CONF-751096
ilte
The ENVIRONMENTAL IMPACT
of WATER CH LORI NATION
OAK RIDGE NATIONAL LABORATORY
OAK RIDGE, TENNESSEE
OCTOBER 22-24, 1975
Oak Ridge National Laboratory
Energy Research and Development Administration
U.S. Environmental Protection Agency
-------
NOTICE
This report was prepared as an account of work sponsored by the United States
Government. Neither the United States nor the United States Energy Research and
Development Administration, nor any of their employees, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or implied, or assumes any
legal liability or responsibility for the accuracy, completeness or usefulness of any
information, apparatus, product or process disclosed, or represents that its use would not
infringe privately owned rights.
This report has been reproduced directly from the best available copy.
Available from the National Technical Information Service, U. S. Department of
Commerce, Springfield, Virginia 22161
Price: Paper Copy $12.00 (domestic)
$14.50 (foreign)
Microfiche $2.25 (domestic)
$3.75 (foreign)
-------
CONF-751096
UC-11,41,48
THE ENVIRONMENTAL IMPACT
OF
WATER CHLORINATION
Robert L. Jolley, Editor
July 1976
Proceedings of the Conference on the
Environmental Impact of Water Chlorination
Oak Ridge National Laboratory
Oak Ridge, Tennessee
October 22-24, 1975
The conference was organized and conducted by members of
the U.S. Environmental Protection Agency, the Energy
Research and Development Administration, and the Biology,
Chemical Technology, and Environmental Sciences Divisions
of the Oak Ridge National Laboratory. The Oak Ridge
National Laboratory is operated by Union Carbide Corpora-
tion for the Energy Research and Development Administration.
Sponsored by
Oak Ridge National Laboratory
Energy Research and Development Administration
U.S. Environmental Protection Agency
-------
FOREWORD AND ACKNOWLEDGMENTS
The formation of chlorine-containing organics during the chlorina-
tion of waters for drinking water sterilization, cooling water treatment,
and sewage processing and disinfection is receiving much national atten-
tion. There exists a need for a thorough examination of the environ-
mental effects of water chlorination. To meet this need the Conference
on the Environmental Impact of Water Chlorination was conceived by
Dr. Carl W. Gehrs and myself. Major objectives of the conference were
to set down in the form of proceedings the present state of knowledge
concerning aqueous chlorination with particular emphasis on chlorinated
organic compounds and associated biomedical and environmental effects.
The conference results and proceedings should establish a better under-
standing of both practical and theoretical aspects of water chlorination
and perhaps set the stage for solving some of the concomitant problems.
I wish to express appreciation to my fellow concerned scientists and
conference registrants for helping make this conference a success and to
the speakers and authors of papers for their many hours of labor and
preparation, for the quality of the conference and proceedings is truly
dependent on their efforts. From the early stages of conference con-
ception, through planning, to accomplishment, Dr. Stanley I. Auerbach,
Director of the Environmental Sciences Division at Oak Ridge National
Laboratory, and Mr. Donald E. Ferguson, Director of the Chemical Tech-
nology Division at Oak Ridge National Laboratory, have encouraged and
advised Dr. Gehrs and me. I wish to express my appreciation to them for
to a great extent the conference is attributable to their leadership.
I wish also to give credit to the capable conference planning committee:
Dr. William A. Brungs, U.S. Environmental Protection Agency; Dr. Robert
B. Gumming, Oak Ridge National Laboratory; Dr. Carl W. Gehrs, Oak Ridge
National Laboratory; Dr. D. Heyward Hamilton, Jr., U.S. Energy Research
and Development Administration; Dr. Sidney Katz, Oak Ridge National
Laboratory; and Dr. W. Wilson Pitt, Jr., Oak Ridge National Laboratory.
In early Spring 1975, Dr. Brungs met in Oak Ridge with Dr. Gumming,
Dr. Gehrs, Dr. Pitt, and myself. At this initial planning session a
tentative program was formulated and subsequently Dr. Brungs enlisted
the speakers for the environmental session, Dr. Gumming enlisted the
speakers for the biomedical session, Dr. Gehrs obtained those for the
modeling and prediction session, and I arranged for the general and
chemistry sessions. The logistics in conducting a conference of this
iii
-------
magnitude were greatly eased by the competent assistance of Mr. Charles
E. Normand and Ms* Wendell H. Martin of the Conference Staff at Oak
Ridge National Laboratory. The assistance of Ms. Joanne Sanford and
Ms. Brenda Taylor was invaluable in taking care of the many details
required to keep the conference running smoothly.
A brief word concerning the proceedings is in order. Because of
the interdisciplinary nature of the papers, I have attempted as much
as possible to have the papers presented in a style which would be com-
patible to .the reading tastes of the many disciplines. For this reason
I selected a style very similar to that used in ecological literature.
The discussions were reproduced with as much fidelity as possible from
the original tapes. However, because a few portions did not tape well
there are several omissions. I have exercised minimal editorial preroga-
tives within the constraints of transfering the intended meaning from
the spoken and written word to the printed word and, also, with respect
to keeping this document concise. I gratefully acknowledge the assistance
of Ms. Martha Stewart for patiently answering my many editing questions,
of Ms. Mary Ellen Smith and Ms. Sandi Lyttle for supervising and typing
the major portion of this final manuscript, of Ms. Rosonna Forkland,
Ms. Ogene Gentry, Ms. Susan Sampson, Ms. Fauna Stooksbury, and Ms. Brenda
Taylor for typing certain papers, of Ms. Hallie Nidiffer for make-up of
the final manuscript, and of Ms. Doris Jolley for many hours of proof
reading.
This conference was sponsored by the Oak Ridge National Laboratory,
the U.S. Energy Research and Development Administration, and the U.S.
Environmental Protection Agency. I especially wish to acknowledge the
advice and assistance of Dr. Charles L. Osterberg, Dr. Robert L. Watters,
and Dr. D. Heyward Hamilton, Jr., of the U.S. Energy Research and Devel-
opment Administration, and Dr. Andrew J. McErlean and Dr. J. David Yount
of the U.S. Environmental Protection Agency. Without their foresight
concerning the need for this conference it would not have been held. I
also wish to thank Dr. Herman Postma and Dr. Chester R. Richmond for
their advice and securing the sponsorship of the Oak Ridge National
Laboratory. A final thanks to Dr. Charles D. Scott, my supervisor, for
permitting me to spend so very many hours on this challenging and
rewarding undertaking.
Robert L. Jolley
Conference Chairman
and Proceedings Editor
iv
-------
TABLE OF CONTENTS
FOREWORD AND ACKNOWLEDGEMENTS, Robert L. Jolley ........ ill
WELCOME, Herman Postma ..................... 1
CONFERENCE ON THE ENVIRONMENTAL IMPACT OF WATER CHLORINATION :
PERSPECTIVE AND OBJECTIVES, Stanley I. Auerbach ........ 3
CURRENT CHLORINATION AND DECHLORINATION PRACTICES IN THE
TREATMENT OF POTABLE WATER, WASTEWATER, AND COOLING
WATER, George Clifford White .................. 7
SESSION I. AQUEOUS CHEMISTRY OF CHLORINE
Joseph E. Draley, Session Chairman ............... 25
THE CHEMISTRY OF AQUEOUS CHLORINE IN RELATION TO WATER
CHLORINATION, J. Carrell Morris ................ 27
MEASUREMENT AND PERSISTENCE OF CHLORINE RESIDUALS IN NATURAL
WATER, J. Donald Johnson .................... 43
ORGANO-CHEMICAL IMPLICATIONS OF WATER CHLORINATION,
Robert M. Carlson and Ronald Caple . .- ............. 73
CHLORINATION OF ORGANICS IN DRINKING WATER, Alan A. Stevens,
Clois J. S locum, Dennis R. Seeger, and Gordon G. Robeck .... 85
CHLORINATION OF ORGANICS IN COOLING WATERS AND PROCESS
EFFLUENTS, Robert L. Jolley, Guy Jones, W. Wilson Pitt,
and James E. Thompson ............. • ....... 115
ANALYSIS OF NEW CHLORINATED ORGANIC COMPOUNDS FORMED BY
CHLORINATION OF MUNICIPAL WASTEWATER, William H. Glaze,
James E. Henderson, IV, and Garmon Smith ............ 153
CHEMISTRY OF HALOGENS IN SEAWATER, James H. Carpenter
and Donald L. Macalady ..................... 177
EXTENDED DISCUSSION ...................... 195
After-dinner Address:
DECISION MAKING IN REGULATION OF CHEMICALS,
Edward M. Brooks ........................ 199
-------
Page
SESSION II. BIOMEDICAL EFFECTS OF CHLORO-ORGANICS
Robert B.,Gumming, Session Chairman 211
HALOGENATED ORGANICS IN TAP WATER: A TOXICOLOGICAL
EVALUATION, Robert G. Tardiff, Gary P. Carlson, and
Vincent Simmon . 213
ORIGIN, CLASSIFICATION AND DISTRIBUTION OF CHEMICALS IN
DRINKING WATER WITH AN ASSESSMENT OF THEIR CARCINOGENIC
POTENTIAL, Herman F. Kraybill .... 229
THE POTENTIAL FOR INCREASED MUTAGENIC RISK TO THE HUMAN
POPULATION DUE TO THE PRODUCTS OF WATER CHLORINATION,
Robert B. Gumming ..... .... 247
THE EPIDEMIOLOGIC APPROACH TO THE EVALUATION OF WATER-
BORNE CARCINOGENS, Kenneth P. Cantor 259
SESSION III. ENVIRONMENTAL TRANSPORT AND EFFECTS
William A. Brungs, Session Chairman . 275
THE TOXICITY OF CHLORINE TO FRESHWATER ORGANISMS UNDER
VARYING ENVIRONMENTAL CONDITIONS, Arthur S. Brooks and
Gregory L. Seegert 277
A REVIEW OF THE IMPACT OF CHLORINATION PROCESSES UPON
MARINE ECOSYSTEMS, William P. Davis and
Douglas P. Middaugh 299
CHLORINATED COMPOUNDS FOUND IN WASTE-TREATMENT EFFLUENTS
AND THEIR CAPACITY TO BIOACCIMJLATE, Herbert L. Kopperman,
Douglas W. Kuehl, and Gary E. Glass 327
INVESTIGATING THE EFFECTS OF CHLORINATED ORGANICS,
Carl W. Gehrs and George R. Southworth 347
SESSION IV. MODELING AND PREDICTION
Carl W. Gehrs, Session Chairman 363
MODELING RESIDUAL CHLORINE LEVELS: CLOSED-CYCLE COOLING
SYSTEMS, Guy R. Nelson 365
VI
-------
Page
A KINETIC MODEL FOR PREDICTING THE COMPOSITION OF CHLORINATED
WATER DISCHARGED FROM POWER PLANT COOLING SYSTEMS,
Milton H. Lie* zke 387
ASSESSING TOXIC EFFECTS OF CHLORINATED EFFLUENTS ON AQUATIC
ORGANISMS: A PREDICTIVE TOOL, Jack S. Mattice 403
SESSION V. ROUNDTABLE DISCUSSION
Carl W. Gehrs, Moderator 423
William A. Brungs 424
Robert B. Gumming 425
Joseph E. Draley 427
D. Heyward Hamilton 428
J. Carrell Morris 429
George Clifford White 431
Audience Participation 432
vii
-------
WELCOME
Herman Postma, Director
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
Welcome to the Oak Ridge National Laboratory. With approximately
300 conference registrants, we have overloaded the system. If problems
arise, be sure to let us know and we will try to resolve them.
The objectives of this conference, to see what is known, what is
being done, and what should be done concerning the chlorination of
various waters, are very important because the real endpoint of this is
to establish a base of data, models, and understanding that will permit
those who have the awesome responsibility of setting standards and
regulations to do so based on facts.
This area of concern, the environmental impact of water chlorination,
shares with so many other areas of modern technology in the last few
years a high degree of visibility. It has the potential for being
emotion-laden and provides a framework for those who prefer to be sub-
jective and for those that consider personal gain to enter the battle.
Therefore, it is very important that this conference has to be held to
objectively assess where we are and what needs to be done. I hope you
make the best of progress toward fullfilling the conference objectives
because, in modern society and the technical areas that we are concerned
with, the impact upon people and our way of life is vitally important.
-------
CONFERENCE ON THE ENVIRONMENTAL IMPACT OF WATER CHLORINATION:
PERSPECTIVE AND OBJECTIVES
Stanley I. Auerbach, Director
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
In a three-day symposium, bringing together scientists of diverse
backgrounds to focus attention on the impact of one of man's activities,
chlorination, on his environment, I believe it is appropriate to discuss
ecology and the role it has come to occupy in our society.
In the last decade, ecology has become a household word, having
different meanings for individuals of differing backgrounds. Although
this general interest in ecology has been beneficial, it has also
brought negative publicity to the science through the activities of
nonecological instant experts and others who convey an aura of "leave
everything like it is," and "no progress is good." Ecology, concerned
with man's relationship to his environment, has made the general public
aware of the unity of mankind and the biosphere. Even though ecology
can trace its lineage to the writings of Hippocrates and Aristotle, it
was not until early in the present century that it became recognized
as a distinct field of science. Part of the reason for this lack of
identity lay in the nature of the discipline. Ecology deals with
organisms and their relationship to their abiotic environment. It is
also concerned with the unique relationships which exist between groups
of different organisms (community ecology) and, ultimately, with the
relationships, interactions, and attributes of groups of organisms in
and to their abiotic surroundings (ecosystem ecology). Because of this
complexity of relationships and the essentially holistic approach of
-------
ecologists to their problems, in general parlance and in the public's
view, ecology is regarded as the totality of the environment—a view
which is somewhat misleading, but which does recognize that environmental
science has physical as well as biological attributes and components.
Ecology is somewhat of a unique science—unique in that while it
looks at the traits of an organism, with respect to itself, or with
respect to other organisms and to its environment, it also looks for
the distinct attributes of organism grouping. In other words, it is
concerned with the occurrence and value of those traits distinctive of
the group as well as the characteristics of the individual components.
In the mid-1970's it is refreshing to know that in our modern world we
are beginning to realize that, as the popular advertisement goes,
"we're all in this together." Or, more scientifically, all of man's
activities in some manner affect his future home or environment.
This conference is called the "Environmental Impact of Water
Chlorination." We have brought together distinguished scientists of
diverse backgrounds; chemists, biomedical biologists, toxicologists,
ecologists, and engineers. We have come together to make an assessment
of the available knowledge of the aqueous chemistry of chlorine, to
examine effects of chlorine and chlorinated organic products on aquatic
ecosystems and man, to assess the biological and ecological implications
of chlorine usage for treatment of natural and process waters such as
cooling waters and sewage effluents, and, through publication of the
conference proceedings, to provide a permanent record and reference
document containing the most recent relevant scientific data. This
forum presents the opportunity for free interchange of information,
both formally and informally, thereby permitting the development of
ideas and concepts concerning areas of research needing emphasis. A
principal result anticipated from the symposium is an increased under-
standing and better definition of the problems associated with
chlorination.
We have gathered together through a common interest in the subject,
but we must recognize that we do not all speak the same scientific
language. This can, of course, cause some problems in communication.
-------
Important, however, is the realization that in coming to this symposium,
each of us could learn from individuals of different scientific back-
grounds. There are unique, valuable attributes in the scientific
community apart from or in addition to those of our own specific
disciplines.
Thus, engineers and chemists discuss the technology of water
chlorination and many aspects of chlorine chemistry in a variety of
waters of environmental concern. Others will discuss biomedical effects,
the epidemiological evaluation of trace concentrations of chemicals,
ecological transport and bioaccumulation of chemicals, toxicity studies,
and predictive tools.
This conference was designed to cover most major aspects of water
chlorination, particularly relating to the chlorination of organic
constituents and possible biomedical and environmental effects resulting
from the chloro-organic products. Each paper surveys its topic, pre-
senting a summary of the known information in addition to individual
author's research. I hope this conference will be a landmark in the
field of environmental effects of water chlorination, and that the
future results emanating from this symposium will be cognizant of the
efforts presented here.
-------
CURRENT CHLORINATION AND DECHLORINATION PRACTICES
IN THE TREATMENT OF POTABLE WATER, WASTEWATER, AND COOLING WATER
George Clifford White
Consulting Engineer
San Francisco, California 94118
ABSTRACT
The present annual production of chlorine in North America is
estimated at 10.5 million tons. Only 3 to 4 percent of this is used
for sanitary purposes such as potable water and wastewater treatment,
swimming pools, household use, cooling water circuits, and food pack-
aging process water. The chemical industry accounts for at least 80
percent and the pulp and paper industry accounts for the rest, that is,
about 15 to 16 percent. This is the perspective of chlorine in the
environment.
The use of chlorine for sanitary purposes dates back to 1854 when
it was used to deodorize London sewage. The first known use as a
disinfectant was in 1879, also for sewage, and its first use as a
potable water disinfectant on a continuous basis was 1903. Since that
time the use of chlorine in potable water treatment has expanded to
include (in addition to disinfection) several other important functions
such as taste and odor removal; iron and manganese removal; hydrogen
sulfide removal; color removal; prevention of water quality degradation
in distribution systems; control of biofouling in long transmission
systems, thereby preventing friction factor deterioration; prevention
and control of biofouling in filter media; restoration of well capacity,
and sterilization of water mains and reservoirs. In sewage treatment
chlorine is being used primarily for disinfection, but it is also used
very effectively for the prevention of septicity and control of hydrogen
sulfide generation in collection systems and treatment plants. Chlorine
is used in limited applications such as control of activated sludge,
bulking, sludge thickening, the destruction of cyanides, and foul air
scrubbing. It is also used in the manufacture of ferric chloride which
is used as an effective coagulant for both potable water and wastewater
treatment. There are many variations and combinations of chlorine
application to potable water. These include chlorination followed by
partial dechlorination or complete dechlorination followed by rechlorina-
tion; and ammoniation followed by chlorination and vice versa for
-------
specific odor control requirements. Chlorine dioxide generated on the
site is used for the destruction of phenols. Chlorine is also used along
with chlorine dioxide, ozone, and activated carbon, for the treatment of
grossly polluted waters. The variations in selective treatment for
potable water are many as is the amount of chlorine applied, for example,
potable water, 5 to 100 pound per million gallon; wastewater, 50 to
400 pound per million gallon; cooling water, 20 to 200 pound per million
gallon (intermittently). In the final analysis the required use of
chlorine in potable water and wastewater depends on local regulatory
requirements with an Environmental Protection Agency override. The
requirements for cooling water are for most economical heat transfer in
the production of power.
Wastewater disinfection in California is being practiced on a total
coliform destruction requirement for various receiving water conditions.
Likewise the necessity to detoxify the chlorine residuals in these
effluents is being accomplished by dechlorination with sulfur dioxide.
Owing to the lack of necessary analytical equipment, proof of continuous
dechlorination to zero chlorine residual has presented some problems.
Dechlorination of cooling water discharge is being taken care of by
dilution of cooling water blow down. This is accomplished in the basic
design of the cooling water system.
Dechlorination of potable water is strictly for enhancement of the
product.
INTRODUCTION
The perspective of chlorine and chlorine compounds in our environ-
ment is amply demonstrated by the percentage distribution of the
annual chlorine production in North America, estimated at 95 x 1011
kilos (10.5 million tons) (Chlorine Institute 1975). The chemical
industry accounts for at least 80 percent of the chlorine use. Most of
this amount is used for making plastics, pesticides, antifreeze fluids,
synthetic fibers, gasoline additives, solvents, and paint removers.
The pulp and paper industry uses about 16 percent of the chlorine. The
remaining 3 to 4 percent is used for "sanitary" purposes which includes
but is not limited to potable water and wastewater treatment, swimming
polls, household use, cooling water circuits, and food packaging
process water (White 1972). The following discussion will be limited
to potable water, wastewater, and cooling water.
-------
We do not know to what extent the housewife's use of sterilizing
and bleaching compounds, encouraged by TV commercials and soap operas,
reaches the environment via the sewage collection system. However, a
quick survey of the detergent and soap manufacturers could give us a
good idea of the amount of chlorine entering our environment by this
route. It might be substantial.
POTABLE WATER
The most important use of chlorine for so called sanitary purposes
is for the disinfection of potable water. Because of chlorine's
oxidizing powers, it has been found to serve other useful purposes
in water treatment such as taste and odor control, prevention of algae
growths in water treatment structures, maintaining clean filter media,
removal of iron and manganese, destruction of hydrogen sulfide, color
removal by bleaching of certain organic colors, maintenance of distribu-
tion system water quality by controlling slime growths, restoration and
preservation of pipeline capacity, restoration of well capacity,
sterilization of water mains and reservoirs.
Current chlorination practice of potable water varies in dosage
from 1.0 mg/1 to about 16 mg/1. This variation is directly related to
the quality of the raw water supply, that is, degree of pollution, con-
centration of nutrients, temperature, and pH. Winter ice cover on a
water supply will almost automatically double the summertime chlorine
demand. To combat the pollution and taste and odor problems, various
combinations and dosages of chemicals are in use today in both the
United States and Canada. These include combinations of chlorine and
chlorine dioxide, ammonia, potassium permanganate, activated carbon,
sometimes followed by dechlorination with either sulfur dioxide or
activated carbon.
Prechlorination of low quality water is most often the operator's
salvation. It is one of the most important tools for maintaining the
efficiency of a water treatment plant. In these situations, which are
numerous, it would be virtually impossible to turn out an acceptable
-------
10
water if it were not for the unique ability of chlorine to maintain a
persisting residual throughout the process. In these cases disinfection
is just a side effect.
Two examples of low quality water that require chlorine dosages up
to 16 mg/1 are the Grand River at Brantford, Ontario, Canada, and the
Passaic River at Little Falls, New Jersey (Inhoffer and De Hooge 1974,
Williams 1974). A notable foreign situation comparable to these two
American cases is the Seine River at Paris, France. The Choisy-le-Roi
treatment plant located on the outskirts of Paris has a filtration
capacity of 900,000 cu meter per day (238 million gallon per day). The
chemical treatment consists of a prechlorine dose of about 16 mg/1 to
reach a 1 to 1.5 mg/1 free residual. Then chlorine dioxide at about
4 mg/1 is added ahead of the filters. After the filters, ozone is added
in accordance with its predetermined 10 minute ozone demand. This is
about 2 to 3 mg/1. Sometimes chlorine is added to the finished water
in the event there is trouble in the distribution system.
In North America poor quality water is treated by prechlorination
sufficient to provide a substantial free residual in the flocculation
or sedimentation basins. This may be followed by an intermediate dose
of 2 to 5 mg/1 chlorine or 0.5 to 4.5 mg/1 chlorine dioxide to carry a
residual through the filters. Posttreatment might consist of the addi-
tion of ammonia to convert the remaining residual to chloramines as the
water enters the distribution system. Alternatively sulfur dioxide might
be used to trim the final residual. Ozone could be used in these situa-
tions for taste and odor control and/or color removal. Granular
activated carbon may also be used for dechlorination of the finished
water.
Now let us consider the case of a clean water where chlorine is
used solely for disinfection. Two typical examples are the supplies for
San Francisco and the East Bay (Alameda and Contra Costa counties).
These waters are derived directly from melted snow in separate run-off
areas high in the Sierra Nevada Mountains. The San Francisco supply is
transported by a 165-mile aqueduct and the East Bay supply travels about
90 miles to various local storage reservoirs. The chlorine dosage
-------
11
required for disinfecting these waters is from 0.8 to 1.2 mg/1. The
San Francisco supply was plagued from the start by assorted difficulties.
The most humiliating one occurred a few months after the system was put
in operation. Part of the 165-mile aqueduct consists of a concrete-
lined tunnel 25 miles long through the Coast Range Mountains. Although
the concrete tunnel lining is up to 12 inches thick, many cracks
developed in it. The groundwater is laden with filamentous bacteria
(Crenothrix sp.) which infiltrated the tunnel through the cracks causing
luxuriant growths on the tunnel lining. This created a debacle in the
distribution system. Industries were forced to shut down because of
this biofouling and the sloughing of the filamentous debris. To combat
the problem, the city of San Francisco, after having investigated
several treatment methods, concluded that a persisting chlorine residual
was the only answer. Therefore, 2 mg/1 of chlorine is applied at the
beginning of the tunnel. This produces a 1 mg/1 free residual at the
end of the tunnel 25 miles away.
Now let us consider those areas not as fortunate as the San
Francisco Bay Area. St. Louis, which uses Mississippi River water,
prechlorinates at an average 5 mg/1. This is supplemented by inter-
mediate doses of chlorine to preserve the efficiency of the filter
system. Kansas City, which uses Missouri River water, prechlorinates
at about 12 mg/1. These chlorine applications are not for disinfection.
This is to keep the water treatment process from deteriorating and, also,
for taste and odor control.
The City of Chicago, which uses a cleaned up Lake Michigan water,
prechlorinates at only 1 to 1.2 mg/1. This is comparable to the melted
snow waters of California. This prechlorination dose is designed to
provide an absolute minimum of 0.25 mg/1 free chlorine through the
filters. It is supplemented by a postchlorination dose of 0.25 to
0.36 mg/1 to give a 0.75 mg/1 free chlorine residual entering the dis-
tribution system. However, to exemplify the effect of pollution on
chlorination, whenever there is an upset with the Chicago Ship Canal,
which amounts to a reversal of flow and dumps the canal flow into Lake
Michigan, the chlorine demand escalates to 10 mg/1. Even at this
-------
12
chlorine dosage there is so much ammonia-nitrogen in the raw water that
a free chlorine residual is not attainable.
Surface waters in areas of Pennsylvania require prechlorination
doses of 7 to 8 mg/1. The City of Philadelphia provides sufficient
chlorination equipment capacity to dose as high as 30 mg/1 at the raw
water basin outlet (ice cover situation) and up to 4 mg/1 chlorine
dioxide for pretreatment.
It is significant to note the quality of the water provided by
the billion dollar California Water Plan. This water is a mixture of
several rivers, primarily the Sacramento and San Joaquin. This water
requires only 2 to 4 mg/1 prechlorination dose to produce a 0.7 mg/1
free residual in the filter plant finished water without any inter-
mediate chlorination. This is a tribute to the diligence and surveil-
lance of both the California State Department of Health and the
California Water Resources Quality Control Board for their programs
designed to preserve the water quality of the receiving waters.
There are many other situations where the water responds to the
various treatment processes but the quality subsequently deteriorates
in the distribution system. These systems have no other choice but to
maintain a chlorine residual in the distribution system. Many large
cities maintain a residual as a matter of policy on the basis of
preventive medicine. The following examples are cited to illustrate
this practice. Brantford, Ontario, converts the remaining free residual
to chloramine by postammoniation. Water leaving the plant carries a
0.4 to 0.6 mg/1 combined residual in winter and 1.5 to 1.7 mg/1 in
the summer. Chicago water enters the distribution system at 0.75 mg/1
free chlorine residual (Willey et al. 1974). The Washington Suburban
Sanitary Commission processes water from the Potomac and Patuxent rivers.
The water leaving their treatment plants enters the distribution system
with a 1.0 to 1.5 mg/1 free residual. There are other systems that go
as high as 2.0 mg/1.
In many groundwaters the bacteriological quality is of no concern
but the presence of a combination of iron and manganese causes dirty
water in the distribution system. A California city afflicted with this
-------
13
problem receives water from some 50 wells with a manganese concentration
varying from 0.1 to 0.5 mg/1. Each well is treated with chlorine and
Calgon. The chlorine dose varies from 0.25 to 2.2 mg/1, depending on
the well. This treatment provides control of the dirty water problem
whereas other solutions have failed.
Other groundwaters require the addition of chlorine as a signifi-
cant part of the iron removal process. Chlorine dosages in these waters
are typically 2 to 7 mg/1. The chlorine applied also controls the
growth of iron-bearing filamentous organisms.
A large number of supplies have to contend with taste and odor
control at the source — usually an impounded supply (AWWA 1974). The
city of Winnipeg, Canada, is an example. Chlorine is their main treat-
ment for taste and odor control of the'impounded water which is
separated from the treatment plant by a 90-mile long aqueduct. Chlorine
is applied as the water enters the aqueduct at a rate of 4 mg/1. Forty
hours later the water arrives at the treatment plant with about 0.3 mg/1
total chlorine residual. Many other systems use chlorine application
to a transmission line solely for the purpose of maintaining carrying
capacity.
A great many water systems in the British Isles resort to high free
residual chlorination (1 to 3 mg/1) followed by residual controlled
dechlorination with sulfur dioxide so that the residual entering the
distribution system can be maintained at the desired level. In some
cases the high free chlorine residual is used to compensate for short
contact times. This procedure is beginning to be used in the United
States (White 1968).
There are some waters that require an artificially induced break-
point to control taste and odors. This is accomplished by adding
ammonia-nitrogen and subsequently chlorinating to a free residual.
One of the most important uses of chlorine in the purveying of
potable water is main sterilization. This is also extended to steril-
izing water reservoirs, ship tanks, etc. Dosages involved are usually
50 mg/1 with a retention time of 24 hours. These parameters have
evolved after a half century of experience by the water industry.
-------
14
The recent publicity given to the formation of chloro-organics in
potable water due to chlorination and to the possibility the chloro-
organics may be detrimental to health has raised considerable interest
in the use of ozone as an alternative to chlorination. This is not a
practical approach because ozone cannot do all the things that chlorine
can do. Besides, the equipment problems present practical considerations
that always favor chlorine.
Ozone is an extremely reactive oxidant. It is a much better
viricide than a bactericide. It reacts so quickly that the residual
die-away is not much longer than five to eight minutes. It is excellent
in certain situations of color removal and taste and odor control. Most
ozone installations are backed up by chlorination equipment (White 1974).
Ozone must be applied on a demand basis. This means that pzone
demands must be performed at frequent intervals. Operating ozone
equipment from either a flow proportional or a residual control signal
i
is not an established fact. Ozone equipment is either on or off.
Another disadvantage is the readout of ozone being applied. This is not
a routine measurement as in a gas chlorinator because the gas leaving
the ozonator is a mixture of air or oxygen and ozone. This has been
the source of operating problems.
It is interesting to note that a recent installation at Strasburg,
Pennsylvania involves a small domestic water supply consisting of
13 springs. This water could probably be handled by a chlorine dose
of 0.8 mg/1. The flow range is 0.26 to 0.45 m3'/min (70-120 gpm) . The
installation cost $23,000 not including the equipment which is leased
for $12,000 per year. The ozone generator has a capacity of 3.7 Ib/hr.
The dosage used is about 1.5 mg/1 (Harris 1975). An automatic hypochlo-
rinator installation would have cost about $5,000 at the most.
The attributes of ozone should not be overlooked. I have and
always will strongly advocate the exploitation of the properties of
ozone to complement the chlorine compounds in the quest for a better
quality potable water. This is particularly significant in view of the
virus problem and the water reuse consideration (White 1975a, 1975b).
We need more information relating to the chemistry of combinations of
oxidants. It is folly to try and pit one against the other.
-------
15
Finally there is the question of proof of disinfection. The
existence of even a trace of chlorine residual, regardless of form,
drastically reduces or eliminates total coliforms from the distribution
system samples (White 1975a). The presence of a chlorine residual as
proof of disinfection would eliminate surveillance problems and eliminate
risks for those small but numerous water systems, for example, ski and
other resorts, roadside restaurants, bus stops, motorway rest areas,
trailer camps, farms, suburban homes, isolated institutions, organized
summer camps, and other similar water systems. This would not be
possible with any other disinfectant than chlorine.
WASTEWATER TREATMENT
The first use of chlorine in wastewater treatment was for the
control of odors. This dates back to 1854 when it was used to deodorize
the London sewage. It was first used to disinfect sewage in North
America in 1893 at Brewster, New York (White 1973). This application
was for the protection of the Croton watershed, which is a part of the
New York City water supply. A review of the literature indicates a
trend of active interest in wastewater disinfection beginning about 1945.
Up to that time the main interest in chlorine was for odor control,
hydrogen sulfide destruction, and prevention of septicity. Most of the
rewage treatment plants practicing disinfection during that time belonged
to the U.S. Armed Forces. It was a matter of policy that sewage efflu-
ents had to be chlorinated at all army bases in the United States during
World War II. Today, as a result of the 1970 Federal Water Pollution
Control Act, all wastewater treatment plants in this country are
subjected to some definitive disinfection requirement.
Chlorine also plays an important role in the treatment of cyanide
wastes which are highly toxic. Cyanide wastes must be treated before
being discharged to either a sewage collection system or a receiving
water. When discharging to a sewer the cyanides need only be oxidized
to cyanates but when discharging to a receiving water the cyanides must
be completely destroyed to elemental carbon and nitrogen. The new
-------
16
federal regulations for industrial discharges make treatment of cyanide
wastes imperative. The state of the art for on-site cyanide waste
treatment has developed to a point where packaged systems are readily
available for the individual discharger.
Today the emphasis is on disinfection of all effluents. The
objectives of wastewater disinfection are: to prevent the spread of
disease; to protect potable water supplies, bathing beaches, receiving
waters used for boating and water contact sports; and to protect
shellfish growing areas. So far, the most efficient way to accomplish
these objectives is by chlorination. Over the last 25 to 30 years the
various regulatory agencies, local, state, and now the U.S. Environ-
mental Protection Agency (EPA), have been seeking a set of ground rules
as proof of disinfection for various receiving water situations.
At one time it was thought that disinfection could be accomplished
by showing a 0.5 to 0.75 mg/1 orthotolidine residual at the end of
30 minutes contact time. Contact chambers were built to give a
theoretical 30-minute detention time at average flow based on the volume
of the chamber. Effective mixing and contact chamber short circuiting
was never considered. The California State Department of Health, Bureau
of Sanitary Engineering, made several intensive studies of wastewater
disinfection. As a result of these studies the chlorine residual-
contact time concept was discarded. Some years ago California adopted
the concept of disinfection evaluation based on compliance with a pre-
scribed most probable number (MPN) of coliform organisms. The numbers
now in effect are 80 percent of samples less than 1000/100 ml for
coastal bathing waters (equivalent to a median of 240/100 ml), a median
of 70/100 ml for shellfish growing areas, and a median of 23/100 ml for
confined waters used for bathing or other water contact sports, assuming
the dilution is at least 100 to 1.
An essentially coliform free effluent (i.e., a median MPN not
greater than 2.2/100 ml) is the requirement for discharge into ephemeral
streams, negative estuaries or other areas where the public is exposed
to effluents receiving little dilution. There is a subtle implication
of the necessity for good operation and adequate (secondary) treatment
-------
17
to achieve the 23/100 ml requirement. The severe effluent standard
of 2.2/100 ml implies the necessity of some type of advanced treatment,
for example, filtration or nitrification. Such a water quality coliform
requirement suggests some virus removal or destruction capability for
the system beyond that which normally occurs.
The EPA requirements for disinfection are in general less
stringent than the requirements of California and other states. The
EPA has temporarily committed itself to a 200/100 ml to 400/100 ml
fecal coliform requirement as consistent with adequate' disinfection,
depending upon the classification of the receiving waters. This figure
is comparable to a total coliform concentration of about 2000/100 ml to
4000/100 ml (White 1975).
The stringent requirements for disinfection as set by California
focused much needed attention on the proper design of chlorination
facilities for this purpose. The result of this attention has been
well documented in the literature from 1970-1975. I have concluded that
if a chlorination facility is to be an optimum design it should consist
of: (1) good mixing of chlorine with wastewater (1 to 3 sec); (2) a
minimum of 30 minute contact time at peak flow in a contact chamber with
superior plug flow characteristics; (3) a good chlorine control system
which should include flow proportional plus residual control followed
by effluent residual monitoring; and (4) competent operators who under-
stand the process chemistry and instrumentation.
On the basis of my own research and a personal investigation of
some 60 plants over the past five years, 1 have made the following
observations: A good secondary effluent, dosed in the range of 10 to
15 mg/1, can achieve a total coliform count of 23.2/100 ml on a consistent
basis. If the same effluent is filtered this same dosage will achieve
an MPN of 2.2/100 ml. If this same effluent is nitrified, but not filtered,
it is possible to achieve a 2.2/100 ml MPN by virture of the presence of a
free chlorine residual due to the nitrification process. It follows
that if the secondary effluent is nitrified, coagulated, and filtered
prior to disinfection, virus removal can be achieved. The residuals
from such optimum systems will be on the order of 2 to 4 mg/1.
-------
18
Quite a few plants are operating in this range of dosage and
residual. A surprising number are achieving a total coliform MPN less
than 9/100 ml. There are also a number of plants whose disinfection
systems are inefficient and which, consequently, require dosages up to
25 mg/1 resulting in residuals as high as 8 mg/1. Then there are the
days when the best of plants suffer a biological upset. These periods
require higher doses of chlorine resulting in higher residuals.
