfill was used as the site  for the in-
 vestigating of the  leaching properties of a typical
 municipal refuse (State of Calif., 1954a; Univ. of
 Calif., 1955;  Univ.  of Calif., 1956).  Two  often
 quoted conclusions from this study regarding the
 effects of landfills on ground-water quality are:
      1.  "A sanitary landfill, if so located that no portion
 of it intercepts ground water, will not cause impairment of
 the ground water for  either domestic or irrigational uses."
      2.  "A sanitary landfill, if so located as to be in inter-
 mittent or continuous  contact  with ground water, will
 cause  the ground water in the vicinity of the landfill to
 become grossly polluted and unfit for domestic or irriga-
 tional uses."
 In 1961 another report was written entitled, Effects
 of  Refuse Dumps on Ground-Water Quality  with
 the main purpose  of collecting all available data on
 the extent of pollution of ground water from dump
 leachates, and to make recommendations for future
 research programs to  fill any  gaps in  knowledge
 available  (State of Calif.,  1961).  The report  in-
 cluded  a good literature review of such topics  as
 vertical water movement, decomposition process,
 gas production and movement, leaching, and travel
 of pollution. The  concern about gas production in
 recent years was  the impetus for the initiation  of
 another study  entitled, In-situ Investigation  of

 98
Movements of Gases Produced from Decomposing
Refuse (State of Calif., 1965; State of Calif., 1967:
Bishop,  Carter and  Ludwig, 1966). This  study
proved to  be of interest since up to that time most
work had been  done  on the question of  refuse
leachates affecting ground-water supplies, whereas,
practically  no  work  had been done  on  refuse-
produced  gases  as  potential  ground-water  pol-
lutants.  Merz and coworkers (Merz,  1969)  of the
University of Southern California conducted a five
year study on  sanitary landfills using four specially
constructed cells at the Spadra Landfill operated by
the Los Angeles County Sanitation District. Finally,
in 1965, the California Legislature directed that a
study be made of  water quality problems in the San
Francisco  Bay-Delta area including water contami-
nation  and  pollution  resulting from  disposal of
solid wastes. This study resulted in the preparation
of a report entitled, A Study of Solid Wastes Dis-
posal and Their Effect on Water Quality in the San
Francisco  Bay-Delta  Area  (Calif. Water   Board.
 1968).  One chapter in the report entitled,  "Influ-
ence of Solid  Wastes on Water Quality" presents a
short review of some of the literature on leachates
and gas production from disposal sites. Reference
is particularly made to past California  studies. In-
 cluded  also is a rather extensive survey of all land
 disposal sites in the Bay-Delta area from the stand-
 point of surface and ground-water quality.  A water
 quality  evaluation scheme was worked  out and ap-
 plied to each disposal site.
      Andersen and  Dornbush  (1967,  1968)  of
 South Dakota State University have been studying,
 over a period of almost ten  years,  the effects on
 ground-water  quality of dumping refuse from the
 city of Brookings in  an abandoned gravel pit lo-
 cated 2 miles south of the community. The ground-
 water  table  at  the site is about 6!/2  feet below
 ground surface and the principal geological feature
 of the  area is a sandy-gravel outwash covered at the
 surface by about a one foot clay and silt alluvium.
 The cation  exchange capacity of  the area was felt
 to be very low. The estimated ground-water veloci-
 ties are in the range of 1 to 3 feet per day. Through-
 out the study period, a total of 45 test wells have
 been  constructed "upstream" and  "downstream"
 of the dump  area. After some preliminary  studies,
 they found that chloride, sodium, specific conduct-
 ance,  and total and calcium  hardness  were the
 inorganic parameters of ground-water quality which
 could   be used  most  effectively to  denote any
 changes attributable to leachates from the disposal
 area. On  the  basis of statistical studies they found
 that chloride level is  the most sensitive parameter
 with wells, for example, in the center of the dis-
 posal area experiencing a 50 fold increase in con-

-------
centrations in comparison to the unaffected ground
water.  Isoconcentration  lines drawn for hardness
and specific conductance demonstrated clearly that
the significant leaching effects of these two param-
eters  on  ground-water  quality  remained in  the
immediate vicinity  of the disposal area.  High con-
centrations of these parameters were confined to a
relatively narrow band but extended as far as 1000
feet from the landfill disposal area. The most recent
work at the Brookings site has been directed toward
an evaluation of the effects of constructing a long
trench  on the quality of the degraded water flow-
ing from  the  disposal area.  The  trench  was con-
structed along the downstream edge of the disposal
site. Upon an evaluation of the water quality data
above and below the  trench, it was concluded that
the intercepting trench improved the quality of the
degraded ground water flowing from the fill area.
It was  felt that this beneficial effect was caused by
the dilution and  photosynthetic activity afforded
by the  surface water.
    A number  of studies have also been conducted
in  England through the  years. One often referred
to study  is called,  Pollution  of Water by Tipped
Refuse (Ministry of  Housing and Local Govern.,
1961)  in which the leaching properties of a landfill
were compared under "dry" and "wet" conditions.
In the  case of the former, the refuse was dumped
under dry conditions and a known amount of water
was applied to the refuse; whereas in the case of the
latter,  the refuse was  dumped  into water. Con-
siderably more pollutional matter was leached from
the refuse  under the  wet conditions and during a
shorter period  of time. The  publication also in-
cludes  some work on the possibility of setting up a
landfill site to collect  all leachates, direct them to a
central point and provide treatment prior to  dis-
posal  just  like  any  other  waste  water,  and  the
possibility  of  removing  stabilized refuse from an
"ideal" site for the reuse of  a fresh refuse. More
recently, a symposium was held on "The Effects of
Tipped Domestic Refuse on Ground-Water Quality"
in  which the results of four investigators were re-
ported (Water Treatment and Examination, 1969).
None presented evidence of any serious degradation
in  ground-water  quality  resulting from  the  land
disposal of solid wastes.
    Some of  the  most  useful  studies  in recent
years on  landfill site selection and evaluation from
the standpoint  of  practical  applicability of  the
information have been conducted by investigators
of the Illinois State  Geological  Survey. Hughes
(1967), for example,  wrote a  very helpful publica-
tion  on a  methodology for evaluating  a disposal
site considering the hydrologic environment of the
site and the method  of disposal. He stressed that
climatic, hydrologic,  and geologic factors strongly
influence  the production and spread of contami-
nants from landfill sites, and therefore it is danger-
ous to over-generalize the findings from one area to
another. Cartwright  and McComas (1968)  con-
ducted  earth resistivity  and soil temperature sur-
veys around four sanitary landfills in northeastern
Illinois and concluded that.
     "Geophysical surveys are not a substitute for hydro-
geologic studies, but can be used with moderate control as
a preliminary tool in the investigation of sanitary landfills,
and can be extremely useful in the location of  piezometers
for detail studies."
Bergstrom (1968a,  1968b) discussed the feasibility
of  disposing of industrial wastes into  deep  geo-
logical  formations in  the  State of Illinois. He
(Bergstrom,  1968c) also  discussed  in general the
disposal of wastes of all types in the ground from
the broad  standpoint of waste  management. He
stated that  instead of  viewing  waste disposal  on
land immediately as a ground-water pollution prob-
lem, many ground-water workers are beginning to
take a more positive role by studying and classify-
ing hydrogeologic environments relative to waste
disposal. Cartwright and Sherman (1969) wrote an
interesting and very useful document entitled Eval-
uating Sanitary Landfill Sites in Illinois in which
they point out, among other things,  that limestone
quarries and gravel pits are rarely, if ever, accepta-
ble refuse disposal sites from a hydrogeologic stand-
point, but that sanitary landfills can be  located in
relatively  impermeable, or slowly permeable, ma-
terial so that movement of refuse leachate will be
retarded. Hughes, Landon  and Farvolden (1969)
have  reported on a  study  describing the hydro-
geologic environments  in  the  vicinity of  four
existing landfill  sites  in northeastern  Illinois, in
order to determine the controls on  the  movement
of the ground water and the solids dissolved in the
ground  water. The intention was that this informa-
tion can then be used by  regulatory agencies to
help  determine   environments  most  suitable for
near-surface disposal of waste  insofar as contamina-
tion  of ground  water  and surface  water is  con-
cerned. The  report has a great deal of practical
information to aid the reader in this regard.
     Remson, Fungaroli and  others  of  the Drexel
Institute of  Technology,  have  been involved in
studies  on  ground-water  pollution potential  of
sanitary  landfills  for several years.  Two recent
reports were published containing useful informa-
tion on the  design of a laboratory lysimeter for
sanitary landfill  investigations, and the design of a
sanitary   landfill   field  experiment  installation
(Fungaroli, Steiner and  Remson,  1968;  Fungaroli,
Steiner, Emrich  and Remson, 1968). Though both
                                              99

-------
of these publications do not contain any actual
operating data and results, they do provide useful
hints for anyone interested in instigating an investi-
gation   of  this  type.  In  another  publication
(Remson,  Fungaroli  and Lawrence, 1968) the
authors proposed  a method of moisture  routing
first through the soil cover and then through the
underlying compacted refuse.  Their method was
illustrated  by  its  application  to a hypothetical
landfill.
     A series of papers have  been  published  by
LeGrand (1964a, 1964b, 1965, 1968) of the U. S.
Geological Survey on the general topic of ground-
water contamination from various sources. In these
papers he  presents a point evaluation system  for
assessing the contamination potential of a waste
disposal site, a very basic  and informative discussion
on  the management aspects of ground-water con-
tamination, the use of "malenclaves" for estimating
the areal extent of contaminants in the ground, and
a  methodology of approaching  a  ground-water
monitoring program.

Other  Studies
     In addition to the groups or States which have
been reported  above,  a number of other studies
have  been  conducted   on  the general topic  of
ground-water  pollution   from  sanitary landfills.
Qasim  and  Burchinal (1970a, 1970b) studied the
chemical and pollutional  characteristics of leachates
from different heights of refuse columns containing
similar fill materials of approximately the same age,
and operating under similar conditions of percola-
tion and leaching.  An interesting conclusion from
their  study was the suggestion  that deeper fills
pose  less  of a pollution problem   than  do the
shallow fills simply because  the rate of pollution
production is greater  initially  in  the case  of the
latter.  Kaufmann   (1969a,  1969b)  has  written a
good review on the  topic of hydrogeological aspects
of the disposal of  solid  wastes on the ground, in
addition to presenting some preliminary findings on
his current investigation on the over-all hydro-
geology  of two sanitary landfills in  the Madison,
Wisconsin area. Landon  (1969), who feels that site
selection for final  disposal of solid wastes is one of
today's most critical solid wastes problems, wrote
an  interesting  paper to show the  necessity and
application of hydrogeologic knowledge and con-
cepts to the selection  of refuse disposal sites. The
use of resistivity  measurements  for economically
obtaining hydrogeological information on a poten-
tial landfill site and operating landfill sites appears
to have merit, according to findings of Page (1968)
and Warner (1969). Finally, Hart (1967) reported
briefly on  a study which has been going on for
three years at the West Berlin, Germany landfill to
determine the effect of the compaction of refuse
upon the water regime within the fill.

General References
     There are a number of references which in-
clude a general discussion of  the  relationship be-
tween sanitary landfills and ground-water pollution
problems.  Most  of these sources  present  brief
summaries of the studies which have already been
presented in  this  paper, particularly  the California
studies. Some are fairly complete and helpful while
others  are quite brief and of limited usefulness
regarding the topic at hand. The main point is that
by reviewing several of these sources one can obtain
a general overview on the ground-water pollution-
solid waste disposal relationship.
     Cummins (1968) wrote  a short  review-type
report  entitled, Effects of Land Disposal of Solid
Wastes  on Water Quality.  Most of the important
studies  are included in the 15  references cited, but
very little detail on  results is included.  Golueke
(1968)  wrote a 300 page report which includes ab-
stracts  and excerpts from the  literature on the
broad topic of solid waste management. Because of
the  nature of the topic, it is not surprising that
some of the important investigations in the area of
ground-water  pollution  do not  appear  in this
publication. Two  of the ASCE—Manuals  of  Engi-
neering Practice  (ASCE, 1959;  1961) include a
short discussion  on ground-water pollution. The
same is true for the APWA book, Municipal Refuse
Disposal (APWA,  1966). All three  of these dis-
cussions draw mainly from the California studies.
Sorg and  Hickman (1968) have  written  a  small
semitechnical report for  the  U.  S.  Public Health
Service  entitled,  Sanitary  Landfill  Facts  which
includes some discussion  of water pollution prob-
lems. A bibliography is  added to the end of the
report.  The  U.  S. Public Health  Service has also
made available a number of helpful  bibliographies
on  the  topic of sanitary landfills  specifically, and
refuse  collection and  disposal in  general (Steiner
and Kantz, 1968; U.  S.  Dept.  of H.E.W., 1954-
1963).  The sanitary landfill bibliography was pre-
pared by  Steiner and Kantz of Drexel Institute of
Technology and covers the literature  for the period
1925 to  1968.  Finally Weaver (1956) and Black
(1965)  have  briefly  discussed  the  ground-water
pollution  problems in some of their  writings  on
sanitary landfills.

Health and Nuisance Problems
     Considering the tremendous  amount of solid
wastes  which have been deposited on  the  land,
there are still relatively  few recorded  instances of
 100

-------
serious ground-water pollution problems linked to
leachates  from  landfills and dump grounds. No
doubt  there  have  been  probably many  small
localized  nuisance conditions which have  never
been reported. There have also been and still are
many cases in which impairment of water quality
has not been detected because there have  been no
noticeable deleterious effects traceable to the water
being used. A review of some of the  health and
nuisance  problems  recorded is  presented below.
Some of the information on instances reported are
quite detailed and  very specific; whereas,  many
others  are similar to the following statement taken
from a 1953 Joint Study of the  HEW Department
and APHA entitled, Refuse Collection and  Disposal
for the Small Community (1953): "Landfills should
be so located that seepage from them will not cause
hazards or nuisances."
     Forty years ago,  Calvert  (1932) reported on
the deterioration of a well water caused by the pit
disposal  of liquor  drained  from  cooked  garbage.
The  well water  before and after  contamination
showed a substantial increase  in  iron,  total hard-
ness, total solids,  CO2  and total organic nitrogen.
The  University of California (1952) in  1952 con-
ducted  field studies on refuse collection  and dis-
posal operation  in 13 California cities. In the
section  on  public health  problems, no  mention
was  made  of ground-water pollution problems
other than a minor reference that the possibility of
ground-water contamination should be considered
in selecting a site for  landfill disposal of refuse.
Publication  No. 24  of  the  California State  Water
Pollution Control  Board  (State  of Calif.,  1961)
includes  several pages on the reported experiences
from the literature  on health  and  nuisance prob-
lems. The ASCE publication Ground Water Basin
Management  (ASCE,  1961) makes reference to
potential  ground-water  pollution  problems  from
disposal of wastes of any kind on the land. A listing
of typical industrial wastes together with common
characteristics affecting ground water are presented
in the document. Anderson  (1964) (APWA, 1963)
has written  on the general topic of the public
health  aspects of solid waste disposal and includes
potential  water  pollution problems as  one of the
matters of public health concern. Hanks (1967) in
1967 wrote  a very detailed and  well referenced
report for the Solid Wastes Program of the  U.S.P.H.
S. on the Solid Waste-Disease Relationships. Under
the  topic of diseases  associated  with  chemical
wastes, several references are cited on the pollution
of ground water  from the  leachates of  sanitary
landfills.  A frequently  mentioned  report on the
subject of ground water is  one entitled, Ground
Water Contamination which presents the  proceed-
ings  of a 1961  symposium (Robert A.  Taft  San.
Engrg. Center,  1961). In the  session dealing  with
specific incidents of contamination of ground wa-
ter,  Weaver discussed the  significance  of refuse
disposal in this regard, using to a great extent data
from the California studies of  Merz  and others.
Chapter five of Sanitary  Landfill, (ASCE, 1959), a
publication of ASCE, is entitled "Public Health and
Nuisance  Considerations," and includes a section
on water pollution. Walker (1969) recently wrote a
very interesting paper on ground-water pollution in
Illinois,  and  one  of the pollution categories he
considered was that of garbage disposal.  He makes
the statement:
     "Serious contamination of the ground-water reservoirs
near  these dumps (garbage dumps) can readily  occur  if the
bottom of the depressions is below  the water table, or if the
earth material  separating the dump from the aquifer  is
primarily silt, sand, or other relatively permeable material."
He  presented two  actual  cases of pollution  of
ground-water supplies traceable to leachates from
garbage dumps.  As part of the City of Santa Clara
demonstration landfill study,  Stone and Friedland
(1969) conducted a survey of American  cities with
populations greater than  10,000.  They received
replies from  120  landfill sites operated  by 102
governmental  agencies serving a combined popula-
tion  of  17,800,000. Ground pollution problems
were reported at 11 of the sites or approximately 9
percent of the total 120 sites. Williams (1969), who
discussed the over-all topic of ground-water pollu-
tion, claimed that;
     "Sanitary landfill seepage into sand or an overly deep
excavation for a lagoon so that shallow subsurface water in
the stream alluvium is intercepted, are two of the most
common ways of polluting shallow ground-water supplies."

An  interesting  case  of the  contamination  of  a
ground-water  supply  by an  industrial  waste has
been reported (Mpls. Trib., 1968). A large company
in the Minneapolis-St. Paul  area dumped isopropyl
ether in  a disposal site for several years before  it
was realized that the industrial solvent contami-
nated the aquifer.  The  company  had to spend
$600,000 to remedy the situation. Another exam-
ple of how an industrial waste landfill can affect a
public water  supply is  the case of Kansas  City,
Missouri,  reported  by  Hopkins  and  Popalisky
(1970).
     Two additional interesting publications should
be noted at this point, though they are  not specif-
ically addressed to the topic of ground-water pollu-
tion from refuse decomposition on the  land. One of
the  reports  entitled  Investigation  of  Travel  of
Pollution (State of Calif., 1954b) is concerned with
the  artificial recharge of aquifers with sewage treat-
ment plant effluents, and thus the principal item of

                                             101

-------
concern is the fate of microorganisms, and organic
and inorganic chemicals as the liquid passes through
the subterranean soil. The report includes a litera-
ture review  on the  subject. The second  report
entitled  Status  of Knowledge  of Ground Water
Contaminants (Stanley and Eliassen, 1961) includes
a  very  comprehensive literature search including
better  than  700 references  on  the characteristics
and status of knowledge on the various contami-
nants which can be found in ground water.  This
report  can  serve as  a  valuable  reference  when
specific information is desired on  the effects of a
particular leachate chemical.

DISCUSSION OF THE LITERATURE FINDINGS
     After some review of the literature on sanitary
landfills, one point  becomes clear almost immedi-
ately; that  is, there are very few case histories of
serious   or even  troublesome  contamination  of
ground water which are  directly attributable to the
leachates from  sanitary  landfills. To be sure there
may have been  many such instances which  were
simply  not reported. Even more  probable, there
may be many unknown cases currently of people
routinely using water impaired in quality somewhat
as the  result of the land disposal of solid wastes.
The writer, for example, has not  learned  of a
single person who has died as the result of ground
water being contaminated by  a landfill. Consider-
ing the number of  landfills past  and present  and
the amount and variety of solid wastes generated
in our  modern technological  society, this is a
remarkable situation. This statement is not made
at the  outset  of this  discussion to  belittle the
potential for serious harm to many people through
this mechanism, nor to  cast any aspersions on the
fine research which has been and  currently is still
being done in this area. But, nonetheless, this point
is most critical when considering recommendations
for future action in  this area, since it is this type of
statistic which  ultimately motivates action in any
similar  activity. For example, the greatest impetus
for improved public water treatment and distribu-
tion works at the turn of the century was no doubt
the 20 to 30 typhoid deaths per 100,000 popula-
tion per annum occurring at that time.
     This situation attests to the almost miraculous
capability of most soils to attenuate the leachates
generated  from sanitary landfills.  From the results
of the  literature, there  is no  question that these
concentrated leachates are of extremely high pollu-
tion strength. There are few industrial waste flows
that would match this material and without doubt
no responsible governmental agency would tolerate
the discharge of a material like this untreated into
a  surface  body  of  water.  There  is  an important
difference in the subterranean regime, however. The
soil provides the site for active microbial degrada-
tion of  the organics which  are present in  the
leachates. The  inorganics are adsorbed to the  soil
surface and many of the more undesirable ions are
exchanged for  the more desirable ones. The  ex-
tremely  low velocity of  the  underground-water
resource   provides  the  necessary time  for  these
activities to reach a fair degree of stabilization, thus
confining most of the degradation processes to the
immediate vicinity  of the landfill. The soluble  end
products  are attenuated even further by the sheer
vastness  of the  underground-water body by  the
simple mechanism  of dilution. The highly soluble
chloride  ion provides a useful tracer for the situa-
tion  described  above.  In most of the research
studies examined, the chloride concentration in the
leachate   directly  below  the  landfill  was  always
extremely  high.  The  chloride   concentration
dropped  drastically in water samples taken only a
short distance  from the landfill  operation. At  dis-
tances of several hundred feet  the concentration
drops down to almost native or background levels.
     Unfortunately the described process does not
hold true to  the  same  degree  for all  geological
formations, and therein lies the crux of the prob-
lem. The above  will usually hold true for uncon-
solidated formations consisting of varying propor-
tions of clay, silt,  fine  sand and loam with low to
medium permeabilities. For unconsolidated materi-
als of coarse sand and gravels with high permeabili-
ties or consolidated materials such as limestone or
shale with fissures, faults or fractures of any kind,
the protective  mechanism breaks down because of
one important reason,  that is, time. In formations
of the latter type there is much less time available
for the  degradation process to  take place within
the vicinity of  leachate  generation  because  the
underground velocities  are much higher. Thus  par-
tially "treated"  and poorly diluted leachates can
appear at greater distances from the  landfill.  The
assumption made  here is that  the ground-water
flow is  through and away from the landfill  site.
If  all the flow lines are directed  toward the site
this situation will  not  necessarily  occur and what
probably will happen is that the ground water will
discharge at the surface somewhere nearby. Such a
situation could then have a deleterious effect on
the surface supply.
     It  is convenient to  think of a mass of refuse
stored in a  landfill  site as representing a certain
mass or  quantity  of pollutants. Some of the re-
searchers have  in  fact done  just  this when they
express  specific  leachate constituents in terms  of
weight  per cubic  yard or per  ton of deposited
refuse. This mass of pollutants  will eventually be
 102

-------
generated from the landfill, since the processes of
weathering and biological degradation always take
place. The important variable again is time. In order
to speed up the degradation and weathering proc-
ess, moisture and favorable temperature are neces-
sary. Usually temperature is not a restricting factor
even in northern latitudes  since the interior of the
refuse mass is insulated from the ambient tempera-
ture and the degradation process is exothermic in
character.  Available  moisture then  becomes the
limiting factor. Rapid degradation will  occur in a
more loosely packed landfill where surface waters
are  permitted  to  percolate  freely  through  the
refuse. The degradation process can  be  slowed up
considerably  by allowing less surface water to pass
through the fill. It is virtually impossible to abate
completely this activity. Obviously if the degrada-
tion process is retarded by restricting  passage of
water in any way, more time will be allowed for
the natural attenuation process  to take place be-
yond the landfill  site.  A combination of retarded
degradation in a geologic formation of ideal attenu-
ation provides  the  least  likelihood   of serious
ground-water deterioration. Speeding up the degra-
dation process imposes a  greater load  on the sur-
rounding  geologic formation;  and if the geologic
formation  is  a poor one, the  problem is  com-
pounded. Again it must be remembered that the
same pollutional mass is involved in both cases.
     It is possible  to engineer and operate a landfill
operation  with  the  intent  of  minimizing  the
amount  of  percolation  through the  deposited
material.  Bottom  liners  of  various types,  high
degree of refuse compaction, shredding, and highly
impervious earth  covers are examples of what can
be done. These measures would, on the other hand,
be  highly  ineffective   in  situations  where  the
moisture sources originate from below the landfill.
There is  no question that  in areas where wide fluc-
tuations  in the ground-water table  occur to the
point where the  refuse becomes repeatedly  satu-
rated with water and then  drained, the degradation
process is intensified to probably its optimum  level.
If the attenuation capability of the geologic forma-
tion surrounding  this  site is  limited,  a situation
again exists  for  serious pollution of the ground
water.
     Some in the field of solid waste  disposal argue
that a  landfill  should be  designed  for optimum
degradation and weathering to occur.  This means
that an  ample amount of  water should be  per-
mitted to percolate through the fill. It  also means
that the  leachates must then be collected in a drain
system and treated  prior to discharge to a surface
water body or possibly back through the fill. After
a  reasonable degree of stabilization has occurred
the leachates  will no longer  be collected  in  the
drain system but allowed to pass into the surround-
ing soil. In this way they argue that the degradation
process can be controlled as desired and the possi-
bility  of  future  pollution  problems  is reduced
considerably.  Some  for example argue  that huge
quantities of  stored refuse  located  in  the earth
close to large population centers are  akin to geo-
logic  pollutional  "time bombs" which could be
very troublesome to  future generations. There is no
doubt that landfill sites can be engineered to speed
up the degradation process and collect and treat the
leachate. However, the economics of  this arrange-
ment  may favor  other disposal methods  which,
relative to traditional landfills, were formerly con-
sidered too expensive.
     Another  point  becomes  quite apparent after
reviewing the literature in this area. Much more
geologic, or  more specifically,  hydrogeologic ex-
pertise should be employed  prior to the selection
of a landfill site. Someone with training in hydro-
geology7 can establish with a fair degree of accuracy
upon an examination of  the site and  often with a
limited amount of field testing what the leachate
attenuation potential of  a site  will be and if the
cover material will permit a slow or rapid percola-
tion of water into the fill. Too  many landfill sites
are selected on a purely political, economic  or con-
venience basis with  no or little attention given the
geology of the site.  It is  surprising that  regulatory
agencies have been  somewhat  lax in this  regard,
also. The writer is convinced  that attention to this
matter alone  will minimize  many future ground-
water pollution problems attributed to landfills.
     The  major threat to ground-water quality in
the  future will likely be from the land disposal of
industrial wastes.  Many of these wastes decompose
slowly or are  nondegradable,  which means the
protective mechanism of attenuation afforded by
the  soil is  no longer available.  Many industrial
wastes can  impart  odor,  taste  and even toxic
problems to ground waters at extremely low con-
centrations. With advances in technology7 and the
increase in over-all affluence there will undoubtedly
be an  increase  in the  amount and  complexity of
solid  industrial wastes produced. Many new com-
pounds will also be synthesized in the  future which
can  pose either acute or chronic threats  to the
health of future users of ground water. The  present
trend  toward more  stringent  surface water  quality
standards will cause  some industries to look toward
land disposal for the  solution to their industrial
waste problems. It is important to keep in mind in
this regard  that much  empirical evidence is availa-
ble to substantiate the generally innocuous effects
to humans resulting  from the decomposition of
                                             103

-------
ordinary municipal  refuse in the ground. On the
other hand,  practically  each new solid industrial
waste completely  nullifies the dependence on this
past evidence as the primary basis for establishing
guidelines to dispose of  the new material. For this
reason alone it  is  imperative that extreme caution
be exercised  in the land disposal of all solid indus-
trial wastes in the future.
       SURVEY OF STATE PRACTICES
     As a means of becoming acquainted with the
numerous  problems associated  with ground-water
pollution  from  disposal of solid wastes, it was de-
cided  to  conduct  a survey  of the  activities and
policies related  to  this area  in  21 of the States in
the  U. S. The  States  selected comprise a  total
present population of approximately 140,000,000
which amounts  to  70 percent of the U. S. popula-
tion. The  basis for selection was somewhat arbi-
trary,  but the  list was intended  to include the
larger  States,  States in the midwestern part of the
country  and  States  in  which it was known that
some  activity in this area was taking place. The
States included in the  survey were: California,
Florida,  Illinois, Indiana,  Iowa, Kansas, Maryland,
Massachusetts, Michigan, Minnesota, Missouri, New
Jersey, New York, North Dakota, Ohio, Oklahoma,
Pennsylvania, South  Dakota, Texas, Washington
and Wisconsin.
     Letters were sent in June, 1969 to the regula-
tory officials in  the various States who were known
to be  familiar with the solid waste practices within
their particular  State. The letters first explained the
reasons for the information requested and specif-
ically  solicited  comments on the  following four
questions:
      1. Are you aware of any research activities in
your State that are  concerned  with ground-water
pollution  from  sanitary landfill and open  dump
type of operation?
     2. Does your  State have  published codes or
guidelines regarding site selection for sanitary land-
fills and dump  grounds?  Do you have regulations
pertaining to the operation of such areas?
      3. Does your State  specify a minimum dis-
tance  that a water well can  be located from  a
landfill or dump ground?
     4. Do you anticipate that your State will be
engaged  in some  aspect  of this question in the
immediate future,  such as writing  new or revising
old codes, field  research programs, etc.?
      Upon an examination of the replies received,
which varied in detail from "ves" and "no" answers
on the original letter mailed to voluminous replies
containing complete laws, codes and research re-
ports, it became readily apparent that there has not
been a great deal of research activity in the area of
ground-water  pollution from the land disposal  of
solid  wastes.  A little  more  than half the States
queried responded that no research activities of this
type  were taking  place  in  their  own  particular
State.  Probably the most extensive work in  the
past has  been confined to three States,  namely,
California, Illinois and South Dakota. Most of the
recent work which has been published in this area
has originated from   universities  and public and
private research agencies within these States. A few
other States are beginning to become a little more
active in  this  regard. They are Maryland, Pennsyl-
vania, and Wisconsin.  Another interesting  observa-
tion was  that a number  of States are beginning to
or are making plans to  monitor the ground-water
quality in the vicinity  of landfill  sites. It is the
writer's  opinion  that other  States will  become
involved in this activity in the future.
     It was also readily apparent that there is much
variation  in the details given to codes and guide-
lines  pertaining to the  selection of landfill sites,
particularly with reference to the possible ground-
water pollution problems. These range from a few
States which  do not  have any published codes or
guidelines to others like California where one of the
Water Quality Control Boards recently published a
ten page  "Statement of Policy" going into numer-
ous details on classification of wastes and disposal
sites  primarily  to  guard against the pollution of
ground and surface waters. Many of the State regu-
lations merely included broadly worded statements
of the type "Sanitary  landfill and other solid waste
disposal  activities shall  not pollute the  ground
waters and surface waters of the  State." Of the
States queried, California,  Illinois, Maryland, Michi-
gan,  Minnesota, New York,  Ohio  and Wisconsin
have  made the most detailed reference to ground-
water pollution potential. On the most part,  these
States  require  as  a   matter of  policy, geological
data, water table  elevations and other hydrological
data  prior to the approval of landfill sites. Other
States  may have  local  health departments which
require  the  same information,  but  it  was not
expressed as  a matter of State policy. It was also
evident  that  the  States with the  most  stringent
regulations in this regard were the  ones  with  the
most recently enacted laws.
      The  policy on landfill to water well distance
employed by the  States surveyed appeared for the
 most  part to be  a very tenuous one. Eight States
would not commit  themselves  to  a specific dis-
 tance,  stating in  effect that each  case was  con-
 104

-------
sidered  individually before a specific distance was
set. For the States that gave a value, the distances
varied from 50 to 1000 feet with most of the values
in the 100 to 500 feet range. Some States indicated
that no specific value was used for sanitary land-
fills, and  thus  they gave the values used for the
location of water wells from any known sources of
contamination.  Most  of  the  States  which estab-
lished a distance cautioned that the value was only
used as a rough guide and it was by no  means a
rigid one. The  tenor of the remarks in this regard
was  that no one really knows what a "correct"
value is and empirical evidence from the  past
indicates that a particular value was used in the
past without any  adverse effects.  The   lack  of
ground-water monitoring in the vicinity of landfill
sites  is  probably  the primary reason  for this
dilemma.
     Very  little research  activity  was  being con-
templated in the States surveyed in  the immediate
future. As was noted previously, some ground-water
quality  monitoring  around landfill  sites  will  be
conducted  in a few of the States. Most  of the
States, however, will  be active  in the area of rules
and  regulations  pertaining to solid waste  disposal.
Many States were either in the process of revision
of current laws or  have new laws pending. It also
appeared  that  more  emphasis will  be placed  on
ground-water pollution problems in the new regu-
lations. There also  appeared to be a trend toward
more rigid  State control of these  activities than
there had been in the past.
  RECOMMENDATIONS TO A REGULATORY
                   AGENCY
     A  review of the literature on the subject of
the  relationship  between land disposal  of solid
wastes and ground-water  pollution, plus the survey
of practice employed by 21  States in this regard,
have suggested certain steps that regulatory agencies
can  take  to  minimize problems  in  this area. The
term "regulatory agency" in the recommendations
to  follow pertains to the  governmental  entity,
agency  or  department   which has  the primary
responsibility of regulating and licensing sanitary
landfill  operations  within the State.
     1.  The  regulatory agency should have availa-
ble a geologist on its staff, preferably one trained in
the area of hydrogeology, to assist in the sanitary
landfill  site selection  processes within the State.
     2.  The  geologist on  the  staff should begin to
accumulate  geological data  within  the  State  and
broadly outline  areas considered to  be either good
or poor potential landfill sites. The activities in the
State of Illinois  serve as a good example in this
regard.

     3.  The trend should be  towards the require-
ment of more hydrogeologic  and hydrologic field
data for sites  that  are  questionable for  landfill
operations. The burden of proof should be placed
on the landfill operator or owner. The staff geolo-
gist should be given the responsibility of deciding
when additional  field  data  are required. It is  the
opinion of the writer that much useful information
can be  obtained even with a modest  amount of
field testing.
     4.  The  regulatory  agency  should  be very
cautious when  it comes to the approval of  the
ground  disposal of industrial wastes. An up-to-date
file  should  be  maintained  on   various types  of
industrial wastes, their degradation properties and
their effects on the aquatic environment. A litera-
ture search  should be made  periodically  on this
topic.
     5.  The use of ground-water monitoring wells
should  be  considered  in those  cases where  some
doubt  exists as to future  effects  of a particular
landfill  operation.  This  is  somewhat akin  to  re-
quiring  that  water samples be  periodically taken
downstream of an effluent discharge  to maintain a
check on waste water disposal operations.
     6.  The regulatory agency should follow as its
basic policy the concept of trying to slow down the
refuse  degradation process by  minimizing water
percolation through the refuse mass. Slowing down
degradation provides more time for leachate attenu-
ation.  Past experiences have demonstrated  that
longer  times  provide the  most effective  safety
measure when it comes  to separating sources of
ground  - water  contamination   from  points   of
ground-water use.
     7.  The  regulatory  agency  should  not  dis-
courage  novel  methods  of  collecting and treating
refuse  leachates for  certain installations  where
proper  monitoring and  control  can  be exercised.
When considering facilities  of this type, an impor-
tant lesson to  be learned from  waste water treat-
ment plant operations is that the smaller and the
more remote  the treatment facility is, the greater
the likelihood  of poor operation regardless of the
original design and the degree of automation.
     8.  It is virtually impossible  to hold to a speci-
fied distance between  a point of water use such as
a  well  and the site  of a sanitary  landfill. Tre-
mendous variations in the hydrogeology surround-
ing each site precludes the establishment of such a
published figure. However, lacking any field data
                                             105

-------
 the distance should be as long as possible in order
 to have the built-in safety factor of greater time as
 stated previously. Figures of 500 to 1000 feet  are
 not unrealistic if adequate field data are insufficient
 to prove otherwise.
      9. The  regulatory agency should encourage
 the practice  of regional or  district approaches to
 solid   waste  collection and  disposal.  Economic
 incentives  should be  available  to provide funds to
 make  area-wide  feasibility studies. This approach
 will  reap  great  benefits  in  the control  of  solid
 waste disposal practices.
      10. As a general  rule,  the regulatory agency
 should prohibit the use of abandoned rock, gravel
 or sand quarries as sites for the disposal of refuse of
 any type.  Standing  water in such depressions is
 usually nothing more than a  visible direct link to
 the ground-water supply. The leachate attenuation
 mechanism under such conditions is  completely
 lost. If extensive hydrogeologic studies demonstrate
 that the depression is in a discharge ground-water
 zone it is possible that such a site  can be used for
 landfill disposal.  However, nearby  future ground-
 water  withdrawals may change the flow network
 around such  a site considerably.   The  burden  of
 proof  plus any  remedial  safeguards   should  be
 placed on the owner  of such a site. As a rule,  such
 sites should not  be used unless a thorough hydro-
 geologic study is made.
     11.  The   regulatory   agency  should  support
 some research work in  this area. Some good exam-
 ples are the studies on existing landfills which  have
 been conducted  in California, Illinois  and  South
 Dakota.

             ACKNOWLEDGEMENTS
     This investigation was  supported  . by  funds
 supplied  by the Wisconsin Department  of Natural
 Resources. A report  on the complete  findings  of
 this study was submitted to the  Department  in
July, 1970. The writer wishes to express his grati-
 tude to E.  D.  Cann and C. D.  Besadny of the Wis-
 consin  Department of Natural Resources, as well as
 the  officials from the other  States which partici-
 pated   in   the survey  for  their   assistance  and
cooperation.

                  REFERENCES
Andersen,  J. R. and J.  N. Dornbush. 1967. Influence of
     sanitary  landfill  on  ground-water  quality.  Jour.
     AWWA. v. 59, p. 457.
Andersen, J. R.  a:id J. N. Dornbush. 1968. Quality changes
     of shallow ground water resulting from refuse disposal
     at a gravel pit. Report on Project  SRI 3553, Brook-
     ings, South Dakota. 41 pp.
Anderson, R. J.  1964. The public  health aspects  of  solid
106
      waste disposal. Public Health Reports, v. 79, p. 93.
 APWA. 1963. Proceedings—National conference on solid
      waste. Special Report no. 29. 228 pp.
 APWA. 1966. Municipal refuse disposal. 2nd Ed., Interstate
      Printers and Publishers, Inc., 111. 528 pp.
 ASCE. 1959. Committee on sanitary landfill practice of the
      sanitary engineering division, sanitary landfill. Manuals
      of Engineering Practice no. 39. 61, pp.
 ASCE. 1961. Ground water basin management. Manuals of
      Engineering Practice no. 40. 160 pp.
 Bergstrom, R. E. 1968a. Feasibility of subsurface disposal of
      industrial  wastes  in Illinois. Illinois State Geological
      Survey, Circular 426. 18 pp. '   i
 Bergstrom, R. E.  1968b. Feasibility criteria for subsurface
      waste disposal in Illinois. Ground Water, v. 6, no. 5,
      p. 5.
 Bergstrom, R. E.  1968c. Disposal of wastes: scientific and
      administrative considerations. Illinois State Geological
      Survey, Environmental Geology Notes, no. 20, 12 pp.
 Bishop, W. D., R. C. Carter, and H. F. Ludwig.  1966. Water
      pollution  hazards from refuse-produced  carbon di-
      oxide. Advances  in Water  Pollution Research.  Proc.
      3rd  Intern.  Conf., Munich,  Germany,  v. 1, pp. 207-
      228.
 Black, R. J. 1965. A review of sanitary landfill practices in
    ,  the  United  States. Proceedings,  3rd International
      Congress.  International  Research Group  on Refuse
      Disposal, Trento, Italy, May 24-29. pp. 40-47.
 California  State  Water  Resources Control  Board.  1968.
      Special study—solid waste disposal. Task VII-IC, San
      Francisco Bay-Delta Water Quality Control Program.
 Calvert, C. K. 1932. Contamination of ground water by
      impounded garbage waste. Jour. AWWA. v. 24, p. 266.
 Cartwright, K. and R. McComas. 1968. Geophysical surveys
      in the vicinity of sanitary  landfills in northeastern
      Illinois. Ground Water, v. 6, no.  5, p. 23.
 Cartwright, K. and F. B.  Sherman. 1969. Evaluating sanitary
      landfill sites  in Illinois. Illinois State Geological Survey
      Environmental Geology Notes, no. 27, 15 pp.
 Cummins,  R. L.  1968.  Effects of land disposal of solid
      wastes on water quality.  Report no. SW-2ts, Dept. of
      Health,  Educ. and Welfare, U.S.P.H.S.  Cincinnati,
      Ohio, 29 pp.
 Fungaroli,  A. A.,  R.  L. Steiner,  G. H. Emrich, and I.
      Remson.  1968. Design  of  a  sanitary landfill  field
      experiment installation. Drexel  Institute of Technol-
      ogy,  Dept. of Civil Engineering. Series 1, no. 10, 32
      PP-
 Fungaroli,  A. A.,  R.  L. Steiner, and  I. Remson.  1968.
      Design of a sanitary  landfill  laboratory lysimeter.
      Drexel Institute of Technology, Dept. of Civil Engi-
      neering. Series 1, no. 9, 27 pp.
Golueke, C. G. 1968. Comprehensive, studies of solid wastes
      management—abstracts and excerpts from the litera-
      ture.  Univ. of Calif.,  Sanitary  Engineering Research
      Laboratory. SERL Report no. 68-3, 308 pp.
Hanks, T.  G. 1967.  Solid  waste/disease relationships; a
      literature survey.  Public  Health  Service  Publication
      no. 999-UIH-6.  U. S. Government  Printing  Office.
      179pp.
Hart,  S. A. 1967. Solid wastes management in Germany;
      report of the U. S. study team visit, June 25-July 8.
      Public Health Service Publication  no. 1812. Govern-
      ment Printing Office. 18 pp.
Hopkins, G. J. and J. R. Popalisky. 1970. Influence of an
      industrial waste landfill operation on a public water

-------
     supply. Jour. WPCF. v. 42, p. 431.
Hughes, G. M.  1967. Selection of refuse disposal sites in
     northeastern Illinois. Illinois State  Geological Survey,
     Environmental Geology Notes, no. 17, 18 pp.
Hughes, G. M.,  R. A. Landon, and R. N. Farvolden. 1969.
     Hydrogeology  of solid  waste disposal  sites in north-
     eastern Illinois. U. S. Dept. of Health, Educ. and Welf.,
     U.S.P.H.S. Bureau of Solid Waste Management, Dem.
     Grant no. DO1-U1-00006, 137 pp.
Kaufmann, R.  F.  1969a.  Hydrogeological aspects of solid
     waste disposal. Univ. of Wisconsin Extension, Solid
     Waste Management. Aug. 13-14.
Kaufmann, R. F.  1969b. Personal communication, Universi-
     ty of Wisconsin, Dept. of Geology, Madison, Wisconsin.
Landon, R. A.  1969. Application of hydrogeology to the
     selection  of refuse disposal sites. Ground  Water, v. 7,
     no. 6, p. 9.
LeGrand,  H. E.  1964a.  System for evaluation of contamina-
     tion  potential of  some waste disposal  sites. Jour.
     AWWA. v. 56, p.  959.
LeGrand,  H. E.  1964b.  Management aspects  of ground-
     water contamination. Jour. WPCF. v. 36, p. 1133.
LeGrand,  H. E. 1965.  Patterns of contaminated zones of
     water in the ground. Water Resources Research, v. 1,
     p. 83.
LeGrand,  H. E. 1968.  Monitoring of changes in quality of
     ground water. Ground Water, v. 6, no. 3, p. 14.
Merz, R. C. 1969. Special studies of a sanitary landfill. Final
     Report to Bureau of Solid Waste Management, Grant
     no. U1005 18-08, Univ.  of Southern California.
Ministry of Housing and Local Government. 1961. Pollution
     of water  by tipped refuse. Her Majesty's Stationery
     Office, London. 141 pp.
Mpls. Trib. 1968. 3M pumps polluted water from dump.
      Feb. 28,  1968.
Page, L. M. 1968. Use of the electrical resistivity method for
      investigating  geologic  and  hydrologic  conditions in
      Santa Clara County, California. Ground  Water, v. 6,
      no.  5, p.  31.
Qasim,  S. R.,  and  J.  C. Burchinal.  1970a. Leaching of
      pollutants  from  refuse beds. J. of S.E.D. of ASCE.
     v. 90, p. 49.
Qasim,  S.  R.,  and J. C. Burchinal.  1970b.  Leaching from
      simulated landfills. Jour. WPCF. v. 42, p. 371.
Refuse  collection .and  disposal  for  the small community.
      1953. A  joint study of U. S. Dept. of Health, Educ.
      and Welf. and  the Am.  Public Health Assoc.
Remson,  I., A.  A. Fungaroli, and A. W.  Lawrence.  1968.
      Water movement in an unsaturated sanitary landfill.
      J. of S.E.D. of ASCE. v. 94, p.  307.
Robert A. Taft  San. Engrg.  Center. 1961.  Proceedings of
      1961 symposium—ground-water contamination. Tech-
      nical Report W61-5. Cincinnati, Ohio. 218 pp.
Sorg, T. J., and H. L. Hickman. 1968. Sanitary landfill facts.
      U. S. Dept. Health, Educ., and Welfare, Public Health
      Service. Publication no. 1792, Cincinnati, 26 pp.
Stanley, W. E., and R. Eliassen. 1961. Status of knowledge
     of ground water contaminants. Published by Federal
     Housing Administration, Technical Studies Program.
     465 pp.
State of Calif. 1952. Report of the investigation of leaching
     of ash dumps. SWPCB Publication no. 2. 100 pp.
State of Calif. 1954a. Report on the investigation of leach-
     ing of a sanitary landfill. SWPCB Publication no. 10.
     92 pp.
State of Calif. 1954b. Report on the investigation of travel
     of pollution. SWPCB Publication no. 11. 218 pp.
State of Calif.  1961. Effects of refuse  dumps on ground-
     water quality. SWPCB Publication no. 24. 101 pp.
State of Calif. 1965. In-situ  investigation of movements of
     gases produced from decomposing refuse. State Water
     Quality Control Board. Publication no. 31. 211  pp.
State of Calif. 1967. Final report on in-situ investigation of
     movements  of gases  produced  from  decomposing
     refuse.  State Water Quality Board. Publication  no. 35.
Steiner, R. L.,  and R. Kantz. 1968.  Sanitary  landfill: a
     bibliography. Public  Health  Service  Publication no.
      1819. U. S. Government Printing Office, 37 pp.
Stone, R., and H.  Friedland. 1969.  National survey  of
     sanitary landfill practices. Public Works, v. 100, p. 88.
U. S. Dept. of H.E.W. 1954-1963. Refuse collection and dis-
     posal—an annotated bibliography, supplements B, C,
     D, E, and F. Cincinnati, Ohio.
Univ. of Calif. 1952. Sanitary  engineering research project,
     an  analysis of refuse collection and  sanitary  landfill
     disposal. Tech. Bulletin no. 8, Series 37. 1 33 pp.
Univ. of Calif. 1955. Report on continuation of an  investi-
     gation  of  leaching of rubbish dumps. Sanitary Engi-
      neering Research Laboratory. (Supplement to Publi-
      cation no. 10), 26 pp.
Univ. of Calif. 1956. Report on continuation of an investi-
      gation  of  leaching of rubbish dumps. Sanitary Engi-
      neering Research Laboratory. (Supplement to Publi-
      cation no. 10), 29 pp.
Vaughan,  R.  D. 1968. The national solid wastes survey, an
      interim report. Presented at the 1968 Annual Meeting
      of  the  Institute for  Solid  Wastes of the American
      Public  Works  Association, Miami Beach, Florida. 53
      pp.
Walker, W. H. 1969. Illinois ground water pollution. Jour.
      AWWA. v. 61, p. 31.
Warner, D.  L. 1969. Preliminary  field studies using earth
      resistivity  measurements for delineating  zones of
      contaminated ground  water. Ground Water, v. 7,  no.
      1, p. 9.
Water  Treatment  and Examination. 1969. Symposium on
      effects  of tipped  domestic refuse  on ground-water
      quality, v. 18, part 1, pp. 15-69.
Weaver, Leo. 1956. The sanitary  landfill-part  II: selection
      of site. American City. v. 71, p. 132.
Williams,  J.  H.  1969. Can ground-water pollution be
      avoided? Ground Water, v. 7, no. 2, p. 21.
                                                                                                                107

-------
                                           DISCUSSION
The  following questions were answered by A.  E.
Zanoni after delivering his talk entitled "Ground-
Water Pollution  and Sanitary Landfills—A Critical
Review."

Q. These  questions pertain to the closing remarks
that  have  to do with new research:  What research
would you  suggest regarding ground-water pollu-
tion  due to  sanitary landfills? Do you feel that we
now  have sufficient knowledge of the bydrogeology
of sanitary landfills to design, operate, and manage
solid waste  disposal using this method? Also,  why
no more  research activity in the future? Do the
States feel  we know enough now?  This conflicts
with  your  recommendation for new  concepts,
research, etc.
A. To answer the first question, I think the studies
of the type  Illinois is doing right now are certainly
in the right direction. Also more studies on the
hydrogeology of sites would be in the right direc-
tion. I think that with the data available now, we
can  do  a  better job  if more is known about the
hydrogeology of the  specific site  involved. As far
as the degradation of the refuse is concerned, this
mechanism  is  fairly  well  understood.  The big
question is  what happens  to the  leachate once it
leaves the immediate vicinity of the disposal site.
To extrapolate information and data from one site
to another  is very dangerous. We need to know
more of the hydrogeology of the site in question
in order to be able to  predict more accurately what
the progression of the leachate will be and what the
impairment  in ground-water quality will  be  at a
specified  distance from  the landfill site. Did  I
contradict myself?

Q. Someone thought you said, "We need no more
research activity in the future. " The question  is,
"Why no more research activity in the future?"
A. No,  what I said was the survey of States  indi-
cates little research activity in this area will be done
in future  years.  My personal feeling is that there
should be more. There is no question about that.
As for State agencies, there doesn't seem to  be
much interest in this research area.  I feel  it is be-
cause we  have had no long history of troublesome
pollutional  situations  and, quite frankly,  this is
what we  normally respond  to. I  am certain  that
people in the  political  area would be more recep-
tive  to research  activities if  they  were  hearing  of
numerous cases of public health and nuisance prob-
lems resulting from sanitary landfill leachates. Very
candidly,  these   are  the  types  of  things which
motivate politicians and others to act.
Q.  Let me take advantage of the moment and ask
a question. In any of the literature work, have you
found anyone working on  such things  as trace
organics or perhaps carcinogens?
A.  No. I'm not aware of any such work. Trace ele-
ments,  yes. We have some indication of this from
work in Pennsylvania and Brookings. There is very
little work on the degree  of bacterial travel from
landfill sites.

Q.  Couldn 't your miracle be explained by the fact
that most landfill sites are deliberately located far
from habitations?
A.  I  think that it is just simply the environment
provided by the soil regime. The active sites, the
availability  of bacterial degradation activity, the
combination of moisture and air; it is just an ex-
cellent  environment  for the breakdown and degra-
dation  of organics. It  is like a good  trickling filter
in  a sense and  then  beyond  that, it's merely  a
dilution of the inorganics. Most soils provide an
excellent  site for this  active microbial degradation
activity, and  perhaps,  if the permeability  is too
high, then  this attenuation mechanism drops off
because the active  sites are less available.  Under
this situation  we haven't got as much  time for the
degradation  to take place. Of course, when the
leachate starts getting into the rocks and crevices,
none of the attenuation mechanism is available. No
doubt  the distance  that landfill sites are  placed
from habitation  helps, but we should not forget
that nature does an  excellent job of treating these
leachates.

Q.  Might one not  wish to maximize rather  than
minimize  the  rate of refuse degradation if leachate
is being collected for treatment?
A.  Absolutely. No question about it.

Q.  This should be  the most economical approach
if collection of leachate is practiced,  correct?
A.  Absolutely. If the design is such that you are
going to collect leachate, then you bring it into the
treatment plant at maximum rates. In other words,
you maximize biological activity in order to save on
dollars. This is a treatment cost we  are not accus-
tomed  to spending for this  type  of municipal
function,  and,  I would  be  very skeptical about
approving  something  like  this  unless it was  con-
trolled  and regulated properly. We all  know of the
sophisticated waste water treatment plants built in
small towns and  2 weeks after the engineer leaves,
the rural policeman  operates it once in awhile and
the  treatment completely  breaks  down. This  is
what might happen  with this type  of installation
unless it is monitored properly, so  I would never
108

-------
 favor this without good control. I would favor the
 traditional approach with slow degradation because
 then  time is  on your  side.  If you  produce these
 leachates and don't know what to  do with them
 after you produce them, then you have a problem.

 Q. Do you believe that surface water pollution by
 landfills may  often  be  masked by other kinds of
 pollution already present so that  surface  water
 pollution by landfills may often be undetected?
 A. That is  partly true.  I have  not  run across too
 much data on this.  Hughes  of Illinois had some
 data on this situation. He showed colored slides of
 leachate oozing out of the sides of  landfills  and
 then  running into streams.  There  was a case in
 Kansas City where an industrial waste deposited  for
 many  years began  leaching  into a river supply.
 Some of this  happens and goes undetected. Some
 of it  has happened and has been detected but I
 haven't run across  a great  deal of this  type of
 information in the literature.

 Q. Wasn 't  the California leachate data  collected
 under conditions of forced leaching?
 A. Yes. In one of the studies when the water was
 added  to the refuse bins in an amount equal to
 precipitation in the area, no leachate was generated
 so the researchers had to turn on the faucet to get
 some  data. This is the main reason why in southern
 California there is not generally a leachate problem.
 There  is not  sufficient precipitation  to  generate
 significant  amounts  of  leachate. Remember, how-
 ever, that a serious leachate problem can still occur
 if the ground-water table rises  up  in the refuse
 mass. This situation really causes active degradation
 of organics with subsequent leachate production.

 Q.  Can the data that  were collected in California
 be applied to a true sanitary landfill?
 A.  No.  You have to  be very careful in extrapo-
 lating some of these data. This is one point, I think,
 that comes out very clearly from all these studies.

 Q.  Will all sanitary landfills produce a leachate?
 A.  It  depends on  the  amount of  water passing
 through  the refuse. There  have been some sites
 that have  remained relatively  unaltered for   20
 years.  But  we must always keep  in mind that
 anything organic must degrade eventually. It is just
 a matter of time. And if you maintain these semi-
 dry conditions, it  may  not happen  in 2  years
 but it  certainly will happen  in 200 years. It  has
 to degrade if it is organic. It's only the rate that is
 the big variable here.

 Q.  What are the moisture holding capabilities  of
solid -waste?
 A. I don't know a number to give you. This would
 depend very much on the degree of compaction.

 Q. Would you  describe how solid waste disposal
 can  be designated  to operate at a normal rate of
 pollution so that solid  waste decomposition and
 resulting leachate production is manageable.
 A. Well,  I think with  proper hydrogeologic data
 inputs it is possible to arrive at a fairly reliable idea
 of the degree of leachate production expected as
 well as what eventually will happen to the leachate
 once it gets  into the surrounding terrain. I  don't
 think  we'll  be  able  to predict  accurately  the
 progression of the leachate but at least we'll  be
 able to determine if potentially we are in a problem
 area. The type of cover material and degree of its
 compaction are extremely important, for example.
 Another  important item is how  we  handle  the
 surface water.  If you grade  most of  the surface
 precipitation away from  the landfill site and you
 have  a relatively  impervious  top  layer, you  are
 cutting down the degradation rate to a very mini-
 mal  extent and the impact of leachate  production
 will  be quite minimal. There is no wonder that we
 have potential  water pollution problems after you
 see how some  of these landfill sites are operated.
 With the  type  of data  Illinois is beginning to put
 out, we can  approach  this design a  little  more
 rationally than we have in the past. At the present
 time  sufficient  funds  are not being expended in
 the process of  selecting a suitable site for disposal
 of refuse. Some government officials think if you
 spend  more than several hundred dollars for this
 purpose you  are being extravagant. It is just  not a
 very popular governmental activity.