It has been convincingly demonstrated that any chlorine residual
greater than 0.05 mg/1 is toxic to a considerable variety of aquatic
life. This has brought about a requirement for dechlorination of these
effluents at the end of the contact chamber. This is now an accomplished
fact in many California plants. The use of sulfur dioxide as a dechlo-
rinating agent has proved satisfactory. It is easy to handle and
utilizes the same equipment and instrumentation as that used for chlorine.
This simplifies the engineer's design problems. Each 1 mg/1 of chlorine
residual requires 1 mg/1 of sulfur dioxide. Mixing is not as critical
as for chlorine and since the reaction is almost instantaneous, contact
chambers are not required.
There are many ramifications in wastewater treatment which have a
profound effect on the disinfection process. Industrial wastes are the
worst offenders. They usually cause an increase in chlorine demand or
render the wastewater septic or both. The latter causes all kinds of
problems. The biological oxygen demand (BOD) and suspended solids do
not have as great an effect on the disinfection process as does the
reduction of colifonns in the process before disinfection. The lower
the concentration of colifonns the easier the wastewater can be
disinfected.
COOLING WATER
Chlorination of cooling water was first tried in the United States
in 1924 at the Commonwealth Edison Company, Chicago, Illinois. Up to
that time power generating stations were plagued with biofouling of the
heat exchangers. This caused plant shutdowns for cleaning the slime
and debris in the condenser tubes. This is a costly and messy business.
-------
19
Now, almost 50 years later the power industry experience has proved
that the chlorination of cooling water is the most efficient and
economical method for the control of biofouling in the water cooled
heat transfer equipment (Cole 1975).
All but about 10 percent of the electric generating plants chlo-
rinate the cooling water on a programmed basis. The few who do not
are in areas where there is excessive scouring by sand in the cooling
water. The total amount of chlorine used in this sector is about
90 x 106 kilos per year (98,915 tons per year).
The most important factor in cooling water chlorination is the
consideration that the cooling water is the vehicle for transporting
the disinfectant to the condenser tubes. Therefore current practice
consists of intermittent doses of chlorine, for example, 1 to 2 mg/1
for 20 to 30 minutes for two or three times each 24 hours. The magni-
tude and frequency of dosage varies with local conditions. There are
two types of systems. The most common is the once-through cooling
water flow (as from receiving water and back again on a continuous basis)
These systems chlorinate on an intermittent basis. The EPA complains
about two things: (1) the chlorine residual in the effluent which may
persist in a plume for two to three hours is objectionable, and; (2)
during the period of chlorination the chlorine is destroying the biota
in the cooling water taken from the source.
Once-through systems now under design are arranging chlorination
so that the effluent from a group of condensers is diluted with the
effluent from the rest of the condensers in order that the residual in
the chlorinated condenser effluent is diluted sufficiently to disappear
by the time it reaches the receiving water.
The other system of cooling water use is the closed recirculation
system using atmospheric or forced draft cooling towers. There are
many advantages to this system from a conservation point of view. How-
ever, a large amount of organic matter does develop on the tower and in
the sludge accumulation of the tower basins. This not only increases
the chlorine demand but prevents the formation of a free chlorine
residual. Therefore practically all tower systems operate on the less
-------
20
efficient chloramine residual. These tower systems do have a distinct
advantage over the once-through systems insofar as protecting the
environment is concerned. This is because the blowdown from the tower
can be turned off during periods of chlorination, or it can be diverted
to the ash sluice system or to a lagoon, or the blowdown can be dechlo-
rinated. The blowdown system on many towers would only have to dechlo-
rinate the 60-minute residual which would be less than 0.2 mg/1 in
most situations.
SUMMARY AND CONCLUSIONS
It is demonstrably clear that there is no alternative to chlorine
as a disinfectant and chemical tool in the treatment of potable water,
wastewater, and cooling water. While there appear to be some disadvan-
tages such as the formation of some undesirable chloro-organics, no
other oxidant can combine all the attributes of chlorine. The popu-
larity of chlorine is deserved because of its potency and wide range of
effectiveness as a germicide, and it is easy to handle, apply, measure,
and control. Most important of all it can be applied to effect a
predictable persisting residual over long distances and extended
periods of time. This turns out to be somewhat of a disadvantage in
wastewater treatment. The other halogens such as iodine, bromine, and
bromine chloride are so much more expensive and cumbersome to handle
that they could only be considered for special situations. Ozone has
some distinct advantages over chlorine, not the least of which is its
superior viricidal efficiency. However, it is such a rapid acting
oxidant that it is impossible to maintain a persisting residual. Any
remaining residual has a very short half-life.
Instead of looking for an alternative for chlorine we should be
investigating the chemical attributes of combinations. For example, in
Europe combinations of chlorine, chlorine dioxide, ozone, and activated
carbon are used, and in Great Britain chlorination is followed by
dechlorination and sometimes this treatment is followed by rechlorination
and ammoniation.
-------
21
We should investigate the difference in efficiency of various
strengths of applied chlorine solutions and at different pH levels.
We should investigate the use of preformed chloramines for
special situations.
We should investigate various combinations of dechlorination with
both sulfur dioxide and activated carbon.
We should investigate all possible combinations of sequences using
chlorine, chlorine-dioxide, and ozone.
Lastly, we should investigate the merit of certain types of
aeration sequences and dechlorination followed by rechlorination with
and without postammoniation.
REFERENCES
American Water Works Association (AWWA). 1974. Special Taste and
Odor Research Committee Meeting. Chicago, Illinois. November 12-13.
Chlorine Institute. 1975. Chlorine — alkali production in North
America, Phamphlet No. 10. New York.
Cole, S. A. 1975. Chlorination for the control of biofouling in
thermal power plant cooling water systems. Presented at Biofouling
Workshop, Johns Hopkins University, Baltimore, Maryland. June 16-17.
Harris, W. C. 1975. Ozone disinfection of the Strasburg, Pennsylvania
water supply. Presented at Pennsylvania Section Meeting, Am. Water
Works Assoc., Champion, Pennsylvania. April 27-29.
Inhoffer, W. R., and F. J. De Hooge. 1974. Free residual chlorination
of Passaic River water at Little Falls, New Jersey. Presented at
Am. Water Works Assoc. Annual Conference. Boston, Massachusetts.
June 16-21.
White, G. C. 1972. Handbook of chlorination. Van Nostrand Reinhold,
New York.
White, G. C. 1974. Unpublished notes from a field survey of water
supplies in France and England.
White, G. C. 1975a. Disinfection: the last line of defense for
potable water. Am. Water Works Assoc. 67: 410-413.
White, G. C. 1975b. Disinfection Committee Report. Presented at Am.
Water Works Assoc. Annual Conference. Minneapolis, Minnesota.
June 8-13.
-------
22
Willey, B. F., C. M. Duke, and J. Rasho. 1974. Chicago's switch to
free chlorine residuals. Presented at Am. Water Works Assoc.
Annual Conference. Boston, Massachusetts. June 16-21.
Williams, D. B. 1974. Presented at Am. Water Works Assoc. Research
Committee on Taste and Odor Control. Chicago, Illinois.
November 12-13.
DISCUSSION
John R. J. Sorenson, Quad Corporation. As I understand it, the
chlorine used in water purification is essentially the dregs of the
chlorine production process. Is that correct? Also, as I understand
it, these dregs contain many different chlorinated aliphatic and
aromatic hydrocarbons including methylene chloride, chloroform, carbon
tetrachloride, hexachloroethane, hexachlorobenzene, etc., which are
being measured in chlorinated water. Could we be mistaken about their
production and they result from impurities in the chlorine itself? I
also have another question concerning the alternative water purification
process, ozonization. Shouldn't there be some concern with regards to
ozonides and epoxides produced in this process and their potential
car cinogenicity?
White. Do you mean are these impurities in the chlorine itself?
Oh, no. Somebody has done a chemical balance study and determined that
none of these chloro-organics are necessarily from the impurities in
the chlorine. So we can't indict the impurities in chlorine, not yet
anyway. Incidentally, you notice that I mentioned the amount of chlorine
that is used in the manufacture of gasoline. Now, since you brought this
subject up, I'll just expand a little on something that is interesting.
If any of you were in Minneapolis, you may have heard Dr. Rook from
Amsterdam. Somebody asked him a question about this catalytic action
of the bromide ion or bromine. The question was, "How does the bromine
get into the environment?" He answered that it was the run-off from the
highways because of the bromine that is used in chlorine additives.
Now bromine was found to be the best lead scavenger in gasoline, and so
this is emitted in automobile discharges. To confirm this idea even
further, the people at East Bay have a catchment high up in the Sierras,
very near a freeway, and they're detecting chloro-organics in the water
before its even treated. So they suspect that the bromide possibly
comes from gasoline additives.
Albert Dietz, Jones Chemical Company. Have you had experience
with the problem of regrowth of bacterial organisms after dechlorination?
White. Yes. There is regrowth after dechlorination. This is, of
course, due to the nutrients that are in the effluent. The health
department's posture on this particular question is that the pathogens
have already been killed. I don't know how valid that is. We have to
play by the rules that are laid down for us.
-------
23
Arthur S. Brooks, University of Wisconsin-Milwaukee. You have
advocated dechlorination in various water treatment processes. Is
dechlorination 100% effective in removing active chlorine species? I
ask the question because biological studies have indicated that chlorine
levels on the order of a few ppb will produce adverse biological effects.
White. So far as I know, it is as complete as will be registered
when you make an amperometric titration. In other words, you cannot
titrate a residual if you are dechlorinating properly and in the
stoichiometric ratio. There is a little loss in the system but the
reaction is very rapid and complete. This is with sulphur dioxide. Now
there are a lot of things we don't know about activated carbon. You
can't say the same for that and I haven't had enough experience with it.
Richard F. Unz, Pennsylvania State University. You stated that
the nonchlorine halogens were too expensive and cumbersome for practical
application in water and wastewater treatment. What is the basis for
this conclusion?
White. Yes. Let's take bromine as an example. The only case of
bromine use in water treatment, that I know of, is in the little town
of Irvington, California, which was very near a West Virginia Chemical
Company's plant. I quess they had bromine to throw away at one time.
The Irvington Water Company was experiencing a lot of troubles with
their distributor system. Their water was supplied by wells. So the
water company made atmospheric brominators using liquid bromine. Two
things happened that were devastating. First, it produced the foulest
tasting water that you could believe, a medicinal taste and odor, and
it couldn't do the job because it reacted so fast with the slime growths
in the distributor system that the residual wouldn't persist more than
800 feet from the well. Then, on top of that, one of the operators
burned himself with the bromine. With the combination of the two, it
just didn't get off to a good start. There has been a lot of experi-
mentation done by Dow Chemical with bromine chloride. I think that
there are other ways to handle bromine although it is very expensive.
I think that there are some possible applications. I'm working with
some people in France that are patenting a device where bromine and
chlorine are used together. It looks like it has some promise for water
reuse because the die-away of the residual in that combination is very
rapid. Iodine is just too expensive.
Unz. Lack of interest in exploring the utility of halogens other
than chlorine for disinfection purposes reflects the perpetuation of
early taboos having little scientific credibility. I hope the con-
ference will not result in an admiration society for chlorine in matters
of pollution control. Iodine, for example, possesses interesting
properties which make consideration of its use in certain applications
highly attractive. It is a poor algicide. However, we have been
working on the discovery of a compatible algicide and have found two
which work well in the presence of iodine and which provide satisfactory
combined bacterial-algae control in test swimming pools.
-------
24
White. Well, my experience with iodine is not the same. I don't
have that same feeling. For example, if you want to learn about a
chemical as a disinfectant and as an algicide, get yourself a swimming
pool and operate it. Now, Dr. Marx patented the Iodine Bank System
whereby you put potassium iodide in a swimming pool and you activate it
by releasing the iodine with intermittant applications of chlorine. It
sounded swell on paper and it was supposed to work. But two things
happened, namely: (1) it colored the water and, (2) it wasn't a good
algicide. They just had to give up on it. Several pools in California
tried it. Now that is not true with bromine. They have bromine sticks
which made it easy to apply. But if you talk to the people in Illinois,
I think you'll find that they say that it is limited to small pools.
I think we first need to learn more about chlorine before we go to
bromine and iodine. That is my feeling. I agree not to shut my eyes
at any of these things.
C. Sengupta, Public Service Electric and Gas Company. I want to
comment on bromine. Our company did some experiments with BrCl. Our
experiments were abandoned for other reasons, but one problem was that
of handling the BrCl. Cost was an important factor, but handling it
was not easy, although it can be done.
White. Yes. There are ways that you can handle bromine that are
very interesting. Now, one of the things that I didn't mention is this,
that when you are chlorinating a cooling water condenser with sea water
and you're putting the sea water through the chlorinator injector, you're
going to release the 60 mg/1 of bromide ion that is in the sea water
and get a hotter residual than if you just used fresh water.
-------
SESSION I. AQUEOUS CHEMISTRY OF CHLORINE
Joseph E. Draley, Session Chairman
Assistant Director
Argonne National Laboratory
Argonne, Illinois 60439
I am impressed by how many people are here who want to hear about
the aqueous chemistry of chlorine. I think for the purpose of the
conference it would be worthwhile saying that we are interested in the
chemistry of chlorine for two reasons. One is to be certain that the
chlorine will do the good things that we would like to have it do. This
includes assuring potability of water supplies, disinfection of waste-
waters, and the defouling of surfaces. You have heard some things
about those already. The other thing we would like to assure is that,
when we discharge from systems, the chemistry of the products is enough
known so that we are able to deal with it. We need to know what the
reactions of chlorine are with the water and with the impurities in the
water. I use the word impurities loosely to mean everything that is in
the water, because they will influence the possibilities of the reactions
that you generally want to occur. We will need to know the reactions
with the system. These have to do with the walls and the components, and
the substances that form on the walls and components. Often times
those are the things that will interfere with what we want and, at other
times, they will be the very reactions that we want. Again, with respect
to discharging safely and acceptably, we will need to know toxicities of
substances that formed because chlorine was used. I have chosen those
words in an effort to avoid saying chlorinated compounds. In other
words, I think that the subject is very broad and I do not want to use
any specific language. So we need to know toxicities and persistencies
25
-------
26
of these substances. It is really to that kind of question, to those
questions, that we address our attention in today's split session on
chemistry.
-------
THE CHEMISTRY OF AQUEOUS CHLORINE IN
RELATION TO WATER CHLORINATION
J. Carrell Morris
Division of Engineering and Applied Physics
Harvard University
Cambridge, Massachusetts 02138
ABSTRACT
When chlorine is dispersed in water, a variety of molecular and
ionic species is produced, including H20C1"1", HOCl and OC1~ as well as
Cl2« Of these the reactively dominant species for most aqueous chlorina-
tion reactions is HOCl. Other species generally are present in too
small a concentration or have specific reactivities too low to be
significant.
The HOCl may act as an electrophilic reagent at either the oxygen
or the chlorine atom. When reaction is at oxygen, chloride ion is
formed by displacement, as has been shown for certain inorganic oxida-
tions. For reactions at amine-N or at carbon the electrophilic attack
is by the chlorine atom which acts as Cl+. Reactions of chloramination,
of chlorophenol formation or other aromatic substitution, of addition
to double bonds, and of haloform formation are all examples of this
form of electrophilic attack.
It is well known to almost everyone, I presume, that the term
aqueous chlorine is a misnomer when it is applied to the usual conditions
of water and wastewater treatment. For, despite the statements in
numerous elementary chemistry textbooks that elemental chlorine, Cl2>
is just partially or only slightly hydrolyzed in water, already at a
formal concentration of 10~3 M in pure water C12 is hydrolyzed more than
99%. Moreover, the hydrolysis increases with decreasing concentration or
with neutralization of liberated H+ to greater pH values.
27
-------
28
The hydrolysis is in accord with the equation
C12 + H20 = HOC1 + H+ + Cl" (1)
with hydrolysis constants ranging from 1.5 to 4.0 x 10""^ (mol per liter)2
for the temperature range from 0 to 25 °C (Jakowkin 1899; Connick and
Chia 1959). Also, the hydrolysis is rapid, with equilibrium conditions
being established within a few seconds or less (Shilov and Solodushenkov
1945; Eigen and Kustin 1962). So, the major oxidizing species present
in dilute acidic "aqueous chlorine" solutions is hypochlorous acid, HOC1.
Hypochlorous acid is a weak acid that dissociates in accordance with
the equation
HOC1 = H+ + OC1~ (2)
with a dissociation constant ranging from 1.6 to 3.2 x 10~8 for the
temperature range, 0 to 25°C (Morris 1966). At pH values in excess of
7.8 to 7.5 for 0 to 25°C hypochlorite ion becomes the dominant species
in aqueous chlorine solutions.
Table 1 shows the distribution of principal oxidizing species for
aqueous chlorine solutions at even pH values between 5 and 9 for a tem-
perature of 15°C with a chloride content equal to 350 mg/1 (10~2 M).
The fraction present as C12 is only in the range of parts per million for
most of the conditions tabulated, as can be seen. Even the presence of
0.5 M chloride in seawater, as shown in the last line of the table,
increases the fraction of C12 only to about 10 parts per million of
total oxidizing chlorine.
Other transient species that have been considered to be formed in
some circumstances in aqueous chlorine solutions are H2OC1 , Cl and
C13~. Only the first of these, H2OC1 , seems likely to be of signifi-
cance in the dilute aqueous media characteristic of water and wastewater
treatment. This H2OC1+, however, is the species probably responsible
for acid catalysis of many reactions of HOCl.
-------
29
Table 1. Distribution of aqueous chlorine species, 15°C
PH
5
6
7
8
9
7.8
pCl
2
2
2
2
2
0.3
Fraction of
Cl2(xl06)
360
36
2.9
0.10
0.001
11
Oxidizing
HOC1
0.997
0.975
0.797
0.280
0.038
0.382
Chlorine as
OC1~
0.003
0.025
0.203
0.720
0.962
0.618
The issue of the major reactive pathway or species, however, is
not determined solely on the basis of the constituent that is present
in predominant concentration. There is also the question of the specific
reactivity of each of the forms , for generally it is the product of
concentration times specific reactivity that determines the contribution
of a given mechanism or pathway to the overall reaction.
Define and express the reactivity of a given reaction pathway by
the equation
Reactivity - R. = r.C. (3)
where C. is the concentration of a particular species and r^ is its
specific reactivity toward the substrate under consideration. The T^
are, of course, equivalent to specific reaction rates with the inclusion
of concentration of substrates other than the Ci.
The overall reaction rate in a given situation is then the sum of
the reactivities of the different pathways, that is
Rate = 2R± = Sr^ . (4)
Now, if this overall rate is divided by the total stoichiometric con-
centration of reactant, the quotient, which may be termed the observed
specific reactivity, r, exhibits the relationships
-------
30
where the f. are the fractions of the total stoichiometric reactant con-
centration in specific molecular forms. Variations in r with pH, ion
content and solution conditions other than temperature and pressure may
then be associated with changes in the f., so long as the substrate
remains unchanged.
Unfortunately the r. are not universal constants; their absolute
and even their relative values depend upon the substrate with which
reaction is occurring. So, no widely applicable tables of absolute or
relative specific reactivities can be constructed. Nonetheless, when
similar types of reaction are under consideration, relative specific
reactivities often maintain themselves sufficiently fixed that at least
approximate extrapolation from one substrate to another is feasible.
Table 2 presents a sample evaluation of relative reactivities for
several forms of oxidizing chlorine in dilute aqueous solution at pH 7.
The tabulated specific reactivities are relative to r„„,-,, = 1; they are
rlULJ.
estimated values for exemplification, based principally on the reactiv-
ities of forms of aqueous chlorine toward nitrogenous compounds, and
are not to be regarded seriously or universally in any quantitative
sense. However, the order of reactivities is likely to remain the same
for many reactions in which electrophilic or oxidizing properties of the
forms of aqueous chlorine are involved.
Table 2. Estimated net reactivities of forms of
active chlorine; pH 7, 15°C
Species
C12
HOC1
OC1~
H2OC1+
Estimated
specific
reactivity
103
1
ICT1*
10 5
Fraction of
total Cl
3 x 10~6
0.80
0.20
10~8
Net
relative
reactivity
0.003
0.80
0.00002
0.001
-------
31
It is clear from a table such as this that HOC1 must be regarded
as the major reactive species for most oxidizing reactions of aqueous
chlorine in dilute solutions (<10~3AO at pH values between 5 and 9.
This is very different from the situation in typical nonaqueous solvents
where very little hydrolysis occurs and where the dominant reactive form
is C12. It is also very different from strongly acidic solutions where
C12, H2OC1+ or even Cl may dominate and from strongly basic solutions
where essentially only OC1~ is present.
So, the types of reactions with organic substances found for
chlorine in organic solvents, where C12 is dominant and free radical
reaction mechanisms are common, are not to be expected in dilute aqueous
solution unless photochemical generation of radicals occurs. Instead
the reactions and reaction patterns occasioned by the electrophilicity
of HOC1 are to be anticipated.
REACTIVITY OF HOC1 WITH INORGANIC ANIONS
Either the chlorine atom or the oxygen atom of HOC1 may act as a
center of reaction. It appears that the Cl is the more electropositive
atom, so that electrophilic processes proceed by way of the chlorine
atom. On the other hand, the attraction of the Cl for electrons is so
great that it may split from the molecule directly as chloride ion.
An instance is the reaction of nitrite with aqueous chlorine
according to the elementary reaction
N02~ + HOC1 -*• N02OH + Cl~ (6)
with the nitrite displacing Cl~ (Anbar and Taube 1958; Lister 1961).
The oxidation of sulfite may proceed similarly, at least in part
(Halperin and Taube 1952).
Somewhat similar transfers, with displacement of either Cl or
I i i I
OH", may occur as initial steps in the oxidations of Fe and Mn and
in the reaction with H202. The situation with regard to the exact
elementary step is less clear in these latter instances, however.
-------
32
HOC1 AS ELECTROPHILE
The usual reactant behavior of HOC1 with organic carbon and with
amino nitrogen is as an electrophilic agent in which the chlorine
atom takes on partially the characteristics of Cl and combines with an
electron pair in the substrate. Simultaneously or subsequently the
hydroxyl ion is split off, often with the assistance of H4" from the
solvent or of reactive centers from other parts of the substrate.
This type of behavior serves to account fundamentally for the
reactions of dilute aqueous chlorine with ammonia and amines, with
phenolic and other aromatic substances, and in the formation of chlo-
roform from organic substrates. Each of these classes of reaction will
be considered in some detail in the following sections.
Clearly, Cl+ itself is the strongest electrophilic species, related
to HOC1, but it occurs only when the solvent has dehydrating properties.
The ion, H^Od"*", is also a stronger electrophile than HOC1, partly
because of its positive charge, and partly because of the easier release
of H£0 as compared with OH~.
Reaction of electrophilic Cl can be viewed analogously to reactions
of H*. Naturally, transfer of Cl+, like that of H+, is facilitated by
negative charge, basicity or nucleophilicity of the recipient atom.
REACTIONS WITH AMMONIA AND AMINES
Reaction of HOC1 with the nitrogen atom of ammonia and amines
follows the pattern that has just been described. The electrophilic
chlorine atom can be visualized as attaching itself to the bare electron
pair of the nitrogen atom with concurrent release of H"1" from the ammonia
or amine and of OH~ from the HOC1. These latter ions would appear to
react forming water with the participation of solvent water as a part
of the overall pattern of the single elementary process.
Conversely, the reaction may be viewed as a displacement reaction
in which the nucleophilic ammonia or amine displaces OH~ from HOCl.
The nucleophilicity of NR£~ as compared with OH~ is then of importance.
In accord with either of these models of the reaction, the specific
rates of chloramination have been found to vary directly with the basic
-------
33
strength (or nucleophilicity) of the nitrogenous substrate (Morris 1967).
Catalysis by H"*", indicating participation of the more electrophilic
H20C1+, has been observed only when the basicity of the receptor
compound is very low.
Although amides appear also to accord with the previous reaction
pattern with regard to basicity and overall rate of reaction, there is
another possible mechanism available to these latter compounds. Since
acidic dissociation of amides occurs relatively readily to give anionic
structures resembling those of enolates, it is possible that chlorina-
tion of amides proceeds by a pattern similar to that of the haloform
reaction.
REACTIONS WITH THE AROMATIC RING
That dilute aqueous chlorine reacts with phenolic compounds to
form chlorinated derivatives has been known for a long time. Already
by 1926 Soper and Smith had shown that the mechanism of the reaction is
electrophilic attack of HOC1 on the phenoxide ion. Here, as in other
attacks by electrophilic agents, reaction is facilitated by the presence
of negative charge on the nucleophilic substrate.
Benzene itself and many of its derivatives require a more active
electrophile that HOC1, such as Cl+ or ^OCl*, for measurable chlorina-
tion at ambient temperature. Only when the ring is "activated" by
substituents such as the oxide ion does chlorine or other halogen
substitution occur easily. It then may proceed at the activated ortho
and para positions until these are occupied fully by halogen atoms;
following this, rupture of the aromatic ring may occur by mechanisms
that are not as yet fully understood.
The accepted detailed mechanism involves, first of all, an addition
of Cl"*" to the aromatic ring to give a transitory intermediate which, in
the formation of p-chlorophenol, for example, can be represented by the
formula
-------
34
0 =
This then becomes ClCgH^O by loss of H~*~ from the para position.
When there is more than one activating group in meta positions on
the aromatic ring, then reaction with HOC1 should be much accelerated.
There is evidence in the very recent work of Rook that such accentuated
reactivity is observed with resorcinol, /rz-dihydroxybenzene, and that
the reaction may lead ultimately to the production of chloroform after
ring rupture. He found, for example, that millimolar resorcinol
treated for 4 hours at pH 7.5 and 10° C with 8 millimolar aqueous chlorine
gave 0.38 moles of CHCls per mole of resorcinol (Rook 1975).
Heterocyclic aromatic rings may be either activated or inactivated
toward electrophilic attack as compared with benzene. Pyridine, for
example, is greatly deactivated, so that reaction in the ring with
HOC1 in dilute aqueous solutions is not to be expected. On the other
hand the a-hydrogens of pyrrole are greatly activated, so that electro-
philic chlorine substitution by HOC1 in dilute aqueous solutions is
quite likely (Sykes 1961).
ADDITION TO DOUBLE BONDS
Addition of halogens or hypohalous acids to double bonds of organic
compounds in polar solvents like water is another example of electro-
philic attack. In this case attack by the hypochlorous acid is more
vigorous than that by Cl2, for the chlorine in the former is more
electropositive.
The initial rate-determining step is believed to be transfer of
Cl+ to the double bond to give a chloronium ion, as shown by
C = C + C10H -> C - C + OH~ (7)
-------
35
Reaction is completed by addition of OH~ or other anion from the solvent
at one or the other of the chloronium bonds.
Reactions of this sort appear to be too slow in general to be of
significance in water or wastewater chlorination unless the double bond
is strongly activated by substituent groups. The dynamics and products
of HOC1 reactions of this sort with conjugated molecules like terpenes,
carotenoids and xanthophylls needs investigation.
THE HALOFORM REACTION
Reaction of aqueous hypohalites with methyl ketones or compounds
oxidizable to methyl ketones to yield haloform is a classic reaction of
organic chemistry that has been known since 1822 (Fuson and Bell 1934).
Two forms of the reaction are known, acid-catalyzed and base-catalyzed.
The base-catalyzed reaction pattern is the one that is predominant for
reactions in dilute aqueous solution at pH >5.
The pattern of the reaction is the successive replacement of hydrogen
by chlorine on carbon alpha to a carbonyl group followed by eventual
hydrolysis to produce CHX3 and, generally, a carboxylate. The mechanism
is believed to be an initial proton dissociation from the a-carbon,
giving an enolate carbanion which is then subject to electrophilic attack
by HOC1 or OC1~. The rate of proton dissociation is ordinarily the slow,
rate-determining step for the conditions under which the reactions have
been studied extensively. It should be noted that most investigations
have either been at pH >11 or with quite strong buffer solutions. There
are few results in simple aqueous media.
Figure 1 gives a detailed diagrammatic representation of the course
of the haloform reaction with hypochlorous acid and acetone. The
electrophilic attack of HOC1 on the enolate ion is very similar to the
attack on the phenoxide ion. Indeed, as was pointed out by Bartlett and
Vincent in 1935, "It would seem, then, that the chlorination of aliphatic
enols and that of phenols are essentially alike, differing only in the
magnitude of some of the constants."
-------
36
-C-CH, ^[RH?-^— R-?=CH J
fast
H20 • R-C—
CX3
Fig. 1. The Haloform Reaction
It seems peculiar, in view of this statement and the well-known
formation of chlorophenols in water chlorination, that no one until
Rook (1974) should have looked for or found the haloform reaction
occurring in connection with water chlorination.
Actually, the reaction of aqueous chlorine with acetone or other
simple methyl ketones is too slow at lO"4 M concentration and pH values
of 5 to 9 to account for any of the observed chloroform formation in
water chlorination. Moreover, the oxidation of simple secondary
alcohols to methyl ketones is also slow for these concentration and pH
values.
More highly activated structures than those of simple methyl ketones
are required to account for the formation of chloroform in water chlorina-
tion. These must involve more acidic carbons that those alpha to a
single carbonyl group, ones for which dissociation of a proton is
several orders of magnitude faster than from a simple alpha carbon.
Acidic carbons like this are found in methylene groups between two
carbonyl groups, such as that in acetyl acetone, CH3COCH2COCH3. Rook
-------
37
(1975) has investigated three such compounds: indanedione,
cyclohexanedione-1,3, and 5,5-dimethylcyclohexanedione-l,3. All of
them produced substantial chloroform, 0.38 to 0.52 mole per mole of
substrate, when allowed to react with 6 to 8 millimolar aqueous chlorine
at pH 7.8 and 10°C for 4 hours. The mechanism of rupture to yield
chloroform after chlorine substitution is still obscure.
Other acidic carbons that yield carbanions without forming an
enolate structure are also subject to electrophilic attack by HOC1.
An instance is chloroform itself, which is a relatively strong carbon
acid. Following loss of proton, the CC13~ carbanion is susceptible to
electrophilic attack by HOCl to yield CClt,. However, it appears that the
rate of this reaction is too slow to be of significance in water
chlorination.
EFFECTS OF BROMIDE ON AQUEOUS CHLORINE REACTIONS
It has become apparent in the past few years that bromide is a
common constituent of natural waters at concentrations in the range of
fractions of a part per million. The importance of these minute con-
centrations of Br~ is great, however, especially in connection with
the reactions of aqueous chlorine.
Whenever chlorine or hypochlorite is added to a water containing
Br~, there is rapid formation of HOBr according to the reaction
Br~ + HOCl -»• HOBr + Cl~ . (8)
The resulting HOBr is also an electrophilic agent, but one that tends
generally to react much more rapidly than HOCl. Much of the HOBr that
reacts will be reduced to the bromide ion, to be reoxidized to HOBr
by residual aqueous chlorine in the water. The overall result is an
enhanced reactivity exhibited by aqueous chlorine in the presence of an
original concentration of bromide ion.
It is quite likely, therefore, that there will be apparent incon-
sistences in the reactions of different water supplied and wastewaters
with aqueous chlorine, depending on the bromide content of the waters.
-------
38
Already it appears that the rate or pattern of the breakpoint reaction,
in which halogenation of ammonia is involved, may vary for different
water supplies depending on the concentration of bromide. Effects on
reactions with organic carbon may also be expected. It may be noted that,
although the absolute rates of haloform formation depend on the rates of
carbanion formation, the relative amounts of bromine and chlorine sub-
stituted into the haloforms depend on the relative races of the halogena-
tion reactions.
There are two types of needed information. First, there should be
efforts to determine the bromide content of numerous waters to provide
a background for assessing observed differences in reactivity with aqueous
chlorine. Second, there should be examination of the reactivity and
reaction pattern of several waters when spiked with small concentrations
of bromide to determine the practical catalytic effect of this ion on
the reactions of aqueous chlorine.
CONCLUSION
The most important conclusion to be drawn from this paper is the
observation that the reactions of aqueous chlorine in water chlorination
are not indiscriminate and unpredictable, but rather that they follow
quite well-defined pathways in accord with general principles of organic
reaction mechanisms.
So, even when the exact composition of the organic material in a
water or wastewater is not known, it is still possible to predict
something of the nature and extent of the reactions with aqueous chlorine
to be anticipated.
REFERENCES
Anbar, M., and H. Taube. 1958. The exchange of hypochlorite and hypo-
bromite ions with water. J. Am. Chem. Soc. 80: 1073-1079.
Bartlett, P. D., and J. R. Vincent. 1935. The rate of the alkaline
chlorination of ketones. J. Am. Chem. Soc. 57: 1596-1600.
Connick, R. E., and Y. Chia. 1959. The hydrolysis of chlorine and its
variation with temperature. J. Am. Chem. Soc. 81: 1280-1285.
-------
39
Eigen, M., and K. Kustin. 1962. The kinetics of hydrolysis of C12, Br2
and I2. J. Am. Chem. Soc. 84: 1355-1358.
Fuson, R. C., and B. A. Bull. 1934. The haloform reaction. Chem. Rev.
15: 278-309.
Halperin, J., and H. Taube. 1952. The reaction of halogenates with
sulfite in aqueous solution. J. Am. Chem. Soc. 74: 375-379.
Jakowkin, A. A. 1899. On the hydrolysis of chlorine. Z. physik.
Chem. 29: 613-657.
Lister, M. W., and P. Rosenblum. 1961. Kinetics of the reaction of
hypochlorite and nitrite. Can. J. Chem. 39: 1645-1653.
Morris, J. C. 1966. The acid ionization constant of HOC1 from 5 to
35°C. J. Phys. Chem. 70: 3798-3802.
Morris, J. C. 1967. Kinetics of reactions between aqueous chlorine
and nitrogenous compounds, p. 23-:53. In S. D. Faust and
J. V. Hunter, Principles and Applications of Water Chemistry.
John Wiley and Sons, Inc., New York.
Rook, J. J. 1974. Formation of haloforms during chlorination of
natural waters. Proc. Soc. Water Treatment and Exam. 23: 234-243.
Rook, J. J. 1975. Formation of haloforms during chlorination.
Presented at Am. Water Works Assoc. Annual Conference, Minneapolis,
Minn., June 9-13.
Shilov, E. A., and S. M. Solodushenkov. 1945. The velocity of hydrolysis
of chlorine. J. Phys. Chem. (U.S.S.R.) 19: 405-407.
Soper, F. G., and G. F. Smith. 1926. The halogenation of phenols.
J. Chem. Soc. 1926: 1582-1590.
Sykes, P. 1961. A guidebook to mechanism in organic chemistry.
John Wiley and Sons, Inc., New York. 247 p.
DISCUSSION
David H. Rosenblatt, U.S. Army Medical Bioengineering Research and
Development Laboratory. I wonder, with all this interest now about
chlorination of water producing small chlorinated hydrocarbons and plural
brominated hydrocarbons, whether anybody has given thought to what
might happen when you brominate the same type of water. Would you in
fact get bromination in the absence of chlorine? Also, things might be
a little more complicated than you show here. You're going entirely by
two electron reactions. The question in my mind is whether there aren't
some free radical paths that would require the presence of HOC1 or other
active chlorine species in order to introduce bromine into the molecule?
-------
40
Morris. So far as I can tell, when one is operating in these
dilute aqueous media, the occurence of free radical mechanisms is
relatively remote. The tendency is almost always for a split with
these hypohalous acid molecules, with one being positive and one
negative.
Robert B. Dean, U.S. Environmental Protection Agency. I think it's
worthwhile considering that in real chlorination reactions, adding
chlorine to a little bit of water produces a low pH, a lot of local
acidity. Now are the reactions at the acid site fast enough to be
useful for substitution, in other words, to attack the organic compounds
via the acid flow before that chlorine gets neutralized by the alkalinity
of the water?
Morris. I think that depends on whether you have used Cliff White
as your engineer in designing your chlorination facilities or not. If
you designed it to mix the chlorine with the water very rapidly, within
a period of 1 to 3 seconds, then there is not time for any extensive
reactions to occur. But, if you simply bleed the chlorine in and let
it drift in a plume downstream, almost anything could happen.
Herbert S. Posner, National Institute of Environmental Health
Sciences. You mentioned three reactions that would yield chloroform:
(1) the degradation of resorcinol; (2) the successive chlorination of
methyl ketone; and (3) substitution into a molecule like benzylacetone
or decenylacetone. Does this exhaust the possibilities or are there
other possibilities?
Morris. It does not exhaust the possibilities except when I gen-
eralized and said I thought that any time you got a carbanion you could
begin to get the substitution. Then, whether or not this will eventually
result in chloroform, depends on the structure of the rest of the mole-
cule. So you would have to look at the individual molecule to see
whether it could eventually end up as chloroform or not.