 Q. I'm glad you  said  that  because  it is very
 important  to  the  next  question,  which I  think
 is  a gig. What profession, geologist, hydrologist or
 engineer,  is presently best suited to conduct site
 evaluation and design studies?
 A. I would say a geologist.  If I had $100 and had
 to pick one, I would pick a geologist, or a hydro-
 geologist.  If  I  could  afford  it, I would use both
 because some of the planning, some of the surface
 soil  quantity calculations, the grading, things  of
 this nature, are more in the engineer's bag, but with
 limited funds I'd stick to a geologist.


 Q. Is there a general tendency to locate landfills in
quarries, ravines and valleys that are very important
resources of ground-water recharge? If so, would a
change in approach help to protect ground water?
For example, bailing and stacking the solid waste.
                                                                                                109

-------
A. Absolutely. I think there is this trend  simply
because these areas—just as they  are—are  totally
worthless  for any other application, and if you can
show someone a restoration plan with these pretty
pictures I told you  about before,  where you can
show usable land being developed from some of
these sites,  it's very attractive to the city  fathers
and  political leaders, particularly from a tax base
standpoint.  So these sites are looked at first and 1
agree with the comment that oftentimes  these are
recharge areas, and you have to be extremely care-
ful about using  these sites. I had some experience
with  one  site northwest of Milwaukee that was a
gravel deposit site. It  was used to mine gravel for
about 10 years  and I was  asked to look at it for
possible use as a landfill site and I was convinced
that it could be used. After taking 30 to 40 borings
we found there was a very thick, dense clay layer
underlining the  whole area. The site was simply a
gravel glacial outwash sitting  on  top of a very
impervious  lense of clay.  This  is an example of a
gravel-sandy area that could be used for the land
disposal of  solid wastes. In  contrast to the previous
example,  deep  limestone  quarries with standing
water are generally very poor landfill sites because
this water is usually a direct outcrop of the ground-
water supply.
Q. Do you feel tbat the \\isconsin Department of
Natural Resources  ivill respect your recommenda-
tion to hire a geologist?
A. There has been discussion of obtaining a geolo-
gist for  assistance in two areas: water wells and
refuse  disposal.  Nothing has been done as of this
date. I've talked to  some of the people there and
they agreed with me but Wisconsin, like all States,
is  presently  going through  fiscal  problems  that
demand  just about a total moratorium on hiring.
In summation,  they  have  not listened to  that
particular recommendation yet, but I suspect they
will. The  people I  have  talked  to seem  to be
favorable to that direction.

Q. Could you add to your  recommendations the
desirability of each State establishing ground-water
quality standards to add to those established by
HEW for surface waters?
A. That has been talked about a great deal. I've
talked  to people personally about this. The enforce-
ment  and the  monitoring of  something like this
would be awesome  now. Maybe in the future we
can  look in that direction or  try  this in isolated
cases,  but the degree of variation in ground-water
characteristics from area  to  area precludes  setting
up any kind of rigid standards.  As you can imagine,
monitoring of  the  water quality  would be very
expensive. Nonetheless, there  are people who feel
that this  is the  direction we have to go.  Eco-
nomically I  can't see  how  we can justify  it yet.
It's much easier to sample a moving stream than it
is to get to a ground-water table 100 feet below the
ground surface.
 110

-------
 Effect  of Early  Day Mining  Operations  on
                 _                                  o
 Present  Day  Water  Quality

 by Leland L. Minkb, Roy E. Williams0, and Alfred T. Wallaced
                  ABSTRACT
     Mining operations within the Coeur d'Alene District
 of northern Idaho have been continuous for over 85 years.
 Data presented herein demonstrate that early day mining
 and  milling wastes are  now affecting  the ground-water
 quality in several locations. One of the affected areas is, the
 lower Canyon Creek Basin located in  the Coeur d'Alene
 District near Wallace, Idaho.
     Ground-water pollution of the Canyon Creek Basin
 results  from leaching of old  mine tailings chat are inter-
 mixed with the upper part of the sand and gravel aquifer.
 High zinc, lead  and cadmium  concentrations occur  in
 ground water and soil samples taken from the portion of
 the sand and gravel aquifer containing old mine tailings.
     Analysis  of water  samples from a settling pond lo-
 cated in the upper portion of the study area indicates that:
 the pond water is not the source of the heavy metal con-
 centrations found in the ground water. However, the water
 from the pond's decanting system provides recharge  to
 ground water and compounds the problem.

                INTRODUCTION
 Purpose and Scope of Study
     Interest in  the  environment during recent
 years has focused considerable  attention  on the
 quality of several streams  in  industrialized areas of
 northern Idaho. Much of the attention is centered
 around one of Idaho's  most important industries
 (mining).  Consequently, researchers  at  the  Uni-
 versity  of Idaho and the Idaho Bureau of Mines
 and  Geology have established a program under the
 support of the U. S.  Bureau of Mines to collect,
 analyze, and interpret water quality information so
 that industries involved would have  a  basis for
 corrective measures where  needed.
     Much of  Idaho's mining activity is in the
 Coeur d'Alene River Basin  located in the panhandle
 region  of northern  Idaho  within  the Bitterroot
 Range  of the  Northern  Rocky Mountains.  This
 paper is concerned with a tributary of the South
      Presented at the  National Ground Water Quality
Symposium, Denver, Colorado, August 25-27,  1971.
     "Research Fellow,  College of Mines, University of
Idaho, Moscow, Idaho 83843.
     cProfessor of Hydrogeology, College of Mines, Uni-
versity of Idaho, Moscow, Idaho 83843.
     "Professor of Civil  Engineering, College of Engineer-
ing, University of Idaho, Moscow, Idaho 83843.
 Fork of the  Coeur d'Alene River called  Canyon
 Creek,  the  scene of mining activity for over 85
 years (see Figure 1).
     This study was designed to: (1) report on the
 water  quality of Canyon Creek; (2) report on the
 ground-water  quality of the Canyon Creek Basin;
 and (3) attempt to gain an understanding of the
 causes and sources of high heavy metal concentra-
 tions known to exist within the ground water and
 surface water  of the Canyon Creek Basin. Initially,
 point number 3 was particularly perplexing because
 metal  concentrations in ground water and surface
 water  are considerably  higher than metal  concen-
 trations  in  the effluent from the only significant
 tailings disposal system in the valley.

       DESCRIPTION  OF STUDY AREA
     Canyon  Creek flows in a relatively  narrow
 canyon for approximately 12 miles. Its headwaters
 are  in  the  Bitterroot  Mountains  at  the  Idaho-
 Montana border; and its mouth is at the confluence
 with the South Fork of the Coeur d'Alene River at
 Wallace,  Idaho (see Figure 1). Canyon Creek basin
 has a maximum elevation of 6785 feet and a mini-
 mum elevation  of 2760  feet, giving a maximum
 relief of 4025 feet. Many of the slopes of the  area
 are inclined at an angle of 30°;  the terrain is very-
 rugged. Most  of the Canyon Creek valley is steep
 and  narrow; only one portion of it approaches a
 width of one-half mile. The entire basin occupies
 an area of 21 square miles.
     The population of the Canyon Creek drainage
 basin is approximately 250 (Driscall, 1970).
     Tue main  industry  along  Canyon  Creek is
 mining. At  present, there is one large operating
 mine and several smaller working  mines in  the
 basin.

               BASIN ANALYSIS
Climate
     The climate  of  the Canyon Creek  basin is
strongly seasonal  with mild temperatures prevalent
during the summer months and below-zero temper-
atures common during the winter months.
     Most of  the  precipitation is in the form of
                                            111

-------
            CANADA
WASHINGTON
                      MONTANA
                 , COEUR D'ALENE
  1. CANYON CREEK ABOVE PONDS
  2. POND INFLOW
  3. SOLUMWELL
  4. SEEP1
  5. POND OUTFLOW
  6. SEEP 2
  7. SEEPS
  8. LOWER TRENCH
  9. SEEP 4                 '
 10. LUDWICKWELL
 11. BILTEWELL
 12. CANYON CREEK
       BELOW PONDS
                                   MONTANA
    CANYON
       CREEK
      WAllACE
       SOUTH FORK COEUR D'ALENE  RIVER
                       B
                                            'WOODLAND
                                                PARK
                                                         r\
                                                        _J \
WALLACE
                                                                1OOO
                              "STATUTE MILES

Fig. 1. Map of Canyon Creek Drainage Basin and sampling points near Woodland Park, Idaho.
snow which falls during the winter months. Thun
dershowers persist from late June through part ot
August. In the latter part of Scpicmher and early
spring a short rainy season can usually be expected.
Snow  at  higher  elevations accumulates in large
drifts and,  where  protected  from  the  sun,  may
remain  until  Lite  August.  Some deep  snows in
cirque b.isms  may  persist  until  covered  by  the
next  winter's  snowfall.  Subsurface and  surface
runoff from these drifts  supply runoff to Canyon
Creek and other .streams throughout the year. The
average annual precipitation ot the Canyon Creek
basin at Woodland Park is 44.7 inches; consequent-
ly, ample  water  is  available lor recharge  ot ground
water in buried valley aquifers such  as that along
Canyon Geek.

Geology
     The  rocks of  Canyon Creek  basin  consist
mainly of the Precambrian Belt Series. The rocks
are fine-grained  argillites  and quartzues associated
with  smaller amounts of carbonate-bearing, dolo-
mitic rocks. Quartz and  sericite are  the principal
minerals within the  Belt Series; accessory minerals
          include  teldspar,  muscovite,  magnetite,  ilmenitt,
          zircon, tourmaline, rutile, and titanite. Representa-
          tive chemical analysis of the Belt Rocks is presented
          in Table 1.
            Table 1. Chemical Analysis of Belt Rocks in Percent of
                 Unaltered Rock in Coeur d'Alene District'
O.VJiiV.s /Viv. /;,'
SiO,
AKO,
Fe:O,
FeO
MgO
CaO
Na:O
K;O
TlO;
P,O<
MriO
CO;
H,O
67.4
I ? . 8
2.4
1 . 2 X
2.1 I
1.76
3 '.i
0.49
0.12
0.1)5
2.51
1.35
100.15
            From: Hobhs and others, 1965, p.28.
112

-------
3800'—


3600'_


3400'—


3200 _


3000'_


2800'_
                   LACIAL
                    TERRACE
               CANYON
                CREEK
          -ALLUVIUM
 BED  ROCK
(BELT  SERIES)
                 2345
                DISTANCE (X 1000 FT.)
                      I
                      6
 Fig. 2. Cross section B-B1 of Canyon Creek Basin at Wood-
 land Park showing valley configuration.

     The river valleys are partially filled with alluvi-
 al deposits  which vary in thickness from less than
 one foot to several hundred  teet. Canyon Creek is
 filled with  alluvium  to an  unknown depth. The
 alluvium consists  mostly of unconsolidated  sand
 and  gravel; large rounded boulders are prevalent.
 Tailings from  early  mine  milling  operations are
 mixed  with river gravels within many valley areas.
 Figure  2 is  a cross  section (B-B1 in  Figure 1) of the
 Canyon Creek valley near  Woodland Park which
 shows the configuration of the valley and alluvium.
     There  is evidence of Pleistocene glaciation at
 the head of the  Canyon  Creek basin.  Glacial ma-
 terial is scattered throughout the basin (Hobbs and
 others.  1965, pp. 68-69).

 Hydrology
     Stream flows  within the Canyon Creek  basin
 are extremely variable. High  flow occurs during the
 spring  months coincident with snow  melt and a
 rather low flow occurs during the fall when ground-
 water discharge and  melt from remnant drifts are
 the major source of water. Spring floods caused  by
 rain and melting snow occur.
     The water table along Canyon  Creek is  shallow
 with  springs in evidence  throughout the year, but
 particularly during the spring months. The valley
 areas  of Canyon   Creek produce  an  adequate
 ground-water supply for  domestic use because  of
 the high permeability of  the sand  and gravel  aqui-
 fer. However,  much  of the domestic water supply
 comes  from the  convenient  springs and creeks in
 the basin or from  adjacent  basins because of the
 poor  quality of ground water. Water costs in the
 valley are relatively high.
         PREVIOUS WORK IN  AREA
     Mink, Williams and Wallace (1971) initiated
the first effort at unraveling the pollution problem
on Canyon Creek. Prior to that study, speculation
had attributed high metal  concentrations  directly
to tailings pond  effluent. An increase in zinc, cad-
mium, and fluoride concentrations were observed
in the South Fork of the Coeur d'Alcne River
downstream from the entry of Canyon Creek. A
mean concentration of 3.07 ppm zinc was observed
below the  confluence of Canyon Creek with the
South Fork while upstream portions of the South
Fork  of the Coeur d'Alene River showed  a mean
zinc  concentration of  0.18 ppm zinc. In short,
Canyon Creek adds considerable zinc to the South
Fork.
     Mink, Williams and Wallace (1971) noted that
the concentration ot zinc in  Canyon Creek  was
approximately three times  that found in the efflu-
ent from  the settling pond located near  Canyon
Creek. Springs and seeps  analyzed  downgradient
from  the settling pond gave preliminary evidence of
an increase in zinc concentration in ground water
with distance downgradient from  the settling pond.
     More  recently,  soil  and  sediment  samples
(much of  which consist of old tailings) collected
from  the Canyon Creek bed and  along the  sides ot
the stream have  been lound to contain up to 6.0
percent lead and 4.4  percent zinc. Copper concen-
trations of the  sediments have  a mean value of
approximately 430 ppm.
     Analyses of plants collected from the Canyon
Creek area have  shown mean zinc, lead, and copper
concentrations of 80 ppm, 60 ppm, and   10 ppm
respectively.  These  values  are  about four times
higher than  respective  concentrations in  plants
growing in areas which did not  receive tailings in
early mining days.
     More  detailed  analysis  ot  the  samples of
intermixed soils  and  old mine  tailings will be pre-
sented in a subsequent paper by Galbraith, Williams,
and Siems.

           PROBLEM  DESCRIPTION
      In August,  1968 a series of two settling ponds
were  established along  Canyon Creek near Wood-
land  Park.  The  mill and  mine  wastes are piped
approximately six miles to these  ponds. The valley
area where the  ponds were established is  approxi-
mately 2.5 miles long and 0.4  miles wide at its
widest point.  The ponds are located on the upper
portion of this wide valley area (see Figure 9). The
lower Canyon Creek  valley is filled with extensive
alluvial deposits. These are the deposits on which
extensive  "jig tailings" or coarse  mine wastes from
early day  mining operations have been deposited.
                                                                                                 113

-------
Because  of inefficient early recovery practices,
much of the "jig tailings" are high in metal content.
     Tigging is a mechanical gravity concentration
process.  It is  used to separate heavy grains from
light grains. In its simplest form, equipment con-
sisted of a box without a top and with a perforated
bottom.  A  shallow  bed of grains was formed by-
fluctuating water currents. The heavy grains passed
to the bottom; intermediate mixtures remained in
the middle; and the lightest rose to the top of the
bed. Some of these  devices were hand-operated or
power-driven  units in which the jig was moved up
and down  in the water-, others were stationary jigs
in which the  water was pulsated by plungers or
paddles.
      The jig  tails along  Canyon  Creek were pro-
duced by the concentrators upstream in the canyon
from the late 1880's until the late 1920's and early
 1930's when  it first became economically feasible
 to recover the zinc.
      During  the  years  mentioned, there was no
 apparent attempt made to recover zinc and it all
 went to the  tailings which display the high values
 in zinc that occur in the deposits along this valley
 bottom. Typical tails for this period showed  lead
 assays of  0.90 percent to 1.5+ percent. No assays
 were  performed for zinc in the feed, concentrate,
 or tails, but  the majority of  the zinc obviously
 went with the tails.
      Residents of the area below the tailings pond
 reported  a  rise  in water  levels of  their  wells
 shortly  after  the settling ponds were put  into use.
 Because of this observation and  the  findings re-
 ported above, a  more thorough study of the basin
 was clearlv needed.
            SAMPLING  PROCEDURE
      Twelve  sampling points were established and
  monitored from June 1969 through April 1970
  with major emphasis on the summer months. These
  dates include only the period  after settling ponds
  were  put  into  operation  by  the  major  mining
  company located  in  the  river basin. During  the
  study  no satisfactory sampling point for  ground
  water  in the sand and  gravel  aquifer was found
  above  the ponds. The water table does  not "crop
  out"  at the ground  surface as it does  below the
  ponds,  and no wells are available at an appropriate
  location. However, the  mining company involved
  collected and had  analyzed one sample of ground
  water intercepted during the construction of one of
  the tailings ponds. The  results of this analysis are
  presented in Table  2.
      The samples collected include both the inflow
  and outflow from the  settling pond. Seepages,
   Table 2. Concentrations of Metals in Ground Water
          Above Tailings Ponds'  (in ppm)

Sample :Vo.   Copper   Lead  Zinc  Iron   Manganese

   SP-1       0.008   0.137  5.96  0.067     0.180

1  Data contributed by mining company owning tailings
  ponds; only one sample available.

springs, and wells were analyzed to give informa-
tion on the ground-water quality below the pond.
All  springs  and  seepages  are  listed as seepages;
however,  most such discharge  points are  actually
springs throughout most of the year. Seepage 1 is
located at the lower base of the settling pond em-
bankment which is approximately  50  feet thick.
Seepage 2 is located 80 feet to the east side of the
lower settling pond between the settling pond and
Canyon Creek. Seepage 3 is located 350 feet south
 (downgradient) of the settling ponds and Seepage 4
 is located  2300  feet south of the settling ponds.
 Three  wells analyzed  within  the  basin were  the
 Solum Well located 200 feet  west  of  the settling
 ponds,  Ludwick Well  located  2450  feet  south
 (downgradient)  of the settling ponds and Bilte
 Well located 3750 feet south of the settling ponds.
 Samples  were also  taken  from a trench 2100 feet
 downgradient  from the settling pond.  Samples of
 Canyon Creek above and below the settling ponds
 were analyzed to  determine the effect of settling
 pond effluent and discharging ground water on the
 stream (see Figure 1). All wells and seepages obtain
 water  from the  sand and  gravel aquifer occupying
 the valley.

                DATA ANALYSIS
      The samples  were collected in one-liter poly-
 ethylene bottles which were washed  in a dilute
 HCl solution and  then rinsed with deionized. dis-
 tilled water prior to sample collection. The sample
 bottles were  rinsed thoroughly with  the .sample
 water  at the time of collection. Immediately after
 collection  the samples were analyzed tor tempera-
 ture, pH. and electrical conductivity.
      The samples were then  analyzed for arsenic
 (As), calcium  (Ca), cadmium (Cd), chromium (Cr>.
 copper (Cu). iron (Fe), potassium (K). magnesium
 (Mg),  manganese  (Mn), sodium (Na),  nickel (Ni).
 lead (Pb), and zinc (Zn) using a Perkin-Elmer model
 303 atomic absorption spectrophotometer.

 Acidification  of  Samples Containing Iron, Lead
 and Zinc
       Experimentation with acidified  and unacidi-
 fied samples  prior to analysis for lead  and zinc
  114

-------
revealed that acidification increased the concentra-
tions of both ions. Two explanations are possible:
acidification is known  to minimize adsorption of
these ions onto the walls of the container;however,
experimentation  revealed  that  acidification  also
strips off ions adsorbed to the solids suspended
in the  samples, especially when those suspended
solids are mine tailings. All samples analyzed during
this  project were collected in polyethylene bottles;
consequently, adsorption of  ions  onto container
walls would be expected to be minimal. Therefore,
the high concentrations of lead and zinc observed
in acidified samples are interpreted as being caused
by removal of  ions from suspended matter  in the
sample. Filtering of samples was avoided because of
the fear of the adsorption of heavy metals onto the
filter paper.
     If  the suspended  solids  settle   out  in  the
natural  environment (as in stream channel deposits)
the  ions  adsorbed  thereon  may   never  become
available to aquatic life. On this basis, and because
of the above complications, the decision was made
to utilize unacidified  samples  for analysis. The sus-
pended solids  were  allowed  to  settle  and  the
supernatant was  analyzed.  Inherent   in this ap-
proach  is the assumption that the partition of ionic
species between water and solid was  at equilibrium.
The  data represented herein are believed to be as
representative as  possible of  ion   concentrations
actually in the water, rather than a combination of
ions  in  the water and ions added  to,  or removed
from,  the  water  by a change in the  physical-
chemical  environment   within  the sample after
collection.
     Finally,  extraction techniques were not  uti-
lized in this stud\\ Such techniques are frequently
used to lower the  detectability  limit  of  dissolved
ions  by the atomic absorption  spectrophotometer.
However, it was observed that results were difficult
to reproduce when this  technique was employed so
the decision was  made  to settle for slightly  higher
limits of detectability.

              WATER  QUALITY
     Because of  the  toxic  nature  of  several  ele-
ments  in  high concentration, notably cadmium,
copper, lead, and zinc, emphasis has been placed on
these elements. The area is nationally famous for
its silver,  lead  and zinc production, with copper
and  cadmium being  important  by-products from
the mining operations.
     Interviews with  residents  of  the Woodland
Park area  indicated that water levels in wells have
risen since the installation of the tailings ponds by
the  mining  industry.  Inspection   of  the   valley
downgradient from the settling ponds revealed that
 septic tanks  and water  meters are under  water
 during a portion of the year. Preliminary analysis of
 the ponds revealed that the ponds were not directly
 contributing the  high zinc concentrations found in
 Canyon  Creek. The  pond  water  contains a zinc
 concentration . of approximately  1.4 ppm  while
 Canyon  Creek has a concentration of approximate-
 ly 4.8 ppm zinc (Mink, Williams and Wallace, 1971).
 Since this preliminary study the outflow from the
 tailings pond decant system has been sampled. This
 water has a mean value of 1.1 ppm zinc (based on
 5  samples) which is slightly lower than that found
 in water within the pond.  The seepages analyzed
 from the basin's sand  and gravel  aquifer show a
 high  zinc  concentration  in  ground water a con-
 siderable distance from the  ponds (see  Figure 3).
ZN
PPM
38


36

34


32


30


28


26 -


24


11


20


18


16
                     LOWER

                     T«ENCH°HUDW,CK
                           SEEP 4
                °SOLUM WELL


                   SEEP 1
     CANYON CREEK
     ABOVE  POND
            POND
            INFLOW
                POND
               'OUTFLOW
       1
              3456789
              DISTANCE (X 1000 FT.)
                                      CANYON c
                                      CREEK
                                      BELOW
                                      PONDS
                               BUTE WELL

                                     1O   11
 Fig. 3. Graph of mean zinc concentration vs. distance for
 Canyon  Creek Basin near Woodland Park, Idaho  (1969-
 1970).

 Samples from seepage 2, seepage 4, lower trench,
 and  Ludwick Well are significantly higher than the
 rest  of the samples  in zinc concentration.  These
 concentrations range from a mean of  31.0 ppm
 zinc for seepage 4 to 37.2 ppm zinc for seepage 2.
 Solum Well, which is located about 200 feet west
 of the  final (lower)  pond, displays  a mean zinc
 concentration of  13.6  ppm.  It is not possible  to
                                             115

-------
determine whether this relatively  high  concentra-
tion is related to recharge from the tailings pond
system. We  suspect  (but  cannot prove)  that  it
reflects ground water which has moved through old
tailings above the pond system; the limited availa-
ble analyses of this water are presented in Table 2.
Bilte Well, located downgradient from the settling
ponds, had a mean zinc concentration of 0.6 ppm
which  can  be interpreted to mean that ground
water in the lower portion of the basin is recharged
in  a direction perpendicular  to  the  axis of the
valley (see Figure 2).
     The zinc, lead  and  cadmium concentrations
in ground water not associated with old mine tail-
ings in the Coeur d'Alene  basin is <  .1 ppm (Mink,
Williams and Wallace, 1971). Additional data col-
lected since  that study  support  this  conclusion.
Based  on 14 samples, an  increase in zinc concen-
tration was  found in Canyon Creek when  mean
concentrations  from  samples above  the settling
ponds were compared to mean values below. Mean
zinc concentration  in Canyon Creek above  the
ponds (4 samples) was 3.1 ppm  while below the
settling  ponds mean zinc concentration (14 sam-
ples) was 5.9 ppm.
.34-


.33-


.30 .




.16-

.14 _


.11-
                      Lowe*
                      TIENCH
  Cd
 (PPM),
                            : IUOWICK Will
     Cadmium  concentration follows  much  the
same  trend  as  zinc concentration (see  Figure 4).
Mean cadmium  concentrations ranged  from 0.40
ppm to 0.20 ppm for seepage 2, seepage 4, lower
trench, and LudwickWell. Cadmium concentrations
were below the detectable limit of 0.02 ppm for
Canyon Creek  above the settling ponds, while a
mean cadmium concentration  of 0.04 ppm  was
observed in Canyon  Creek  7800  feet  below the
settling ponds at the  lower end of the valley.  The
inflow  and outflow  of  the settling ponds  was
always  below  the detectable limit of  0.02 ppm
cadmium.
     Lead  concentrations are  also higher in the
seepages than  in  the inflow or  outflow of the
settling ponds. Except for 0.1  ppm lead observed
in one  sample  from Ludwick Well, no detectable
lead (detectable  limit of 0.1 ppm) was found in
Canyon Creek, any of the wells, or in the settling
pond outflow. Mean  lead concentrations (13 sam-
ples) in seepage 1, seepage 2, seepage 3, seepage 4,
and the lower trench range  from 0.6 ppm to 1.6
ppm lead (see Figure 5).
   U-

   1.6-


   14-

   LJ_

 LEAD
   U>_
 ("«)
   Jt-


   .6-


   .4-
                                                                    °SEEP 1
                                                                    cSSEP ^    OSEEP 4
                                                                      °SEEP 3
                                                                            1OWIB
                                                                           c TRENCH
    -       INLOW
    CANTON CK.  »  SOLOM
    A.O«     =  wm
                                                        A.O
                                                                                          CANTON CK.
       I   1  3  4  5  6 " 7   •»   10  11  'I  13
              DISTANCE (X 1000 FT.)
 Fig. 5. Graph of mean lead  concentration vs. distance for
 Canyon Creek  Basin near Woodland Park, Idaho  (1969-
 1970).
.06 _


.04 _
                 OSOLUM  Wilt
                    OSIEP 3
                   CSIEP I
                                        CANTON CK.
   at—                                         -
      CANTON CK.  POND   POND
      AIOVE     INFLOW  OUTFLOW    llltl WELL
     Q.   i   loi    l°;   t   t   a   I   i   I   f   -
         I   13456   7  •  9   10  11  11
               DISTANCE 
-------
ground water decreases to a mean value of 4.27 at
seepage 4 located 2300 feet dovvngradient from the
settling pond. The  pH of Bilte Well is higher than
the other ground-water samples, presumably be-
cause of recharge of uncontaminated ground water
entering the aquifer near the valley sides.
  1.0 _
 Cu
(ffM)
  0,8 -
    POND
   OUTFLOW
               SOLUM
                Will
            POND
            INFLOW
     StfP 4

SffP 3	] LOWER
              LUDWiCK
               Wflt
                                       CANTON CK.
                                        BflOW
              DISTANCE (X 1000 FT.)
Fig. 6. Graph of mean copper concentration vs. distance for
Canyon  Creek Basin near Woodland Park, Idaho (1969-
1970).
  8-
            POND
           INFLOW
   CANTON CK.
  7*. ABOVE
                    POND
                 o OUTFLOW
                            CANTON CK.
                             BELOW
PN
 SOLUM WELL


SEEP
    2
    EP 3
                               ollLTE  WELL


                          0 LUDWICK WILL
                           JEEP 4
                             R TRENCH
              3   *   i   t   7   S   9
               DISTANCE (X  1000 FT.)
                                      10  11   12
 Fig. 7. Graph of pH vs. distance for Canyon Creek Basin
 near Woodland Park, Idaho (1969-1970).
                                             [VIUM
 2700 _
 2600
                DISTANCE (X 1000 FT.)
 Fig. 8. Longitudinal cross section (A-A1) of Canyon Creek
 Basin near Woodland Park showing proposed flow system.
           GROUND-WATER FLOW
     Figure  8 shows a longitudinal cross section
(A-A1) of Canyon Creek  with a  proposed flow
system which is influenced by recharge from set-
tling ponds. The depth of  the aquifer is estimated
because of lack of data on depth of the alluvium
in the valley.
     Before the establishment of the settling ponds,
it is believed the major discharge area  was at the
lower end of the valley where the valley narrows
and  the  alluvium thins out.  The  proposed flow
system shown in Figure 8 indicates that  a discharge
area exists from a point directly below  the settling
pond to  the  lower end of  the valley. Evidence for
this  conclusion  consists of several  springs which
now emerge between the ponds and the  downgradi-
ent end of the  valley even  during dry weather. The
discharge area below  the settling ponds brings
ground water into contact with the upper portion
of the sand  and gravel aquifer containing the jig
tailings.  Virtually  all ground water discharges into
Canyon  Creek before  leaving the valley  because
thickness of the sand and gravel  aquifer thins  to
zero in a narrow constriction above the  confluence
of Canyon Creek and  the South Fork of the Coeur
d'Alene  River.  A water table map of the Woodland
Park area shows the water table is near  the surface
throughout the valley (see  Figure 9).


   DISCUSSION AND RECOMMENDATIONS
     The data  presented herein indicate a ground-
water pollution problem  with respect  to low pH
and high concentrations of cadmium, copper, lead,
and zinc in the sand and gravel aquifer of Canyon
Creek. The concentrations of cadmium, lead and
zinc  are  in  some  cases above the limits set for
domestic water supply by the U.  S. Public Health
Service.  The concentrations  observed  would also
be  detrimental  to many  aquatic organisms. Con-
centrations  of cadmium,  copper, lead,  and zinc
found in Canyon Creek below the settling ponds
are  well above that reported toxic to trout and
other forms of aquatic life  (McKee and Wolf, 1963,
pp.  149, 169, 209, and 294). Of the four elements
above, only copper (which is  not particularly high)
appears  to  be derived directly from  the tailings
ponds within the basin. Cadmium, lead, and zinc
show a  much  higher concentration in  the ground-
water samples  than that observed in samples from
the  tailings ponds. The high concentrations of lead
and zinc in  soil samples, along with the high con-
centration of these elements in ground water in the
sand and gravel aquifer above and below the ponds
in the valley, indicate that the source  of metals is
within the aquifer located below the water table.
                                                                                                  11:

-------
  TOPOGRAPHY    AND   WATER   TABLE   CONTOURS

       FOR   BASIN  OF  CANYON   CREEK, IDAHO
                                                                            1000
                                                                    CONTOURS
                                                               TOPOGRAPHY	
                                                              WATER TABLE	
1.
-->
3.
H .
5.
6.
7.
8.
9.
10.
11.
1 2 .
13.
14.
15.
16.
17.
18.
CANYON CREEK
CONFLUENCE
CANYON CREEK
BELOW
BILTE WELL
SEEP 5
BUSSEL WELL
LUDWICK WELL
SEEP 4
LOWER TRENCH
WOODLAND PARK
WELL
SELP 3
UPPER TRENCH
SEEP 1
SEEP 2
POND OUTFLOW
SOLUH WELL
LOWER POND
UPPER POND
INFLOW
CANYON CREEK
ABOVE
Fig. 9. Topography and water table contours for basin of Canyon Creek, Idaho.
Essentially all of the ground water moving through
the aquifer discharges into Canyon Creek before it
enters the South Fork of the Coeur d'Alene  River
because the thickness of the sand and gravel aquifer
diminishes to /cro in a narrow constriction immedi-
ately upstream from the South Fork.
     The decrease in pH of the ground water down-
gradient  indicates  that  a mechanism  is  present
which  solubilizes the  metal  sulfides of old  mine
tailings. The  sulfide ion then becomes oxidized to
sulfate ion. This  reaction creates sulfuric acid which
accounts tor  the increase in H+ ion concentration
and  the decrease in pH,  The H+ ions can then
undergo  cation exchange  to allow the metal  to go
into solution. This process, along with an evaluation
of  the  role  played by   microorganisms,  will be
discussed  in greater detail in a subsequent paper by
Galbraith,  Williams  and  Siems.   This  proposed
mechanism could account for the anomalous pH
values  and metal concentrations  observed in  the
ground  water of the Canyon Creek basin. Recharge
caused by the establishment of the settling ponds
has apparently raised the water table of the valley
into the mixture  of  tailings and  sediments. This
rise has resulted in additional metal sulfides from
old jig tailings  being  subjected to the mechanism
described above.
     Trenches  from three to fourteen feet deep
have  been  constructed by the mining  company
concerned and  have alleviated the problem to some
extent by lowering the water table approximately
one to four feet. This drop eliminated some of the
seepages in the  valley.
     To minimize  the problem, it is  believed that
discharge of  the settling pond effluent should be
diverted  from  the  unlined ditch  (upper  ditch  in
Figure 9) where it now flows, to Canyon  Creek, a
distance of approximately 400 feet. A similar result
might be realized by lining the ditch.
     The seepage  from the settling ponds  should
be evaluated from the point of view of recharge ot
ground water by the ponds. At present, the amount
of  recharge  being contributed  by this  source  is
unknown.  Experience in  other parts  of the Coeur
d'Alene valley  has shown  that a peripheral tailings
discharge  system  minimizes seepage. The  slimes
118

-------
from the milling operation  are relatively effective
in sealing the ponds. An appropriate technique is
described by Kealy  and Busch  (1971) and Kealy
and Soderberg (1969).
     Deep trenches constructed at an  angle across
the valley and discharging into  Canyon Creek
would further lower the water table below the old
mine tailings (jig tailings) and also help  solve the
problem.
     Immediate measures needed include ascertain-
ing that the  ground  water emanating from  the
springs and seeps in  the lower part of the valley is
not consumed by animals or humans.  The concen-
trations of lead observed  constitute  the primary
reason  tor this recommendation. It is  also advisa-
ble that the ground water along a band downgradi-
ent from the  ponds not be consumed.
     It must  be noted that  it is  not possible to say
with  certainty  that elimination  of  recharge  of
ground water by the tailings pond system will com-
pletely eliminate the pollution of ground water. It
is apparent that the critical factor is whether or not
the old jig tailings in the upper portion of the sand
and gravel are saturated. On the  basis of testimony
by area residents, the pond  discharge has raised the
water table considerably; however, it is not certain
that the water table did not periodically extend up
into the layer of old  tailings prior to the installation
of the tailings ponds. No data are available under
original  conditions on  which to  base  an opinion.
However, it is certain that the tailings pond effluent
is contributing to the problem.
     In addition to  providing an explanation  for
the  heavy metal  ion concentrations  observed in
ground water in the  sand and gravel aquifer of Can-
yon Creek, and in addition to  accounting for the
heavy  metal concentrations in Canyon Creek, this
study  has demonstrated  the need for thoroughly
evaluating all aspects of a waste disposal site  prior
to its use. Had the conditions in Canyon Creek been
understood in advance of  pond construction, the
portion  of  the problem attributed to the ponds
could  have been essentially eliminated by one of
several  design procedures.  The  industry has now
adopted a process of careful site evaluation prior to
use, including  sampling of initial water quality.
This procedure not only facilitates proper design;
it also enables the  company  or other  entity in-
volved to specify precisely what effect its operation
has had on the environment in the event that any
post-installation problems arise.


                   SUMMARY
     1.  The Canyon Creek basin has been the loca-
tion of mining activities for over 85 years. Prior to
1968,  most of  the waste generated by concen-
trating plants associated with these operations was
discharged into Canyon Creek. This study includes
the period of time after the installation of tailings
ponds  to  improve  water  quality. The  study was
designed to  evaluate and  explain the heavy metal
concentrations known to occur in the ground water
and surface water of Canyon Creek basin.

     2. Water samples were  collected  from  12
stations within the Canyon Creek basin  during a
period from June  1969 through  April  1971 and
analyzed for 13 metals along with electrical con-
ductivity, pH, and temperature.
     The analyses showed that the concentrations
of  the elements zinc and  cadmium are  high  in
ground water and  surface water. Lead  is high  in
ground water discharging  at some locations. Lead,
zinc,  and cadmium concentrations are  lower  in
tailings pond  effluent than in  surface water  or
ground water. Zinc and cadmium increase in Can-
yon Creek  as ground water  which  has moved
through the valley's  sand and  gravel aquifer dis-
charges into Canyon where the aquifer thickness
diminishes to zero in the stream channel.
     Soils were  sampled in an effort to obtain  an
estimate of the amount of heavy  metals  contrib-
uted  to the upper portion of the basin's sand and
gravel aquifer by the emplacement of a layer  of
old "jig" tailings during the early days of mining in
the basin. The data indicate that the upper portion
of the aquifer contains up to 6.0 percent lead and
up to 4.4 percent zinc.
      3. The settling pond, established on the upper
end of an extensive sand and gravel alluvial deposit
which  contains old  mine tailings, is acting as a
recharge area for the  basin aquifer and  has appar-
ently  resulted in a rise of the water table within
the basin below the  settling pond.  This rise has
contributed to the raising of the water table into
old mine  tailings  deposited  during  early mining
days. When saturated,  these tailings become subject
to lixiviation by ground water.
      4. The ground-water  quality of  the  lower
portion of  Canyon Creek  is the  direct result of
leaching of the old mine tailings (jig tailings) which
have been deposited by the past mining operations.
As a result of this process, high concentrations of
cadmium, lead, and zinc occur in the ground water.
The process also  results  in a lowering  of pH with
distance away from the settling ponds.
      5. Solution  of  the  problem can be accom-
 plished by utilization of different design techniques
for tailings ponds in this type  of  environment.
Methods described by Kealy and Soderberg (1969)
                                             119

-------
and Kealy and Busch (1971) should prove success-
ful.

     6. It is  not possible to determine what the
natural conditions were because no data are availa-
ble. The water table may occasionally rise into the
layer  of old tailings even under natural conditions.
It is  certain, however, that the ground-water re-
charge  contributed by the tailings pond effluent
intensifies the problem.

     7. This  study emphasizes the advisability of
thoroughly evaluating an  area before disposal of
wastes is  initiated. Without such  an examination it
is difficult to accurately delineate the role played
(or not played) by the waste disposal facility.

             ACKNOWLEDGMENTS
     The  writers are grateful to Mr. Gordon Graig
and  Mr.  William McKee, Hecla Mining Company,
Wallace,  Idaho for their assistance  in  investigating
the ground-water problems in the  vicinity of Can-
yon  Creek.  This project was supported  by U. S.
Bureau of Mines Contract  No. HOI 10088, which is
directed  at determining what types of  wastes tail-
ings ponds can be expected to treat  adequately.

              REFERENCES CITED
American  Public  Health Association and others. Standard
     methods  for the examination of  \\aste\vater.  1965.
     Boyd Printing Co., Inc.. Albany, New York, 744 pp.
Driscall, Marie. 1970.  Wallace Chamber  of Commerce.
     Wallace, Idaho (written communication).
Ellis, M. M. 1940. Pollution of the Coeur d'Alene River and
     adjacent  waters by mine wastes.  Special Scientific
     Report 1, U. S. Bureau of Fisheries, 61 pp.
Hobbs,  S.   W.. A. B. Griggs,  R. E. Wallace,  and  A. B.
     Campbell. 1965. Geology of the Coeur d'Alene Dis-
     trict. Shoshone  County. Idaho. Washington.  U. S.
     Govt. Printing Office, U. S. Geol. Survey Prof. Paper
     478,  pp. 1-69.
Idaho Bureau of Mines and Geology. 1964. Mineral and
     water resources of Idaho. Washington, U. S.  Govt.
     Printing Office. Special Report no.  1, pp. 276-278.
Idaho Department of Commerce and Development.  1963.
     Idaho almanac,  territorial centennial  edition,  1863-
     1963. Boise. Idaho, pp. 1-78.
Kealy, C. D., and R. L. Soderberg. 1969. Design of dams for
     mill  tailings. U.  S.  Bureau of Mines, Information
     Circular 8410. 49 pp.
Kealy, C. D.. and R. A.  Busch. 1971.  Determining seepage
     characteristics of mill-tailings dams by  the  finite-
     element method. U. S.  Bureau of Mines, Report of
     Investigations 7477, 113 pp.
McKee, J. E., and H. W. Wolf. 1963. Water quality criteria,
     2nd ed.. California State Water Quality Control Board,
     Sacramento, California,  Pub. no. 3-A, 404 pp.
McKee, William. 1971. Assistant Manager of Mills, Hecla
     Mining Company, Wallace, Idaho (personal communi-
     cation).
Mink, L. L., R.  E. Williams, and A. T. Wallace. 1971. Effect
     of industrial and domestic  effluents on the water
     quality of the Coeur d'Alene River Basin, 1969, 1970.
     Idaho Bureau of Mines and Geology, Moscow, Idaho.
     Pamphlet  149, 95 pp.
U. S. Bureau of Mines. 1969. Minerals yearbook, vol. I-II.
     Washington, U. S. Govt. Printing Office, 1194 pp.
U. S. Department of Commerce. 1970. Climatological data,
     Idaho, annual summary.  Washington, U. S. Govt.
     Printing Office, v. 73, no. 13.


                  DISCUSSION
The following question was answered by Leland L.
Mink after delivering his talk entitled "Effect of
Early Day Mining Operations on Present Day Water
Quality"  (coauthored  by  Roy  E.  Williams  and
Alfred  T. Wallace).

Q. hit't the significance level of lead and cadmium
in  water much  less than your lowest detectable
limit of 0.1 ppm, .005 ppm Public Health Service
upper limit?  Does this affect your  conclusions?
Perhaps  the  problem  may be  even  worse  than
stated.
A. This can  be true. Our detectable limit  is quite
high and again we're hoping to improve  this situa-
tion with our new equipment. Now  it is possible
that when we receive the results  of a few  samples
we have  sent  out,  the  concentrations of some of
these elements, for instance, lead, may prove to be
present  in concentrations  which  are  above  the
Public  Health  Service  limits  but still  below  our
detectable limits. We may  have to revise some of
our ideas  concerning what  is polluted and what is
not polluted  within the basin at that time. How-
ever, our objective  in this study was to unravel the
ground-water  pollution problem with respect to
sources and causes. We did  not set  out to monitor
exclusively for compliance with standards. Our data
did enable us to meet our objective.
120

-------
 Bull  Session  3—Solid Waste—Its Ground-Water Pollution Potential
Session Chairman: Jack W. Keeley, Chief, National
  Ground Water Research Program, Environmental
  Protection Agency, Robert S. Kerr Water Re-
  search Center, P. 0. Box 1198, Ada, Oklahoma
  74820.

Jack W. Keeley, Lead Bull:
     Dr.  Zanoni will not be able to be with us. The
subject is solid waste  and landfill. By  recognition,
get  up to the microphone and ask your question,
and we will discuss it; hopefully, we'll get a little
controversy going. I hope we can make the thing a
two-way discussion. Who wants  to be first? Does
anyone want to start with anything in particular?

Jack  Hennessy,  Brookhaven  Natural  Laboratory,
Upton, New York:
     1 asked this question  earlier  today of Mr.
Apgar; 1  didn't feel it was completely  answered. If
you  collected the leachate from the  cells, would
you  describe what treatment methods you would
use to neutralize it or to make  the leachate non-
polluting so it could be discarded?

Michael Apgar, Roy F. Weston, Inc., West Chester,
Pennsylvania:
     I'm not sure that this is my field of expertise,
but the materials in the  leachate include organics
which have to be oxidized and this could be done
by aeration  and raising the pH. So, lime treatment
and  spraying or something like  that ought to be
able to help it. Then you've got  high salt content,
and I'm not exactly sure what you could  do about
that except dilute it—that is why 1 suggested that
an acceptable way to get rid of this stuff might be
to put it into a municipal  sewer system. I'm not
sure if that is being done anywhere right  now, but
a lot  of people have been talking about it. 1 know
that there are at least one and probably two sites;
but, I know of one in operation in Bucks County,
Pennsylvania,  where  leachate  is collected  and
treated. But this is a very  expensive process because
you're not through treating the leachate when you
are through utilizing  the site for waste  disposal.
That  is, once you're done putting refuse somewhere
and you've got a leachate collection and treatment
system, you've got to keep  operating that treat-
ment system.
     The  regulations  in Pennsylvania have  now
been  made stiff enough  so that in some parts  of
the State permits won't be granted to operate a
disposal site unless you go to collection and treat-
ment. So, as far  as 1 know these two operations in
the southeastern part  of  the State of Pennsylvania
are the only two that are  operating legally now.
Jack Hennessy:
     Could you mention any areas that are actually
collecting leachate and treating it?

Michael Apgar:
     Yes,  the  Warner Co.  is the name  of one  in
Bucks  County, Pa.; I have  the address somewhere
but I haven't visited their site myself.

Russell Stein, Ohio Division of Water, Columbus:
     We are just getting into landfill monitoring in
Ohio, but I know of one case up near Cleveland, a
suburban area, a town  known as Painesville up on
Lake Erie,  where they are collecting leachate from
a landfill through underdrains which is going into a
sanitary  sewer and being treated  by the  City  of
Cleveland Treatment Plant. Other than that, there
is  another  landfill  in one  of the counties  in Ohio
in which they collect leachate and simply retain it
in an oxidation pond before it bleeds out into the
nearby stream, which does a little bit of good, but
not very much.
     I  think one of our problems is this idea  of
whose  responsibility it is to treat a leachate. You
always have this inevitable time lag in  a  landfill,
particularly a small county landfill, where you may
have four or five years of active use in this landfill.
Five years  after that  you may have a leachate
problem develop and in the meantime the landfill
is closed up. Who is impelled to go back and build a
treatment plant? Or better still, if treatment exists,
and the landfill operator is operating the treatment
plant,  what happens to the treatment  plant and
what happens to the leachate when they finish the
landfill operation?

Michael Apgar;
     Well,  I don't  think anybody has reached the
point where they have  stopped operating a landfill
that is operating with a treatment system now.

Russell Stein:
     I  could give you an example of this. Here in
Ohio we  are  faced  with  one situation  where we
have a landfill up near Cleveland which has a really
nasty  leachate  problem.   It is not causing any
particular ground-water problem because it is in  an
area served by city water, but it's causing a very
bad stream pollution problem. Now, all of the local
residents in this area are trying to get an  injunction
against the operator to shut this down. The State
people are trying to impel this landfill operator to
collect and treat the leachate. But our feeling  is
this, if they get an injunction against this man and
                                                                                               121

-------
permanently close this  landfill down, there is just
nobody available to go in there later and try to
correct  a very serious pollution problem, unless it
would be the State.
Michael Apgar:
     This  is why in a lot of cases the State might
have the authority to  do something about a par-
ticular site.  I know this is the case in Pennsylvania,
but they are reluctant  to close the operation down
because either a  private contractor or the munici-
pality itself is running  the landfill and by closing
them down,  even  if   it is not  being  properly
operated,  the operator  just moves to another site
and the  area that was a problem remains a problem.
Everybody has moved away and it just sits there
and it is a sewer.

Roy Williams, University of Idaho, Moscow:
     We have an easy way to solve the problem in
Idaho. The  operator in one  case, or the only case
that has come to court was simply given an  order
by  the  judge to do a  number of things before he
left the site; and  if he  didn't do that, he would be
in contempt of court. It took care of the  matter.

Michael Apgar:
     Yes,  but in  a lot of these situations it is an
expensive proposition  to correct  any damage  or
potential damage that's already built into the site.
And you  can't very well tell a small  operator  to
either dig up the refuse or collect  the leachate and
treat it  or something like that. And in some cases, I
think there  is another way of getting at this, and
that is  to shift  responsibility back  to the people
that generated the leachate in the first place.
     Is  anybody  here from Illinois? I understood
somebody to say that the liability could be traced
back to the people that were putting the material
into the site if  it could be proven it  was their
material, even though they said goodbye to it when
they gave it to a  private contractor or to the local
municipal hauler at their generation point and let
h>m take it  somewhere.  I'm  not sure if that's true
but that is what I heard.

Jack W. Keeley:
     What is objectionable about having built into
a contract at the outset that these matters be taken
care of in whatever means is appropriate?

Michael Apgar:
     That's  a good idea, but I don't think most
States are to that point yet.
Warren  Hofstra, U. S. Geological  Survey, Denver,
Colorado:
     One thing that occurs to me  is that if a State
or even a Federal agency is in the position of giving
a permit, they can make certain requirements on a
thing that has a relatively limited life, say a landfill
that will only last 10 or 15 years. It seems to me it
would be in order for the governing body to require
piping the leachate to a permanent facility, such as
a municipal sewer. And, if the site was finally filled
or finally used, it would seem reasonable to me that
the governing agency—in this case the city or the
State, or whoever had it would at that time take
over maintenance  of the thing as a public health
hazard.  The other thing would be not to allow any
of these things unless they  were run by a  munici-
pality.

Jack W. Keeley.
     Perhaps we  are talking about three  separate
problems. When a landfill is in existence, the permit
to  construct  this facility  was  based on  certain
existing local or State legislations. If we get worried
about the things  and the State or the community
passes legislation with a measure of control, we are
dealing with private or municipal enterprises, which
I think  legally are probably a little different. If it is
a municipal enterprise, the State agency can simply
say cease and desist. If it is a private individual, the
operator, probably a corporation  if he had  any
sense, could take bankruptcy, you see. Then who is
responsible?  Now  these are questions that  are
pretty well established over the country. These then
would fall  back on the State or the municipality.
Can you pass an ex post facto law when dealing
with private enterprise? I don't think you can. The
State can issue a cease and desist order  but the
operator can take  bankruptcy. Is he then responsi-
ble? My point is to open up or maybe guide this
discussion along these lines. Does anyone have any
comment on the legality?
Roy Williams:
     The bankruptcy part  is obviously not in the
realm of what I was talking about earlier, but let's
say  that a  municipality in Idaho  contracts the
operation of a landfill to an individual and he isn't
bankrupt. I can't see  where there would be any
problem with writing any restrictions on the condi-
tion he must leave the site in when he finishes.

Jack W. Keeley:
     You're talking about before the fact though.

Roy Williams:
     Yes, that's the way  we operate them.

Jack W. Keeley:
     That's a different ball game. That is clear-cut.

Roy Williams:
     Maybe the solution should be operating land-
fills on a contract basis.
122

-------
Paul Plummer, Miami Conservancy District, Dayton,
Ohio:
    On the last  statement, all you have to do is
look at the strip mine operators to see what effects
you are likely to  have and the type of reclamation
proceedings you're likely  to  get.  You know, the
law says now that they must clean up the site and
reclaim the site. Go look at some of these reclaimed
sites. I think we as geologists, hydrologists,etc. are
going  to have to make recommendations  to the
legal and  political people so that some of these
problems  can be  solved by legal  means. How we
arrive at these recommendations, 1  don't know.

Truman Bennett, Ranney  Water Systems, Colum-
bus, Ohio:
    I'm sure that in an individual case the con-
tractual relationship would be very successful, but
as a political fact it would become something you
couldn't live with after awhile because you  won't
negotiate in the  same way with every contractor.
Then everyone who has an axe to  grind has a place
to grind it, and the sanitary landfill then is going to
become a  political question instead of the techno-
logical question which it should  be. Maybe what
we are talking about  is the possibility of operating
sanitary-landfills as public utilities.

Michael Apgar:
    What we are looking at  here, as mentioned in
Dr. Zanoni's paper this morning, is that landfills
haven't yet been  fatal to anybody. So, while we're
concerned about  ground-w'ater pollution, the alter-
native is one that results in letting refuse pile up in
the streets which is a much worse alternative than
letting them go ahead and pollute ground water. I
think that laws are coming and pressure is coming
that will force operators to  clean landfills  up,  to
take care of them in  the same way that strip mine
laws are now taking  effect.  There are  still some
mines that  look  like  they are  in bad shape, but
there  is  a terrific improvement  in others. That
wouldn't have been the case several years ago with-
out the new laws.

Robert  Kaufman, Desert  Research Institute,  Las
Vegas, Nevada:
     I'm supposed to be  taking the place  of  Dr.
Zanoni. The contractual part of engaging in landfill
garbage disposal  can go  three  ways—municipal,
municipal supervision with private contractor,  or
straight private   contractor.  In any case,  except
perhaps the first, it is a matter of contractual law.
To guarantee performance as  in any other business
operation  where there's a risk factor or competence
factor, there is simply a matter of posting  a per-
formance  bond. The  most serious problems from
solid waste are the future operations. Last time I
heard it, Utah was not concentrating on the exist-
ing industries and polluters, but rather on control-
ling future polluters or potential polluters. They are
interested  in  future industry and how it is tooling
up with respect to air, land and water degradation
as a by-product of its operation.  So, if we look at
sanitary landfills and the well documented mount-
ing refuse volumes, the object is not so much to
attack with  an ex post  facto law  which  legally
probably wouldn't hold up, as it is to go after the
coming problem.
     And  this is simply a matter of when negotia-
tions come up for contracts for refuse hauling, the
contracts  should be  written  with  some meat in
them.  I know in the Las Vegas  area the contract
from  the  city  and one  carrier who has had the
contract for  over 10 years and wants to negotiate
for another 10 years simply calls for hauling the
refuse  out of the city. Whether he piles it or burns
it—the city says that's the county's problem. This
is childish. Each is trying to pass it on to the other.
So really  it  is a matter of strong contracts with
performance  bonds for any dereliction or abroga-
tion of the  contract. Now as far as going  after
existing landfills and polluters, 1 don't  think you
can tell them to stop. That's ridiculous. New York
City didn't stop  for  too  long; no city has. It's a
matter of putting a compliance  schedule into any
hearing's policy, most commonly with air polluters
and big water polluters. They're  still polluting, but
six months from now they'll be  polluting less, and
still less a year and a half from now.  This is a
reasonable way to go after it. If  a contractor feels
he can no longer provide that kind  of service and
put in this special equipment, then the contract has
to be  renegotiated.  It is  a  bilateral contract. The
contract conditions are changed if the State or city
says you now have to put in this or that for control
or diversion  works.  So what it really comes down
to, when  you want service in solid  waste disposal
and a first-class refuse disposal operation, you are
going  to pay for it. The best way to get into  a bad
contract is to not have competition when the specs
are drawn up and the contracts bid  on.  The worst
condition is to have one bidder, because he names
his prices and the city feels they are stuck with it.
But there is nothing more secure than a municipal
contract and especially with refuse. It is going to be
there always and it is increasing.
     So my feeling is that we know enough now to
operate landfills. It is really a matter of legal and
legislative controls and just  good,  plain  contract
law to get it to operate.  If we  are  working with
details of one site  here vs.  another site there, I
                                              123

-------
firmly believe that there is enough known now to
design a landfill to operate with certain pollution
outputs. And I don't think that digging up the old
skeletons to stop landfills is really the heart of the
problem. The future is the heart of the problem.

Russell Stein:
     I'd like to comment on our situation in Ohio
because  I think  we have a real problem in solid
waste because it is a pretty populous State. I work
primarily in  evaluating new landfill sites  which I
don't think is really that much of a problem now;
we have these under control.  One is required to
obtain  a permit from  the  State before one  can
operate a landfill. So we have some control here.
The problem is this in Ohio—the permit system for
landfills was  effective July  1, 1970. Prior to that
time  we had  about 600 open  dumps  and  a  few
landfills. Now  what have they done since then?
Many  of these open dumps have been  shut down
and there have been about  200 permits issued for
new  acceptable  sanitary landfills,  yet there  are
about 350 open dumps  still  operating  because
they can't come up with a site which is suitable for
the technical people in  the State. They will choose
a site, and we say, no this isn't acceptable. So, con-
sequently they continue to use their open dump.
We have a time lag here of over a year.  I think our
pollution problems go back to  a lot of these old
dumps  in  which  the  local people applied for a
permit to convert an old dump to a sanitary landfill
because  it was the only or maybe it was the best
site they had in an area.
     It still boils down  to economics, and you
simply can't  impel people to pay much more for
solid  waste disposal than they are now because
they just won't accept  it. They'll take it out  and
dump it along the highways like they used to do,
and probably still do in a lot of areas. I think this is
one of our main problems. It seems like we techni-
cal people in the State set our standards pretty high
for selecting a suitable waste disposal  site, and I
found myself kind of  giving in  a little bit from
this respect because there  just aren't  this  many
sites available; or if they are available,  the people
just won't accept the cost of trying to develop the
sites.
Richard Pearl, Colorado Geological Survey, Denver:
     This comment pertains to  Wyoming where a
city had the advantage of expert hydrologic advice,
but disregarded it and located their dump right
over their city water field. The U.S.G.S. office in
Cheyenne pointed this out to them and did almost
everything but go to court. The city fathers felt it
was better to face future water pollution than to
face air pollution because the other site, the dump,
was up-wind from the city of Cheyenne; and there
is some burning at this dump, so they chose to put
it over their well field. The question I want to ask
Bob Kaufman is how are you really going to get
around  it when a city ignores expert advice? There
is no State control up there.
     And then I want to ask Mike Apgar a question.
The water level is about 150 feet or deeper up there
in this  arid to semiarid environment. You didn't
show conclusive  proof  that your  leachate was
reaching the water table this morning. At least that
was my interpretation of it. Is there a chance that
this leachate is going to get down there or is it go-
ing to be purified enough so that it isn't going  to
bother the water quality if it does get down there?

Robert  Kaufman:
     Well, we have a democracy  and for all its bad
and good points, if an  issue comes up, I think the
best route is a public hearing where there are expert
witnesses, not  paid experts, not out-of-town ex-
perts, but groups like the League of Women Voters,
or various environmental groups. There was a de-
velopment  in  Las Vegas recently  where  over  a
million  dollars was spent on site evaluation, just
evaluation, and the return  on the investment was
reported to be somewhere in  the  order of 100
million  dollars. Now the only objections put before
the  State were put forth  by  unpaid,  nonprofit
interest groups who  were in violent opposition  to
changing water quality standards in the Colorado
River, and other variances  that this firm wanted
against  county ordinances regulating the develop-
ment. The point I make is that an informed public
which may consist of just interest groups can bring
out a lot of these sticky little issues like putting a
dump over a well field.
     Now, looking at the other side of the coin  it
might not be  a bad  idea. George Hughes has one
landfill  site in northeastern Illinois on the banks of
the DuPage River,  I  believe it  is,  and he  figures
there is something like a  maximum of  5 or  10
milligrams per liter increase  in chlorides from that
landfill  site. Now there isn't a person in the world
that can taste 10 milligrams per liter of chloride.
I'm not making an excuse for it, but you could put
this over a giant well field and the landfill is dump-
ing out maybe 20,000 gallons a day and you are
pumping 5 million gallons a day out of the well
held. It's inconsequential.
     Landfilling is a community service.  We  have
buses outside now dumping diesel fumes and we are
accepting the service. George Hughes has proved in
Illinois  that 90 cents out of every dollar is spent on
transportation. For every solid waste dollar, trans-
124

-------
portation eats up most of it. Either go to rail haul
and get the economies of mass transfer, or pick a
site close in, and the money saved can be used to
divert the leachate.
     These  are the sorts of issues and the questions
that  have to be generated for each issue, either on
the part of the State or with local groups. I think
the worst cases develop when  things are done in
closed  meetings and  an under the  table contract is
let for 10 years of service which has no concern for
these matters. And I think  in the interim, before
we have State agencies  that are  well  funded to
inspect and regulate  and so  forth,  that we have to
depend on the local semipolitical groups and action
groups to work with. I  don't see any alternative
because city  and municipal leadership is going to
adopt the least cost solution in most cases, because
nobody likes  increased taxes and you  don't get
voted back  into office with increased taxes.  '
     We have \ State law in Nevada now and the
only reason k was passed is because the agricultural
interests were exempted; they  contribute a major
part  of the solid waste in the northern part of the
State.  They said  if you want support for this bill,
exempt agricultural waste. It was  a limited victory
because the legislature only  meets every two years;
and if we wanted any solid waste legislation, it had
to be  without  the  agricultural  people included.
Maybe in the  next two years we will pick them up.

Michael Apgar:
     I  think that Bob already answered my part of
your question—that you can't really be sure in this
type of setting that the landfill is going to have any
more  than  minimal  effect  on  the ground-water
quality. In  a semiarid environment with an aquifer
150  feet below the surface, you probably won't
even notice the effect of the landfill up there, but
I'm guessing.  I wouldn't want to  say that for sure
without looking at the site a little more and know-
ing a little  more about it. If they want to go ahead
and do it that way, it doesn't sound that bad.

Roy Williams:
     Mike,  I  would like to comment a little further
on the problem of saturation. Your paper is a super
important paper  because it is one of the few that
has  reported on the  water quality in  the un-
saturated zone.  I'm  a little  concerned that it may
not  be unsaturated  because the  200 foot water
level  doesn't  mean the water levels are 200 feet,
and  the absence of nitrates  suggests anaerobic con-
ditions which may suggest saturation. I would like
to hear comments on your evidence for  unsatura-
tion.
Michael Apgar:
     Generally there aren't  any  major  perched
systems between the land  surface and the water
table, except that we do have evidence that water is
standing in  one of  the refuse cells in the lower
topographic position.  Sometimes  there  is  water
ponded in the upper refuse cell because  after a
certain  amount of  decomposition,  the  floor  of
these refuse trenches becomes impermeable enough
to hold water. Soil  moisture contents of samples
that we cored beneath these refuse cells only range
from 10 to 20 percent, and there is nobody in the
area that  has  a hole in the ground  that has water
standing in it  anywhere between 200 feet and the
surface.

Don  Langmuir,  Pennsylvania State  University,
University Park:
     We have a water table map which  was not
shown  today  which indicates the general water
levels  throughout  the whole area and  they are
around 230 feet or so with very little variance. It is
quite a flat water table. So as Mike said, there could
be  locally  some  saturation  when  the  soils are
appropriate for this, but in general this is not the
case.

Jack W. Keeley:
     Let  me  interrupt just a moment and ask a
question.  Can  we  necessarily assume  that because
a zone is unsaturated, it is aerobic?

Don Langmuir:
     Whether  the  zone is aerobic or  not is really
going to be a  function of rates. It is a function of
the  rate  of depletion of the oxygen by  organic
material versus the rate of its introduction into the
system by movement through a soil or in fresh
soil  moisture. In most cases if there is much BOD,
say  10 milligrams  per liter  or so,  and  you are
talking about a soil, the depletion rate is faster than
the  rate of introduction and anaerobic conditions
will exist because the rates  are favoring this. We
found in our  system, for soils that are heterog-
eneous, aerobic zones adjacent to anaerobic zones
because of the nature of the soil  in the system. So
it isn't a simple matter.


Roy Williams:
      For  future studies, I think it would be advisa-
ble  to put in  a few well points just 10 dispel all
doubts that you really are dealing with unsaturated
conditions.  If somebody comes  along two  years
from now and wants to use your data in a decision,
there will be absolutely no doubt.
                                            125

-------
Michael Apgar:
    We have a similar study involving a program of
spray irrigation of treating secondary waste effluent
from the  University Sewage Treatment Plant on a
game land area. We are dealing with the same type
of soil, same  underlying bedrock, and same water
table  conditions at depths between  200 and 400
feet below the surface. Pan lysimeters were installed
in the soil,  and they just couldn't get samples in
them. That  is why we are using suction lysimeters.
Now there  are pan lysimeters  in use in  the spray
irrigation  study, but they only collect moisture
after the soil is essentially saturated by heavy spray-
ing directly over where the lysimeters are located.
I am not saying that there aren't positions in the
soil, especially after periods of heavy recharge, that
aren't  saturated. There may  be small perched sys-
tems on  clay  lenses and unweathered bedrock
ledges  and  things  of this nature, but in general
you don't get much water, if you put a hole in the
ground, you're  not going to get much  out of it.
You can  be fortunate enough to end any hole or
piezometer  on an  impervious ledge and have water
standing down there on a regular basis. But it just
depends on where you locate the end of this thing.
And to get samples we went to suction lysimeters.


Jeffrey Sutherland, Williams and Works  Engineers,
Grand Rapids, Michigan:
      I'd like  to ask Don Langmuir and Mike Apgar
 a question  related to their paper this morning. I'm
 unsure of  how you  are  able to determine that
 leachate  from the upper refuse cell contaminated
 the ground underneath the lower one. It seemed to
 have had a rather  shallow  topographic slope be-
 tween  the two, and no perched water. Therefore it
 doesn't seem to me that much of a route for hori-
 zontal transport was present. Maybe I'm missing
 something here.
 Michael Apgar:
      I'm sorry if 1 gave you that impression. The
 figure I used in the talk was a plot of depth versus
 conductance  beneath  the  cells.  But the  instru-
 mented cells in question are on opposite sides of
 this dry stream bed and so we're not saying that
 leachate from one actually goes down and reaches
 the  sampling  points beneath the other.  But rather
 there are a series of cells all along both slopes and
 the  cells located above the valley bottom cell on
 this side of the stream  bed may in fact move down
 and make any soil moisture contamination problem
 beneath the lower cell  look worse than  it might be
 if it received a contribution only from that cell.
Jeffrey Sutherland:
     So there is enough of a stratification  in these
soils or sediments to get some lateral movement
from one cell to the next?

Michael Apgar:
     Well,  I  wouldn't  say  it is  a very regular
stratification. The soil bottoms slope downhill and
the topography is going that way  and there is an
outcrop located about  700 feet away in which the
dip of bedrock is. in the same direction  as the slope
here. But of course the soil itself, even though it is
a residual soil with some lenses, layers, and  regulari-
ties, is fraught with all kinds  of collapse  features
and mix-ups that developed as the rock weathered.
So it's not a simple situation where you  can put
things in at the top and let them float downhill.

Jeffrey Sutherland:
     1 have another question.  I'm new in the land-
fill site evaluation business myself. In Michigan I'm
acquainted with the regulations established by the
Michigan Geological Survey  which  is one of the
two State regulating agencies. We have  a lot of
apparently homogeneous dry  sand terrain without
much in the way of low permeability layers like
 clay  to  afford some kind of a natural collecting
 substrate for gathering leachate into some kind of a
 treatment system. I wonder if anyone  has any
 experience with bentonite or any kind of liner that
 can be harrowed or disked into sand to minimize or
 perhaps  eliminate  any  significant  liquid  leachate
 percolation?
 Paul  Goydan,  Koppers  Co.,   Inc.,   Pittsburgh,
 Pennsylvania:
      Of course today many of your State agencies
 require  bentonite  as the lining to use in lagoons
 and other types of installations where you expect
 contamination of ground water. Bentonite may be
 purchased from a number of agencies. Wyoming
 bentonite, which is generally regarded as one of
 the best bentonites as far as being impervious, is
 pretty much a sodium mommorillonite clay. You
 often get involved in Arkansas bentonite and some
 of these other States that claim to have bentonite.
 You are dealing then with a mixture of clays; and
 although there may be a certain degree  of mont-
 morillonite  this is often calcium montmorillonite
 which is not  as impermeable.  Some  research has
 been going  on  as far as taking  a clay such as a
 calcium-rich  montmorillonite with a mixture of
 kaolinite and chlorite  or whatever else  you may
 have and impregnate this with salt water  supplying
 the sodium.  This  can  be used  as  an  impermeable
 type bed;  and, of  course, through time,  your
 lattice  structures in  the clays will  expand  and
  126

-------
improve your barrier.  Did that answer your ques-
tion or were you  talking about absorption rather
than sealing?

Jeffrey Sutherland:
     I was asking whether or not the properties of
sodium bentonite or sodium montmorillonite actu-
ally  do  provide a maximum seal to minimize or
absolutely prevent the formation of leachates.
Paul Goydan:
     I've found a statement that 1(T7 to 1(T9 centi-
meters per second permeability is expected. This is
the permeability range of a good Wyoming benton-
ite. This is considered "virtually" impermeable; and
without going to  some cementing  material  or
manufactured material this  is pretty good;  and
with an expanding lattice, with water, you don't
have too much to worry about.
Don Langmuir:
     Might  you than  have  to worry about  the
nature of the  leachate  itself altering the character-
istics of the montmorillonite if it  is a calcium or
magnesium  leachate? Ion exchange can  alter the
permeability  of  your  liner  if the leachate-liner
combination is not compatible.

Joe Miller, New Jersey  Geol. Survey, Princeton:
     We have  a similar problem. I was talking with
Mike about  it. The County of Camden is going to
make a county sanitary landfill on  450 acres in an
area where  it  is 27  feet to ground water. We are
trying to  figure out what the heck  we can  put
down to keep the stuff from going down. So far I
don't know whether your bentonites are going to
stand up, because we don't know what they are
going to  put  in this landfill.  Any  ideas on some-
thing like this? What about putting an impervious
layer with perforated  pipes and then transporting
the leachate back into a municipal system. Is this
acceptable?
Grover Emrich, A. W. Martin Associates, Inc., King
of Prussia, Pennsyfvania:
     This raises a very interesting question because
we are now completing the  engineering  and  con-
struction of a limestone quarry that will be turned
into a sanitary landfill. The  quarry had about 30
feet of water in  the  bottom of it.  Submersible
pumps have  been installed  to  lower the water
table and maintain  it  below  the  bottom  of the
quarry.  The quarry is being filled with rock  and
earth to bring it up  about 15 feet above the water
level and then an impermeable liner will be put in.
We are disking soil cement in the upper foot of the
soil and then  we will  put on a specially prepared
tar pitch material that will be sprayed across the
bottom of the quarry. It will be sloped to a central
sump from which the leachate will be pumped to
the surface; it will be treated and  then discharged
into a nearby sanitary sewer.  This quarry will be
reported at the AAAS meeting in Philadelphia this
winter.  If any of you are interested in details I'll
be here for the rest of this evening and tomorrow
morning.
     The quarry should  be in operation toward the
end  of September. Unfortunately  Murphy's Law
has come into play and we  are now 9 months be-
hind in completion of the quarry. Murphy's Law
says everything that can go wrong will go wrong,
which included an  operating engineer's strike for
three months. Power lines  were not there and it
took three months to get them  in and things like
that. We are also  doing  the same thing for another
quarry which  is a cement quarry  north  of Allen-
town,  Pennsylvania.  Cement rock  up  there  is
fantastically tight. All the people use cisterns, not
because their  water is polluted  but because they
don't have any  water in their wells. We will go
through  the  same sequence  of  dewatering the
quarry which has about 5 feet of water in it, build-
ing  the  quarry  floor  up,  putting in the  liner,
collecting the leachate, treating it and  then dis-
charging it to a sanitary sewer. Another landfill is
in operation  in  southeastern Pennsylvania with a
liner in it. This  has  been in  operation since last
October. They also will be collecting leachate and
treating it. In that case  they will discharge directly
to the Delaware  River  Estuary.  The  technology I
think  has  reached the point where  we  can hy-
draulically isolate landfill leachate from the ground
water, treat the  leachate and handle it  just like
another industrial waste.
     There is a possibility of putting a. cover on top
of the landfill. This has  been tried many times. I've
worked with Pennsylvania regulatory agency setting
up  the regulations and doing all  the hydrogeologic
investigations. In fact,  this  Penn State study was
something that the Pennsylvania Dept. of Health
initiated by going to Penn State. There was another
one we initiated at Drexel University. My staff and
I have looked at  100 or  150  landfills  designed
specifically in an attempt to put  clay covers over
them to keep the water out and no leachate from
forming. They have been notoriously unsuccessful-
I'd say 100%. There are several reasons for this—the
decomposition of organics and other  materials  in
the  refuse will produce a liquid. Of course if you
have no moisture in  it, you might delay  this.  In
most cases  they use a clay  cover which is im-
properly installed. It  is wet, it swells, water runs
off, then  it dries off  and it cracks. I've seen  some
cracks that are 2 or  3 inches  wide and many feet
                                            127

-------
 long so that the next time it rains all the water runs
 down  the cracks. There is  very little that runs off.
 Another problem you have is  that even with an
 impermeable clay layer on top you  still have  your
 gas generation. Some of this gas may move laterally;
 unfortunately, it  will build up  pressure and will
 crack  your liner. I cannot recommend the idea of
 making it impossible for leachate to form. I've yet
 to  see it  work; and, unfortunately, it has  been
 sold. Maybe somebody else has seen it work. I'd be
 interested in anybody that  could fill me in on this.

 Ted Hudson, Radian Corporation, Austin, Texas:
     Has anyone done any work on the equilibria
 that are set up between the absorption of the vari-
 ous  organic compounds that are going to be pro-
 duced  in the  leachate and what effect this has on
 the  permeability  of a clay liner?  What are  the
 equilibria that are set up and what does  this do to
 the  permeability of the clay liners that people are
 talking about  using?  Which is preferentially ab-
 sorbed, the organic, the sodium, or  the inorganics
 that are in the landfill?

 Don Langmuir:
     To my knowledge, there has been no detailed
 study  of  sorption of the  organics.  The charged
 species you would primarily be concerned with are
 going  to be anions, so it's anion exchange.  Most
 clays are  cation exchangers so  that they're not
 going to play an important  role. It is more likely to
 be a factor of the accumulation of these materials
 on trench bottoms. Mike can tell you a little about
 this. There is  a physical clogging effect that these
 organics will have on clay liners,  where leachate is
 trying  to  flow down through the bottom of the
 trench. You'll  get slimes and gels which prevent the
 flow of the leachate, which will decrease the per-
 meability. Since in most cases  cation exchange is
 predominant,  the inorganic cations will be the
 favored species in exchange. The sodium, calcium
 type thing will affect the permeability, of course.

 Ted  Hudson:
     Mike,  can you refresh my memory on the pH
 of some of these leachates?

 Michael Apgar:
     The pH in leachate ranged from about 5 to
 6.5  normally.  However, in  one  case where we dis-
 solved  a cement floor and maybe due to some other
 causes, it got as high as 8.0 or 8.5.

 Ted Hudson:
     I know there is some work on some chromium
oxide absorption on sodium montmorillonite clays
that  MacAtee  and Shaw did several years ago  and
these things are quite  strongly absorbed  on clays.
 128
 Also Hawthorne did some work in which he looked
 at  some positive nitrogen containing organic com-
 pounds. They were  involved  in looking at these
 montmorillonites  as  cracking catalysts for various
 operations. Therefore, they  are quite strongly ab-
 sorbed by these  sodium  montmorillonite clays.
 They did  not look  at  the  equilibria  that were
 established. They  were not concerned with it and I
 was just wondering if anyone here was.
      The other question I have is concerned with
 your  Eh  values.  I  would  like for  you or  Dr.
 Langmuir  to  elaborate a little more fully on  the
 oxidation reduction potentials and what value you
 placed in these redox potentials in your interpreta-
 tion of your data.

 Michael Apgar:
      I don't think anybody that has been studying
 leachates from landfills has made the jump to try-
 ing to figure out any kind of equilibria absorption
 relationships  between organics and  soils because
 most people don't even bother with identifying the
 organics. I can remember reading one study where
 this was done and we're thinking of getting  into
 this at Penn State,  but ground-water geologists with
 a  chemistry  background  are  generally  inorganic
 chemically oriented. It is kind of a sticky business,
 and not much fun  either,  to collect this  highly
 organic contaminated slop and take  it in  the  lab
 and play with it. To fool with the organics is some-
 thing that  hasn't appealed to many people yet.  I
 don't even know if they have considered doing it.
     I can remember a study in West Virginia where
 column experiments  were  carried out.  Leachates
 were put into concrete  cylinders and water was
 leached through this refuse and collected from the
 bottom at a spigot or drain and then the organics
 were identified or separated  chromatographically
 and identified as  being mostly acetic  or butyric
 acids. As for carrying  this further, I don't know.
     I'd  start  answering the second  half of your
 question but  I see Dr. Langmuir is getting very-
 excited about  doing this so I'll let him take it.

 Don Langmuir:
     I was  going to comment a little on  what you
 were just talking about, Mike. This organic stuff in
 leachates is so complicated that you  can find any-
 thing you like in there. You'll find what you look
 for.  There are not just organic acids.  There are
 carbohydrates, fats, proteins, you  name it. You
 could spend a lifetime  trying to  discover which
 things were there. And that's why it's so difficult to
figure out  what anything is doing in terms of its
effect on, for  example, trace  metal transport. This
kind of thing we're going to try and get into in the
next  few years.

-------
     I'd like to talk about the Eh thing for a min-
ute—as I said this morning in an answer, we worried
about this quite a bit. If I can get back to the be-
ginning on  Eh, it might be useful. An Eh measure-
ment is going to be a measurement of a zero cur-
rent flow  at  a particular potential.  There  is no
current flow and you find that potential at which
this current will not flow using a platinum electrode
and  a reference electrode. For this to occur there
are two voltages involved; there's an oxidation and
a reduction voltage potential. You need reactions
which are reversible to  get a meaningful Eh meas-
urement and this can be an awful problem if you've
got a whole  series of different species,  some of
which can  be  reduced but can't be oxidized, and
conversely.
     Now this is what a leachate is in many cases.
You've got nitrogen species which can be oxidized,
you've got sulphur species which  may  be or may
not, and you've got a variety of reactions which are
not reversible.  But there is  a way out  of this. In
many cases a leachate is very, very rich in iron, and
the total iron  values  Mike  was talking about are
chiefly ferrous iron.  It has  been  found  that the
reaction between ferrous iron and ferric hydroxide
is ubiquitous in the leachate, because of the iron in
the soil. This reaction is reversible and it goes fast.
There is  a  measurable potential from ferrous iron
to ferric hydroxide and back. And  this is what
seems to be controlling our measurements. It makes
good thermodynamic  sense  in  spite of  the mess a
leachate  is  to  make this assumption. We can  com-
pute things about  this reaction from  the measure-
ments we make of pH,  Eh, of our assumed solubil-
ities for  the oxides, and it works.  We didn't show
the plots today. We have plots of measurements in
the leachate and they parallel right on boundaries,
ferric hydroxide, ferrous iron, if you know what
those look  like on an  Eh, pH diagram. They make
pretty good  sense. There's  one  up  there Mike's
waving if anybody wants to look at it later on. So
we don't want to take these numbers and say they
are thermodynamics  in the  pure  sense, but they
work reasonably well. For example, they seem to
tell us that  we will not have nitrate when we're way
out of the nitrate field in that diagram—for example
when we're way down in the ammonia field. They
give us a pretty good semiquantitative  view of what
is in there and what to expect.
Paul Goydan:
     Did you  look at the reduced and oxidized
species  of  nitrogen? Were they  able to give  you
similar type indications of the Eh, and also of the
oxidized and reduced sulphur species?
Don Langmuir:
     We  didn't see oxidized species  of nitrogen
except at the very beginning of the  study when
anaerobic conditions were setting in,  and at  that
time I don't think we had Eh  measurements. So
most of  the study  conditions  were  anaerobic
enough so that we had no nitrate or nitrite.
     We don't know exactly what the organics are
but there is a tremendous literature now of organic
interactions, as you know, with clays; and at Penn
State, George Brindley  has  done a great deal of
work of  this kind. Of course,  the gap between
knowing what the organic is in  a pure system and
what it really is in leachate is a long way.

Ted  Hudson:
     My  only thought here, as  Dr.  Langmuir has
pointed  out,  is  he  doesn't really have a pure
thermodynamic ferrous, ferric system  in which he
makes his Eh measurement or his redox potential
measurements, which is  fine.  It  doesn't really mat-
ter because he gets the information he is interested
in. I would not like to trudge out there and collect
these samples either. Nor would I like to use them
in the laboratory. But  if one  made  model com-
pounds  and  then  made these studies  you would
have some basic understanding  of the competitive
nature  of the  reactions  between  the cationic
organic  materials and the cationic  inorganic ma-
terials. This is the point I was trying to make.

Don Langmuir:
     In  the  future, what we plan to do is look at
some of the  organics as they move  down through
the  soil—look at  the  BOD and break  it down in
terms of certain types of organic chemical species—
and  see how they vary in amount  and type with
depth. This should give  us some kind of an idea of
what is going on.  We'll relate this to  trace metals
and to absorption if we can.

Jack W. Keeley:
     I'm going to get in trouble here. We've talked
about how we like to go out and measure pH and
chloride  and sulphide,  etc.  I asked this question
this morning  and I think it's one that we very often
ignore, probably because we are afraid of it or we
don't know how to handle it. I can at least imagine
situations of zero BOD, zero chlorides, zero sul-
phates, zero  this, that and the other, and we have
some highly carcinogistic materials, or their degra-
dation products,  that  totally escape  us.  I know
DDT degrades anaerobically. We don't know what
the degradation products are. They may be more
soluble. They may be more toxic. The world is full
now of antibiotics and  estrogens, of birth control
pills, and  diethylstilbesterol which  we feed to
                                           129

-------
animals.  I think we tend to  completely overlook
this  field because it's difficult. Do  we  have  the
latitude to be that pious to simplicity?

Ted  Hudson:
     Also overlooked are things that are precursors
to carcinogens.  There are many compounds that
react with other things that then create a carcino-
gistic compound. So I call these, or lump  these,
into  what is called a precursor to a carcinogen. An
example of this is nitrite. If nitrate is present, you
have the possibility of nitrite. Nitrite can react with
various  and sundry  compounds to form nitrosa-
mines which are known as carcinogenic compounds.
I just throw this out to say that the subject matter
is even  more  complicated in that you  have also
these precursors to carcinogens.


Jack W. Keeley:
     Then we throw all of the degradation products
of all of these together and we've got a mess. Does
anyone  wish to comment on this before we move
to other questions?

Paul Plummer:
     We are probably sampling what we can meas-
ure.  A lot of these things, even  sampling for trace
metals, gets extremely complicated. And once you
get outside of the laboratory into a practical field
situation, it becomes almost impossible to get what
you  consider to be a valid measurement of some of
these substances. There is room for a lot of work
on sampling techniques and methods for measuring
these many  trace  elements  which we are intro-
ducing into the system.

Jack W. Keeley:
     I can't accept that because something is diffi-
cult  we in the ground-water industry can ignore it.
I have been in the field a good deal myself. There
are techniques available and  there are possibilities
to develop new techniques. What I am suggesting is
that we  in the industry should be forward thinking
enough to start  getting into these.  I realize you
can't go out in the field with a  bucket and a rope
and  concentrate  samples to where we can shoot
them into a gas  chromatograph and look at these
things. I'm suggesting here that  we start looking at
them. Let's  lead the surface boys a little bit.

Don Langmuir:
     There are a couple of ways we can go at this
whole thing.  From major species  we have various
defenses; for example, if we have charged species to
see if we've got everything in there, we measure the
total anions and the total cations. They've got to
equal each other. If something is in there we don't
know about, we'll find out that way. So insofar as
we  have  major  species over 2,  3  or  4 percent,
perhaps, we can spot them often  this way.  This is
the  inorganic charged  species.  We've  got total
dissolved  solids and we've got conductance. We've
got a variety of  total measurements as a means of
getting  at the summations of other species in the
system. Beyond  that point, when it comes to trace
species, we've got to go to other techniques  such as
spectrophotometric scanning or neutron activation.
There  are various new  devices and new ways of
getting  at a variety of things that may be in there.
We ought to develop more of these techniques; for
example,  scanning for the trace species. That's the
big problem, as you say. What is there in the trace
amount which can hurt us? That's where we need
the most advancement, the most research.

Bob Kaufman:
     I don't want to drag down the scientific level
of  our preparation  for Dr.  Greenfield. I'm  not
arguing with  any research goals. The point  I'm
making though is we have to service a community
problem.  NWWA has come up  with well  drilling
specs, completion specs and so forth; maybe they
are not the best but they are the best the  NWWA
could come up with at  this time. The point is that
there has been enough  work done on the  hydro-
geology and chemistry of landfills that  we  can
provide a lot of service and advisement to practical-
ly any  major  problem  in  the country. And I dare
say  if we had an interchange like this tonight in
front of a community  of concerned citizens at a
local hearing, these people would be very astray as
to which course  of action to take.
     I'm  not  arguing with anything you've  said,
Don, or the gentleman  from Texas; but I think as
others have pointed out today, Al Zanoni for one,
the  garbage  trucks are coming down the street.
Now we  can bury it or we can burn it. Those are
the  only two  alternatives we have, and  there's
enough knowledge to provide a reasonable factor
of  safety for burial, the  sanitary landfill.  I can't
overemphasize that enough because to go and say
we need  more study, we've got  to find this out,
we've got to find that  out, is not doing a service
now. It may be in 10 or 15 years but it's not doing
it now. I don't  see why we should go into multi-
million dollar landfilling operations over a 20 or 30
year time frame with a  $500 site investigation. The
techniques are available to design landfills to meet
certain performance criteria and when all the other
information comes along  later  as to exact knowl-
edge, fine. Maybe I'm just a marginal scientist, but
we have a today problem  and more study, the data
 130

-------
of which and conclusions of which are available in
five years, won't meet today's problems.

Warren Hofstra:
     One thing that has impressed me a lot is that
people are  looking at these things from  a  very
practical, today standpoint.  But I question whether
the water system  in the humid crowded portions
of the country will stand the total impact of solid
waste disposal, septic tanks, industrial wastes, and
all of the other things that are going into the water
system.  I  think that as a practical matter we will
have to develop the best sanitary landfills that we
can for  the time being.  But I do feel that we are
allowing a five percent change in the water quality
with this landfill. I feel that we are, in the name of
practicality  and economy,  perhaps overlooking a
more thorough  and a more complete control  of
solid  waste  and all waste. Perhaps we are being a
little  short  sighted considering a  tremendous  in-
crease in population, and I think we may have to
reclaim these materials.

Bob Kaufman:
     I agree, the average site investigation  today is
absolutely  minimal. 1 don't think  we are in dis-
agreement. What I am trying to say is we are not
using the tools that are available. This is the prob-
lem. If a community goes for any kind of a major
capital investment  such  as  landfilling,  it's a 10 or
20-year operation in practically every situation. A
town of 250,000 will spend a million dollars a year
in landfilling. Now, a million dollars a year is a
capital investment that should be handled carefully
with  long-term goals.  For instance, if we put in a
landfill in a human zone, we can control infiltration
to a certain  degree. There can be tile systems put
in as the refuse goes. It can be incremented over the
years as the landfill expands. After  the infiltration
is reduced, either through control of runoff  or
proper vegetative cover or using it with a very tight
asphalt  cover  for  parking lots,  we  use  the tile
systems to collect  leachate; and if things still get
out of hand, diversion wells can be put in. We have
talked about dewatering before refuse is put in. We
can lower a water  table after it's affected.  I don't
see anything wrong with this. An engineer, for all
his skills, plays every bet he can.  And when he is in
doubt, he  puts in a factor of  safety.  Geologists
think that they're  putting their  tail between their
legs if they  start talking factors of safety. And yet
geologists should be the last ones to be exact. Who
can  say how a bentonite line is going to  behave
after  10 years of leachate  contact? We can't wait
10 years to find out.
Dick  Rhindress,  Pennsylvania  Dept. of Environ-
mental Resources, Harrisburg :
     I am directly involved with our sanitary land-
fill program. I just walked in so I'm not sure what
has been  said. But picking up the shreds of the
conversation here, we've tried to tackle  this by the
horns and run with it.  We agree that there is  no
single system.  You can come to us with  practically
any plan under the sun to run a landfill. Yes,  we
have guidelines which say  you should run them in
these basic manners.  But  when it  comes to engi-
neering  and geology and hydrology, if  you've got
some new plan,  we'll listen. We're fully  open to
anything,  whether it's underdrains, lining a quarry
with whatever you want  to line  it with, putting
down an asphalt layer on flood plain gravels, what-
ever you have to do to collect leachate, or  if you
have a good site where there is a reasonable zone of
aeration like Mike's site up at Penn State. We still
feel you can probably get away with the natural
renovation system. The sites are limited but this is
a possibility. We're also open to all the new research
we  can  get. We've been in on  the support  of the
Penn State study, and we've  had input  into the
Drexel  study  and gotten output.  We're  in  close
contact with the Illinois people and everyone else
that's doing  any of this  type of  work.  We take
every little fragment we can get, and we're going to
be a clearing house to help you people  around  the
State come up with some good sites.
     In  the idea of site investigations, we have what
we call  modular forms. They are forms  to be filled
out in great detail on geology,  ground water, soils,
engineering plans and so forth. The geology section
of those  forms has been written  so that only  a
geologist can understand it. There's about 17 pages
of geologic data  that we require before we'll even
look  at your site. So we are at least forcing  the
landfills in Pennsylvania, the new ones and the  old
ones that  have to be upgraded, to come in with one
raft of geologic data before we'll even talk to them.


Jim  Urban,  Agricultural  Research Service, Uni-
versity Park, Pennsylvania:
     It  seemed to me throughout much of this
discussion this common  thread  of getting  rid of
waste by  dilution seems to occur. I wonder if we
might  consider  a  pollution potential  somewhat
analogous to  a dam.  If a  dam  is close to  a city, it
requires a higher degree of site investigation than
one in  the middle of a strip  mine- area  50 miles
away from the nearest population. What do you do
when you have  an  area  where  the ground-water
quality  becomes  impaired to the point  where you
reach, say nitrate nitrogen  at  15  ppm? Where do

                                             131

-------
you stop your inputs? Do you stop putting in land-
fills at that point? Just where do you draw the line
on inputs? Can you allow these inputs to occur?

Michael Apgar:
     I don't think you really know what the impact
on ground-water quality is going to be until you go
ahead  and do it.  It's just like aquifer modeling
programs  right now,  or in the development  of  a
model for any kind of a hydrogeologic system you
want to play with. You really don't know how the
model is doing until you  go ahead and do it and
check  the results. Once you see that you're  con-
taminating water from a landfill in  an area,  well,
you're contaminating the water. I  don't know  if
that answers your question or not, but in a lot of
cases  you  really can't be  sure how much  con-
tamination you're going to get.
     I  want to make a comment also on some of
the implications of the Penn State study that we've
been doing. Sure, beneath at least one of the cells
that we're instrumenting we see that the leachate  is
not being very effectively renovated as it percolates
down  through the soil. It is of horrible  quality to
begin with and it seems to be going down to about
50  feet now. But in  other settings the leachate  is
being effectively aerated and diluted so  that really
it is not a problem by 20 or 30 feet. And you have
different things going on within the same site. This
depends on how much water you get in the cell and
what kind of refuse materials you've got there any-
way because it is heterogeneous all over the place;
and, as far as an impact of this landfill on the water
quality immediately  beneath the site, there's  a de-
terioration in ground-water quality that we feel
confident is due to the landfill. Now, of course, we
only have the one monitor well which is a problem.
     In any case, whether or not landfills are going
to affect  anybody beyond the  landfill at the next
water supply or something like that, I'm pretty sure
that it's not. The next water supply might be for
watering  a  golf course half a  mile away and  is
probably  not even in the same flow direction. By
the time you get the water that far away it will
probably be mixed up and diluted with other things
so that it's not going to be a big deal. They're not
going to kill anybody further away.
     I'd also like to take this opportunity to change
the subject.  Here we are  out in Colorado and no-
body  seems to be very worried  about leachates
from mine tailings, when it has already been shown
that such leachates have killed a horse and maybe a
man, even though he doesn't want to know what's
killing him. Whereas leachates  from landfills  are a
more widespread problem, we  haven't really  dem-
onstrated that they are of a toxic nature. We're just
talking about increasing the solids' content of water
or making the water kind of smelly, and you can
get around this. If leachates foul up your water
supply to the point where people won't drink it
anymore, they  can always treat  it and  this will
cost them money on that end rather than the other
end. And that's not a good thing. I think we ought
to give Leland Mink a chance to talk a little more
about his studies with the metals from the mine
dumps.

John Rold, Colorado Geological Survey, Denver:
    There's been  a lot more concern about the
problems of leachates  from mine tailings than is
apparent on the surface. Yet I think we  have less
problems from mine tailings here in Colorado than
we have from  sanitary landfills.  Zanoni's paper
today gave the  impression that sanitary landfills
were not really much of a problem in general. The
same comment  has been made again tonight. When
I was down at  the office  this evening doing some
work,  I saw a  news release from  U.S.G.S. which
was quite  interesting  in  light of our discussion
today. It's from Earth Science dated August  22.
Many of you will  probably get this. On  the third
page it says "Although there  are several hundred
documented cases  of pollution caused by  leachates
from solid waste disposal sites, few contain data on
the magnitude  of  the  pollution and its effect on
the hydrologic  cycle." We  have several  hundred
solid waste disposal sites  here in Colorado in  the
alluvial gravel of the South Platte River.  We can't
say it's in the past because there is a  new one
starting in Adams  County right now. It's  the same
thing,  a water-filled gravel pit that's going to be
used as a sanitary  landfill. How much damage  has
this done? I don't  know except that below Denver
and downstream from Denver for quite a few miles
you can't drink the ground water. Now  what  has
caused it? I know of two monitoring attempts on
sanitary landfills. One  of  them was an old landfill
and all they could  tell was that the water above the
landfill was just as polluted as it was below  the
landfill. So you'd have to  trace it farther  upstream
to  find out what  the first landfill really  did to it.
And another  one, of course, just  went  in and
proved what we thought;  leachate was coming out
of the landfill and going into the gravel, but nobody
seemed to want to do anything about it because we
couldn't prove  that it killed anybody. But nobody
really  dies of drinking polluted  water;  they just
quit drinking it, whether it's surface water or sub-
surface water. So I don't think that it is a very good
argument that either Dr. Zanoni or you used. Water
is our lifeblood here in Colorado. And we  look very
longingly at the East where you get dilution. There
 132

-------
isn't anything  we  do  with our water here that
doesn't hurt it. We  just can't do anything with wa-
ter but what the  water is worse  when  we get
through with it than it is when we start. Water from
the South Platte doesn't start  to  clean up  again
until it gets to Grand Island, Nebraska, and that's
too far to bring it  back up to Colorado to use it.
So we  have no dilution factor at all. Going out the
west slope to California  is even worse. They are
beginning to complain even more than Nebraskans
about the type of water we're giving them.

Don Langmuir:
     Leland,  can you give me some details on the
techniques that were used in sampling for the trace
metals in  your study? We've had some difficulty,
and  we're trying to learn techniques ourselves at
Penn State on  trace metal  sampling and  analysis,
and  I'd be curious  if you could elaborate  on how
you did this.

Leland Mink, University of Idaho, Moscow:
     We ran into some problems in  sampling, espe-
cially the mine waters. The natural waters within
the environment didn't pose too much of a  prob-
lem. As one probably  noticed, I didn't  mention
anything  about acidification of the sample  upon
collection. We  collected in  a polyethylene bottle
and analyzed for a  few  parameters in the field and
then transported them to a  lab. Within  the labora-
tory we just allowed the sample to settle, took the
supernatant from the top and analyzed on the AA.
The  problems we  ran into on  acidification  were
numerous. First off, we experimented on acidifica-
tion  and found that even though the acidification
will  minimize  the  absorption  on  the container
walls, we  also were stripping ions off of the sus-
pended solids contained in  the sample. And  upon
filtration  we  removed a lot  of the material within
this.  Also we found  acidification of some of the
mine waters  caused a precipitate to form which
completely eliminated some of the parameters we
were looking for. So we  found that acidification
posed more problems  than nonacidification;and by
allowing it  to  settle, we assumed that we  were
approximating  conditions  we  would  find  in  a
natural environment. When  these sediments would
settle  out they would  not be available  to the
aquatic life.

Don  Langmuir:
     The settling was before acidification?

Leland Mink:
     Right—the acidification was experimental. We
never did acidify the sample.
Don Langmuir:
     You have another problem of absorption with
pH changes in plastic bottles because of gas transfer
through the bottle  walls. A pH change will affect
the absorption and can change the trace metal con-
tent in the supernatant before you even acidify it.
This is a thorny business. People have argued against
using plastic bottles  because they  are so  much
worse than glass bottles in this regard.          #

Paul Plummer:
     I have a question on standards. We've talked
a lot about sampling and various levels which have
been set by the U. S.  Public Health Service, Drink-
ing Water Standards and others. Are these standards
based on  total water, or are they  all on filtered
water? This  is a  problem  we ran  into. We  are
sampling from wells.  Do we sample it the  way it
comes out  of the well, or do we filter it through an
0.45 micron filter which, again, is very arbitrary.
And how  then, if it is filtered, does this relate to
what the standards may be?  Does anyone have any
comments on this?

Don Langmuir:
     I think  the  U.S.G.S. normally refers  to  the
0.45 micron filter when filtration is considered, and
it is normally on a whole sample at that point. You
take it through  that filter, then you are considering
what is in the sample.  But  that's an awfully  im-
portant point to think about too. For example, are
you concerned about what  is transported  in sus-
pended  sediment in stream load? Because  it is a
genuine pollutant if it  is toxic and on the sediment.
It  should be considered. Although it may come out
of the water supply where  you have coagulation
and filtration before it goes into the water distribu-
tion system, it still is transported  and it  still is
worth considering. What you  really  have to do is
say that the trace metal or toxic material is smaller
than 0.45  micron or  bigger than 0.45 micron. If
you don't do this you're in trouble.

Bob Kaufman:
     I'd like to know  how many people are drink-
ing 0.45 micron filtered water out of their wells?

Roy Williams:
     I'd like to comment on that  problem. The
EPA currently is suggesting that Idaho, for instance,
write standards in  terms of what they call total
metals which means acidifying thf  sample, which
strips things off and  busts  up sulfides and gives
phenomenal concentrations  of zinc, let's say, in
water. Very few mines can operate with standards
that are being proposed when analyses are done in
                                                                                               133

-------
this fashion. The problem of transporting is not so
cut and  dried. The ions that are  attached  to the
sediments are undoubtedly transported, but we've
got analyses of both the sediments and the water,
and there is very little relationship. For instance,
lead which is high in the sediments, is zilch in the
water. So  when you're  setting  water  quality
standards it makes very little sense to say let's take
the total. Let's go pick up a sample during flood
stage  when things are  murky. You will get a wild
reading which is not indicative of what is really in
the water.
     The other thing that bothers  me a little bit is
that  the  procedures  being  used  now are  quite
variable throughout the country. They need to be
standardized  in  terms of monitoring  procedures.
We need some nationwide conferences among the
monitoring agencies because you  get things you
can't  compare from one  part of  the country to
another as a result of the different procedures that
are being used.

Jack W. Keeley:
     I think Bob Kaufman made a very profound
introduction and  I'm surprised that someone didn't
follow it up. Are we not really talking about what
use is to be made  of this water? We  take it as we get
it. If  we're  talking about a  streamflow that is
variable in terms  of its silt load which in turn gives
variance  to the amount of material that is sorbed,
and this  is  what we're drinking, aren't we really
getting back to perhaps a  statistical base of setting
quality standards?

Roy Williams:
    Well, most people are no longer (in the North-
west anyway) talking  about drinking water stand-
ards.  Those are  way  too high for aquatic life,
particularly the kind of aquatic life that the pres-
sure groups in  the  Northwest are interested in
preserving. And even if you do consider a statistical
base you're still  stuck with  some  technique that
gave you a number with which you have to live.
     I might  make  one more  comment on this
problem. Personally, I  feel that the way to go is to
let the solids  settle out. One thing  we didn't do,
which we should have done, was let them settle out
for a  definite, constant period of time and then
analyze the supernatant.  You've got the contact
between  the solid and the water  to  give  you a
constant  equilibrium  which  you  don't have any
other way.

Don Langmuir:
     Can  I comment on the  chemistry of that? If
you have a  plastic bottle with a sample in it and
you let it sit on the shelf for a week, it will breathe.
134
Most ground waters are higher in carbon dioxide
than surface waters by a factor of 10:100 or more.
It is going to go out through the bottle walls, and
the pH change  can commonly be half a unit or
more.  If you're talking about absorption  on the
sediment at the  bottom of the bottle equilibrating
with the liquid above it, it's going to be messed up
by  this pH change because  the  hydrogen  ion  is
competing  with the metals.  So this is kind of a
risky way to go. If you can get it to settle in a day
and refrigerate it while it is settling, that may work.
Or if you use glass bottles and fill them to the top
that will work.

Paul Plummer:
    The criteria which I used was if this is the way
the people  drink it, this is the way we will analyze
it. In other words, if the people are drinking water
from the well and this has not been filtered through
an 0.45 micron filter, this  is the way I want to ana-
lyze. Now the problem I have run into is, does the
data that I am getting correspond with the limits
which  have been set?  I feel they probably don't. I
suspect that many  of these standards were set at
some   detectable limit, whatever year they were
established. We could detect cadmium to 0.01 mg/1
so that's the  standard. Since that time we have
added  a zero to the thing and now we can  detect
cadmium  to 10 mcg/1 but I don't think  this  is
necessarily the case in a lot of analysis. Still, it is
the total consumption of say lead  or zinc that is
important.  If the material leaves the sediment, it
becomes available to the biota and their processes.
It is well documented that fish and other organisms
concentrate metals; then maybe we really want to
know what the total is, including the sediments as
well as the filtered water.

Roy Williams:
    I  think  the significance  of the precipitation
business that Leland  mentioned can't  be  over-
looked. Not many people have messed around with
mining wastes, and I  can  take you to an effluent
which,  if acidified, you  could watch  the solids
precipitate  out of solution into the bottom  of the
bottle. I can only take you to one place where you
can watch it. I can  take you  to other places where
you will see changes take place. Somebody needs to
look into this in  terms of a research project.

Jack W. Keeley:
    We seem to be burning out after two  hours.
Are there any questions?

Joe Miller:
    To get back to our so-called sanitary landfills,
I haven't heard  any mention  at all of what these
operators bring into the landfills between midnight

-------
and reveille in 5,000  gallon  tank trucks. This is
rather a common practice. Any comments?

Michael Apgar:
     I wouldn't say that  most municipal landfills
are guilty of having exotic industrial wastes dumped
into  them  while  nobody is watching.  It depends
upon what type of industries you have  in the area.
Who knows what they're doing with their wastes?
This can be a  problem because some of the wastes
that  come  out of industrial sites  are extremely
toxic and tricky to handle.

Joe Miller:
     It's not only the industrial, Mike. I'm thinking
of a pharmaceutical waste  which I know was put
into  the outcrop of a formation, as I said, during
the designated hours between midnight and reveille.
There are no monitoring wells.

Dick Rhindress:
     I  think I can comment at  least on a partial
solution to this. We have recently surveyed by mail,
and we're following up with on-site inspections by
our industrial waste people, finding out what prod-
ucts are made, what  by-products are  likely,  and
what the wastes are going to be. And we're asking
what they  are doing with this specific waste. If we
know it is the kind of plant that always generates
phenol, for example, we go in and ask  specifically
what they  are doing with the phenol. Half the time
they  tell  us they're  taking it  to New Jersey or
something  like this. At least this is a partial  solu-
tion.  I  realize we're not going to get the gospel
truth out of all these companies, but  it is a start
at preventing  the midnight  to dawn dumper. They
still exist though, I realize.

Bob Kaufman:
     Jack  has asked me to  take  over for a  few
minutes. Any more questions  on the  landfills? I
have a comment for John Rold. How many publi-
cations has the U.S.G.S. had on sanitary landfills?
I know of one. You made the comment that the U.
S.G.S.  had documented 700  cases of  pollution
from sanitary landfills.

John Rold:
     They said they were documented.  They didn't
say they had done it.
Bill  Seevers, Geraghty and Miller Inc., Port Wash-
ington, New York:
     The only point I raise is the fact that sanitary
landfills are associated with our populated areas.
This is where  we live. Mining problems are in the
relatively unpopulated parts of the country. There
are regions in  the Northeast where we are simply
running out of space; and in  a sense, this is part
of the problem with sanitary landfills. We've got
the physical problem of the landfill, but we also
have the space problem. So this is why 1  view it as
a rather serious situation.

Dick Rhindress:
     Aren't we really saying  that  we've  got to
dedicate some land to landfilling, and are we saying
let's choose the best techniques  we have and use
them in landfill design. Or, are we saying that we
have a fairly good handle on how to clean up the
leachate from  landfills by collection and treatment
systems or natural renovation at very good sites.
We do have a pretty good handle on it, and it is
something we can control.  In  the case of  mining,
we still don't really have our technologies down
anywhere near where we do on landfills.

JackW. Keeley:
     Are there any pertinent questions?  The hour
is getting  late. I'd  like  to make  an attempt at
summarizing very briefly. First of all, with respect
to landfills, it appears that we do have  sufficient
technology available to install  sanitary landfills. We
should use this technology as  best we can  because
the garbage truck is coming. We can make  an anal-
ogy here  to all types of engineering designs where
we enter  into them knowing  that there  is  a possi-
bility of failure. We can, with the knowledge availa-
ble, reduce this probability. Further, we realize that
there are things yet unknown,  and we will con-
tinue to improve both our techniques to measure
and our abilities to control. We have pointed out
that there are  several techniques available to handle
leachates  from landfills. One  may  or may not be
appropriate at any one location and we are  mindful
of this. We are also  mindful that in  dealing with  a
ground-water  resource  we must be  a little more
careful because a ground-water resource once pol-
luted remains polluted for a  protracted period of
time in comparison to a surface-water resource.
     With respect to mine pollution, it seems that
we dealt  mostly with our ability and inability to
sample, measure the contents of that sample, and
understand what that measurement describes.
     Now then, are there any additions,  omissions,
or corrections? Then let's adjourn.
                                                                                                 135

-------
Methods of  Geologic  Evaluation  of
Pollution   Potential  at  Mountain  Homesites
by James P. Waltzc
                   ABSTRACT
     Development of mountain homesites is accelerating
in the Rocky Mountains of central Colorado. These home-
sites often require individual water wells and sewage dis-
posal systems. Unfortunately, the widely used septic tank-
leach field system generally is not suited for use in the
mountainous  terrain where soils  are  thin  or  missing.
Although current federal  regulations call for six feet  or
more of soil at the leach field site, many of the individual
sewage disposal  systems now in  operation in the Rocky
Mountain Region of Colorado fail to meet this requirement.
Sewage effluent  at these sites may directly enter bedrock
fractures and travel large distances without being purified.
As  a consequence, contamination of  streams, lakes, and
ground water  from these  malfunctioning leach fields has
become a problem of increasing magnitude.
     Investigations of geologic, topographic, and hydro-
logic conditions at over 100 homesites in the Rocky Moun-
tains of north-centra! Colorado have resulted in the develop-
ment of objective criteria for evaluating pollution potential
at mountain homesites. In addition, the results  of these
investigations  indicate that contamination of water wells
may be  decreased significantly where geologic conditions
are  considered in the selection of sites for leach fields and
wells. Although  the results of these  studies should  be
considered preliminary, they do tend to confirm that the
orientation of jointing surfaces in the bedrock significantly
affects the travel  path of contaminants.

     The press of our growing population and the
increasing  affluence of our  society have resulted in
more and  more people buying homesites in the
scenic and relatively isolated mountainous area  of
central Colorado. These mountain homesites often
require individual  wells and sewage disposal sys-
tems. Because  many of the privately owned sewage
disposal systems are poorly situated or improperly
designed, wells and streams in the mountains are
susceptible to  contamination. Current federal regu-
lations  provide the principal constraints on design
and construction of septic tank-leach field systems.
These regulations  are  intended for use in  regions
where soil is six feet or more in thickness beneath
the leach field area. Many of the  individual sewage
     Presented at the  National  Ground Water Quality
 Symposium, Denver, Colorado, August 25-27, 1971.
     ^Assistant Professor, Department of Geology, Colo-
 rado State University, Fort Collins, Colorado 80521.
disposal  systems now in operation in the  Rocky
Mountain  region  of Colorado  are  in technical
violation of  these  federal regulations  because of
unsuitable soil conditions.
     The mountain homesites are usually character-
ized by exposures of bare rock and areas of thin or
discontinuous soil. This condition permits water
and any contaminants which are present to directly
enter fractures in the rocks. The filtering and puri-
fication of leach field waters can occur where soil
or similar  material  receives the effluent,  but the
fractured crystalline rocks which underlie many
mountainous areas  do  not effectively filter the
percolating  effluent. In addition, the rate  and
direction of contaminant movement  through frac-
tured rock may be difficult to identify. Therefore,
the location of a leach  field on a tract of land may
be  critical if contamination of well water is to be
avoided. Current procedures, however, base selec-
tion of a leach field site largely on topographic and
convenience factors.  Where the subsurface consists
of fractured crystalline rock, this is not  a  satis-
factory practice.  It should be clearly understood
that the rule of thumb which calls for placement of
a leach field downslope from a water well does not
provide much  insurance against  contamination of
the well water in mountainous terrain.
     Investigations of over 100  homesites in the
Rocky Mountains  of north-central Colorado indi-
cate  that  contamination of  water wells  may be
decreased  significantly  where  geologic conditions
are considered  in  the  selection  of sites for  leach
fields  and  wells. Two  separate  but  related field
studies in Precambrian  igneous  and  metamorphic
terrain, supported by the Office of Water Resources
Research, have provided data which illuminate the
role of geologic conditions  in the travel of ground
water and contaminants in fractured media.

    DISCRIMINANT FUNCTION ANALYSIS
     In the first study (Freethey, 1969), 28 moun-
tain  homesites  in north-central Colorado  were
selected for detailed analysis of variables thought
to  be related to pollution  potential.  See Figure  1
for general locations of study  sites. The homesites
 136

-------
were separated into  two groups: the contaminated
sites (Group  I) consist  of homesites where con-
tamination of the well water  can  be  documented
and a pollution source can  be identified; the un-
contammated sites (Group II)  consist  of homesites
where no contamination can  be detected but where
a  pollution  source  is  present.  Placement  of a
particular horncsite into either Group 1 or Group II
was done on  the  basis of a water quality test  for
Escbericbia  coii bacteria. Seven of the  twenty eight
sites studied  had  wells  which  were contaminated
with E.  coli bacteria.
     LARIMER
     i   CO.
                                 WYOMING _
                                "COLORADO
RED FEATHER
 LAKES AREA
 BASE  LINE  40*N
                         BOULDER
 Fig. 1  Location map  of  areas  studied. Areas are  cross-
 hatched.