George Clifford White, Consulting Engineer. Dr. Morris, I am very
interested in this oxidation of the bromide ion situation. I have done
a lot of reading about it. You say that the bromide ion is readily
oxidized by hypochlorous acid. I am under the impression that a
stoichiometric reaction of oxidizing bromide ion with hypochlorous acid
only occurs if the pH is below 5, and that you have to have an excess of
bromide ion to make the reaction go at all if the pH is near neutral.
Is that correct?
Morris. When we were doing some studies on the bromamines, the
experimental and theoretical work was carried out which indicated that
this reaction of hypochlorous acid with bromide increased with decreasing
pH, or increased linearally with the hydrogen ion concentration of the
solution and became more rapid than the reaction of hypochlorous acid
with ammonia at a pH of about 7.8. So on the acid side of pH 7.8, you
-------
41
would expect the reaction to be more rapid than the reaction with
ammonia. Thus if you put hypochlorous acid into say equal concentra-
tions of bromide and ammonium ion, you would form hypobromous acid
rather than forming chloramine in such a situation.
White. But how about low concentrations of bromide, say 10 to 15
mg/1, in a similar or equal concentration of chlorine?
Morris. The absolute concentration would make no difference. It's
only the relative concentration that would make a difference. You would
have to go a pH unit lower to get predominance if you had had 1/10 as
much bromide as ammonia. Every power of 10 would reduce you that much.
But, otherwise, it's very rapid.
Walter J. Blogoslawski, National Marine Fisheries Service. In view
of the presence of about 50 to 55 ppm bromide in seawater, with a pH of
about 7.8 to 8.0, would you feel that there is a significant reaction of
hypochlorous with that bromide in forming a hypobromide compound?
Morris. Yes, I would.
Blogoslawski. Would you feel that would be significant from a
power plant which might be considering chlorinating large volumes of
seawater in effluent?
Morris. Yes, indeed I would.
-------
MEASUREMENT AND PERSISTENCE OF CHLORINE RESIDUALS IN NATURAL WATERS
J. Donald Johnson
Department of Environmental Sciences and Engineering
School of Public Health, The University of North Carolina
Chapel Hill, North Carolina 27514
ABSTRACT
Selective measurement of only the good disinfectant chemical species
of chlorine will minimize the concentration of chlorine which must be
added to produce a microbiologically safe water. The lower dosage and
better control permitted by selective analytical methods minimizes the
formation and persistence of toxic chlorination products, while destroying
pathogenic microorganisms.
Chlorine residual, as normally measured, is the concentration of
all oxidizing agents produced by chlorination of natural water and
remaining after sometime. These oxidizing agent products of chlorination
may contain no chlorine but are still measured and referred to as chlorine
residual. Of those which do retain chlorine in the plus one oxidation
state, all have radically different ability to disinfect. The really
good disinfectant, hypochlorous acid, HOC1, is never measured selectively.
New methods are needed to distinguish HOCl from the poor disinfec-
tant hypochlorite, OC1~, in drinking water. Monochloramine, NH2C1, is
the good disinfectant in wastewater and poor quality cooling waters.
N-chloro organics are poor disinfectants measured along with NH2C1. The
formation of toxic residuals and their persistence results from the
excess chlorine required for reliable disinfection because our analytical
methods are not selective.
The nature and persistence of chlorination products classed as
chlorine residuals are discussed by type of compound. Actual measure-
ments of chlorine residual decay for free and combined chlorine in water,
wastewater, and estuarine water are discussed. The decay of chlorine
residuals by apparent first order specific rate constants of 0.03 hr~
to 7.5 hr"1 are related to residual type, demand, volatility and photo-
chemical effects.
43
-------
44
Present field, laboratory and continuous methods for free and com-
bined chlorine residual are compared for specificity, reagent stability,
accuracy and simplicity. Included in this discussion are the acid
orthotolidine methods, DPD, SNORT, LCV, FACTS, amperometric titration,
copper-gold amperometric cells and the NBS flux monitor.
A new analytical method specific for HOC1 or NH2C1 in the presence
of the poor disinfectants OC1~, organic chloramines and other inter-
ferences will be presented. The advantages and disadvantages of this
method will be discussed in terms of selectivity, sensitivity, and other
effects in disinfection efficiency measurements.
INTRODUCTION
Chlorine residual measurements are made to determine the efficiency
of disinfection, the objective of chlorination. If only the good dis-
infectant species of chlorine are measured, the disinfection process can
be done with small concentrations of chlorine. Low dosages, resulting
from careful control of the disinfection process, will minimize the
formation of toxic chlorination byproducts. For this purpose, selective
measurement of the good disinfectant HOC1, hypochlorous acid, is needed
to the exclusion of OC1~, hypochlorite ion, and other poor disinfectants
commonly measured.
Residual measurements are also needed for toxic chlorine compounds
particularly NH2C1, monochloamine. Monochloramine is important because
of its toxicity, rapid formation rate, stability and widespread appear-
ance due to the ubiquity of ammonia in natural water.
Monochloramine is, also, the principal chlorine disinfectant in
any water containing as much as 0.1 mg/liter NH3~N and less than 1.0 mg/
liter Cl2» Disinfection control in water of this type requires selective
measurement of Nt^Cl to distinguish it from other chlorine oxidizing
agents formed in natural water which are less persistent and are poor
disinfectants.
False chlorine residuals are commonly measured, especially in the
methods available for monochloramine. These methods are universally
based on the reaction of monochloramine with iodide to give iodine. The
selectivity of this approach is particularly poor fpr low concentration
levels at which monochloramine is toxic to fish and, also, in the low
-------
45
quality waters in which this measurement is important for determination
of environmental toxicity. It is this problem of poor selectivity in
monochloramine measurements which makes interpretation of much of our
fish toxicity data difficult. Lack of selectivity and sensitivity makes
control of a maximum combined chlorine residual for the purpose of
limiting toxicity to fish nearly impossible.
This paper reviews measurements of persistence and decay for free
and combined chlorine residual in both estuarine and fresh water and,
also, in the dark and sunlight. The selectivity of analytical methods
for chlorine to determine disinfection efficiency and environmental
toxicity are discussed. The meaning of chlorine residual is presented
in terms of what we should be measuring as contrasted with what we are
currently measuring. Common interferences will be discussed by inter-
ference type and effect of procedure. The basis of the iodide procedure
for measurement of various combined chlorine forms is described and
new data presented. Electrode methods, both laboratory and field pro-
cedures, are reviewed and a new membrane electrode method of obtaining
analytical selectivity is described for HOC1 and NH^Cl amperometric
measurement.
FREE "AVAILABLE" CHLORINE
When added to water, chlorine gas reacts rapidly and practically
completely with water to form a mixture of HOC1 and OC1~ depending on
the pH, temperature, and dissolved solids or chlorinity (Morris 1966,
Sugam and Helz 1975). This mixture is referred to as free available
chlorine as shown below:
FAC = HOC1 + OC1~ . (1)
Free available chlorine (FAC) is hypochlorous acid plus hypochlorite
ion. These two forms of free available chlorine are produced when either
chlorine gas, Cl2» or hypochlorite solution, NaOCl, or solid, Ca(OCl)2
are added to water:
-------
46
C12 + H20 + HOC! + H+ + Cl~ , (2)
OC1 + H20 - HOC1 + OH . (3)
A small amount of acid, H , is produced in dissolving chlorine gas (2)
and a small amount of alkali, OH~, comes from adding hypochlorite to
water (3). However, the quantities are small compared to the HC03 con-
centration of natural water. These small amounts of acid and alkali
in equations (2) and (3) are neutralized by HCOa. The pH of the solu-
tion determines the proportion of the FAG present as HOC1 or OC1~ as
shown by equilibrium (3) or (4):
HOC1 - OC1~ + H+ . (4)
The FAC is half HOC1 and OC1~ at pH 7.5 and 25°C, while at 0°C the
50:50 value is at pH 7.9. At higher pH values the FAC shifts to more
OC1~ and at lower pH to more HOC1 for the same concentration of FAC.
HOC1, REAL FREE CHLORINE, THE ONLY GOOD DISINFECTANT
Figure 1 shows the minimum safe bactericidal free chlorine residual
necessary after ten minutes at 20 to 25°C or 70 to 77°F (Am. Water Works
Assoc. 1973). The basis of this data is the work of Butterfield at the
U.S. Public Health Service (USPHS) as reinterpreted by the National
Acacemy of Sciences. Note that disinfection requires twice as much
chlorine at pH 8 as it does at pH 7, while it takes 4 times as much at
pH 9. The reason higher concentrations of free chlorine, FAC, are
required is the shift from HOC1 to OC1~ at higher pH values, making it
necessary to increase the level of FAC to maintain a constant level of
the only good and effective disinfectant hypochlorous acid, real free
chlorine. The other form of FAC, hypochlorite ion, is not a good dis-
infectant and is not "available" to kill microorganisms. But it is as
strong an oxidizing agent and so is available to the analytical reagents
for measuring free available chlorine.
-------
~ 1.0
o>
v»
JVI Q Q
o
o»
E
0.6
Q
O
0.4
-------
48
Figure 2 shows the increasing concentration, C , of FAC, as HOC1
+ OC1 , required to maintain a constant level of hypochlorous acid, the
effective disinfectant form of FAC, as pH increases (McClanahan 1975).
The lines are calculated from the equilibrium shown in equations 3 and
4 so as to keep the HOC1 level constant, independent of pH. As more
OC1 is formed at the higher pH values as shown by the positive slope of
the concentration of OC1 line above pH 6, more free chlorine, CT, is
required to maintain a constant level of HOC1. The data points plotted
on the diagram are the actual free chlorine, FAC, measurements of
residual necessary to produce 99.6% virus inactivation, 99.999% coliform
kill, and 99.999% cyst disinfection. This comparison shows that disin-
fection is not pH dependent in the neutral range but that we have been
measuring the wrong chemical or rather sum of chemicals, HOC1 + OC1~.
If we could measure only real free chlorine, HOC1, and do not confuse
it with hypochlorite in a FAC measurement, we would not need to change
the concentration of residual HOC1 required as a function of pH.
Table 1 summarizes the primary factors affecting disinfection
efficiency. As we normally measure the sum of HOC1 + OC1~ as FAC, free
available chlorine, higher quantities of chlorine residual are required
at higher pH and lower temperature. The effect of temperature is a
combination of the shift in the HOC1 — OC1~ equilibrium constant for
equation (3) with temperature and the increase in the quantity of HOC1
required because disinfection is slower at low temperature. These two
effects are in opposite directions. As temperature decreases the pK for
HOC1 ionization increases from 7.5 to 7.9 increasing the HOC1 fraction at
given pH. At the same time the lower temperature markedly increases the
concentration of disinfectant required. This produces a complex total
effect when FAC is measured. If HOC1, real free chlorine, is measured
the temperature effect is simply the slower rate of kill at lower tempera-
ture requiring higher concentration.
COMBINED AVAILABLE CHLORINE, CAC
In wastewater and even in many drinking and river waters sufficient
ammonia is present to make it difficult to add enough chlorine to produce
-------
49
(T
Z
UJ
O
0 CYSTS 22-25°C
7 -
pH
Fig. 2. Log diagram for the system HOC1-H20 with the concentration
of HOC1 constant. Data points show effect of pH on the concentration
required to inactivate virus (99.6%), E. Coli (99.999%) and E. Histolytica
cysts (99.999%).
-------
50
Table 1. Primary factors affecting disinfection efficiency
1. Residual FAC measured
pH effect
Temperature effect — complex
2. Residual HOC1 measured
No pH effect
Temperature effect — simple
free available chlorine. Combined available chlorine, CAC, forms
rapidly from pH 7 to 10 to give NH2C1, monochloramine:
HOC1 + NH3 -»• NH2C1 + H20 . (5)
Reaction 5 may proceed further to replace another IT*" from the nitrogen of
ammonia to give dichloramine, NHC12, and trichloramine, NC13- These
latter compounds are formed more slowly, at higher concentrations of
chlorine relative to the ammonia present and at lower pH. The slowness
of their formation and the instability of NHC12, dichloramine, make
trichloramine concentrations generally very small. The instability of
NHC12 to oxidation to nitrogen gas and nitrate produce the effect known
as the breakpoint. This breakpoint reaction is so fast above pH 7.5 that
NHC12 is not found because it is oxidized as rapidly as it is produced.
Figure 3 shows a typical curve of chlorine residual as a function
of chlorine added or dosage in the neutral to alkaline pH range. The
first bit of chlorine added to any water is reduced to chloride or
unavailable chlorine by its reaction with chlorine demand. By definition
chlorine demand is the sum of the reducing agents producing this reaction
with chlorine in given time.
This initial loss of disinfectant shown in Fig. 3 is due to easily
oxidized organic compounds, ferrous ion, sulfide, nitrite and other
reducing agents. After the demand is satisfied, combined available
chlorine is formed that is primarily monochloramine, NH2C1. As addi-
tional chlorine is added beyond that needed to react with ammonia
nitrogen (5 mg Cl2/liter for each 1 mg NHs-N/liter) dichloramine, NHCl2»
-------
51
UJ
z
QC
s
o
UJ
QL
UJ
CHLORINE DOSAGE
Fig. 3. Relationship between chlorine dosage and residual chlorine
for breakpoint chlorination.
-------
52
begins to form and decompose. This decomposition results in the loss of
residual CAC, until the breakpoint occurs (near 8 mg Cl2/liter per mg N)
as the ammonia present in the CAC is oxidized to N2 and NO3 and the
chlorine becomes chloride, Cl . Beyond the breakpoint excess FAC, free
chlorine, is added and remains in solution except as it is consumed by
difficult to oxidize organic and inorganic compounds. Seldom does the
breakpoint actually return to the zero residual chlorine level since CAC
includes difficult to oxidize chloro-organic compounds, RNC1 where R
z
represents many possible organic compounds. Figure 4 shows monochloro-
glycine, curve B, is measured as monochloramine. Combined available
chlorine, CAC, is thus a complex mixture of chloramines of ammonia,
NH Cl , and organic chloramines, RNC1 , where x, y and z are small whole
*t y z
numbers between 0 and 3:
CAC = NH Cl + RNC1
x y z
By far the most important of these chloramines is monochloramine,
NH2C1. In general the ammonia chloramines are good disinfectants com-
pared to the organic chloramines but are poor compared to HOC1 (Table 2).
In wastewater, however, monochloramine is generally relied upon as the
desired disinfectant. The high stability and fish toxicity of monochlo-
ramine is a problem. Because of the much greater disinfection efficiency
of HOC1 than NH2C1 (Table 2), the measurement of the latter as an inter-
ference in the free available chlorine measurement has been a major
problem.
PERSISTENCE
Free Chlorine
The stability of free chlorine in natural water is very low,
expecially below pH 7, because it is a strong oxidizing agent, for
example, E° is 1.49 V at pH 0 or 1.28 V at pH 7. Hypochlorous acid
rapidly oxidizes inorganic compounds such as Br and I in seawater.
The rate of reaction with iodide is too fast to follow. The rate of
formation of OBr from Br is also fast but has been measured (Farkas,
-------
53
o
j= 2.0
<
cr
LU
o
0 I
o «
LL)
_J
O
LJ
CC
1.5
1.0
0.5
0
2 4 6 8 10
CHLORINE CONCENTRATION ADDED (mg/l CI2)
Fig. 4. "Breakpoint" reaction with glycine (2 mg/l C12) at pH 7.0
after 20 hours. A-free chlorine; B-monochloramine; C-dichloramine.
-------
54
Table 2. Relative germlcidal activity of N-chloro
compound compared to HOC1
(Marks and Strandskov 1950)
Compound Relative activity
Monochloramine 1/36
N-chlorosuccinimide 1/13
N-chloropiperidlne 1/300
N-chloro-p-toluenesulfonamide 1/27,000
Lewin, and Black 1949). From the E° value it is concluded that hypo-
chlorous acid should be capable of oxidizing many compounds which
actually can reduce chlorine to chloride only slowly. The oxidation of
water to 02 is an obvious example. This reaction does not occur at a
significant rate unless the solution is irradiated with ultraviolet
light (Hancil and Smith 1971). Other compounds especially organics
react only slowly with chlorine and at rates strongly dependent on the
pH and chlorine concentration. Manganese(II), iron(II), and CN are
important inorganic compounds in natural water which react slowly. Some
of these reactions produce new oxidizing agents that are mistaken by
even the best analytical method for free chlorine.
Oxidation of organic compounds with chlorine is generally slow
especially at the low concentrations of chlorine used for disinfection
except in the presence of strong ultraviolet light (Hancil and Smith
1971). The reactions of chlorine and oxychlorine species in natural water
has recently been reviewed by Rosenblatt (1975). Atkinson and Palin
(1972) have reviewed chemical oxidation in water treatment.
The wide variety of rates of loss of free available chlorine
reacting as an oxidizing agent in demand reactions make generalizations
difficult. Recent measurements by Snoeyink and Markus (1973 and 1974)
give decay rates for free chlorine measured in nitrified secondary efflu-
ents at pH values of 8.2 to 8.5. These rates depended on sunlight, depth
of stream, turbulence, temperature, and type of residual. Samples
exposed to prevailing winds and daylight, and dosed with 3.12 mg/liter
C12, gave first order decay rates from 2.1 to 7.4 hr"1 with half-lives
-------
55
of 8 to 28 minutes. The chlorine residual in the samples was predomin-
ately (80%) OC1~ measured with the DPD, FAS titration (Palin 1957). For
samples kept indoors, without the stirring and ultraviolet light present
in the outdoors samples, a ten-fold greater persistence was measured.
The range of first order decay constants were 0.19 to 0.77 hr"1 for OC1
after an initial reduction of half of the added initial dose at a rate
similar to the outdoors samples. The first half-life required only
10 to 30 minutes while the second half-life was 1.3 to 5 hours.
The initial rapid reduction of chlorine or "demand" is due to
the easily oxidized reducing agents present. These might be iron(II),
sulfide, and the low concentrations of easily oxidized organic compounds
contained in these well nitrified effluents. The presence of sunlight,
copper, bromide, or other redox catalysts may be effective in increasing
the rates of these demand reactions. Sunlight promotes the oxidation of
organic compounds and even water to oxygen by free chlorine. Volatility
of chlorine would not be expected to be large for the OC1 predominate
in these alkaline samples studied by Snoeyink and Markus. In the more
acid pH range where HOC1 predominates, volatility losses would also be
expected.
Combined Chlorine
Ammonia at 1 mg N to 8 mg or greater C12 weight ratios acts as a
demand or reducing agent with the classical breakpoint reaction shown in
Fig. 3. Beyond the breakpoint where chloramine decomposition has
occurred, ammonia has acted simply as a reduced species consuming
chlorine. At higher concentrations of ammonia than 1 mg N to 5 mg Cl2,
mono chloramine is formed as a stable combined chlorine. Monochloramine
is a weaker oxidizing agent than free chlorine. The E° value of mono-
chloramine recently calculated by Rosenblatt (1975) is 1.16 V. At pH
7 this value would decrease to 0.95 V. These values are both approxi-
mately 0.25 V more positive than the values reported by White (1972) but
both sets of values show NH2C1 is 0.25 to 0.33 V less positive than
HOC1.
-------
56
The much lower oxidation potential of NH2C1 compared to HOC1
predicts NH2C1 would be less able to participate as an oxidant in
chlorine demand reactions. This is found in the measured first order
decay rates of 0.03 to 0.075 hr"1 for NH2C1 determined indoors by
Snoeyink and Markus (1973). They found higher rates of 0.28 to 0.31 hr"1
for these same high ammonia trickling filter effluents when subjected to
sunlight and turbulence, but these rates are still a factor of 10 slower
than the much less persistent free chlorine. Monochloramine would be
expected to persist for hours to days compared to the minutes to hours
persistence expected for free chlorine.
In estuarine water Bender et al. (1975) found decay rates for "free
chlorine" in the York river of 0.046 to 0.052 hr"1 where the pH was
7.5 to 8.0, ammonia concentrations were 0.1 to 0.5 mg N/liter, salinity
was 18 to 20 parts per thousand, and temperatures were 17 to 28°C.
Although hypochlorite solutions were added in the "free chlorine" test,
the pH and concentrations of ammonia were sufficient to form monochlo-
ramine. They also prepared solutions of monochloramine from chlorine
and ammonia. When this monochloramine solution was added to York River
water the linear portion of the first order decay curve gave a k of
0.052 hr"1, the same rate seen for "free chlorine" addition. In both
cases monochloramine is responsible for the observed persistence. Their
results are shown in Fig. 5. The slow monochloramine decay rates with
20 hour half-lives were reached after an initial rapid decay in the
first hour of contact. The major difference in the two curves is the
more rapid initial loss from free chlorine before it formed combined
chlorine. In seawater with its significant concentration of bromide
it is possible that bromine and bromamines may also be formed. These
compounds although slightly weaker as oxidizing agents than their
chlorine analogs are likely to react even quicker than chlorine and be
even less persistent.
A similar pattern of a rapid initial decay followed by a persistent
residual was measured by Baker (1970) in Passaic River water at Little
Falls, New Jersey. When dosed with 20 mg C12 per liter this water gave
77% of its total demand after 2 hours contact in the first 4 minutes.
-------
57
ORNL DWG 76-576
CHLORINE HALF-LIFE
MONOCHLORAMINE (20%)
FIRST ORDER
InC* -kt+lnCn
FREE CHLORINE
k = 0.046 hr"1
12 16
TIME IN HOURS
Fig. 5. Persistence of chlorine in York River water.
Bender et al. 1975.)
-------
58
This sample contained about 1.5 mg N/liter as ammonia. A majority of
this ammonia should be completely oxidized by this high concentration
of chlorine. In spite of this, the amperometric titration measurements
showed a persistent of 0.5 mg/1 dichloramine fraction. Baker points
out this dichloramine label is "likely inaccurate since it could include
other available chlorine compounds besides dichloramine."
ANALYTICAL METHODS
Several compounds are commonly present or formed on chlorination
which are falsely measured by one or more of the usual methods as
chlorine. These compounds are then incorrectly interpreted as chlorine
residual in either the free or combined available chlorine fractions.
The most common interference in free chlorine measurements is manganese
dioxide formed by chlorine oxidation of soluble manganous ion. Even
though manganese dioxide is a solid, it interferes stoichiometrically in
almost all methods including the continuous amperometric monitors. Only
the free chlorine amperometric titration is free of manganese dioxide
interference. Even this method, however, measures manganese dioxide in
the combined available chlorine test. Other strong oxidizing agents
such as ozone, peroxide, bromine, iodine, and chromium(VI) or dichromate
universally interfere in chlorine (disinfection) measurements.
The purpose of this portion of the paper is to describe the major
limitations of each of the common methods for chlorine measurement.
Finally a new instrumental method developed in my laboratory will be
described.
Common methods for chlorine measurement
A. Acid orthotolodine (OT. OTA, flash, drop dilution). This method
is the simplest but least accurate of all the methods. It is being
removed from Standard Methods (AWWA, WPCF, APHA 1971) in its next edition.
The yellow color fades rapidly. The free chlorine test has a combined
chlorine, NH2C1, interference of 3% per second at 20°C or 77°F. Nitrite
and iron(III) interfere as well as Mn02 which interferes stoichiometrically.
-------
59
B. Diethylparaphenylenediamine (DPP) (Palin 1957). This method
is available in colorimetrie and ferrous titration procedures. Its
major disadvantage is the instability of its reagents. The interference
from combined chlorine is small compared to the OT methods but much
larger than the SNORT, LCV, and FACTS methods discussed below. Combined
chlorine, NH2C1, interference in the free chlorine measurement is 1%
per minute at 25°C. Mn02 also interferes stoichiometrically in the free
chlorine measurement. It is the best developed method with many modifica-
tions and is readily available in kit form. It is a Standard Method.
C. Leuco crystal violet (LCV) (Black and Whittle 1967). This
method is a colorimetric procedure with low, less than 0.1%, NH2C1 inter-
ference in the free chlorine test. It is a complex procedure requiring
that reagents be added down side of the tube. The reagent stains con-
tainers. Mn02 interferes stoichiometrically in the free chlorine measure-
ment. It is available in kit form from Taylor Chemical and Hach. It is
a Standard Method.
D. Stabilized neutral orthotolodine (SNORT) (Johnson and Overby
1969). This is a colorimetric method with low, 0.1% per minute, NH2C1
interference at 25°C in the free chlorine test. Figure 6 compares the
NH2C1 interference at 35°C to the DPD method. The blue color fades
slowly. Mn02 interferes stoichiometrically in the free chlorine measure-
ment. It requires two liquid reagents which are quite stable. It is
available from La Motte in kit form (#7846). It also is a Standard
Method.
E. Syringaldazine (FACTS) (Sorber. Cooper, and Meier 1975). This
is a new method. The color fades slightly and the reagents are colored
which makes interpretation difficult at low concentration expecially
using reagents capable of measuring above 1 mg/1. It has little if any
NH2C1 or Mn02 interference. It is not yet a Standard Method but will be
published as a tentative standard in the next edition. It is not avail-
able yet in kit form.
F. Amperometric titration. This method is generally considered
to be the laboratory standard. Its main problems are its complexity and
the operator experience required. To obtain good results, blank
-------
60
40
32
24
16
8
SNORT
8
10
TIME (min)
Fig. 6. Relative interference error from monochloramine (4.9 rag/
liter Cl) in free chlorine measurement at 35°C.
-------
61
corrections should be made. Because of the fast stirring rate used with
commercial instrumentation, volatility and reduction losses can be
significant unless the majority of the titrant is added quickly. It
has low, 0.1%, NH2C1 interference and no Mn02 interference in the free
chlorine measurement. Electrode poisons and films are a problem and
careful cleaning to remove iodide between titrations is important.
Equipment is available from Fisher and Porter, Wallace and Tierman, and
others.
G. Continuous amperometric cells, bare electrode. These instruments
are on-line amperometric measurements. They should not be confused with
the amperometric titratlon. They use no titrant but do consume buffer.
They are claimed to measure in either a free chlorine mode with acetate
buffer or in a combined chlorine test.with iodide added to the buffer.
These methods have stoichiometric NH2C1 and MnC>2 interference even in
the free chlorine mode (Morrow and Roop 1975). These methods are also
subject to electrode films and poisons. Equipment is available from
Wallace and Tierman, Fisher and Porter, Capital Controls, Honeywell,
and others.
A modification on this procedure which also uses an amperometric
cell is the National Bureau of Standards Flux Monitor (Marinenko,
Huggett, and Friend 1975). It has the novel feature of a built in
coulometric standardization but, also, has the same problems as above.
H. Summary. The selective free chlorine methods such as SNORT,
DPD, LCV, FACTS, or amperometric titration can be used only with small
interferences from Mn02 and other strong oxidizing agents. For combined
chlorine these methods with added iodide or the iodometric back titration
and the continuous amperometric monitors or the NBS Flux Monitor can be
used. A determination of interferences can and should be done especially
in the combined chlorine methods. This can be done by adding arsenite
to the sample which will react with free and combined chlorine, and
leave manganese dioxide, nitrite, and other compounds whose reaction or
the blank reading should be subtracted. Use fresh reagents especially
if they contain color or growths. Iodide solutions should also be
fresh each week when measuring low concentrations.
-------
62
None of the methods above are capable of measuring HOC1 in the
presence of OC1 . Cyanuric acid and other acidic nitrogen functions
such as those found in amides and peptide bonds when they form N-chloro
derivatives have an acidic and easily hydrolyzed chlorine (O'Brien,
Morris, and Butler 1974). These compounds are in equilibrium with HOC1
but have lower and probably no inherent ability in themselves to act as
disinfectants (Table 2). Because of this equilibrium, any analytical
method which removes or reacts with the HOCl from these compounds will
measure them stoichiometrically as free chlorine, FAC. For this reason,
even the measurement of free chlorine with the best of current methods
must be interpreted with care with respect to disinfection.
The measurement of combined chlorine, CAC, and its interpretation
in terms of fish toxicity or disinfection from NH2C1 is difficult in
natural water and impossible in wastewater. The CAC methods are gen-
erally based on the use of iodide and/or acid to give an equivalent of
12 from the chlorine associated with nitrogen. In general, organic
chloramines are measured by this method (Palin 1950) with iodine gen-
erated in the same way as with NH2C1. As shown in Table 3 methods which
employ acidification are especially poor because of the decomposition
produced by acidification. Thus the acid orthotolidine method, OT,
gave much lower readings in all the samples. After only 30 seconds at
pH 1.5 the chlorine residual determined by amperometric titration in a
chlorinated sewage dropped from 4.75 to 0.58 ppm. Table 3 also shows
N-chloropiperidine is measured as if it were NI^Cl in the amperometric
titration. Table 2 shows, however, this compound is only 1/9 as effective
as NH2C1 as a germicide.
The HOCl membrane electrode
An amperometric membrane electrode similar to the oxygen probe has
been perfected for chlorine, bromine, and iodine (Johnson and Edwards
1975). The response of this electrode is in proportion to the decreasing
effectiveness of the halogens in their various chemical forms X2, HOX,
and OX". Figure 7 shows the cell design developed for the halogen elec-
trode. This cell is similar to the amperometric oxygen electrode. It
-------
63
Table 3. Organic chloramines measured as residual chlorine
15 minute contact at pH 7.0
(Marks and Joiner 1948)
Base
1-Cystine, 25 ppm
Piperidine, 61 ppm
Above piperidine + S02
Raw sewage, pH 7
Above sewage + S02
Raw sewage, pH 7
Titration of above sample after
30 seconds at pH 1.5
Chlorine
dose, ppm
15
6
-0.56
10.5
-0.84
12
12
Amperometric
titration, ppm
1.85
5.60
5.20
3.95
3.15
4.75
0.58
OT, ppm
0.04
0.20
0.17
0.90
0.50
0.10
uses a platinum or gold cathode separated from the solution to be mea-
sured by a microporous film that permits the transport of X2, HOX, and
OX species with decreasing sensitivities. Several membranes were
screened for use in the system. A microporous polypropylene film and
two microporous fluorocarbon films were found to be the most useful.
The microporous polypropylene and one of the flurocarbon films used were
hydrophobic in character. The other flurocarbon film, an ion exchange
membrane, absorbed approximately 18% water, had a microporous structure
of 10 mil thickness, and was used in the hydrogen form.
Chlorine, bromine, and iodine, all three, exist in several chemical
forms as determined by the following equilibria:
X2 + X~ = Xs , (6)
X2 + H20 = HOX + H+ + X~ , (7)
HOX = OX~ + H+ . (8)
Of these forms the X2 form is generally the best disinfectant while HOX
is also highly effective depending on the biological system being dis-
infected. For chlorine only the HOCl form is important. The OX and
-------
64
CATHODE TIP IS GOLD
POLYETHYLENE
COLLAR
POLYETHYLENE
COLLAR
SILVER ANODE
MEMBRANE
FILL PORT FOR
ELECTROLYTE
CABLE TO
INSTRUMENT
i '„ . PRESSURE
LJl j COMPENSATOR
Fig. 7. The HOC1 membrane electrode.
-------
65
X3 species are relatively poor disinfectants compared to the HOX and
X£ species. The halide ion itself, X , is completely ineffective as
a disinfectant.
Figure 8 shows calibration curves for HOCl taken with several elec-
trodes over a six-month period. This curve shows the response of the
electrode system is linear as a function of concentration and therefore
gives a current signal in direct proportion to the concentration of the
sample as indicated by Equation 9:
I = SC , (9)
where S is the sensitivity in microamperes per ppm and C is the concen-
tration in ppm or milligrams per liter,.
Table 4 gives the sensitivities in microamperes per ppm determined
from calibration curves like those in Fig. 8.
The ion exchange fluorocarbon electrode response to the HOX species
has approximately twice the sensitivity it shows to the OX species,
while the X2 species gives twice as much response as HOX. Using the
hydrophobic fluorocarbon or polypropylene, the halogen electrode
response is approximately 30 times greater for the X£ species than for
the HOX species and is completely insensitive to the OX species.
The temperature dependence for the response of the electrode system
obeys Equation 10:
(10)
where Ep is the energy of permeation, R is the gas constant in cal/mole
degree and T is the absolute temperature in degrees Kelvin. For the
ion exchange fluorocarbon film Ep averages 2.36 Kcal/mole with a standard
deviation of 0.45 for the six chlorine and bromine species given in
Table 4. The value of Ep averages 4.26 Kcal/mole with a standard devia-
tion of 0.90 for the microporous polypropylene with the five chlorine,
bromine, and iodine species which can be measured. This compares with
17 Kcal/mole reported for oxygen diffusion through homogenous polyethylene.
-------
66
0.50
0.45-
0.40-
0.35-
0.30-
£ 0.20-
w
O 0.13-
O.tO-
0.05-
0.00-
0.00
0.50 1.00 1.50
Concentration mg/l as Cl,
2.00
2.50
Fig. 8. Composite of nine months of HOC1 measurements at pH 5 and
337 mv of applied potential using the microporous hydrophobia membrane
at 20°C and at an ionic strength of 0.02.
-------
67
Table 4. Sensitivities for halogen species with ion exchange
fluorocarbon and polypropylene membranes
Species
C12
HOC1
OC1~
Br£
HOBr
OBr~
12
Sensitivities
Ion exchange
fluorocarbon
4.27
2.98
1.43
1.22
0.98
0.40
-
in microamperes /ppm
Polypropylene
13.64
0.41
-
4.13
0.15
-
0.37
Thus the oxygen electrode has approximately 3 to 7 times greater tempera-
ture sensitivity than the halogen electrode. This gives the halogen
electrode a sensitivity to temperature as Q^Q at 25°C of approximately
1.26 for the polypropylene film and 1.14 for the ion exchange fluoro-
carbon membrane. This is slightly less temperature sensitivity than has
been reported for the disinfection of bacteria by halogens.
The halogen electrode like the oxygen electrode can be placed
directly in the solution or sample stream and the current calibrated in
terms of effective chlorine residual or disinfection efficiency. There
is an important difference between the halogen electrode response and
typical concentration measurement systems. The electrodes responds
with sensitivities for various halogen species in proportion to their
efficiency as disinfectants. The current measurement, then, gives not
a measurement of the sum of the total halogen concentration with com-
plete disregard of the lower effectiveness of OCl as a disinfectant,
for instance, but gives a measurement of effective residual halogen con-
centration. Thus the response of the electrode does not suffer from
the OC1~ interference typical of total concentration measurements but
reads current which is in proportion to the efficiency of the chemical
species as water disinfectants.
-------
68
If desired, FAC concentrations can be determined by calibrating
at the pH of the sample. Calibration factors would then be used which
were appropriate to the chemical form desired as a function of pH. The
FAC concentration measurement is not the measurement which is really
needed. The measurement needed is the determination of residual effec-
tive disinfectant. On this basis FAC concentration measurements
determining the sum of HOC1 and OC1 are in error to the extent that
they measure OC1~ as if it were as effective as HOC1. This electrode
gives us for the first time a method capable of measuring only HOC1,
real free chlorine.
The electrode is not sensitive to manganese dioxide which is a
common interference in all current analytical methods for chlorine. The
electrode can be placed directly in the inplant flow downstream from
the point of application of chlorine to determine chlorine residual at
any contact time without the necessity for the time delay and sample
loss typical of remote systems. The electrode is not as subject to
electrode fouling which is a common problem with other amperometric
electrode systems. Depending on the applied voltage, it will measure
either HOC1 alone for use in drinking and cooling waters or HOC1 plus
NH2C1 for use in distribution systems and wastewater. Table 5 summarizes
the electrode characteristics. The electrode is available commercially
from Delta Scientific Corp., Llndenhurst, N.Y.
Table 5. Characteristics of the halogen membrane electrode
1. Measures only (HOC1) independent of pH,
0.4 yA/ppm
2. Insignificant interferences from poor
disinfectants:
Monochloramine - 1% NH2C1
Hypochlorite ion - 1% OC1~
Nitrite ion - 1% N03~
Manganese(IV) - 0% Mn02
Iron(III) - 0% Fe(OH)3
Color, organics, oxygen - 0%
3. Temperature dependence -
2 to 6% per °C, 1 to 3% per °F
4. Sensitive to 0.03 mg/1
5. Accurate, = 0.40 to 0.42 uA/ppm at 95% C.I.
6. Independent of pH (FAC depends on HOC1/OC1"
ratio)
-------
69
ACKNOWLEDGEMENT
This work was supported in part by the U.S. Army, Medical Research
and Development Command, Contract No. DADA 17-72-C-2053.
REFERENCES
American Water Works Association. 1973. Water chlorination principles
and practices. AWWA Manual M20, New York.
Atkinson, J. W., and A.^ T. Palin, 1972. Chemical oxidation in water
treatment, p. El-E^. In Proceedings of international water supply
conference, special subject No. 5. New York.
AWWA, WPCF, APHA. 1971. Standard methods for the examination of water
and wastewater, 13 ed. American Public Health Association. New
York.
Baker, R. J. 1970. Engineering considerations in disinfection, p. 685-
697. In Am. Soc. Civil Eng., Proceedings of the national specialty
conference on disinfection. Amherst, Mass.
Bender, M. E., M. H. Roberts, R. Diaz, and R. J. Huggett. 1975.
Proceedings technology and ecological effects of biofouling,
Maryland Power Plant Siting Program. Baltimore, Md.
Black, A. P., and G. P. Whittle. 1967. New methods for the colorimetric
determination of halogen residuals. Part II: free and total
chlorine. J. Am. Water Works Assoc. 59: 607.
Farkas, L., M. Lewin, and R. Black. 1949. The reaction between hypo-
chlorite and bromides. J. Am. Chem. Soc. 71: 1987.
Hancil, V., and J. M. Smith. 1971. Chlorine-sensitized photochemical
oxidation of soluble organics in municipal wastewater. I&EC
Process Design and Dev. 10: 515-523.
Johnson, J. D., and J. W. Edwards. 1975. A halogen membrane electrode.
Presented to Annual Conference, American Water Works Association.
Minneapolis, Minn.
Johnson, J. D., and R. Overby. 1969. Stabilized neutral orthotolodine
method for chlorine. Anal. Chem. 41: 1974.
Marinenko. G., R. J. Huggett, and D. G. Friend. 1975. An internal
calibration instrument for monitoring chlorine residuals in
natural waters. J. Fish. Res. Board Can. In press.
-------
70
Marks, H. C., and R. R. Joiner. 1948. Determination of residual
chlorine in sewage. Anal. Chem. 20: 1197.
Marks, H. C., and F. B. Strandskov. 1950. Halogens and their mode of
action. Ann. N.Y. Acad. Sci. 53: 163-171.
McClanahan, M. 1975. Recycle —what disinfectant for safe water then?
p. 49-66. In J. D. Johnson (ed.) Disinfection-water and wastewater.
Ann Arbor Science. Ann Arbor, Michigan.
Morris, J. C. 1966. The acid ionization constant of HOC1 from 5 to
35°C. J. Phys. Chem. 70: 3798.
Morrow, J. C., and R. N. Roop. 1975. Advances in chlorine residual
analysis. J. Am. Water Works Assoc. 67: 184.
O'Brien, J. E., J. C. Morris, and J. N. Butler. 1974. Equilibria in
aqueous solution of chlorinated isocyanate, p. 333. In A. J.
Rubin (ed.) Chemistry of water supply treatment and distribution.
Ann Arbor Science. Ann Arbor, Michigan.
Palin, A. T. 1950. A study of the chloro derivatives of ammonia and
related compounds with special reference to their formation in
the chlorination of natural and polluted waters. Water Water Eng.
1950: 248.
Palin, A. T. 1957. The determination of free and combined chlorine in
water by the use of diethyl-p-phenylenediamine. J. Am. Water
Works Assoc. 49: 873-880.
Rosenblatt, D. H. 1975. Chlorine and oxychlorine species reactivity
with organic substances, p. 249-276. In J. D. Johnson (ed.)
Disinfection-water and wastewater. Ann Arbor Science. Ann Arbor,
Michigan.
Snoeyink, V. L., and F. I. Markus. 1973. Chlorine residuals in treated
effluents. Report prepared for Illinois Institute for Environ-
mental Quality, Urbana, Illinois.
Snoeyink, V. L., and F. I. Markus. 1974. Chlorine residuals in treated
effluents. Water Sew. Works. 121: 35-38.
Sorter, C., W. Cooper, and E. Meier. 1975. Selection of a field method
for free available chlorine, p. 91. In J. D. Johnson (ed.)
Disinfection-water and wastewater. Ann Arbor Science. Ann Arbor,
Michigan.
Sugram, R., and G. R. Helz. 1975. Apparent ionization constant of
hypochlorous acid in seawater. Department of Chemistry, University
of Maryland, College Park, Md. Submitted for publication.
-------
71
White, G. C. 1972. Handbook of Chlorination. Van Nostrand Reinhold,
New York, p. 212.
DISCUSSION
George R. Helz, University of Maryland. On your chart you show
essentially 2 ppb as the sensitivity of your electrode. On your
previous slide the calibration curve was essentially 1 to 10 ppm. Can
you clarify that? Can you really measure at the ppb level?
Johnson. No, the electrode will not measure 3 ppb, thirty maybe,
and then only if you're willing to accept 100% error. That is, it mea-
sures plus or minus .03 ppm or 30 ppb. It's not as sensitive as I'd
like it, that's for sure. The NBS flux monitor is nice and sensitive
but then you don't get selectivity for chlorine. So you're sort of
between the selectivity devil and the sensitivity devil.
George Clifford White, Consulting Engineer. Don, you sound very
confident about being able to selectively measure monochloramine. This
is very exciting news because this would be very helpful in wastewater
treatment obviously. Now, have you done anything measuring monochlo-
ramine in the presence of all the rest of the garbage that normally
interferes in monochloramine measurements in wastewater?
Johnson. I just made an experiment last week with N-chloroglycine.
We ran a curve with a constant concentration of glycine adding chlorine
up to where we have just N-chloroglycine without cleaving off the amine
and doing oxidative deamination. So we have this compound present and
it measures with the standard analytical methods, for example, SNORT,
DPD, amperometric titration. It measures like monochloramine. Then we
put the electrode in and we measure with the lower voltage where the
electrode responds to monochloramine. We don't get any response. So
the electrode does respond to monochloramine under proper conditions
but it does not measure the chloroaminoacid, N-chloroglycine.
-------
ORGANO-CHEMICAL IMPLICATIONS OF WATER CHLORINATION
Robert M. Carlson and Ronald Caple
Department of Chemistry
University of Minnesota
Duluth, Minnesota 55812
ABSTRACT
The desire for structural information on the specific organic com-
pounds present in a given sample of renovated water makes it imperative
that there be a basic understanding of those principles of mechanistic
organic chemistry that apply to the situation in question. The process
will be illustrated for aqueous chlorination, where the observed
chloroorganics can be readily explained on the basis of commonly recog-
nized reactive intermediates and those stereochemical and electronic
features associated with the organic moiety. The relationship of these
mechanistic processes to pH, product distribution, BOD, oxidative
capacity, and chloramine formation are considered.
INTRODUCTION
The common thread that has woven its way throughout the entire fabric
of research into the chemical and biological implications of water
renovation processes is a desire for structural information regarding
the specific "parent" and "second-order" compounds under investigation.
This awareness of potential structural implications may be as subtle as
the negative feeling synonymous with polychlorinated organics or as
direct as the development of new GC mass-spectral techniques. Although
most of the "awareness" is necessarily "after the fact," it should be
reaffirmed that a major goal of this effort is to provide a predictive
capability. That is, what structural features — and thereby ultimately
what biological response — should be anticipated under a given set of
circumstances? If indeed this goal is to be met it is imperative that
73
-------
74
a greater number of those individuals concerned with water renovation
be aware of those basic elements of mechanistic organic chemistry that
apply to their particular situation.
In the case of aqueous chlorination, we should be concerned with
two distinct electron deficient species of chlorine — the "chloronium
ion" and the chlorine radical. As might be expected from the charged
nature of the chloronium ion, it is the major contributor to reactions
occurring in aqueous media. On the other hand, significant product
formation resulting from the chlorine radical should be anticipated in
the presence of light and the absence of a significant polar reaction
pathway involving the loss of the parent chlorine molecule.
"Cl+" = Chloronium Ion "ci»" = Chlorine radical
Outer shell has 6 e~ Outer shell has 7 e~
Electron deficient Electron deficient
Positively charged Neutral
In predicting the mode of action of these reactive electrophilic
intermediates we should examine those potential electron sources that
would satisfy this inherent electron deficiency. Indeed, such electron
rich centers as the unshared electrons of the nitrogeneous bases and
the ir electron systems of olefins and aromatic compounds appear particu-
larly vulnerable.
Amines represent a major class of compounds recognized to be
present in waters subjected to renovation with the result that chemistry
of chloramine formation has received considerable attention. However,
such chemistry remains as a topic for the current discussion only in
that within the chloramine there remains a potential that has yet to
be determined for the subsequent formation of carbon-bound chlorine.
RESULTS AND DISCUSSION
The reaction of chlorine with aromatic systems results in "electro-
philic aromatic substitution" (Fig. 1). In this process the general
principles of relative substrate reactivity and the orientation of
-------
75
Fig. 1. Electrophilic aromatic substitution.
substitution is documented to the extent that the elements of predicta-
bility are particularly attractive. This was confirmed in a limited
study conducted in our laboratories on the chlorination of aromatic
compounds possessing substituents of diverse electronic properties.
Under the conditions of dilute aqueous chlorination phenol was found
to be the most reactive, with the other compounds decreasing in chlorine
incorporation with the increasing electronegative nature of the sub-
stituents (see Table 1).
Table 1. Electrophilic aromatic substitution: chlorine
incorporation in selected aromatic compounds
(9.5
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Compound
± 0.6 x 10~4 M)
phenol
anisole
acetanilide
toluene
benzylalcohol
benzonitrile
nitrobenzene
chlorob enzene
methylbenzoate
benzene
Percent
after
pH 3
2.2 ± 0.1
19.3 ± 0.2
44.7 ± 0.5
88.9 ± 0.1
97.7 ± 0.2
97.9 ± 0.2
98.2 ± 0.1
98.2 ± 0.1
98.2 ± 0.2
98.5 ± 0.1
chlorine remaining
reaction timea
pH 7 pH 10
2.4 ± 0.1 2.4 ± 0.2
88.6 ± 0.4 97.2 ± 0.3
98.6 ± 0.2
97.1 ± 0.4
- -
-
-
-
-
— —
Chlorine (7.0 x lO'4 M), 20 minutes, 25°C.
-------
76
It should also be noted that in three cases (anisole, acetanilide,
and toluene) the extent of chlorine incorporation increased with
decreasing pH. This observation clearly appears to be related to an
existing equilibrium in the potential chlorinating species. This
result also closely parallels the varying disinfecting capability of
chlorine with pH, where low pH values promote effective disinfection.
HOC1 - H+ + OC1~, pKa =7.5
A related point of interest is illustrated by a detailed examina-
tion of biphenyl chlorination conducted in our laboratories and those
of Monsanto (Dr. Harold Weingarten and coworkers) where the major pro-
ducts were those containing limited numbers of chlorines incorporated
into the biphenyl nucleus (Fig. 2). Although this observation can be
partially ascribed to the dilute nature of the reaction medium, it also
illustrates the decreasing reactivity of the aromatic nucleus with
increasing chlorine substitution.
o
o
C1
o
o
ci
+ OTHER MONO- AND DI-CHLORO
BIPHENYLS
Fig. 2. Chlorination of biphenyl.
The relationship of aromatic chlorination to such commonly used
parameters as BOD was ascertained by determining the BOD of a given
molar amount of parent.phenol relative to the mono- and dichloro progeny.
The results indicate that at low concentrations the chlorophenols are
not significantly metabolized over the test period and that at high
-------
77
concentrations the chlorophenols are actually adversely affecting
the microbial population (Table 2).
These results have the obvious environmental implications that:
(1) chlorinations at low pH (e.g., certain industrial wastes) have a
greater probability of generating "second-order" chloroorganics, (2)
polychlorinated aromatics isolated from a given effluent are most
likely not derived via a dilute aqueous chlorination process, (3) an
a priori estimate of the product type and distribution can be made if
the parent aromatic organic content of a waste is known, and (4) quali-
tative criteria such as reduction of BOD upon chlorination must be viewed
with some suspicion.
Olefins and acetylenes represent an additional group of compounds
that are potentially vulnerable to attack. However, in this situation
each molecule has its own complex set of electronic and stereochemical
features that dictate the ultimate set of products. Although this
complexity makes a detailed analysis of these diversified groups of com-
pounds more difficult, several principles are apparent from the examina-
tion of some representative olefinic materials.
For example, the aqueous chlorination of oleic acid provides a
mixture of chlorohydrins (Fig. 3), a result anticipated on the basis of
the participation of the solvent (i.e., water) in the second stage of
the addition sequence and the "isolated" nature of the disubstituted
double bond. This successful competition of water with chloride should
have been expected considering the dilute nature of the reaction but
also raises the question of the possible implications of using chlorina-
tion processes in substantially different media such as seawater which
contains significant amounts of "impurities" (e.g., Cl~).
In another instance, the anticipated stereochemistry of the inter-
mediates in the aqueous chlorination of the environmentally ubiquitous
a-terpineol assisted in the determination of the products and their
variation with pH (Fig. 4). For example, it should be expected that
epoxides would be observed at higher pH values where the nucleophilic
substitution of the alkoxide (RO ) on the carbon bearing the chlorine
would be promoted by the trans-anti parallel arrangement of the
-------
Table 2. Summary of phenol results (BOD)
Test
number
2
3
4
5
6
7
8
Phenol 0-Chlorophenol p-Chlorophenol (2,4-)Dichlorophenol
Concentration
1.06 x 10- 5 M (1 ppm)
7.78 x 10~6 M (1 ppm)
7.78 x 10-5 M (10 ppm)
7.78 x 10-1* M (100 ppm)
1 x 10~5 M (1.29 ppm)
1 x 10-" M (12.9 ppm)
1 x 10~3 M (129 ppm)
2 x 10~6 M (0.19 ppm)
6 x 10-6 M (0.56 ppm)
1 x 10-5 M (0.94 ppm)
1 x 10~5 M (1.29 ppm)
1 x 10~3 M (129 ppm)
2 x 10~6 M (0.19 ppm)
6 x 10~5 M (0.56 ppm)
1 x 10-5 M (0.94 ppm)
2 x 10"6 M (0.33 ppm)
6 x 10~6 M (0.99 ppm)
1 x 10~5 M (1.65 ppm)
5 (day) 10 20 5 (day) 10 20 5 (day) 10
312
191
195
157
226
225
94 140
168 215 284
214 306 290
229 335 335
165 207 229
187 223 245
183 233 269
227 290 290
243 278 273
20 5 (day) 10 20
135 159 166
154 156 173
151 149 147
•-4
00
-------
79
CH3-(CH2) -CH - CH -(CH2)
H H C1 OH
\ / DH 2-10
CH3-(CH2) - C = C -(CH2) -C02H — > CH3-(CH2) -CH - CH -(CH2)
7 ? OH Cl 7
Fig. 3. Chlorination of unsaturated fatty acid.
chlorine relative to the hydroxyl in chlorohydrin IV that is not present
in compound II (i.e., II -> III at high pH) . Likewise, the presence of
the bridged system can be explained by the involvement of the 3°-OH in an
intramolecular attack on the "down" epoxide (Fig. 5). The other observed
products can similarly be derived by examining alternative reaction
pathways from these or related intermediates (Fig. 4).
The search for those products derived from "free-radical" processes
should begin where the molecule in question can provide a relatively
stable odd electron system. An example of such a system would be that
obtained from the abstraction of a hydrogen atom from a benzylic system
where the resulting radical owes its stability to the delocalization of
the odd electron throughout the molecule by resonance.
Examples of such "free radical" processes appear to have been
observed among the chlorination products of the resin acid, dehydro-
abietic acid (Fig. 6).
CONCLUSIONS
The investigation of the dilute aqueous chlorination of typical
compounds known to be present in water subjected to chlorine renovation
indicates that chlorine is readily incorporated into the carbon frame-
work by a pathway that is predominately ionic. Furthermore, the use of
the basic principles of mechanistic organic chemistry is helpful in
elucidating the structure and/or evaluating the distribution of the
aqueous chlorination products. A major result of such a capability is
thereby the provision for an apparent element of predictability of the
environmental impact of variations that might occur during the chlorina-
tion process.
-------
II
III
IV
OH
Cl OH
QH
OH
OH
(42*)
(39*)
Base
AND I (8%)
,/ II (27%)
cl (51J5) III (1%)
VI
VII DICHLORIUES (2%)
Fig, 4. Chlorination of ct-terpineal.
-------
81
HO
CH-, •=•
Fig. 5. Formation of bridged system by intramolecular rearrangement.
pH 2
10-100 ppm C12
Cl
Fig. 6. Chlorination of resin acids (abietic and dehydroabietic acid).
In addition, the current study has generated several major questions
that require immediate attention:
1. How much of this structural information can be translated to a
"real world" situation where the presence of ammonia and a
complex combination of organic and inorganic materials are
involved?
2. Are the present requirements (i.e., standards) for chlorination
meaningful, especially where the environmental "trade-off"
would mean the production of significant quantities of degrada-
tively resistant chloroorganics?
3. Is the present effort to discredit the use of chlorine to be
matched by a concurrent investigation of the potential con-
sequences of using an alternative procedure?
-------
82
REFERENCES
American Water Works Association Inc., 1971. Water quality and treat-
ment. McGraw-Hill, New York.
Carlson, R. M., R. E. Carlson, H. L. Kopperman, and R. Caple. 1975.
The facile incorporation of chlorine into aromatic systems during
aqueous chlorination processes. Environ. Sci. Technol. 1975: 674.
De La Mare, P. B., and J. H. Ridd. 1959. Aromatic substitution:
nitration and halogenation. Academic Press, New York.
Johnson, J. D. 1975. Disinfection: water and wastewater. Ann Arbor
Science, Ann Arbor, Mich.
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants. Ph.D.
Thesis, U. of Tenn., Oak Ridge National Laboratory TM-4290.
Kopperman, H. L., R. E. Carlson, R. Caple, and R. M. Carlson. 1974.
Structure-toxicity correlations of phenolic compounds to Daphnia
magna. Chem. Biol. Interactions 9: 245.
Liberles, A. 1968. Introduction to theoretical organic chemistry.
Macmillan, New York.
Manufacturing Chemists' Association. 1973. The effect of chlorination
on selected organic chemicals. EPA Water Pollution Control
Research Series Project //12020EXG.
Morris, J. C. 1975. Formation of halogenated organics by chlorination
of water supplies. Environmental Health Effects Research Series,
EPA-600/I-75-002.
White, G. C. 1972. Handbook of Chlorination. Van Nostrand-Reinhold,
New York.
DISCUSSION
George R. Helz, University of Maryland. You stated that multiple
substitution in aromatic compounds is unlikely. Won't this depend on
the initial conditions, especially the chlorine to aromatic ratio? You
might get one answer at the sewage treatment plant and another answer
in drinking water treatment. Have you investigated this?
Carlson. Yes. We have in a sense and we still are convinced that
unless your aromatic molecule is highly reactive you are still not
going to incorporate a lot of chlorine into that molecule. The only
way we could get a highly chlorinated byphenyl was to go to 250 ppm at
a pH of 2. That's pretty high. I think that relates to your question.
-------
83
Carl W. Gehrs, Oak Ridge National Laboratory. Bob, I'm not sure
I followed you correctly. What pH were you working with? Were they
strictly pH 2 and 10?
Carlson. We went through the whole series. We looked at the
extremes in that one case. But we studied a variety of reaction condi-
tions and, depending on the conditions, there were varying combinations
of all those different products.
Gehrs. You did go across the pH range?
Carlson. Right.
J. Donald Johnson, University of ;iorth Carolina-Chapel Hill. Did
you look at the effect of uv catalysis on any of your reactions?
Carlson. No. That would be very interesting to do. I think you
would expect on the basis of what is known to see a lot more free radical
reactions and probably a lot more chlorine incorporation. I think a
good example of the use of these mechanistic principles in product
prediction is in this business of polychlorinated or halogenated methanes.
I have been convinced by looking at the problem and talking with people
about polychlorinated methane, that we're going to find that the incor-
poration of chlorine is going to be enhanced quite readily by having
multiple carbonyl groups present. We've talked about the iodoform reac-
tion and the ability to have multiple carbonyl groups. Activating that
carbon is going to provide a very facile mode of incorporation of
chlorine. Any molecule that can generate an oxygenation pattern in a
1-3 relationship is going to give these halogenated methanes. I'm
convinced of it.
-------
CHLORINATION OF ORGANICS IN DRINKING WATER
Alan A. Stevens, Clois J. Slocum, Dennis R. Seeger, and Gordon G. Robeck
Water Supply Research Division
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
Halogenation of organic compounds occurs during chlorination of
drinking water. The major known products of these reactions are tri-
halomethanes. Recognition of the potential health significance of these
compounds has led to a search for alternatives to present treatment
practice for potable water supplies. Some factors influencing trihalo-
methane production are precursor compound concentration, pH, type of
disinfectant used (e.g., free vs combined chlorine), and temperature.
Appropriate control of these factors reduces concentrations of
trihalomethanes in the finished water. The influence of these factors
as determined by bench- and pilot-scale experiments is demonstrated and
application of some appropriate control measures at a full-scale treat-
ment plant is discussed.
INTRODUCTION
Recently there has been great interest in the study of organic com-
pounds in drinking water — interest that steins largely from the results
of (and the publicity that followed) a 1974 study of New Orleans drinking
water (U.S. Environmental Protection Agency 1974). About the same time,
two studies (Rook 1974, Bellar et al. 1974) called attention to the
presence in finished drinking water of some trihalomethanes (mostly
chloroform) which were not found in the respective raw waters at the
locations of study. Conclusions were drawn in both reports that the
trihalomethanes were formed during the chlorination step of the water
treatment process.
85
-------
86
Because of the interest in the various organic contaminants and
the concern as to their significance, the U.S. Environmental Protection
Agency (EPA) undertook a survey of 80 selected cities to measure the
concentration of six halogenated compounds in raw and finished water.
Those six included four trihalomethanes (chloroform, bromodichloro-
methane, dibromochloromethane, bromoform) suspected of being formed on
chlorination, plus carbon tetrachloride and 1,2-dichloroethane, known
contaminants at New Orleans, but not necessarily formed on chlorination.
During this National Organics Reconnaissance Survey (NORS) a more
comprehensive organic analysis was also performed in five of the 80
cities and has just been completed in another five.
The occurrence of trihalomethanes in finished drinking water was
demonstrated to be widespread and a direct result of the chlorination
practice. No hard evidence was found in this regard with respect to
1,2-dichloroethane or carbon tetrachloride.
Based on the survey results, a theoretical finished water with the
median concentration of each compound would contain about 21 yg/1 of
chloroform, 6 yg/1 of bromodichloromethane, 1.2 yg/1 of dibromochloro-
methane, and an amount less than the detection limit for the method
used (Symons et al. 1975) of bromoform (Fig. 1). Although most of the
finished waters tested demonstrated this decreasing order of concentra-
tion, this was not always the case. The finished water at one location
had a chloroform concentration of only 12 yg/1, but a bromoform concen-
tration of 92 yg/1. It was speculated that this concentration reflected
a relatively high bromide concentration in the raw water, with oxidation
of bromide to hypobromite by hypochlorite, and subsequent reaction of
hypobromite with precursor compounds to form the bromine substituted
trihalomethanes. Recently workers at another EPA laboratory (Bunn et al.
1975) have adequately demonstrated this effect by experimentally adding
the halides fluoride, bromide, and iodide in the form of salts to
Missouri River water and subsequently chlorinating that water. The
detected reaction products included all 10 possible nonfluorine mixed
and single halogen-containing trihalomethanes. Finally, the range of
chloroform concentrations was less than 0.1 to 311 yg/1; bromodichloro-
methane, none found (NF) to 116 yg/1; dibromochlormethane, NF to 100 yg/1;
and bromoform, NF to 92 yg/1.
-------
87
300
2 5 10 30 50 70 90 99
PERCENT EQUAL TO OR LESS
THAN GIVEN CONCENTRATION
Fig. 1. Frequency distribution of trihalomethane data (NORS),
-------
88
While the health significance of trihalomethanes produced during
chlorination of drinking water has not been completely evaluated at
this time (1975), understanding the factors affecting the ultimate
formation of the trihalomethanes was considered prudent. The goal was
then to develop general conclusions applicable to rational modification
of water treatment processes if removal of trihalomethanes was finally
deemed important for public health reasons. Basic approaches to affect
finished water trihalomethane concentrations considered for study were:
reducing precursor compound(s) concentration, changing disinfectant
(e.g., ozone, chlorine dioxide, etc.), and removing trihalomethanes
after formation. The last of these is being studied as an alternative
and has been discussed elsewhere (Love et al. 1975). Changing disin-
fectant without an intense research input to study other public health
ramifications could be a catastrophic step. Therefore, because a chlorine
residual must be maintained within the distribution system, removing of
precursor compounds or controlling their reactions with chlorine was
considered the most logical approach.
The foremost consideration in adjusting a series of water treatment
processes to remove an organic precursor is identifying the compound(s).
Bellar et al. (1974) proposed ethanol as the compound with oxidation by
hypochlorite to acetaldehyde, or acetaldehyde itself, followed by the
classical haloform reaction to be the mechanism of trihalomethane pro-
duction. Organic chemistry texts typically cite acetone as the simplest
example of a methyl ketone that undergoes the haloform reaction. Indeed,
Fairless et al. (1975) have investigated the reactivity of simple methyl
ketones in water supplies and consider them to play a major role in
trihalomethane production. Glaze and Henderson (Glaze 1975) have
identified chlorinated acetone derivatives that could be haloform reac-
tion intermediates in super-chlorinated sewage effluents. These theories
are attractive because the precursor compounds mentioned have been
qualitatively identified during gas chromatographic-mass spectrometric
(GCMS) analysis of Ohio River water that contains the unknown precursors
which react to form trihalomethanes upon chlorination.
-------
89
In December 1974 Rook proposed that natural humic substances were
responsible. Later Rook (1975) discussed the probable role of the fulvic
acid fraction in trihalomethane production, elaborating with examples of
very reactive tfJ-dihydroxy aromatic compounds suspected to be basic
building blocks of the humic (fulvic) acid structure.
Some clarification of the relative roles played by these two groups
of compounds (humic materials vs acetyl derivatives of low molecular
weight) is paramount to the eventual understanding needed to predict the
success of any water treatment process change designed to bring about
a reduction in the ultimate trihalomethane concentrations. The roles of
other treatment parameters such as NHs addition with chlorine (free vs
combined chlorine), pH, and temperature should also be clarified.
METHODS
Reagents
Chlorine was obtained from Union Carbide Corporation (high purity
grade, Ohio Valley Sales, Cincinnati, Ohio). Stock solutions were
prepared by passing the pure gas through nitrogen purged distilled
water. Freshly prepared stock solutions were standardized by ampero-
metric titration as described in Standard Methods (American Public Health
Association 1971). Experimental mixtures were prepared by appropriate
volumetric dilution of the stock solutions in the test media.
Water for the various experiments was obtained from the pilot plant
facility of the Water Supply Research Division (WSRD), Municipal Environ-
mental Research Laboratory (MERL), U.S. Environmental Protection Agency,
Cincinnati, Ohio. This plant has previously been described in detail
(Love et al. 1975). "Raw" water was that obtained directly from the
Ohio River intake at the Cincinnati Water Treatment Plant, Cincinnati,
Ohio. This water was used as an untreated source water in all pilot
plant work. "Settled" water was that obtained from the pilot plant
after alum coagulation and sedimentation. "Dual-media filtered" water
was the "settled" water after anthracite-sand filtration. "Activated
carbon filtered" water was the same settled water after passage through
1.5 meters of Filtrasorb 200 (Calgon Corporation, Pittsburgh, Pennsylvania)
-------
90
granular activated carbon (GAG). Filtration rates through this plant
were similar to those found in a conventional water treatment plant:
2 to 2.5 gpm/ft2 (5 to 6.25 m/hr).
Blank water for analytical purposes was obtained by exhaustively
purging laboratory distilled water with helium gas.
The test precursor substances [hurnic acid (Pfaltz and Bauer,
Flushing, New York, or Aldrich Chemical Company, Milwaukee, Wisconsin),
acetone (Mallinckrodt, "Nanograde," St. Louis, Missouri), acetaldehyde
(Aldrich), acetophenone (Fisher Scientific, Fairlawn, New Jersey)] were
used as obtained from the suppliers.
Standard analytical solutions of chloroform (Fisher Scientific,
"Spectro-analyzed"), bromodichloromethane (Aldrich), dibromochloromethane
(Columbia Chemical Company, Columbia, South Carolina), and bromoform
(Fisher Scientific) were prepared as described in the NORS 80-city
report (Symons et al. 1975).
Procedures
Analyses for the trihalomethanes were performed by a modification
of the "volatile" organic gas chromatographic technique described by
Bellar and Lichtenberg (1974) using specific halogen electrolytic con-
ductivity detection (Stevens and Symons 1975) as described in the NORS
80-city report (Symons et al. 1975).
Nonvolatile total organic carbon (NVTOC) was measured using the
method and apparatus described in the NORS 80-city report (Symons et al.
1975). Samples were acidified with nitric acid and purged with carbon-
free air for about 10 minutes to remove carbon dioxide before the actual
analysis. Some volatile organic materials were lost during this step.
NVTOC was defined as that organic carbon remaining in the sample after
this treatment.
Briefly, the experimental procedure was as follows: All reactions
described were carried out in the presence of phosphate buffers. Reac-
tion solutions were made up at pH 7 and adjusted to the desired pH with
the addition of either hydrochloric acid or sodium hydroxide. Reaction
mixtures were prepared with the appropriate source water, buffer was
-------
91
added, and pH was adjusted. The mixtures were then spiked with the
test compounds, and chlorine stock solution was added. The reaction
mixtures were typically one to two liters. Immediately after mixing,
zero time samples were taken by pouring from the larger vessel into a
50-ml serum vial containing an appropriate quantity of 0.1 N sodium
thiosulfate (Fisher Scientific) to halt the reaction by removing chlorine.
Samples for storage (extended reaction time) were taken in a similar
manner without sodium thiosulfate. All vials were sealed headspace-
free with Teflon-faced septa immediately after filling as described in
the NORS 80-city report (Symons et al. 1975). The sealed samples were
stored at the indicated temperature in either a water bath or incubator
controlled at ±0.5°C. At the appropriate time, the vials were opened,
and aliquots were quickly transferred to a 30-ml vial containing sodium
thiosulfate. The smaller vial (headspace-free) was then sealed as
described above. All preserved samples were then stored under refrigera-
tion until analysis.
RESULTS AND DISCUSSION
Precursor at pH 7
General. Trihalomethanes must result from a reaction or series of
reactions of chlorine with a precursor material. Simple methyl ketones
react through the classical haloform reaction mechanism. More complex
substances, such as humic materials, also react by this mechanism or by
some other mechanism that includes an oxidative cleavage step. Because
control of trihalomethane production by precursor removal or control of
precursor reaction rate was considered the best approach, some knowledge
of precursor identity was required. Suggestions, as mentioned above,
as to identity of precursor varied from complex humic materials to simple
methyl ketones or simple compounds with the acetyl moiety.
The earliest work at this laboratory with precursor removal was
simply an experiment to determine whether GAG adsorption had any effect
on precursor concentration. In this work, samples of water taken from
the pilot plant were chlorinated at a dose of 8 mg/1 — that used at that
-------
92
time by the Cincinnati Water Treatment Plant on the same raw water to
satisfy chlorine demand and maintain a free residual in the distribution
system. In this experiment, not only were settled and activated carbon
filtered water samples chlorinated to determine the effect of the
carbon, but dual-media filtered and raw water samples were also chlo-
rinated at the same concentration for comparison. All four samples
were buffered at pH 7. The results in Fig. 2 show that when the result
of chlorination of fresh GAG filtered water was compared with the result
of chlorinating the settled water, removal of precursor was indicated.
The effectiveness of GAG filtration, however, was shown later to be
relatively short-lived — a matter of only a few weeks under conditions
of pilot plant operation (Love et al. 1975). The other important aspect
of this experiment was the observed dramatic change in rate of chloroform
formation when the results of raw and settled water chlorination were
compared. Conventional alum coagulation and sedimentation caused the
removal of most of the precursor material from the raw water.
Particulates. The above experimental results indicated that
precursors are one or more of the following: some sort of particulate,
a substance associated with particulates, a substance reacting in associa-
tion with the particulates, or possibly a substance that could be com-
plexed with the alum and precipitated with the floe. The nature of the
role of the particulates was, therefore, further investigated. A simple
vacuum filtration of raw water through filter paper (Whatman No. 1) was
carried out. The filtrate, particulates with filter resuspended in GAG
filtered water, original raw water, and GAG filter effluent with and
without filter paper were each chlorinated and subsequently analyzed for
trihalomethane content after varying periods of storage.
Comparison of the reaction rate curves for raw and filtered raw
water shown in Fig. 3 illustrates a reduction of the rate of trihalo-
methane production caused by removal of particulates. The rate curve
for GAG filter effluent with resuspended filter paper and particulates
from the raw water indicates that essentially all of the difference
between the raw and filtered raw rate curves can be accounted for by
the substances trapped on the resuspended filter paper. The curves
-------
93
200-r
CI2 EXHAUSTED
z
SETTLED
FRESH GAC FILTERED
60
STORAGE TIME, HRS.
-h-
90
Fig. 2. Effect of treatments on chloroform production.
chlorine dose, 8 mg/1; 25°C; pH 7.
120
Conditions:
-------
94
120-r
CARBON FILTER EFFLUENT + FILTER PAPER
CARBON FILTER EFFLUENT
30 40 50
STORAGE TIME, MRS.
Fig. 3. Effect of simple filtration on trihalomethane production.
Conditions: chlorine dose, 10 mg/1; pH 7; 25°C.
-------
95
for GAG filter effluent and GAG filter effluent plus filter paper are
simply the appropriate controls and are nearly identical. They indicate
essentially no reaction interference or enhancement by the filter paper
itself. According to these results, simple filtration either removed
some trihalomethane precursor from the raw water or the removal of some
of the particulate matter reduced the reaction rate of dissolved
precursor. The particulate matter, therefore, played some direct role
in trihalomethane production when Ohio River water was chlorinated.
To determine which of these mechanisms was important, the effect
of potentially active surfaces was investigated by spiking two sets of
GAC-filtered water samples with simple acetyl derivatives and then
suspending Bentonite clay in one set and powdered activated carbon in
the other set. Neither of the two added particulates caused any detect-
able increase in rate of trihalomethane formation. Therefore active
surface effects were not considered significant but particulate matter
or substances strongly sorbed on the particulate matter were found to
be important precursors to trihalomethane production at pH 7.
Humic acid. Because humic substances are more likely to be found
in natural waters as small particulates or sorbed on clay particles
(Schnitzer and Khan 1972) than are soluble simple methyl ketones, a
direct test of Rook's (1974) hypothesis was attempted using commercially
available humic acid, both suspended at pH 7 and dissolved at a higher
pH but later readjusted to pH 7. At concentrations of humic acid
representing an NVTOC concentration similar to that found for Ohio River
water (approximately 3 mg/1 of NVTOC), the rate curve for formation of
trihalomethanes was observed to be very similar to that seen for chlo-
rination of the natural water (Fig. 4). In addition, a filtration
experiment (0.2 pm pore filter) similar to that carried out on the raw
water described above was conducted on suspensions and solutions of
humic acid. The results (Fig. 5) observed were similar to those
reported for the raw water filtration experiment (Fig. 3). Thus, in
terms of rate of trihalomethane formation on chlorination, the physical
and chemical characteristics of humic acid in suspension and solution
at these concentrations were found to be similar at pH 7 to those of the
unknown precursor substances present in the Ohio River.
-------
96
180 -r
160--
10
20
30 4O 5O
REACTION TIME (hrs)
6O
7O
80
Fig. 4. Comparison of humic acid, raw water reaction rates at
similar NVTOC concentrations. Chlorine dose, 10 mg/1.
-------
97
180T
DISS°READJUSTED, F.LTERED
SUSPENDED, FILTERED
30 40 50
REACTION TIME (hrs)
Fig. 5. Chloroform production from filtered and unfiltered humic
acid mixtures, 5 mg/1. Conditions: chlorine dose, 10 mg/1; pH 7.
-------
98
Finally, attempts to react chlorine at pH 7 with simple acetyl
compounds (acetone, acetaldehyde, and acetophenone), when these com-
pounds were spiked at 5 pmol/1 into raw- and GAC-filtered water, failed
to produce trihalomethanes at rates significantly higher than those
observed for the blank samples (Fig. 6). Therefore, for chlorination
of natural waters at pH values near 7, the humic acid precursor hypothesis
of Rook seemed the most valid.
Effect of pH on reaction rate and precursor identity
General. Because the rate determining step of the classical halo-
form reaction is enolization of a ketone, the rate of trihalomethane
formation is pH dependent. For example, the reaction of acetone with
hypochlorite to form chloroform proceeds at a faster rate at pH 11.5
than at pH 6.5. Experimentally, a sample of settled water was buffered
at pH 6.5 and another at pH 11.5 and both were chlorinated at an
initial concentration of 10 mg/1. The results (Fig. 7) illustrate that
the rate of formation of chloroform increases with an increase in pH.