 Selection of Variables
      Quantitative measures  of geologic, topograph-
 ic, and hydrologic variables were collected at each
 homesite.
      Geologic variables were devised to show the
 probable direction of effluent movement in relation
 to the position of  the  leach  field and  the  well.
 Field data included measurements of the strike and
 dip of all major joint sets, foliation directions, and
 other fractures.  These  data  were combined to
 obtain a "resultant" strike and dip which could be
 used as an approximation of the direction in which
 fluids would move  through the unsaturated frac-
 tured media. To take into  account the effects of
spacing and width of fractures on permcabilitv,  a
weighting procedure was devised by which fracture
sets which appeared to be relatively impermeable
would be assigned  a ranking  ol one and fracture-
sets  which  appeared to  have high  permeability
would be  assigned  a  ranking of  five.  Fractures
              o             C"
which appeared to have a moderate permeability
were  assigned  rankings of two, three or four de-
pending  on their apparent permeability. The "re-
sultant" dip direction is obtained graphically using
the rankings given to each measurement. Figure  2
shows an example of this procedure. The vector
which represents  the   first measurement  with  a
ranking of one is drawn  only halt as long as the
vector for the second  measurement, which  has  a
ranking of two. After the resultant of one and two
is obtained, vector number three is drawn, which
because  of its ranking  of three,  is three times as
long  as number one. The final resultant dip  direc-
tion is obtained by drawing another parallelogram,
the sides being measurement three and the resultant
of measurements one  and two. The dip angles of
each  of the  original surfaces are  then projected
onto  the vertical plane of the hypothetical resultant
dip direction.  These apparent dips can be quickly
determined with the use of  a special  protractor
(Lahee, 1952). The apparent dips are then averaged
according  to  their  rankings,  and a hypothetical
resultant dip  angle  is  obtained. This procedure  is
shown in part B of Figure 2.
                                                       Measurement  Dip direction  Dip Ranking
                                  Fig. 2. Example for calculation of the "resultant" strike and
                                  dip of fractures in rocks (after Freethey, 1969).
                                 Geologic Variables
                                      Three  geologic  variables used ;i this  study
                                 involve angular relationships  between the "result-
                                 ant" strike  and dip of fractures at each site and the
                                 orientation  of  a  reference  plane.  The  reference
                                 plane  at each homesite is the vertical plane which
                                                                              137

-------
passes through the leach  field and the well. Figure
3 illustrates the geologic variables. A description ot
the variables follows:
Variable * 1. Angle  between the azimuths of the
   "resultant" dip and the leach field-well plane.
Variable *2. Acute  angle between the azimuths of
   the "resultant" strike  and the leach field-well
   plane.
Variable * 3. Vertical angle  between the line con-
   necting the leach field with the water level in the
   well and  the projection of the "resultant" dip
   onto the leach field-well plane.
Variable  *4. Degree of rock weathering. (.Because
   weathered rocks may be effective  in filtering and
   cleansing  leach field effluent, some assessment of
   this variable was  considered essential in evalua-
   tion of pollution potential at mountain  home-
   sites. A ranking procedure was  devised. A rank of
   one represents  relatively  unweathered  rocks,  a
   rank of  five represents a highly weathered rock
   mass.)
                       projected "Resjltant"
  water table gradient and the  leach  field-well
  plane. (In several cases the azimuth of the water
  table gradient was estimated  from land surface
  topography.)
Variable *6. Depth to static  water level  in  the
  well.
Variable #7. Vertical angle between the well head
  and the leach field.
Variable *8. Angle between the  azimuths of the
  topographic gradient in the  immediate vicinity
  of the leach field and the leach field-well plane.
Variable #9. Magnitude of the topographic gradi-
  ent  at the leach field  as projected  onto  the
  leach field-well plane.
Variable *  10.  Horizontal  distance between  the
  well and  the center of the leach  field.
 Fig. 3.  A. Diagram  illustrating  variables  *1  and  *2;
 B. Diagram illustrating variable *3.
 Fig. 4. A. Diagram illustrating  variables  *6,  #7, #9, and
 #10; B.  Diagram illustrating variables #5 and #8.
 Hydrologic and Topographic Variables
      "Ihe remaining variables  used in this study
 were devised  to  represent those  hydrologic  and
 topographic  conditions  at  the various homesites
 which are considered  to be related to  pollution
 potential.  Figure 4 illustrates  these  variables.  A
 description of the variables follows:
 Variable  »5. Angle between  the azimuths of the
 Statistical Analysis
      A discriminant function analysis was done to
 determine if any of the ten variables used were re-
 lated to or sensitive to  the incidence of pollution.
 Briefly,  the  discriminant function  analysis  is a
 standard  statistical technique  which computes a
 linear function of n  variables  measured  on each
 individual  of two groups.  This  linear function
  138

-------
provides a criterion for discrimination between the
groups. The function is computed so that the dif-
ference between  the  mean indices for the two
groups divided by their pooled standard deviation
should be as  large as  possible. Fisher (1946) dis-
cusses the computational steps and data require-
ments for the discriminant function analysis.

Results of Discriminant Function Analysis
     The  two  most significant  variables were varia-
ble # 1  (angle between the azimuths of the "re-
sultant"  dip  and' the- leach field-well plane)  and
variable  #8 (angle between the  azimuths of the
topographic gradient  at the leach field and the
leach field-well plane).  The significance of variable
# 1 apparently stems from the tendency of leach
field effluent  to follow the dip of fractures in the
zone of aeration. The significance of variable #8 is
possibly   due  to  the  tendency for near  surface
fluids to percolate laterally down slope in response
to gravity and zonation of weathering. If lateral
movement of effluent  from the leach field has a
component toward the well, pollution potential is
increased.
     Using only variables # 1 and * 8, a discriminant
function significant at the 0.1% level was obtained.
The  values of the discriminant function for all 28
sites are given  in Table 1. Values of the discriminant
function are computed using the following formula:

         D(f)  = -.00196(V,)- .0015(V8)
where D(f) is  the  discriminant function, V\  is the
value for  variable  one, and  V8  is the value for vari-
able  eight. The discriminant function D(f) can be
calculated for a site if estimates  can be obtained
for variables # 1 and #  8.

       FLOW  FIELD DIAGRAM  METHOD
     The  second study  (Millon, 1970) included 85
mountain homesites  in  the  Red Feather  Lakes
Area, Larimer County, Colorado (see  Figure  1).
Sixty two percent  of the  wells tested at these
homesites were found to  be  contaminated  with
E. coli  bacteria.  Data  collected for 65 well sites
established the position of each well relative to the
nearest leach field. Data were also collected on the
orientation  of bedrock jointing  surfaces and  on
water  levels  in wells  so that  ground-water  flow
directions could  be  evaluated.  Figure 5 shows
contours  drawn on water level elevations in wells.
The  study area was divided into 9 subareas in order
to obtain a  single  generalized direction for the
hydraulic gradient in  each of the  subareas. The
generalized directions of the hydraulic gradient are
indicated  by the wide arrows in Figure  5. Orienta-
tions of major high angle joints in the Red Feather
        Table 1.  Discriminant Function Values
               (after Freethey, 1969)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Values for
polluted
group
-.09678
-.12491
-.14253
-.18165


-.24230


-.29549


-.35614















Values for
unpolluted
group




-.22372
-.24201

-.28114
-.28536

-.30788
-.32189

-.36201
-.37507
-.37766
-.38321
-.39665
-.40115
-.40573
-.41615
-.41615
-.41714
-.42856
-.44093
-.46410
-.48433
-.51041
Polluted
site no.
5
2
3
1


4


6


7















Unpolluted
site no.




20
18

19
21

12
13

17
10
6
3
14
15
11
8
9
2
16
7
1
4
5
Lakes  Area  are summarized  in  Figure  6.  The
ground-water flow pattern at any point in the area
was  assumed to be  determined by the combined
effects of the hydraulic gradient and  the orienta-
tions of both major  and minor jointing surfaces in
the rock.
     The data  on well and  leach field locations,
joint orientations, and ground-water flow potential
were all combined by means of a specialized "flow
field" diagram  (see  Figure 7).  This diagram is in-
tended to graphically  illustrate how bedrock frac-
ture orientations may affect transport of leach field
effluent by  ground  water. A region of high pollu-
tion potential can be  identified on the flow field
diagram.

Results
     The  study  showed that wel's   in the  Red
Feather Lakes  Area which, because of their posi-
tion relative to nearby leach fields, plot on the flow-
field diagram within the  region of high pollution
potential have  about four times the  incidence of
pollution as do wells which plot outside the region.
                                            139

-------
                                                            Red Feather
                                                                 Lakes
                                                              Colorado
           Fig, 5. Map showing subareas and hydraulic gradients for the Red Feather Lakes Area (after

           Millon, 1970).
              joint orienta-
              tions
              sf»ear zones
                                           L.Nafcwnis
                                                                        1500   	3000
                                                                      tcaie (ft.)
                            i. Hiawatha
                                             ^ I
                                         .Ranncna
                                      Red Feather L.
/
     o
                                                                      Shajwa L.
                                              s
      Fig. 6. Map of Red Feather Lakes Area showing generalized orientations for major joints and shear zones

      (after Millon, 1970).
140

-------
  location of non-
  polluted well
                             location of
                             polluted veil
  strikes
  prominent
  joi
  in sub area
nt systems
 Ground water  flow direction imposed by the
 hydraulic gradient in the water producing
 zone of the sub area (marked with  10° range
 in each direction).
 Zone of ground water flow direction imposed
 by the confinement of direction within high
 angle joints  of the sub area.
Fig. 7. Hypothetical  flow  field diagram showing locations
of  wells  in relation to  neighboring leach  fields (after
Millon, 1970).
     A   normalized  significance  test  employing
Bernoulli data  (i.e.,  the percentage of  polluted
wells in  the flow field and  the percentage of non-
poiluted wells in the flow field) was of the follow-
ing form:
      t =
                   7Tp - 7Tl
                                      = 2.31
                JTT
                            1
                           rr
                           A i
where
               number of
       polluted wells in flow field
                                 25
                              •= — = .555
?TT =
     total number of polluted wells   45

     total number of wells in flow field   30
                                    — = — = .462
     total number of nonpolluted wells in flow field
           total number of nonpolluted wells
                                                               = .25
           total number of wells
                                    65
                                                            20
N, = number of nonpolluted wells = 20

N2 = number of polluted wells = 45.

The  "t"  value obtained  by the  above formula
showed statistical significance at the one percent
level.

                CONCLUSIONS
     The  analyses which have been described can
be reliable only if geologic conditions are adequate-
ly represented.  A hidden fault or igneous dike can
invalidate a site evaluation  based  on the discrimi-
nant function  or flow field  criteria. Hence, these
methods  for  evaluating  pollution potential  at
mountain homesites*  should be considered reliable
only in the statistical sense. Much  additional data
must be acquired before an acceptable criterion for
evaluation of pollution potential can be developed.
     Although results of these two studies should
be considered preliminary, they do tend to confirm
that  the  orientation  of jointing  surfaces  in the
bedrock does  affect the  travel path of contami-
nants.  Thus, evaluation of pollution potential at
mountain homesites similar to those investigated in
these studies must include detailed geologic analy-
sis of bedrock underlying the thin soils.

                 REFERENCES
Fisher, R. A. 1946. Statistical methods for research workers.
     Oliver and Boyd, London, 10th ed., 356  pp.
Freethey, G.  \V.  1969. Hydrogeologic evaluation of  pollu-
     tion potential in mountain dwelling sites. M. S. thesis,
     Colorado State University, 90 pp.
Lahee, F. H. 1952. Field geology. McGraw-Hill Book Co.,
     New York, Appendix 14, 583 pp.
Millon,  E.  R. 1970. Water pollution, Red Feather  Lakes
     Area, Colorado.  M. S.  thesis,  Colorado  State Uni-
     versity, 68 pp.
                                              DISCUSSION
The following questions were answered by James
P. Waltz after delivering his talk entitled "Methods
of  Geologic  Evaluation of Pollution Potential at
Mountain Homesites."

Q.  \Ybat  is  the  range  of values  of the diversion
radius of the well compared to the distance be-
tween well and waste disposal sites?
A.  The problem  of proper spacing between leach
fields  and  wells  is a  very real one. I  showed  a
diagram which illustrated that fluid can travel from
                                                  the  leach field toward  the  well  along fractures in
                                                  the  zone of aeration. I  showed another portion of
                                                  the  diagram which demonstrated that  local topo-
                                                  graphic  reverses might  cause  percolating fluids to
                                                  enter  the ground  water upstream  of the well  be-
                                                  cause the regional ground-water gradient is different
                                                  from the local topographic  slope. There is also  the
                                                  very common problem, which the questioner  has
                                                  raised,   where  the leach field  is  at the  ground
                                                  surface  and is  recharging the ground water while
                                                  the  well is  perhaps 200 feet  away pumping water
                                                                                                     141

-------
and  creating  a large  cone  of depression  in  the
ground  water around  the well. This condition re-
verses the ground-water gradient so that even if the
well was  originally  upslope both  topographically
and with respect to  the  ground-water gradient the
flow direction can change very quickly.
     I don't have any figures for the shape and size
of the drawdown cone. It will probably be a while
before we know very  much about  that. The prob-
lem  can  be  approached on  a theoretical  basis,
assuming  fractures  are like parallel plates. We've
done some experimenting with this type of model
to determine the interaction  between a series of
wells and a series of  leach fields as, for example,
would occur in a subdivision. The results from this
type  of analysis are useful but fairly generalized.
The fact of large diversion  radii of mountain wells
certainly  contributes  to the  pollution  potential
because the storage capability of fractured crystal-
line rocks is minuscule compared to most aquifers.
We're talking about less than 1% porosity, even in
fairly fractured  rock.  Tremendous drawdowns are
associated with  the withdrawal of water and this
certainly  is going to affect the direction of ground-
water flow in the vicinity of the well.

Q.  Arc there E. coli  in  the lakes and streams of
the Red Feather Lakes Area?
A.  Yes.

Q.  What u-ere their percolation rates?
A.  The percolation  rates vary tremendously. Even
within  an individual  half  acre or  acre lot it  is
possible to get a wide range of percolation rates.

Q.  Were  the  contaminated  wells evaluated for
sanitary construction?
A.  No, the homesites  that we checked were home-
sites which had  been  in  operation for quite some
time, and we assumed that the wells were in a state
of  equilibrium  with their  environment, including
the nearby leach field. I think that in some cases
there might have been poor construction practices
associated with the contamination.

Q.  Was any attempt made to clean  up the contami-
nated wells with  Marine?
A.  Certainly. Many of  the people in these  areas
that we studied have gone to chlorination systems.
This is  one certain way they can fight the problem.
We had to make special arrangements  to obtain
our samples from their nonchlorinated supply. We
sometimes tapped into their water system upstream
from the chlorination.
Q.  Were tests run for other enteric bacteria?
A.  There  was a facet of our study done  by our
microbiologist which  included  additional testing
but no extensive program of testing has been done
other than on the E. coli.

Q. Do you believe values of variables  V-l and  V-8
can be reliably obtained in the vicinity of most new
homesites?
A. Data for these variables must be collected by a
geologist. The data collection presents an interest-
ing problem  because many  of the homesites are
built  on  somewhat  flat ground where  thin  soils
commonly obscure the bedrock. It is necessary to
go to the periphery  of the  homesite to find  the
rocks that are in place where orientations of frac-
ture systems can be recorded. We usually would
measure at least several dozen fracture orientations
to  establish  the  uniformity  of  these  fractures
throughout the  homesite area. Where we could do
this,  I feel the  measures are reliable,  but in some
cases we had exposures on maybe only 2 or 3 sides
of the property. In these cases there is room for all
sorts of geologic features which can invalidate the
criterion. As  I have said, we're hoping here only for
the statistical validity. There  is no absolute security
in the discriminant function criterion because there
are often going to be  complicating geologic factors.

Q.  How do you evaluate  the reliability of j rela-
tively variable discriminant function?
A.  We hope to  evaluate this discriminant function
after we have extended the  data base to include a
large number of sites. Then  we wish to apply it to
a number of unknown sites  and test the success of
our  classification. I think this can be done. Relia-
bility can be  tested in  a  statistical  sense.  The
discriminant function  is  variable  because of the
nature of the function. If it were not highly variable
then  its value  as a  classification  criterion would
be less.

Q.  Why not simply require a chemical home treat-
ment system where there is less than 6 feet of soil?
A.  Economics appears to be the main reason. How-
ever, there must be alternatives to the septic tank
leach field system in these areas where there is an
inadequate soil. Here in Colorado, the Water Pollu-
tion  Control Commission  has issued new rules on
septic tank practices which include the "designa-
tion" of pollution prone areas in the  State.  For
these  "designated"  areas, special  engineering and
geologic studies will  be required  before approvals
for  private  sewage  disposal systems can be ob-
tained.  It looks to  me like  this is  a  great  step
forward. What it will likely  mean is that a detailed
investigation will show that a septic  tank system
cannot  be installed  in most areas and  alternative
systems will have to  be considered.  I'm not a
sanitary  engineer  and  I  don't know  about the
 142

-------
success of many of these chemical systems. I have
heard that  many of  the present alternatives  to
septic tank systems are not entirely reliable because
they require  a  degree  of maintenance  which is
often lacking. The leach field system has worked
beautifully where conditions are  appropriate be-
cause it doesn't really need much maintenance.

Q. Do you have any minimal lot size regulation for
lots  hi Precambrian crystalline rocks where both
wells and septic tanks are used?
A. In  order  to  assess  the minimum  size of lots
from mountain homesites  we need to know a little
more about  what happens to  the  bacteria that
enter the  fracture bedrock.  We  know that they
can  easily  travel  hundreds ot  feet,  and we may
even be talking about travel of U of a mile or more
without being appreciably filtered out or absorbed.
We are presently  involved  in studies where we've
taken  different rock  types,  crushed  them,  sized
them,  and added  water containing known quanti-
ties  of E.  coli. The fluids were periodically tested
for two weeks to determine a die-off curve for E.
coli. We found remarkable variations in the chemi-
cal  effect  of the rocks  on the  viability of the
organisms. I  think this sort of data will ultimately
be  useable in determining minimum  lot sizes  for
various types of  geological terrain. We  have also
attempted to learn a little more about the rate of
travel of fluids through crystalline rocks, because in
addition to  the  time that is required  to  kill  off
these organisms in the rock you must be concerned
about  how fast the fluid  is going to move in that
period of  time. These data are necessary- to deter-
mine what the safe radius from a  leach  field is
going  to be.  So we have taken areas  such  as the
road cut shown in this picture, where well defined
joint sets can be seen,  and injected fluids into the
rock fractures. We put 55 gallon drums filled with
bacteria-inoculated water at the top of the cliff and
siphoned  the fluid into PVC pipe that had been
cemented  into the rock.  In this case the bacteria-
laden  water surfaced  down here at approximately
 15 to 20  feet  away  from the injection point.  I
just wanted  to  show  that we get effluent coming
out of this  rock, this highly fractured  rock con-
centrated  at this location. We were able to collect
the fluid and  sample the  bacteria. We found that
essentially nothing was filtered  out. The same con-
centrations were at the bottom  as at the top. Little
seeps occurred at 3 or 4 other places, but this was
the only place where any real  flow occurred and
the rate was quite high. We had a garden hose, and
as fast as it would siphon, the fluid would go into
the rock.

Q.  Who  is working  on development of  a  better
septic tank system?
A.  1 have nothing more to say about alternatives.  I
know there are  lots of people who recognize that
the septic tank needs to be  improved or  replaced
for installations  in mountainous areas. 1 am confi-
dent that there are better solutions presently avail-
able and that  they will be used only if the pressure
is put on people to  not use septic tank-leach field
systems any more where terrain is not suitable. We
just have to cut them off.
Q.  Here's a question about your testing  methods
for E. coli. Do  you mean coli form by  standard
method?  What technique did you  use for E. coli
analysis? Did  you rim any tests for f-cai  c.n'nform
since  these are  a  better indicator  of domestic
sewage contamination?
A. I  wish I had the microbiologist associated with
this project here to answer that question. We,  for
the studies I've recorded here, did test only for the
E.  coli. We considered a site to be contaminated  if
large numbers of E.  coli were present in the water
samples taken. We did find variations in the content
of E. coli in the samples we obtained.  We took
several samples at each site and replicate determina-
tions were made. We went back  on different days,
different weeks, and in some  cases as many as  a
dozen or more  samples were repeated so that we
were relatively sure  of the status of contamination.
And, in some cases, definitely contaminated wells
would  be retested  and  turn out to be okay.  We
tried to tie this  in with precipitation patterns and a
 number of other  things. There are obviously  a
 number of factors  which  influence  the changes
 with time in  the concentration of E. coli in ground
 water.
                                                                                                  143

-------
Nitrate  in  Ground  Water  of the  Fresno-Clovis
                                                        a
Metropolitan  Area,  California
by Kenneth D. Schmidt
       Member N\VWA
                   ABSTRACT
     N.itu'.il  concentrations of nitrate are  quite  low  in
most ground waters in the eastern part of the San Joaquin
Valley. High nitrate contents are related to sewage effluent
percolatu n  ponds, septic rsnk  disposal systems, industrial
waste-waters, .mil agricultural fertilizers. Hydrologie factors
are closely related to  the occurrence of nitrate. Transmissi-
bilitv of the aquifer, hardpan development in  the soil, canal
recharge, and cobble zones in the subsurface are the primary
factors of importance. Nitrate is stratified in the  aquifer
beneath unsewered metropolitan areas and highest contents
occur in the upper 50 or 60 feet. Water quality hydmgraphs
were used to  show long- and short-term trends in  nitrate.
Chloride  and  nitrate  hydrographs, triiinear  diagrams, the
distribution of other constituents, and hydrologic data were
used  to  efiVcnvch  delineate  sources of  nitrate in  areas
where  numerous potential  sources were present.  Conclu-
sions fr 'tn •_'!''>und-\\.ner c!a;,j  .igic'  -.veil with previous
studies tn other areas beneath unsewered tracts and near
sewage treatment plants in which attention was  focused
pnrnanfv  on the soil or the unsaiurated /one.

                 INTRODUCTION
      Fresno County was formed in 1856 out of the
Mariposa  territory  and currently  produces  the
greatest  amount of agricultural wealth of  any coun-
ty in the nation.  The Fresno  Irrigation District
(.Figure  1) supplies surface water to much  of the
agricultural  area  through  canals  from the Kings
 River, which drains part of the Sierra  Nevada east
of Fresno.  The Fresno-Clovis  Metropolitan Area
 (F.CM.AJ comprises about  145 square  miles be-
 tween the Kings  and San Joaquin  Rivers  in  the
 east-centra!  portion of  the  San  Joaquin Valley.
 Rapid industrial growth has occurred in the  past
 decade and  population of the urban area reached
 310,000 in 1969.
      Water  lor  municipal  and  industrial  use is
 pumped  trom  wells, whereas in  the  surrounding
 agricultural  area surface water and  ground water
 are used.  Ground  water occurs  in the alluvial-fan
 deposits  derived  from  the Sierra  Nevada.   The
 principal  subsurface  geologic units were divided
 into consolidated  and unconsolidated  materials by
      Presented  at the National Ground Water Quality
 Symposium, August 25-27, 1971.
      I'HydroIogist, Harshbarger  and Associates. 1525 E.
 KSemdale Road. Tucson, Arizona 85719.
     Si
     ,->
     v
      f
    '*§->.
Fit). 1. Fresno County and related features.
Page  and  LeBlanc  (1969).  Consolidated rocks
comprise the basement complex ot pre-Tertiary age
and the marine and continental rocks of Cretaceous
and Tertiary age.  The basement complex is about
3,000  feet deep  beneath Fresno. Unconsolidated
deposits are  the  major source of  ground water
pumped by wells  in the Fresno area. Deposits ot
Quaternary age crop out over most ot the area and
comprise the principal  aquifer. These deposits  in-
clude numerous intercalated  strata ot sand, gravel,
silt, sandy  clay, clay,  and  localized  cobble /ones.
Most of the ground water moves laterally in coarse-
grained beds, and the vertical movement ot \\atei is
restricted in places  by thick  layers of  clay anil
silty clay.
     The climate of the F.C.M.A. is characterized
by  a  long,  warm,  and dry summer  trom  May
through October,  and a cool,  rainy winter season
from  November  through April. Annual precipita-
tion at Fresno averages about eleven inches,  Ihe
amount of precipitation in the foothills and moun-
tains  of the Sierra  Nevada  directly affects the
streamflcnv, recharge to ground water, and surface
water  supply. Much of the precipitation  at higher
elevations  falls as  snow,  and runott  in major
streams usually reaches a peak  in  June or  July,
following periods of high snowmelt.
 144

-------
     Depth of ground water averages about 70 feet
 beneath the F.C.M.A., and the regional direction of
 flow  is  southwest.  A ground-water  cone of de-
 pression  exists beneath part of  the urban area due
 to large  pumping extractions. Sources of recharge
 include  seepage  from  streams  and  canals  and
 ground-water  inflow  from  the east.  Well  depths
 range from less than  100 to more  than 500 feet.
 Most wells were drilled by  the cable-tool method
 until  recently,  and  unperforated   casings  were
 installed with open  bottoms. Recently, wells have
 been  drilled  by  the  reverse-rotary  method  and
 gravel  packed.   The average specific  capacity  is
 about 1 30  gallons per minute per foot  (gpm/ft) and
 the average transmissibility exceeds 220,000 gallons
 per day per foot (gpd/ft).
     A  detailed  correlation of  subsurface  rock
 types was  made as part of the development of the
 lithologic framework of the area. Drillers' logs were
 correlated  with several electric  logs available. Ex-
 tensive deposits ol cobbles  and  coarse sands occur
 to depths  greater than  500 feet in the northern
 portion ot the I-'.C.M.A. An impermeable hardpan
 ot the San Joaquin  Series  also  occurs over much
 ot the northern part ot the area. Other soils vary in
 hardpan  development  and  the Madera  sand has
 virtually  no  hardpan  development.  Correlations
 generally suggested little continuity of clay strata,
 but  some  cobble zones  were  traced tor  several
 miles.  Clays predominate  below  350 to 400  feet
 depth in the southern  portion of the F.C.M.A. The
 depth of Quaternary alluvium ranges  from  350 to
 400  feet, and the underlying continental deposits
 are finer grained except near the major rivers.
     High  nitrate in drinking water  is important
 because of an infant disease, mcthemoglobmemia.
 Concentrations exceeding  the  1962  U.  S. Public
 Health Service limit  of 45  parts per million (ppm)
 were noted in the F.C.M.A. by the California De-
 partment of Water Resources (1965).  The primary
 sources  of  ground-water nitrate were  identified as
 sewage effluent percolation ponds  at the  Fresno
 Sewage  Treatment  Plant and septic tank disposal
 systems in  unsewered areas.  Nitrate concentrations
 ranged from 0.8 to  58.5 ppm.  and averaged 15.1
 ppm,  for the  F.C.M.A. Most high nitrate contents
 were from  samples  collected from  shallow wells.
 Analyses of surface  waters showed low concentra-
 tions of all constituents,  including nitrate.  Analy-
 ses  of wastewaters  at sewage  treatment  plants,
wineries  on independent disposal systems, and  a
 meat-packing plant were presented. Although high
concentrations of most constituents were apparent,
total nitrogen was generally unreported.
     Behnke and Haskell (1968) studied nitrate in
ground water  near the Fresno Air Terminal. They
noted the importance of septic tanks and the Clovis
Sewage  Treatment  Plant  as  nitrate  sources.
Nightingale (1970)  reported on the salinity and
nitrate  trends for ground water of the  urban area
and the surrounding agricultural  lands. Although
urban zone well water nitrate  was almost  twice
that ot the agricultural zone for the period  1950-
56, for the period 1962-67 the two zones had an
almost equivalent nitrate content. The mean nitrate
content  of the agricultural zone  well water in-
creased by  300 percent from year lot  1950-56 to
year lot 1962-67, but no reason for this increase
was given.
     Despite these  studies and  others, many  ques-
tions have  remained unanswered with  regard to
                                        c
ground-water  nitrate in the  F.C.M.A. Questions as
to source, effect of lithology and well construction
on  nitrate,  and  relation to  the  hvdrologic svstcm
                                      C?   -
often have arisen. It is my intent to establish a firm
hydrologic  framework  for  the present study.  A
model can be built upon this framework  to describe
the sources and distribution of nitrate in ground
water of the F.C.M.A.
     Five sites were selected for detailed analysis
(Figure 2). The  Figarden-Bullard Area was selected
because  of  the  prevalence  of one nitrate source,
namely septic tank effluent. A second site selected
was the  vicinity of  the Fresno Sewage  Treatment
Plant,  which  comprises the majority  of sewage
treatment  on  a volume  basis in  the F.C.M.A.
Sewage effluent disposed  ot by  percolation ponds
is the primary nitrate source to ground water in
this area. The Tarpey  Village and Mavfair-Fresno
Air Terminal  Areas  are both unsewered and have
      R. 19 E.
      : SEWAGE TREATMENT
Fig. 2. Study areas in the Fresno-Clovis Metropolitan Area.

                                            145

-------
some of the highest nitrate contents reported in the
F.C.M.A.  Sources  other than  septic  tanks are
present,  including   sewage  effluent  percolation
ponds,  winery wastewater disposal  ponds,  and
agricultural fertilizers. The  fifth study site  com-
prised the Downtown Fresno area, which is sewered
and  has no obvious nitrate source. Although high
nitrate contents have been monitored by several
wells in this area, virtually no study had previously
been undertaken.

         FIGARDEN-BULLARD AREA
     Schroepfer and Polta (1969) reported  on
nitrate in ground water  of  septic tank areas near
Minneapolis and related  ground-water contamina-
tion  to suburban growth. These authors found that
septic  tank effluent had   an  average ammonia-
nitrogen  content  of 28  ppm orCTzTj^m^nitrate
equivalent.  The Figarden-Bullard Area  is in the
northern part of the F.C.M.A. Although the area is
unsewered, only three wells have ever had nitrate
over 25 ppm. Ground-water depth ranges from 90
to 100 feet and water levels have declined about 2
or 3 feet per year for the  past decade. Lithologic
sections show an abundance of hardpan, sand, and
cobble zones in this area. Much of the hardpan is
impermeable and  is in the San Joaquin sandy loam
soil type. About 80 or 90 percent of the area was
agricultural in  1950,  whereas less than 20 percent
was  agricultural by  1970.  Population density  is
lower than  for most other urban  parts  of the
F.C.M.A.
     Chemical hydrographs  were prepared to illus-
trate long-  and short-term trends in water quality.
The  predominant historical trend in the northern
portion  of the Figarden-Bullard  Area  was con-
stancy of  nitrate.  This portion  has  the largest
transmissibilities, the  most  abundant cobble zones,
and the greatest development of impermeable hard-
pan  in the  Figarden-Bullard Area. About one-half
of the hydrographs in the  southern portion had a
constant nitrate trend, whereas the remainder had
increasing trends. The southern  portion  had the
earliest  septic  tank development and maximum
septic tank density of the area. Cobble zones are
locally absent,  transmissibilities  of the aquifer are
lower than in  the  northern portion, and the im-
permeable hardpan is locally missing.
     Nitrate and  chloride hydrographs (Figure 3)
were drawn for all  wells where more than four
separate analyses were available. Short-term trends
indicated similar nitrate  and chloride patterns for
16 out of  19 hydrographs.  This similarity prevails
throughout septic tank parts of the F.C.M.A. and
is apparently due to  the derivation of nitrate and
chloride from septic tank effluent and the mobility
of both ions in ground water. Three hydrographs
with  opposite  nitrate  and  chloride  short-term
trends were for wells apparently affected by sewage
effluent from  plants north of  the Figarden-Bullard
Area.
                  1958   I960
                                        1968   <970
Fig. 3. Nitrate and chloride hydrographs for a selected well.


     Water  samples  from  wells have  been taken
semiannually for nitrate analysis in several Fresno
County  Waterworks  Districts  in  the  Figarden-
Bullard Area. Nine wells had greater nitrate con-
tents in  the summer than in the winter, and had
cased  depths  less than  150 feet.  Higher nitrate
contents in the  summer  are  due  to  downward
movement  of shallow waters,  which  are high in
nitrate, to producing zones during heavy pumping.
Higher temperatures in the summer also favor nitri-
fication  of ammonia-nitrogen retained  in the soil
from septic tank  effluent. Five wells had higher
nitrate contents in the winter than in the summer,
and  had cased  depths  greater than 150 feet. This
group of wells was on  canal banks or nearby. Low
nitrate contents in the summer are attributable to
canal dilution by low nitrate Kings River water, as
several major unlined canals traverse the area. Low
nitrate  contents  in  the summer occur for  wells
where the impermeable hardpan is absent through-
out the F.C.M.A.
     The concept that nitrate is stratified in the
aquifer was tested in  the Figarden-Bullard Area.
Only open-bottom wells were used in this analysis,
as they tend to draw water from specific portions
of the  aquifer.  The difference  between depth to
water and cased depth is a reflection of the amount
of aquifer penetrated and is generally related to the
zone of production. Depth to water was taken from
the  Fresno  Irrigation District Map for  December,
1970.  The difference between depth to water and
cased depth is referred to herein as "aquifer pene-
tration," and values for specific wells were plotted
graphically with  1970 nitrate analyses  (Figure 4).
146

-------
  200
           5      10     15     20    25    30    35
           NITRATE IN  PARTS  PER  MILLION
Fig, 4. Relation  between  aquifer penetration and  1970
nitrate for wells in Figarden-Bullard Area.


The  number  of  wells with lo\v  nitrate contents in
each  major depth zone  shows that there  is  no
continuous "nitrate cap"  in the  Figarden-Bullard
Area, a  fact  also shown by contours drawn tor
nitrate  content   on an areal  basis.  The  highest
nitrate contents  are in  the  upper 60 feet of the
aquifer,  and  about 5 to 7 ppm approximates the
"background" nitrate  in the ground water. The
decrease  in nitrate content with  depth in the upper
80 feet of the aquifer  averages about  I ppm per 4
feet  of depth. Wells with  the highest nitrate con-
tents for their  specific aquifer penetration  were
usually within the zone of lowest transmissibility
and  in the area  where  septic tanks had  been used
for the longest time.
     The low nitrate contents in the aquifer of the
Figarden-Bullard  Area are  attributed  in large part
to hvdrogeologic  factors.  Extensive  cobble  zones
permit more  dilution and dispersion of high nitrate
recharge  from septic tank  disposal systems. Few-
wells are extremely shallow  or  perforated in shal-
low  intervals, and the water table is relatively deep.
The  presence of an impermeable hardpan prevents
percolation of nitrogen forms in some cases, unless
it is  penetrated by leach lines or seepage pits from
septic tanks.  The low density of septic tanks com-
pared to  other unsewered areas of the F.C.M.A. has
                                                    also minimized  nitrate contents.  Although  the
                                                    Figarden-Bullard  Area  was  converted from  pri-
                                                    marily an agricultural area in 1950 to an urban area
                                                    at present, there is little evidence of greatly increas-
                                                    ing nitrate contents. Little downgradient movement
                                                      O                           t.
                                                    of nitrate  from septic tank areas to sewered areas
                                                    has occurred.  The Figarden-Bullard  Area is perhaps
                                                    the ideal location in the F.C.M.A. for operation of
                                                    septic  tank disposal systems. With proper construc-
                                                    tion of wells and septic  tank disposal systems, little
                                                    problem with  high nitrate contents in ground water
                                                    should occur in the future.
    FRESNO SEWAGE TREATMENT PLANT
     The Fresno Sewage Treatment Plant is about
3 miles southwest  of downtown  Fresno and has
been  in  operation  since  1891. Plant No.  1 has
provided primary treatment since  1947 and  Plant
No. 2. which was built in 1960, provides secondary
treatment.  Final disposal of most of the effluent is
by percolation on  1,440 acres of ponds. About 10
to  15 percent of  the effluent is used  for  crop
irrigation  on nearby  lands. Wastewater  flow cur-
rently averages almost 30 million  gallons per day.
The total nitrogen in domestic wastewater effluents
ranges from about 18 to 28 ppm when there is no
specific treatment for nitrogen removal. Records at
the Fresno  Sewage Treatment Plant suggest that the
average total nitrogen content is  about  25  to  30
     t_            O
ppm,  about 8  to  10  ppm of which  is in the
ammonia form. Nitrate contents averaged about 6
ppm  in 1969. Although industries comprise only
about 10  percent of  the  sewage  influent to the
plant, wineries  contribute  significantly  to waste
strengths during the period from  September 1  to
December  1.
     Jcnks and Adamson (1968) determined  pond
percolation rates and ground-water quality  at the
Fresno Sewage Treatment Plant. Fight auger  holes
were  drilled  to reach the ground water and water
samples  were collected. Five of  the eight water
samples  had nitrate  contents  less than  4  ppm,
whereas nitrate contents tor the remaining samples
were  76,  160, and  213 ppm.  The holes  with high
nitrate content were  in a newer area where ponds
were  periodically  dried out in order to achieve
higher infiltration rates. These holes were apparent-
ly in  an area recently dry and  just flooded. High
nitrate water is common at the beginning of flood-
ing cycles (Stout, Burau, and Allardicc, 1965), and
the  water   is readily leached  to  ground water.
Ponds develop anaerobic conditions when kept full
and  nitrogen is retained in the soil in the organic
and  ammonia forms.  Aerobic  conditions develop
when ponds are periodically dried out and nitrifica-
                                                                                                 147

-------
tion can occur.  Nitrate is  then  leached upon the
next wetting of the ponds, and this period is fol-
lowed by low nitrate recharge water once anaerobic
conditions develop.  Denitrification can  also sub-
stantially  reduce nitrate contents under anaerobic
conditions.
     Typical percolation pond water  was sampled
for a portion of 1968. Analyses for nitrate-nitrogen,
ammonia-nitrogen, and pH demonstrate the influ-
ence of winery  wastewater at the Fresno Sewage
Treatment Plant. One  pond  had  an  average  am-
monia-nitrogen content of  about 1.2 ppm prior to
September 1, and an average of 8.5 ppm during the
winery season of September  1 to December 1. The
average ammonia-nitrogen content  for  another
pond was 4.1 ppm prior to September 1, and 8.5
ppm during the  winery season. Nitrate-nitrogen for
the first pond during  the same period  increased
from an  average of  0.3  to 2.1  ppm.  Nitrate-
nitrogen for the second pond increased from 0.4 to
1.7 ppm for the same period. Winery wastewaters
in the  F.C.M.A. have a total nitrogen content rang-
ing from  4 to 10 times that of typical domestic
sewage  effluent.
     Chemical  hydrographs  were  prepared for 6
monitor wells near the Fresno Sewage Treatment
Plant.  All  hydrographs showed similar  seasonal
fluctuations (Figure  5) for nitrate and ammonia.
Nitrate contents sharply decrease  after maximum
values in  September.  Contents gradually     -ease
to  a minimum in March and  gradually inci ~,e to
July. Contents   sharply increase  in  August  and
September. High nitrate contents occur in ground
water at the beginning of the winery  season, when
the nitrogen  contents of pond waters  are very
high. However,  pond  waters  also have the highest
biochemical oxygen demand for the entire year at
this time.  These  highly  anaerobic   wastewaters
readily utilize  oxygen, and nitrification cannot
occur. Nitrogen is thus regained in the soil and not
leached to  the ground water; hence ground-water
nitrate content sharply decreases. Nitrification does
not occur as rapidly during the winter and nitrate
contents gradually decrease until the  spring. Nitri-
fication readily occurs with the warming  tempera-
tures of the spring  and summer, nitrate is readily
leached to ground  water, and contents gradually
increase to the  fall. Maximum nitrate contents in
the ground water each year  may  actually be due
to  nitrogen supplied to the soil during the previous
year. Pond operation during the summer includes
more drying out cycles, as the wastewater flow is
smaller, and aerobic  conditions tend to develop,
which also favor nitrification. Waste  flows are the
largest  in  the winter and ponds are generally kept
full, allowing anaerobic conditions to develop. The
Q.
Q.
z
LU
O
o
o
        WINERY
        SEASON
                                      WINERY
     	1	1	T	1	1	1	1	1	1	1	1	1	1	1   I  I   I
     ASONDJFMAMJJASOND
Fig. 5. Nitrate concentrations in ground water near Fresno
Sewage Treatment Plant.
seasonal variation in nitrate (Figure 5) clearly indi-
cates the inherent limitations of comparisons made
from year to year based on only one analysis per
year.
     Ammonia-nitrogen contents of ground water
show opposite  trends to nitrate-nitrogen near the
Fresno Sewage Treatment  Plant. The highest con-
tents occur in March and the lowest contents occur
in September.  The  soil retains  nitrogen  on ad-
sorptive  and  cation-exchange  sites when  little
nitrification occurs. Ammonia gradually tends to be
leached in small amounts  as the capacity of the
soil to retain nitrogen is approached. The ammonia
contents of ground water are high when nitrate
contents are low,  because  little nitrate is  formed
under  these  conditions. Much of the capacity of
the soil to retain nitrogen is unused when abundant
nitrification occurs, as much of the nitrogen is con-
verted to nitrate and leached. Little ammonia is
leached in this case, as it is either retained by the
soil or converted to nitrate.
     A number of chloride  and nitrate hydrographs
for wells near the  Fresno Sewage Treatment Plant
exhibit opposite nitrate and chloride  short-term
trends. That is, as  the nitrate  content increases for
a period,  the chloride content simultaneously de-
creases. This trend was  thus in contrast  to that ob-
served in septic tank  portions  of the F.C.M.A.
Chlorides are an effective tracer in ground water of
the F.C.M.A.,  because  natural concentrations are
quite low. Ground water affected by sewage efflu-
ent generally has chloride concentrations exceeding
 148

-------
45  ppm.  High chlorides in ground water near the
treatment plant suggest that little dilution of sew-
age  effluent occurs,  a fact also suggested by the
amount of  water  recharged on an annual basis. A
trilmear diagram for various ground-water samples
showed a  consistent chemical type, with sulfate less
than 10 percent and chloride ranging from 10 to 20
percent on a reacting value basis.

           TARPEY VILLAGE AREA
     Nitrate has substantially  increased in ground
water of the Tarpey Village Area (Figure 6) in the
past, and  there are a  number  of potential sources.
Thirteen  wells have had  nitrate  over  25 ppm at
least once,  and 6 of  these wells have  had  nitrate
over 45 ppm. Tarpey  Village  proper presently has
municipal  sewerage  connections  east  of  Clovis
Avenue,  whereas  septic  tanks are used west of
Clovis Avenue. The area is bounded on the east and
northeast  by extensive acreages of vines, peaches,
and  pastures.  Johnson et  al.  (1965) conducted a
subsurface  drainage  study of the  San Joaquin
Valley.  Variable portions of  applied nitrogen in
agricultural areas were lost with the drainage wa-
ters. The nitrogen content of drainage water ranged
from 1.8  to 62.4 ppm, with a weighted average of
25  ppm,  or 110  ppm nitrate equivalent. Sewage
treatment plants and wineries use percolation ponds
tor waste disposal  and are other potential nitrate
sources.
     The  Clovis Sewage Treatment  Plant has pro-
E  SHAW iVEWUF
Fig. 6. Tarpey Village Area and wells of interest (circles)—
waste dischargers are cross-hatched.
vided primary treatment for essentially domestic
sewerage  from  the municipality  of Clovis since
1955. Disposal of sewage effluent is by percolation
on  about  30 acres west of Tarpey  Village.  Total
nitrogen content of the effluent was 28.0 ppm, or
123 ppm  nitrate  equivalent,  in  1968. About 70
ppm nitrate  were sampled in  the effluent in 1963
and were  apparently related  to  operation  of  the
trickling filters. Nitrate in the past several years has
averaged 10 to 20 ppm.  The average effluent flow
is about 1.0 million gallons per day and in  recent
years 2  wells near the plant have  monitored  the
highest  nitrate contents  reported in the F.C.M.A.
for pumping wells. Water from these wells has ex-
ceeded  120 ppm nitrate. Domestic wastes from the
eastern  part of Tarpey Village are treated by Fresno
County  Sanitation  District  No.  1,  which was
formed  in  1955. The average effluent flow is about
0.2  million  gallons per day  from an activated
sludge unit.  The  total  nitrogen  content  in 1963
was 29.0 ppm, or 128 ppm nitrate equivalent. The
pond effluent is percolated from 5 acres of ponds
southeast of the Italian Swiss Colony Winery.
     The Italian  Swiss  Colony Winery, or  Allied
Winery, had a wastewater flow of 40 million gallons
in 1968. Disposal of wastewater is by land flooding
of  about 110 acres south  and east of the winery,
and disposal of clear water into  the Gould  Canal.
Wastewater analyses by the California Department
of  Water  Resources  (1965)  indicated potassium
content  was 920 ppm,  sulfate content was 177
ppm, sodium content was 102 ppm, and chloride
content was  41  ppm. However, no determinations
were reported for total  nitrogen during the winery
season.  The  author took 2 samples of wastewater
from ponds of the Italian Swiss Colony Winery in
November,  1970. Total  nitrogen content for one
sample  was  100 ppm, whereas that for a second
sample  was  250 ppm.  These values have nitrate
equivalents of 440 and 1,100 ppm, respectively.
Winery  wastewater seems to have a  total  nitrogen
content  similar  to that for meat-packing plants.
Winery wastewaters are a potential source of nitrate
in ground  water  that have been  neglected in past
studies in the F.C.M.A.
     Static water levels declined in the Tarpey
Village  Area from the  1950's to 1962 or  1963.
This trend  was reversed and rising water levels have
occurred from 1963 to the present. Depth to water
in 1969  ranged from less than 30 feet in the eastern
portion  to more  than 70 feet in the western por-
tion. The predominant  direction of ground-water
flow beneath the Tarpey Village Area is  southwest,
and the hydraulic gradient averages  25  to 30 feet
per  mile. This area is in a transition zone between
the  cobble and sand zones of the northern F.C.M.A.
                                                                                                149

-------
and the abundant clays to the south. Cobbles are
generally absent and a soil with no hardpan occurs
throughout  much of  the  area. Transmissibility  is
rather uniform and averages 180,000 gpd/ft.
     Nitrate hydrographs show long-term increases
for most wells. However, wells perforated in deeper
parts  of the  aquifer had  a constancy  of nitrate.
Historical records indicate a substantial  increase  in
nitrate content occurred between 1958 and 1963.
The  septic tank  part of Tarpey Village was devel-
oped, the Clovis Sewage Treatment Plant was re-
activated, and the Fresno County  Sanitation  Dis-
trict No.  1 treatment plant was built during  or
shortly prior to this period. Five wells with chloride
and nitrate hydrographs had a similarity of nitrate
and chloride short-term trends, as in the Figarden-
Bullard Area.  However, 5 other well hydrographs
had  nitrate short-term  trends opposite to those  of
chloride. The latter  group of wells were at or near
the  Clovis  Sewage  Treatment  Plant  and  were
apparently affected by sewage effluent.
     Analyses of  samples  of  wastewaters  and
ground water  were  plotted on a trilinear diagram.
Typical ground water in septic tank areas contains
less  than  10  percent sulfate  and less than 10 per-
cent  chloride on a reacting value basis. Ground
water  near the Clovis Sewage  Treatment Plant has
sulfate  ranging from  10 to 20 percent and chlo-
ride from  10  to  20 percent. Fresno County Sanita-
tion District  No. 1  and Clovis Sewage  Treatment
Plant  effluent have similar  compositions to the
latter ground-water samples.  Italian Swiss Colony
Winery effluent has about 10  percent chloride and
 30 percent sulfate.  A trilinear diagram  for ground
water in the Tarpey Village Area showed that septic
tank effluent, the Clovis Sewage Treatment Plant,
and the  Italian  Swiss Colony  Winery were  the
primary sources of nitrate.  Some  wells  had one
primary source, whereas other wells apparently had
more than one nitrate source.
     Seasonal nitrate analyses for District 8 wells in
 1967, 1969, and  1970 showed substantial decreases
 as the  summer  progressed. Virtually  all of these
 wells are on  canal banks  or  nearby,  and with the
 limited hardpan development of the area, abundant
 recharge  from canals is a certaintv. An analysis ot
      cr*                         "
 aquifer penetration revealed a nitrate stratification
 as in the  Figarden-Bullard Area. However, equiva-
 lent  nitrate  concentrations  were  about 40 feet
 deeper in the aquifer  beneath the Tarpey Village
 Area. The  primary reason for higher nitrate con-
 tents  at  greater depths is the presence of  nitrate
 sources  other than septic tanks.  Ground  water
 moving from beneath these sources to wells in the
 area follows  flow paths  that  take nitrate  deeper
 into the aquifer.
     On  the basis of available data,  wells at the
Clovis Sewage  Treatment Plant,  Clovis No. 8  and
No. 10, Well No.  8-1, and wells at the Palm Lakes
Golf Course have been primarily affected by sewage
effluent. Four of the 6 wells with nitrate over 45
ppm at  least once in the Tarpey  Village Area  were
near the Clovis Sewage Treatment Plant. Wells No.
8-2,  8-6, and 8-4 have been primarily affected by
septic tank effluent. Wells No. 8-3 and 8-7 and a
well at the Italian Swiss Colony Winery have  been
primarily affected  by winery wastewaters.  Non-
urban land immediately upgradient of Tarpey Vil-
lage is  idle and  underlain by ground water  with
nitrate contents less than  15 ppm. There is a  small
agricultural contribution to ground-water nitrate in
this area, because of the lack of nitrogen  fertiliza-
tion for several decades. Nitrate  from agricultural
areas further east is diluted by ground-water inflow
before reaching the Tarpey Village Area. The septic-
tank  part of  the  Tarpey Village Area  and the
-Italian Swiss Colony Winery are planned to be con-
nected  to the  Fresno  sewerage system within the
next several years. After these connections, nitrate
content will decrease in most ground water ot the
area, but the Clovis Sewage Treatment Plant proba-
bly will continue to be an  important nitrate source.

   MAYFAIR-FRESNO AIR TERMINAL  AREA
     The Mayfair-Fresno Air Terminal Area (Figure
 7) has  a  high  septic tank density  and  previous
 studies have related high nitrate contents in ground
 water to septic tank effluent. However, a number of
 potential sources are upgradient ot this area. The
               R. 20 E. i R. 21 E.
 Fig. 7. Mayfair-Fresno Air Terminal Area—waste dischargers
 are cross-hatched.
 150

-------
Clovis Sewage Treatment Plant, Italian Swiss Colo-
ny Winery, and Fresno County Sanitation District
No.  1  treatment  plant were previously discussed.
Anhydrous ammonia is widely used on agricultural
lands immediately east of the Fresno Air Terminal
and ground water in this area had a nitrate content
ranging from 15 to 25 ppm in 1970. About  500 to
600  acres were also irrigated in or near the Fresno
Air  Terminal in  1970.  Nineteen wells  have had
nitrate contents over 25  ppm and 6 of these wells
have had  nitrate  over  45 ppm at least once in  the
Mayfair-Fresno Air Terminal Area.
     The  Guild Winery or Alta Vineyards Winery
had  a  wastewater flow  of 16 million gallons in
1968. About four million gallons were disposed of
in Mill  Ditch and through  dry wells. This  waste-
water had the lowest waste strength of samples
from the  3 wineries on Clovis Avenue sampled by
the  California Department of  Water Resources
(1965). Sampling of wells at the winery and within
one-half mile downgradient of the disposal ponds
indicated  low nitrate  contents in 1970. Although
part  of this was due to the existence of moderately
deep wells, apparently the relatively small volume
of percolated wastewater minimized the contribu-
tion  to ground-water nitrate.
     The  Gallo Winery had a wastewater flow of
50 million gallons in  1968. Disposal of all waste-
water prior to 1970 was by land  flooding of about
260  acres south and east of the winery. This  winery
has the highest wastewater flow  of the 3 wineries
on Clovis Avenue; all of the water  was formerly
percolated; and the waste has the greatest strength
of all 3  wineries. A sample of pond water was taken
by the author in June, 1970 prior to connection of
Gallo Winery to the Fresno sewerage system. Total
nitrogen was 410 ppm, or about 1,800 ppm nitrate
equivalent. Two  of 3  wells sampled near the  dis-
posal ponds had nitrate over 45 ppm.
     Ground-water  flow is  primarily southwest in
the  northern portion and west  in  the  southern
portion. Depth to water  in 1969 ranged from 60
feet  in the  southeastern part to  90 feet  in  the
northwest. Water levels  declined  from  1958 to
1965 at about 4 feet per year, but have risen 1 to  3
feet  per year thereafter. Lithologic sections demon-
strate that this area is  in a transition zone from the
sands  and cobbles of the  northern F.C.M.A. to
clays of  the south. The area  is generally  free of
cobbles in the subsurface except for a lens in the
central  portion. A shallow,  thick clay occurs from
10 or  20 to about 60 feet beneath much of the
area.
     Soils of the  Mayfair-Fresno Air Terminal Area
are of  the Madera Series. Most of the soil north of
East McKimey Avenue is Madera sand, which has
no hardpan. Hardpan in the remainder of the area
lacks the impermeable nature of that in the older
San Joaquin soils to the north and east. Transmissi-
bility of most  of the aquifer is  about 180,000
gpd/ft. However, a lens-shaped area in the vicinity
of East McKinley Avenue has a transmissibility of
370,000 gpd/ft.
     Historic nitrate analyses show that all wells
sampled in 1951 had concentrations less-than  15
ppm. Nitrate  contents had  greatly increased  by
1963.  The number of wells with nitrate  content
over 25 ppm remained  relatively  constant from
1963 to 1969. Eighteen hydrographs had long-term
increases in nitrate. Most  of these hydrographs
were for wells with a cased depth of less than 150
feet. Six hydrographs  had  nitrate  decreases,  but
most of these  records were for short periods. Wells
with decreasing nitrate trends were  probably  af-
fected by canal recharge  during the wet years of
1967 and 1969. Fifteen hydrographs had constant
nitrate  trends and  most  of these  were  for wells
cased  to  at  least  150 feet depth.  Fifteen well
hydrographs had nitrate short-term trends similar
to those of chloride. These well hydrographs have
the dominant  trend of the Figarden-Bullard Area, a
trend related to septic tank effluent. Only  3 wells'
hydrographs  had  opposite  nitrate and  chloride
short-term trends,  and  these wells were down-
gradient from  the Clovis Sewage Treatment  Plant.
     Seasonal nitrate analyses for 1967, 1969, and
 1970 indicate  consistent decreasing contents during
the  summer,  as in the Tarpey Village Area. This
feature is also due to canal recharge. An analysis of
aquifer penetration revealed a step-like pattern of
nitrate as in the Figarden-Bullard and Tarpey Vil-
lage Areas. However, few shallow wells in the May-
fair-Fresno  Air Terminal Area have  low nitrate
contents. Thus, there is some  indication  of a
"nitrate cap" as described by Stout, Burau, and
Allardice (1965) in San Luis Obispo County, Cali-
fornia. Several wells in the area have nitrate con-
tents too high to fit the  step-like pattern.  Two of
these wells are downgradient of the Clovis Sewage
Treatment Plant and 2 others are downgradient of
the Gallo Winery.
     A trilinear diagram was prepared for  samples
of wastewater and ground water  in the  Mayfair-
 Fresno Air  Terminal Area.  The anion portion was
most  useful and suggests that septic tanks have
 contributed the majority of nitrate to the ground
 water.  This diagram  and chloride concentrations
 indicate downgradient movement of sewage efflu-
 ent  for at least 1 or 2 miles from the Clovis Sewage
 Treatment Plant. Similar data indicate downgradi-
 ent movement of winery wastewater from the Gallo
Winery ponds for  at least  1 or  2 miles. Ground-
                                             151

-------
water velocities are estimated to range from several
hundred  to 1,000 feet per year in this area. The
downgradicnt  effect of  recharged wastewater is
limited by dilution and dispersion in the aquifer.
     Most wells with  high  nitrate contents  are in
the southern part of the area, and no wells with
high nitrate are in the central  part. Hardpan exists
in some  parts  ot the northern portion  and may
limit the percolation of high nitrate waters. Cobble
zones disperse  and dilute the high nitrate recharge
waters in the central portion. Ground water in the
Mayfair-Fresno Air Terminal Area with nitrate over
25  ppm  in  1969 covered an area larger  than any
other part of the F.C.M.A., except near the Fresno
Sewage Treatment Plant. This situation is related to
the high density of septic tanks and the presence of
other nitrate sources. The absence of cobble zones
and impermeable hardpan favor higher nitrate con-
tents  in ground water. The lower transmissibility,
as compared to the  northern part of the F.C.M.A.,
also lessens the  dilution of  high-nitrate  recharge
waters.  Nearby upgradient agricultural lands con-
tribute some nitrate  to the urban area. The Maytair-
Frcsno Air Terminal Area septic tank portions be-
gan conversion  to the Fresno sewerage system in
1970, and  the  Guild and Gallo Wineries  were also
connected  in  summer,  1970.  The future nitrate
situation will  greatly improve  in  this area with
these connections, but the Clovis Sewage Treatment
Plant will  probably continue  to  be an important
nitrate source to the northeastern portion.