This could be explained simply by an increase in the humic acid reaction
rate, as would be expected by the classical mechanism. Another possi-
bility, however, is that other compounds such as acetone in the source
water (settled) that do not react readily at pH 6.5 become significant
contributors to the overall reaction rate (chloroform formation) at
pH 11.5. An indication of the latter possibility was previously noted
in the work of Fairless et al. (1975) in which acetone was shown to react
at a significant rate at pH 9.5, but not at a pH near 7. Because chlo-
rination is carried out at high pH in some water supplies, especially
where lime softening or excess lime softening is practiced, further
investigation of the effect of pH was necessary.
Humic acid. Figure 8 illustrates the reaction rate curves for
formation of total trihalomethanes (TTHM) from three concentrations of
humic acid (0.1, 0.5, and 1.0 mg/1) spiked in GAC-filtered water in
presence of excess chlorine (10 mg/1 with less than 10% change during
the course of the experiment). An apparent first-order rate dependence
on initial humic acid concentration is graphically demonstrated; that
-------
99
200T
180 .
160
140-•
^120
o>
3
0100
o
oc
o
O 80
60--
40--
20
* ACETONE
• ACETALDEHYDE
o ACETOPHENONE
• BLANK
CARBON FILTERED
10
20
30 4O 50
REACTION TIME (hrs.)
60
70
80
Fig. 6. Chloroform production from raw and carbon filtered water
spiked at 5 uM with low molecular weight acetyl compounds. Conditions:
chlorine dose, 10 mg/1; pH 7.
-------
120--
o
o
10
20
30
40 50
REACTION TIME (hrs)
60
70
80
90
Fig. 7. Effect of pH on chloroform production from settled water.
Conditions: chlorine dose, 10 mg/1; 25°C.
-------
0.3--
0.2--
o
I
10
Fig. 8.
production.
30
40
50
TIME (hrs)
60
70
80
90
100
Effect of humic acid concentration on trihalomethane
Conditions: chlorine dose, 10 mg/1; pH 6.7; 25°C.
-------
102
is, at any given time between any two curves, the ratios of concentra-
tions of TTHM produced are equal to the respective ratios of initial
humic acid concentrations. The change in rates with apparent exhaustion
of reaction sites can also be seen- as nearly constant TTHM concentrations
are approached.
In Fig. 9 the pH dependency of reaction rate at one of these concen-
trations (1 mg/1) is illustrated. The same curve characteristics were
observed at all pH values. As noted above, one can assume from the
shape of the curves that the reaction was nearly complete at pH 6.7 or
was proceeding very slowly relative to the initial rate. Because the
reaction is essentially complete at pH 6.7 at the end of the experiment,
the nearly two-fold increase in final product concentration at pH 9.2
can only be explained by the presence of certain reactive sites on the
complex humic acid molecule that react at insignificant rates at the
lower pH but are reactive at higher pH. The concentration of signifi-
cant reactive sites in the reaction mixture, when expressed as equiva-
lents per liter, is therefore at least twice as high at the higher pH.
Based on this analysis, and considering humic acid to be 60% carbon,
0.7% and 1.4% of the carbon present reacts ultimately to become trihalo-
methane at the low and high pH values, respectively.
Acetone. Reactions of acetone with chlorine can be compared
quantitatively with those of humic acid in an evaluation of the potential
role of acetone as a precursor because the similarity of the humic acid
reaction to that of the natural material in the source water has already
been demonstrated (see Fig. 4). Figure 10 shows the pH dependency of
the rate of reaction of 1 mg/1 acetone. At pH 6.7 the TTHM concentra-
tion from acetone after 96 hours is about one-third of that observed
from 1 mg/1 humic acid in the same 96-hour period (see middle curve,
Fig. 9). These numbers might seem to indicate that acetone could be a
significant precursor at pH 6.7. Because the rate of trihalomethane
production from acetone through the classical haloform reaction mechanism
is known to be proportional to acetone concentration, however, 3 mg/1
of acetone would be required to give the same TTHM concentration at
96 hours as that from 1 mg/1 of humic acid. Therefore, approximately
-------
0.7 -
40 50
TIME (hrs)
60
70
80
H
O
U)
90
100
Fig. 9. Effect of pH on trihalomethane production from humic acid,
1 mg/1. Conditions: chlorine dose, 10 mg/1; 25°C.
-------
.7 •
.6
.5--
5 -4--
o
-3 +
.2--
.1--
10
20
30
40 50 60
TIME (hrs)
80
90
100
Fig. 10. Effect of ptl on trihalomethane production from acetone,
1 mg/1. Conditions: chlorine dose, 10 mg/1; 25°C.
-------
105
15 mg/1 of acetone would be required to give the concentration of
chloroform observed for the raw water (Fig. 4). Thus, if acetone were
the important precursor at pH 6.7, sufficient acetone would be required
in solution to account for over 9 mg/1 of NVTOC, which far exceeds the
2 to 3 mg/1 NVTOC usually found in the source water (acetone is not
easily lost in the C02 stripping during NVTOC sample preparation).
Furthermore, the reaction rate curve for acetone at pH 6.7 is
nearly a straight line which indicates no change in rate during the
experiment. By again using the assumption that acetone reacts by the
classical haloform reaction mechanism and from the final trihalomethane
concentration observed, less than one percent of the acetone initially
present was calculated to have reacted. Because this change of acetone
concentration was insignificant, its effect on reaction rate was not
observed in this experiment. An insignificant change was expected,
based on calculations using a reported rate expression for acetone in
the haloform reaction (Manufacturing Chemists Association 1972). There-
fore, if acetone was the most important precursor and if its concentra-
tion was high enough to account for the observed rate of trihalomethane
production from the source water, the characteristic rate curve would
be linear as plotted. For these two reasons acetone is not likely to
be a significant precursor at pH 6.7.
At pH values much higher than 6.7, however, the situation could be
different. Figure 10 has been plotted on the same numerical scale as
Fig. 9, so that a direct comparison of reaction rates between acetone
and humic acid at the various pH values is possible. A comparison of
the curves on these figures, representing the trihalomethane formation
rates at the higher pH values, reveals a much larger increase in reaction
rate of acetone with changing pH than that observed with the same con-
centration of humic acid. The 30-fold observed increase (graphically
measured) in acetone reaction rate was also expected from calculations
based on the reported rate expression (Manufacturing Chemists Association
1972). A rate increase of this magnitude could allow as little as
500 yg/1 (i.e., 15 mg/1 divided by 30) of acetone to account for the
trihalomethanes formed on raw water chlorination at pH 10.2. Therefore,
-------
106
low molecular weight compounds containing the acetyl moiety that have
haloform reaction rates similar to that of acetone can become signifi-
cant contributors to total trihalomethane production where chlorination
is carried out at high pH. Thus, both possible explanations for the
effect of pH on reaction rate noted in the discussion of Fig. 7 are
valid.
The question of precursor identity is, therefore, complicated
because the "precursor" is actually a mixture of compounds with differing
reactivities at varying pH values, solubilities, and other physical and
chemical characteristics. The relative contributions of the various
constituents of a given water depend somewhat on the treatment practiced
as well as on the source of the water. The probable diverse nature of
precursor also may hamper efforts to find a single general organic
parameter for unit process control that will predict effective removal
of precursor.
Temperature
The effect of temperature on the rate of reaction of precursors
present in Ohio River water was investigated to assess the potential
effect of wide seasonal temperature variations in raw and treated waters.
The winter-to-summer water temperature variation in the raw and finished
water at Cincinnati, Ohio, is approximately 26°C (from less than 2°C to
greater than 28°C). The results presented in Fig. 11 show that this
temperature differential could easily account for most of the winter
to summer variation in chloroform concentration (less than 30 ug/1 to
greater than 200 ng/1) observed in Cincinnati tap water over the past
year when raw water chlorination with a 3- to 4-day chlorine contact
time was practiced. Some other factors, such as seasonal variation in
precursor concentration, certainly have some additional effect, however.
Disinfectant
Work is progressing with measurement of the effects on trihalo-
methane production of the use of oxidants other than chlorine as disin-
fectants (e.g., 03, C102). When completed, the results of these
-------
107
225--
200--
O)
3
Z150--
O
UJ
O
O
O
100 -
cc
O
u.
O
cc
o
x
u
50--
20 40 60
TIME (hrs)
80
100
120
Fig. 11. Effect of temperature on chloroform production from
raw water. Conditions: chlorine dose, 10 mg/1; pH 7.
-------
108
experiments will be the subjects of future reports. The work reported
herein was confined to a study of the effect of chlorination practice,
given the presently recognized need for maintenance of a chlorine
residual in the distribution system. Chlorination in the presence
of added ammonia is practiced in some locations in an attempt to main-
tain residuals (as chloramine) for extended periods of time. Figure 12
illustrates the result of an attempt to form trihalomethanes with chlo-
rine added in the presence of added ammonia. Chlorine was added at
5.5 tng/1 (measured) to raw water and to raw water spiked with 20 mg/1
NH^Cl (ammonia-nitrogen, 5.2 mg/1). The results of the measurements for
trihalomethane production and free- and combined- (mostly NH2C1) chlorine
residuals in Fig. 12 show that when combined chlorination was practiced,
trihalomethane production was minimized. Therefore, during chlorination
of water where ammonia breakpoint is not achieved, trihalomethane pro-
duction may not be a problem. At this time, however, ammoniation is not
recommended as a technique to avoid trihalomethane formation because of
the relatively poor disinfecting power of chloramines when compared with
that of free chlorine.
Full-scale plant operation
The reduction of ultimate trihalomethane concentration in finished
drinking water is the primary goal of on-going field research at a number
of water treatment plants in the United States. Preliminary results of
this work indicate that the conclusions drawn above with regard to the
role of coagulation and settling in reducing precursor concentration
are valid, although analysis of data are not yet completed. The details
of this field work will be the subject of future papers.
SUMMARY AND CONCLUSIONS
The precursor to trihalomethane production during the chlorination
process in drinking water treatment is probably a complex mixture of
humic substances and simple low molecular weight compounds containing
the acetyl moiety. The relative importance and contribution to trihalo-
methane production of each of the specific precursor compounds are pH
-------
109
1.35 -
1.2 -
1.05 -
0.9 •-
0.45 -
0.15 "
TTHM, RAW -i- NH3
-4-
+
+
10 20 30 40 50
REACTION TIME (hrs)
60
70
T10
9
•8
7
o>
6*
c w
5 LU
oc
UJ
z
E
49
•3
•2
- 1
80
Fig. 12. Effect of free vs combined chlorine on TTHM production
at pH 7.
-------
110
dependent. Where chlorination following clarification is carried out
at pH values near 7, effective coagulation and sedimentation may be
sufficient to reduce the precursor concentration to levels where ulti-
mate trihalomethane concentrations are below the yet undefined adverse
health effect levels. Where chlorination is carried out at high pH
(as in a lime- or excess lime softening plant), treatment for precursor
removal is more complicated. In these cases, removal of relatively
water soluble low molecular weight compounds (concentrations of which
would not be expected to be significantly affected by coagulation and
settling processes) is also necessary before chlorination. Thus, the
point of chlorination in the treatment process, being a significant
factor in trihalomethane production, probably represents the most
important variable to be considered for change in attempts to reduce
ultimate trihalomethane concentrations in finished drinking water.
To date, GAG has been used with only limited success to remove
precursor compounds. Because its effectiveness is limited to only a
few weeks after being placed in filters, its use would require frequent
activation or replacement cycles.
Work is continuing in an effort to determine ways to reduce the
extent of trihalomethane reaction through precursor removal or control
of reaction rates. The final evaluation of the success of this work
must, however, await more precise health effect information regarding
the significance of the presence of trihalomethanes in drinking water.
ACKNOWLEDGMENTS
We acknowledge the assistance of the Research Sanitary Engineers,
0. T. Love and J. K. Carswell and accompanying staff, who were responsible
for pilot-plant aspects of this work; B. L. Smith, Physical Science
Technician, for NVTOC analyses and some chlorine residual measurements;
J. M. Symons and J. K. Carswell for review of the manuscript, and
Mrs. M. Lilly for its preparation.
REFERENCES
American Public Health Association (APHA). 1971. Standard methods for
the examination of water and wastewater, 13th Ed., New York.
-------
Ill
Bellar, T. A., and J. J. Lichtenberg. 1974. The determination of
volatile organic compounds at the yg/1 level in water by gas
chromatography. EPA-670/4-74-009. U.S. Environmental Protection
Agency, National Environmental Research Center, Cincinnati, Ohio.
Bellar, T. A., and J. J. Lichtenberg. 1974. Determining volatile
organics at the ug/1 level in water by gas chromatography. J. Am.
Water Works Assoc. 66: 739.
Bellar, T. A., J. J. Lichtenberg, and R. C. Kroner. 1974. The occurrence
of organohalides in chlorinated drinking water. J. Am. Water Works
Assoc. 66: 703.
Bunn, W. W., B. B. Haas, E. R. Deane, and R. D. Kleopfer. 1975. Forma-
mation of trihalomethanes by chlorination of surface water. Environ.
Lett. 10: 205.
Fairless, B. 1975. U.S. Environmental Protection Agency, Region V,
Central Regional Laboratory, Chicago, Illinois. Personal
Communication.
Glaze, W. H. 1975. North Texas State University, Denton, Texas.
Personal Communication.
Love, 0. T., Jr., J. K. Carswell, A. A. Stevens, and J. M. Symons.
1975. Treatment of drinking water for prevention and removal of
halogenated organic compounds (An EPA Progress Report). Presented
at the 95th Annual Conference of the American Water Works Associa-
tion, June 8-13, Minneapolis, Minnesota.
Manufacturing Chemists Association. 1972. The effect of chlorination
on selected organic chemicals. Project 12020 EXG 03/72, U.S.
Environmental Protection Agency, Washington, D.C.,
Rook, J. J. 1974. Formation of haloforms during chlorination of
natural waters. Water Treat. Exam. 23(2): 234.
Rook, J. J. 1975. Formation and occurrence of chlorinated organics
in drinking water. Presented at the 95th Annual Conference of the
American Water Works Association, June 8-13, Minneapolis, Minnesota.
Schnitzer, M., and S. U. Khan. 1972. Humic substances in the environ-
ment. Marcel Dekker, Inc., New York.
Stevens, A. A., and J. M. Symons. 1975. Analytical considerations for
halogenated organic removal studies, p. XXVI-1. In Proc. Am. Water
Works Assoc. Water Quality Technology Conference, December 2-3,
Dallas, Texas.
-------
112
Symons, J. M., T. A. Bellar, J. K. Carswell, J. DeMarco, K. L. Kropp,
G. G. Robeck, D. R. Seeger, C. J. Slocum, B. L. Smith, and
A. A. Stevens. 1975. National organlcs reconnaissance survey for
halogenated organics in drinking water. Water Supply Research
Laboratory and Methods Development and Quality Assurance Laboratory,
National Environmental Research Center, U.S. Environmental Protec-
tion Agency, Cincinnati, Ohio. J. Am. Water Works Assoc. 67: 634.
U.S. Environmental Protection Agency (EPA). 1974. Lower Mississippi
River facility, New Orleans Area Water Supply Study (Draft
Analytical Report), Slidell, Louisiana.
DISCUSSION
John R. J. Sorenson, Quad Corporation. I am still concerned with the
purity of chlorine used in water purification. What grade of chlorine did
you use in your studies? Isn't that a purer grade of chlorine than used
in waste purification? Finally, would you comment on the possibility
that the concentration of chlorinated organic compounds in drinking water
might be reduced by using purer chlorine gas, without substantial amounts
of CH2C12, CHC13, CCli^, C2Clg, chlorinated aromatic hydrocarbons, etc.?
Stevens . We are using a high purity grade and, of course, we still
get the trihalome thanes . They are produced on reaction of the chlorine
with the precursor compounds. This is clear. It was one time suspected
that in some commercial chlorine carbon tetrachloride was a contaminant .
If this were the case you should see not an increase with time of carbon
tetrachloride, but as soon as you chlorinate, you should see an immediate
increase that stays constant. That is the same with any other contaminant
compound. We've never observed this, I'm not sure we've ever really
carefully looked for it. In the Nationwide Organics Reconnaissance
Survey, we did look for it with the two compounds that weren't formed
in chlorination that I mentioned, carbon tetrachloride being one. We
don't feel we have strong evidence or any significant results that would
show that any finished waters contain higher carbon tetrachloride levels
than raw waters. We see variations in the data that might look, if you
don't think about them in the context in which the numbers were taken,
that carbon tetrachloride was higher in some cities. But those numbers
are all very small values and usually approaching or even below the
detection limit on the day the raw water was run. So we have to be
careful about interpretation of that. In any case, we saw no significant
increases of carbon tetrachloride. Now, our study was for 80 different
cities. I'm not saying what you say is impossible. We just haven't
seen it.
Joseph J. Delfino, University of Wisconsin. After the EPA survey
last fall, there was a lot of talk about what these 300 micrograms per
liter meant and then a lot of discussion came up about cough syrup con-
taining 1% chloroform. What is the significance of 1% CHC13 in cough
syrup compared with 100 ppb in drinking water in terms of health? Does
-------
113
the EPA have any toxicological data as to the significance of say
somebody drinking a certain amount of chloroform that way in terms of
relating it to drinking water consumption?
Stevens. Well, as I said in my last statement, we don't really
know the health significance of the chloroform levels we have observed
in the water. We don't know what the health ramifications are of
drinking that water. You can obviously calculate what you take from
cough syrup, if you take a lot of cough syrup, and what you get in
drinking water at so many hundred micrograms per liter. The chronic
versus the acute problem hasn't been evaluated. We don't have those
answers. We're waiting for them from those who are working on them.
Keith Lawson Murphy. McMaster University. Were you working with a
straight humic material compound or did it contain fulvic acid as well?
S tevens. We were working with what Aldrich Chemical Company
markets as humic acid. It is relatively insoluble in water. I think
that if it contained fulvic acids, as you call them, you would have seen
a soluble portion and an insoluble portion. I think it's humic acid
according to the classical definition which is pretty much insoluble at
pH 7. As I said, we went to alkaline solutions, say around 0.01 N
sodium hydroxide, to dissolve it and then diluted and readjusted the pH.
We got essentially the same curves for the reaction of that mixture as
we did when we just suspended it by ultrasonification. However, filtra-
tion had a different effect. So, going to the base first rendered the
molecule more soluble. Whether it was the rate of going into the
solution or some kind of alkaline hydrolysis, we can't really say.
David Friedman, Food and Drug Administration. Has anyone compared
the ratio of chlorinated haloforms to brominated haloforms to the
chloride — bromide ratio in the water?
Stevens. I'm not surprised the question has come up. The problem
is getting the data on bromide content of the water. The extremes we
saw were extremely low concentrations of bromoforms, that is, none
found in well over a hundred micrograms per liter of chloroform, versus
the other extreme for the finished water at Brownsville, Texas, which had
12 micrograms per liter chloroform, 116 of bromodichloro, 106 dibro-
mochloro and 92 bromoform. A lot of the Southwestern cities and South
Texas cities, if you look in the report which will be available in the
November issue of the Journal of the American Waterworks Association —
a lot of those cities had high bromocompound concentrations. We suspect
their ground water sources are probably high in bromide. We suspect
that's probably the reason. I think Bill Glaze is going to talk about
some of this. We have also done work which gives presumptive evidence
such as that at our Kansas City EPA Laboratory in which Missouri River
water, which normally gives the usual thing that I was describing, was
spiked with KI, KBr, KC1, and KF and all ten possible haloforms were
observed, except there were no fluorocompounds. But all 10 possible
combinations of chlorine, bromine, and iodine were observed. It is
-------
114
clear to us that is how it happens. We just don't have a good, fast,
easy method for bromide determination in the presence of all that
chloride.
Max Eisenberg, Maryland State Department of Health. In your com-
parison of the chloroform concentrations of the chlorinated raw waters
versus the filtered waters have you carried your chlorinated raw water
through filtration (activated carbon) to determine the removal of
chloroform?
Stevens. We have used granulated activated carbon filtration and
it does work. We can run Cincinnati tap water through — I keep saying
Cincinnati tap water only because that is where we are, O.K.? We remove
the chloroform rather easily with fresh carbon for a few weeks. Then
it will start to come through fairly quickly. So carbon used in that
mode would have to be reactivated or replaced rather frequently. As for
following chlorinated water through a plant, we can do that anyway and
we have done it in many plants where they do chlorinate their raw water.
Of course you get a continual increase with reaction time.
-------
CHLORINATION OF ORGANICS
IN COOLING WATERS AND PROCESS EFFLUENTS
Robert L. Jolley, Guy Jones, W. Wilson Pitt, and James E. Thompson
Chemical Technology Division
Oak Ridge National Laboratory
Oak Ridge, Tenner see 37830
ABSTRACT
Many water-soluble chlorine-containing organic compounds of low
volatility were found to be present in samples of chlorinated cooling
waters from electric power-generating plants and chlorinated effluents
from domestic sanitary sewage treatment plants. Both types of samples
had been chlorinated to milligram-per-liter chlorine concentrations in
the laboratory under conditions similar to those used for treatment of
cooling waters and disinfection of sewage effluents. The chlorinated
constituents were separated from concentrates of the water samples by
high-pressure liquid chromatography. Chlorination yields (as Cl) of the
chloro-organic compounds, determined by using the radioactive tracer 36C1,
ranged from 0.5 to 3.1% of the chlorine dosage. The formation of chloro-
organics and the reaction yields correlated with the chemical compositions
of the water samples. Several chloro-organics were identified in the
typical domestic sewage effluents and were quantified at the microgram-
per-liter level. Comparison of the chromatograms of the chlorinated
constituents in the cooling water samples with those of the sewage process
effluent samples revealed a high degree of correspondence with respect to
the elution positions of the separated constituents. A compilation of
relevant data concerning organic constituents in natural waters and sewage
process effluent is presented. The chemical species subject to chlorination
during the cooling water and sewage treatment are discussed.
INTRODUCTION
For several decades, chlorine has been the principal means for
controlling water-borne diseases via disinfection of drinking waters
and sewage effluents (Morris 1971, White 1972). More recently, with
the rapid expansion of the electric power industry, chlorine has
achieved major importance as an antifoulant for the cooling systems of
115
-------
116
electric power-generating plants (Draley 1972, White 1972). Concomitant
with the increased understanding of the ecological problems associated
with the ubiquitous chlorinated pesticides and polychlorinated biphenyls
(Van Middelem 1966; RLsebrough et al. 1968; Veith and Lee 1970) has been
the increasing concern that chloro-organics may be formed in the environ-
ment through the chlorination of various waters and process effluents
(Ingols and Jacobs 1957, Dugan 1972, Weber 1972, Jolley 1973). Evidence
to confirm that chloro-organics are produced during the chlorination of
sewage treatment plant effluents (Glaze et al. 1973; Jolley 1973, 1974,
1975), cooling waters (Jolley, Gehrs, and Pitt 1975), and potable waters
(Rook 1974; Bellar, Lichtenberg, and Kroner 1974) has been presented only
recently, because of the need to develop new methodologies for carrying
out the required analyses.
In this paper we will review the available information concerning
organic constituents in natural waters, including effluents from sewage
treatment plants, and discuss selected aspects of aqueous chlorination.
We will also summarize previously detailed chlorination studies with
sewage effluents (Jolley 1973, 1974, 1975) and cooling water (Jolley,
Gehrs, and Pitt 1975), and present results from a recent study with a
sample of cooling water. Finally, several general conclusions will be
drawn concerning the total environmental impact of water chlorination
and possible ecological ramifications.
SOLUBLE ORGANIC CONSTITUENTS
Assessment and prediction of chlorination effects on organic
compounds present in waters of environmental interest require a
knowledge of the identity, concentration, and nature of such compounds.
Several major efforts have been made to collect the known and published
data relative to the soluble organic constituents in a variety of waters.
Vallentyne (1957) comprehensively summarized the information available
at that time concerning organic matter in natural waters, sewage, and
soil. Little (1970) prepared a report which listed the organic compounds
present in freshwaters principally as a result of pollution. Hood (1970)
and Faust and Hunter (1971) edited proceedings of conferences which made
-------
117
significant contributions to the understanding of the complex nature
and activity of organic substances in the aquatic environment. The
EUROCOP report (1973) compiled qualitative and quantitative data on
naturally occurring organic compounds and industrial pollutants in
process effluents, natural waters, and tap water. Pitt, Jolley, and
Katz (1974) and, later, Pitt, Jolley, and Scott (1975) presented
quantitative information about the soluble organic constituents of low
volatility in effluents from both primary and secondary stages of
municipal sewage treatment plants. The report by the WHO International
Reference Centre for Community Water Supply (1975) tabulated qualitative
and quantitative data for 289 organic constituents that have been
identified in wastewater, river water, and drinking water. The U.S.
Environmental Protection Agency (1975) has prepared an exhaustive
compilation of the known organic constituents in water supplies. Symons
et al. (1975) recently presented quantitative data about the volatile
halogenated organic compounds present in the drinking water of 80
metropolitan areas. In a recent compilation, Junk and Stanley (1975)
summarized data collected from the literature published through April
30, 1975, for potable and river waters but did not include organics in
sewage effluents.
Of these sources of information, only the conference proceedings
(Hood 1970, Faust and Hunter 1971) deal in any significant way with
humic substances. These materials are reported to comprise 90% of the
soluble organic matter in sewage effluents (Rebhun and Manka 1971, Manka
et al. 1974) and 90% of the soluble organic matter in surface waters
(Junk and Stanley 1975). Humic materials are a generic type of organic
substances classified according to solubility. These complex polymers,
which range in molecular weight from several hundred to many thousand,
are composed of a variety of subunits such as aromatic and alicyclic
moieties containing alcoholic, carbonyl, carboxylic, and phenolic
functional groups (Steelink 1963, Christman and Minear 1971, Schnitzer
and Khan 1972, Stevenson and Goh 1972, Wershaw and Goldberg 1972).
Voluminous tables of data are available in the cited references
and will not be reproduced in this paper. Only relevant data for
selected chemical species and groups will be considered during the
discussion of aqueous organic chlorination reactions.
-------
118
AQUEOUS CHLORINATION REACTIONS
The aqueous chemistry of chlorine and possible organic reactions
in aqueous systems have been already treated in considerable depth by
previous speakers at this conference. Therefore, we will only briefly
summarize the chemistry germane for understanding chlorination effects
on organic constituents in cooling waters and sewage effluents.
The reactive chlorine-containing species formed by the addition of
chlorine to natural waters or process effluents that are at near-
neutral pH values and also contain ammonia and organic nitrogen compounds
are generally considered to be the following: hypochlorous acid (HOC1),
hypochlorite ion (OC1~), monochloramine (NH2C1), dichloramine (NHC12),
organic monochloramine (RNHC1 or R2NC1, in which R represents any
organic moiety), and organic dichloramines (RNC1-). At the concen-
trations of chlorine normally used for the chlorination of cooling
water, molecular chlorine (Cl ) and the chloramine nitrogen trichloride
(NC1_) are not significant except at pH values less than 4. Equilibrium
conditions are established very rapidly for HOC1, OC1~, and NH2C1, but
somewhat more slowly for NHC1». The equilibrium concentrations of these
species are dependent on pH, temperature, and initial chlorine and
ammonia concentrations (Morris 1967, Jolley 1973). The nature and
concentrations of the reactive chlorine-containing species are of critical
importance in determining the formation and yield of chloro-organic
compounds. HOC1 is known to be an effective chlorinating agent for
aromatic organic compounds in aqueous solution (Burttschell et al. 1959,
Lee and Morris 1962). Morris (1967) estimated HOC1 to be four orders
of magnitude more effective as a chlorinating agent than NH Cl. Several
authorities indicate that NH-C1 is principally an aminating agent in aqueous
solutions (Colton and Jones 1955; Drago 1957; Theilacker and Wegner 1964;
Kovacic, Lowery, and Field 1970) but no information concerning the
effectiveness of NHC1- as a chlorinating agent in aqueous solution is
available. The presence of organic chloramines is, of course, dependent
on the presence of organic nitrogen compounds, such as amines and amides,
in the water. Although amides form the corresponding chloramine slowly,
amines are known to react rapidly; in turn, the chloramine product is a
very effective chlorinating agent for phenolic compounds (Morris 1967).
-------
119
Hence the presence of organic amines in some waters may be significant
with respect to the formation of chloro-organic compounds. If HOC1
is the principal chlorinating agent, then it follows that the formation
and yield of chlorine-containing organics in the cooling waters and
process effluents should be proportional to the HOC1 concentration.
Conversely, all other things being equal, a smaller quantity of
chlorinated organics would be anticipated in solutions with high ammonia
concentrations (Jolley, Gehrs, and Pitt 1975) because of the extremely
rapid reaction of HOC1 with ammonia to form NH Cl (Weil and Morris 1949).
Assuming that HOC1 is the major chlorinating agent, the possible
chemical reactions with organic constituents in aqueous solution may
be grouped into several general types according to Jolley (1973), namely:
(1) Oxidation
(2) Substitution
a. Formation of N-chlorinated compounds
b. Formation of C-chlorinated compounds
(3) Addition
Oxidation may be the predominant type of reaction occurring between
HOC1 and organic compounds in natural and process waters (Jolley 1973),
although some conflicting evidence has been reported (Zaloum and Murphy
1974; Murphy, Zaloum, and Fulford 1975). Many organic compounds are
subject to oxidative reactions (Hoist 1954). The carbohydrates and
carbohydrate-related compounds are examples of organic constituents in
natural waters and process effluents which are probably subject only to
oxidative reactions in aqueous chlorine solutions (and not to substitution
and addition reactions). These constituents are present in parts-per-
billion concentrations (Vallentyne 1957; Pitt, Jolley, and Scott 1975).
Table 1 lists the identities and concentrations of carbohydrate-related
constituents determined in a typical effluent from the primary stage of
a municipal sewage treatment plant (Jolley, Pitt, et al. 1975). Although
these types of compounds probably contribute to the chlorine demand of
cooling waters and process effluents, they would not result in the
formation of environmentally signficant chloro-organic compounds. .
The chemistry of the formation of N-chlorinated compounds has been
detailed by Morris (1967). Various organic nitrogen-containing compounds
-------
120
Table 1. Carbohydrate and carbohydrate-related organic
constituents in primary domestic sewage effluent
Constituent
Identification
methoda
Concentration
Ca rbohy dra t e s
Galactose
Glucose
Maltose
Polyols
Erythritol
Ethylene glycol
Galactitol
Glycerin
Aliphatic organic acids
3-Deoxyarabinohexonic acid
3-Deoxyerythropentonic acid
2-Deoxyglyceric acid
2, 5-Dideoxypentonic acid
3, 4-Dideoxypentonic acid
2-Deoxytetronic acid
U-Deoxytetronic acid
Glyceric acid
4-Hydroxybutyric acid
2-Hydroxyisobutyric acid
Oxalic acid
Quinic acid
Ribonic acid
Succinic acid
AC,
AC,
AC,
AC,
AC,
AC,
AC,
MS
MS
MS
MS
MS
MS
MS
MS
GC,
GC,
AC,
MS
MS
AC,
GC
GC
GC
GC, MS
GC, MS
GC, MS
GC, MS
MS
MS
GC, MS
GC, MS
-
-
0.5
5
3
2
15-19
7
4
7
6
13
6
6
5
6
k
2
50
k
21+
a
AC - anion exchange chromatography; GC - gas chromatography on two
columns; MS - mass spectrometry.
3Based on Flame lonization Detector response during gas chromatography.
-------
121
have been found to be present in natural waters (Vallentyne 1957) and
sewage effluents (Jolley, Katz, et al. 1975). Table 2 summarizes the
identities and concentrations of nitrogen-containing organic compounds
determined in either primary or secondary sewage effluents (Pitt, Jolley,
and Scott 1975). Many of these compounds would be subject to the generic
N-chlorination reactions shown in Fig. 1 (Morris 1967). Furthermore,
proteinaceous material such as that in bacterial cell walls, for example,
would probably be subject to chlorine substitution reactions, which lead
to the formation of N-chlorinated proteinaceous material. However, the
kinetics for the formation of N-chloro compounds from amides is considerably
slower than the formation of such compounds from amines (Morris 1967).
H
R-NH2+ HOCI >R-NCI + HOH
o o H
R-C-NH2 + HOCI >R-C-NCI + HOH
Figure 1. Chlorine substitution reactions with organic nitrogen-
containing compounds.
The substitution of chlorine into organic compounds and the
resulting formation of C-chlorinated compounds have been summarized by
Jolley (1973), Carlson et al. (1975), and Morris (1975). The two
principal reaction types of interest are substitution into aromatic or
heterocyclic compounds and the haloform reaction. As previously mentioned,
humic substances comprise the major portion of the soluble organic matter
in cooling waters and sewage effluents. These complex molecular substances
contain aromatic moieties as indicated by such degradation products as the
following: benzoic acid, catechol, 3,4-dihydroxybenzoic acid, 4-
hydroxybenzoic acid, 2-methylphenol, 4-methylphenol, resorcinol, syringic
acid, and vanillin (Christman and Minear 1971). Aromatic compounds such a
as phenols and aromatic acids are readily chlorinated in aqueous solution
-------
122
Table 2. Organic nitrogen containing constituents
in sewage effluents
Constituent
Identification
method3-
Concentration
(ppb)
Amides
Urea
Amino acids
AC, GC, MS
16-1*3
Phenylalanine
Tyrosine
Indoles
3- Hydroxyindole
Indican
Indole-3-acetic acid
Pyridine derivatives
N1 -Methyl-2-pyrir'one- 5-carboxamide
N1 -Methyl-4-pyridone- 3-carboxamide
Purine derivatives
Adenosine
Caffeine
1, 7-Dimethylxanthine
Guano sine
Hypoxanthine
Inosine
1-Methylinosine
1-Methylxanthine
3-Methylxanthine
7-Methylxanthine
Theobromine
Uric acid
Xanthine
Pyrimidine derivatives
5-Acetylamino-6-amino-3-methyluracil
Orotic acid
Thymine
Uracil
AC, GC, MS
AC, GC, MS
MS
AC, GC, F
MS
AC, CC, UV, GC
AC, UV, GC
AC, CC, UV, GC, MS
AC, CC, GC, MS
AC, CC
AC, CC, UV, GC, MS
AC, GC, MS
AC, CC, UV, GC, MS
AC, CC, UV
AC, CC, UV
AC, CC,
AC, CC, GC
AC, CC
AC, GC, MS
AC, CC, UV, GC, MS
AC, CC, UV, GC
AC, UV, GC, MS
AC, CC, GC, MS
AC, CC, UV, GC, MS
50-90
34
2
1, 2C
13
20*;, 25
10C, 14
/. 13
10°, 29-46
_
4-28, 50C
12-42, 25°
11-23, c50c
70C
_
2, 90°
-
20°
2-7, 70C
I40cc
2, 5C
7C, 9-28
16-58, 40C
AC - anion exchange chromstography; CC - cation exchange chromatography;
UV - ultraviolet spectroscopy; GC - gas chromatography on two columns;
MS - mass spectrometry; F- fluorometry.
Based on Flame lonization Detector response during gas chromatography, unless
otherwise designated.
%
'Based on uv absorbance of anion exchange chromatographic peak.
-------
123
by HOC1 or OC1~. Thus, if humic substances do contain such aromatic
moieties, they should chlorinate readily at activated sites. Furthermore,
free phenols and aromatic acids have been detected in natural waters and
sewage effluents (Vallentyne 1957; Pitt, Jolley, and Scott 1975). Data
concerning the identities and concentrations of aromatic organic compounds
found to be present in sewage effluents are presented in Table 3.
Chlorination of such compounds with HOCl should readily produce their
chlorinated analogs. For example, the reactions of phenol and resorcinol
are presented in Fig. 2 (Chandelon 1883, Lee and Morris 1962), and the
substitution reactions of benzoic, salicylic, and phthalic acids are
given in Fig. 3 (Hopkins and Chisholm 1946, Goodrich 1949).
Some organic nitrogenous compounds may also undergo chlorine
substitution reactions (Jolley 1973, Morris 1975). For example, the
pyrimidines cytosine and uracil (Table 2) react with aqueous HOCl to
form the 5-chloro analog (Patton et al. 1972; Jolley, Pitt, and '• .C'mpson
1975) as shown in Fig. 4, in addition to forming more complex degradation
products (Ramage and Landquist 1959). Purines (Table 2) may react with
aqueous HOCl to form the chlorinated analog, as shown for xanthine and
theobromine in Fig. 5. Actually, the formation of the chloro-compound
in the case of these two purines has not been corroborated in dilute
HOCl solutions; however, the reactions have been studied in nonpolar
solvents and in acetic acid (Howard 1969). The chlorination of purine
and pyrimidine bases in nucleic acid material of bacteria, plankton,
and decomposing plant and animal matter resulting in chlorinated nucleic
acid fragments or polymers may also occur. Prat, Nofre, and Cier (1965)
identified 5-chlorouracil and 5-chlorocytosine in acid hydrolysates of
nucleic acids separated from bacteria after disinfection with parts-per-
million chlorine concentrations.
A major pathway for chlorine substitution may be the haloform reaction
with humic materials, as postulated by Rook (1974). The haloform reaction
(see Fig. 6) has been discussed in detail by Morris (1975). Thus, a
principal product of the chlorination of sewage effluents and cooling
waters may be the volatile chloro-organics such as chloroform.
-------
124
Table 3- Aromatic organic acids and phenolic compounds
in sewage effluents
_ ... . Identification Concentration
Constituent ., ,n / , \
method8 (ppb)
Aromatic organic acids
Benzoic acid
2-Hydroxybenzoic acid
3-Hydroxybenzoic acid
4-Hydroxybenzoic acid
U-Hydroxyphenylacetic acid
3-Hydroxyphenylhydracrylic acid
3-Hydroxyphenylpropionic acid
Phenylacetic acid
o-Fhthalic acid
AC, GC, MS
AC, GC, MS
AC, GC, MS
AC, GC
AC, UV, GC, MS
AC, UV, GC, MS
AC, GC, MS
AC, GC
AC, UV, MS
3 „
2, 7CC
7, UOC
1
16-52, 190
10-22
6-20C
10c
200
Phenolic compounds
Catechol MS 1
p-Cresol AC, GC, MS 20 , 29
Phenol AC, GC, MS 6°, 12
aAC - anion exchange chromatography; GC - gas chromatography; UV - ultra-violet
spectroscopy; MS - mass spectrometry.
Based on Flame lonization Detector response during gas chromatography, unless
otherwise designated.
°Based on uv absorbance of anion exchange chromatographic peak.
-------
125
OH
+ HOCI
OH
Cl
+ HOH
OH
+ HOC!
OH
OH
OH
+ HOH
Cl
Figure 2. Chlorine substitution reactions with phenolic compounds.
-------
126
C02H
+ HOCI
C02H
Cl
+ HOH
C02H
OH
+ HOCI
C02H
OH
+ HOH
Cl
C02H
C02H
C02H
+ HOH
Figure 3. Chlorine substitution reactions with aromatic organic
acids.
-------
127
NH2
^
H
NH2
+ HOH
(2)
O
n
,U *
H
HN
*J,
Cl
+ HOH
Figure 4. Chlorine substitution reactions with pyrimidines.
-------
(I)
H
H
+ HOH
(2)
OH
CH
00
Figure 5. Chlorine substitution reactions with purines,
-------
129
0 0~
R-C-CH3 - > R-C = CH2 + H
9" S
R-C = CH2 -I- HOCI - > R-C-CH2CI +OH
0 0~
R-C -CH2CI - > R -C = CHCI + H+
R-C = CHCI + HOCI - >R-C — CHCI2+OH
0 0
R-C-CCI3+ OH~ > R-C — OH + CCIJ"
CCIJ + H"*" > HCCI3
Figure 6. Haloform reaction.
-------
130
The last generic type of organic reaction, i.e., addition, may
occur with organic compounds containing reactive double bonds, for
example, unsaturated fatty acids in sewage effluents and natural waters
(Vallentyne 1957; Pitt, Jolley, and Scott 1975). The reaction products
are chlorohydrin and dichloro compounds (Houben and Weyl 1962).
CHLORINATION OF SEWAGE EFFLUENTS AND COOLING WATERS
The detection and examination of chlorination effects on organic
constituents that are present in natural waters and effluents of sewage
treatment plants at parts-per-billion concentrations represent a
challenging and formidable task. Definitive information concerning
these effects (Glaze et al. 1973, Jolley 1973) has been obtained only
recently as a result of the adaptation and development of new method-
ologies. In our laboratory we developed the following stepwise procedure
(Jolley 1973, 1974) to provide realistic, reproducible results for
samples chlorinated under conditions simulating those used for disin-
fection of sewage effluents and/or antifoulant treatment of cooling
waters of electric power-generating plants:
(1) Chlorination of the water sample in the laboratory with
o/c
Cl-tagged chlorine gas or hypochlorite solution.
(2) Concentration of the radioactive chlorinated solution via
low-temperature vacuum distillation.
(3) High-pressure, high-resolution anion exchange chromatographic
separation (high-pressure liquid chromatography or HPLC) of
the chlorination reaction products in the concentrate of the
reaction mixture.
O f
(4) Detection and quantitative measurement of the Cl-tagged
chlorinated constituents using sensitive liquid scintillation
counting.
Using this analytical technique we examined the chlorination effects
on the organic constituents in both primary and secondary effluents from
a municipal sewage treatment plant and two cooling water samples. The
experimental results from the analyses of sewage treatment plant effluents
and one cooling water sample have been previously presented (Jolley 1973,
-------
131
1975; Jolley, Gehrs, and Pitt 1975). The major aspects and conclusions
from these studies will be summarized below. In addition, previously
unpublished results from the recent analysis of a second sample of
cooling water will be discussed.
Sewage treatment plant effluents
Over 50 chloro-organic constituents were separated during each
HPLC analysis of 500- to 1000-fold concentrations of primary and secondary
sewage effluents which had been chlorinated in the laboratory with parts-
per-million concentrations of chlorine. The effluent samples, which were
grab samples collected at the Oak Ridge Waste Treatment Plants, contained
essentially no industrial pollutants. Each sample was chlorinated in
the laboratory, using either chlorine gas or hypochlorite solution, to
chlorine dosages of 2.5 and 6 mg/liter for secondary and primary
effluents, respectively; the corresponding chlorine residuals (orthotolidine)
were 1 and 2 mg/liter. The chlorination contact.times were approximately
equivalent to those used for disinfection at the sewage treatment plants.
O£
The chlorination reagents contained 0.03 to 0.12 mCi of • Cl (Jolley 1973,
1974, 1975).
Q f
A typical HPLC chromatogram of the Cl-tagged chlorine-containing
constituents separated from chlorinated secondary sewage effluents is
shown in Fig. 7. Seventeen chloro-organics, prinicpally chlorinated
purines, pyrimidines, phenols, and aromatic acids, were tentatively
identified and are indicated at their respective elution positions in
the chromatogram. The identifications were made by: (1) comparing
anion exchange elution positions of the chromatographic peaks with
chloro-organic reference standards, and (2) establishing the presence
of an unchlorinated analog or progenitor organic compound in the
unchlorinated sewage effluent (Jolley 1973, 1975). We have also obtained
corroborative evidence for the presence of 5-chlorouracil by comparison
of cation exchange elution positions. More than 99% of the radioactivity
was associated with the chloride ion peak, which was eluted at 18 hr.
This high chloride concentration is supporting evidence that a major
reaction mechanism of chlorine with organics in sewage effluents is
-------
fLUTION VOLUHf.nl
4MON CX^MflNGf CH"OVaT-='1HflPH RUN COND*1"!')*!?
£ Vcm- 10 » ISO-cm STAINLESS STEEL COu^WN *r*
81? „ OlAM AMiNtX A-27 *.E!$IN; TEMPf^ATuBL
PPOOBAM. AM8lENT TO 55 "C AT 9 4 h^. f^JENT GOADIE
INCREASING iNEAfltT IN CONC ENT»flTIOr; *RCM
aWMOWUM :. .ETflTE.pM 4 4, ELJENT FLOW *?ATE
PRESSURE HOOo
ELUTION TIME ,h< 30
ELUTION VOLUME.ml
280 Ml Of EFFLUENT CHLORINATED IN LABORATORY (I Omj/lltlr COMBINED CHtOR'Nt R t S fDUAL ,0»T HOTOL 101 NE I 90 mm REACTION TIME ,
Figure 7. Chromatogram showing the Cl-tagged chlorine-containing
constituents in a sample of chlorinated secondary effluent from a
domestic sanitary sewage treatment plant.
ro
-------
133
oxidation. However, this has not been proved because chloride may
also result from the decomposition reactions of monochloramine. About
1% of the radioactivity was associated with the other chromatographic
peaks and some residual activity not removed from the resin by the
chromatographic eluent. These peaks and the residual activity represent
stable chloro—organic compounds as deduced from an extensive study of
chlorinated secondary effluents (Jolley 1973, 1975). In that study it
was shown that less than 1% of the activity associated with the chroma-
tographic peaks other than chloride was contributed by chloride-metal
or inorganic chloramine-metal complexes, or by isotopic exchange with
chlorine-containing constituents present in the effluents prior to
chlorination. Apparently, these constituents are quite stable since
their separation and detection are accomplished only after a rigorous
sample preparation procedure and subsequent chromatographic separation.
O£
Assuming complete isotopic dilution of the Cl-tagged chlorinating agent
with the inert chloride in the effluent samples, the reaction yields of
chloro—organics (as Cl) during the chlorination of both primary and
secondary sewage effluents were calculated to be about 1% of the chlorine
dosage for the reaction times customarily used for disinfection (Jolley
1973, 1974, 1975).
The following major conclusions were deduced from the analytical data
obtained during these studies of chlorinated sewage treatment plant
effluents (Jolley 1973, 1975):
1. Stable chloro-organic compounds are formed during the chlorination
of sewage effluents at parts-per-million chlorine concentration.
2. The chlorination yield of chloro-organic compounds (as Cl) is
about 1% of the chlorine dosage when disinfection reaction conditions
are used.
3. The types of organic products formed included chlorinated phenols,
purines, pyrimidines, and aromatic acids at the parts-per-billion concen-
tration level.
-------
134
Watts Bar Lake: Cooling water for Kingston Steam Plant
Over 50 chloro-organic constituents were separated by HPLC from a
1500-fold concentrate of Watts Bar Lake water which had been chlorinated
O£
in the laboratory with Cl-tagged chlorine gas (Fig. 8). The water
sample was collected at the cooling water inlet of the Kingston Steam
Plant, which is a coal-fired electric power-generating plant operated
by the Tennessee Valley Authority and located on Watts Bar Lake at
Kingston, Tennessee. A 1.5-liter aliquot of the grab sample of cooling
water was chlorinated, at a Cl dosage of 2.1 mg/liter, to a chlorine
residual (orthotolidine) of 1 mg/liter. The chlorinating reagent was
O£
tagged with 0.01 mCi of Cl per mg of Cl. After a 75-min reaction
time, the chlorine residual was destroyed by using thiosulfate (Jolley,
Gehrs, and Pitt 1975).
As in the chlorination of sewage effluents more than 99% of the
radioactivity was associated with the chloride peak which eluted at
19 hr. About 0.5% of the Cl activity was associated with the other
52 chromatographic peaks, which represent stable chloro-organic consti-
tuents as deduced by analogy with the extensive study of chlorinated
sewage effluents. Examination of the chromatogram (Fig*. 8) indicates
considerable similarity in chromatographic profile with many of the
separated chloro-organic constituents corresponding in elution position
to those of compounds separated from sewage treatment plant effluents
(Fig. 7). The chlorination reaction yield (as Cl) of chloro-organics
in this experiment was 0.78% of the chlorine dosage. The reaction
yield under actual plant antifoulant operating conditions was estimated
to be about 0.5% (Jolley, Gehrs, and Pitt 1975).
The following major conclusions were deduced from the analytical
data obtained in this study (Jolley, Gehrs, and Pitt 1975):
(1) Stable chloro-organic compounds are formed during the
chlorination of cooling waters at parts-per-million chlorine concen-
trations .
(2) The chlorination yield of chloro-organic constituents (as Cl)
is about 0.5% of the chlorine dosage under reaction conditions simulating
those used for antifoulant treatment of the cooling system of an electric
power—generating plant.
-------
ELUTION VOLUHE.nl
ANION EXCHANGE CHROMATOSRAPH RUN CONDITIONS'
DUAL 0.3-tm-10 « ISO-cm STAINLESS STEEL WITH B-12,. DIAM AMINEX A'27 RESIN,
TEMPERATURE PROGRAM,AMBIENT TO 33'C AT 93hr;ELUENT GRADIENT INCREASING
LINEARLY IN CONCENTRATION FROM 0 DISK TO 6 DM AMMONIUM ACETATE, pH 4.4;
ELUENT FLOW RATE lOml/Hr EACH COLUMN; COLUMN PRESSURE 1500 pug TOTAL.
ELUTION TIME.hr
29 30 31 32 33 34 35 36 57 38 39 40
48 49 50 51 52 53 54 55
300
CLUTIOH VOLUME, ml
3*0
._
550
2(0*1 OF WATTS BAR LAKE WATER CONCENTRATED I500X, 294n«
ZTO^I Of WATTS BAR LAKE WATER CHLORINATED IN LABORATORY (APPHOX. l.4m|/l{ltr COMBINED CHLORINE RESIDUAL .ORTHOTOLIDINE ) CONCENTRATED I500X,{ "54 n^,CT'V'TT
O (i
Figure 8. Ghromatogram showing both Cl-tagged chlorine-containing
and uv-absorbing consitutents in a sample of Watts Bar Lake water collected
from the Kingston Steam Plant cooling water inlet and chlorinated in the
laboratory. A chromatogram of the uv-absorbing constituents in the
unchlorinated cooling water is included for comparison. Both samples
were chromatographed simultaneously on the dual-column UV-Analyzer
(Jolley, Gehrs, and Pitt 1975).
u>
Ln
-------
136
(3) The HPLC chromatographic profile and peak elution positions were
similar to those obtained for chlorinated sewage effluents.
Mississippi River: Cooling water for Allen Steam Plant
As with the previous cooling water sample, over 50 stable chloro-
organic constituents were separated by HPLC from a 1470-fold concentrate
of Mississippi River water which had been chlorinated in the laboratory
(Fig. 9). This concentrate had been prepared from a sample collected
from the cooling water inlet (prior to chlorination) of the Allen Steam
Plant, which is a coal-fired electric power-generating plant operated by
the Tennessee Valley Authority and located on the Mississippi River near
Memphis, Tennessee. The total chlorination yield of chloro-organics
(as Cl) was 3.1% of the chlorine dosage after 15-min contact time. This
O£
yield was calculated assuming that isotopic dilution of the Cl-tagged
chlorinating agent occurs rapidly and essentially completely with the
inert chloride in the cooling water sample.
After collection, the water sample was stored at -60°C and thawed
just prior to use. Analytical data on the sample are presented in the
Discussion section and compared with the results obtained for sewage
effluents and the previous cooling water sample. The chlorine demand
of the sample was determined using standard methods (American Public
Health Association 1971). The linear equation expressing the relation-
ship between the chlorine dosage and the chlorine residual of the chlori-
nated cooling water after 5-min contact time is as follows:
Y = 0.38X - 0.09,
in which Y is the chlorine residual in rag/liter and X is the chlorine
dose in mg/liter. A 1.0-liter aliquot of the cooling water was
36
chlorinated with 3.4 mg of chlorine gas containing 0.033 mCi of Cl
radioactive tracer. After a chlorination contact time of 15 min, the
chlorine residual of 1 mg/liter was destroyed with a slight stoichiometric
excess of thiosulfate solution. The sample was then concentrated 1470-
fold by vacuum distillation at temperatures ranging from ambient to 35°C
-------
ELUTION TIME.hr 0 I 2 3 4 5 « !
12 IJ 14 15 16 17 18 19 20 21 22 23 21 25 26 27 28 29 3O 31 32 33 SI 35
ELUTION VOLUME .ml 0
—I—
75
100
125
175
200
250
275
I06
ANION EXCHANGE CMROMATOGRAPM RUN CONDITIONS:
OJ-cm- 1.0. > ISO-cm STAINLESS STEEL COLUMN WITH
6-12 p OIAM AMINEX A-27 RESIN) TEMPCRATURC
PROGRAM, AMBiCNT TO 55 "C AT 9.5 It,; ELUENT GRADIENT
INCREASING LINEARLY IN CONCENTRATION FROM 0.015 .M
AMMONIUM ACETATE, pH ««; ELUENT FLOW RATE
10 iil/ltr EACH COLUMN; COLUMN PRESSURE 1500 pll?
TOTAL.
U)
ELUTION TIME.ftf 36 37 SB 39 40 41 42 43 44 45 46 47 40 49 50 51 SZ 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 «9 TO 7
ELUTION VOLUME,ml
I 1 1 1 1 1 1 1 1 ,1 1
325 350 375 400 4Z5 450 475 50O 525 550 575
2BO pi OF WATER CHLORINATED IN LABORATORY 3.4 mt/tiltr CHLORINE 15 min REACTION TIME, CONCENTRATION 1470 X
MISSISSIPPI RIVtR WATER
O f
Figure 9. Chromatogram showing the Cl-tagged chlorine-containing
constituents in a sample of chlorinated Mississippi River water collected
from the Allen Steam cooling water inlet and chlorinated in the laboratory.
-------
138
(Jolley 1974). The chromatogram shown in Fig. 9 was obtained by HPLC
analysis of 0.28 ml of the radioactive concentrate.
Comparison of this chromatogram (Fig. 9) with the chromatograms
obtained for sewage effluent (Fig. 7) and the Watts Bar Lake cooling
water (Fig. 8) reveals a considerable similarity in profile. The
chromatographic peak elution positions show much correspondence; however,
some significant differences exist. For example, the peak eluting at
29 hr is much larger in the chromatogram of the Mississippi River water
sample (Fig. 9) than in either of the other chromatograms. As in the
chlorination of sewage effluents and the sample of Watts Bar Lake water,
•j/:
about 99% of the Cl activity was associated with the chloride peak.
The remaining 1,3% was associated with the other chromatographic peaks
and the strongly sorbed material on the resin. By analogy with the
O£
previous studies, this activity represents stable Cl-tagged chloro-
organic products.
The following major conclusions were deduced from the analytical
data in this study:
1. Stable chloro-organic compounds are formed during the chlori-
nation of cooling waters at parts-per-million chlorine concentrations.
2. The chlorination yield of chloro-organic products (as Cl) is
about 3% of the chlorine dosage under reaction conditions simulating
those used for antifoulant treatment of the cooling system of an electric
power-generating plant.
3. The HPLC chromatographic profile and the peak elution positions
were similar to those obtained for chlorinated sewage effluents and Watts
Bar Lake water.
DISCUSSION
Similarities in the HPLC chromatographic profiles of the chlorinated
sewage effluents and cooling waters suggest that many of the same chloro-
organic products are formed in each medium during chlorination. A possible
explanation is that natural waters must contain many of the progenitor
organic compounds that occur in sewage effluents. This could be reasonably
expected since the natural waters used for cooling systems contain a vast
complex of microflora and microfauna, animal excreta, plant and animal
-------
139
metabolites, and decomposing plant and animal matter. In many ways,
the composition of such waters resembles that of dilute sewage effluents.
Based on a comparison of chromatographic profiles and peak elution
positions, the peaks for the samples of Watts Bar Lake and Mississippi
River cooling water corresponding to the chloro-organic constituents
identified in the sewage effluents were quantified. The quantitative
data for the chlorinated secondary sewage effluent and cooling water
samples are presented in Table 4. Two compounds, 5-chlorouracil and 4-
chlororesorcinol, were chosen for initial assessment of toxicity and
possible environmental effects. The chloropyrimidine was selected because
of possible incorporation into genetic material; the chlorophenol was
selected because of probable toxicity. Results of these studies have
been previously reported (Gehrs et al. 1974, Gehrs and Jolley 1975) and
will be discussed later in this conference. It is interesting that the
estimated concentrations of chlorinated purines and pyrimidines are
comparable for chlorinated secondary sewage effluent and chlorinated
cooling waters. The concentrations of chlorinated phenols and aromatic
acids are highest for the sample of Mississippi River water. They may
be the result of a higher concentration of humic material in that water.
Because of the large number of unidentified chloro-organic consti-
tuents and the variety of those that have been tentatively identified,
it appears that a number of complex chlorination reactions take place
during the disinfection of sewage effluents and the antifoulant treatment
of cooling waters. The yields of chloro-organic products are determined
by the concentrations of the available chlorinating agent and organic
reactant, the thermodynamics and kinetics of the individual reactions,
and reaction parameters such as time, temperature, light, and catalysts.
36
Most of the chloro-organic compounds separated by the coupled Cl
tracer—HPLC analytical technique used in these studies are considered
to be nonvolatile or, at the least, to have a relatively low volatility.
Their molecular weights are probably less than several thousand. Volatile
chloro-organics, such as chloroform, which might be anticipated as
reaction products from the chlorination of natural waters and which
have been identified by Glaze and Henderson (1975) in chlorinated sewage
effluents, would not have been detected or measured by our analytical
-------
140
Table k. Chloro-organic constituents in chlorinated secondary
sewage effluent and cooling waters
Identifications
Concentration (ug/liter)
Secondary
sewage ,
effluent
Watts Bar
Lake
sample
Mississippi
River
sample
Nucleoside
5-Chlorouridine
Purine
8-Chlorocaffeine
6- Chloro- 2 -aminopurine
8-Chloroxanthine
Pyrimidine
5-Chlorourac il
Aromatic acid
2-Chlorobenzoic acid
3-Chlorobenzoic acid
4-Chlorobenzoic acid
3-Chloro-4-hydroxybenzoic acid
If-Chloromandelic acid
U-Chlorophenylacetic acid
5-Chlorosalicylic acid
1.7
1.7
0.9
1.5
0.3
0.6
1.1
1.3
1.1
O.U
0.2
0.6
1.1
1.0
3
0.6
1.1
0.2
0.3
0.8
1.8
3
3
6
3
10
8
8
3
6
20
18
Phenol
U-Chloro- 3-methylphenol
2-Chlorophenol
3-Chlorophenol
U- Chlorophenol
U-Chlororesorcinol
1.5
1.7
0.5
0.7
1.2
0.2
0.2
0.2
0.2
0.5
0.7
1*
6
2
7
Calculations based on assumption of complete isotopic exchange of the Cl-
tracer in the chlorinating agent with the nonradioactive chloride in the water
samples.
Volley 1973, 1975.
-------
141
technique. Our analytical methodology would also have been inadequate
for detecting or quantifying the chlorination effect on large polymers
such as nucleic acids and humic acids. As previously discussed, chlorine
substitution in the organic bases uracil and cytosine might be anticipated
in nucleic acid polymers. In addition, humic substances might be expected
to chlorinate at active sites on the aromatic rings.
Several general conclusions may be drawn concerning reaction yields
of chloro-organic products from the disinfection of sewage effluents and
the antifoulant treatment of cooling waters. Table 5 presents selected
analytical data for the water samples used in these studies and gives the
chlorination results in terms of chlorine dosage, chlorine residuals, and
reaction yields. The chlorination yield (i.e., that fraction of the
chlorine dosage associated with the chloro-organic products after
termination of the reaction) is relatively much lower for the sewage
effluent, which has an ammonia (as N) concentration of 11 mg/liter, than
for the Mississippi River cooling water, which contains 0.15 mg/liter.
This supports an earlier observation that, when all other factors are
equal, smaller quantities of chloro-organics would be anticipated in
waters containing higher concentrations of ammonia. Two factors - the
extremely rapid formation of monochloramine and the resulting low
equilibrium concentration of HOC1 - are involved. The significant effect
of ammonia on equilibrium concentrations of the reactive chlorine-
containing species is shown in Table 6. If HOC1 is the effective chlori-
nating agent for organic compounds in aqueous solutions, as we suspect,
then a relatively greater reaction yield should be obtained for Watts
Bar Lake and Mississippi River water samples than for the secondary sewage
effluent. The data obtained for Mississippi River water samples in our
study bear out this expected relationship. On the other hand, the yields
we obtained for the secondary sewage effluent were higher than those for
the Watts Bar Lake water; however, we believe that this apparent discrepancy
is due to the much higher concentrations of soluble organic constituents
in the sewage effluent (e.g., see the organic carbon values, Table 5).
The progressive decrease of NH^Cl and concomitant increase of NHC1. as
the ammonia content of each water sample decreased may be significant.
Unfortunately, little is known concerning the aqueous chemistry of
-------
142
Table 5- Comparison of selected analytical data and chlorination
results for secondary sewage effluent and cooling water samples
PH
Chloride
Organic carbon
Organic nitrogen
Ammonia (as N)
Chlorine dosage
Chlorine residual (OT)
Chlorination yield of Cl as
chloro-organics
Secondary Watts Bar
sewage , Lake
effluent sample
7.U 7.5
22 0.5
12 2.7
5.8 2.7
11 0.5
3.2 2.5
1 1
1.0^/^5 min 0.8^/75 min
(0.5^/15 min)a
Mississippi
River
sample
7.3
8
9
<0.05
0.15
3-^
1.2
3.1^/15 min
Concentrations are given in ing/liter.
bJolley 1973, 1975.
cJolley, Gehrs, and Pitt 1975.
estimated value after a chlorination contact time of 15 min.
-------
143
Table 6. Equilibrium concentrations of selected constituents
and reaction yields of chloro-organic products for chlorinated
secondary sewage effluent and cooling water samples
Chlorine residual (OT)
NH (as N)
NOC1C
OC1"°
NH2C1C
NHC12C
Reaction yield after 15 min
Secondary
sewage ,
effluent
1.0
11
<0.0001
-------
144
dichloramine. If NHC19 is a chlorinating agent for organics in aqueous
solutions, then the relatively high equilibrium concentration of dichlora-
mine in the Mississippi River water sample may have contributed to the
high reaction yield of chloro-organics. This aspect of the chemistry of
NHC1 should be studied.
ENVIRONMENTAL SIGNIFICANCE
If we assume that 100,000 to 200,000 tons of chlorine are used
annually in the United States for disinfection and antifoulant purposes,
and that the reaction yields determined in our studies are representative,
we can reasonably estimate that several thousand tons of chloro-organics
are produced each year and released to aquatic ecosystems. Although
these compounds may be present individually at only parts-per-billion
concentrations, their environmental effect on a collective basis may be
very significant. A large gap exists in our knowledge concerning the
nature and concentrations of the compounds produced by these water
treatment processes. Acute toxicity, chronic or low-level toxicity,
and mutagenicity studies are currently being conducted in several
laboratories. The results obtained in some of these studies will be
presented in this conference.
Another area of concern is the creation of noise in or interference
with chemical communications and pheromone systems in aquatic ecosystems
exposed to chlorination. It is possible that this may represent a major
biological effect accompanying the chlorination of cooling waters. For
example, anadromous fish apparently use chemical sensing to find breeding
areas (Sutterlin and Gray 1973, Sutterlin 1974). If this homing instinct
is dependent on sensing organic constituents, the chlorination of cooling
waters could cause confusion because of the large variety of possible
chlorination reactions. Biological changes such as species shifts have
been documented below chlorinated sewage outfalls (Tsai 1968, 1970).
Although these effects are usually attributed to the toxicity of the
chlorine residual and to the decreased oxygen concentration associated
with an increased concentration of organic matter, a contributing cause
might be interference with pheromone systems of other chemical communi-
-------
145
cation systems. Phthalates have been found to be either crowding factor
pheromones or mimics of natural crowding factors (Pfuderer, Williams, and
Francis 1974; Pfuderer and Francis 1975; Pfuderer, Janzen and Rainey 1975).
Phthalates are also subject to chlorine substitution reactions in aqueous
solutions (Goodrich 1949).
One of the goals of this conference is to define environmental
problems associated with water chlorination. To help establish the
significance of the problem, we recommend that the efforts to identify
and quantify the chloro-organics (and possibly the bromo-organics) that
are formed during water chlorination be intensified, that the ecological
consequences of possible pheromone or communications effects be inves-
tigated, and that analytical methodology for detection, isolation, and
identification of organics down to the parts-per-quadrillion level be
developed to facilitate these efforts. Obviously, if the environmental
aspects of water chlorination are found to be unfavorable, we must either
live with the most suitable compromise or develop alternative biocides and/
or disinfection techniques.
CONCLUSIONS
The major conclusions of the experimental studies summarized in
this paper may be stated as follows:
1. The chlorination reaction yields of chloro-organic products
(as Cl) in chlorinated cooling waters and sewage effluents range from
0.5 to 3.1%.
2. Annually, the environmental impact of water chlorination on
the aquatic ecosystems of the United States is estimated to include
the introduction of several thousand tons of chloro-organic compounds.
3. Any or all of a large number of possible aqueous chlorination
reactions may occur during water chlorination, depending on the presence
of organic constituents, reaction kinetics and thermodynamics, and
other reaction parameters.
4. Complex mixtures of chloro-organic compounds are produced during
chlorination, each at parts-per-billion concentration or less.
5. The nature of the chloro-organic products formed during the
chlorination of sewage effluents and cooling waters suggests a variety
-------
146
of possible biological effects relative to (a) genetics; (b) toxicity;
and (c) population, through altered chemical communications in aquatic
ecosystems.
ACKNOWLEDGMENTS
The authors wish to thank C. D. Scott and S. Katz for their encour-
agement and J. E. Attrill, C. W. Rancher, S. Katz, and M. G. Stewart for
critical analysis of this paper.
The original research reported here was sponsored by the Energy
Research and Development Administration, U. S. Environmental Protection
Agency, and National Science Foundation-RANN. The work was carried out
at Oak Ridge National Laboratory, which is operated for the Energy
Research and Development Administration under contract with the Union
Carbide Corporation
REFERENCES
American Public Health Association. 1971. Standard methods for the
examination of water and wastewater, 13th ed. Washington, D.C.
874 p.
Bellar, T.A., J. J. Lichtenberg, and R. C. Kroner. 1974. The occurrence
of organohalides in chlorinated drinking water. J. Am. Water Works
Assoc. 66: 703-706.
Burttschell, R. H., A. A. Rosen, F. M. Middleton, and M. B. Ettinger.
1959. Chlorine derivatives of phenol causing taste and odor.
J. Am. Water Works Assoc. 51: 205-213.
Carlson, R. M., R. E. Carlson, H. L. Kopperman, and R. Caple. 1975.
Facile incorporation of chlorine into aromatic systems during
aqueous chlorination processes. Environ. Sci. Technol. 9: 674-675.
Chandelon, T. 1883. Ueber die durch Einwirkung alkalischer Hypochlorite
auf Phenol gebildeten Chlorphenole. Chem. Ber. 16: 1749.
Christman, R. F., and R. A. Minear. 1971. Organics in lakes, p. 119-
143. In S. J. Faust and J. V. Hunter (ed.) Organic compounds in
aquatic environments. Marcel Dekker, Inc., New York.
Colton, E., and M. M. Jones. 1955. Monochloramine. J. Chem. Educ. 32:
485-487.
Drago, R. S. 1957. Chloramine. J. Chem. Educ. 34: 541-545.
-------
147
Draley, J. E. 1972. The treatment of cooling water with chlorine.
ANL/ES-12. Argonne National Laboratory, Argonne, 111. 11 p.
Dugan, P. R. 1972. Biochemical ecology of water pollution. Plenum
Press, New York. 159 p.
EUROCOP. 1973. List of substances which have been identified in various
waters. EUROCOP-COST Project 64B "Micro-pollutants." Appendix to
first annual report from International Co-ordinating Laboratory,
Hertfordshire, England. September.
Faust, S. D., and J. V. Hunter. 1971. Organic compounds in aquatic
environments. Marcel Dekker, Inc., New York. 638 p.
Gehrs, G. W., L. D. Eyman, R. L. Jolley, and J. E. Thompson. 1974.
Effects of stable chlorine-containing organics on aquatic
environments. Nature 249: 675-676.
Gehrs, G. W., and R. L. Jolley. 1975.. Chlorine-containing stable organics:
new compounds of environmental concern. Verh. Internat. Verein.
Limnol. 19: 2185-2188.
Glaze, W. H., and J. E. Henderson. 1975. Formation of organochlorine
compounds from the chlorination of a municipal secondary effluent.
J. Water Pollut. Control Fed. 47: 2511-2515.
Glaze, W. H., J. E. Henderson, J. E. Bell, and V. A. Wheeler. 1973.
Analysis of organic materials in wastewater effluents after
chlorination. J. Chromatogr. Sci. 11: 580-584.
Goodrich, B. F. Co. 1949. Nuclear chlorination of aromatic carboxy
acids. Brit. Pat. 628,401. Aug. 29. (Chem. Abstr. 44: 2561 g).
Hoist, G. 1954. The chemistry of bleaching and oxidizing agents.
Chem. Rev. 54: 169-194.
Hood, D. V. 1970. Organic matter in natural waters. Institute of
Marine Science, University of Alaska. 625 p.
Hopkins, C. Y., and M. J. v,uisholm \946. Chlorination by aqueous
hypochlorite. Can. J. Res., Sect. B 24: 208-210.
Houben, J., and T. Weyl. 1962. Halogen-verbindungen, p. 760-811.
In J. Houben and T. Weyl, Methoden der Organischen Chemie (Houben-
Weyl) Vierte Auflage, Band 5, Teil 3. George Thieme Verlag,
Stuttgart.
Howard, G. 1960. Purines and related ring systems, p. 1635-1759.
In E. H. Rodd (ed.) Chemistry of carbon compounds, Vol. IV, Part
C. Heterocyclic compounds. Elsevier Publishing Co., New York.
-------
148
Ingols, R. S., and G. M. Jacobs. 1957. BOD reduction by chlorination
of phenol and amino acids. Sewage Ind. Wastes 29: 258—262.
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants. Ph.D.
Diss., University of Tennessee, Knoxville. (ORNL/TM-4290. Oak
Ridge National Laboratory, Oak Ridge, Tennessee). 340 p.
Jolley, R. L. 1974. Determination of chlorine-containing organics in
chlorinated sewage effluents by coupled ^6C1 tracer—high resolution
chromatography. Environ. Lett. 7: 321-340.
Jolley, R. L. 1975. Chlorine-containing organic constituents in sewage
effluents. J. Water Pollut. Control Fed. 47: 601-618.
Jolley, R. L., C. W. Gehrs, and W. W. Pitt. 1975. Chlorination of
cooling water: a source of chlorine-containing organic compounds
with possible environmental significance. Proc. Fourth National
Symposium on Radloecology, Corvallis, Oregon, May 12-14. In press.
Jolley, R. L., S. Katz, J. E. Mrochek, W. W. Pitt, Jr., and W. T. Rainey.
1975. Analyzing organics in dilute aqueous solutions. Chem.
Technol. 1975: 312-318.
Jolley, R. L., W. W. Pitt, Jr., C. D. Scott, G. Jones, Jr., and
J. E. Thompson. 1975. Analysis of soluble organic constituents in
natural and process waters by high-pressure liquid chromatography.
Proc. 9th Annual Conference on Trace Substances in Environmental
Health, Columbia, Missouri, June 10-12. In press.
Jolley, R. L., W. W. Pitt, and J. E. Thompson. 1975. Synthesis of
36ci-tagged 5-chlorouracil, p. 162-165. In C. D. Scott (ed.)
Experimental Engineering Section semiannual progress report,
March 1, 1974 to August 31, 1974. ORNL/TM-4777. Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
Junk, G. A., and S. E. Stanley. 1975. Organics in drinking water,
Part 1. Listing of identified chemicals. IS-3671. Ames Laboratory,
Ames, Iowa. July 84 p.
Kovacic, P., M. Lowery, and K. W. Field. 1970. Chemistry of N-bromamines
and N-chloramines. Chem. Rev. 70: 639-665.
Lee, G. F., and J. C. Morris. 1962. Kinetics of chlorination of phenol-
chlorphenolic tastes and odors. Int. J. Air Water Pollut. 6: 419-431.
Little, Arthur D. Inc. 1970. Organic pollution of freshwater. Water
quality criteria date book, vol. 1. U.S. Environmental Protection
Agency. Washington, D.C.
Manka, J., M. Rebhun, A. Mandelbaum, and A. Bortinger. 1974. Characteri-
zation of organics in secondary effluents. Environ. Sci. Technol. 8:
1017-1020.
-------
149
Morris, J. C. 1967. Kinetics of reactions between aqueous chlorine and
nitrogen compounds, p. 22-53. In S. D. Faust and J. V. Hunter (ed.)
Principles and applications of water chemistry. John Wiley and Sons
Inc., New York.
Morris, J. C. 1971. Chlorination and disinfection—state of the art.
J. Am. Water Works Assoc. 63: 769-774.
Morris, J. C. 1975. Formation of halogenated organics by chlorination
of water supplies. EPA-600/1-75-002. U.S. Environmental Protection
Agency, Washington, B.C. 54 p.
Murphy, K. L., R. Zaloum, and D. Fulford. 1975. Effect of chlorination
practice on soluble organics. Water Res. 9: 389-396.