         DOWNTOWN FRESNO AREA
     The  Downtown  Fresno Area (Figure  8) has
been  sewered since  1891  and no sewage treatment
plants are  within several  miles of the area.  Settle-
ment occurred  in the late nineteenth  century, and
most  of the early industrial development occurred
here.  Upon  careful examination, a  number of
                             R ZOE.
 Fig. 8. Downtown Fresno Area—present and former waste
 dischargers are cross-hatched.
potential nitrate sources  can  be found in or near
the Downtown Fresno Area.  Urban parts of the
city surrounding the area used septic tanks many
years  ago.  The townsite ot Calwa,  which covers
about  500  acres,  presently  is unsewered and is
one-half mile southeast of the area. The Mayfair-
Fresno  Air  Terminal  Area is  also unsewered  and
extends to within one-half mile of the  Downtown
Fresno  Area. Ten  wells have  had  nitrate  over 25
ppm at least once and 2  of  these wells  have had
nitrate over 45 ppm. Industry and agriculture also
comprise potential sources of nitrate to  the area.
     Meat-packing plants have high nitrogen wastes.
The concentration of organic  plus ammonia-nitro-
gen in such waters is quite high, ranging from 100
to 300 ppm, or 440 to 1,320 ppm nitrate equiva-
lent (American Water Works  Association, 1967).
Two meat-packing plants are within 1.5 miles and
southwest  of  the  area. Historically, a number of
industries used independent waste disposal facilities
prior to connection to municipal sewerage systems.
Roma Winery has a wastewater flow of 1.0 million
gallons per  day  and is at the southeastern corner
of  the  Downtown  Fresno Area.  This  winery was
unconnected until  1936.  Mont LaSalle Vineyards
or Christian Brothers Winery is one mile east ot the
Roma Winery and was unconnected prior to 1953.
Both wineries  may have  contributed substantial
amounts of nitrate  before connection to municipal
sewerage systems. Several chemical industries used
independent waste  disposal facilities ir. 1969, but
wastewater flows  were too  small  to  affect wells
thousands of feet distant.
     Many cooling water return wells are  in or near
the  Downtown Fresno  Area. A previous study
stated  that water quality was unaffected by  the
cooling process (California Division of Water  Re-
sources, 1952).  However,  the highest  nitrate con-
tents in the F.C.M.A. in  1951 were sampled near 2
recharge wells. High nitrate ground water also was
found  near the Fresno  County  Hospital  cooling
return  well in 1970.  Apparently,  either recharge
water   is enriched  with  nitrate derived  from  the
refrigerant, or  other  sources of nitrate have been
injected in some of these wells. The Fresno Sewage
Treatment Plant is about 3 miles  southwest of the
Downtown Fresno  Area.  The California  Depart-
ment  of Water Resources (1965)  assumed  that
sewage effluent from this plant had  not  reached
the Downtown Fresno Area by 1963. Most histori-
cal water level contour  maps have shown a slight
gradient towards the treatment plant  from down-
town Fresno. However, these maps have been based
on only a few wells, and  were usually drawn for
December conditions. Ground-water elevation con-
tours  for  the fall  of some  years, such  as  1951,
 152

-------
show either a flat water table or a gradient toward
the Downtown Fresno Area.  Northeasterly move-
ment of ground  water from the Fresno  Sewage
Treatment  Plant toward the urban depression cone
was possible during summer pumping conditions.
     Agricultural areas southwest of the Downtown
Fresno  Area may  contribute nitrate to  ground
water,  as anhydrous  ammonia is in wide use. Large
dairies  are  additional sources of  nitrate  in this
vicinity. The Fresno County Disposal Yard is within
1  mile  of  the Downtown  Fresno Area,  and solid
waste disposal  sites  are  potential nitrate sources.
Several  old cemeteries lie northwest of the area and
may have  contributed nitrate decades ago  when
waterlogging occurred over much  of the F.C.M.A.
     The  extreme southern  portion  of the urban
cone of depression in the  F.C.M.A.  is beneath  the
Downtown Fresno Area. Ground water can poten-
tially enter the area from almost any  direction.
Depth to water in 1969  ranged from 50 feet in the
southwest  to  83  feet in the  northeast.  No major
canals  traverse the area except for Dry Creek Canal
at the  northwest  corner.  Water levels declined until
1966,  and have stabilized  up  to  the present. The
Downtown Fresno Area  has  abundant clays in the
subsurface, especially below  several hundred feet
depth.  A thick clay occurs over much of  the area
at a depth interval  of 10  or 15  to  about  75 feet.
Soils are  of the Madera Series,  and Madera sand
occurs  throughout much  of  the  area. Transmissi-
bility  is highly variable and ranges from  60,000
gpd/ft  in the nortrmest to  370,000 gpd/ft in the
northeast.
     Almost all wells in the F.C.M.A. with nitrate
contents over 15 ppm in 1951 were in or near the
Downtown Fresno Area.  This was prior to extensive
development of suburban  areas on septic tanks and
nitrogen  fertilization on  agricultural areas. Wells
with nitrate  over  25  ppm  in  1958 were in an
elongate  area! distribution,  which  trended  north-
east-southwest. Significant amounts of chloride are
present in ground water of the Downtown Fresno
 Area.  This is in  contrast  to  the  Figarden-Bullard,
 Tarpey Village,  and Mayfair-Fresno Air Terminal
 Areas.
     Most wells with iong-term nitrate hydrographs
 have either decreasing  or constant  trends. Short-
 term chloride and nitrate  trends were similar for 5
 weii hydrographs, but opposite  for  5 others.  Sea-
 sonal  nitrate  analyses for 1969  and 1970 showed
 both  increasing and decreasing contents as  the
 summer progressed. An  analysis of aquifer penetra-
 tion showed no  stratification of nitrate as in septic
 tank areas of the F.C.M.A. However, there was a
 lack of data for aquifer penetrations of less than 8O
 feet. Many wells with aquifer penetrations ranging
from less than 60 to more than 160 feet had nitrate
contents between 15 and 25 ppm. This distribution
suggests that nitrate did not  originate from a sur-
face source  in the  immediate  vicinity  of the
Downtown Fresno  Area.  A  trilinear  diagram for
ground water in the area demonstrated at identical
composition to ground water  at the Fresno Sewage
Treatment Plant.
     Several  wells  with nitrate content over 15
ppm are  within  one-half mile of cooling return
wells. Two wells with nitrate  content over 1 5 ppm
are near a meat-packing plant. One well with nitrate
content over 15 ppm  is located between the Roma
Winery  and Mont LaSalle Vineyards Winery. Two
wells with nitrate content over 15 ppm are  near
septic tank areas. However, there are at least 5 wells
with nitrate  over 15  ppm for which a source is
not immediately apparent.  Several facets of  the
nitrate distribution in ground water of the Down-
town Fresno Area should be explained. Key factors
are  the  relatively  high  nitrate  contents of  the
1950's. the high chloride content, the decreasing
long-term nitrate trends, the abundance of opposite
nitrate and chloride short-term trends, the presence
of nitrate in concentrations above the background
level at considerable depth in the  aquifer, and the
trilinear plot for ground water.
     Water samples were collected from many wells
in July  and August,  1970,  in the  area  between
downtown  Fresno and the Fresno Sewage Treat-
ment Plant. Chloride was analyzed by the Mohr
titration  method utilizing a Hach field testing kit.
 Zones of  chloride content greater than  25  ppm
were delineated  (Figure  9).  The  largest zone ex-
tended from the Fresno  Sewage  Treatment Plant
northeast to the Downtown Fresno Area. Smaller
zones were near the Clovis Sewage Treatment Plant,
 the Gallo Winery, and the Fresno County Hospital.
 Chloride content was also greater than 45 ppm in
 the smaller zones.  The highest chloride contents in
 F.C.M.A. ground water in 1970 were monitored at
 the Fresno Sewage Treatment Plant.
      Historical records in Fresno  state that water-
 logging occurred at the treatment plant site in the
 earlv  1900's This problem was solved by ground-
 water pumpage at the plant  site and  in the nearby
 urban area in the 1920's. it is significant that the
 original treatment plant  site was one mile east of
 the present plant, and thus  one mile closer to the
 Downtown Fresno Area.  Sewage effluent from the
 Fresno  Sewage  Treatment  Plant has played an
 important part in  recharge to ground water of the
 F.C.M.A. since 1891. The northeasterly movement
 of ground  water beneath the treatment plant was
 initiated with the  greatly increased pumpage in the
 urban area between  1910 and 1920. Some wells in
                                             153

-------
Fig. 9. Distribution of chloride in ground water in 1970—
contours are in parts per million and circles are locations of
sampled wells.
the Downtown Fresno Area were monitoring fairly
hiilh chloride contents by 1930.  Historical chloride
analyses  from  that  time to  the present  show  an
elongate  pattern  ol  wells \\ith high chloride con-
rents.  'The northeast-southwest  trend  of  the high
chloride  zone in the Downtown Fresno Area im-
plies .1 source TO  the northeast or southwest. How-
c\er,  as  no hisih  chloride source ot large volume
occurs to the northeast, the source must he to the
southwest.
     The relatively   high  nitrate contents  ot the
195()\  were  related  to ground-water intlow for
several  decades  from the  vicinity  ol  the  Fresno
Scwaee  Treatment Plant. 1'his inflow was accentu-
ated  bv  urban  water level declines. However,  by
!°65 v. ,iter A-\-, .s ;-.ad stabilized ,n  the Downtown
l-rcsjio Area and  : "it low from t!u southwest dimin-
 -iied. Ihus.  decreasin_  trend-,  have occurred ior
m.mv  nitrate  hvdroijraphs. espcuallv over the past
uec.uk.  I he hi.,a ch!o<->dc contents of the Down-
   \\ n Fresno  ,\;c;: .a c a direct ici.e^uon ot r-C'1. age
efliuent. as chloride i> ncarlv ai. ideal  tracer in this
cm ironmcm.  I'he  abundance ot  opposite nitrate
and  chloride  short-term  trends and  the  trilinear
,-/ )t  of  ground  w.rer  confirm that  the  Fresno
Srw.iuc  I reatment  Plant  has  been  the  primary
source of  nitrate. Nitrate at  considerable depths in
the aquifer  was  caused by  movement  of nitrate
from  a  source several miles  distant. Flownet con-
siderations demonstrate that sewage effluent from
the mound beneath the plant would tend to move
 into deeper portions ot  the aquifer.
     Septic tank effluent from the Mayfair-Fresno
Air Terminal  Area has reached  the northeastern
portion of the Downtown Fresno  Area and  has
been the dominant nitrate source since about 1963.
Septic  tanks  at Calwa have  contributed  some
nitrate to the  southeastern part of the area. Winery
wastewaters  have  likely   contributed  nitrate  to
ground water in the southeastern portion  of the
area.  Meat-packing  plants have contributed  nitrate
near the southwestern portion. However, the domi-
nant source has been  the Fresno Sewage Treatment
Plant.
     With  the planned conversion  of septic rank
areas and industries on independent waste disposal
units to municipal sewerage systems,  it is inevitable
that nitrate conditions will greatly improve \\ ;th--i a
period of years  in most ot the F.C.M.A.  However.
in the vicinity of the urban depression cone, Citrate
contents will likely  increase.  This is due  to th<'
planned concentration ot almost  jll ot  the rot.il
nitrogen load of the  urban area at one site. I nU 
-------
sewage  effluent in the  aquifer and water samples
from wells  plot in a  characteristic part of the tri-
linear diagram. Short-term trends for chloride and
nitrate  hydrographs  are commonly  opposite near
sewage  treatment  plants in the F.C.M.A.
     The highest  nitrate  contents  occur  in  the
upper  50 or 60 feet of the aquifer beneath septic
tank areas.  Nitrate stratification is obscured when
other sources are  present,  due to the derivation of
nitrate  from nearby areas  by ground-water  inflow.
Trilinear diagrams for constituents in ground water
and short-term nitrate and chloride trends in septic
tank areas  differ from those for  ground  water
affected by sewage effluent.
     An analysis  of two high nitrate areas near the
Fresno Air  Terminal  indicates  septic  tanks con-
tributed most of the ground-water nitrate. How-
ever, wineries and the  Clovis Sewage Treatment
Plant are significant  local sources.  Winery  waste-
waters  were sampled  and had  a total nitrogen con-
tent ranging from 4  to  10  times  that found in
typical  domestic  sewage. Wineries  have been im-
portant sources  of ground-water nitrate, as they
have existed since the early 190()'s. The agricultural
contribution of  nitrates to ground  water  in the
urban area  is small due to  dilution as downgradient
movement occurs.
     Ground water beneath the  Downtown Fresno
Area is in  a significant hydraulic depression cone.
Extensive northeasterly  flow of ground water from
the mound  beneath the Fresno  Sewage Treatment
Plant   took place  for  many  decades. Moderate
concentrations of  nitrate occur  at  depth  in the
aquifer and are  related  to the  inflow of sewage
effluent  from a  source several  miles  distanr.  Al-
though other sources have  contributed nitrate to
the ground  water, sewage effluent is the dominant
source  as  indicated  by  chemical and  hydrologic
data.
      Twenty-six  wells in the F.C.M.A. have sam-
pled nitrate over 45 ppm at least once since 1947.
 Seven  of these  wells have been near the  Fresno
 Sewage Treatment Plant.  Six wells west and south-
west  of the  Fresno  Air Terminal  have sampled
nitrate greater than 45 ppm at least once. Five wells
in the  vicinity of  the Clovis  Sewage Treatment
Plant had nitrate  exceeding 45 ppm at least once.
Other ground  water with  nitrate greater than 45
ppm in the F.C.M.A. was in the Downtown Fresno
Area,  in septic tank portions of Figarden-Bullard
Area,  in the  agricultural areas north and  east of
Fresno, and near the Gallo Winery.

                  REFERENCES
American Water  Works  Association.  1967.  Sources of
     nitrogen and phosphorus in water supplies. Journal of
     American  Water Works  Association,  Task Group
     Report, March, 1967, pp. 344-366.
Behnke, J. J., and E. E. Haskell. 1968. Ground water nitrate
     distribution beneath Fresno,  California. Journal of
      American Water  Works Association, v. 60. no. 4, pp.
     477-480.
California Department of Water Resources. 1965. Fresno-
      Clovis metropolitan area water quality investigation.
      Bulletin 143-3.
California Division of Water Resources. 1952. Investigation
      of cooling  water return,  City of Fresno. Report to
      Central  Valley  Regional Water Pollution Control
      Board, Project no. 52-5-10, 44 pp.
Jenks, J. H., and P. L. Adamson. 1968. Metropolitan area
      wastewater treatment  and disposal  facilities. Report
      to City of Fresno, Department of Public Works.
Johnson, W. R., F. Illichadieth, R. M. Daum, and A.  F.
      Pillsbury.  1965. Nitrogen and  phosphorus in tile
      drainage effluent. Soil  Science, v. 29, p. 287.
Nightingale, H. I. 1970. Statistical evaluation of salinity and
      nitrate content  and trends beneath urban  and agri-
      cultural areas—Fresno, California. Ground Water, v. 8,
      no. l,pp. 22-28.
Page, R. W., and R. A. LeBlanc. 1969. Geology, hydrology,
      and water quality in the Fresno Area, California. U. S.
      Geological  Survey  Open-File Report, Menlo  Park,
      California,  189 pp.
 Schroepfer, G. J., and R. C. Polta. 1969. Travel of nitrogen
      compounds in  soils.  Sanitary Engineering Report
      172-5, University of Minnesota.
 Stout, P. R., R. G. Burau, and W. R. Allardicc. 1965. A
      study of the vertical movement of nitrogenous matter
      from  the  ground  surface to the water table in the
      vicinity of Grover City and Arroyo Grande, San Luis
      Obispo County. Report to Central Coastal Regional
      Water Pollution Control Board, 51 pp.
                                               DISCUSSION
 The following questions were answered by Kenneth
 D. Schmidt after delivering his talk entitled "Nitrate
 in Ground Water of the Fresno-Clovis Metropolitan
 Area, California."
 Q.  Can you describe the hardpan geologically?
 A.  This hardpan has been described by M, G. Croft
 and R. J.  Janda of the U. S.  Geological Survey as
 part of a  study of the alluvial formations of the
 eastern San Joaquin Valley. It is in the oldest soil of
 the area, is cemented by iron salts, and belongs to
 the noncalcic brown soils that occur in the eastern
 part of  the  San  Joaquin Valley. The hardpan  is
 composed of firmly cemented soil particles,  and is
 dense and impervious.  It has a low humus content
 and contains no alkali.
 Q.  How  did  the nitrate  (NO~2), ammonium, and
 organic nitrogen  vary with the nitrate distributions?
 A.  These are important parameters to consider in
 the ground water besides nitrate. There is less than
 0.1 ppm ammonium or nitrate in the aquifer under-

                                               155

-------
neath the septic tank areas. At the Fresno Sewage
Treatment Plant there are some ammonia contents
between 0.5 and 1 ppm. There is an inverse relation
between ammonia content and nitrate content on a
seasonal basis. Some wells have shown as much as 5
ppm  ammonia, which is  an indication  that the
ability  of these  soils to  continue retaining the
ammonia  form has perhaps been  reached. Nitrite
contents have generally been less than the limits  of
detection  in  all ground water of this area, as sam-
pled by pumping  wells.  Organic nitrogen has not
been tested for to a large extent, but contents are
below the limits of detection.

Q. Did you determine any oxidation potentials?
A. No, I didn't, and perhaps this would  be a very
important  thing  to do. There are several  areas
where  investigators have  actually  worked  with
this,  such as  Stewart  and others in Colorado.
Apparently,  there is  good  correlation  between
oxidation  potential and nitrate above the water
table  in some areas and a poor correlation in other
areas. Obviously,  studies in the unsaturated  zone
would  be necessary  to completely  understand
nitrate in  ground water of  the Fresno area. Oxida-
tion  potential would be an important parameter;
however,  my analysis was  confined primarily  to
water in the aquifer.

Q. Has there been any plan to  line evaporation
ponds at  the Fresno  Treatment Plant? Are many
of these ponds already lined?,.
A. No ponds are lined at present.  About 80 to 85
percent of the effluent disposal at the Fresno City
Sewage Plant is by percolation. If the ponds are
lined, then a much Larger  waste disposal problem
will occur. Approximately 1440 acres of percola-
tion ponds are  necessary  at present and on the
average of 10 percent of the effluent is  used for
irrigation  on nearby farms. There is some evapora-
tion  also, which makes up  the remainder. I think
the best solution will not be to line the ponds but
to continue percolation of the water. The solution
involves the  nitrogen  compounds themselves, be-
cause recent literature suggests that with refined
technology we  are learning how  to control the
total  nitrogen  content  of  effluent.  The City  of
Tucson Sewage Treatment Plant has been able  to
substantially decrease  total nitrogen contents of the
effluent in recent years.

Q. Ken, you mentioned the  use of agricultural
fertilizers.  Are super-phosphates used in  this area?
Also,  what fluoride levels were found in the ground
water?
A. About  90  to  95  percent  of  the agricultural
fertilizer used in this  area is anhydrous ammonia.
Farmers  use virtually no  significant amounts of
super-phosphates to my knowledge. Fluoride levels
are below  0.1  ppm in this whole  area except in
two  septic tank areas where they use fluoridation
units. Fluoride  levels ranging from about 0.1 to
almost 1 ppm have shown  up in the ground water
within 3  to 5  years since installation of these
fluoridation units.

Q. Here's another one on agricultural sources and
nitrates.  To what extent have agricultural sources
of nitrates  been examined in the  San Joaquin  Val-
ley?
A. Many studies along this line have been done. If
one looks at the nitrogen literature  not only in the
San Joaquin Valley but  elsewhere in the country,
one may conclude that prejudice  has been evident
among some investigators.  In the  Fresno area, agri-
cultural workers have neglected the wineries, and
have generally  not  considered the  effect of ferti-
lizers on ground water near urban areas. One com-
mon argument  is that plants use all  of the nitrogen
in the fertilizer. I think we have to recognize  that
the true  position is somewhere between no effect
and  a  substantial effect.  That  is,  most farmers
don't want  to  apply  fertilizer  that  would  be
leached  in great amounts. A farmer is going to try
to be efficient,  but he usually is not going to be
100  percent efficient and recover all of the nitro-
gen added. Agriculture contributes to nitrate in irri-
gated areas near my urban study  area. The nitrate
contents beneath agricultural areas varied with the
transmissibility and other factors, and ranged from
15 to 25 ppm where the transmissibility was lowest.
So in the Fresno area, agriculture  near the urban
area  is not really causing nitrates that are over 45
ppm.
     One of the important problems now in Cali-
fornia is how to get rid of the nitrate that is going
to be in the California master drain when it is built.
They have done experiments at Firebaugh, which is
northwest of Fresno,  and have been quite success-
ful in denitrification.  About 25 ppm nitrate is the
average  nitrate  content in irrigation return water
in the San  Joaquin Valley. Now,  in some areas
nitrates  occur in amounts  over the  limit and  it is
due  to  agriculture. Many factors  must  be con-
sidered, such as fertilizer practice, irrigation, aquifer
characteristics, and well construction.

Q. / have 2 questions here I think  are similar to
what you're discussing. Has spray irrigation been
considered  to recycle the nitrate  back to the agri-
cultural area?  Aren't these  high  nitrate  waters
worth something as an irrigation water source?
A. This  is  something that people  working near
Phoenix, Arizona have considered. Herman Bouwer
156

-------
presented a paper on this at the Arizona Academy
of Science meeting in April. It  seems logical to
consider using sewage   effluent  for its  nutrient
value. One major problem is that in  the future we
are going to, in places such  as Arizona and Cali-
fornia, grow  more and  more high yield crops. A
high nitrogen load all the time is not desirable on
these crops because of rank growth  of the vegeta-
tion.  However,  effluent may be used on  field
crops such as safflower or may be diluted and used
on many crops.
     The idea in Phoenix is to have  sewage plants
along a dry stream bed, with a number  of percola-
tion basins.  Water  will be  recharged  and  later
pumped by wells after moving a sufficient distance
through the soil. They believe that by operating the
ponds  in a certain  manner, such as controlling the
aerobic  vs.   anaerobic  conditions,  one  can get
almost any nitrate content desired in the recharged
effluent. Hopefully,  they will be able to deliver
the nitrate load that the farmer wants when he
wants  it. One can imagine that with different crops
growing and a number of different farmers, this is a
formidable task. The concept of utilizing the nutri-
ent value of nitrate in well water or sewage effluent
is worthwhile and will eventually be perfected.

Q.  In  your discussion, one person wants to know
how did you eliminate lawn fertilizers as a nitrate
source in the suburban septic tank areas? Was there
no early fertilizer usage in the Fresno area?
A.  Let me answer the second part of that question
first. The important point about fertilizers is the
drastic  change that  occurred between  1950 and
1970, when nitrogen fertilizer use increased by 600
percent. Fertilizer use in this area was  small prior
to  1950.  As regards lawn fertilizers in the septic
tank  areas,  that's a very good  question. In the
septic  tank areas where  there is hardpan,  nitrogen
compounds  will not percolate and  generally will
not enter the ground water, in other areas where
the impermeable hardpan is absent,  nitrates could
build up in the ground water. I did not scientifically
eliminate lawn fertilizers as a source, but  I think a
simple calculation, if we knew the amounts used,
would show  only  small  accretions to the ground-
water nitrate.

Q.  Shouldn't  the  inverse relationship you noted
be between nitrate content and the specific ground-
water  discharge rather  than the transmissibility
alone?
A.  This is a  point that is theoretically  sound. My
concept is that we are  adding nitrate at  the land
surface, and  that it  will be diluted in relationship
to  the transmissibility  of the aquifer.  Of course,
water  flows through the aquifer based not only on
transmissibility, but on the hydraulic gradient and
width of flow. However, the gradient tends to be
uniform in  each one  of the small areas studied.
Thus nitrate content on a local scale does depend
primarily on the transmissibility, as the gradient is
rather uniform.

Q.  What percent  of irrigation water applied to
fields is returned to ground  water by infiltration?
What percent is this of the total recharge?
A.  Let's consider  eastern Fresno County and the
agricultural  areas in it. The  best estimates, which
are fairly reliable, indicate that farmers are apply-
ing about 4 to  5  acre-feet per  acre per year. The
consumptive use  of  the crops growing  there  is
about 2.5 acre-feet per acre  per year. The percent
of applied water that returns to the ground water
is thus about 30 to 40 percent.

Q.  Let  me  finish  the  other  part of  the  last
question—what percent is this of the total recharge?
A.  I don't  have the vaguest idea. We don't really
have valid estimates of the canal recharge and other
components of the water  budget.  Return flow
comprises one of the  major unknowns  in many
water budget studies in  semiarid and arid environ-
ments.

Q.  How about naturally  occurring  nitrate as in
some of the salt pans in the inner mountain basins,
such as  Owens Valley, Death  Valley, etc.?
A.  J.  H. Feth, in 1966,   summarized  data on
nitrogen compounds in  water. He points out that
natural  sources of nitrate are associated with caves,
playa deposits and caliche. None of these occur in
or near  Fresno's urban area.

Q.  Do  you think  that your  sampling frequency is
sufficient to  really establish confidence  in water
quality  trends?
A.  The frequency  you're probably referring to is as
shown  on the water quality hydrographs. I more or
less financed this  study myself; for this and other
reasons, I wanted to take an area where the  data
was available and see what could be done. Now it's
probably  easier to go  out with a large sum of mon-
ey, take a large number of water samples,  and have
good data  to interpret. However,  as  a  practical
problem,  we should sometimes force ourselves to
us.e available data.  I think the trends are  meaning-
ful. I looked at all available analyses for specific
wells and used  judgment to cast out erroneous or
atypical values. There  are many problems, such as
laboratory  errors,  pumping  time,  contamination,
and time between sampling and analysis. Of course,
different  methods have been used at  different
times.  Some  of these problems cannot be solved
                                             157

-------
because these are the only historical data available.
Because  these  trends  could  be related  to  actual
waste disposal  operation at  the  land  surface and
hydrogeologic factors,  I would argue that they are
meaningful. However,  they  are by no means  the
ultimate—and  more  detailed  sampling is necessary
for precise  interpretation.

Q.  This might  be along the same line, Ken.  It says
here  the  almost perfect  positive correlation  of
nitrate and chloride in septic tank areas and  the
almost perfect negative correlation for sewage plant
effluent  hardly appear  to be just chance  occur-
rences.   Can  you  speculate  about  any  casual
mechanisms?
A. The  similarity of chloride and nitrate behavior
in  water,  of course,  is well documented  in  the
literature and would be expected from geochemical
and other  considerations.  Chloride and nitrate are
both soluble  and  mobile in  most  ground-water
systems. There is a correlation between the chloride
and nitrate in  septic  tank areas because chloride
and nitrogen  compounds are in the  effluent,  and
these ions have similar geochemical characteristics
in  the aquifer; they are both soluble and they are
not adsorbed or exchanged.
     The opposite trends are more difficult to ex-
plain.  An  explanation may be somewhat theoreti-
cal, but it is  interesting.  There was evidence of
denitrification between the Fresno Sewage Treat-
ment Plant. I believe that many workers have over-
emphasized denitrification and have  used it when
the other  items didn't balance out. I believe that if
one considers chloride or some other ion that does
not undergo  denitrification or nitrification, valu-
able conclusions  can  be drawn.  Chlorides-in the
ground  water  for  quite  a distance around  the
sewage treatment plant are generally the same as in
the effluent. This indicates that there is little dilu-
tion of recharged effluent near  the plant.  This is
true because  the quantity of effluent is so large,
about 25,000 acre-feet per year. However, nitrates
are not correspondingly  high  in the same  area,
although some high contents do occur. Denitrifica-
tion completely  eliminates nitrogen compounds, as
they go off as gaseous forms and  are no longer in
the water. Nitrate would  tend to be lower, gener-
ally due to the anaerobic conditions  developing in
the area, and denitrification.
     Another alternate could be related to a con-
cept discussed  last year  in the Journal AWWA.
Malhotra  and  Zanoni discussed  the interference
effect of  high chloride ion contents  on  nitrate
determination. High chloride contents near sewage
plants  could  cause interference with the phenoldi-
sulfonic method  for nitrate determination.  Thus
with higher chlorides, nitrates may be erroneously
low in  some cases.
Q. Were  bacterial  analyses made  on  water so
obviously contaminated by primary waste effluent
from the Fresno Treatment Plant?
A. Bacteriological sampling is something I did not
get involved with to a great extent. There are only
1 or  2 wells in  the urban area that have shown
biological  contamination.  This is largely, in my
opinion, due to the regulations that are imposed
by the county for septic tank construction, that
the wells are generally  deep, and  are not usually
open in the upper 40, 50 or 60 feet of the aquifer.
I am not familiar with biological sampling near the
Fresno Sewage Treatment Plant.

Q. Could a source  of nitrate in downtown Fresno
be leaking sewers? Could uniform vertical distribu-
tion be attributed to large vertical flow components
associated with pumping, particularly in the central
portion of the cone of depression?
A. As for  the first  question, there were some areas
not near sewage plants where chemical data point-
ed to  sewage effluent  as  the source.  A possible
explanation  would  be  related to  leaking sewers.
However, I made no positive correlation; in the
future, it might be possible to do this.
     In regard to the uniform vertical distribution—
my reasoning is  that effluent has moved from  a
mound beneath the Sewage Treatment Plant toward
downtown  Fresno. This water would spread  out in
a vertical  sense,  as one can visualize  by flownet
consideration. Hence, no stratification occurs as in
the septic tank areas. The idea that the uniform
distribution of nitrate is due to vertical flow is a
good one, except for the lithology. The alluvium in
that area contains  a number of clay layers,  much
more  than in any  other part of the metropolitan
area. Vertical flow  would be greatly restricted, but
still  possible. There is  a  lot  of pumping in the
urban  depression cone,  and there may be a relation
to distribution. But there is overwhelming evidence
for the case I have presented. Historical nitrate and
chloride data, water levels, trilinear diagram  plots,
and  chemical hydrographs indicate  movement of
effluent from the Fresno Sewage  Treatment Plant.
Perhaps  the most  conclusive data  are the high
chloride contents between the plant and the  down-
town Fresno area.
 158

-------
The  Use,  Abuse  and  Recovery of  a  Glacial   Aquifer
by Edward M. Burr
                  ABSTRACT
     The inter-relationships  between an industrial  plant
and a shallow sand aquifer of glacial origin are described.
These relationships include industrial and  potable  water
supplies, industrial and human waste water disposal systems,
the hydraulics of the pollution of the ground-water aquifer
and the types of corrective actions taken to re-establish the
wise use of the ground-water resource. Also reviewed are the
ground-water movement-quality relationship, veil designs,
method of drilling, well redevelopment  and ground-water
recharge.

                INTRODUCTION
     In the oast few years, since the general public
discovered ti:_ v.se, if not the  literal definition, of
the  words "ecolog}'"  and  "environment"  many
industries of this nation have received "black eyes."
Some are deserved, but as in everything else,  there
needs to be a balance of  perspective. A purpose of
this  paper is to illustrate  the efforts of one manu-
facturer  to maintain  a peaceful coexisicnre with
the environment. The efforts have  spanned almost
fifteen  years—during  which time  scientists   ;/>m
public agencies and  private concerns, administra-
tors, engineers, architects and production personnel
have  worked  together, agreed and disagreed, but
have  achieved the goal of preserving  the  environ-
ment.
     Another  purpose is  to  illustrate  the   inter-
relationships among the ground disposal of wastes,
the ground-water supply and well design, construc-
tion, and  methods  necessary  to affect rehabilita-
tion. While the specific information given is related
to a  single site, some of the  conclusions drawn
reflect experiences with other similar projects.
     The firm involved is a manufacturer who pro-
duces chemical intermediates  related  to  the  phar-
maceuti<" i! anJ agricultural industries. Some of the
products -trf basic, while others are finished chemi-
cal nrocacts  of hlihiv complex, composition. The
corporation  began on a  small  scale about fifteen
years  ago. The corporate capability  and product
diversifies ion rapidly increased and  in  time the
companv nv.Mcea \\ith a major  national corporation
of international repure.
         GEOHYDROLOGIC SETTING
     The site is in Michigan and is situated within
a surface  watershed of approximately  12 square
miles (see  Figure  1). This shed  is tributary to a
larger  watershed,  shown  as  Creek  "B," which
eventually drains into one of the Great Lakes.
     aPreiv;-,,ed at  uie  NiatK.i.a!  CrounU  Water Quality
 Symposium,  Denver, coluuido, August Js-J/  19< 1.
     ^Geologist, Williams & '\,"ks, 250 Michigan Street
 N.E., Grand Rapids, Michigan 49503.
Fig. 1. Schematic location map.
     The company chose the sire because it offered
access to  a major population center, a large labor
market, an abundance of ground water and reason-
able  isolation  from  intense  development.  The
original facilities are shown in Figure 2. The major
structures  consisted of one building  to  house ad-
ministrative  offices,  laboratory and  maintenance
facilities;  another buila;ng for production; and a
series of ponds  for the ground disposal of cooling
and waste waters.
     The potable and industrial water-supply needs
were  originally  met by a single well tapping the
unconsolidated glacial sands. This  tubular well was
6  inches in diameter and  supplied approximately
50 gallons per minute. Water storage and pressure
modulation was accomplished by a hydropneumat
ic tank.
     A septic tank and tile field system adjacent to
the administration building accommodated the san-
itary  wastes. Since  both  industrial  and human.
Bastes were disposed to the  ground, shallow ob-
servation   wells  were  installed  :--'tween  the ad-
ministration building and the nenivst residence. The
responsible regulatory agency required the periodic
Dimpling of the observation well waters tor possible
ground-water pollution.
     The  growth of the company led to a staged
expansion  program to develop the site as shown in
                                            159

-------
Fig. 2. Schematic of original site development.
n
                        .
                        u
                               ca*t.*e <#u*r
                         ...  00s-'
     0
Fig. 3. Schematic of current site development.

Figure 3. However, during this expansion, pollution
of the ground water was detected.
     At that point in time, it was necessary that the
company accomplish the following:
     1. Develop a bacterially and  chemically safe
water supply for their personnel and the production
processes.
     2. Contain the  pollution within  the bound-
aries of their property.
     3. Prevent additional pollution by developing
suitable waste water treatment processes.
     4. Restore the  quality  of damaged ground
waters.
     5. Maintain  and expand the manufacturing
capabilities of the firm.
     In the past, the  priority  of these needs would
assume  a   different  order,  depending upon  the
individual  firm involved. In today's complex soci-
ety and  in  light of modern science and technology,
it is necessary that all be accomplished simultane-
ously.
     The plant site  lies  within a  large lake plain
ancestral to the present Great Lakes. As such, the
surface soils are sandy lake beds. "The sand deposits
range from 60-120 feet thick,
     The surface watershed and  'he ground-water
watershed are nothydrologically identical. Original-
ly the ground watershed was larger than the surface
watershed. The initial static water level of the water
table was  approximately  5  feet below  the land
surface.
      Figure  4  indicates the  original  direction of
ground-water flow within the unconsolidatcd glacial
aquifer.  The ground disposal of industrial  waste
water up the hydraulic gradient from the original
well  in  conjunction with the  company's  use of
water, modified the original direction ot  flow and
resulted in the pollution of the water supply. Thc
initial observation wells did not  show the  pollution
of the  aquifer because the waste water was inter-
cepted  by the  well which furnished the company's
total water supply.
      The records of deep wells surrounding the site
indicated  the  presence  ol another aquifer. This
aquifer lies within the bedrock formation and is of
limited thickness  and  area! extent. A few  nearby
domestic wells tapped  this aquifer. Water samples
from those wells indicated  a higher mineral content
than was found in the water from the  sand aquifer.
                                      Fig. 4. Original direction of ground-water flow.
 160

-------
However,  the  degree of mineralization was within
United States  Public Health Service limits for pota-
ble drinking water purposes. The capability of the
consolidated aquifer to supply water was explored
and found to  be significantly less than that of the
unconsolidated aquifer.
     The deeper bedrock formations include shales,
limestones, dolomites  and  sandstones. Where satu-
rated, these rocks contain saline water. Local de-
posits  of petroleum   and  natural  gas  have been
found in some of these sedimentary rocks.  Beneath
the sequence  of sedimentary formations,  igneous
and/or  metamorphic  rocks  occur. The basement
rocks lie approximately one mile below the surface
of the land.
     From the standpoint of potable and industrial
water supply,  the  upper  120 feet of the earth's
mantle offer the only reasonable supply at  this site.
     The  deeper bedrock  formations are  of value
only  for  consideration as disposal reservoirs for
highly concentrated industrial  waste  waters.  The
relationship between  the  static water level, high
specific gravity of the  native brines within  the bed-
rock formations and the low specific gravity of the
waste waters to be  discharged make it theoretically
possible  for  disposed waste waters  to   migrate
upward  by means  of unrecorded or  improperly
plugged  wells  into  the fresh  water aquifers.  The
uncertainties related to natural fracturing  between
the various formations also cause a logical hesitancy
on the part of regulatory agencies  and consultants
to  recommend  deep well  disposal. For these rea-
sons, waste water disposal into the bedrock forma-
tions less than 3,000 feet  below the land surface is
not recommended  in  this area. However,  the  bed-
rock formations below 3,000 feet are suitable for
limited storage of waste waters.
               INVESTIGATIONS
     Studies of the industrial wastes involved de-
tailed, in-plant surveys of the chemical character of
the various waste streams  and their treatability. As
is often  the  case, the studies revealed that  more
thari 90% of  the  waste  water was cooling water
which was unimpaired except for  a slight increase
in  temperature.  It was judged that this water was
readily  amenable  to  return to  the  ground for
reuse.  Approximately  5%  of  the  waters  were
 impaired and  could be treated by available  methods
and discharged to surface waters. Approximately
 2'/2% of the waste waters were of human origin and
 could be treated with ease.  The remaining 2l/2% of
 the water was highly concentrated, highly  complex
 industrial waste water. Further investigation  indi-
 cated the latter could be treated by  incineration
 and/or deep well disposal.
     At the time of the study, the volume of the
most obnoxious waste water totaled 5 to 10 gallons
per minute. Projections based on the  anticipated
expansion indicate the volume will not exceed  50
gallons per minute.
     Concurrent with the in-plant waste water  in-
vestigations, treatability studies  of the contami-
nated ground  waters were conducted,  and new
potable and industrial water supplies  were  devel-
oped. The investigation  of  the extent of ground-
water pollution and development of new supplies
involved construction of a series of \1A  inch diame-
ter observation wells tapping the upper and lower
portions of the sand aquifer. These wells permitted
measurement  of water levels,  sampling of the
ground  waters, delineation of  the  area  of the
ground-water contamination and definition  of the
aquifer  hydraulics.  The  latter data, together with
the  separation  of the various waste streams, per-
mitted  construction of a new  ground disposal
recharge pond for the cooling waters and individual
treatment  of the contaminated  industrial  waste
waters.  The locations of purge  wells,  observation
wells,  production wells, cooling water return site
and waste water ponds are  shown in Figure  5. The
basic head  and ground-water flow conditions  are
also shown.
                              A INOUST
 Fig. 5. Basic head and ground-water flow conditions.
                                                                                                  161

-------
                                                                                           NO 5
                                                                  n
                                                                  I	1 U-J
                                                              RECHARGE  PONDS
                                                                    ~
. / if 	 ^> '
1
1
1
" 1
1

1 °*"
/




LEGEND
O OBSERVATION WELL
• DEEP
O SHALLOW
A INDUSTRIAL SUPPLY WELL
D POTABLE SUPPLY WELLS
/\ PURGE WELL

CHLORIDES IN PPM.
••• 1000 & ABOVE
1.' •• .'.-I 500 - 1000
1 l 100 - 500
1 1 0-100
Fig. 6. Range and concentration of dissolved chlorides.


162

-------
     Ground disposal  of concentrated  industrial
waste waters continued after purging of the con-
taminated ground waters was instituted. The origi-
nal production wells (1,2 and 3) were used in  the
initial  purging operations.  Purging confined and
removed the  contaminated  waste waters. These
waters were delivered to the suitable receiving body
of  water  until  on-site waste treatment facilities
were  built. Figure  6  shows the range and con-
centration of dissolved chlorides, a parameter found
to  be  proportional  to the total concentration of
contaminants. The  Figure reflects the areal extent
of  the  ground-water pollution and illustrates  the
combined effects of natural  dilution and pumping.
The greatest concentration occurs about the  waste
disposal ponds  while  natural  factors disperse  the
contaminants  away from the ponds.  Purge wells 2
and 3  provided an  effective  barrier between  the
contaminated water and the fresh ground waters
tapped by wells 4 and  5.
     Figure  7  illustrates  the  vertical  extent of
ground-water pollution and the effect of the cool-
ing water  recharge ponds   and purge wells.  The
flow of waste waters and their vertical distribution
were  governed by the natural recharge to the area
and the effect of pumping. The top portion of the
                                       aquifer  was the first zone to show improvement
                                       after cessation of disposal of wastes to the ground.

                                                    WELL STRUCTURES
                                            The  potable  supply, industrial supply  and
                                       purge wells were all  constructed by the cable tool
                                       method of drilling.  This  method was chosen be-
                                       cause efficient  well  structures  could be  accom-
                                       plished  with  ease and  economy. However,  the
                                       economies and well efficiencies were balanced with
                                       the immediate  needs for production capacity.  The
                                       basic physical characteristics of  the production
                                       wells are  summarized  in  Figure  8.
                                            Well  No.  1 supplied  the initial potable  and
                                       industrial  needs. The well is  tubular, 6 inches in
                                       diameter, with 15 feet  of screen. The screen was
                                       set in the lower portion of the  sand aquifer bottom-
                                       ing at approximately 60 feet. According to records
                                       the well screen is a wire wound, silicon red brass
                                       screen with 0.010 inch slot openings designed for
                                       50%  retention of  the formation. Based on subse-
                                       quent wells of similar construction, the  specific
                                       capacity of the well was estimated to be from 5 to
                                       8 gallons per minute per foot of drawdown.
                                             The construction of the  many wells through-
                                       out the site  revealed that the  aquifer  sands are
       z
       o
       1
110 .

100 .

 90.

 80.

 70.

 60

 50

 40.

 30

 20.
                              Waste
                               Ponds
Recharge
 Ponds	;
                                                                  /  < V
                                                                   / X
                                                                  ,
                                                                          PW 4	
                                                                             16-
^
V
\
\
\






I-
        LEGEND

        	WATER  LEVEL ELEVATIONS

        	  FLOW  LINES

        VERTICAL EXAGGERATION = 10

        PW = PRODUCTION  WELL
        P   = PUKGE  WELL
        OB = OBSERVATION  WELL
                               CHLORIDES  IN  PPM.

                                       1000 &  ABOVE

                                       500-1000

                                       100-500

                                       0-100
 Fig. 7. Schematic profile.
                                                                                                 163

-------
   MO I «Q I t
                         MO14147
Fig. 8. Well constructions.

uniformly fine. This characteristic led to standardi-
zation  on gravel pack type well construction. Wells
3 through 7, and replacement wells 2A and 3A are
all gravel packed well structures of the type shown
in Figure 8. The original wells 1 and 2 and  the
potable  supply wells are tubular  wells, also as
shown in Figure 8. The potable wells and industrial
wells 4,  5,  6 and  7 have met  and  surpassed  the
water  needs of the  company.  The  original yield
characteristics, as specific capacity, is summarized
in Table  1.
          Table 1. Original Specific Capacity

     Wells                     Specific Capacity
                             (gallons per minute per
                               foot of drawdown)
Industrial Well No.:
4
5
6
7
I
Potable Wells:
East
West

13.5
14.2
28.1
14.2

6
6
               WATER  QUALITY
     The original water quality of the shallow sand
aquifer was approximately 110 milligrams per liter
(parts  per  million) of hardness as calcium carbon-
ate,  5  milligrams  per liter of chloride, and 0.25
milligrams  per liter of iron. These native ground
waters are  not as highly  mineralized as ground wa-
ters  elsewhere in  Michigan.  Even so,  with time
some deterioration in the production characteristics
of the  well by chemical precipitation of carbonate
deposits  occurs and occasional redevelopment is
required.
     The complex  chemical nature of the waste
waters originally disposed to the ground has not
only adversely affected the quality of the ground
waters within  the area but has created maintenance
problems with  the  purge  wells. It is necessary  to
redevelop the wells as often as once a month. The
wastes are  more viscous than water, move slowly
through the fine sand aquifer and continually plug
the aquifer,  gravel  packs,  well  screen  openings,
pumps and pipeline facilities associated with the
purge well  system.  The chemical characteristics  of
the waste waters disposed to the ground over the
years in which this disposal method was used varied
from moment to moment, day to day, and month
month, according to the  types of products made
and the market demand.
    The industrial waste waters vary in pH from
extremely acid  (pH 2±)to highly alkaline (pH 12±).
Some of these waters are organic compounds. BOD
and COD  characteristics  are also present  and ex-
tremely variable.
    Part of the fouling of the purge wells appears
to be related to the interaction of  alkaline wastes
with the  bicarbonate  component  of the natural
ground waters. Such mixing leads to supersatura-
tion with calcium carbonate. Precipitation of the
latter  within  the turbulent flow regions of the well
is  natural.  Organic residues arc  another  cause  of
purge well  fouling. Soil bacteria are suspected but
unproven fouling agents.
    Despite  the complex properties of the waste
waters disposed to  the ground over a period of ten
years, persistence and the  application of hydrologic
principles  has  assured recovery of  the  aquifer.
Figure 9 summarizes this recovery. The  contami-
nated ground waters have been steadily removed  by
pumping and diluted by  natural and artificial  re-
charge.  Extraction  of ground waters at a rate
slightly in  excess of the natural recharge  confined
the contamination and permitted flushing of the
wastes through the  soils and into the purge wells.
As anticipated, the upper  portion  of the  aquifer
was improved  first. The improvement of the zone
is shown by well 7a. Natural recharge, coupled with
the effect of pumping, drove the wastes to  the low-
er portion of the aquifer as shown by well 19.

           WELL REDEVELOPMENT
    The fouling of well structures, particularly the
screen openings  and  the gravel packs,  result  in
marked reduction in the production capabilities of
the wells.  The loss of capacity and need for fre-
quent  redevelopment  is  particularly  marked  in
connection with purge well systems. In that system,
redevelopment is  accomplished by  various means
such as air lift surge, pumping surge and jetting,
assisted  by  the addition  of  chemicals.  Various
chemicals used include acids, alkalines, chlorine and
phosphates. The more tightly plugged the screen
and gravel  pack openings, the greater the necessity
164

-------
        4MO

        mm
      O  !soo
      X
      u  •
         10W

         SMj

i        r
i        i   i
i        i    T
                                                  <500
                                                E 35C5
1850


 MO

  0
       IT
I  1  II
I  I  II
|
                                                                          T
                  WELL   NO.  7q.  J[SHALLOW)                    WEIL  NQ.J9 1PMPJ

                          CHLORIDES IN  SELECTED  OBSERVATION  WELLS
Fig, 9. Chlorides in selected observation wells.
to use several  applications of  chemicals  and the
most vigorous hydraulic surging  procedures.
     The purge wells tapping  the ground waters of
the highest concentration of pollution, require the
most frequent redevelopment. Even so, gradual loss
of capacity to yields of one-fourth to one-half the
original specific capacity occurs.  The less concen-
trated  the waste  handled  by  the  well,  the less
frequent the need for redevelopment and the more
successful the redevelopment. The potable supply
wells and the industrial supply  wells 4,  5, 6 and  7,
al! of which  produce waters nearly the same as the
natural ground  waters, seldom require redevelop-
ment.  When redevelopment  of these  wells  is  re-
quired, it is rapidly  accomplished by acidizing and
surging with the  pump,  followed with thorough
disinfection bv chlorination.
                 CONCLUSION
     The  preceding  has  summarized  the  gross
aspects involved in the continued use of a vital
natural ground-water resource while simultaneously
correcting a misuse of that resource. The multi-
plicity  of interdisciplinary responsibilities was in-
                             ferred more than described, but constituted a major
                             contribution in accomplishing the end result.
                                  The scientific principles,  technical methods
                             and skills associated with the location and develop-
                             ment of a ground-water  supply are equally appli-
                             cable to the  solution of ground-water contamina-
                             tion  problems. However, the role of the geochemist
                             and  the sanitary  engineer  are more  vital in the
                             solution of  ground-water pollution  problems  if
                             successful recovery of the damaged portion of the
                             environment  is to be  realized.  For instance, the
                             services of the geochemist are needed to assess the
                             degree of hazard involved and the effect  of con-
                             tamination on the geologic  environment.  The ser-
                             vices of the  sanitary  engineer are necessary for
                             evaluation of alternate methods of treatment and
                             the  associated economics.  Depending upon the
                             characteristics of  the geohydrologic  environment
                             and  the nature of the contaminants, the services of
                             other specialists may be necessary  for a proper and
                             economical  solution. It is essential that all parties
                             involved recognize that if man is capable of utiliz-
                             ing the many technical aspects of his environment,
                             he  is  also capable  of  meeting the  challenge of
                             correcting the misuse of the environment.
                                                                         165

-------
                                           DISCUSSION
The following questions were answered by Edward
M. Burt after delivering his talk entitled "The Use,
Abuse and Recovery of a Glacial Aquifer."

Q. What is  the  temperature of the water in the
recharge  ponds and what  effect did these ponds
have on  the pumped water temperature from the
two wells immediately east  of those ponds?
A. The temperature  was checked and the  effect
was thoroughly investigated. I am sure all of you
are aware of the problems inherent in the operation
of any  type of a  facility, whether it's water or
waste water  treatment.  It  seems  that we human
beings take the path of least resistance. If a gallon
of water will keep a reactor cooled  to the proper
temperature, then the maximum hydraulic flow will
do even better. The average flow of cooling waters
vary from 0.5 to 1.0 mgd. The temperature of 90%
of this flow does not change more than 2 or 3 de-
grees F. By  the  time the  recharge  water reaches
Wells 4  and 5 and is blended with fresh ground
water, there is no measurable difference in temper-
ature between those  two wells and  samples taken
from Wells 6 and 7.

Q. You  mentioned  that  in order to receive  a
permit from the Water Resources Commission, the
industry must  comply  with certain regulations?
What are these?
A. Michigan does  not have a rigid  set of regula-
tions, but rather establishes specific criteria tailored
to each  individual  site. These standards are set in
this manner. Any industry that wishes to make use
of either the surface or ground-water resources in
the State must apply for a  permit. This application
must indicate the nature of the business  and the
waste they wish to discharge. The staff of the Water
Resources  Commission  confers with other  State
 igencies,  such  as  the  Michigan  Department  of
Public Health and  the Geological  Survey, and to-
gether they develop  a set of standards based on
existing  uses of the  particular water  resource in
question. The most common limits imposed pertain
to pH, BOD, COD, solids and specific elements.

Q. Since the pollution showed up after a few years
of operation,  were the regulations  given serious
consideration by the industry prior to the building
of the plant?
A. This question precedes our involvement in the
project, but our experience with the firm indicates
that they have always given serious consideration
to pollution problems, both  real and potential.
From  what  we  have ascertained, the  problem
evolved because a detailed study of the ground-
water hydrology was not originally required, there-
fore was not originally conducted. Another impor-
tant consideration is the fact that between the time
the plant was built and the time  the problem was
discovered,  the  types  of products and  processes
employed changed. The regulations under which a
discharge  permit is  issued are  not like the law of
the Medes and Persians—they are just the  opposite;
they are in a constant state of continuance. This is
reasonable because, in establishing regulations, it is
not  possible to stop  time. Each time  a manu-
facturer considers the addition of a new product
that introduces a new waste, it is necessary for the
firm to apply for an amended order and the review
process is repeated.  This is the  manner  in which
the presence of a problem was discovered.


Q. What was the character of the material clogging
the purge wells? Was it comprised of  corrosion
products,  organic growth, etc.?
A. That's the $64,000 one! Let  me describe  what
happened. When the redevelopment  of  the  wells
was  attempted,  the  greatest problem occurred in
the  pump impellers.  As in  all situations of this
type, the first step was to remove the pump. The
contractor's report indicated  that after the bowls
were disassembled, cleaning could be accomplished
most efficiently by  allowing  the material to dry
and then to remove it with chisels. Soaking in acid
or alkalis was partially effective, but it made  quite
a  gooey  mess.  There  have  not been  complete
analyses made of the composition of the material.
The  reason a complete analysis has not been made
is  that over the years of plant operation, varying
quantities of many different  types of waste have
been produced. As a  result, during a given time
interval today the waste extracted will reflect those
products of several years ago, but next month the
waste  will  reflect  a different  combination  of
products  and the composition  of the  extracted
material will be entirely different.
166

-------
           Bull  Session 4—Aquifer Protection and  Rehabilitation
Session  Chairman: Leslie G. McMillion, Geologist,
  National Ground Water  Research Program, En-
  vironmental Protection Agency,  Robert S. Kerr
  Research  Center,  P.  O.  Box 1198, Ada, Okla-
  homa 74820.

Leslie G. McMillion, Lead Bull:
     Our  session is  "Aquifer Protection and Re-
habilitation." All of the speakers of this session are
present; they are Ed Burt, Ken Schmidt, and James
Waltz.  Also, Tom Ahrens  is  sitting in to  answer
questions that  might  be asked on the topic on
which  Joe Mogg was  to  have given a paper this
afternoon; as you  are  perhaps aware, Mr. Mogg
could not attend the Symposium and consequently
he did not make his presentation.
     This afternoon we received  more questions
concerning James Waltz's talk than James was able
to answer during the  discussion period that fol-
lowed  his talk.  I have  these unanswered questions
with me, and there are about a dozen of them. I
hope we will have time to enter  these questions
into the record of this bull session.
     I  appreciate the excellent attendance tonight.
The subjects are excellent, and I surely anticipate
lots of spirited  and interested discussions. We are
ready  for your questions  and discussion, and we
certainly invite your participation.