Patton, W., V. Bacon, A. M. Duffield, B. Halpern, Y. Hoyano, W. Pereira,
and J. Lederberg. 1972. Chlorination studies. I. The reaction of
aqueous hypochlorous acid and cytosine. Biochem. Biophys. Res.
Commun. 48: 880-884.
Pfuderer, P., and A. A. Francis. 1975. Phthalate esters: heartrate
depressors in the goldfish. Bull. Environ. Contain. Toxicol. 13:
275-278.
Pfuderer, P., S. Janzen, and W. T. Rainey, Jr. 1975. The identification
of phthalic acid esters in the tissues of cyprinodont fish and their
activity as heartrate depressors. Environ. Res. 9: 215—223.
Pfuderer, P., P. Williams, and A. A. Francis. 1974. Partial purification
of the crowding factor from Carassius auratus and Cyprinus carpio,
J. Exp. Zool. 187: 375-382.
Pitt, W. W., R. L. Jolley, and S. Katz. 1974. Automated analysis of
individual refractory organics in polluted water. EPA-660/2-74-076.
U.S. Environmental Protection Agency, Washington, B.C. 98 p.
Pitt, W. W., R. L. Jolley, and C. B. Scott. 1975. Betermination of trace
organics in municipal sewage effluents and natural waters by high-
resolution ion-exchange chromatography. Environ. Sci. Technol. 9:
1068-1073.
Prat, R., C. Nofre, and A. Cier. 1965. Effet de 1'hypochlorite de sodium
les constituants pyrimidiques des bacteries. Comp. Rend. Acad. Sci.
Paris 260(May): 4859-4861.
Ramage, G. R., and J. K. Landquist. 1959. The pyrimidine group, p. 1257-
1298. In E. H. Rodd (ed.) Chemistry of carbon compounds, Vol. IV,
Part C. Heterocyclic compounds. Elsevier Publishing Co., New York.
Rebhun, M., and J. Manka. 1971. Classification of organics in secondary
effluents. Environ. Sci. Technol. 5: 606-609.
-------
150
Risebrough, R. W., P. Rieche, D. B. Peakall, S. G. Herman, and M. N. Kirven.
1968. Polychlorinated biphenyls In the global ecosystem. Nature
220: 1098-1102.
Rook, J. J. 1974. Formation of haloforms during chlorination of natural
waters. Water Treat. Exam. 23(2): 234-243.
Schnitzer, M., and S. U. Khan. 1972. Humic substances in the environment.
Marcel Dekker, New York. 327 p.
Steelink, C. 1963. What is humic acid? J. Chem. Educ. 40: 379-384.
Stevenson, F. J., and K. M. Goh. 1972. Infrared spectra of humic and
fulvic acids and their methylated derivatives: evidence for
nonspecificity of analytical methods for oxygen-containing functional
groups. Soil Sci. 113: 334-345.
Sutterlin, A. M. 1974. Pollutants and the chemical senses of aquatic
animals—perspective and review. Chem. Senses Flavor 1: 167-178.
Sutterlin, A. M., and R. Gray. 1973. Chemical basis for homing of Atlantic
salmon (Salmo salar) to a hatchery. J. Fish. Res. Bd. Can. 30: 985-998.
Symons, J. M., T. A. Bellar, J. K. Carswell, J. Demarco, K. L. Kropp,
G. C. Robeck, D. R. Seeger, C. J. Slocura, B. L. Smith, and A. A.
Stevens. 1975. National organics reconnaissance survey for
halogenated organics in drinking water. J. Am. Water Works Assoc.
67: 634-647.
Theilacker, W., and E. Wegner. 1964. Organic syntheses using chloramine,
p. 303-317. In W. Foerst (ed.) Newer methods of preparative organic
chemistry. (Transl. by H. Birnbaum). Academic Press, Inc., New York.
Tsai, C. F. 1968. Effects of chlorinated sewage effluents on fishes in
Upper Patuxent River, Maryland. Chesapeake Sci. 9(2): 83-93.
Tsai, C. F. 1970. Changes in fish population and migrations in relation
to increased sewage pollution in Little Patuxent River, Maryland.
Chesapeake Sci. 11(1): 34-41.
U.S. Environmental Protection Agency. 1975. Suspect carcinogens in water
supplies. Office of Research and Development. Interim report with
appendices. April 1975.
Vallentyne, J. R. 1957. The molecular nature of organic matter in lakes
and oceans with lesser reference to sewage and terrestrial soils.
J. Fish. Res. Bd. Can. 14: 33-82.
Van Middelem, C. H. 1966. Fate and persistence of organic pesticides
in the environment,-p. 228-249. In Organic pesticides in the
environment, Advances in Chemistry Series No. 60. American Chemical
Society, Washington, B.C.
-------
151
Veith, G. D., and G. F. Lee. 1970. A review of chlorinated biphenyl
contamination in natural waters. Water Res. 4: 265-269.
Weber, W. J., Jr. 1972. Physiochemical processes for water quality
control. Wiley Interscience, New York. 640 p.
Weil, I., and J. C. Morris. 1949. Kinetic studies on the chloramines.
I. The rates of formation of monochloramine, N-chloromethylamine,
and N-chlorodimethylamine. J. Am. Chem. Soc. 71: 1664-1674.
Wershaw, R. L., and M. C. Goldberg. 1972. Interaction of organic
pesticides with natural organic polyelectrolytes , p. 149-158.
In S. D. Faust (symp. chairman) Fate of organic pesticides in the
environment. Advances in Chemistry Series, Number 111. American
Chemical Society, Washington, B.C.
White, G. C. 1972. Handbook of chlorination. Van Nostrand Reinhold
Company, New York. 744 p.
WHO International Reference Centre for Community Water Supply. 1974.
Current knowledge of concentrations of organic contaminants in
waste water, river water, and drinking water. Working Document
No. 4 . Working Meeting on Health Effects Relating to the R'—Use
of Waste Water for Human Consumption. Amsterdam. January j.j-16,
1975.
Zaloum, R., and K. L. Murphy. 1974. Reduction of oxygen demand of
treated wastewater by chlorination. J. Water Pollut. Control Fed.
46: 2770-2777.
DISCUSSION
Alan A. Stevens, U.S. Environmental Protection Agency. I would like to
point out that Bob's estimate of chlorination yields are probably low
because the HPLC technique does not detect the volatile compounds, for
example, the trihalomethanes. In our work, roughly 1-3% of the chlorine
used, when Ohio River water is chlorinated.to meet the 96-hr demand,
becomes trihalomethanes.
Jolley. I think that is very significant. Although I did not show
any material balances for our chlorination experiments, the number 3% is
about the yield of volatile chlorinated compounds that would have been
estimated from the Watts Bar water experiment. Essentially, we had 97%
material balance and so, indeed, maybe this did represent the formation
of chloroform and other haloform compounds.
Robert S. Ingols, Georgia Institute of Technology. May I ask whether
you believe that you have any proteins or any organized organic matter,
not the monomers as it were, involved in any of your analytical procedures?
Have you hydrolyzed or modified any of the large molecules such as proteins
in which tyrosine or tryptophan are present? I assume you meant this was
present as free tyrosine.
-------
152
Jolley. That is correct.
Robert S. Ingols. And the tryptophan would also be readily chlorinated,
if present as the free amino acid. I have good reason to believe, on the
basis of some of the work done in the early fifties, that chlorine goes
on to the tyrosine moiety within the protein molecule because it will
not develop color in the presence of chlorine dioxide.
Jolley. Thank you, Dr. Ingols. No. We have not looked at proteinaceous
materials. We have separated some rather large molecules by HPLC but we
are dealing essentially with the compounds that are less than 1000 in
molecular weight. We have not analyzed many free amino acids in these
waters. I included in my tables only those confirmed by mass spectral data.
J. J. Nelson, Energy Research and Development Administration. I
would like to first comment for the record that, if our instrumentation
for detection of chloro-organics approached the order of magnitude of
sensitivity available in radioactivity detection equipment, then our
questions on impact would be much more well defined. My question is,
what was the basis for your suggestion that organo-chlorides may have an
impact on the pheromone communication among animals?
Jolley. I am speculating. However, phthalic acid esters are known to
be crowding factor pheromones, or at least to mimic them, in several aquatic
species and they are known to be susceptible to chlorination. This leads
me to think that such organic compounds and others which may be pheromones
may be chlorinated during cooling water treatment. Thus, one effect of
chlorinating cooling waters may be that of disrupting the natural chemical
communications system.
J. Carrell Morris, Harvard University. It would be interesting to
note while we're on the question of tonnage production here, that the use
of chlorine in the bleaching of pulp and paper is about 10 times the use
in water and sewage treatment. And probably, there is, resulting from
this, a production of the order of 10 tons of halogenated methanes per
year. This is of the order of magnitude, but possibly only about a third,
of the total global production of chloroform from all sources. So it is
still moot as to whether marine bacteria or man is producing the most
chloroform.
-------
ANALYSIS OF NEW CHLORINATED ORGANIC COMPOUNDS FORMED
BY CHLORINATION OF MUNICIPAL WASTEWATER
William H. Glaze, James E. Henderson, IV, and Garmon Smith
Institute of Applied Sciences and Department of Chemistry
North Texas State University
Denton, Texas 76203
ABSTRACT
The effect of chlorination on secondary municipal wastewater
effluents has been investigated using two analytical techniques. Total
organic-bound chlorine (TOC1) is measured before and after chlorination
by a microcoulometric procedure. Concentrated extracts of the effluent
before and after chlorination at various chlorine dose levels are
pyrolyzed and titrated in the Dohrmann halide analyzer. The TOC1 results
show a significant increase in the level of organic-bound chlorine after
chlorination, particularly using large doses of chlorine (2000 to 4000
ppm). More explicit information regarding the nature of the new
organic chlorine-containing compounds is obtained by gas chromatography/
mass spectrometry studies of the concentrates obtained by XAD-2 resin
extractions of the effluents. The GC/MS results confirm that chlorina-
tion causes the formation of many new chlorinated organics, the struc-
tures of over 50 of which have been identified. Whereas the majority of
the compounds are aromatic halides, many are not derivatives of "activated"
aromatics such as phenol but are simple derivatives such as chlorobenzenes,
-toluenes, and -alkylbenzenes. Nonaromatic chlorides have also been
identified. Particular attention has been focused in this work on the
effect of heavy doses of chlorine in the range of 2000 to 4000 ppm. The
TOC1 and GC/MS results indicate that treatment of this type causes a
very large increase in organic-chlorine content. Also reported are GC/MS
data on XAD extracts of Denton, Texas, drinking water. Of particular
interest are the occurrence of three iodine containing compounds, viz.
dichloroiodomethane, dibromoiodomethane, and bromochloriodomethane in the
finished water.
153
-------
154
INTRODUCTION
The use of aqueous chlorine as a water and wastewater treatment
agent has come under scrutiny lately, primarily due to the discovery
that chlorination of residual organics in such waters causes the forma-
tion of new chlorinated organics of unknown toxicities (Glaze et al.
1973, Jolley 1973, Bellar and Lichtenberg 1974, Rook 1974). Of course,
the notion that organic compounds of certain types will react with
aqueous chlorine is not new; workers as early as 1883 recognized this
fact. Morris (1975), one of the contemporary pioneers in this area, has
recently reviewed the chemistry of aqueous chlorine including typical
reactions with organic functionalities. The more recent attention of
environmental chemists to the phenomena of chlorination, therefore, may
be veiwed as a result of increased federal participation in the water
and wastewater treatment arena, as well as the wider availability of
sensitive instrumentation with which to study materials and processes
at the trace level.
Under the auspices of an Environmental Protection Agency grant,
the Trace Analysis Laboratory at North Texas State University has been
involved for the past two years in a study of the effects of chlorination
of municipal wastewaters. Our early work under the sponsorship of the
NTSU Faculty Research Fund was published in 1973.
In this paper we describe more recent data on the formation of new
chlorinated organics, particularly with the use of large doses of
chlorine (1000 to 4000 mg/1). We also report on the development of a
method for monitoring total organic-bound halogen (TOC1) in chlorinated
wastewaters, and show the application of this method to the analysis of
TOC1 in superchlorinated municipal wastewaters.
EXPERIMENTAL METHODS
Total organic-bound chlorine determinations. Figure 1 is a scheme
for the concentration of organics from water samples. The scheme consists
We shall refer to the TOC1 method as if only chlorinated organics
were being measured; in fact, the method measures organic bromides and
iodides as well, and should be termed a TOX (total organic-bound halogen)
method.
-------
155
SAMPLE COLLECTION
Gravity Filtration
Division into Two Aliquots
Chlorination
Quench
Quench
XAD Extraction
XAD Extraction
Elution
Et20 Elution
Concentration
Concentration
ANALYSIS
Chromatographic: FID, CECD, GC/MS
Total Halogen: Pyrolysis/Microcoulometry
Fig. 1. Scheme for the extraction and analysis of chlorinated
organic compounds from water and wastewater samples.
-------
156
of a concentration step involving Rohm and Haas XAD macroreticular
resins, and subsequent analysis of the organic eluate from the resin.
A two liter wastewater sample is obtained from the treatment plant
before chlorination, usually after final clarification. The sample is
filtered through coarse paper and divided into two equal portions. One
portion serves as the system control (BEFORE); the other portion (AFTER)
is chlorinated to whatever level desired. Chlorine contact time is
usually one hour. Both BEFORE and AFTER samples are treated with sodium
sulfite to quench active chlorine, and then allowed to flow through an
XAD-2 resin bed at approximately 30 ml/min. The scrupulously cleaned
resin (ca. 2 grams) is contained in an 8-cm x 1-cm-ID glass column
fitted with a 24/40 I outer joint at top and a Teflon stopcock at the
bottom. The water sample is introduced into the column from a one liter
separatory funnel, after which the separatory funnel is rinsed with a
few milliliters of purified diethylether. These washings and more
ether, totaling approximately 20 to 30 ml are used to elute the organic
materials trapped on the XAD-2 column. The ether eluate is collected
in a flask of the design recommended by Junk et al. (1974) and concen-
trated to final volume (1 to 5 ml) using a three-ball Snyder column.
Microcoulometric analysis of the organic halides (excluding fluo-
rides) in the eluate is accomplished using a Dohrmann-Envirotech C-300
microcoulometer with S-300 pyrolysis furnace. Aliquots (1 to 25 ul)
of the ether eluate are injected directly into the furnace through a
septum or using a platinum or quartz boat sample injection system. In
each case, the sample is entrained in a stream of argon into an inlet
furnace operated at 800°C (lower temperatures are used in some cases),
where the gas stream is mixed with oxygen. Oxidative pyrolysis is
completed as the sample components flow through the middle and outlet
areas of the furnace which are at 800°C. The pyrolyzed sample plus
carrier gas then flows into the coulometric cell where halides other
The resin is cleaned and regenerated using the methods described
by Junk et al. (1974).
Ether (Analytica
column and the center cut taken.
fEther (Analytical Reagent Grade) is distilled in a 1.5 m Oldershaw
-------
157
than fluoride are titrated. Readout of the system is digital and is
given in terms of nanograms of chlorine. A readout of the signal vs
time is also displayed on a strip-chart recorder. Minimum detectable
limit of the system is quoted as 2 ng chlorine (S/N = 10).
Injections of samples to be titrated may also be made directly into
the coulometer cell, both for calibration of the system and for monitoring
of inorganic halide levels in aqueous samples.
Mass spectrometry/gas chromatography. Wastewater samples are also
treated according to the scheme shown in Fig. 1. After elution from the
resin with ether and subsequent concentration to 1 ml, the organic
extract is surveyed by gas chromatography using both flame ionization and
Coulson electrolytic conductivity detectors. The Coulson detector is
coupled to a Hewlett-Packard Model 5710 chromatograph; the FI detector
is a part of the Finnigan Model 3200 Mass Spectrometer/Model 8500 Gas
Chromatograph System with Model 600 Digital Interactive Control and
Graphic Output System. The chromatograph column is usually 6-ft x
3-mm-ID glass packed with 3% Dexsil 300 GC coated on 100/120 mesh
Supelcoport (Supelco Inc.). The helium flow rate is 30 ral/min. The
temperature program conditions are as follows: (1) isothermal at room
temperature (ca. 27°C) for four minutes after injection; (2) program
ballistically to 50°C; (3) 50° to 300°C at 6°/min; (4) isothermal at
300°C until completion of run.
The injector temperature was 255°C; the detector temperature was
300°C. The Coulson block temperature was 300°C, and the furnace tempera-
ture was 830°C. The detector was operated in the reductive mode with
80 ml/min of hydrogen added to the GC column effluent prior to pyrolysis.
The bridge current was 30 V. The FID detector was run at a range of
10"11 amps/mv and an attenuation of X8 to X32.
Extraction of drinking water samples. The procedure followed that
used for wastewater samples except that no laboratory chlorination was
carried out. Tap water (1 to 50 1) was passed directly into the column
containing XAD resin; workup was as described for wastewater samples.
-------
158
RESULTS
Evaluation of the TOC1 method. The applicability of the TOC1
procedure depends on several factors, viz.: (1) the efficiency of
adsorption of organics on the XAD-2 resins; (2) the recovery of the
organics from the resin by ether elution; (3) the efficiency of the
pyrolysis/coulometric titration procedure; and (4) the magnitude of
interferences in these processes. Junk et al. (1974) have addressed
themselves to items (1) and (2) and have shown that overall recovery of
a large number of organics, including organochlorides, is quite good
(average recovery of 80 to 90% from 10 to 100 ppb aqueous samples). It
should be noted that no data is available in their paper on volatile
organics such as chloroform, nor on substances of high molecular weight
such as polymeric materials. In both of these cases, one may assume
that the recovery efficiencies would be less than those obtained on the
test compounds used by Junk et al. (1974).
With respect to item (3), the efficiency of the pyrolysis/
microcoulometric titration procedure, there is limited data published
although apparatus similar to ours have been utilized by other workers
for TOC1 determinations of liquid/liquid extracts (Greve and Haring
1972).* The manufacturer (Dohrmann, private communication) reportedly
has found good efficiency for a wide variety of organic halides, but it
is clear that the system should be checked with a variety of compounds
before the TOC1 method is widely applied. Further information on the
efficiency of the Dohrmann pyrolyzer/microcoulometer system will be
published by us in a forthcoming paper. However, we may report on the
basis of these results that it would appear that one should limit injec-
tion volumes to 5 to 25 yl and chloride amounts to something less than
1000 ng. Both conclusions are consistent with the recommendations of
the manufacturer. Our data also show that 1 to 2% precision is possible
Kuhn and Sontheimer (1973, 1975) have described TOC1 methods based
on adsorption of organic halides by activated carbon. A sample of the
carbon is then pyrolyzed in a steam/air atmosphere and the resulting
hydrogen halides measured either by microcoulometry or by the halide
specific-ion electrode.
-------
159
(3 injections); however, it should be noted that the injection process
requires care and there is considerable variation from one operation to
another.
Figures 2 and 3 show recorder traces of the coulometric integral for
a chlorobenzene/ether standard and a wastewater extract in ether. It
is of particular interest to note the shapes of these curves, both of
which were obtained using the same platinum boat injection procedure.*
The shape of the chlorobenzene integral is typical of volatile organic
halides which apparently are swept into the furnace as a narrow plug. The
wastewater extract curves in Fig. 3 show a much different shape which
apparently reflects the presence of higher molecular weight organohalides
in wastewater extracts. Thus, the XAD resin extraction procedure does
not appear to be limited to low molecular weight materials, although the
data here do not give any evidence of the overall efficiency of the
extraction process.
Because of the apparent presence of high molecular weight materials
in the XAD extracts, it was important to determine the time required
for complete combustion of the sample using the platinum boat procedure.
Figure 4 is a plot of integrator response (ng Cl) vs time for the
injection of a 1 yl sample of two different "superchlorinated" waste-
water extracts. Inlet furnace temperatures were 800° and 200°C. As
indicated in the figure, the 200°C curve has not reached a limiting
value after 960 seconds (the present integration limit), although it
appears that the signal is leveling off. For the data obtained at 800°C,
however, only 420 seconds are required for the integral to reach a value
of 95% of the 960 second value.
Table 1 shows the results of a limited number of runs using two
inorganic halides. In these runs aqueous solutions of NH4C1 and NaCl were
injected into platinum boats and analyzed as usual. The results in
Table 1 clearly show the interference of inorganic chloride in the
pyrolysis/microcoulometry system. Of particular interest are the results
The boat is pushed into the edge of the pyrolysis furnace until
the recorder signal "peaks," then the boat is pushed into the furnace
to its mechanical limit.
-------
160
Fig. 2. Response of microcoulometric halide analyzer: three
repetitive samples of chlorobenzene in isooctane (10 yl of 100 ng/yl).
Integral values: 965.4, 955.7, and 966.7 ng Cl. Inlet furnace tempera-
ture 200°C.
-------
161
Fig. 3. Response of microcoulometric halide analyzer: three
repetitive samples of Denton, Texas, wastewater extract after chlorina-
tion. 25 ppm (20 yl injections). Integral values: 363.6, 356.5, and
361.5 ng Cl. Inlet furnace temperature 200°C.
-------
162
Fig. 4. Response of microcoulometric halide analyzer. Upper trace;
Denton, Texas, wastewater extract after chlorination (2000 ppm). Lower
trace: system blank.
-------
Table 1. Microcoulometer response: direct aqueous injections into platinum boat
NH(+C1/H20 (200°C
Quantity injected
(nano grams)
1000
100
100
Inlet Furnace)
Quantity detected
(nanograms)
860
104
107
NaCl/H20 (800°C
Quantity injected
(nanograms)
100
100
100
100
100
100
Inlet Furnace)
Quantity detected
(nanograms)
119
92
84
93
87
85
U>
-------
164
with ammonium chloride where quantitative detection is observed even at
an inlet furnace temperature of only 200°C.
On the basis of the results shown in Table 1, it became imperative
to test the recovery of ammonium chloride using the XAD procedure des-
cribed earlier. Table 2 shows the results of such experiments. One
liter samples of ammonium chloride in water (100 and 500 mg/1 concentra-
tions) were processed using the scheme shown in Fig. 1. Also processed
was a distilled water blank. The total quantities of ammonium chloride
passed through the resin were 1 x 108 and 5 x 108 ng respectively for the
100 and 500 ppm solutions. As shown in the figure, the ether extract
contained only traces of chloride which could not be distinguished from
blank levels. Thus, it appears that while inorganic halide would repre-
sent a significant interference if one attempted direct aqueous pyrolysis,
the use of the XAD/TOC1 procedure would seem to eliminate this interference.
Table 2. Retention of NH^Cl by XAD-2 resin
Nano grams NHi^Cl Nano grams recovered in Et£0 eluate
(1 liter H20) (5 ml volume) % Recovery
Distilled H20 4.9 ± 1.0 x 102
1 x 108 (100 ppm) 5.5 ± 0.5 x 102 5.5 x ICT1*
5 x 108 (500 ppm) 1.2 ± 0.05 x 103 2.4 x 10~4
Finally, it should be noted that the Dohrmann pyrolysis/microcoulo-
meter system is sensitive to other elements, particularly sulfur and
nitrogen, which may be in wastewater extracts. However, the relative
sensitivities for halogens, sulfur and nitrogen are reported to be
1 : 0.01 : 10"4 respectively. Thus, it would appear that the XAD/TOCl
method is applicable for the determination of TOC1 values in both waste-
water and potable waters, since interferences of these magnitudes are not
expected to occur in XAD extracts.
Table 3 shows the results of the application of the method for the
analysis of municipal wastewaters after chlorinations . Figure 5 is a
recorder trace for a typical wastewater extract and a system blank. As
indicated in Table 3, system blanks (using distilled water) are quite
-------
165
Table 3. Total organic-bound chlorine:
Denton, Texas, wastewater effluents
TOC1 (ug/D
Pyrolysis
temp (°C)
800
800
800
800
800
200
200
Chlorine
dose
2000
1200
3500
1200
1400
2000
25
Before
chlorination
10.96
22.4
35.4
36.9
23.0
13.0
16.2
± 0.2
± 0.3
± 0.4
± 0.8
± 0.2
± 0.1
± 0.5
After
chlorination
906.3 ±
608.95 ±
163.8 ±
337.7 ±
402 . 8 ±
479.7 ±
18.8 ±
73.6
5.3
1.9
20.2
1.0
6.8
0.1
Blank
4.4 ± 0.2
21.8 ± 0.2
1.7 ± 0.1
variable and reflect the impurity levels of distilled water as well as
the ether. Nonetheless, it is clear from these data that chlorination of
wastewaters, particularly using "superchlorination" conditions, results
in a significant quantity of organic-bound halogen as compared to non-
chlorinated samples. Further work in this area will attempt to extend
these measurements to survey various wastewaters and potable waters, and
to improve the overall efficiency of the XAD/TOC1 procedure.
GC/MS studies on chlorinated waters. More explicit information
concerning the new chlorinated organic compounds in "superchlorinated"
wastewaters has been obtained by combined gas chromatography/mass
spectrometry. This method should be viewed as complementary to the
high-pressure ion exchange chromatography method used by Jolley (1973,
1974, 1975), in that XAD-2 is more efficient at trapping non-polar
compounds. Thus, Jolley characterized several new chlorinated aromatics
including carboxylic acids and phenols. Of particular interest was the
finding of chloroderivatives of uracil, uridine, caffeine, guanine, and
xanthine, which may possess significant physiological activity.
The results of GC/MS investigations of "superchlorinated11 waste-
waters have been published by us recently and will be reviewed here
-------
166
INTEGRATION TIME vs COULOMETER OUTPUT
I i I
j I
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960
TIME (SEC.)
Fig. 5. Response of pyrolysis/microcoulometry system. Denton,
Texas, wastewater extract. Upper curve, Sample I: 200°C inlet furnace
temperature. Lower curve, Sample II: 800°C inlet furnace termperature.
-------
167
(Glaze and Henderson 1975). As shown in Fig. 6, the FID chromatogram
of a typical wastewater XAD extract contains innumerable species, only
part of which are chlorinated. To distinguish chlorinated species, the
Coulson electrolytic conductivity detector was employed, the result of
which is shown in Fig. 7. The very large number of chlorinated species
in the upper spectrum is consistent with the TOC1 results discussed in
the earlier portion of this paper.
Figure 8 is a total ion monitor of the gas chromatograms of another
superchlorinated extract and control. In this case, the time axis is
replaced by a spectrum number axis, indicating that a total of 450 mass
spectra have been taken and stored in computer memory during the
chromatographic run. Analysis of these mass spectra is then accomplished
by manual interpretation and comparison with standard spectra (Registry
of Mass Spectral Data, 1974 and EPA Mass Spectral Search System). Some
of the compounds identified by this procedure are listed in Table 4.
It is clear from an inspection of Table 4 that most of the compounds
identified thus far are aromatic derivatives. However, the compounds are
by no means derivatives of "activated" aromatics in every case. The
chloroderivatives of benzene, toluene, and benzyl alcohol are evidence
that superchlorination may lead to substitution of unactivated aromatic
moieties. Moreover, we are particularly interested in the formation of
several nonaromatic derivatives, such as, chlorocyclohexane, a chloroalkyl
acetate, and, perhaps most significant, three chlorinated acetone deriva-
tives. The latter may be precursors of chloroform which we have shown
in previous work to be formed in wastewater superchlorinations and which
has been shown by other workers to result from the chlorination of
organics in drinking water.
Finally, we note that the concentrations of the compounds listed in
Table 4 are in the yg/1 range (ppb) in agreement with earlier works in
this area.
*
Figures 6 to 8 published by permission of the Journal Water
Pollution Control Federation.
-------
CO
Fig. 6. Gas chromatograms (FID detector, 10"11 x 16 amps/fs) of
Denton, Texas, wastewater extract (11-12-74). Bottom, before chlorina-
tion: top, after 2000 mg/1 chlorine dose for one hour contact period.
Analytical conditions described in text.
-------
Fig. 7. Gas chromatograms (Coulson electrolytic conductivity
detector, halogen mode, x 4) of Denton, Texas, wastewater extract
(11-22-74). Bottom, before chlorination: top, after 2000 mg/1 chlorine
dose for one hour contact period. Analytical conditions described in
text.
:
-------
8 18 W 39 •« Sfl 69 M 99 9fl 199 !|8 ISO 130 IKJ ISfl 160 110 199 198 298 S|8 228 338 5KJ 2S« ?W ZM »9 399 398 Slfl 3» 33S 318 358 388 3TO 399 339 109 119 428 «3B 119 ISO
H
VJ
o
IB 29 38 « SB 88 VB SB 98 IW IID ITO I3U 110 ISO 160 110 118 130 '2M 210 3CC C3P SW ESB W3 219 280 7JP 3BB JIB 3?1 330 3*1 3S3 36B J« am ifl 109 110 «9 139 113 1SB
Fig. 8. Gas chromatograms (mass spectrometer total ion monitor) of
Denton, Texas, wastewater extract (11-22-74). Peaks normalized by com-
puter to make largest peak full scale.
-------
171
Table 4. Compounds identified in superchlorinated
municipal wastewaters
Nonaromatics
Chloroform Chlorocyclohexane
Dibromochloromethane Tetrachloroacetone
Dichlorobutane Pentachloroacetone
3-Chloro-2-methylbut-l-ene Hexachloroacetone
Aromatics
o-Dichlorobenzene Trichloroethylbenzene*
p-Dichlorobenzene Trichlorocumene*
Chloroethylbenzene Dichlorotoluene*
Dichloroethylbenzene* Chlorocumene*
N-Methyl-trichloroaniline* Trichlorophenol*
& A
Trichlorodimethoxybenzene Tetrachlorophenol
Tetrachloroethylstyrene* Tetrachlorodimethoxybenzene*
Trichloromethyls tyrene* Trichlorophthalate*
Chloro-«-methylbenzyl alcohol* Tetrachlorophthalate*
Dichloro-oc-methylbenzyl alcohol*
Reference: Glaze and Henderson 1975.
*
Isomer not specified.
Mass spectra of volatile organohalides in drinking water. Figure 9
is a gas chromatogram of an ether extract of drinking water obtained from
a tap in the NTSU Chemistry Building. In addition to those compounds
found in earlier work, it is interesting to note the presence of three
iodine-containing compounds, dichloroiodomethane, bromochloroiodomethane,
and dibromoiodomethane. The source of these compounds in Denton, Texas,
drinking water is unknown; presumably they are formed by chlorine treat-
ment. Further work in this area is currently in progress.
ACKNOWLEDGMENTS
This research was supported in part by Grant No. R803007 from the
Environmental Protection Agency and in part from a grant from tne NTSU
-------
172
NTSU DRINKING WRTER: 8-13-75
ex
D,
100
50 100 150 200 250 300
Fig. 9. Gas chromatogram of tap water taken at UTSU Chemistry
Building 8-13-75. Finnigan Model 3200 MS/GC System. 3% Dexsil 300 GC
column on 100/120 mesh Supelcoport.
-------
173
Faculty Research Committee. We are grateful to Dr. Craig Shew of the
Robert Kerr Water Research Laboratory in Ada, Oklahoma, for assistance
in the GC/MC work.
REFERENCES
Bellar, T. A., and J. J. Lichtenberg. 1974. The determination of
volatile organic compounds at the ug/1 level in water by gas
chromatography. J. Am. Water Works Assoc. 66: 739-744.
E.P.A. Mass Spectral Search File, Cyphernetic Corp., Ann Arbor,
Michigan.
Glaze, W. H., J. E. Henderson IV, J. E. Bell, and V. A. Wheeler. 1973.
Analysis of organic materials in wastewaters after chlorination.
J. Chromatogr. Sci. 11: 580-584.
Glaze, W. H., and J. E. Henderson IV. 1975. Formation of organochlorine
compounds from the chlorination of a municipal secondary effluent.
J. Water Pollut. Control Fed. 47: 2511-2515.
Greve, P. A., and B. J. A. Haring. 1972. Die Mikrocoulometrische
Bestimmung von organische gebunden Halogen in Oberflachen-und
anderen Gewassern. Schriftenr. Ver. Wasser. Boden Lufthyg. Berlin-
Dahlem 37: 59-64.
Kuhn, W., and H. Sontheimer. 1973. Einige Untersuchungen Zur Bestimmung
von organische Chlorverbindungen auf Aktivkohlen. Vom Wasser
41: 65-79.
Kuhn, W., and H. Sontheimer. 1975. Zur analytischeu Erfassung
organischer Chlorverbindungen mit der temperaturprogrammierten
Pyrohydrolyse. Vom Wasser 43: 327-341.
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants.
ORNL/TM-4290. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Jolley, R. L. 1974. Determination of chlorine-containing organics in
chlorinated sewage by coupled 36C1 tracer — high resolution
chromatography. Environ. Lett. 7: 321-340.
Jolley, R. L. 1975. Chlorine-containing organic constituents in
chlorinated effluents. J. Water Pollut. Control Fed. 47: 601-618.
Junk, G. A., J. J. Richard, M. D. Grieser, D. Witiak, J. L. Witiak,
M. D. Arguello, R. Vick, H. J. Svec, J. S. Fritz, and G. V. Calder.
1974. Use of macroreticular resins in the analysis of water for
trace organic contaminants. J. Chromatogr. 99: 745-762.
-------
174
Morris, J. C. 1975. United States Environmental Protection Agency
Report. EPA-600/1-75-002.
Rook, J. J. 1974. Formation of haloforms during chlorination of natural
waters. Water Treat. Examin. 23(2): 234-243.
Stenhagen, E., F. Abrahamsson, and F. W. McLafferty. 1974. Registry of
mass spectral data. Wiley-Interscience.
DISCUSSION
George Clifford White, Consulting Engineer. This has a very signifi-
cant practical application in wastewater disinfection. Normally chlo-
rinators operate their injectors with effluent. To put it in perspective,
an 8000 Ib/day chlorinator that would dose at the rate of 8 to 10 mg/1
chlorine uses 200 gal/min backwater that would be producing the organics
that you showed there.
Glaze^ You mean the make-up water?
White. The water going through the injector. In other words, for
an 8000 Ib/day chlorinator the ratio of chlorinated effluent injector
water to sewage effluent would be about 200 gal/min effluent containing
3500 ppm chlorine solution to about 80,000,000 gal/day sewage. So you
could work out what the organic pickup was on account of the use of
effluent for injector water.
Robert B. Dean, U.S. Environmental Protection Agency. I am glad
you are going after the amino acids. It is very important to look at
chlorinated proteins after hydrolysis to amino acids. It is easy to
graft chlorine onto a protein and produce a product that is resistant
to bacterial attack. This production is based on the extreme ease of
producing iodinated tyrosine and thyroxine hormone by adding iodine to
dry proteins. We have been looking for compounds only where we have a
convenient analytical technique and have not looked at the rest of the
compounds that are not volatile or elutable from XAD resin.
George Helz, University of Maryland. I have several questions about
the XAD resin. First of all what happens to the chloramine? Second, do
you know if a very large chlorine concentration generates chloro-organics
on the resin?
Glaze. Well, we destroy the chlorine with sodium sulfite before we
use the resin. So the answer to the last question is presumably no. The
more general question of what happens to chloramines on XAD, again they
are presumably destroyed by sulfite. So we don't see them. Whether if
they were not destroyed they would be trapped, eluted, and preserved in
this method, I don't know. We have not looked into that.
-------
175
Max Deinzer, Oregon State University. What do you think is the
mechanism for formation of symetrically substituted chloroacetones in
water supplies?
Glaze. I suppose we are really under very acid conditions. I
don't have a mechanism. Perhaps some other expert in the audience has
a mechanism.
Sidney Katz, Oak Ridge National Laboratory. Since recovery of
unknown compounds from XAD resins is an unknown, how do you quantify
your results?
Glaze. Any kind of concentration, whether it would be carbon or
XAD, suffers from that disadvantage. I don't even want to apologize for
it. That is a fact of life. What one can do, and what Fritz's group
has done, is to look at a wide variety of model compounds and look at
efficiencies. In the Journal of Chromatography in 1974, p. 745, there
is a very very fine paper by that group which lists efficiencies for
many compounds. The efficiencies are all high. They didn't go to high
molecular weight ranges. They didn't look at the volatiles. So there
are some holes in that work, obviously. They just didn't have that
much time, I guess. So you do what you can do and, fortunately for us,
they have done it at Iowa State, and it looks like efficiencies for most
of the compounds that we're identifying are really not bad, not bad at
all. So we have quantified these, by the way, and if you want the numbers
I'll give them to you. But I agree with you that they are probably only
good to within a factor of two or three or so.
David Friedman, Food and Drug Administration. Just a comment on XAD
resins. We have tried the resins with highly chlorinated materials. They
are very good at taking material out of the water, but you can't get
them back off the resin.