Steven  IM. Goldstein, Mitre Corporation, McLean,
Virginia:
     Mr. Schmidt, we have a great  deal of interest
in the  technological opportunities that individual
home  sewage disposal  systems might  offer. With
respect to coming up with an optimal national plan
for  sewage  treatment, there are  quite obviously
areas  where  septic tanks  and other individual
systems are very competitive and  a better  choice
than centralized  sewage  collection systems. Of
course, quite the opposite holds  true in densely
populated localities. It's most important then that
proper  septic  tank and  other individual  system
practices be reintroduced as opposed to the current
situation where many builders just dig a hole and
throw them in the ground and the things fail in 5
years,  15 years, or something like that. So, we've
been studying this problem for the Office of Water
Programs of EPA and one of my reasons for coming
to this Symposium is to get a better feeling for how
the  people  who are actually doing field work in
water  resources regard these systems. Could you
comment on what you regard  to  be  some of the
future problems in this area of individual home dis-
posal systems on the basis of the work you've done
in the Fresno area?
Kenneth  D. Schmidt, Harshbarger and Associates,
Tucson, Arizona:
     Organic nitrogen  and  ammonia  nitrogen,
which  occur in both sewage effluent and septic
tank effluent, can potentially form nitrate. In the
case of septic tanks the conditions  are generally
more favorable for conversion to the nitrate form.
This is due to the fact that the waste is disposed of
over a large area and anaerobic conditions don't
tend to develop as readily. Under anaerobic condi-
tions, the nitrate may never be  formed from the
other nitrogen that forms. Also, denitrification may
be lost as gaseous forms. More control of the nitrate
situation in ground water should be  possible with
centralized  sewage treatment facilities.  However,
each site must  be  carefully studied.  The waste
disposal  operation, the soil, and the  ground-water
conditions are key factors.
     There are areas near Fresno today where plans
exist for new septic tank developments in the foot-
hills. It is  obvious  from previous studies  that the
foothill area has  generally  poor aquifer character-
istics.  The  aquifer is thin, the  transmissibility is
low, and locally major faults  separate the  develop-
ments  from sources of recharge. Thus, these areas
may  be unfavorable for waste disposal of this
type.  However, in favorable areas, properly con-
structed  septic tanks and disposal systems will be
practical and not harm the ground  water. In the
Figarden-Bullard  Area, septic tanks have been used
for 20 years. Yet, only 2 wells have ever had nitrate
contents over 25  ppm. Thus, in a favorable environ-
ment,  if the wells, septic tanks, and septic  tank dis-
posal  systems  are  constructed  properly, a nitrate
problem may not occur. Similarly, biological con-
tamination may also not be present.

 Hank  Baski, Wright Water  Engineering, Denver,
Colorado:
     We have  a pending water quality study  of
ground water in an area where  nitrates have been
measured up to 50 or 60 ppm. Upstream  from the
problem site there's a municipal sewage treatment
 plant  with secondary treatment, and some people
 say that the sewage plant is going to  contribute
 nitrates  which  may enter  the aquifer. Do the
 secondary  treatment sewage plants actually take
 out nitrate?

 Kenneth D. Schmidt:
     As  far as the amount of nitrogen that enters
 and leaves a conventional plant you  have to re-
 member it's not all in  the form of nitrate. We're
 usually talking about the  sum of organic  nitrogen,
 ammonia,  and  nitrate. Most conventional plants
                                            167

-------
convert some nitrogen to the nitrate form and also
remove some organic nitrogen. Thus, total nitrogen
is usually significantly lower in the effluent than
influent. I seriously doubt that besides settling out
of solids there is any drastic difference between the
influent and effluent. I think the real opportunity
to control nitrogen  occurs  after the effluent  is
discharged  to ponds.  Recent studies have shown
that by controlling factors such as pH and tempera-
ture, loss of ammonia  from ponds can be enhanced.
Nitrate contents of 50 or 60 ppm would be ex-
pected near sewage treatment plants. Most sewage
effluent contains from 25 to 30 ppm total nitrogen.
If all of this nitrogen was converted to nitrate,
concentrations exceeding 100 ppm could occur in
ground water near sewage plants.
Edward M. Burt, Williams & Works, Grand Rapids,
Michigan:
     This nitrogen  discussion is interesting. We're
doing  many  waste water  treatment  projects in
Michigan that involve oxidation lagoons and spray
irrigation.  Concerning  oxidation ponds or lagoon
systems, we have built over the years a number of
ponds where  the  first cell was anaerobic  on the
bottom and aerobic on top and then the following
cells  were totally  aerobic. The  initial pond  was
designed and built so that the basal portion of the
pond operates under  anaerobic digestion and the
surface operates under aerobic digestion which  is
typical of oxidation  ponds.  The balance  of the
ponds following this are oxidation ponds of aerobic
digestion.  Both types  of digestion  accomplish
things that are  desirable and that's why we have
used them as raw sewage or initial ponds,  though
this is no longer allowed by the Michigan Depart-
ment of Health. We designed some of these and got
very high degree treatment. There's a great deal of
discussion and evaluation in Michigan by the Health
Department;  and  some people,  including some
sanitary engineers  argue  that  oxidation  lagoon
systems  without the  anaerobic facility or with  it
accomplish what is commonly known as secondary
degree  sewage  treatment.  This  process then  is
comparable to a mechanical activated sludge plant.
Along the idea of nitrogen removal, this past winter
we obtained permission from the Michigan Depart-
ment of Health to experiment  and  have winter
drawoffs. The normal procedure in Michigan is that
you store and discharge in the fall and in the spring.
This past winter the  Health  Department gave us
permission to use continuous discharge from about
9 ponds. And while we had ice cover we drew off
from these at a  regular rate while  not breaking up
the  ice  cover,  and  we had significantly  higher
removal of the nitrogen than has been accomplished
by ponds or secondary degree treatment. We don't
have any conclusions but the suspicion is that the
oxidation ponds rely upon algae which feed on the
nutrients in the waste water and in the winter the
algae  are  not as active as in the growing season;
therefore  they die and go to the bottom as sed-
iment  with the  result that under  the ice we pull
off very high quality waste waters.

M.  G. Croft, U.  S. Geological Survey, Bismarck,
North Dakota:
    I'd like to direct this question to Ken Schmidt.
In your discussion this afternoon you mentioned a
background nitrate count of several ppm in the
ground  water. I  was wondering if you'd care to
comment on the origin of this nitrogen.

Kenneth D. Schmidt:
    As you probably know, nitrate has been ana-
lyzed  from samples of the Kings River and the San
Joaquin River, and  it often ranges from 1  to 3
ppm.   Evaporation would  slightly  increase  this
content before percolation of surface waters to the
water table. However, I do not envision the 6 or 7
ppm as all natural, but rather  background to the
urban  area.  Flow  net  considerations and  travel
times  in ground water suggest  that much of the
urban  ground water was  actually  beneath  agri-
cultural lands several decades ago. Thus, some of
the background  content could have been supplied
by  agriculture.  Agricultural lands  surround  the
urban areas,  but were probably not heavily ferti-
lized prior to 1940.
    Another  natural source besides runoff would
be related  to  soils near the  foothills, northeast of
Fresno. Stout, Burau, and Allardice point out that
areas  with a  Mediterranean climate are  favorable
for natural nitrate accumulation. Strahorn.who did
the early soils work of the Fresno area, noted that
high grasses grew on the alluvial plains. In  areas of a
Mediterranean climate,  precipitation occurs in the
winter,  and natural plants grow and take nitrogen
out of the atmosphere.  In the hot summer these
plants  die and  most of the nitrogen remains as
residue in the soil. This nitrogen is then converted
to nitrate by bacteria. With the coming of the next
winter, abundant rainfall leaches  the nitrate past
the root zone and into the ground water.
     The  Agricultural Research Service  has done
studies  on a small  basis in eastern Fresno County
and has found natural nitrates. Near the east edge
of the San Joaquin Valley  rainfall may  approach
16  inches  per year and vegetation  may  be more
plentiful. Natural nitrates have been found beneath
dry farmed areas of eastern Fresno County and are
apparently related primarily to  the soil. The soils
themselves have  some nitrogen  even though they
are very  low in organic  matter. They contain
168

-------
something  like  1,000  ppm of  nitrogen  on the
average. However, there is no guarantee that any of
this  nitrogen  can  be  converted to nitrate and
leached to  the  ground water.  It is obviously a
source worthy of consideration.

Robert E. Pendergast, Geotechnical  Engineering
Corporation, St. Paul, Minnesota:
     My question is with regard  to installing ob-
servation wells,  sampling  wells and techniques for
sampling. I'm aware of various methods that can be
used, but  I'd like to take  advantage of your expert
knowledge on  the  state  of the  art. What is the
most efficient method  for  installing the well and
obtaining  the  samples for reliability and accuracy
on a normal  industrial  job, rather than on a long-
term research study?

Edward M. Burt:
     Let's talk about observation wells—monitoring
wells in broad generalities. I will make reference to
the industrial investigation that my firm had going
for 15 years, and the  work we've been doing to
correct the  situation  that I discussed in my  paper
this afternoon. Since the  aquifer of concern  in my
paper  consisted  of  unconsolidated sand with rela-
tively  good homogeneity throughout  the water-
bearing zone,  we were able to use the auger ma-
chine for installing  1V4  inch diameter casing in our
observation wells. The auger that we used  was  a
continuous worm auger that stays open with rota-
tion.  We did not find  it  necessary to use mud to
get good results. On the bottom end of the  1'4 inch
casing we used  standard  household well points of
24-30 inches long. In some instances we went to 5-
foot well  points. I mentioned that we preferred to
sample the upper 15 feet of the formation and the
basal  15  feet; and this sampling approach  was  a
judgement decision based on the geohydrologic con-
ditions of the site and the soil conditions.
     In connection with the construction of waste
water  lagoons, oxidation ponds,  and spray  irriga-
tion, we again tailored  the method of investigating
for establishing  monitor wells because there  has to
be an  initial investigation to establish that the soil
conditions and the geologic conditions are proper,
and  there  has  to  be a continuing monitoring
program established, especially in connection with
irrigation  projects. We normally go in with the
auger machine first. However, we utilize the readily
available well records;  for several years the water
well drillers in  Michigan  have  been licensed and
they are required to file logs; and, in addition to
this,  the  filing  of logs on  oil  and gas wells are
required.  We take this information and go on the
site  and make preliminary estimates of the direc-
tion  of ground-water  flow  and what  the basic
hydrogeologic conditions are.  We  take  an auger
machine and check out our original ideas, and then
we  make a predesign of  how we are going to ana-
lyze these  conditions and how we are  going to
finalize the monitoring well system.  In more com-
plicated geologic situations we go to either cable
tool drilled  wells  with  casing  or  rotary drilled
wells with mud  and we  use electric logs. In using
cable tool drilling, we normally collect our geologic
samples at  5-foot intervals and at each change in
formation.  The  final well for monitoring is really
designed to  the geologic  conditions and the use of
the site. We  establish the  initial ground-\\aujr gradi-
ent  and we have to set  up sufficient  observation
wells to permit  continuous monitoring of this by
the people who operate the facility.

George Taylor,  U.  S. Geological Survey, Washing-
ton, D. C.:
     At the  present time in West Pakistan, there is
quite a bit of controversy going on over the fall-off
in yields of wells that have been put down during
the last 8 years, using fiber glass slotted pipe. They
have a standard  well design. I'm sure Mr. Ahrens is
familiar with this  area.  I  understand  he's  a well
design  expert, going by reputation.  He  doesn't
know  me but I know him. I wonder  if he could
comment on the relative importance of the chemi-
cal precipitation of carbonates in the filter pack
around these fiber  glass  slotted casings as against
the  physical clogging of the slots  through incom-
plete development.

Thomas P. Ahrens, U. S. Bureau of Reclamation,
Denver, Colorado:
     I'm talking pretty  much from scuttlebutt.  I
have tried  to get  some  definite information  and
have talked to several people from Pakistan when
they were  in the  office over  the  last few years.
None of them know anything about it. I was fortu-
nate enough to have a few old friends who were
over there  working for contractors; and from the
letters I have received from them, I have been read-
ing  between  the lines.  When  Harold Smith  was
there the first time, they put down  all these wells
and tested them, and they turned  out  to be 2 or  3
second-foot wells, but it was 9 months before they
got  the pumps to put in. When they put the pumps
in they turned them on, and they sucked air. They
tried shooting them with  primacord, thar would
bring them  back for about  2 months and then they
would go out again. So they pulled  some of them
and found a  combination of sulfate reducing bac-
teria and encrustation; thus,  what you had was the
corrosion products mixing with the encrustation
products to give a complete blocking of the well in
a couple of months. So then they started out on
                                             169

-------
this fiber glass, reinforced  epoxy. The prime con-
tractor on  the job called me and wanted to know
what  I thought about it, and I said we had been
looking at it for years and were afraid of it. He said
they were going to use it over there and asked me
to recommend a minimum wall thickness. I said  I
wouldn't go less than 1A inch. So they went in with
.180  inch for the casing and .200 for the slotted
pipe.  Well, from  the stories I've received from the
contractors over there,  they  can't develop those
wells  completely. If they  go in with air or with a
surge block,  they collapse. So  what  they do  is
merely go  down  to near the top  of the screen and
overpump  them.  They are  running anywhere from
60 to 90 feet of perforated casing in those  things
and  all  of you  probably  know that  when  you
develop  a  well  by  overpumping all  you  do  is
develop possibly the upper quarter or maybe third
of the perforated section. The rest of it—nothing
happens to it.  I  have  always suspected  that the
change to the glass  reinforced epoxy  would cut
down on the  corrosion; in fact, it would eliminate
the  corrosion.  But  I  always  thought  that the
encrustation would still occur but that it would be
extended over a long period of time. Now you've
provided the first concrete evidence I've heard that
this is probably happening.

George Taylor:
      I think there is still some controversy over this
 point; as a matter of fact I'm in the minority. All I
 know is  also hearsay. I've been in touch with  a
number of people, and we have a man still out
 there. I think everybody agrees that the fiber glass
 is essentially inert. There are no corrosion  products
 forming directly in the slotted pipe or in  chemical
 combination  with the material.  But, these wells
have  fallen off in yield. I understand that television
 cameras have been lowered  down some of these
 wells and they have been able to look at the  insides
 of these wells. They're finding that there is a little
 bit of slime on  the inside  which may be a result of
 bacterial action.  I have the feeling that a lot of the
 precipitation  of  carbonates may be still occurring
 in the filter pack outside the casing contributing to
 the fall off in the yields. There is some evidence in
 some of the wells that there is quite an accumula-
 tion  of material  in the bottom of the well that has
 come through the slots, fine material.

 Thomas P. Ahrens:
      Just sand? Is it actual material or what?

 George Taylor:
      Well, apparently, yes, sand has come through.
 Through,  as  you point out, incomplete develop-
 ment, you see. So it may be a combination of these
things  but most  people  I've talked to don't give
much credence to this idea that  there is  actually
carbonate precipitation taking place  in the filter
pack. I understand that much of the filter pack
material they use is quite high in carbonate materi-
al, native carbonate material.

Thomas P. Ahrens:
     If your screen was not developed for its entire
length, you had your  concentrated flow at the top
where it was developed and you probably run into
inflow velocities  there,  possibly a  few  feet  per
second. So you'd have an ideal  situation for the
deposit of carbonates if they were in there. Now
this  other slime you  mentioned. Do  you know if
they are pulling any of those wells down below the
top of the screen?

George Taylor:
     I  really don't know. I understand  they  do
have high  drawdowns  in some of the wells.

Thomas P. Ahrens:
     When the water levels are pulled down  and
you get Crenothrix, Leptothrix, Gallionella, and all
of those,  they'll get down  in there and every time
you pull that down a  little bit they'll start  growing.
And, as they grow down, they block it; the water
table or pumping levels keep going down, so they
just keep  moving on  down the screen. That might
account for some of that slime you mentioned.

George Taylor:
     I  understand they only lowered this thing in
one or two wells but they did find some kind  of
growth that  seemed  to  be a precipitate  or slime
inside the casing.

Thomas P. Ahrens:
     Incidentally, do  you know if they  can  pull
that fiber glass  pipe or will the joints pull apart
when they try?

George Taylor:
     Well, that's another problem.  They did,  I
understand, get one out more or less intact, but it
was clean. I mean they found no evidence of slime
or precipitates.

Thomas P. Ahrens:
     Well, it's so slick that I think  if you did have
any carbonate deposits on there, just pulling it out
would probably strip  it off.

Jerry  Frick, Walkerville Well Drilling Co., Walker-
ville, Michigan:
      I'd like to ask Mr. Ahrens if he has any infor-
 mation or experience on the relationship between
 intake velocity and buildup of encrustation?
 170

-------
Thomas P. Ahrens:
     Let us say "no observation." We have some
beautiful examples at the office that I could show
you  of high intake velocities and  very high corro-
sion; and also  some similar information on encrus-
tation  but I have just observed these latter myself.
Now we just use a 10th of a foot per second intake
velocity. We have never had any bad' problems with
that criteria except when we were in a terrifically
corrosive aggressive water; but the problems would
have occurred, I think, whether  we  would have
been pumping or not.

Edward M. Burt:
     I've played  with wells too,  and  so  I'll share
some top of the  hat recollections. I'm not sure of
the  details of the construction of these wells, but I
have been involved in experiences  with bacterial
plugging where the best result on redevelopment
was not accomplished  by chemical  treatment but
by heavy chlorination.  Another slimy sort of thing
that we can run into is, for instance, in a double
screened well, with the suction intake set down in
the  screen or  between the two screens, when  you
pull the pumping water level down below  the upper
screen you cause a cascading, you have a beautiful
in-the-well  iron precipitation plant. And this  will
foul them very rapidly, quite completely. As a
 practical guide, I haven't used it too much, but by
 my experience I'm convinced that you get plugging
 beyond the screen face, and perhaps beyond the
 gravel pack. A lot has to do with the geology of the
 formation. When we talk about gravel packs it's a
 misnomer to  say gravel because it is really filter
 sand in wells  that are developed in a silty, fine sand
 formation. The City of Kalamazoo  recently deter-
 mined that the distance out to where the plugging
 extended  in their wells was between 7 and 9 feet.
 These determinations were made by running a step
 test in reverse (that is making an injection test) and
 calculating the volumes  and  decrease of injection
 rates.

 Thomas P. Ahrens:
      You  mentioned a heavy chlorination. There's
 been  quite a bit of research done on that and one
 item  in particular, that being sulfate reducing bac-
 teria. It takes at least 400 ppm and about an 8-hour
 contact to kill sulfate  reducing bacteria. We started
 out using the HTH  calcium  hypochlorite but our
 labs called us on it because when you run over 300
 ppm   on  HTH  you  start precipitating calcium
 carbonate. So we have switched over now toClorox
 and 2 gallons of Clorox to 100 gallons of water in a
 well will give you about 1,000 ppm and we figure
 we're getting a  certain amount of dilution so it's
 probably  coming down to around 400 ppm.
George Taylor:
     How long would it be do you think before
this bacteria would again build up to the concentra-
tion where they would cause a plugging problem?

Thomas P. Ahrens:
     We have a terrific problem with that down in
Wellton Mohawk in Yuma. We started out steriliz-
ing those wells and we found out right after steri-
lization we got a negative reading on our sampling
for anywhere from 2 to 6 weeks, and then it would
start  coming back. I think that there are just so
many little crevices and cracks and places for a bug
to hide in a well  that you can  get  99 percent of
them but that 1 percent  that are left are right back
building families as soon  as the stuff is gone.

 Leslie G. McMillion:
     Tom, what is your  procedure on taking sam-
 ples in checking for bacterial growth?

 Thomas P. Ahrens:
      That I will have to claim ignorance on, Les.
 We turned that over to the  microbiologists and
 they would go out and pump the well.  1  do know
 they had some sort of a filter that they passed the
 water through, and collected so much  water or
 passed  so much  water  through  it and then, as I
 understood  it,  the colony actually grew on the
 filter. I can get additional information  for you if
 you'd like it.

 Leslie G. McMillion:
      Yes, thank you. I  will contact you later for
 the. information,

 Steven N. Goldstein:
      Maybe I can clarify that procedure somewhat
 for you. I'm  no expert  in the area but I have gone
 through some of  those  tests.  The filter in question
 is probably something like a membrane  filter with
 pore size small enough  to trap  bacteria. What you
 do is take a measured quantity of liquid, say 100
 milliliters, draw it through on the vacuum, the bac-
 teria then will not pass through, they will be caught
 on the  filter paper and  then you transfer the filter
 paper to the top  of some kind  of a nutrient medi-
 um  like an agar  gel. The nutrients can diffuse up
 through the filter  and the colonies can begin to
 grow.  The idea  is that each viable bacterium or
 other microorganism that can live in that medium
 will begin to divide and form colonies and after a
 suitable amount  of time you can actually see the
  colonies as little spots on the filter paper. Now
 what you're getting from that kind of a reading is
  called a total  viable count.  In other words, how
  many bacteria or  other  microorganisms did you
 have in that initial sample that were viable enough
                                              171

-------
to form a colony. But at that particular point you
have almost no idea as to their identity, except for
the fact that you may know something about the
medium which you are using to feed them and you
know that various media  are selective for various
species. But you only have a rough idea of what
they might be.

Alfred H.  Harder, U.S.G.S., Boise, Idaho:
     Are these bacteria indigenous to the aquifer
or are they exotic?

Thomas P. Ahrens:
     I  was just going to ask Mr. Goldstein if he
could answer  a question on that. One of the bac-
teria we're having trouble with is Desulfovibio, a
sulfate  reducing organism   which,  by  the way,
causes tremendous problems in pipelines too. But
from what I've been able  to read  concerning the
pipeline problem, the Desulfovibio generally don't
occur  below a depth of  10 to 12 feet. But here
we're getting them down in wells that are 125, 150
and 200 feet.  I've had more darned arguments with
various people, and they'll swear up and down they
are in  the aquifer. But if you check the size of the
things, they are bigger than any of the pathogenic
bacteria which are filtered  out in the aquifer in a
relatively short distance. How could these  larger
ones travel through  and affect the whole aquifer?
At Wellton Mohawk we have two wells put down
by the same contractor, % mile apart. One of them
has been  in for 14  years and has never given any
trouble at all, the other went out in 18 months.
Now if it's in the aquifer, they both should have
gone out.

Steven N. Goldstein:
     I can't give you a direct answer but maybe I
can bring to light a few things which might add to a
total picture.  First of all, as far as travel of micro-
organisms in  aquifers are  concerned  I have  no
first hand experience. I've  done some  reading on
the subject lately. This was mainly  investigations
done on  the  Santee project in California, and  if I
can recall the details correctly there  were some
movements on the order of 200 feet, even 400 and
600 feet  where the  bacteria were picked  up in
sampling  wells. I believe the situation here was one
of a saturated aquifer—thus, travel through saturat-
ed zones. I'm also led to believe, however, that if
you're not going through  a saturated  zone that
most of  the things  are usually filtered  out in the
first 7 to  10 feet. As far as the size of the bacteria, 1
don't think that they vary that much in size. I may
 be wrong,  but I don't think  the  size  variation  is
 very great. The other thing about bacteria is that
 they are  all over the place. They're all over your
skin  and  all  throughout  your  body; and you
couldn't live without some of  them. They are on
the floor of the ocean: most of them are concen-
trated  in the  first  millimeter of sediment  there.
They are on the floor of the bays, and they are in
the air. Bacteria have even been found in the upper
atmosphere  where  it was  thought  there were no
living forms. So you just can't get away from them
and  about all you  can hope to do  is reduce their
concentration in a clean  room so that the chances
of one of  these  things falling  and  contaminating
some research work you're doing is kind of nil.  I
imagine that they'd be  fairly  ubiquitous in soils
too. It only takes one of these  things to get some-
thing going, one or a very  few. And you never kill
them all off.  The  dose  of antibiotic that is now
necessary to kill off the gonococcus bacterium that
causes  gonorrhea, has  gone up  quite significantly
because people were getting a  dose of antibiotic
that would  kill off 90 to 99 percent of the popula-
tion, and  those which weren't killed off were the
highly  resistant ones. Then, of  course, these things
build up in  the population and  it will take a little
more,  and a little more and then a little more anti-
biotic  each  time, and so this  can  now become  a
significant public health hazard.
     Another thing which  is very interesting is the
work on clogging that was conducted by Thomas,
Schwartz, and Bendixen at the Robert A. Taft Sani-
tary Engineering Center in the middle 60's. They
were studying the clogging of soil by secondary
sewage effluent with lysimeters. They reported that
most of the soil clogging occurred in the first centi-
meter  at  the soil-water interface and that severe
clogging was associated  with anaerobic conditions
in the zone of clogging.  The black  color of a slime
which developed at the soil-water interface was due
to  the presence  of ferrous sulfide  but the ferrous
sulfide did  not contribute measurably to the clog-
ging mechanism. What happens is  that ferric ions
are  reduced to ferrous ions by anaerobic bacteria.
In  a similar manner, sulfate is reduced  to  sulfide
and the reduced forms combine to precipitate out
as ferrous sulfide. Addition of acid or restoration of
aerobic  conditions rapidly  eliminate the ferrous
sulfide but the layer remains clogged to the same
degree. It  takes  a longer time of drying  under
 aerobic conditions to  destroy the surface mat and
 to eliminate much of the clogging agents. In sum-
 mary, the  clogging mechanism in applying sewage
 to the soil through seepage beds, trenches, etc., is
 probably due  to development of anaerobic condi-
 tions  right  at  the  soil-water interface.  It has been
 reported by Mitchell and  Nevo that accumulations
 of bacterial polyuronides  or polysaccharides in the
 slime  layer may be a major cause of the clogging. I
 172

-------
really don't have any idea what happens in the deep
wells, but  it may be some of these surface phe-
nomena that occur in seepage beds, trenches, etc.

Hank Baski:
     I'll comment some more about bacteria. Here
north of Denver we've worked on  Fox Hills wells
that range  in depth anywhere  from 700 to  1200
feet. We chlorinate  these wells since chlorination
is  required by  the  Health Department. We try to
have it  in the  well itself in  order  to prevent any
bacteria from becoming  excessive.  It sounds as if
we've got bacteria all over and  during normal con-
struction of a well. It would be impossible to pre-
vent bacterial contamination. We found that chlo-
rination  in  a  well seems to keep  a well in good
shape.  One possible solution in an area which has a
problem is you just chlorinate whenever you pump
so there is a residual chlorine.  However, we've got
one water Association which is against well chlo-
rination. It's an odd situation, too, because they
have a well which has a problem; sometimes it will
pump  out a slime which may be  5 or 10 percent
of the water volume. I've never looked at the stuff
pumped out, but they say it  looks like fish eggs
when  coming  out  of the well. Pipes were coated
with it when they pulled the pump. It seems that
preventative maintenance in the nature  of chlorina-
tion of the well would  be  a possible  solution to
these problems.

James P. Waltz:
     What is the concentration  of chlorine that you
use in this routine treatment of the Deep Fox Hills
wells?

 Hank Baski:
     It would be in the order of 1 ppm.

 James P. Waltz:
     What  is  your  procedure  for  introducing the
 chlorine into a well?
 Hank Baski:
     Whenever the pump operates, it's designed to
 operate so the pump starts,  the  solenoid switch
 opens, you've got the pressure differential of the
 water pressure at top, just run it through the chlo-
 rinator and have a plastic pipe, 1A  or % inch which
 goes down. We've  found it's best to have it go
 below  the pump because sometimes the  stronger
 chlorination solution if operators don't watch it
 can corrode the pump.

 Thomas P. Ahrens:
     We  had  a somewhat  similar problem  at the
 Brush  substation  at Beaver Creek. What we did
 was just take a cylinder of chlorine and about an
 eighth-inch diameter  pipe  and ran  the pipe clear
down to  the  bottom  of the  well, then  we just
cracked that  cylinder  and we let  the chlorine
bubble up. However, it wasn't for potable use, it
was just for cooling water; but it did lick the prob-
lem.

Alfred H. Harder:
     I have three things that I'd like to get some
comment on. One is, I had an opportunity to work
on a well from the very first time it was constructed
and the water superintendent of this town was very
concerned  about  an iron problem in his wells so he
decided that iron bacteria were being introduced.
So  from the initiation of drilling of the well he
had  everything  chlorinated.  The workers wore
gloves that they  had to dip in chlorinated water,
the  drill stem was chlorinated, the mud was chlo-
rinated, and everything was chlorinated. He  still
had  iron problems in his well. I wondered if any of
you had  ever  heard of such  an approach as  this
and how it had come out. The other thing I had an
experience with was on an industrial well in which
the  pump  failed. They pulled out the pump; it had
a stainless  steel  pump shaft then, of course, but
around the pump shaft were layers of encrustation.
The  peculiar  thing was that it was  layered-light
layer, dark layer, light layer,  dark layer, etc.  So I
cut  this off and preserved it, and my first reaction
was that it must be 3 years old, and sure enough in
checking  with  the company the well was 3 years
old. It might be  a result of water level fluctuation.
 But strictly speaking, water level fluctuation didn't
 seem to  make a whole lot of  sense.  So then I
decided, well, it  must  be they pumped more water
 in  the  summer  than  they  did in the winter and
 maybe this caused the encrustation.  But I'd like to
 know if any of you have heard of a similar occur-
 rence, and what  would be your interpretation of a
 multilayer deposit like this? The stainless steel was
 not pitted, it was just as clean and nice as if it had
just  arrived from the  factory. The third  thing I'd
 like  to mention is  that I am really interested  in
 comments  previously  made regarding the relative
 size  of iron bacteria; I believe somebody said that
 all bacteria were about the same size while some-
 one  else  said  that iron bacteria were  larger  than
 others. I know nothing about bacteria, really. I'd
 like  a little elaboration on this because what I'm
 immediately thinking of is if the bacteria are all the
 same size, they must be able to move through the
 aquifer. How do they move away from the well?

 Thomas P. Ahrens:
      Well, I'm not a bacteriologist, I'm  just talking
 from what I  have  deducted from my  actual field
 observations. I think a lot depends on the nature of
 the  aquifer under study. If  you have an aquifer
                                             173

-------
 consisting of clean well-worked gravel, for instance,
 I think bacteria can travel for miles in it. But where
 you have a fine sand aquifer, and as Mr. Goldstein
 mentioned, you are making tests under unsaturated
 conditions, I think they will find that the bacteria
 are  filtered  out  rapidly. Now, looking  at one of
 your  other  areas of concern—the  deposition of
 alternating  layers  of  light  and  dark substances
 (perhaps organic substances) in wells—I have never
 seen anything quite  like that in  wells. I have seen
 it where you had  the filamentous iron bacteria
 where it just looked as though somebody had taken
 a bunch of horse hair and wrapped it around the
 shaft.

 Steven N. Goldstein:
      Just one comment about the size of bacteria. I
 guess a  lot of them will range  anywhere from a
 micron up to 10 microns; there might be some that
 are  even bigger. Some will travel in groups attached
 in a long  chain, so that will be even bigger. Some
 bacteria have flagella that make them capable of
 motion under their own effort. There are all kinds
 of different conditions. When you're talking about
 something  the size of the 1 to 10 micron range and
 then  you look  at  the  size of the pores in your
 aquifer  to  see what can get through, I think also
x whether  it is saturated or not makes  a difference.
 So  when  I said  they are all about the same size I
 mean "sort of in that  range."

 Thomas P. Ahrens:
      Along with what Mr. Goldstein said,.I did
 check on  sizes  of  sulfate reducing  bacteria.  The
 average  size was about 1.7 microns.  Most patho-
 genic bacteria are a  little less than a micron. That's
 what I was referring to when I was talking about
 sizes.

 Hank Bask i:
      Speaking  of iron  in  water, I  don't  believe
 everything is known about it yet, especially with
 respect  to well  water.  I've  personally had experi-
 ence with two  cases—one  a  Fox Hills  well  near
 Denver  where the wells when new do not have any
 iron in the water and perhaps 1 to 5 years later will
 develop  iron and  get  progressively  worse.  1 am
 inclined  to believe  from reports of  other people
 that something in the water is converting the casing
 to  soluble iron  in the  water. They ran a pumping
 test and  there was about 8 ppm iron; then  they
 heavily chlorinated it, pumped it again, and found
 the iron  content had  dropped  to about  2 ppm,
 which is quite a drop. The  other situation involves
 Dakota wells in  the Pueblo, Colorado, area; these
 wells are all characteristically high in iron content.
       Let me make reference  to  a well in this area
 that I field tested last summer;  I checked the iron
  174
content of the well water as it came out of the well.
The  iron concentration of the water varied as we
varied the pumping rate of the well. We discovered
that  the water produced  at the higher  pumping
rates contained lower concentrations of  iron than
did water which  was obtained from low pumping
rates. Thus, the differences in iron concentration
could  be due to  the  contact time,  which the
water has in the well casing. It should be noted that
the water was warm, being almost  100 degrees F,
casing  string was 1800 feet long, pump was set at
probably 400 feet; so it had a difference in travel
time and distance sufficient for that water  to pick
up iron from the casing. I'm seriously  thinking of
trying  to case the next well we drill in the Pueblo
area with  fiber  glass casing,  and  see  what the
difference in iron content is.
Edward M. Burt:
     I  agree with the statement that not everything
is known about  iron. When you start calculating
the  pounds per day produced by a well and how
many  pounds of iron  you've got in  the casing,
look out!
Jerry Frick:
     I'd like to ask Ed a question since he's  familiar
with the sands in Michigan on our side of the  State.
We have had in the older, small residential wells the
old, type two metal  screens that are supposed to be
subject to encrustation and dielectric action.  These
older wells also are  of small diameter and have low7
pumping rates. We have put all stainless steel points
in them which would have lower intake velocity,
and we've had a great increase in  the amount of
encrustation on  several of these wells. I  wondered
if you  would  have  any  calcium  deposits, and
encrustation, just  cementing over of the screen.
Some of the older screens have lasted  for 10 to 15
years, but  these new screens are  completely ce-
mented over in 2  years' time.  I  wonder if you
would know anything about that.

 Edward M.  Burt:
     Are you talking about replacing the screen?
 You've pulled the screen?

Jerry Frick:
     Yes, we've  pulled the old type out-of course
 they've been plugged up  but they had been in for
 10,  15, or 20 years.  Then we put in the newer type,
 and maybe in 2 years' time they're encrusted.

 Edward M. Burt:
     When  you make this change do you redevelop
 the well—acidize it,  clean  it up?

 Jerry  Frick:
      Not acidizing the well—it's in the small 2-inch
 diameter. In the same area, we'll put in 4-inch wells

-------
which  would have a lot more intake velocity, and
we  don't  have trouble;   it's  just  in the  small
diameter wells which have slow pumping rates.

Edward M. Burt:
     My point was that if you simply pulled  the
screen  and set a  new one in its  place  and you
haven't redeveloped,  you haven't  done  anything
about  the  plugging that is occurring in the forma-
tion outside of the screen.

Jerry Frick:
     Well,  in  our area the formation all  caves in
when we pull the old  screen out. We have to drive
the new one or drill the hole  out and  install a
new one;  sometimes  a formation  may even have
heaved up  the  pipe and we have to clean it out  and
install  a new screen just as if a new well were being
drilled. The only development that the small diam-
eter wells  get  is only that which  is incidental to
their being pumped; and unless  we have problems
where  there's  low capacity,  we sometimes surge
them  and develop  them a little  more.  But  not
normally, no.

Edward M. Burt:
     Well, then I'm  afraid I can't  answer your
question.

Leslie G. McMillion:
     We had some questions left over today regard-
ing Jim Waltz's paper which he didn't have time to
answer.

James P. Waltz:
     One of the questions was, "could you put  on a
slide showing a rock formation and point out dips,
joints, sets,  etc. for  those nongeologists who are
not familiar with the  jargon." Well, obviously  over
the microphone I can't put on a slide, but I might
at least define what a joint is. If there are still any
nongeologists, and  for the  record, the  joint  is a
crack  in the rock, a  fracture  in the general sense,
along  which  no movement has occurred. A  very
 interesting characteristic  of  joints  is that  they
 often  occur in geometric patterns that are fairly
 predictable  in  their  orientation.  In  other words
 there  is seldom a single  crack, there are usually a
 number of cracks  that are parallel to one another,
 all of them oriented in approximately the same
 direction.  And usually there is  more than one set
 of these  cracks.  Secondary  and  tertiary sets  of
joints can usually be found and  each of these has a
 characteristic direction and dip, at least  character-
 istic in a  statistical sense. Dip is simply the direc-
 tion of inclination. The direction of dip defines the
 direction  of inclination.  The  magnitude  of dip is
 the angle from the horizontal,  the vertical angle. So
you  can have  a gentle dip or a steep dip, and the
direction is the direction that water would run off
if you were to pour water on one of these fracture
surfaces.  Strike  is a term that is commonly found
in  the geologic literature. It isn't necessary to use
the term  strike.  If you define  the direction  and
magnitude of dip of a fracture you have completely
defined its geometric characteristics.
     A second question: "For the first study, was
the  degree or amount of pollution in  the wells
simply the amount of E. coll bacteria?"  For most
of  the  wells  that we  sampled  in this  mountain
pollution  study, we considered  the well to  be
polluted if we were able to get several samples from
the  well  which showed amounts  of more than 1
E. coli per 100 ml sample. The health standards are
consistent with this criteria—that if you  find more
than 1 per 100 ml, it's considered contaminated. If
we were able to reproduce this result we considered
the  well  to   be contaminated.  I considered  the
arrangement  as sort of a black and white situation,
either there are E. coli present or there are not, and
it wasn't of  much  concern  to  me whether  there
were 2, because it's a matter of degree, like preg-
nancy and the common joke.

 Russell E. Larson, ARS, USDA, St. Paul, Minnesota:
      This afternoon you showed a slide  of a cabin
 down in  this flat, gravel plain,  and you indicated
 that they had pollution problems. Did you say they
 corrected the problem by treating their well water,
 or treating the effluent that  was getting into the
 stream, or did you say that the stuff just goes into
 the stream? What are they doing about this?

 James P.  Waltz:
      To my knowledge, they are not really  doing
 anything about it. Many of them are drinking con-
 taminated water. A number of people who recog-
 nize the problem have chlorinators on their wells.

 Russell E. Larson:
      On  their wells,  yes. But,  they don't worry
 about what  happens to the effluent that is going
 out into the  stream. Is that it?

 James P. Waltz:
      That's right. I don't know of any efforts by
 these people to treat the water before it leaves their
 leach-field system.

 George Taylor:
      I'm asking a question that  should probably be
 answered by  a sanitary engineer. I'm not a sanitary
 engineer; I'm a ground-water geologist. As I under-
 stand, they  are using, more or less, one standard
 type of leaching field, septic tank design for most
 of these homesites. Is that  correct?  Is it pretty
                                             175

-------
much a  uniform  design  or is  any consideration
given to  alternative designs that  would  be more
effective  in this kind of geologic terrain?

James P.  Waltz:
     Well, again, I'm groping a little bit as far as
what all  the alternatives are.  I  know there are
different designs for leach fields, different types of
distribution systems, drain tiles, and of course dif-
ferent sizes,  depending on the amount of effluent
that is anticipated.  I firmly believe that there are
perhaps  better  methods  of designing these leach
fields-, primarily what I have in mind is that addi-
tional material  could  be brought in  to  serve as a
filtering medium.

George Taylor:
     Do  you mean sand filters or something of that
kind?
James P. Waltz:
      Right—this sort  of approach could certainly
improve the  situation.

George Taylor:
      1 would certainly think so, rather than relying
on the natural material itself.

James P. Walt/:
      There  is an  interesting battle going  on in
Colorado  right now. There  is a fellow who has
developed an individual sewage disposal system that
 involves letting the effluent, still kind of ripe, go
out on the surface of the ground, and as a matter of
 fact, I think this would be an improvement over
 the present  situation because the small amount of
 soil  that  is  present  in  the mountainous terrain
 would be defective as a filter  and if the bacteria
 are brought in  contact with the soil, with the sur-
 face, I think that many more of the bacteria would
 be removed than the present situation where they
 scrape away all of this good soil, the 6 inches,  8
 inches, 1 foot  or  2 feet, and  then they put their
 leach field  bed right on the bare rock. So there
 certainly are improvements in the design that could
 be applied right now.

 Paul Goydan. Koppers Co., Inc.,  Pittsburgh, Penn-
 sylvania:
      Under  present Colorado landfill practice, do
 you permit  septic  tank construction in gravel or
 alluvial deposits today?

 James P. Waltz:
      I assume that the answer is "yes." I am not
 directly involved in approval of septic tank loca-
 tions. The county sanitarians are usually the agents
 who say, "Yes, you can put your septic tank here
 or your  leach  field here." I keep using the term
 septic tank  when really  we're talking about  the
 176
leaching field that is the critical element. The septic
tank functions fine because, in terms of releasing
effluent, it does it just as well  in the mountains as
anywhere else; it's the leaching field end of  the
treatment that is failing,  in my estimation, in  the
mountain.  And, it fails where  leach field effluent
can pass into clean gravels because the gravels do
not adequately filter and  purify the effluent, and I
have seen a number of instances where very clean
gravels have comprised the ''soil" into which these
leach fields  have  been  constructed.  I had  some
argument at  dinner this evening that some of the
mountain streams, which  I generalized about, have
perhaps much tighter alluvial materials than I have
admitted; and  I think certainly  there must  be all
kinds of conditions, because if  you get very high in
the mountain streams you begin to get into glacial
materials  or you  have  all sorts of permeability
characteristics.

Paul Goydan:
     In an alluvial type deposit where you  essen-
tially have flat terrain, and your distribution field is
2  to  3 feet below  the surface, and in  this very
permeable material and no surface cover, you may
expect a  fairly aerobic type of condition to exist.
Now would  this  perhaps enhance  the over-all
degradation, the ultimate goal of getting a fairly
safe by-product?  I'm  curious  because under ideal
conditions where  we  do  have the necessary  soil
depth and the necessary  degree  of degradation
occurring we're in fine shape.  But what happens in
a  very permeable  type material?
James P. Waltz:
     Let me comment first and I can be followed
by Steve. As long as the effluent is passing through
very permeable materials I question the amount of
filtering and adsorption that can occur. I'm sure it's
a  matter  of degree and that if you go far enough
through highly permeable material you can achieve
the same effect as passing the fluid through a few
inches or  a  few feet of appropriate soil material.
 But this is really outside my area of expertise.

Steven N. Goldstein:
      I'll try and  place  a little  perspective on  it.
 First of  all, when you look at  the soil system
 itself it's thought that most  of the degradation is
 being accomplished by microorganisms, and so you
 must have the  makings of a soil, some kind of an
 environment in which these organisms  can live in
 number and sustain themselves. If you just  have
 gravel I'm not sure  that when effluent is not being
 applied the  environment would  be all that great to
 keep them alive.  Now remember  if you're talking
 about  just  gravel,  they're going  to swish  right
 through  there. It's not  going to  be a  very  long

-------
percolation.  I  was  reading  some  minutes of  a
Symposium held up at Dartmouth in March of this
year and I believe I'm quoting now from Professor
Ryan  Locke from  Connecticut University.  He
came up with an interesting point, if indeed it's him
that I'm quoting. He indicated that while clays perc
much more poorly, much more slowly, much more
impermeably than sands—and sands just seem to be
great—it's really the clays that are doing—what's the
jargon word—"beneficiation" to the water, actually
purifying it a lot more.
Paul Goydan:
     You mean adsorption?

Steven IM. Goldstein:
     Well,  whatever it is that the soils do, clays are
doing it better than  sand; in other  words, they're
taking a lot more of the  objectionable quantities
out  of  the water by perking slower,  and a lot of
people  say they don't want  to put a leach field in
clay. But, if you design a leach field of sufficient
size to  handle  your load,  clays  apparently  are
treating the water a lot better than sands. And it
may be because they give a better climate to the
microorganisms.
Paul Goydan:
     This perhaps gets a little bit into the  ethics of
the  true purpose of the septic tank, or landfill or
anything  you  are speaking about. Do we really
want to  absorb  or adsorb,  or whatever the term
may be, onto  something  such  as a clay or do we
really want  a degradation?  I mean, are we just  in
fact perhaps storing this?   I  realize you desire
permeability so  that you  may hold this particular
material that is giving your substance BOD or COD
so that microorganisms have sufficient time to act
upon it. But this is not the  same thing as physical
adsorption and perhaps the distinction here should
be  made. The clay will adsorb perhaps, but unless
it has the necessary biological population or organic
matter  in the vicinity to function, it will not really
degrade. And perhaps this is really ultimately what
we want. So that's why I  asked the question.  Per-
haps we can still achieve the desired ultimate goal
of  degradation  through   a  little more permeable
type material,  something almost  like a  trickling
filter, although in trickling filters you do have algae
and perhaps insects and this type of thing that helps
the system  along. But in a sense,  maybe we can
have an underground type of trickling filter and not
adsorption as such.
 Russell B. Stein, Ohio Division of Water, Columbus:
     I  came in  here a little late and I wanted  to
inquire  about the use of septic tanks.  In Ohio,
septic tanks in recent years  have been under quite
a bit of scrutiny and, as  a  matter of fact, I think
that in some counties they are actually outlawing
them on  new  installations.  Now,  of course,  in
established farms and things of this nature, if you
have a septic tank you  continue to use it. The
reason for the  scrutiny,  I suppose, at least in our
part of the country,  is that in the glacial till soils
and  the high percentage  of clay, septic tanks just
simply do not work. We were talking about alterna-
tives for individual waste disposal and I wanted to
ask  the question, has anybody discussed in this
section the use  of aerobic units, aerobic digesters or
egg  beaters,  or whatever you want to call them?
This is something which, in a few counties in Ohio,
if you develop a housing development in a subur-
ban area, you  either  have  to go into this type of
disposal or you have to  go into a central sewage
treatment  plan. Our  experience with a tight clay
soil indicates aerobic units  are a better  means of
sewage disposal for the individual homes. There are
disadvantages,  the main one being that people
pull the plug on them and they just discharge raw
sewage. But if there  was some regulation on that,
these would do a pretty fair job of aeration and
sewage treatment.

Edward M. Burt:
     Earlier we talked about the different types of
sewage treatment and the discussion of septic tanks
is  like  primary  degree  treatment which accom-
plished certain functions. You have perhaps heard
of secondary  degree treatment  which takes this
criteria of primary treatment to a higher degree of
treatment.  But I'd like to point out, both to you
and to the other speaker, that when you're talking
about septic tanks, you're really talking about the
mechanisms involved in secondary  degree sewage
treatment plants; further, we should realize that
the next  step  after that is tertiary level treatment
which in the Midwest now  consists of phosphate
removal,  but pretty  soon  it's going to include de-
nitrification. Let's all bear in mind that when we
talk about any one of these treatment methods,  if
we don't take  a look at the total ecological picture
we're just pushing the marble around. We say that
E. coli in certain concentrations are indicative of
pollution  so we want to chlorinate and  kill them
down to an acceptable  level. We do not like high
biological oxidation demands (BOD); so, we satisfy
the oxidation of bacteria to accomplish an accept-
able degree of treatment. But  the  nutrients  are
still there but may be in another form. The reason
Michigan  is going to oxidation  lagoons, aerobic
ponds, and  spray irrigation  is to utilize-even use
 up—the nutrients, the effluent from a treating proc-
 ess, in irrigation of crops.  Thus, the cropping is an
 important  part  of  the  treatment;  for  instance,
 phosphate is one that is  readily picked up by many
                                             177

-------
of your soil particles, especially your clays, by ad-
sorption. The growing  crop can again pick that
phosphate up and utilize it and so you are com-
pleting the  cycle of changing it around.  In like
manner, the crops use nitrates in their growth cycle.

Russell B. Stein:
     I think you misunderstood me or I got off the
track. I was referring primarily to individual home
sewage treatment. Whatever method you choose for
sewage disposal,  of course  depends on individual
conditions,  but I've seen, particularly  in Ohio, so
many cases  where people have a septic tank system
with as many as four leaching tile fields on a half-
acre lot. Their whole yard  is a leaching tile field
yet in the spring of the year they have raw sewage
coming up through  their grass.

Edward M. Burt:
     They   don't  have  raw  sewage,  they  have
primarily treated sewage, which isn't acceptable.

Russell B. Stein:
     I  don't  know  what  it  is,  but it's pretty
raunchy  stuff. It's black and evil smelling.

Steven IM. Goldstein:
     There are individual system aerobic treatment
units which in many ways are similar to the septic
tank in construction except that air is mechanically
pumped  into the tanks, thus attempting to keep the
conditions aerobic  rather than anaerobic; a septic
tank  is  anaerobic  digestion,  so  now you  have
aerobic organisms taking over and  it's more than
just bacteria. You  may have other kinds of bugs
involved there. It's still  not entirely clear; however,
a  lot  of investigations show  consistently  better
performance for properly operated aerobic  units
as  opposed  to  septic  tanks.  There  is  only one
hooker.  The hooker is you have to keep that pump
working. If the pump doesn't work then the whole
thing goes "kaput." And that turns the individual
homeowner  into a waste  treatment system  oper-
ator; he must check on the pump, he must make
sure everything works  properly and this is some-
thing which in general  he doesn't want to do nor
does he know how to  do it. Secondly, there are
very few qualified  people  who can maintain the
system  even when  the  homeowner finds out that
something  is going  amiss.  In general,  some places
that do  require that you have aerobic systems  as
opposed to  septic  systems also  require that you
have  a  maintenance contract with  a qualified
maintenance person. So it does show some promise
but there are a lot  of problems in terms of proper
maintenance which you never have with a septic
tank.
James P. Waltz:
     One quick comment regarding your situation
in Ohio. As I understand the aeration system, you
still have the effluent to  get rid of, so you haven't
solved  anything other than making the fluid more
acceptable  by switching to the package aeration
unit.

Russell B. Stein:
     This is true, and then it's simply a matter of
the fluid either going into a leaching field or more
frequently  into an open ditch or the nearest water
course.  The  comment I started to make,  and as
this  gentleman just  emphasized,  is the fact that
some people don't want to spend  the money for
the power  to operate another motor; and, they do
pull  the plug on  these  things. So, if you're dis-
charging into the low part of your land, then you're
strictly  discharging raw  sewage. If you had some
control that prevented this from being done these
might  perform satisfactorily.  I'm  sure it's a legal
problem, and it's certainly a moral problem. But if
they maintain them, it's  my feeling that in certain
geologic conditions it's a much better sewage dis-
posal system because the effluent is  of a much
higher quality.

Edward M. Burt:
     We seem  to be  evolving around anaerobic
sewage  treatment and aerobic sewage treatment.
Your septic tank is anaerobic and  all your anaero-
bic systems  characteristically are  the odoriferous
ones. The aerobic ones are not so odoriferous. Even
if you had an aerated septic tank system and then
went back into  the soil with a leaching tile  field,
the nutrients are  right there, the anaerobic bacteria
in the soil  would take over and you would be right
back to your anaerobic thing  that came out of the
septic tank in the first place.  It would just change
the  type  of bugs that thrive on  it, back to the
smelly mess.

Roy E. Williams,  University of Idaho, Moscow:
     Mr. Burt, what happens to a State or federal
agency which issues a permit to dispose, and which
disposal results in a gross pollution problem such as
the one you were discussing today?

Edward M. Burt:
     Under  the  permit  system of Michigan, the
Michigan Water Resources Commission has jurisdic-
tion over all of the surface and ground waters  of
the  State. This  agency's permits are not fixed  in
time. In other words, each one is  a contract with
either  a public  or  private corporation  that says
"O.K., you want to  do this, these are the limits,
this is  the degree of treatment that you have  to
 178

-------
accomplish; this surface or ground water is accept-
able for receiving this type of waste." The party
receiving the permit has to monitor the waste dis-
charge and  has  to  submit  reports  and proofs
monthly that it is living within its stipulated range.
I mentioned this afternoon that there is a "hooker"
clause  at the bottom of all these contracts which
states  that at  such  time  that a higher degree  of
treatment or a nuisance or any  other  thing is in-
volved that requires re-evaluation and re-establish-
ment, the  State agency can order  a change.

Roy E. Williams:
     In other words, nothing  happens to the State
agency if pollution results? Is that correct?

Edward M. Burt:
     Yes,  the  responsibility rests on the user of
the resource.

Paul Goydan:
     Mr.  Burt,  if you  only  have the  funds  to
construct   a  few monitoring wells, where would
you put them, and to what depth below the water
table to learn the vertical extent of contamination?

 Edward M. Burt:
     We design the monitoring wells after we have
 drilled exploration  wells to evaluate  the geologic
 and hydrologic  conditions,  and to establish the
 observation well  pattern. To illustrate this with the
 example which I discussed this afternoon, we first
 found out the basic  conditions, the direction of
 flow, area of contamination;  and we worked from
 the surface of the water table on down. In other
 words, we sampled as we went down. From this
 we came  up with a final design  of the monitoring
 system with shallow wells that were in the upper 15
 feet of the water table, and others within the bot-
 tom   15  feet. (This  was a  water table aquifer,
 incidentally.)  But,  the business of limited funds
 goes back to the responsibility of having sufficient
 proofs that  you are complying,  or you have suffi-
 cient data and it doesn't stick. It's the responsibility
 of the party  that's going to use this resource to
 protect it. And  if he does cause contamination or
 pollution, that is unacceptable  and he has to cor-
 rect it. As far as the regulatory agency is concerned,
 they don't care what it costs.

 Russell B. Stein:
     Mr.  Schmidt, we had a relatively small nitrate
 contamination problem  in  a rural area of Ohio
 which I worked  on a number of years ago. It only
 involved  about 4 or 5 square miles and possibly 25
 or 30 individual farms,  but we had some pretty
 high nitrate values. We had  a situation  similar to
 what  you  described  in Fresno  where there were
several potential sources of nitrate in the water. For
example, there were  agricultural fertilizer,  septic
tanks and  feedlot operations.  In our  chemical
analyses we tried  to  make  a distinction between
some of these sources by analyzing for chloride
and also potassium. We had  pretty good base data
in this area and  we  knew  what was normal for
most of these  values;  and, in some instances where
we  had rather high nitrates, we would get high
potassium values and  low chlorides. In these cases
we suspected chemical fertilizers. In other instances
we had very low potassium (maybe zero potassium),
high nitrate,  and  high  chloride and this  led us
toward  sewage sources.  Did you  do any analysis
with potassium  and  do  you think this  is a valid
tool to use in this kind of study?

Kenneth D. Schmidt:
     In my study area  there is one outstanding
source of potassium,  namely winery waste water.
One of the by-products of the winery operation is
potassium  bitartrate.  So, using potassium would be
logical near wineries.  The problem with potassium
is that it is not very mobile in these alluvial ground-
water  systems. Potassium is also high in  sewage
effluent as compared to  most ground water. If one
wants to distinguish between agricultural fertilizers
and other  sources, chloride is a very good constitu-
ent  to consider.  There  are difficulties  in using
potassium  in  agricultural areas  such as  in  Fresno
County. Little potassium fertilizer is used as com-
pared to anhydrous ammonia. Anhydrous ammonia
when  applied to the  soil is known to have  a great
replacing  power.  Thus, cations such as calcium,
sodium, and  potassium  may be  replaced by am-
monium ions from the fertilizer. However, in areas
other  than alluvial  basins,  potassium may  be  a
useful cation. People should be encouraged to use
all available chemical data, not  only chloride and
 potassium, but boron, fluoride,  and other constitu-
 ents.

 Russell B.  Stein:
      I  probably  should  have clarified this  a little
 better. This was  an  area in northwest Ohio which
 was in a very shallow carbonate area; the highly
 cavernous limestone  aquifer was exposed at the
 surface  in the area; it was  likewise an  area  of
 extremely poor well construction. Most of the wells
 were between 50 and 75 feet deep but were only
 cased to depths of 10 feet or less and for this reason
 we  felt we were  getting extremely rapid drainage
 from the surface  directly into  some  of these wells.
 We corrected most   of  the  nitrate-contaminated
 wells by deepening them and double casing them,
 to  100 feet  or so, and it  did  solve the problem.
 Now, concerning the potassium, we  have observed
                                             179

-------
that if ground water in Ohio has more than 1 ppm
or maybe 2 ppm something unusual is happening;
thus, we were alarmed because we were getting 20
to 30 ppm  in  some  of these wells. Something
caused this and in this instance  it wasn't due to
wineries.

Paul Goydan:
     From hearing the talks and discussions the
last 2 days, it seems to me that we're largely talking
about the soil as a means of disposal, not only in
landfills  or  in septic tanks; but irrigation has also
been  mentioned. I would like to emphasize this.
Much of the research today has been  directed
towards  surface-type  treatments, and  we  have
really neglected this natural resource we have in the
soil and its ability to do an effective job. One of the
things that should come out  of this conference is to
direct research and funds to this area. I believe this
has a lot of promise. We're  not only talking about
contamination, we're talking about disposal, degra-
dation,  and available means for  solving our  prob-
lems. Let's look more closely at the  soil, and one
area I'm particularly interested in is soil irrigation.
We haven't had  too much discussion there but this
is an area  I'd  like to  see the various  control
officials, industry, and organizations get  involved
in. Perhaps we  shouldn't be so  ignorant when it
comes to subsurface means to dispose of our waste.
Let's get down  to the nitty gritty, face the facts,
and really  try to solve the problems; and hopefully
this will be  something that will  come out of the
conference.

Edward  M. Butt:
     I agree with you. I'd like to point out a couple
of studies that  are going on.  One that has  many
years of background research is at Penn State Uni-
versity.  We  are participating in  an EPA  financed
study. The City of Belding, Michigan, with a  popu-
lation of around 5,000 has  a series of waste  water
lagoons. A portion of this effluent is being used in
two irrigation areas, summer irrigation and winter
irrigation.  This  is  all  layed out for 2, 3, and  4
inches of application, the type  of soils,  and it's a
complete type of study where you analyze the soil,
how much penetration, and  effect on various  crops.
In addition to these 2, something about the sewage
treatment operation known as the Muskegon  Coun-
ty project—this is also in Michigan. There again they
are  talking  about oxidation  ponds,  lagoons  and
irrigation,  and  some questions were raised  today
about this.  There was comment  that under the
irrigation area they were  speaking of underdraining
it with  tiles; I  do not know the details but my
understanding is that while this is another  sandy
lake  plain  with unconsolidated  sands, it  will run

 180
down to maybe 50 or 75 feet and their intention is
to ring  the irrigation fields with dewatering wells
and pump  (in the terminology of Penn State) the
renovated waters captured by these wells that are
intended to collect anything that seeps through and
isn't used up in the soil or taken off by the vegeta-
tion, and discharging that to a surface water body.

Paul Goydan:
     We do need to look  at the specific variables
involved. It is fairly well established that it is feasi-
ble and practical in many  instances to treat waste
fluids,  particularly  municipal  sewage, by spray
irrigation and production of vegetative crops with
it. The  Penn State  studies in particular  showed
these systems to be feasible. But, these studies fail
to get down  to what the real mechanisms involved
are, and to show how the  data can be  extended to
and used in other areas. They showed  some of the
characteristics  of their system  and they  showed
that the system can work, and they were working
with a domestic waste at that. But, when you start
talking  about industrial  waste and various special-
ized types of chemistry, here again is another story.
I think we are moving in this direction and I'll be
looking forward to getting the data and the results.

Leslie G. McMillion:
     Along this line of using  the soil as a living
filter and  treating  media  for  sewage and waste
waters, I have recently become convinced that such
waste treatment methods could have a place in the
over-all plan for sewage treatment and disposal in
such  heavily populated areas as Long Island, New
York. If you're  familiar with the ground-water
situation on  Long Island you realize that past and
present methods of handling the Island's municipal
and domestic sewage have resulted in some publi-
cized cases of ground-water contamination. These
cases involve local buildup of relatively high  con-
centrations of nitrates and other substances such
as chlorides. The apparent cause of the local prob-
lems is downward seepage of contaminating fluids
from septic  tanks,  sewage lagoons, and the  like.
There are literally thousands of individual domestic
sewage disposal systems on the Island since most of
it is not served by public sewer lines.
      The soil treatment method could  be especially
useful  on  Long  Island  since a principal  nutrient
source  which  can be removed by this process  is
nitrates; and,  as you probably know, contamina-
tion by high levels of nitrates is one of the primary
concerns of the Long Island situation. However, we
 cannot  disregard the  difficulties  that could  be
 involved in  instituting the actual treatment opera-
 tions in such areas  as this one; the main  problem
 could be the availability and price  of land for the

-------
operation.  Another  difficulty  could  be technical
know-how itself; and in this regard, we are thinking
of the actual design and operation of a unit.
     While  these  concepts of  soil  treatment of
wastes are  in their infancy, it is interesting to note
that  they  are  not  being  totally ignored  in the
country. A considerable  amount  of money has
been released to the  U.S. Army Corps of Engineers
so that the agency can make studies of waste water
collection, treatment, and disposal in  5 of our large
metropolitan areas:  San Francisco,  Chicago, De-
troit, Cleveland-Akron, and Boston. It is my under-
standing that what prompted the initiation of these
studies was concern  over the possibilities for using
the so-called "land  treatment" methods (such  as
spray irrigation of waste water) as a way of treating
municipal  sewage. While many are skeptical about
these methods providing all the answers, we do see
a place  for them in  the  total waste treatment and
disposal picture. It  may be inconceivable that all
the waste  fluids from these 5 metropolitan areas
could be treated by the land treatment approach;
but it is reasonable to assume that some of it could
be treated  in that fashion. So, there may be possi-
bilities for broad applications of this concept. If
possibilities exist for its application in the northern
areas of the nation, the opportunities  for success of
the method should be much greater in the southern
and southwestern States  where the growing season
is longer.
Edward M. Burt:
     In  the conversations  we've had  in  the last 2
days, while the emphasis is on utilizing the soil for
waste treatment and what happens from the land-
fill and from the sanitary waste water aspects, and
industrial   aspects,   another  thing that  has been
paramount is the general realization  that it takes
the combination of  many types of skills to accom-
plish a workable solution to these problems. It isn't
just the geologist or  the hydrogeologist or the sani-
tary engineer. When you  start working with the soil
you should be looking for the agricultural specialist
because the type of crop that you  utilize in an
irrigation system can make a great deal of differ-
ence. For  instance,  the crops that we're most fre-
quently talking about in Michigan are  corn, because
it  has a very good extractive characteristic; Reed
canarygrass;  and  brome-alfalfa mix. Now some
types of alfalfa instead of taking nitrogen  out  of
the soil, take it out of the  air and put it in the soil.
We're looking for those things that take the nutri-
ents back  out  because to a  degree  the soil  is a
storage  reservoir for those things that  you want
captured by adsorption  and  then taken back out.
Clay  minerals have generally very good  capability
for removing phosphates, but  you want to get it
back out of there; otherwise you're storing it.
     This brings up another thought—I mentioned
today  that  this industrial company in  the final
treatment methods  that  were  used,  the  highly
concentrated combustible waste which was in  the
order of IVa to  2% went  to incineration, and  the
question was asked,  "What happens to the waste
from the incineration?" This is collected and goes
along with the other 2%% of the wastes that require
additional treatment, into a concrete stabilization
lagoon  or  pond with  aeration, for removal  of
ammonia (for instance), and where we are trying
to develop a growth  of bacteria to prove the feasi-
bility of going  into a standard secondary  degree
treatment process. I  mentioned that the cost of
incineration was about the same as the cost of dis-
posing of the same volume via a deep disposal well;
thus, some of you are perhaps wondering why we
didn't go to deep subsurface disposal. I mentioned
we'd been  considering casing to a depth of about
3,000 feet and disposing into formations between
3,000 and 5,000 feet. I, as a geologist, worked on
the design  and the method of deep well disposal
and one of our sanitary engineers, along with some
combustion engineers,  studied the economics of
the incineration; then we sat down and argued the
pros and cons of each method. In those discussions,
I expressed a personal philosophy that I still adhere
to—that the more things we can handle on the  sur-
face  of the  earth where we can readily get at the
mechanical  workings to make proper repairs,  the
better off we are in the long run.
Leslie G. McMillion:
     I  have been notified that all three of the other
sessions have adjourned. In light of that and since it
is past  ten o'clock,  we will very shortly conclude
this session.
     It is most difficult for me, as session chairman,
to appropriately express my appreciation for such
tremendous  participation as  we  have  had  here
tonight. I  thought  that  our  panel  really  outdid
themselves in  answering  the  questions  from  the
floor and providing other comments for the record
and for our edification; I thank each one of you for
your excellent contribution. This session would not
have been possible without the presence and active
participation of you who have attended tonight.
We  thank you for  your attendance  and your
enthusiastic participation. We  were very glad that
several of you helped answer some of the questions
that  were asked, and  provided information on some
knotty technical topics.
     This session has been especially interesting and
stimulating. I am  sure that we all hope that the
enthusiasm of this meeting will carry  far beyond
Denver and  this point in time.
                                            181

-------
 A   Systems  Approach  to  Management  of  the
 Hanford  Ground-Water  Basin"
 by D. B. Cearlockb
                    ABSTRACT
      Mathematical models for simulating ground water and
 radionuclide movement as a function of time and space are
 being developed. In addition to the models, a man-machine
 interactive computer  system is under development for use
 in model applications. In total, the research and develop-
 ment program will produce a management and engineering
 too! for use in analysis, decisions and policy formulations
 relative to management of ground-water systems.
      The system is  separated  into sequential  or  parallel
 components that can be modeled independently of each
 other. This results in maximum capability to simulate all
 combinations of situations that may be encountered and in
 ease of  modifying or refining the models independently
 without  having  to  reformulate the entire  system. The
 system is composed of three major categories of models:
 (1) data models; (2)  hydraulic  models; and  (3) water
^quality  (transport) models. Data models calculate input
 characteristics  required  for operation of the hydraulic and
 water quality  models from  a minimum of field measure-
 ments. The Transmissivity Iterative Routine for calculating
 transmissivity distributions and the Sorption Transmissivity
 Routine  for calculating  sorption coefficient  distributions
 are two types  of data models.  The  hydraulic  models
 predict  the  flow of  ground water in  saturated and un-
 saturated soils. The  Partially-Saturated Transient  Model,
 which describes unsaturated, transient flow, and the Vari-
 able Thickness Transient Model, which describes saturated,
 transient flow,  are included in this category. The water
 quality models predict the movement of the waste through
 the subsurface soils. The Macro-ion Transport Model, which
 describes macro-ion movement, and the Micro-ion Transport
 Model, which describes micro-ion movement, are included
 in  this category. The assumptions used  in developing the
 system of models, justification  of these assumptions, the
 interrelationship  of each of the models, and  the intended
 application of the entire system are presented.
      The man-machine  interactive computer system pro-
 vides an efficient means for the engineer to interact in the
 problem  solving  functions using the previously  discussed
 models. The system allows the engineer to rapidly scan a
       Presented  at  the  National Ground Water Quality
 Symposium, Denver, Colorado, August 25-27, 1971. This
 paper  is based on  work performed under  United  States
 Atomic Energy Commission Contract AT(45-1)-1830.
      "Manager, Water Resources Systems Section,  Water
 and Land Resources Department, Battelle Pacific Northwest
 Laboratories, P. O. Box 999, Richland, Washington 99352.
 182
large  number of alternatives  and  use  his experience in
rapidly converging on a solution. The components of the
computer system and their functions are described. Exam-
ples showing how the computer system  is being used with
models that have been developed are presented.
                INTRODUCTION
      Since  the beginning  of the  Atomic  Energy-
Commission Hanford  Project in the  mid  1940's,
geological and hydrological investigations have been
conducted  to  establish and evaluate management
alternatives with respect to storage and disposal of
waste materials in the vadose  zone above  the re-
gional  water table. Initially  these investigations
consisted  mainly of analyzing  (1) well  logs  and
driller's samples for  stratigraphic information and,
(2) ground-water potentials for ground-water move-
ment patterns and rates. The efforts concentrated
on defining  and  understanding  the water-table
aquifer because of  its direct relationship to waste
management activities. Historically, Hanford waste
management practices  have been based on conserv-
ative  assumptions  and  an extensive monitoring
program to  evaluate and  measure the effects of
these practices.
     Investigations have continued in an effort to
improve the understanding of the water-table aqui-
fer. The  aquifer has been classified  into textural
types and pump tests  have been conducted to de-
termine  the permeabilities of each textural type.
This data showed that the Hanford ground-water
system was highly variable.
     In an attempt to  develop a practical tool
for   managing  the  ground-water basin,   simu-
lation techniques  were employed. Initially, simple
graphical  flow net analysis  and analog models were
tried. With the increase of capability and flexibility
of the  digital  computer, the emphasis shifted to
mathematical simulations. The modeling effort was
directed toward simulating steady, ground-water
flow.  These  models  were theoretically sound, but
the output was less reliable than desired because
the data input requirements could not be economi-
cally satisfied.  It was, therefore, necessary to make

-------
compromising  assumptions in order  to  use the
models. In addition, the models could not simulate
transient conditions or waste  transport, which are
required to answer many of the pertinent problems
associated  with  management  of a ground-water
basin.
     Recent pressures for multiple uses have posed
important questions concerning the management  of
the Hanford ground-water basin. Although signifi-
cant  time  and  effort have  been  expended on
collecting field data and conducting field investiga-
tions,  methods have not been available to utilize
these data  to produce  the  answers required for
waste management decisions.
     To provide  the methods for answering ques-
tions concerning management of complex ground-
water basins,  a major  program was undertaken  to
develop  a  system   of mathematical  models for
simulating ground water and waste movement as a
function of time and space. Selection and develop-
ment of the  system of  models was based on the
economics of data input requirements, precision of
results required, and the practicality of application.
In addition to the models, a man-machine interac-
tive computer system is being developed to provide
an efficient means for  utilizing the models and ana-
lyzing the results. In summary, the program is pro-
ducing a management  and engineering tool for use
in analysis,  decisions and policy  formulations rela-
tive to management of ground-water systems. A brief
description  of the  models, assumptions and  justi-
fications of the  assumptions, the interrelationship
of each of the models  and the intended application
of the entire system are presented. Also included is
a description of the interactive  computer display
system.

   DESCRIPTION OF SYSTEM  COMPONENTS
     The objective  of  this program  is to develop a
ground-water resources management tool which can
be  applied  practically to both local and regional
ground-water systems  and which will provide reli-
able answers to those  pressing problems associated
with the exploitation  of ground-water systems.  To
accomplish this objective, the ground-water system
was separated into sequential and/or parallel com-
ponents that could be modeled independently of
each other:  saturated  flow, unsaturated flow, and
contaminant  transport. This resulted in:  (1) maxi-
mum  capability to simulate all combinations of
situations that could possibly be encountered,  (2)
generally applicable models  that could be applied
to  any ground-water system in which  the  basic
assumptions  were  satisfied or where the assump-
tions were not compromising, and (3) ease of modi-
fying or refining the models independently without
having to reformulate the entire system.
     The  model  development  effort  is presently
separated  into  three major categories of  models:
data models, hydraulic models, and water quality
(transport) models. Data models calculate, from a
minimum  number  of  field measurements,  input
characteristics required  for operation of  the hy-
draulic and water quality  models. The hydraulic
models predict the ground-water velocities or rate
of ground-water movement in  the unsaturated and
saturated  sediments. The water quality model com-
bines the ground-water  movement calculated  by
the hydraulic models with soil-waste reactions that
contaminants  would undergo during  a  traverse
through the soil to predict the contaminant move-
ment. In effect, the ground water  is a transporting
mechanism and can be simulated independently of
waste movement. A simplified schematic  showing
the  interrelationships between the  data, hydraulic
and transport models is shown in Figure 1. Table 1
lists  the models and their respective functions as
well  as input  data requirements  and output  in-
formation.  Each of these categories will be dis-
cussed in  the  following narrative.
 Fig. 1. Schematic  diagram showing interrelationships and
 data flow between the various categories of ground-water
 models.
 Data Models
      Data models include those developed to calcu-
 late  (from  a minimum number of field measure-
 ments)  input  characteristics  required  for  the
 operation  of  the  hydraulic  and  water quality
 models.  In the  initial  planning  stages of  the
 program, it was realized that it was not economical
 to  obtain  input  characteristic  data  using  the
 classical method of extensive field data collection
 programs. Therefore, models were developed which
 would  use  data that were available  or  could  be
 obtained economically, and from these, calculate
 the  desired model  input.  The  most  important
 model  that comes under this category is the Trans-
 missivity Iterative Routine.
      It  is  common  in ground-water systems for
 permeabilities  to  vary  considerably over  study
 areas;  thus,  the assumption  of a  homogeneous
 permeability  distribution  will  usually  result  in
 erroneous predictions  in changes in ground-water
 elevations  and  flow patterns.  Because of these
 variations in permeabilities, the cost of establishing
                                             183

-------
                Table 1.  List of Ground-Water Models, Functions, Input Data  Requirements and Outputs
         Models
         Function
           Input
          Output
Data Models:
Transmissivity Iterative
Routine

Hydraulic Characteristics
Routine

Sorption Transmissivity
Routine

Micro Sorption Routine
 Hydraulic Models:

 Partially-Saturated
 Transient (PST) Flow
 Model

 Variable Thickness
 Transient (VTT) Flow
 Model
Transport Models:

Macro-ion Transport
Model
Micro-ion Transport Model
Calculation  of model input
from  minimum  number of
field measurements.
Calculates transmissivity dis-
tributions.

Calculates permeability func-
tion (permeability vs. capillary
pressure relationship).
Calculates s o r p t i o n  coeffi-
cient distribution.

Determines  the  coefficients
for the Kd-macro-ion relation-
ship.
Simulation of subsurface wa-
ter movement.
Simulation of unsaturated
flow.
Simulation of saturated flow
(approximates  three - dimen-
sional systems).
Simulation  of subsurface
waste movement.
Simulates macro-ion m o v e -
ment in saturated and unsatu-
rated soils.
Simulation  of  micro-ion
(trace) movement in saturated
and unsaturated soils.
1. Ground - water  potentials;
2. Transmissivity  measure-
ments.
1. Desorption curve (moisture
content vs. capillary pressure);
2. Saturated permeability.
1. Sorption-transmissivity re-
lationship;
2. Transmissivity distribution.
1. Experimental data for spe-
cific soils and wastes.
1. Hydraulic characteristics;
2. Boundary conditions;
3. Initial conditions.

1. Transmissivity distribution;
2. Initial conditions;
3. Boundary  conditions:  in-
cludes quantities and location
of recharge calculated  by the
PST Model.
1. Cation exchange capacities;
2. Initial  concentrations  on
the soils;
3. Grou nd-water velocities
from   either  PST or VTT
models;
4. Boundary conditions.
1. Sorption  coefficient distri-
bution;*
2. Initial  concentrations  on
the soils;
3. Macro - ion concentrations
as  a  function of time and
space  from  the   macro - ion
model;
4. G round-water velocities
from   either  PST or VTT
models;
5. Boundary conditions.
Heterogeneous transmissivity.
Permeability function.
Sorption coefficient distribu-
tion.

Coefficients for the Kd-macro-
ion relationship.
Time-dependent  ground  wa-
ter: (1) potentials; (2) veloci-
ties; (3) flow rates; (4)  flow
paths; and (5) travel times.
Time-dependent  ground  wa-
ter: (1) potentials; (2) veloci-
ties; (3) flow paths; (4)  flow
rates; and (5) travel times.
Macro-ion  concentrations in
liquid  and solid phases as a
function of time and space.
Micro-ion  concentrations  in
liquid  and solid phases as a
function of time and space.
*  Input requirements vary depending on the situation being modeled; i.e., either unsaturated or saturated and the importance
of the macro-ion influence on micro-ion movement, etc.
184

-------
an observation well network and making the many
measurements  required to assure a reliability per-
meability distribution quickly becomes prohibitive.
To avoid  the prohibitive cost of a massive number
of field  measurements, a  computer routine  was
developed which allows calculation  of the perme-
ability distribution over study  areas using a mini-
mum number  of field observations. The method
developed, based on continuity of flow in a stream-
tube for a quasi-steady state ground-water  system,
utilizes the measured energy dissipation in a stream-
tube to calculate transmissivity and/or permeability
at any point within the tube.
     The  program was written for two-dimensional
systems where the transmissivity is assumed equal
to the product of the depth and the average perme-
ability  over  the vertical  section of the  aquifer.
Although the method could be extended to include
the third  dimension, the field data available or that
which  could  be obtained  economically does not
warrant such an extension at this time. Figure  2
shows the calculated transmissivity distribution for
the Hanford Reservation ground-water system.
                               HANFORD PROJECT
 Fig. 2. Calculated transmissibility distribution of the Han-
 ford ground-water flow system.
      Using  this transmissivity  distribution along
 with the ground-water potentials, the existing travel
 times and flow rates  from any site on the Hanford
Reservation can be calculated.  Figure  3 shows the
relative travel time from a reference line located on
a 50  square mile block of  the  Hanford ground-
water system. The  predicted three-pronged effect,
a result of high permeability channels, agrees very
well with measured observations of ground-water
contamination  in that area. The effects of hetero-
geneity on waste movement  are  explicitly shown
by  this three-pronged front. It would be impossible
to  simulate waste  movement accurately  in  the
Hanford system by assuming a homogeneous trans-
missivity distribution and  using a large dispersion
factor to mask the effects of heterogeneity. This is
why  it is  necessary to  have an  accurate  trans-
missivity distribution and why considerable effort
has been  expended  in  developing methods with
which  these data  can  be obtained economically.
The photograph in Figure 3 was taken off a com-
puter generated Cathode Ray Tube (CRT) display.
 Fig. 3. Ground-water movement pattern at time = t, origi-
 nating from a  reference line on a portion of the Hanford
 ground-water system.

      This program is presently  being extended to
 include transient systems.  In addition, it has been
 adapted  to  a  man-machine interactive computer
 system (which  is discussed later) so  that the errors
 in  interpreting ground-water  potentials  can  be
 minimized. This  involves  an iterative process of
 reinterpreting  ground-water  contours  from  the
 original  data and recalculating the  transmissivity
 distribution.  The  process  would  be  impractical
 using standard  computer input/output procedures.
 The method is called the Transmissivity  Iterative
 Routine.
      In modeling unsaturated flow, it is necessary
 to have a mathematical description of the hydraulic
 characteristics  of unsaturated soils. The hydraulic
 characteristics  consist  of  two  relationships:  (1)

                                               185

-------
 permeability versus capillary pressure (permeability
 function); and (2) moisture content versus capillary
 pressure (desorption curve). Obtaining the permea-
 bility function from laboratory  measurements is
 extremely time  consuming and expensive relative
 to  the number  of measurements  required  to
 accurately define a system. Methods developed for
 calculating the  permeability  function  from  the
 desorption curve  and the  saturated permeability
 have been modified  and used successfully.  Since
 the  desorption  curve and  saturated permeability
 data can be obtained without too much difficulty,
 this approach to establishing hydraulic  character-
 istics of unsaturated soils represents a considerable
 savings in time and money.
      The  micro-ion  (radionuclide) sorption coeffi-
 cient is a function of both the  soil  type and the
 macro-ion  concentrations.  The  Micro  Sorption
 Routine determines the coefficient for the relation-
 ship between the Kd and macro-ion concentrations
 using a two  factorial design regression technique.
 Experimental data  must  be  collected  for  each
 soil type and range of wastes encountered.
      On   large  systems,  such as Hanford which
 covers in  excess  of 400  square miles, it is im-
 practical  to  mount the field data collection pro-
 gram  necessary to  accurately  define  the spatial
N sorption coefficient (Kd) distribution. Therefore, a
 data model is being developed to calculate the Kd
 distribution for large ground-water systems utilizing
 a minimum  number of  field  measurements. The
 approach being  used is to relate  the Kd to the soil
 type through the permeability/transmissivity distri-
 bution. Since an  accurate transmissivity distribu-
 tion  can  be obtained using the  Transmissivity
 Iterative  Routine, a Kd distribution can then be
 calculated from these data. Preliminary results are
 very encouraging. To use the method, available soil
 samples  are  analyzed  to  obtain  the permeability
 versus Kd relationship. Once this  relationship  is
 established, the Kd  distribution can then be  calcu-
 lated from the transmissivity distribution.

 Hydraulic Models
      It was decided at the initiation of this program
 that the  most  practical  approach  to  modeling
 ground-water flow was to consider the unsaturated
 and saturated ground-water flow system separately
 rather than  as  a  combined system. This decision
 was made in consideration of the field data availa-
 ble and  the  cost  of obtaining additional data, the
 size of core memory in large computers presently
 available, and the computational time associated
 with use of large grid networks to obtain numerical
 solutions.  Considering the systems  separately re-
 sults in  having  to define the hydraulic  character-
istics  only at  sites where unsaturated flow is im-
portant (points  of  recharge or  injection). This
limits the  amount of data required on unsaturated
soils and for most problems is not compromising.
It is anticipated that the unsaturated model will be
used to calculate the quantity of flow entering the
ground-water system. The calculated inflow is then
input to the saturated model to predict the changes
in the ground-water surface elevations.
     The capillary conductivity approach was cho-
sen for modeling the unsaturated portion of the
ground-water  system. The model is based on con-
tinuity of flow into and out of an elemental volume
where the  permeability, hydraulic potential and
moisture  content are all a function  of capillary
pressure. The  model is valid for transient, heterog-
eneous, saturated and  unsaturated flow. The pri-
mary input requirements  to  this model are the
hydraulic   characteristics,   permeability - capillary
pressure,  and  moisture  content-capillary pressure
relationships.
     The   PST  computer program  will pres-
ently  work  in  one  and  two  dimensions, and
in  three   dimensions  for  axisymmetrical
problems.   The  program  can  be  extended  to
include the third dimension.  It  should  be  noted
that  while the  PST  Model  is  used to  describe
combined  saturated and unsaturated transient flow
above regional water tables, it also can be used to
solve combined saturated  and  unsaturated  flow
problems  which include the regional ground-water
system. The reason the PST Model is not used in
this manner is that a  model which  approximates
three-dimensional saturated flow (VTT Model), has
also been  developed.  As mentioned earlier, using
the two models significantly reduces the  amount of
input data and computer core memory required, as
opposed to the entire  system being modeled with
the PST Model.
     The   PST Model has  been  used to  simulate
hypothetical  buried  tank  and  pipeline ruptures.
 Figure  4  shows output  from  two  hypothetical
problems that were analyzed with the PST Model.
 Figure  4(a) shows the  moisture  pattern developed
 at 223 minutes  after  a simulated pipeline rupture.
 The degree of saturation in these Figures is propor-
 tional to  the  intensity of light on the  computer
 driven  CRT from which these photos were  taken.
 Therefore, white indicates saturation while black
 indicates   the  initial  soil  moisture.  The  various
 shades of gray indicate  gradation  of moisture
 between these two limits.
      Figure 4(b) shows the moisture pattern de-
 veloped at 16.55 hours after a simulated storage
 tank rupture. There are five  geologically distinct
 strata which are represented by the four horizontal,
  186

-------
Fig. 4(a). Moisture distribution for a 3 inch pipeline break
simulation  (time is indicated from start of problem, and
seal is in feet).
Fig. 4(b). Moisture distribution  resulting from a simulated
storage tank rupture.
Fig. 4. Computer generated displays showing output from
the  PST Model for two hypothetical problems that were
analyzed.
white lines. The total  distance between the tank
bottom  and  the water-table surface is approx-
imately  230  feet.  Results from  this model  are
encouraging since very little has  been published
relative  to  the numerical  solution  of the equa-
tion involved.  The numerical problems encountered
on  the  tank and pipeline break simulations were
severe because of the low initial saturation (approx^
imately  8% moisture  content),  the  high positive
heads and the  large variations in permeabilities.
     Studies of this nature are continuing in both
field applications and for model verification.  Model
verification  includes  simulation  of both one and
two-dimensional columns under precise laboratory
control utilizing nondestructive testing procedures.
As a result of the testing to date, it appears that the
model can  be used to  simulate any situation that
will be encountered on the  Hanford Reservation.
The above  simulations  demonstrate the utility of
the models in developing  management policies and
evaluating management  alternatives.
     For the saturated portion of the flow system,
a two-dimensional model  that approximates three-
dimensional  saturated,  transient  flow  was devel-
oped. The  model (VTT)  was  developed for  a
heterogeneous  system  where  the  permeability,
aquifer bottom  elevation, and storage coefficients
are a function  of space coordinates,  x and y.
Changes  in  transmissivities resulting from changes
in water  table  elevations are  factored  into  the
model.
     Comparative  figures are presented to provide
a quantitative understanding of the advances which
have been made in simulating saturated flow since
the inception of this  program. Figure  5  shows the
measured Hanford ground-water potential distribu-
tion  compared   with  the  potential  distribution
predicted by the  VTT  Model using a  homoge-
neous permeability distribution and the Hanford
ground-water system  boundary  conditions. As can
be  seen,  the match  in potentials  is  poor
over much  of the Reservation. In addition, waste
movement calculated from such a potential distri-
bution  would be highly  inaccurate relative to
spatial locations and concentrations since the calcu-
lated  flow paths and travel  times would compare
poorly  to  actual  values.  The  assumption  of a
homogeneous permeability  distribution  has  been
used in many areas because of the lack of methods
for economically  obtaining a representative distri-
bution.  This assumption  has also been considered
for past  models  of  the Hanford ground-water
system. However, since an accurate knowledge of
the direction and rate  of the ground-water move-
ment is  a  prerequisite  to  accurately  predicting
waste movements, a  more representative distribu-
tion is necessary.
     Figure 6 shows the same base map as in Figure
5  with  the VTT  Model  output superimposed. In
this  case a  highly variable  transmissivity distribu-
tion calculated from  the streamtube program was
used  (see Figure 2). As can  be  seen, the  match of
ground-water elevations  and  the  shape  of  the
ground-water contours  is very good. The resulting
prediction  of waste movement would  be consider-
ably more  accurate than in the first example and
represents  a significant step toward obtaining the
objectives of this development program. The effect
of grid  size and  other variables  on the match is
                                                                                                  187

-------
Fig. 5. Measured and calculated Hanford ground-water con-
tours—homogeneous permeability.
Fig. 6. Measured and calculated Hanford ground-water con-
tours-calculated transmissivity.


under investigation.  It is planned to use the Trans-
missivity  Iterative Routine along with additional
transmissivity measurements that have been made
to  further  improve the  calculated transmissivity

188
                                                     distribution.  With the new distribution, an even
                                                     better  match than  that  shown  in Figure 6 is
                                                     expected.
                                                          Output  from  the VTT  Model used  on an
                                                     idealized test case is graphically shown in Figure 7.
                                                     The  solution shows the computer generated transi-
                                                     ent  development of a mound that  would result
                                                     from discharge to a ground-water system.
                                                          INITIAL CONDITION
                                                              TIME  • 0
                             TRANSIENT SOLUTION
                               TIME  • 1  DAY
                                                         TRANSIENT SOLUTION         TRANSIENT SOLUTION
                                                           TIME • 3 DAYS              TIME • 13 DAYS
                                                     Fig. 7. Progressive change in ground-water surface resulting
                                                     from discharge of wastewater to the ground.
Water Quality Models
     Development of methods for predicting con-
taminant movement is based on the transport equa-
tion which incorporates fluid movement and con-
taminant reaction components into one interrelated
equation.  The transport equation is the analytical
tool with which the time-dependent, concentration
distribution of contaminants in saturated and un-
saturated soils can be predicted. In the case of the
nuclear industry, radionuclides of interest generally
occur in trace concentrations relative to naturally
occurring cations.  By definition, trace ion sorption
of a specific ion is independent of other trace ion
concentrations.  The trace ion  movement is influ-
enced, however, by the macro-ions in the system.
Therefore, two transport models have  been de-
veloped—one for describing the transient movement
of the macro-ions, and  the other for describing the
transient movement of the trace (micro)  ions. Both
models are valid in  saturated and unsaturated sys-
tems.  The  models  take  into  consideration the
influences  of convective ground-water movement,
hydrodynamic  dispersion,  molecular  dispersion,

-------
sorption and decay. For the sorption reaction used,
it  is assumed that equilibrium conditions between
ions in liquid and solid phases exist.
     Satisfactory numerical  solutions to the  con-
vective transport equation cannot be obtained using
standard  techniques  such  as those used  in  the
hydraulic models. The method  of characteristics
was  used to  obtain  a two-dimensional numerical
solution.  Both models  are  currently working  and
are undergoing further testings and refinement.
     The input data required for the water quality
models  are  the ground-water velocities (obtained
from the hydraulic  models), the  dispersion  and
diffusion coefficients (which  can be obtained from
laboratory and field tests) and the  sorption coeffi-
cient. Other than convective  transport, sorption is
the most important factor influencing waste move-
ment. The sorption coefficient (Kd) at any time or
spatial  location is  a  function  of  the  macro-ion
concentration (for Hanford these areMg2*,Ca2*, K",
and Na"), pH and soil sorptive characteristics.
     Considerable soil-waste reaction work has been
conducted to determine the relationship of the Kd
and   macro-ion  concentrations.  A  soil-sorption
model, which  solves the one-dimensional convective
transport equation  for  steady  flow  fields,  was
developed to  evaluate and  refine these relation-
ships. Na*,  K*, Ca*%  and Mg"  have been success-
fully handled by the use of ion exchange equations
that  have  been  developed.  The  system  can  be
extended to other solution  ions. Known carbonate
and  sulfate  precipitation and  association equilibria
are used to  predict  these  equilibria. The model has
also  been  used  to  evaluate  the  experimentally
determined  sorption coefficients.
     Since laboratory soil column  work has  been
conducted at Hanford  for  many years, the work
accomplished  under  this program  can be  readily
compared to the previous accomplishments. Figure
8 is  intended to graphically indicate  the progress
which has been made on radiocontaminant trans-
port through  research on this program. A column
experiment  was conducted using a fabricated waste
solution  (typical of those encountered at Hanford).
The triangles shown in Figure 8 are the 85Sr relative
concentrations measured  in the column effluent.
The  dashed  and  solid curves represent predicted
values using the soil-sorption model. For curve 1,
it is  assumed that the Kd is constant (the approach
used prior  to the inception  of  this development
program); i.e.  the Kd  is independent of the macro-
ion concentrations. Curve 2 represents the break-
through using a Kd that  is a function of macro-ion
concentrations. The Kd's  can  be  obtained through
either small  column or batch techniques. As can be
seen, significant errors in predicting radiocontami-
nant movement can occur  if the Kd is assumed
independent  of  macro-ion  concentrations.  This
comparison  shows  that the approach being used
does an excellent job of describing the chemistry of
the soil-waste interactions.
Fig. 8. Relative concentration of 85Sr from a Burbank sand
subsoil column as a function of effluent volume.
Model Application
     Concurrent  to  the  development  of  the inte-
grated model system  described  above, a unique
man-machine interactive computer system  has been
developed for use in application  of the models to
field problems. The objective of this system is to
provide the tools with which an engineer, who is
not a computer expert, can analyze waste manage-
ment alternatives through efficient  utilization of
the  models previously  described.  Figure 9 is a
photograph  of this system. From left to  right are
(1) graphic  digitizer, (2) CRT-light  pen,  (3) tele-
type, and (4) PDP-9 digital computer. The PDP-9
computer is being linked to a much larger Univac
1108 computer as shown in the schematic diagram
of the man-machine interactive computer system
in Figure 10. The dashed lines represent proposed
additions to the system.  A  high speed printer-
plotter  is presently being  added to the system.
Problems will normally be input to the computer
through one or more of the devices and transmitted
to the larger computer for calculation. The results
would then  be   transmitted back to  the smaller
system for analysis and output.
     The graphic  digitizer is an input device which
permits input of analog data (maps, charts, etc.)
directly to  storage in  the PDP-9.  This results in
extremely rapid and accurate input of information.
Input data changes can also be readily input with
this  system  without complicated revision ot the
originally stored information.
                                            189

-------
Fig. 9. Photograph  of  the  man-machine  interactive com-
puter system.

       COMPUTER NETWORK AND OAT* DISPLAY SYSTEM
                           UUK cowutei STSTEH
 Fig. 10. Schematic diagram of the man-machine interactive
 computer system.

      The CRT  is a device which displays problems
 visually, and utilizing the light pen. the engineer-
 operator  interacts directly and in real-time with a
 problem solution and analysis. The power of such a
 system to the manager or engineer/user is difficult
 to  convey on paper. The problems are set up and
 input in rapid  and accurate fashion. Solutions are
 obtained  on a  rapid turn around  basis  and dis-
 played for the user. The user immediately is able to
 evaluate the results, make changes and ask for new
 solutions. Of particular importance is the ability to
 inject the experience and knowledge  of the user
 into  the  problem  solution.  When an acceptable
 solution is achieved, the results can be  printed on a
 line printer or plotter. This system development is
 being carried out in conjunction with the model
 development and  will be fully operational at the
 same time as the models themselves. The system is
 already serving an important function in the model
 development efforv .as mentioned earlier.
 190
                                                        Fig. 11 (a). Reference map of the Hariford reservation.
    Fig. 11 (b). Contours of the Hanford water table.
    Fig. 11 (c). Flow paths and associated travel times.

Fig. 11. Computer generated displays associated with the
calculation of ground-water travel times.

-------
     Figures 11 through 13 show some additional
photographs of displays generated on  the CRT
display unit.  Figure  11 shows the sequence  the
engineer uses  to instantaneously calculate ground-
water contours, flow paths and travel times. Figure
11 (a)  shows the  reference display of  the  reserva-
tion.  In  the contour  mode the operator can plot
contours  of either measured  or predicted  water
tables  which are  digitally stored on the disc unit
(Figure  ll(b)). The  values of the  contours that
are plotted  are input  through the keyboard of the
display unit. Contours can readily be added  or de-
leted  by the operator. In the travel time mode, the
starting location of the flow path is input through
the digitizer. Then the flow path is instantaneously
calculated and displayed along with the associated
travel time (Figure 1 l(c)). The travel time displayed
on the CRT is associated  with the last flow path
that was  selected. With this program,  the engineer
can  rapidly assess  the impact  of  management
alternatives  on the water table elevation, flow paths
and travel times.
 Fig. 12. Stereogram of an isometric view of the  Hanford
 ground-water surface.
Fig. 13. Stereogram of the Hanford ground-water surface.

     Figures  12 and  13  are  self-explanatory and
represent another form of output which is available.
These photographs were  taken from a memory
scope display.

             ACKNOWLEDGMENTS
     The author would like  to acknowledge  the
engineers and scientists who  were responsible for
the development of the integrated model and the
display systems: A. E. Reisenauer, hydraulic mod-
els; R. C. Routson and R. J. Serne, soil-waste inter-
relationships; A.  Brandstetter  and  R. G. Baca,
transport models;  D.  R. Friedrichs, L. E. Addison
and H. P. Foote, data models and display programs;
L. H. Gerhardstein, system  software; and R.  D.
Mudd, general  assistance.  Although listed  by  their
major area of contribution, there was considerable
integration  and combined effort  in each of the
categories  by  the individuals  listed above. The
author would  also like to acknowledge the con-
tinued support of the program by the  Atlantic
Richfield   Hanford  Company,  especially R.  E.
Isaacson, D. J. Brown and M.  D. Veatch.
                                            DISCUSSION
The following questions were answered by  D. B.
Cearlock  after delivering  his talk entitled  "A
Systems Approach to Management of the Hanford
Ground-Water Basin."

Q.  Now that you have the models working, sup-
pose  the power company came to you to do an
environmental  impact analysis of waste discharge
from  a planned nuclear  reactor plant.  Assuming
that only routine soil geological maps were availa-
ble for the area, how much time, about how many
man-years effort, would be required to apply your
analysis to this situation?
A.  It  is difficult to  estimate the effort required
for a job without a more detailed  elaboration of
the problem. The cost is directly  associated with
the complexity and size of the ground-water basin,
the various components of the system of models to
be used and the limitation of the assumptions used
in the models. If simplifying assumptions can be
made and only a portion of the system of models
used, application costs would be considerably less
than if it was necessary to use the entire system as
described in  the  paper. As  the  complexity of the
system and the accuracy requirements increase, the
associated costs will increase.

Q.  You  indicated  you have data to verify watei
movement and travel time. \Youid y>ii elaborate on
those data? And allied with that  is this question.
What is your evidence that the impervious strata in
one of your slides arc really impervious?
A. A water quality surveillance program, which has
been  conducted  since the  inception of the AEC
activities, has provided the  data for verifying the
water movement patterns and the travel times. As
                                            191

-------
 mentioned  in  the paper, the agreement between
 observed and measured movement patterns is good.
 However, these verification results are preliminary,
 and detailed laboratory and field verification stud-
 ies  are  presently being conducted.  Emphasis  to
 date  has   been  directed  predominantly toward
 developing the complete system of models.
      Both  geologic  and hydrologic  data indicate
 the existence of an impervious strata over much of
 the project. Geologic data indicate the presence of
 tight  silt-clay  layers although  the continuity  of
 these layers has not been established.

 Q.  Here's another part of a question. Where is the
 natural discharge of the lower aquifer?
 A.  This is being researched by ARHCO and USGS.

 Q.  Can your program solve other kinds or types of
 tank failure in place of the rather unlikely  sudden
 removal of the whole tank bottom and would these
 give significantly different results?
 A.  Other  types  of tanks and tank failures can be
 modeled.  The models were  designed to be com-
 pletely flexible  so  that, in general, any type of
 tank leak  or tank failure could be modeled. There
 are about 25  different shapes of  boundaries that
  can be used in describing a tank facility.
^      After analyzing the results of the  rupture of
  the entire bottom of a tank, we believe a small leak
  would result in faster penetration to the water table
  and higher concentrations of nuclides in the ground
  water if the leak were allowed to persist; however,
  this  is unsubstantiated  by  either  modeling  or
  observations.
  Q. Does your, or did  the pumping well  cause any
  subsidences in the area of the plant and the storage
  tank and, if so, what measures were taken to cor-
  rect it?
  A. The pumping wells we refer to  are  test wells
  used for transmissibility determinations and are not
  water-supply  wells; hence, one  should  not antici-
  pate subsidence.

  Q.  What methods would you use or will you use to
  solve  the  differential equations for  hydraulic dis-
  persions in the natural system?
  A. The transport equation  is  essentially  a first
  order equation since in ground-water systems, con-
  vection is usually more significant than  dispersion.
  Therefore, use of numerical techniques  for second
  order equations can result in numerical  dispersion.
  To minimize numerical dispersion we have adopted
  a hybrid approach first developed in  the petroleum
  industry.  It  involves  using both  Lagrangian and
  Eulerian systems. The convective transport is solved
  in the Lagrangian system and dispersion is solved in
a Eulerian  system.  We  have  tested the  method
against analytical solutions and the agreement was
good.

Q. Are values  of porosity  and dispersion coeffi-
cients  required for  your models  and,  if so,  how
were they determined and  what is the sensitivity
of the model for variations in these parameters?
A.  Both  porosity and  dispersion coefficients are
required in the model. We are presently using dis-
persion coefficients  measured in  laboratory soil
columns.  At specific sites on the reservation, the
effective  porosity has been estimated from travel
time data. We have not conducted  sensitivity analy-
ses with respect  to porosity and dispersion coeffi-
cients.
Q.  Do  you believe  that our presently  available
methods  of obtaining reliable field data justify the
utilization of such elaborate mathematical models?
A. Yes. Otherwise we would not be developing the
models.  The  models are being used to  identify
where additional data are required to improve the
understanding of the aquifer. Use of the models in
this way minimizes the collection of data that are
 not directly useful in  understanding the ground-
water  system. This approach has substantial advan-
 tages  over the  classical  approach of  conducting
extensive  field  measurement programs  prior  to
modeling the system. In  addition, the preliminary
verification results presented in the paper  are very-
encouraging.

Q. Can  your analysis  which  now  considers the
effect  of ion exchange on  ion transport also in-
corporate the effects of such processes as precipita-
tion and oxidation?
A. Yes. The model presently handles some precipi-
tation reactions. For these reactions it is assumed
that the precipitate does not alter the permeability.

Q.  Isn't the porosity important for travel time? It
appears that only transmissivity was used.
A.  Yes, porosity is  important. We have been using
an  effective porosity estimated from  field measure-
ments and tests. We have also assumed the effective
porosity  is  constant over  the entire project because
of the lack of data. The effective porosity has been
observed  to be  considerably less than  the  bulk
porosity  in certain  areas. However, these data are
limited.
Q. How much does the porosity vary?
A. The   bulk  porosity  varies approximately be-
tween 30 and  45 percent.  In contrast, the calcu-
lated permeability varies by 4 to 5 orders of magni-
tude.  We  do  not  have  the  data  available  to
determine how much the  effective porosity varies.
  192

-------
Salty  Ground  Water  and  Meteoric  Flushing  of
                                                                              a
Contaminated  Aquifers  in   West  Virginia
by Benton M. Wilmoth
                   ABSTRACT
     Salty  ground  water  is commonly encountered at
relatively  shallow depths of 100 to 300 feet beneath the
major stream channels in the western half of West Virginia.
Because  of the  wide distribution of salty  ground water
and  connate brine  at  various  depths,  it  is difficult to
distinguish  natural  contamination  from that caused  by
subsurface industrial activities. Natural changes in quality
apparently are minor. The available historical data indicate
no large-scale natural variations in salt content during the
period of  record. Histories of some water well developments
show unnatural  large-scale  increases in  salt content from
various industrial activities that affect the fresh water zones.
Some records also reveal decreases in salt content after the
source of the salt was  eliminated or after the subsurface
activity responsible for artificial migration of the salt water
was stopped.
     Artesian brine contaminated a fresh water aquifer in
Fayette County. Chloride content changed from 53 mg/1 to
more than 1,900 mg/1 in a period of 5'/2 years. When pump-
ing was stopped, chloride content decreased to 55 mg/1 in
10 years.
      Heavy pumping of well fields in  Charleston during
1930 to  1956 accelerated migration upward of salt water.
Chloride  content increased from less than 100 mg/1 to more
than 300 mg/1 in some wells and to more than 1,000 mg/1
at individual wells. Pumpage has declined greatly since 1956
and chloride content has decreased below 200 mg/1 at some
of the contaminated wells.
      In an oil field  of Kanawha County, a water well was
contaminated by salt water accelerated by subsurface activi-
ties. Chloride content increased  from less than 100 mg/1 to
more than 2,900 mg/1 within 2  months. After the oil-field
activity was curtailed, chloride content decreased to 190
mg/1 in  about 2'/i years. Road  salt piles contaminated  a
carbonate aquifer in Monroe County. Chloride concentra-
tions in  wells located 1,500 feet from the  piles increased
from 185 mg/1 to 1,000  mg/1 in 5 years. The greatest change
was  1,000 mg/1  in 1969 to 7,200 mg/1 in 1970 when the
salt storage area was enlarged. All salt piles were removed in
late  1970 and within 2  months chloride content decreased
to 188 mg/1.

      One  of  the  most  difficult problems  in sub-
surface  hydrology is to predict the movement of a
contaminant in ground  water when concentrations
are  decreased  by dilution. The problem is  further
complicated if the contaminant, such as salt water,
      Presented  at the National Ground Water  Quality
 Symposium, Denver, Colorado, August 25-27, 1971.
      ^Geologist, Environmental Protection Agency,  303
 Methodist Building, Wheeling, West Virginia 26003.
also occurs naturally in adjacent aquifers at a wide
range  of concentrations,  velocities, and depths.
Consequently to determine when salt  water has
occurred naturally and when it is the result of sub-
surface industrial activities is a formidable  task.
Predictions  of  future  subsurface  conditions are
likewise extremely difficult.
     Sedimentary rocks  of Pennsylvanian and Per-
mian  (?) ages of the western half of the State form
a complex, alternating series of conglomerate, shale,
siltstone, limestone, and coal  strata. The rocks are
gently folded  along subparallel axes  that trend
generally northeastward.  These flexures tend to
obscure  the general regional dip of the rocks to the
northwest. Because  of the folding and subsequent
erosion,  some  of the older brine-bearing rocks are
close to  the land surface along the axial centers of
major anticlines. Although ground water has been
obtained at depths of more than a mile, test holes
usually  encounter  little water below 2,500 feet.
This  is about the top of the zone  of rock flowage
where the weight  of the overlying rock column
tends to close pore space.
      The majority of the shallow rocks that are
weathered and fractured at depths ranging from 25
to  300 feet  yield ground water. The major devel-
oped  aquifers  and  most of  the  ground-water
pumpage are  within  this zone. Intergranular or
primary  permeability and fracture or  secondary
permeability  both contribute to  shallow ground-
water occurrence but little evidence is available to
indicate  the  predominant form.  Sandstone and
conglomerate usually have the highest permeability
because  either the primary   permeability is pre-
dominant  or  the  secondary  permeability  is de-
veloped  to  a larger degree.  These are  the pre-
dominant  water-bearing  rocks. Highly  fractured
sandy shale  within 100 to 300 feet of the land
surface can also yield moderate amounts of ground
water. Minor beds of thin nodular limestone, fire
clay, and coal are aquifers in local areas only.
      Large areal exposures of limestone and dolo-
mite  in  the eastern part of the State are excellent
major aquifers.  In general, throughout  the  State,
aquifers that are considered practical at the present
time  for development lie within about 500 feet of
                                             193

-------
 the  land  surface; however,  the average depth of
 existing wells is still only about  125 feet.
     Meteoric fresh ground water moves by gravity
 from upland intake recharge areas through fractures
 and interstitial openings in a  very complex flow
 system. The configuration of flow through various
 kinds  of  stratified rocks with complicated vertical
 and lateral  hydrologic boundaries  can  only be
 generalized.  The ground water  may remain only a
 few days in the shallow rocks before it is discharged
 as  springs or  seepage; or it  may recharge  deep
 artesian reservoirs and remain for many thousands
 of  years  becoming  more  mineralized  with  time.
 The velocity of flow may range from a few inches
 to several hundred feet per year.
     Although some ground water in the outcrop
 areas of aquifers is unconfined, artesian conditions
 exist below where the water becomes confined and
 is known to exist at depths of  more than 2,000
 feet. The water-bearing rocks form a series of aqui-
 fer systems, each composed of several hydraulically
 connected beds. The amount of hydraulic connec-
 tion ranges from direct or free to very little, due to
 relatively impermeable intervening formations.
      If the stream channels of  the western half of
 the State act as sumps for the  natural discharge of
^both fresh and salty ground water, the interface or
 top of the salty ground water  apparently tends to
 "cone  up"  toward  the stream  and  then lie at
 successively  greater depths with increasing distance
 away  from the valley beneath  the hills  and ridges.
      The saturated,  unconsolidated  alluvium bor-
 dering the major river valleys also contains salt
 water in  certain areas. Segments of former stream
 channels  in  the bedrock  of present major valley
 floors  are filled with highly permeable sand and
 gravel  of Pleistocene  age. These sand-and-gravel-
 filled  depressions along the Ohio River Valley in
 Cabell, Tyler, Brooke, and  Hancock Counties ap-
 parently are intercepted by the  salt water migrating
 from the underlying bedrock.
      Usually, the shallow saturated  rocks contain
 the fresh meteoric ground water,  and the  rocks
 below contain the salty ground water. The shallow
 salty ground water is essentially a zone of diffusion
 or mixing between the fresh ground water and the
 deep  connate brines. These fluids  generally  cut
 across geologic  contacts  in response  to natural
 hydraulic conditions or  to the changes in condi-
 tions imposed by subsurface industrial activities.
      Chemical character of fresh ground water is
 controlled primarily by the chemical composition
 of the soluble rock minerals of  the aquifers. Conse-
 quently,  both local  and regional lithologic differ-
 ences  can produce  differences  in  the  chemical
 quality of ground water. Fresh ground water con-
tains  less  than  1,000 mg/1  of dissolved  solids,
whereas  shallow salty ground water usually  con-
tains 1,000 to 3,000 mg/1 of dissolved solids. Fresh
ground  water  from  shallow  bedrock  aquifers is
chemically suitable for most domestic  uses. The
water is mostly of the calcium, sodium, or magnesi-
um bicarbonate type and is alkaline,  but the en-
croachment of brine  into the fresh-water aquifers
changes the water to a sodium chloride type.
     In  most areas of the  western half of the State,
the salty ground water occurs 100 to 300 feet be-
low the altitude of the major  stream channels. The
chloride concentration of the ground  water  con-
tinues to increase with depth  as the flushing effect
of  the  circulating  fresh  meteoric ground water
diminishes. At depths of 1,000 to 5,000 feet, highly
concentrated brines containing 30,000 to 100,000
mg/1  of total  dissolved   solids  are  encountered
(Price, Hare, McCue, and Hoskins, 1937, pp. 28-32).
These brines were derived from sea water trapped
in the original  rock-forming  sediments.  At these
depths,  the natural migration is extremely slow.
The various  processes of chemical  concentration
and  modification  of the sea water are not com-
pletely understood.
     In  totally undeveloped areas, where there has
been no deep drilling, the  natural vertical migration
of salt water apparently is extremely slow. Perhaps
any  natural  changes  in the quality  of both fresh
and  salty  ground water  in these  areas  are minor
over several hundred  years. The few historical data
available for specific aquifer areas indicate no large-
scale natural variations in salt content during the
period of record (Wilmoth, 1970). Much salt-water
migration into shallow rocks, therefore, is a natural
phenomenon  from a very long period  of inter-
formational seepage of artesian brine  out of deep
formations. Apparently, this migration has occurred
over several million years of geologic time.
     Salt water seeps  and  springs in several counties
were used by the Indians and the pioneer settlers to
obtain  salt.  The best  known occurrence  of  salt
water  discharged  as natural  springs  is  in  the
Kanawha River valley near the mouth of Campbell's
Creek at Maiden, Kanawha County. Here a major
geologic structure,  the Warfield anticline,  caused
geologically older brine-bearing rocks to crop out.
Historical records (Summers, 1935) show that these
salt water  springs were used by the  Indians in the
1700's.  Archeological evidence indicates they used
brine as early as the year 1000, but they may have
obtained brine or salt in the general area for more
than 10,000 years  (Broyles, Betty, personal com-
munication). The early settlers in Kanawha River
valley improved the  springs and later made crude
drilling  explorations  in the area by means of tilt
 194

-------
pole rigs. Deep drilling methods and  equipment
were  improved  before the  Civil War,  and large
amounts of salt were produced in the valley (Howe,
1845). Most of the brine was obtained from deep
wells  tapping the Pottsville Group at depths  of
several hundred feet.
     All  of  the  conditions which made  fresh
ground-water aquifers vulnerable to the encroach-
ment  of salt water  are not known. The best in-
formation is obtained from records at fresh water
well developments which experienced salt water
encroachment after many years of successful opera-
tion. While  the subsurface environment of these
areas cannot be described accurately, it is possible
to  define  the  general  occurrence of  the salty
ground water and its  relation to the fresh ground
water.  In some areas, deep drilling and related sub-
surface industrial activities have merely accelerated
the vertical  and lateral migration of natural forma-
tion fluids.  Consequently, these fluids  are  now
present  in  certain   shallow  aquifers  in greater
quantity and concentration than under the original
natural conditions. In the past  few decades rapid,
unnatural,  salt  water migration toward  the  land
surface has  been mostly  by vertical leakage along
hundreds of unplugged wells and test bores. These
holes were put down during exploration for brine,
oil, gas, water, and coal. Commonly, the holes were
abandoned  uncased  and improperly plugged. The
formation fluids of brine and related hydrocarbons
flowed up the  holes  in  response  to  the artesian
pressure  under  which they  were confined. These
essentially  open holes can act as nearly infinite
vertical rock permeability and  the natural protec-
tion from  rapid fluid migration offered by thick
sequences of shales and  other  confining strata is
lost. In many areas,  salt water under artesian pres-
sure has  migrated into developed fresh water aqui-
fers when casing was removed from an abandoned
deep well.
     Because of the  extensive distribution of brine
and salty ground water at a wide range of depths in
the western part of  the  State, it  is difficult to
determine  how much of  the  salt water  occurs
naturally and how much has been accelerated arti-
ficially by subsurface industrial activities. Salt water
can leak  and build up for many decades immediate-
ly below a fresh-water aquifer and still not pose a
problem  to shallow  drilled  domestic water wells
pumped  at  low rates. However, waste brine  dis-
posal,  water flooding, repressuring, or hydraulic
fracturing can trigger the salt water encroachment
of  these fresh water  zones.  Solutions to and  re-
medial actions for such  contamination problems
become  more  difficult with greater distances of
aquifer area and longer periods of time involved.
     Most areas of consolidated sedimentary rocks
are underlain  by two or more aquifers in which
water levels are  of  quite different elevation. As a
result,  some  natural  seepage  or leakage  occurs
between  the  various  aquifers. A lower  artesian
pressure in an aquifer indicates that  ground water
from an overlying or underlying  aquifer of higher
pressure may move  into it through confining beds.
The direction of  this  natural  interformational
leakage can  also  be  reversed by  pumping  and
changing of heads  in the  aquifers.  High-capacity
pumping from fresh  water zones may lower the
head in the rocks to levels below the artesian head
of the  salty ground water. The salt water may then
migrate into the fresh water aquifer if the zones are
hydraulically connected.
     Records  of ground-water quality in the bed-
rock aquifers at  Charleston, Kanawha County indi-
cate that  withdrawal of large amounts of  fresh
water  was a significant  factor in the accelerated
encroachment of salt  water. From 1930 to  1956
pumpage of ground water was heavy, and salt water
encroachment was  accelerated  (Figure  1).  The
maximum  withdrawals occurred during 1940 to
1950 (Wilmoth,  1965). Chloride content of ground
water  increased  from  less  than 100'mg/l to more
than 300 mg/1 in several well fields and to  more
than   1,000  mg/1  at individual  wells (Wilmoth,
1970). Pumpage gradually declined from 1956 and
by 1966  was extremely small with  only periodic
use of one  or two wells. Chloride content of ground
water  at some of the contaminated wells had de-
creased below 200 mg/1 by 1970  (Figure 2). This is
about  14 years from the last year of heavy ground-
water pumpage.
     Extensive areas of fresh ground-water aquifers
have  been contaminated by vertical migration of
artesian salty ground water into fresh water through
unplugged test  bores and improperly  abandoned
wells.  In  Fayette County, the chloride content of
ground water in a fresh water sandstone aquifer of
the Pottsville Group  increased from 53 mg/1 to
more than 1,900 mg/1 within 5Vz years (Figure 3).
Pumpage   in  the aquifer was  essentially stopped
when  the  principal water  well was  shut down be-
cause  of  the high salt  content. Pumping of the
well apparently helped to accelerate the contamina-
tion, because within about 10 years natural mete-
oric flushing  had reduced the chloride content to
55 mg/1 in part of the aquifer.
     In an area of Kanawha County the secondary
recovery of oil by water flooding has resulted in a
higher piezometric  level  of the brines in rock units
below the  fresh water aquifers. This has caused an
accelerated migration of the brine upward into the
fresh water.
                                             195

-------
600*-
400'-
                                   rWater well fields in Charleston
                                                                                              1000 feet  -
                                                         Average depressed piezornetric surface 1950
      Kanawho
      River
                                                 Explanation
        Clay and silt        Sand and gravel          Shale              Sandstone
Fig. 1. Changes in elevation of salt water interface, Charleston, Kanawha County.
                                                                                             prni
                                                                                             i	i	i
                                                                                           Limestone
  2OOO
   ISOO —
  1000 —
   soo—
     ~>9ia    I92C     1930     B4O     1950     I960   S7Q
 Fig. 2. Changes  in chloride  in  ground  water from over-
 development of aquifer, Charleston, Kanawha County.
Fig, 3. Changes in chloride in ground water from open-hole
migration of artesian brine, Fayette County.
     Unnatural  fluctuations  in  chloride  content
were observed  in  ground water from several wells
in this area. The fluctuations in quality apparently
are related to the rate and pressure of brine  injec-
tion. In one domestic water well supply the natural
chloride content in 1967 was 32 mg/1. Shortly after
secondary- recovery of oil was started the  chloride
concentrations began to increase. Within 9 months
the chloride content  had  increased  to  250 mg/1,
and within 11 months the concentration was 1,140
mg/1  (Figure 4).  After this maximum  chloride
level  was  reached,  the  concentrations began an
uneven but rather steady decline. After 2l/2 years of
general  decline the chloride content was 450 mg/1
early  in 1971. Apparently changes were  made in
the repressuring  operations in the oil  field. Then
 196

-------
v iOOO —
  500 —
         1967      1968      1969      1970      1971
 Fig. 4. Changes in chloride in ground water from repressur-
 ing of oil field, Kanawha County.
natural meteoric flushing of the shallow aquifers
helped to remove enough of the chloride for the
wells to be utilized.
     Oil field  operations also caused salt contami-
nation of a  fresh  water  aquifer near  Wallace,
Kanawha  County. In late 1967 the natural chloride
content of  the aquifer was 100 mg/1.  Within 2
months after  the start of the subsurface oil  well
operations,  the  chloride  content reached 2,950
mg/1 (Figure 5). The oil well operation was altered
and chloride concentration began to decrease rapid-
ly.  Within 2  months  the concentration  was 950
 ^
mg/1. During the next 2'/2 years the chloride content
declined to 190 mg/1.
      In the oil fields of Roane  County, pressure
injection disposal  of produced waste brine into the
   3000
   2SQO
   ?000h-
   1500
   iOOO
    500I—
           1967        1968        1969        «70
 Fig. 5. Changes in chloride  in ground water from oil well
 operations, Kanawha County.
Salt  Sandstone of the Pottsville Group has raised
the piezometric level of  the natural brine in  this
formation.  Because  the  overlying shallow fresh
water  zones are hydraulically  connected to  the
injection zone, there has  been extensive damage to
the fresh  ground water.  In some areas, brine con-
tamination  has  been rapid with  chloride content
increasing  from less than 25  mg/1 to more than
1,500  mg/1 within several months. Unnatural cyclic
changes in quality  have occurred  in many private
ground-water supplies. In one part of a fresh water
aquifer, the chloride concentration of less than 50
mg/1  increased  to  1,150 mg/1  in a period of 19
months (Figure 6).  After the  subsurface oil-field
brine disposal operations were altered, the chloride
content declined to  68 mg/1 within 17 months.
 Fig. 6. Changes in chloride from pressure injection of brine,
 Roane County.

      Soluble chemicals such as highway deicing salt
 are causing the deterioration of ground-water quali-
 ty  in certain shallow fresh water aquifers in the
 State.  Chemical analyses show  unnatural  cyclic
 changes in chloride content. Although the aquifer
 areas of deteriorated water quality are local, re-
 ported incidents of such contamination are increas-
 ing every year.
      The amount of salt used annually for ice and
 snow control on  highways continues to  increase.
 This  highly  soluble  material is now being recog-
 nized as a  contaminant that  easily migrates  into
 shallow fresh water aquifers with natural recharge.
 In  several States,  water quality control specialists
 are  observing possible  cumulative effects of the
 salt load on  the hydrologic system (Walker, 1971).
 Apparently  most of the salt is transported during
 thaw periods by runoff to the nearest stream and is
 progressively diluted downstream. Chloride concen-
                                              197

-------
trations of more than 10,000 mg/1 have been ob-
served  in  the  runoff from heavily salted areas of
Grant and Raleigh Counties. Salt carried by natural
recharge may  require several months or years to
percolate to aquifers and move laterally to pumping
water wells. Although the actual risk of contamina-
tion from salt spreading is not known at this time,
it is estimated that more than half of all the water
wells in the State are located within 100 to  500
feet of a road that receives some salt treatment.
     Shallow fresh water aquifers in carbonate rock
terrane are developed extensively for rural farms
and businesses  in the  eastern  part  of  the  State.
The principal source of recharge is precipitation
which enters  the  aquifer via solution crevices and
sinkhole depressions. The rate of travel from re-
charge  areas to  water wells may be rapid with very
little filtration  occurring. Potential contamination
of  such aquifers should be suspected it the ground-
water turbidity  increases after heavy rainfall.
     An  unusual ease  of  salt contamination ot
fresh ground water  from the land surface occurred
near Union, Monroe County. The history of water
quality in this  area illustrates the high contamina-
tion risk of shallow carbonate aquifers. There is no
natural near-surface salty ground water in the area,
nor any  gas  wells  or  other  deep  drilling. Partial
chemical  analyses indicate that the natural chloride
content of the fresh  ground  water in the  area
averages about 25 mg/1.
     Several hundred tons of rock salt stored di-
rectly  on *he land  surface were  also unprotected
from precipitation and runoff.  Water wells located
about  1,500  feet laterally and downgradient were
subsequently  contaminated with such high concen-
trations of chloride that the water was unfit  for
use. The  chloride concentration in one water  well
steadily increased to 1,000 mg/1 in about  5 years
(Figure 7).  The  period of  greatest change  was
 1,000  mg/1 in  1969 to a high of 7,200 mg/1 in
 1970. The wells in  this area are 73 to 79 feet deep
and tap highly  permeable fractured and solutioned
carbonate aquifers.  The  natural  gradient of the
water table was from the salt storage area toward
the wells. The  contaminated ground water moved
downgradient through the aquifer toward points of
natural discharge. Interception of the flow by the
pumping  wells  resulted in extreme degradation of
several supplies. In  late  1970, the salt was removed
from the ground and stored in a protective building.
During the  2  months  following  the  salt  storage
removal, chloride concentrations  in the monitored
contaminated well decreased to 188 mg/1.
     In Grant County, an investigation was made of
salt contamination  of a  shale aquifer tapped by  a
drilled  well  160  feet deep.  Apparently the  con-
                                                      8000
            i965                        197O
Fig. 7. Road salt contamination of ground water, Monroe
County.
tamination in this well was from the  spent salty
backwash from a zeolite water softener, which dis-
charged into a dry sump located upgradient from
the  water  well.  Natural  ground-water  recharge
carried the salt downgradient toward the  pumping
well. The natural  chloride content of  the ground
water in  this aquifer was determined in 1962 to be
9 mg/1. Within  16 months  after the zeolite water
treatment unit was installed, chloride  content  had
increased to 580  mg/1  (Figure 8).  Early  in  1970,
some 3 years after the brine was periodically dis-
charged  to  the sump, the  chloride content  had
reached  1,650 mg/1.  Early  in  1970 the treatment
unit was shut down and brine disposal was stopped
for several months. Within  a few weeks,  the  rate
of change in chloride  content was decreased sig-
nificantly. From January 1971 to May 1971. the
chloride  content actually declined from 1,712 mg/1
to 1,650 mg/1.
     Because the records of some salt-contaminated
well developments show decreases in  salt content
after the source of the  salt was eliminated, there is
 2000
      1962          1965                  1970
 Fig, 8. Contamination of ground water from brine disposal,
 Grant County.
 198

-------
reason to consider remedial  work  in  certain un-
naturally contaminated aquifers.
     If the  abandoned wells  and test bores could
be properly plugged  and cemented,  the source of
the salt water should  be effectively stopped. There-
after, the natural meteoric ground-water recharge
to the fresh water zones should eventually flush out
the salt water. Then pumping of the  water  wells
tapping the aquifer should actually  accelerate the
flushing action.
     Because  the ground-water aquifer is not seen
at the land surface,  much pollution often occurs
before it  is known.  The  public  is  often unaware
that certain industrial activities or incorrect pump-
ing can pollute aquifers. The single most important
aspect of ground-water  quality  control is the pre-
vention of  contamination rather than  attempts to
correct problems after they occur. Public education
for constant surveillance of activities  that can  result
in contamination is certainly an important need to
be considered.
                 REFERENCES
Howe,  Henry.  1845. Historical collections of Virginia.
     Babcock & Co., Charleston, South Carolina.
Price, P. H., C.  E. Hare, J. B. McCue, and H. A. Hoskins.
     1937.  Salt brines  of  West  Virginia.  West Virginia
     Geological and Economic Survey, 203 pp.
Summers, G. W.  1935. Pages from the  past. Summers
     Publishing Co., Charleston, West Virginia.
Walker, W. H. 1971. Limiting highway salt pollution of area
     water supplies. Rural and Urban Roads.
Wilmoth, B. M.  1965. Natural  equilibrium in ground-water
     storage re-established  at Charleston,  West Virginia.
     Proc. West Virginia Academy Science, v. 37, pp. 167-
     173.
Wilmoth, B. M. 1970. Occurrence of shallow salty ground
     water in West Virginia. Proc. West Virginia Academy
     Science, v. 42.
                                             DISCUSSION
The following questions were answered by Benton
M. Wilmoth after delivering his talk entitled "Salty
Ground  Water and Meteoric  Flushing of Contami-
nated Aquifers in West Virginia."
Q. In  the example  where  the  water softener dis-
charge was recycled, what was the original hardness
of the water?
A. I don't remember exactly, but it was  less than
200 mg/1. The well at  this  facility  taps a shale
aquifer.
Q. Does West Virginia have a regulatory agency at
present to insure the adequate construction of oil
wells and brine pits  for  the protection of potable
ground and surface water supplies, and, if not, why
not?
A. Yes,  the  State has the Division of Oil and Gas
that regulates oil and gas well construction, opera-
tion, and abandonment. The State Division of Water
Resources is charged with protection of the water
resources.
Q. / think each of us from our own States recog-
nizes that we have a legal authority, but sometimes
we don't have the enforcement personnel to do the
job the law says we can do. Here's a question. Why
were your brine  heads  converted  to fresh water
equivalents?
A. This  was  done for diagrammatic purposes to
illustrate how the head  of the salty ground water
does override the head  of the fresh water zones
under certain  hydrogeologic conditions.
 Q. In one of your slides showing the close relation-
 ship of oil and saline shallow ground water, was the
 oil production  from the  shallow  Pennsylvanian
 sands or from the deeper-zones?
 A. Oil production was from the deeper Big Injun
 formation of Mississippian age. Some abandoned
 holes must  be  hydraulically  connected  to  the
 shallower  Salt  Sandstones of  Pennsylvanian age,
 because some holes  are  uncased and  other holes
 have deteriorated casing.

 Q. Have you been able to map or identify most of
 the natural discharge  area for the brine and map the
 natural regional movement for the brine?
 A. No, at present we merely know general  areas
 where brine is  discharged. There is  a cooperative
 study by the State and Federal Geological Surveys
 to map the  salt water—fresh water interface. This
 should be of significant aid to the Division of Oil
 and Gas in regulating the  casing and cementing of
 new wells.

 Q. What  do you  consider to be the future for
 ground-water use in  the Kanawha River Valley and
 will significant development be possible in view of
 the problems?
 A. At present  we  are limited to  the relatively
 shallow consolidated aquifers for ground water not
 high  in chloride. Yields of 50 to 200 gpm are availa-
 ble at depths of about 150 feet. The unconsolidated
 alluvium  produces  50 to  150 gpm  to  standard
 vertical screened wells in  selected places between
                                             199

-------
Charleston and Point Pleasant. Higher yields of 200
gpm may  be available  in  downstream areas of
Mason County.
Q.  One of your slides showing the occurrence of
saline ground water showed that it was shallow in
the western  half of  the  State and deeper  in the
eastern half.  Is  the  reason that there is no shallow
saline ground water in the eastern half of the State
because the  bedrock  contains no shallow aquifers
there, or because the water is fresh there?
A.  The history of ground-water development shows
that deeper aquifers in the eastern part of the State
contain fresh water.  Apparently the natural salty
ground water is nearer the land surface towards the
Ohio River area which limits fresh ground water to
the relatively shallow zones.
Q. Was the surface soil removed from the chloride
storage sites to speed the recovery of the aquifer?
A. I doubt it. I believe that the storage areas were
swept  clean  of solid salt  with brooms. New salt
storage  facilities will  be protective sheds, perhaps
with impermeable  floors.  Also some salt may be
stored in plastic bags.


Q. Here's a question  that has to do with so-called
impermeable zones. Was the shale heavily fractured,
or did  the chlorides readily  move through these
"impermeable " zones?
A. Most of  the shale was very heavily  fractured.
The  silty shale or  sandy  shale, almost always will
produce water. Certainly not as much as sandstone,
but enough for domestic supplies.
 200

-------
Probable  Impact  of NTA  on  Ground  Water

by William J. Dunlap, Roger L. Cosby, James F. MclMabb, Bert E, Bledsoe, and Marion R.
                   ABSTRACT
     Laboratory studies were employed to investigate the
fate and effect of NTA both in ground waters and in soil
profiles overlying ground waters.
     Studies of the sorption of NTA by sand, loam, and
clay-loam soils indicated that sorption of NTA on soils
could slow its movement into and through ground waters.
Sorption will  probably not be sufficient to prevent or
greatly reduce potential pollution of ground water by NTA
used as a detergent builder.
     Soil column  studies  were employed  to investigate
the degradation and effect  on metals of  NTA infiltrating
through  soils. These studies indicated: NTA infiltrating
through most unsaturated soils likely would undergo rapid
and complete degradation  and contribute only  inorganic
nitrogen compounds and carbonate to ground waters; NTA
infiltrating  through saturated soils would probably experi-
ence only very limited degradation, with a major portion
entering ground water  intact; any NTA  which escaped
degradation during infiltration through soils could transport
such metals as iron, zinc, chromium, lead, cadmium, and
mercury from soils into ground waters.
     Studies with model aquifers constructed from natural
aquifer  sand indicated that NTA would likely undergo slow
degradation in essentially anaerobic ground-water environ-
ments, with production of  CO2, CH4,  and possibly other
organic compounds.

                INTRODUCTION
     Ground waters in the United States are being
increasingly  subjected to possible pollution by a
wide array of synthetic  organic chemicals entering
the environment as a result of the  needs and de-
mands  of  modern  society.  Nitrilotriacetic  acid
(Figure 1) is a synthetic organic compound worthy
of special  concern in regard to its possible impact
on  ground water  because of the  magnitude  of
potential  usage of this  compound, both from the
standpoint  of quantity  and geographical distribu-
tion.
     Nitrilotriacetic acid (called NTA) is currently
the leading  candidate   to  replace the condensed
phosphates now used as "builders" in the formula-
tion of synthetic detergents. Elimination of phos-
phates from detergents  is sought  because of their
probable  contribution to the eutrophication prob-
lem plaguing many of our surface waters. If NTA
            0
            II
      HO-C
     aPrepared for the  National Ground Water Quality
Symposium, Denver, Colorado, August 25-27, 1971.
     ^National Ground Water Research Program, Environ-
mental Protection Agency, Office of Research and Monitor-
ing, Robert S. Kerr Water Research Center, Ada, Oklahoma
74820.
^x
                                       o
                       N-CHo-C
                                     \
               /
        0=C
             i
             OH
             Fig. 1. Nitrilotriacetic acid.

were to replace only half of the condensed phos-
phates currently  used in  detergents in the United
States,  over  a billion  pounds of  this substance
would  appear in our waste waters  annually. Fur-
thermore, the intended use of NTA in detergents
would insure its distribution throughout the Nation
and provide  ample  opportunity  for entry of this
substance into ground waters.
     NTA  from  detergents  could  possibly enter
ground water both by infiltration through the soil
mantle above the water table and discharge directly
into the ground water. The first pathway would be
operative  in  cases  where  waste waters  are  dis-
charged to permeable soils appreciably above the
water table, such as in properly designed and oper-
ating  septic  systems, waste  lagoons,  and related
sewage  spreading systems. The   second pathway
would be followed in the all too frequent situations
where  septic  systems, lagoons,  and similar waste
disposal systems  are connected directly to ground
waters, and also in cases of direct artificial  recharge
of aquifers.
     Both during infiltration through the soil man-
tle  and upon entry to the ground-water  environ-
ment below the water table, NTA would be subject
to possible sorption by soil particles and microbial
degradation.  Also, since it is an exceptionally effi-
cient chelating agent, NTA might also react with
metals present in the soils to form soluble metal
chelates.  Therefore,  answers  to  the following
questions are needed in order to assess the probable
impact on ground water of large scale  use of NTA
in detergents.

     1.  How will sorption affect the infiltration of
                                             201

-------
NTA through soil profiles and its movement within
ground-water aquifers?
     2.  To what extent will NTA be degraded dur-
ing infiltration through unsaturated and saturated
soils and within essentially anaerobic ground-water
environments?

     3.  If  NTA is  degraded  during infiltration
through  soil profiles or within ground-water en-
vironments, what will be the nature and fate of the
degradation products?
     4.  To what extent will NTA affect the trans-
port of metals into and through ground water?

     This report describes a research project under-
taken in an effort  to provide answers  to  these
questions. The project was part of a relatively short
term  but intensive  effort by  the Environmental
Protection  Agency to evaluate the  environmental
acceptability of NTA before its large scale use in
detergents  is  begun, and research  was therefore
limited  to  a laboratory investigation consisting of
three phases conducted more or less concurrently.
The first phase of the investigation involved study
of the sorption  of NTA by typical sand, loam, and
clay-loam soils.  In the second phase,  soil column
studies  were employed to investigate  the possible
degradation and effect on metals of NTA infiltrat-
ing through saturated and unsaturated soils. In the
third phase of study, the possible degradation of
NTA in an essentially anaerobic ground-water en-
vironment was investigated using simulated aquifers
constructed in the laboratory from natural aquifer
sand.

              SORPTION STUDIES
     Sorption of NTA by soils was investigated by
laboratory sorption  studies  employing a  sand,  a
loam, and a clay-loam soil in order to broadly span
the various  soil  types which might be encountered
in natural  situations. The soils used, all  of which
were obtained in the vicinity  of  Ada, Oklahoma,
were identified  by the United States Department
of  Agriculture  Soil  Conservation  Service  soil sur-
veys as Konawa loamy fine sand,  Claremore  loam,
and Burleson clay loam.   :
     Solutions  of   the  trisodium  salt  of   NTA
(NTA Na3) uniformly labeled with carbon-14 were
utilized  in  the  sorption  studies as well  as in the
other phases of this  investigation. Soils were placed
in  contact with various  concentrations  of  14C-
NTA Na3  in  Erlynmeyer flasks and the mixtures
were agitated  at  20° C  until equilibrium was at-
tained.   Concentrations of  NTA Na3  in  the  solu-
tions at equilibrium and quantities of NTA Na3
sorbed  by the soils in contact  with  these solutions
were determined by liquid scintillation spectrome-
try.  Sterile  solutions and  soils  as well as aseptic
techniques were employed throughout the sorption
studies to eliminate the possibility of errors due to
microbial degradation of NTA.
     The data obtained for sorption of NTA on the
soils utilized in  this investigation were  best  de-
scribed by means of the Freundlich adsorption iso-
therm, as shown in Figure 2. Greatest sorption was
evidenced by the loam soil, with the sand being
least effective. At  an  equilibrium concentration of
50 mg/1 of NTA Na3  in water, sorption values for
the loam, clay-loam,  and sand soils were  64,  28,
and 8.7 Mg NTA Na3  per gram of soil, respectively.
Analogous data at  5 mg/1 NTA Na3 were 10.2
3.5A/g/g, and 0.98 Mg/g.
                                  sorption - •
                                          NTA
                                        9 soi I
                                 brium Concentration - mg/1
                                 •e empirical constants
                          40 60 60 100
               C (mg/1 NTANOj)
 Fig. 2. NTA sorption; Freundlich isotherm at 20°C.

     These  data indicate that under some circum-
 stances sorption of NTA by soils could  consider-
 ably  retard movement of  this  compound  into
 ground water and within aquifers. Nevertheless, as
 an individual mechanism, sorption of NTA by soils
 does  not appear to be sufficient to greatly reduce
 the potential pollution of ground waters by this
 compound  over the course of ,its long-term usage
 as a detergent builder. However, by retarding the
 movement  of NTA in soils  and hence effectively
 increasing its residence  time  in regions of high
 microbial activity, sorption may play a significant
 secondary role in the removal of NTA from infil-
 trating waters by microbial degradation.

            SOIL  COLUMN STUDIES
 Studies  Employing Natural Soils
      Initial  soil  column studies were concerned
 with  the possible  degradation of NTA during infil-
 202

-------
tration through unsaturated and saturated Konawa
loamy fine  sand,  Claremore loam, and Burleson
clay-loam soils'and the effect of NTA on the metals
native to these  soils. The soils were  packed into
4-inch O.D. X 4 feet borosilicate glass- columns to
produce  columns of soil which were 3.75 inches in
diameter and averaged  39  inches in  height. Two
columns of each soil type were prepared, with one
being operated under unsaturatedv or  aerobic, con-
ditions with the water table near the bottom of the
column, and the  other  being operated  under
saturated, or essentially' anaerobic, conditions by
maintaining the water table near the surface of the
soil. Loam and clay-loam columns are pictured in
Figure 3.
Fig. 3. Loam and clay-loam soil columns.

     Columns were dosed  daily for 58 days at a
hydraulic loading  of 2  gal/ft2  (580  ml/day for
each column) with a weak synthetic sewage pre-
pared from the liquid diet  food Sego,  14C labeled
NTA Na3, and distilled water. The synthetic sew-
age contained approximately 50 mg/1 both of BOD
and  NTA Na3 and had a specific'activity of 900
pCi/ml. Upon completion of the initial 58 days of
operation, dosing of the  soil columns  was con-
tinued for 19 "additional days with synthetic sewage
                                                  containing no NTA in order to better observe the
                                                  effect of NTA on the soil metals.
                                                       Column  effluents were  analyzed daily for
                                                  NTA Na3  by the automated  zinc-zincon method
                                                  (Thompson and Duthie, 1968) and for 14C by liq-
                                                  uid  scintillation  spectrometry. Selected effluents
                                                  were also subjected to thin-layer chromatography
                                                  and radioautography to detect  intermediate organic
                                                  degradation  products of  NTA.  Effluents  were
                                                  examined  initially by emission spectroscopy for
                                                  evidence of metal increases. Those metals observed
                                                  to increase  significantly were  then  determined
                                                  quantitatively  by atomic  absorption spectropho-
                                                  tometry, employing standard  procedures  (Federal
                                                  Water Pollution Control Administration, 1969).
                                                       As shown by Figures 4  and  5, essentially no
                                                   NTA was detected in effluents from the unsaturat-
                                                   ed loam and clay-loam columns at any time during
                                                   the 77  days  of operation, even  though  approxi-
                                                   mately 29 mg of NTA Na3 were applied daily to
                                                   each column. NTA was present in the effluent from
                                                   the unsaturated sand column from the 2nd through
                                                   the 9th day of operation, with a peak level  near
                                                   feed concentration occurring  on the 5th day, but
                                                   had virtually disappeared from this effluent by the
                                                   10th day and was not again detected as an effluent
                                                   constituent (Figure 6). Significant quantities of 14C
                                                     __ K>


                                                       '
                                                                      jffttt Lnfl, NTANo^ »^ ^C^

                                                                     10    30
                                                                                    5O    60    70
                                                                       Days of Operation
                                                     Fig. 4. NTA, I4C, and metals in effluent from unsaturated
                                                     loam column.                 .

                                                     not  accountable as NTA were found in the efflu-
                                                     ents from all 3 columns during most of the period
                                                     of operation, thus confirming the occurrence of
                                                     NTA degradation-  in the columns. The early pres-

                                                                                                203

-------
ence and subsequent disappearance of NTA in the
sand column effluent undoubtedly reflected a short
acclimation period required for establishment with-
 E

 m
a*no-


/ Xc 	
/ N
/ f™ ^ 	
j iO 20 50 40 50 60 70
Days of Operation
 Fig. 5. NTA, "*C, and metals in effluent from unsaturated
 clay-loam column.
                          30    40     5O     60    ?0
                       of Op«ro!ion
  Fig. 6. NTA, !4C, and metals in effluent from unsaturated
  sand column.
in the soil  of a microbial population capable of
degrading NTA. Because  of the greater  sorptive
power of the loam and  clay-loam soils, the  move-
ment  of  NTA in these  columns  was  probably
retarded sufficiently  that  acclimation and  subse-
quent  degradation occurred before any NTA trav-
ersed the lengths of these columns.
     Thin-layer chromatography  and  radioautog-
raphy  studies  revealed  no  organic  degradation
products of NTA in  effluent samples from any of
the unsaturated soil columns. Radioanalysis  of gas
released when effluent samples were acidified indi-
cated that virtually all of the 14C present in efflu-
ents from all 3 columns was incorporated in CO2 .
     These  data  indicate  that  NTA  infiltrating
through unsaturated soil  systems  probably  will
undergo rapid and complete degradation, with the
only products likely to enter ground waters  as  a
result of such degradation being CO2 and probably
inorganic nitrogen compounds.  The  introduction
of  additional carbon and  nitrogen  into ground
waters is certainly less than desirable, but the mag-
 nitude of  the  addition likely to  result from the
 projected  use of NTA in detergents is  probably
 not sufficient for critical concern.
      NTA  and 14C appeared in  effluents from the
 sand  and loam columns operated under saturated,
 or essentially anaerobic conditions  on the 4th and
  llth  days  of operation, respectively (Figures 1 and
  8). After  attaining  concentrations near  or some-
                  20    30    40     SO     SO
                    Days of Operation
  Fig. 7. nJAt 14C, and metals in effluent from saturated sand
  column.
  204

-------
                   Days of Operation
Fig. 8. NTA, 14C,  and  metals in effluent from  saturated
loam column,

what exceeding the feed level (50 mg Na3 /I) during
the early phases of column operation, NTA Na3 in
effluents from these columns remained within the
range of 65-85  percent  of feed levels until dosing
with NTA Na3  was terminated. Following  initial
peaks,  I4C levels in both  effluents also decreased
and were  generally 80-95  percent  of feed concen-
tration for the remainder of the dosing period.
     NTA  and  14C  first appeared in the  effluent
from  the  saturated  clay-loam column on  the 7th
day of operation, attained relatively low peaks on
the  14th  day. and  then  declined  to minima  on
about the  29th day (Figure 9). This decline was
attributed  to aerobic degradation of NTA resulting
from  the  presence  of  air in the capillary fringe of
the column and possibly to the entrapment of air
pockets in soil interstices  throughout the  column.
After about the 30th day of operation, the saturat-
ed clay-loam column began performing in a manner
similar to  the other saturated columns, probably
both  as  a result  of  intentional  flooding of the
capillary fringe of the  column on the 29th day and
depletion   of  oxygen   in  trapped  interstitial  air
pockets.
     The fact  that NTA Na., and 14C effluent con-
centrations were  less  than  feed  levels and  14C
effluent levels were equivalent  to  2-10  mg/1 more
NTA Na3   than  actually appeared in column efflu-
ents during most of the operating period for both
the saturated  sand and  loam columns and for the
clav-loam column after the 30th day indicate that
                                                     — 600-
                                                     e
                         JO     40     50

                      Days of  Operation
Fig. 9. NTA, 14C, and metals  in effluent from saturated
clay-loam column.

some degradation of NTA was  occurring  within
these columns. It is not evident, however,  if this
degradation was  truly anaerobic in nature, or if it
was  at  least partially aerobic degradation  which
occurred in the capillary fringe regions of the col-
umns and which  was  arrested by lack of oxygen as
the  feed  solutions  infiltrated  deeper  into  the
columns.
     Analysis  of  effluents  from the saturated col-
umns for 14CO2  revealed  that a major portion of
the  14C present  in  form  other  than  NTA was
incorporated in  CO2.  Very small quantities of at
least  one  organic  degradation  product of NTA
were  shown  to  be present in these effluents by
thin-layer  chromatography and  radioautography,
but  the extremely  low  levels  of  the  substance
prevented its identification.
     The data obtained in these studies with satu-
rated soil  columns  show that  NTA  infiltrating
through  saturated  soil systems, such  as flooded
septic tank percolation fields  or under waste water
lagoons, will  likely  undergo at most  only slow
degradation, and hence high proportions will prob-
ably  reach ground  water  essential'}' unchanged.
Some NTA degradation products could possibly be
produced  in  saturated  soils, but  the  quantities
entering ground  water would likely be  extremely
small.
     As indicated in Figures 5-9, there were gener-
                                             205

-------
 ally good  correlations  between concentrations of
 NTA Na3 and of iron, zinc, nickel, cobalt (although
 the  concentration of the latter element was never
 very high), and possibly  manganese  in effluents
 from various soil columns. The decreases in metals
 concentrations  in  the saturated sand  and loam
 column effluents which accompanied decreases in
 NTA Na3  concentrations following .termination of
 NTA Na3  dosing on the  58th day  of operation
 appear particularly significant.  Observation of the
 effects of termination of NTA dosing on metals
 concentrations in the effluent  from  the saturated
 clay-loam column was not possible because clogging
 and resulting  poor flow performance precluded the
 collection  of further meaningful data from this col-
 umn  after about  the  60th  day. However, the
 minima in effluent  metals concentrations  corres-
 ponding to the minimum in  NTA  Na3 concentra-
 tion occurring on the  29th day of operation pro-
 vides significant additional information concerning
 the effect of  infiltrating NTA on the metals in the
 saturated clay-loam soil.
      Reasonable caution must be exercised in inter-
 preting these metals data because of the  lack  of
 comparison data from control columns. Neverthe-
 less, the  results  of  these  experiments indicate
-strongly that NTA which infiltrates through over-
 lying soil without being degraded is likely  to  solu-
 bilize and transport  into .ground water  metals
 native to the soils.

 Studies Employing Sand Enriched with Metals
      In order to further investigate  the possible
 transport  of  metals  into  ground water by NTA,
 additional  soil column  studies were conducted
 employing columns prepared from Konavva loamy
 fine sand  enriched  with lead, zinc,  chromium,
 cadmium, and mercury.
      Six soil  columns 3.75 inches in  diameter and
 approximately 39 inches in depth were used. Two
 were prepared from sand enriched with the follow-
 ing natural ores: galena (PbS); sphalerite (ZnS);
 chromite (Fe(CrO2)2); greenockite (CdS); and cin-
 nabar (HgS). The other 4 columns  were packed
 from  sand enriched  with a  synthetic sludge pre-
 pared by  alkali precipitation  of lead, cadmium,
 chromium, and zinc  from  a  solution of  their
 soluble salts,  with subsequent drying and grinding
 of  the precipitated  sludge. Mercuric sulfide  (cin-
 nabar) was also added  to the sludge-enriched sand
 as a source of mercury.  The  individual columns
 contained 5 g of each of the  5 metals, except that
 only 995  mg of chromium was present in each of
 the. 2 sand-metal ores  columns because the chro-
 mite ore  contained  only  very low levels of this
 element.
     Two of  the columns containing synthetic
sludge were operated under unsaturated, or aerobic,
conditions, while the other pair of sludge-contain-
ing columns and the 2 columns containing the ores
were operated under saturated, or essentially anaer-
obic, conditions.
     The columns were dosed  daily at a hydraulic
loading of 2 gal/ft2 (580 ml/day for each column).
One  column  of each pair was  dosed with  67 mg/1
of 14C labeled NTA Na3  in distilled water (specific
activity 920 pCi/ml), while the 2nd column of the
pair  served  as a control and  was dosed with
distilled water only. Starting  with the 56th and
53rd day of operation,  respectively, for  the col-
umns dosed with  NTA and the control columns,
sufficient Sego to impart a BOD of  50 mg/1 was
added  to the daily feed solutions.  The columns
were operated for a total of 92  days.
     Effluents from columns dosed with NTA Na,
were analyzed for NTA Na3  and for radioactive
carbon as in the  initial soil column studies. Iron,
manganese,  lead,  zinc, chromium, and cadmium
were determined in effluents  from all 6 columns
by atomic  absorption  spectrophotometry,  while
mercury  was determined by the flameless atomic
absorption procedure (Federal Water Quality Ad-
ministration, 1970). Data obtained for the columns
dosed with NTA are presented  in Figures 10-12.
                    Days of  Operation
Fiq. 10. NTA, 14C, and metals in effluent from unsaturated
sand-metals sludge column.
 206

-------
     216

     192

     166
  96O

  84O
  36O

  240-

  1 20-
                   Day s of  Operation
Fig. 11. NTA, 14C, and metals in effluent from saturated
sand-metals sludge column.
  IOSO

  960

  840

  720

  60O-

< 4SQ-
u~
a
— 360-
o
2 240-

  120-
                               -FeedLevtl NTANc3 &  C
 .0


1-30 !


 20


 10
       I    to   20   30   40   50    60   70    60   90
                   Days of Operation
Fig. 12. NTA, 14C,  and metals in effluent from saturated
sand-metal  ores column.
     NTA first appeared in the effluents from the
unsaturated and saturated columns of sand en-
riched  with metals sludge on the 19th and 22nd
day of operation, respectively (Figures 10 and 11).
NTA  concentration  in  the unsaturated  column
effluent attained a maximum on the 39th day, and
then declined  to become relatively stable at about
25 percent of feed level after the  70th day. In the
saturated   column  effluent  NTA  concentration
reached a  maximum  on  the  57th  day  and  then
slowly declined until termination of the  experi-
ment. NTA first appeared  in the effluent from the
saturated column of sand enriched with metal ores
on the 3rd day  of operation, attained a concentra-
tion near the feed level  on the  7th day,  and re-
mained near feed level for the remaining 85 days of
the study (Figure 12).
     Comparison of  these data with similar  data
obtained from the unsaturated and  saturated  sand
columns in the  initial soil  column studies (Figures
6  and  7)  reveals  that the  movement  of  NTA
through sand enriched with metals sludge was very
much slower than through sand alone, while  NTA
movement through sand enriched with metal ores
was  little different than  movement ot  this  com-
pound through  sand alone. Obviously,  the metal
hydroxides which principally constituted the  met-
als  sludge  exerted  some retarding effect on  infil-
trating  NTA  not manifested by  the metal  ores,
which were mostly sulfides.
     As Figure  10 shows, only about 75 percent of
infiltrating NTA was degraded in the unsaturated
sand-metals sludge column, while NTA was  com-
pletely degraded, after a short acclimation  period,
during infiltration through the unsaturated column
containing  only sand in  the  initial soil  column
studies (Figure  6). This effect most  likely  was due
to toxic effects of the heavy metals in the sludge
which  resulted  in partial  inhibition of microbial
activity in the sand-metals sludge column.
     The total quantities of metals eluted from the
columns dosed with NTA were in every case many-
fold greater than the quantities  eluted from the
corresponding control columns receiving no NTA.
     As shown  by Figures  10  and  11, cadmium
levels were very high  in effluents from both the un-
saturated   and  saturated  sand-metals sludge  col-
umns,  attaining maxima of  14 mg/1 and 36  mg/1,
respectively, at  the same time that peak NTA con-
centrations occurred. Iron and zinc  were also pres-
ent in the effluent from the unsatu-rated  column,
and high levels  of iron, which had not been added
to the metals sludge, appeared intermittently in the
saturated  column effluent. In addition, lesser quan-
tities of zinc, chromium, lead,  and mercury oc-
curred in  the  latter effluent during the period that
                                             207

-------
NTA was present, with peak concentrations gener-
ally appearing  in synchrony  with  iron peaks. As
with the saturated sand-metals sludge column, iron
was intermittently present in high  concentrations
in the  effluent from the saturated sand-metal ores
column (Figure 12), together with lesser quantities
of zinc, chromium, cadmium, and mercury.
    Throughout  the period  of operation, concen-
trations of the metals mentioned above were either
very small  or below detectable  levels in effluents
from the control columns.
     Solubilization  and  transport of metals from
the soil by  NTA appears to be the only explanation
for the significantly  greater quantities of metals
eluted  from  the  experimental columns than from
the control columns. The experimental and control
column pairs were prepared from identical quanti-
ties of the same  enriched  soils and were operated
by identical procedures except for the addition of
NTA  to the feed for  the experimental columns.
There  were essentially no differences in pH values
and redox potentials in effluents from experimental
and control columns.
     The data obtained in  these experiments leave
little doubt  that infiltration  of water containing
NTA   through saturated  soils  containing  heavy
metals could result in solubilization and subsequent
transport of such metals into ground waters. These
studies also indicate that infiltration of water con-
taining NTA through unsaturated soils containing
high levels  of heavy metals could result in both
NTA  and  metals entering the underlying ground
water  because of metals toxicity and the resulting
inhibition  of microbial degradation of NTA. The
extent  and  nature  of metals solubilization and
transport by infiltrating NTA in a specific natural
situation will obviously depend on such variables as
the quantities  of various  metals in  the soil, the
anionic species with  which the metals are associ-
ated in the soil matrix, the concentration of NTA
in the  infiltrating solution, and the relative stabili-
ties of the various  metal-NTA complexes under
conditions  prevailing in the  particular soil system
of concern.
    Over-all, the soil  column  studies described
above  indicate some  potential for  pollution of
ground water by metals as a result of the use of
NTA  in detergents.  Accurate evaluation  of the
actual  extent and seriousness of this potential will
likely require thorough evaluation of the complex
factors affecting the formation  of  chelates by
NTA and metals in soils, including metals native to
the soils and those deposited by waste disposal.

       SIMULATED AQUIFER STUDIES
    In order to  investigate the possible degrada-
Fig. 13. Details of simulated aquifer design.

tion of NTA  in the essentially  anaerobic ground-
water  environment, model aquifers simulating as
nearly as possible  conditions prevailing in a natural
aquifer were constructed in the  laboratory.  Figure
13 shows details of the simulated aquifer design.
     Each laboratory aquifer was constructed from
a 4  1  aspirator bottle  containing  3  1  of  natural
aquifer sand. A 6  mm O.D. borosilicate glass tube
with perforations  placed  in a spiral arrangement
below  the surface of the  sand served  as  a well
shaft. This tube was packed loosely with glass wool
and  was topped with a  W-inch Swagelok Zytel  tee
"well head."  The  side  arm of  this tee was con-
nected  to a  closed loop  of Tygon tubing which
passed through a peristaltic pump  and back to  the
aquifer unit. This  system  permitted water to be
pumped from the saturated depths of the aquifer
sand through  the  well head and returned  via  the
closed tubing system to the aquifer unit. The upper
arm  of the well head tee was sealed with a Teflon-
lined rubber septum through which the pumped
stream  could  be  sampled. Ports  for evacuation,
gassing, venting, and gas  sampling were also pro-
vided.
     Natural  aquifer sand for construction  of  the
simulated  aquifers was obtained near Byars, Okla-
homa,  from a shallow aquifer underlying the flood-
plain of the South Canadian River.  Aseptic pro-
cedures were employed to  the extent  possible in
collection  of the sand and in construction of simu-
lated aquifers in order that the microbial popula-
tion  of the models might  approach that of  the
natural  aquifer. Three simulated  aquifers were con-
structed, with one being sterilized by autoclaving
208

-------
 after loading of the sand to  serve as a control.
 Initial  charging  of the  simulated aquifers  was
 achieved by first removing most of the water native
 to the aquifer sand by application of vacuum at the
 bottom outlets of the  aspirator bottles and then
 allowing sterile solutions  of NTA Na3  in distilled
 water to flow into the evacuated units. The NTA in
 the  feed  solutions for  the sterile  control aquifer
 and  one of the other units, designated Simulated
 Aquifer Number  2, contained carbon-14, while the
 NTA  in  the  feed for  the  third unit,  designated
 Simulated Aquifer Number 1,  contained no radio-
 active isotope. Conditions  of very low  oxygen
 tension were  created within each aquifer unit by
 twice thoroughly  evacuating  and  refilling  with
 nitrogen gas.
      The  simulated aquifers were  operated in  a
 darkened room at 20° C for 265 days. For the final
 122  days they were operated  under a  blanket of
 nitrogen  to  further  preclude the  possibility of
 oxygen entering  the units. The aquifers  were re-
 charged with NTA after 52,  143, and 209 days of
 operation by  aseptic addition of small volumes of
 relatively  concentrated  sterile  NTA Na3 solutions
 through the tops of the units. Because the volumes
 of water in the  aquifers and the quantities of
 NTA Na3  sorbed on aquifer sand could only be
 estimated, the concentrations of NTA Na3 actually
 present in the aquifer waters immediately  after the
 initial charging and subsequent rechargings of these
 units varied  from 30-55 mg/1. Sufficient Sego to
 impart a  BOD of 60-90 mg/1 to the water in each
 aquifer was included in  the third recharge in order
 that  the effect of an extraneous carbon source on
 NTA degradation might  be observed.
     The  aquifer units  were  pumped occasionally
 for a few hours  to simulate movement of ground
 water and to maintain as much uniformity as pos-
 sible within the units, particularly after recharging.
 Units were always pumped for a short period im-
 mediately before sampling in order to obtain repre-
 sentative samples from  the depths of the aquifers.
 Samples were obtained at 3-7 day intervals by with-
 drawing required quantities of liquid through the
 well   head septums  by  means of sterile  syringes
 while the units were being pumped. These samples
 were  analyzed  for NTA Na3  and for 14C  as previ-
 ously described.  Selected samples  were also sub-
jected to  thin-layer chromatography  and radio-
 autography,  and   the composition  of the gas in
 each  unit was determined by gas chromatography
 each time a liquid sample was obtained.
     As shown by Figure  14, significant microbial
 degradation of NTA occurred in the experimental
 simulated aquifers during the initial  52-day period
of operation. Beginning concentrations of 30-32
                                 Initial Charge
Z 20
        55   60   65   70  75   80   65   90   95   100
                    Days of  Operat ion
Fig. 14. Degradation of NTA in simulated aquifers; initial
charge and first recharge.
mg NTA Na3/l in the aqueous phases of both Aqui-
fer Number 2, which contained  14C-NTA,  and
Aquifer  Number  1, which contained  only  non-
labeled  NTA, decreased to essentially zero during
this period, while NTA Na3  in the sterile control
aquifer decreased only  from approximately 33 mg/1
to 29 mg/1. In the periods following the first and
second rechargings of the simulated aquifers,  simi-
lar degradation  of NTA occurred  in the experi-
mental  units, with the over-all rates being some-
what higher than in the initial period of operation
(Figures 14 and 15).
     Gas samples taken from  the aquifer systems
were found  to contain low levels (usually less than
0.2 percent) of oxygen at various times during the
initial period of operation, but essentially no oxy-
gen was noted in  such samples at any  time  after
the first recharging. Gas analyses also revealed the
production  of methane and  low  levels  of carbon
dioxide   in  both experimental aquifers, with no
production  of either gas  in  the  control  aquifer.
The methane produced in Simulated Aquifer Num-
ber 2, which contained 14C-NTA, was found to be
radioactive,  thus confirming  it  as  a degradation
product  of NTA. This would appear to show be-
yond  reasonable doubt that  the  degradation of
NTA in  the  simulated aquifers was truly  anaerobic.
                                                                                               209

-------
                               Stcond Rtchorge
                NT*—Aflu.lw
     210  #5  22Q 2 25   250  235  MO  2*5250
                                           7*5-
                 Oojft of Optrotion
Fig. 15. Degradation of NTA in simulated aquifers; second
and third recharges.

     Thin-layer chromatography and radioautogra-
phy  gave  limited  evidence  for  the  presence in
Simulated  Aquifer  Number 2 during  maximum
degradation of  NTA of an  intermediate organic
degradation product with chromatographic proper-
ties  very  similar to NTA. Time  limitations  pre-
cluded extensive study  of  this substance, and the
possibility  exists that  it was a  chromatographic
anomaly, produced by the effect on NTA of salts or
other substances in the sample solutions.
     A large proportion of the 14C introduced into
Aquifer Number 2 with  each NTA charge remained
in the aqueous phase of the aquifer after essentially
all of the  NTA had disappeared (Figures 14  and
15).  This  14C was necessarily  a component of
degradation products of NTA, but hardly any of it
could be accounted  for  on  radioautograms of aqui-
fer samples obtained after NTA had disappeared
from the units. Although a portion of this 14C was
undoubtedly present in the form of carbon dioxide
and methane, it appears  that other volatile degrada-
tion  products, which would have been lost during
sampling and  chromatography and  hence would
not have been detected by radioautography, may
have been  present in the  aqueous phase of  the
aquifer after all of  the NTA had been degraded.
Obviously,  the pathways, intermediates, and final
products of degradation of NTA under anaerobic
conditions need further elucidation.
210
     Degradation of NTA proceeded at slightly re-
duced rates in both of the experimental aquifers
after the third recharging, which included Sego as
a carbon source in addition  to NTA (Figure 15).
The magnitude of these rate decreases were relative-
ly small, however, and there is no certainty that
they were  attributable to the extraneous carbon
source  since  products of microbial  metabolism
which had  accumulated in the aquifers during the
210 days of operation prior to the third recharge
could  also  have  caused  some  decrease in  NTA
degradation.  It should be noted that  addition of
an extraneous carbon source to an aquifer with the
initial charge of NTA, before the microbes present
in the aquifer have adapted to degrade  NTA, could
produce a quite different effect than observed here.
     Over-all,  the  data obtained in the  simulated
aquifer  studies  indicate  that  NTA  entering an
essentially  anaerobic  ground-water environment
will  likely  undergo  degradation, with  methane,
carbon  dioxide, and  possibly other volatile  com-
pounds being the  probable final products of such
degradation.  This  phenomenon could  be   quite
important  in  retarding the development of high
levels of NTA in aquifers receiving waste waters
which contain this substance and in decontamina-
tion of aquifers which might become  polluted by
NTA. It should be noted, however, that the sand
employed for construction of  the simulated  aqui-
fers in these studies was obtained from a very shal-
low natural aquifer which probably contained a
numerous  and  active population  of  anaerobic
microorganisms because of periodic  infusion with
water  of relatively high  organic  content.  It is
probable that anaerobic microbial populations in
many natural  aquifers, particularly deep aquifers,
would be much less numerous and active; hence,
the degradation of NTA, or at least the initiation
of  such degradation, would  likely occur  much
more slowly in such aquifers  than was  observed in
simulated aquifers during these studies.
                CONCLUSIONS
     The investigations described above permit the
following  conclusions concerning the  potential
impact of NTA on ground water.

     1.  Sorption  of NTA  by  some types of soils
may considerably  slow  the  initial movement of
this  compound into and  through ground waters.
However, sorption of NTA by soils does not appear
to be sufficient to prevent or appreciably decrease
the pollution of ground waters by this compound
over the course of its long-term usage as a detergent
builder.
     2.  NTA infiltrating through unsaturated  soils,
as in properly operating  septic tank  percolation

-------
fields, will  probably undergo rapid  and complete
degradation. Only inorganic nitrogen compounds
and carbonates are likely to enter ground water as
the  result  of  such degradation.  Introduction  of
these substances into ground waters is not desirable,
but  the  quantities introduced as a result of NTA
use  in detergents  are not likely to be sufficient to
create critical pollution problems.

     3. The degradation  of NTA  infiltrating
through  unsaturated soils which contain high levels
of  heavy metals, such  as  those receiving  wastes
from metals plating operations,  may be partially
inhibited as a  result of toxicity  of  the  metals  to
soil  microorganisms; hence, some NTA may enter
ground water under such circumstances.

     4. NTA  infiltrating  through  saturated soil
systems, as in flooded septic tank percolation fields
and  under  some  waste  water lagoons,  will likely
undergo  only limited degradation; hence, a signifi-
cant portion will enter underlying ground  water,
possibly   accompanied  by  small   quantities   of
organic degradation products of NTA.

     5. NTA which  reaches the water  table will
likely undergo  slow degradation  in the essentially
anaerobic ground-water  environment.  The  final
products of such degradation will probably include
carbon dioxide and  methane, and possibly other
organic compounds of low molecular weight and
relatively high volatility.  Unknown  intermediate
products of NTA  degradation may also  be present
in an aquifer when NTA is being actively degraded.

     6.  NTA that infiltrates through overlying soil
into ground water  without  being degraded may
solubilize and  transport into  the ground  water
such metals as iron, zinc, chromium, lead, cadmi-
um,  and  mercury which may be present in the
soil. Situations  where NTA might infiltrate through
soils  receiving  wastes  containing high  levels  of
toxic heavy metals, such as metal plating wastes,
would appear to entail considerable potential for
metals contamination of underlying ground waters.
However, comprehensive evaluation of the poten-
tial  threat  to  ground-water  quality posed by
solubilization  and transport of metals in soils by
infiltrating NTA will require further elucidation of
the factors  controlling formation  of NTA-metals
chelates in soils.

     7. Even in the most favorable  circumstances,
use of NTA as a detergent builder is likely to result
in some increases in nitrogen and probably carbon
in ground waters. Phosphates  in waste waters do
not readily  move into or through  ground waters
because  of strong  sorption in soils. Hence,  in
comparison  to phosphates the use of NTA in deter-
gents would likely result in at  least a limited ad-
verse  impact on ground water.

            ACKNOWLEDGEMENTS
     The  authors are indebted  to Bobby Newport,
Physical   Science  Technician,  for  his significant
contribution in  construction and operation of the
experimental systems.  The invaluable assistance of
Richard E. Thomas, Soil Scientist, of Linda Harmon,
Secretary, and  of Jack W. Keeley,  Chief, National
Ground Water Research  Program, is also gratefully
acknowledged.

                 REFERENCES
Federal Water Pollution  Control Administration.  1969.
     FWPCA methods for chemical analysis of water and
     wastes.  Analytical Quality  Control Laboratory, Cin-
     cinnati,  Ohio.
Federal Water Quality  Administration. 1970. FWQA pro-
     visional  method  for mercury (flameless AA proce-
     dure). Analytical  Quality Control Laboratory, Cincin-
     nati, Ohio.
Thompson, J.E. and J. R. Duthie. 1968.The biodegradability
     and  treatability of  NTA. Journal WPCF.  v. 40, pp.
     306-319.
 «U.S. GOVERNMENT PRINTING OFFICE: 1972 484-484/122  1-3
                                                                                                  211

-------