Glaze. What kind of material are you taking about?
Friedman. Chlorinated paraffins.
Glaze. Chlorinated paraffin, yes, because they presumably penetrate
the polymer. That's a valid criticism. I think we can not say that we
get. all of the material out. It's one technique, like carbon, which
has some merit.
-------
CHEMISTRY OF HALOGENS IN SEAWATER
James H. Carpenter and Donald L. Macalady
Rosenstiel School of Marine and Atmospheric Science
University of Miami
Miami, Florida 33149
ABSTRACT
There has not been sufficient research to provide a satisfactory
understanding of the reactions that occur when +1 oxidation state chlo-
rine is added to seawater. However, present information suggests that
the bromide ion is oxidized and, perhaps, disproportionates to several
oxidation states. Formation of brominated or mixed brominated-
chlorinated organic compounds can be expected but the extent and specia-
tion of such reactions remain to be determined.
Our experiments show that present analytical procedures do not
measure all of the inorganic "residuals" present in chlorinated
seawater.
Other presentations at this conference review the chemistry of
chlorine in fresh waters and wastewaters. This paper focuses on the
limited information that is available for reactions that occur when
chlorine is added to seawater. The material that is reviewed here is
intended to supplement the fresh water and wastewater data and those
facts will not be repeated here. The principal reason for differences
between the sets of reactions that may occur in seawater in contrast
to fresh waters and wastewaters is, of course, the particular composi-
tion of seawater.
Since this presentation is addressed to individuals of diverse
backgrounds, a brief and superficial description of the nature of seawater
and oceanography may be helpful. The large mass of liquid that fills
the ocean basins has been produced by processes acting over long periods
177
-------
178
of time and it is, therefore, rather uniform. However, the composition
of this natural solution is rather complex. Elemental analysis has
shown 75 of the chemical elements to be present in concentrations ranging
from grams per liter to picograms per liter. The ionic and molecular
speciation is only partly known for the inorganics and the several
hundred identified organic compounds only account for part of the total
organic content. Furthermore, seawater contains suspended and motile
particules ranging from colloidal inorganic minerals through single-
celled organisms to large mammals. In short, seawater studies would
probably not be attractive to thoughtful scientists and the discipline
of oceanography has been developed to provide a haven for the foolhardy.
Much of the current knowledge in chemical oceanography has been
summarized in the treatises edited by Riley and Skirrow (1975) and this
reference leads to the following generalizations with regard to sea-
water as a reaction medium:
1. The ionic strength is 0.7 molal.
2. The abundant (mg/1 or greater) cations are sodium, magnesium,
calcium, potassium and strontium ions and the abundant anions
are chloride, sulfate, bromide, bicarbonate, borate, silicate
and fluoride ions.
3. The pH is close to 8 if the seawater is in equilibrium with
the atmosphere.
4. The alkalinity is 2 milliequivalent.
5. Dissolved oxygen provides an "apparent" redox potential of
approximately 0.5 volts.
6. The total organic carbon content is approximately 1 mg/liter
or less.
7. Transition metals (Fe, Mn, Cu, Ni, Zn, Mo) are present at
microgram/liter concentrations; i.e., at potentially catalytic
concentrations but substantial fractions of the metals are
present as colloidal hydroxides and organic complexes.
8. The sun shines on the oceans periodically.
Turning to the question of what reactions occur when +1 oxidation
state chlorine is added to seawater, the first alternative is the
-------
179
addition of chlorine gas or solutions of salts of hypochlorous acid.
If chlorine gas is added, the familiar hydrolysis reaction would be
expected,
C12 + H20 - HOC1 + H+ + Cl~ ,
with a concurrent adjustment of the primary proton equilibrium in sea
water involving bicarbonate ion,
2HC03 - CO 3 + H2C03
it
H20 + C02
In order to see the characteristics of these reactions, consider
the addition of chlorine to make the solution 3.5 ppm added chlorine.
The presence of 0.53 molar chloride tends to repress the hydrolysis.
The assumed added chlorine to 50 micromolar is computed to cause a
resultant concentration of dissolved molecular chlorine of 6 picomolar
and a transient pH change of less than 0.1 unit until equilibrium with
the atmosphere occurs.
The addition of a similar quantity of hypochlorous acid or hypo-
chlorite salts would produce approximately the same final equilibrium.
Disequilibrium during the transient initial injection of chlorine where
local higher concentrations would be expected would permit lower pH
and higher chlorine concentrations but cannot be generalized.
The high ionic strength and pH of seawater would encourage the
dissociation of the hypochlorous acid and, assuming an activity
coefficient of 0.6 for hypochlorite ions, the equilibrium would be 90%
hypochlorite ion and 10% hypochlorous acid at 20° C, using pK of 7.54
(Morris 1966). Sugam and Helz (1975) have examined the apparent ioniza-
tion of hypochlorous acid in "seawater" (no bromide ion present) and
found that some ion-pairing of hypochlorite ions with the cations was
present to the extent of approximately 20 percent.
-------
180
Thus, it seems clear that the most pertinent chemistry is that
of the halogen oxyacids. The statements by A. C. Downs and C. J. Adams
in Part 26 of the series, Comprehensive Inorganic Chemistry (Pergamon
Press) on page 1347 provide an excellent introduction to this subject.
"Both thermodynamic and kinetic factors are important in the
chemistry of the halogen oxyacids, and particularly in the interrela-
tionships of the various oxidation states of the halogens in the
condensed phases; moreover, both the thermodynamic and kinetic parameters
which control the reactions of the acids and anlons in solution are
critically sensitive to pH." We may reasonably add the idea that other
aspects of the reaction medium as outlined above will need consideration.
Grounds for intimidation are apparent if reproducible and interpretable
results are sought.
Three broad categories of further possible reactions are:
1. Decomposition
2. Reaction with organics
3. Reaction with inorganics
a. ammonium ions
b. other halogens
The decomposition of hypochlorite solutions can occur through
different competing reactions:
2C1(T -*• 2C1~ + 02 , (1)
2C10" -> C102 + Cl~ , (2)
C10~ + C102 -»• C103 -H Cl~ . (3)
Reaction (1) does not occur readily for the hypochlorite ion as shown
by the stability of commercial bleaching solution. The analogous reac-
tion for the free acid occurs more readily and, if it were predominant,
the environmental effects of our use of chlorine would be greatly
reduced.
Reaction (2) is reported to be slow and solutions of hypochlorite
ions can be expected to contain chlorite ion concentrations that increase
-------
181
with the age of the solution. The use of aged commercial bleaching
solutions or aged reagent grade sodium hypochlorite solutions for
chlorination in bioassay tests or biofouling control usually involves
a determination of the strength of the solutions by the Standards
Methods procedure (1971, sec. 114A). As pointed out by Kolthoff (1922),
rapid titration of solutions buffered with acetic acid-acetate ion
misses the chlorite ion content, since it reacts slowly with iodide
ion. Chlorite ion may be toxic and differences between experiments and
experimenters may arise from the failure of the present "strength"
standardization procedure to measure the chlorite ion content of the
solutions.
For seawater, the high chloride ion concentration may influence
the extent of both reactions (2) and (3). The addition of chlorine may
be expected to yield a mixture of hypochlorite, chlorite and chlorate
ions with the relative abundance of the three species depending on
reaction conditions and time. Photolysis may be especially important.
With regard to reaction with organics, there have not been any
extensive systematic studies using seawater. Dr. Peter Williams of
the Scripps Institution of Oceanography has observed the transformation
of fluorescein to the brominated product, eosin, and the formation of
brominated phenols in chlorinated seawater.
With regard to reaction with inorganics, the formation of chlora-
mines from ammonium ion would be expected and this subject has been
thoroughly reviewed by other participants. It should be noted that the
ammonium ion concentrations in seawater are variable and lower than in
domestic wastewaters. For example, Chesapeake Bay waters vary season-
ally from 50 to 5 micromolar. Florida Current waters contain less than
1 micromolar ammonium ion.
Reactions with other halogens would not involve iodide, since the
abundant form of iodine in seawater is iodate and the concentration is
5 micromolar. Oxidation of fluoride ion would not be expected on
thermodynamic grounds. Reaction with bromide would be the principal
one to be expected, as pointed out by Johannesson (1955) and Lewis (1966)
The resulting bromine would be 99.9 percent hydrolyzed to hypobromous
acid, or tenfold less than the same reaction for chlorine under seawater
-------
182
reaction conditions. The hypobromous acid would be 40 percent dissociated
in seawater. The relative reactivity of bromine in picomolar concentra-
tions and hypobromous acid and hypobromite ion in micromolar concentra-
tions with organics will not be considered here, but poses some
interesting questions.
The problem of further reaction of these bromine species is
analogous to those discussed previously for chlorine, except the litera-
ture is more sparse. Bromamines may be formed. Direct decomposition
will compete with disproportionation reactions. The reaction reported
by Weszelszky (1900) and modified by van der Meulen (1931) is interesting:
Br~ + 3 C10~ ->• BrOa + 3 Cl~ . (4)
The requirement that the solution must contain a large excess of
sodium chloride for complete reaction from left to right is, to use the
vernacular of the American teenagers, somewhat "mind-expanding." The
pH is critical and various authors have recommended buffers such as
bicarbonate, borate, dihydrogen phosphate, etc., with pH in the range
5.5 to 8.5. Szabo and Csanyi (1952) report that the reaction takes
place in two main steps:
2 Br~ + C12 -»• Br2 + 2 Cl~ ,
Br2 + H20 -> BrOH + H + Br~ , (5)
BrOH + C12 -> BrCl + HOC1 ;
and
3 BrOH -»• BrOa + 3 H+ + Br~ ,
BrCl + 2 C12 ->• BrCl5 , (6)
BrCl5 + 3 H20 ->• BrOa + 6 H+ + 5 Cl~ -
-------
183
The first steps are favored by pH 6.5 to 7.5 and the second steps are
favored by somewhat higher pH. The details of the proposed reactions
are probably not in close agreement with current reaction mechanism
theory but clearly illustrate the multistep system with which we must
deal.
Recent unpublished work at the University of Miami has sought the
identification of bromate as a product of the chlorination of seawater.
With 30-minute reaction time, polarographic observations after destruc-
tion of hypochlorite ion with thiosulfate show that bromate is not
present at concentrations as large as 10 micromolar.
Another possible product is bromite, but there is no evidence
cited in the literature for bromite formation in chlorinated seawater.
In addition to the oxyanions, there is the possibility that
interhalogen complexes are formed in seawater. Pungor et al. (1959),
have presented spectroscopic evidence for the existence of chloride
ion complexes of bromine chloride in aqueous solution. Identical
spectra were found in 0.5 N NHj+Cl and 0.5 N NaCl. They suggest the
formula BrCl65~ for the complex, but Gutmann et al. (1968) express the
view that other chloride complexes with less chloride per bromine may
also account for the spectra. The bromine chloride complexes show lower
redox potentials than "free bromine chloride" by 0.2 volts and, if such
species are formed when chlorine is added to seawater, possible additional
thermodynamic and kinetic phenomena will be present. Also, bromine
chloride is reported to react more rapidly with aromatic and olefinic
compounds.
The impression that we have gleaned from the literature is that
chlorination of seawater may be expected to produce several inorganic
species containing bromine (these compounds may be directly toxic to
aquatic organisms) and, in view of the reactivity with organics of the
expectable bromine species, a variety of brominated or mixed chlorobromo
compounds is probably being produced with present chlorination of power
plant cooling waters and chlorinated wastewater discharges.
-------
184
WHAT DO PRESENT ANALYTICAL METHODS MEASURE?
Colorimetric methods are popular because of the apparent simplicity.
The most widely used reagent is orthotolidine and some state agencies
require chlorination of wastewaters to residual levels as determined
with this reagent. As reported by Bellanca and Bailey (1975), the use
of orthotolidine produced a serious underestimation of residual chlorine
and resulted in extensive fish kills in the James River estuary. Eppley
et al. (1975) found that chlorinated seawater samples, judged to be
chlorine-free with the orthotolidine method, were still inhibitory to
phytoplankton photosynthesis. Carpenter et al. (1972) observed that
chlorination of seawater used for power plant cooling, at levels that
produced no detectable residuals at the outfall, decreased phytoplankton
productivity by 79 percent. It seems clear that the orthotolidine
procedures are badly misrepresenting the "residuals" in wastewaters
and seawater.
Various methods have been compared as reported in Standard Methods
(1971). The iodometric method is considered the standard against which
other methods are judged. Standards were distributed to 32 laboratories
and the results show a relative standard deviation of ca. 27 percent for
the iodometric method — an appalling situation. The amperometric titra-
tion method is "rated among the most accurate for the determination of
free or combined available chlorine." The iodometric method includes
the use of starch for visual endpoint estimation in the presence of
0.012 molar potassium iodide and the amperometric method includes the use
of a galvanic cell in a deadstop endpoint procedure in the presence of
0.0015 molar potassium iodide. The amperometric endpoint detection is
more sensitive than visual starch indication, as pointed out by Bradbury
and Hambly (1952). However, the electroactive species appears to be
free diatomic iodine, rather than triiodide ion that is formed in the
presence of high concentrations of iodide ion. The lower concentration
of iodide used in the amperometric method reflects this characteristic.
However, the volatility of free iodine is well known and severe errors
may result from the violent agitation used in commercial amperometric
instruments. The magnitude of the errors depend on the particular ways
-------
185
in which various people carry out the manipulations and, with practice,
precise but erroneous data can be generated by any one individual, as
demonstrated by Carritt and Carpenter (1966). Present methods do not
appear to be satisfactory if close compliance with regulatory statutes
is to be demonstrated by analysis of samples from wastewater treatment
plants and electric generating stations.
We have carried out some tests of methods for estimating "residual
chlorine" in seawater and other solutions in the following manner.
EXPERIMENTAL PROCEDURE
In each of our experiments, sufficient chlorine reagent was added
to a measured volume (usually about 250 ml) of the test solutions to
achieve a chlorine concentration in the range 2 to 12 ppm. In some of
the early trials, Fisher Chlorine Test Solution, U.S.P., was used. For
later trials, a chlorine reagent was made by diluting a saturated
(15°C) solution of C12 gas in distilled water. The dilute chlorine
reagent was prepared fresh daily and buffered with 0.1 g Na2C03 per
100 ml. (Results were observed not to be related to the origin of the
Cl2 reagent.)
The five different test solutions were prepared as follows:
1) Ion-exchanged and distilled water, with 2.0 x 10~3 M NaHC03
added.
2) 1.0 x 10~3 M KBr in 2.0 x 10~3 M NaHC03.
3) 35°/oo NaCl in 2.0 x 10~3 M NaHC03. Due to impurities in
the reagent grade NaCl, this solution was also about 10~5 M
in Br~.
4) 35°loo NaCl in 2.0 x 10~3 M NaHC03 with KBr added to make
the solution 1.0 x 10~3 M in Br~.
5) Florida Current seawater, filtered through a 0.22 y Teflon
Millipore filter.
During a typical experiment, identical quantities of the chlorine
reagent were added simultaneously to each of the 5 test solutions. The
chlorinated test solutions were then allowed to react with gentle
stirring at room temperature for 30 minutes. Analysis for "total
-------
186
residual chlorine" was then made by an iodometric procedure using
standardized sodium thiosulfate solution as the reducing titrant and
5 percent starch solution as an indicator.
Three different chemical conditions for generation of iodine were
created for each set of analyses.
In some tests, analytical conditions were established to duplicate
those prescribed in Standard Methods, Section 114B, the amperometric
titration method. Here 1.0 ml of a 50 g/1 solution of KI and 1 ml of
acetic acid/sodium acetate buffer (pH = 4) were added sequentially to
the 250 ml of test solution immediately prior to titration.
In other test sets, analytical conditions similar to Standard
Methods, Section 114A, were achieved by adding 2 ml of a 50 g/1 KI
solution followed by 2 ml of the pH 4 buffer immediately prior to
titration.
In a third group of experiments, iodine was generated by adding
2 ml of 50 g/1 KI followed by sufficient 10 M R2SOit to achieve a pH of
2.0. Titration was again begun immediately following the addition of
the KI and acid.
After titration to an initial endpoint (disappearance of I3~starch
blue color), the titration flasks were stoppered and allowed to stand,
unstirred, in the laboratory. Any additional blue color which appeared
after 30 minutes was then titrated with additional thiosulfate. In
many of the tests, the flasks were then allowed to stand for longer
periods of time, with subsequent titration of reappearing blue color.
Periodic checks for air oxidation of KI solutions were made by
adding starch directly to buffered KI solutions and noting any
appearance of blue color. If a color developed, fresh solutions
were prepared.
Standardization of the thiosulfate was carried out according to the
biiodate method outlined in Standard Methods, Section 114A.
RESULTS
The results of one experiment in which 1.8 ppm of chlorine was
added to the five different solutions are shown in Fig. 1. The
-------
187
differences observed with titrations of distilled water solutions under
different pH and potassium iodide concentrations are larger than we
have found in subsequent experiments. The data in Fig. 1 were collected
using a piston microburette that permits very rapid titration. With
slower titration (1 to 2 minutes), permanent endpoints are found and the
three titration conditions show differences of less than 8 percent with
the solution containing the lower concentration of potassium iodide
requiring the least titrant.
The seawater solution showed a "residual" of 0.7 ppm chlorine or
an apparent demand of 1.1 ppm, with the recommended potassium iodide
concentration for amperometric titration. The Florida Current seawater
that was used would be expected to contain only 0.1 ppm of total
organic carbon and the "chlorine demand" is seen to be related to the
titration conditions; i.e., the pH 2, higher potassium iodide concen-
tration results indicate a demand of only 0.2 ppm and the apparent
large demand is an artifact of the particular titration conditions. This
viewpoint is strengthened by our observation that when the same seawater
was chlorinated to an added level of 11.4 ppm, the "apparent demand"
was 4.9 ppm, which cannot be accounted for in terms of reducing sub-
stances in such clean seawater. When the titrated solution was stored
in the dark for 48 hours, the starch triiodide blue color reappeared
and titration to the "new" endpoint showed an apparent demand of 2.6 ppm.
The same disappearance of the added chlorine was found in potassium
bromide solutions and in potassium bromide-sodium chloride solutions
when they were titrated at pH 4 with the lower concentration of potassium
iodide. With higher concentration of potassium iodide or lower pH,
these solutions showed no significant "loss of chlorine" during the
30-minute reaction time.
The results for the sodium chloride solution are somewhat perplexing.
In subsequent experiments, the lower concentration of potassium iodide
produced some apparent chlorine demand for additions of 1 to 3 ppm
chlorine, but this effect was smaller when additions of 10 to 26 ppm
chlorine were made. With slower titration of several minutes, results
for the sodium chloride solution were similar to those for distilled
-------
LOW Kl
HIGH KI
HIGH KI,PH*2
DISTILLED WATER
SEA WATER
HIGH KI,pH=2
KBR AND
LOW Kl
HIGH Kl
HIGH Kl.
00
00
KBR IN DISTILLED HzO
NO DATA FOR LOW Kl
HIGH Kl
HIGH Kl, PH=2
NACI
I
O.I
I
0.2
I
0.3
I
0.4
MILLILITERS THIOSULFATE
Fig. 1. Titration of solutions that had 1.8 ppm chlorine added.
Rapid titration shown as the clear bar and additional titration after
30 minutes shown as cross-hatched.
-------
189
water. If the solutions were stored for 24 to 48 hours, there was no
significant chlorine demand. The sodium chloride was not free of
sodium bromide and some of the kinetic effects could be due to the
bromide rather than the chloride content of the solutions.
The slow reaction of the test solutions with potassium iodide can
occur if chlorite is present. As pointed out by Kolthoff et al. (1957),
if the solutions are acidified before adding the potassium iodide,
low results are obtained owing to the formation of chlorine dioxide.
We examined our chlorination solutions for evidence of the presence
of chlorite ipns by comparing the results with acidification before and
after the addition of potassium iodide. As shown in Fig. 2, there was
no evidence of significant chlorite concentrations. Also, with 30
minutes of reaction time after chlorination, no evidence for chlorite
was found.
A possible reason for the smaller titrant volumes required for
chlorinated seawater compared to chlorinated distilled water could be
the formation of substances that oxidize iodide ion slowly but oxidize
thiosulfate ion to sulfate ion, rather than to tetrathionate ion.
Phenylarsine oxide does not have the multiple oxidation state character-
istics that thiosulfate does. We have titrated the iodine in replicate
seawater samples, chlorinated to 12 ppm, with thiosulfate and phenylar-
sine oxide without finding differences that were larger than the
reproducibility of the experiments.
DISCUSSION
Our observations are inconclusive but it is clear that the widely
used procedure for estimating "residual chlorine" by rapid titration
of seawater samples, immediately after adding millimolar potassium
iodide and adjusting to pH 4, underestimates the actual "residual
oxidizing components" to an extent that the results are misleading.
The identification of the chemical species that are formed when chlorine
is added to seawater is a prerequisite to designing proper analytical
methods. However, the lack of understanding of exactly what occurs
when seawater is chlorinated should not distract attention from the
-------
Kl ADDED
BEFORE BUFFER
ACETATE
BUFFER ADDED FIRST
30 MINUTE REACTION TIME
Kl ADDED BEFORE BUFFER
I
.2
I
.3
I
.4
I
.5
MILLILITERS THIOSULFATE
I
.6
30 MINUTE REACTION TIME
ACETATE BUFFER ADDED FIRST
Fig. 2. Results of tests for chlorite in the solutions used for
chlorination. Acetate buffer addition before potassium iodide would
produce lower results if chlorite were present.
-------
191
significance of the present empirical observations on the responses of
aquatic organisms to acute chlorine exposure, as reported in other
presentations at this conference. The need for assessment of the
effects of chronic exposure to the inorganic and organic products of
seawater chlorination seems obvious.
ACKNOWLEDGMENTS
Much of the experimental work was performed carefully by
Cynthia Moore. The research was supported by the U.S. Environmental
Protection Agency, Grant No. R 803893-01.
REFERENCES
Bellanca, M. A., and D. S. Bailey. 1975. A case history of some
effects of chlorinated effluents on the aquatic ecosystem of the
lower James River in Virginia. Abstract, 48th Annual Conference,
Water Pollut. Control Fed., Washington, D.C.
Bradbury, J. H., and A. N. Hambly. 1952. An investigation of errors
in the amperometric and starch indicator methods for the titration
of millinormal solutions of iodine and thiosulfate. Australian
J. Sci. Res., Ser A, 5: 541-554.
Carpenter, E. J., B. B. Peck, and S. J. Anderson. 1972. Cooling water
chlorination and productivity of entrained phytoplankton. Mar.
Biol. 16: 37-40.
Carritt, D. E., and J. H. Carpenter. 1966. Comparison and evaluation
of currently employed modifications of the Winkler method for
determining dissolved oxygen in seawater; a NASCO report. J. Mar.
Res. 24: 286-318.
Eppley, R. W., E. H. Renger, and P. M. Williams. 1975. Chlorine
reactions with seawater constituents and inhibition on photosyn-
thesis of natural marine phytoplankton. Estuarine Coastal Mar.
Sci. In press.
Gutmann, H., M. Lewin, and B. Perlmutter-Hayman. 1968. The ultra-
violet absorption spectra of chlorine, bromine and bromine chloride
in aqueous solution. J. Phys. Chem. 72: 3671-3673.
Johannesson, J. K. 1955. Note on the chlorination of water in the
presence of traces of natural bromide, N. Z. J. Sci. Technol.,
36B: 600-602.
-------
192
Kolthoff, I. M. 1922. Rec. trav. chim. 41: 740, as quoted in Volumetric
Analysis, Interscience Publishers, Inc., New York, 1957, p. 262.
Kolthoff, I. M., R. Belcher, V. A. Stenger, and G. Matsuyama. 1957.
Volumetric Analysis, page 267. Interscience Publishers, Inc.,
New York.
Lewis, B. G. 1966. Chlorination and mussel control. I: The chemistry
of chlorinated seawater: a review of the literature. C.E.R.L.
Report No. RD/L/N 106/66.
Morris, J. C. 1966. The acid ionization of HOC1 from 5 to 35°C. J.
Phys. Chem. 70: 3798-3805.
Pungor, E., K. Burger, and E. Schulek. 1958. Interhaloid complexes
in aqueous solution. J. Inorg. Nucl. Chem. 11: 56-61.
Riley, J. P., and G. Skirrow. 1975. Chemical Oceanography. Academic
Press, New York.
Standard Methods for the Examination of Water and Wastewater, Thirteenth
Edition. 1971. American Public Health Association, Washington, D.C.
Sugam, R., and G. R. Helz. 1975. Apparent ionization constant of
hypochlorous acid in seawater. Research in Aquatic Geochemistry,
University of Maryland, College Park, Maryland.
Szabo, Z. G., and L. Csanyi. 1952. Anal. Chim. Acta 6: 208, as quoted
in Volumetric Analysis, Interscience Publishers, Inc., 1957, p. 255.
Van der Meulen, J. H. 1931. Chem. Weekblad. 28: 82, as quoted in
Volumetric Analysis, Interscience Publishers, Inc., New York,
1957, p. 255.
Weszelszky, von J. 1900. Z. anal. chem. 39: 81, as quoted in Volu-
metric Analysis, Interscience Publishers, Inc., New York, 1957,
p. 254.
DISCUSSION
David H. Rosenblatt. U.S. Army Medical Bioengineering Research and
Development Laboratory. The formation of chlorite in hypochlorous
acid-hypochlorite ion stock solutions is most unlikely in view of the
careful study by D'Ans and Freund, Z. Elektrochem. 61, 10 (1957), on
the formation of chlorate from such solutions. The mechanism is:
2HOC1 + OC1~ = [H2C1303] = 2H+ + 2C1~
This completely bypasses possible formation of C102, which those authors
looked for and did not find.
-------
193
J. Donald Johnson. University of North Carolina-Chapel Hill. How
did you dismiss bromate, BrO"3?
Carpenter. We couldn't find it polarographically after we destroyed
the hypochlorous acid with thiosulfate. It makes a beautiful wave at
pH 7.
Johnson. Did you have enough sensitivity?
Carpenter. Yes. It is not more than 5 x 10~5 molar.
Johnson. Well, I don't know anything about this polarographic
method you mentioned, but I like bromate.
Carpenter. Well I did too. I was very frustrated when I didn't
find it.
Walter J. Blogoslawski, National Marine Fisheries Service. In some
recent work we have found the presence of a long-lived oxidant (measured
by iodometric titration) after ozonization of seawater. This compound
may be a bromite or hypobromite as suggested by the work of McKinney
at Dow Chemical Company, Freeport, Texas. Chlorine when added to
seawater may produce a similar effect.
Carpenter. Did you say hypobromite or bromite?
Blogoslawski. We think it might be hypobromite.
-------
EXTENDED DISCUSSION AFTER SESSION I
Josepji E. Draley, Argonne National Laboratory. Are there comments
concerning the subject of the symposium thus far, namely the chemistry
of chlorine in water. It occurs to me to remark that we have a lot of
confusion about what is needed. I ought to define what I mean by what
is needed. Let's eliminate temporarily the pure research that we do
because we are engaged in fundamental research, and ask the question
from the point of view of the needs for the control of the addition of
substances to our environment or to our drinking water that would be
harmful either to people or to our ecology. What things do we need?
We've found or mentioned a number of uncertainties and unknowns. I'd
like to ask you to offer comments on what our problems are that we could
consider in a kind of a focused way.
Robert S. Ingols, Georgia Institute of Technology. Chlorinated
organic pesticides apparently concentrate in the biota. Therefore I
believe that data should be obtained on the accumulation of chlorinated
organics in the biota below wastewater discharges. This may be a very
important point.
Draley. Are there others who would like to relate to that partic-
ular topic? Do we know that if you accumulate this material in plankton
or in a higher species that it is automatically bad if it contains
chlorine? Is there a tenet that if you get chlorinated organics in these
systems that it will be bad for them or that it will be bad for us? Is
there a threshold level below which you don't have to worry and above
which you do have to worry? Do we know these things or are we speculating?
If we are speculating, do we need to know?
William H. Glaze, North Texas State University. Obviously, I can't
answer that question. But I would like to tell you about some results,
and this sort of anticipates tomorrow. I am reporting the results of a
study that was done this summer in my laboratory. There's a real neat
system close by to us, namely a power plant lake, which is almost dead-
end. That is to say, except at flood stage, there's not any water getting
out except by evaporation, but there is water being brought in from a
river that rivals the Ohio, and Trinity. So there's a good deal of
organic material in the lake both from natural and unnatural sources.
The point of this is that I had a student who did look at accummulation
of chlorinated organics, not specific compounds, but as measured by
195
-------
196
TOC1 in fish, in perch, in that lake as compared to a control lake —
that is, a lake of about the same size, about the same terrain, which
was only getting agricultural run-off as far as we could tell. There
was significant higher concentration as measured by TOC1, the way I
talked about it a minute ago, in the power plant lake. That lake being
such that it takes water through the cooling towers and chlorinates
periodically, and so it's a natural system for the buildup of these
compounds that we've been talking about. What the concentration is in
the lake, I don't know. We are now in the process of looking at the
lake water and the sediments thereof. We are going to look at specific
compounds, and also the TOC1 values, in the lake and the biota thereof.
Draley. I have a comment here. Bob Baker, I am going to ask you
about the measurement of chlorine in seawater in a little while. You
may be thinking about how terrible a question it is going to be.
H_._Sikka, Syracuse University. I think the possibility for bio-
accumulation of the compounds mentioned by Dr. Jolley is going to be
much less compared to the chlorinated hydrocarbon pesticides. This is
because they are much more polar chlorocompounds than the chlorinated
pesticides. As a result the bioaccumulation by fish and animals will
not be much of a concern. Of course, this does not exclude the possi-
bility of doing research in that area.
Draley. No. The reason I asked the question is you can't do
everything you can possibly think of. Somehow we have to find a way to
guide those who fund and those who do research for, shall we say,
practical benefit, in what needs to be done.
Glaze. I think you're quite right. Nevertheless, there is some
significant chance for accumulation. But another thing to understand
now is what I think Jolley pointed out also, that it's not just accumula-
tion that we're interested in. We are also interested in other specific
effects. When you look at 5-chlorouracil and think about the implica-
tion there, and when you think about chlorinated amino acids and we know
that some of the chlorinated amino acids are antagonists for the
unchlorinated amino acids, then it's not necessarily bioaccumulation
that one is concerned about. One is worrying about some of those
specific effects whether they be communication or genetic or whatever,
which were referred to earlier.
Sikka. Yes, but these are two separate areas. We have to distin-
guish between the two areas.
Robert B. Dean, U.S. Environmental Protection Agency. We have been
looking a lot at sewage. We have been looking a lot less at what is
produced when you chlorinate paper pulp. Chlorinated paper pulp and
paper mill wastes can be expected to contain chlorinated terpene com-
pounds similar to known chlorinated pesticides. These would be expected
to bioaccumulate.
-------
197
Unknown. I think bioaccumulation is certainly a thing we need to
look at. Another thing relating to bioaccumulation, which is really more
broad than bioaccumulation, and that is movement. Movement in the
sediment, whether it is facilitated by bacteria or not, should be studied.
Maybe some compounds go into storage. This might be important.
Draley. I would like to ask my question now of Bob Baker about
the measurement of chlorine in seawater. The reason I brought it up
is that in the last talk some problems were discussed with respect to
what one can hope to measure by what might be called ordinary or
present methods. Do you have any comment to make?
Robert J. Baker, Wallace & Tiernan Division, Pennwalt Corporation.
I see you've done a perfect job of putting me on the spot. I've gone
through about three phases where I thought we were measuring chlorine,
and then I thought we were measuring bromine, and now I haven't the
vaguest idea. What actually occurs is that the results we get are simply
recorded as if they were chlorine. That doesn't really answer your
question. In some of the work we have done, obviously, the results
don't seem to be comparable as far as the reaction is concerned. This
comment on the flash back — I've spotted it in some cases, and in some
cases I have not. That's all I can offer.
Draley. Other comments?
Robert L. Jolley, Oak Ridge National Laboratory. Relative to
chlorination of paper waste effluents, Dr. Larry Keith at the South-
eastern Environmental Research Laboratory, Athens, Georgia, has done
some work. He has analyzed several chlorinated phenolic ethers.
Glaze. The Fisheries Research Board of Canada has done a whole
series of publications where they have looked for chlorinated compounds
in paper effluents and are beginning to look at specific toxicities.
In one of the last editions there was a good bibliography for those
interested in chlorination at paper mills. This is a bleaching process.
There is some good research and they are getting into it in the same
way that you have done here.
-------
DECISION MAKING IN THE REGULATION OF CHEMICALS
Edward M. Brooks
Chief, Policy Review Staff
Office of Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
I'm delighted to have the opportunity to talk to you this evening.
I must confess that, being neither scientist nor engineer, I approach
the opportunity with no little trepidation. I am somewhat consoled by
the fact that my topic has to do with decision making in the regulation
of chemicals, since most decision-makers in this area are also not
scientists — a fact which must no doubt frustrate many of you from time
to time. Of course there is that body of opinion which holds that the
regulation of toxic substances is too important to leave to toxicologists.
In any event, the information exchanged — or lack thereof — between
scientists and the decision-maker is what I want to discuss tonight.
When he approached me to give this talk, Dr. Jolley was kind
enough to suggest a number of questions I might address, among which I
found four particularly provocative:
1. What type of toxicity data are considered in making regulations?
2. Can cost-benefit analyses be applied in developing regulations?
3. What is the mechanistic procedure or protocol for developing
regulations? and
4. What can data generators do to make decision making easier?
199
-------
200
All four questions touch upon a more general proposition I now put
to you, namely that regulations to^ control serious chronic toxicants are
not developed within a. consistent logical framework. I will first
demonstrate the truth of this general proposition, and then explore a
few reasons why it is the case and what, if anything, ought be done.
Before embarking let me define, for purposes of discussion, two key
phrases in Brooks' proposition. By "serious chronic toxicants" I n*»an
those that are of concern because they may cause death or illness to
humans — either after many years of continuous exposure, or after a
latency period of many years, — at levels markedly below the dosage at
which the tolerable levels of risk can be detected in laboratory test
animal experiments. Such toxicants include but are not, in my opinion,
limited to human chemical carcinogens.
The term "consistent logical framework" refers to a system that
imposes a stipulated set of values, principles and rules upon the manner
in which data are evaluated and exploited to reach decisions — such that
different players, operating independently with the same information,
or lack of information, reach essentially the same conclusions. A major
function of such a system is to compel consideration of costs, risks and
benefits across the range of available regulatory options.
The value of such a rational approach to regulatory decision-making
is generally appreciated. The National Academy of Sciences (NAS)
recently completed a study entitled Decision Making for Regulating
Chemicals in the Environment (NAS 1975) the fruit of which is a series
of 34 recommendations — from four of which I quote in part.
"17. The quality of chemical regulatory decisions is dependent
largely upon the adequacy of the available information.
To develop an adequate data base, research efforts in
basic clinical and environmental toxicology and epidemi-
ology and in economic analysis must be strengthened ..."
"30. Highly formalized methods of benefit-cost analysis can
seldom be used for making decisions about regulating
chemicals in the environment. However, benefit-cost and
decision frameworks can be useful in organizing and
summarizing relevant data on regulatory alternatives
which the decision maker must review."
-------
201
"31. Value judgments about noncommensurate factors in a decision
such as life, health, aesthetics, and equity should be
explicitly dealt with by the politically responsible
decision makers and not hidden in purportedly objective
data and analysis."
"32. The decision process should require the agency's technical
staff to present a full set of options with a corresponding
range of cost-benefit-hazard data and explicit statements
on the confidence limits of each analysis."
The Academy thus expects staff scientists to identify several
regulatory alternatives and objectively estimate and present to the
politically accountable decision-maker the costs, risks and benefits
associated with each — together with explicit probability statements
regarding the reliability of those estimates. From these analyses the
decision-maker selects a regulatory option and, in proposing and promul-
gating the decision, explicitly, sets forth the value judgments he applied.
'Tis a consummation devoutly to be wished.
It will be instructive to look at a few proposed or promulgated
regulations to see how closely they approach this ideal. I emphasize
two points. First, nothing that follows is intended as criticism of any
given regulatory agency or decision. Further, where discrepancies are
found between different regulations controlling the same substance I
offer, in these remarks, no opinion regarding their relative merits.
My purpose here is to examine what has been done to assess possible
weaknesses in the process rather than in any particular regulation.
Secondly, I trust it goes without saying that any weaknesses found hardly
constitute grounds for not continuing to aggressively implement our
diverse authorities as best we can. I will now discuss four proposed or
promulgated standards — two for asbestos and one for vinyl chloride and
aflatoxins.
Clean Air Act — Sec. 112 — Asbestos;
In April of 1973, the U.S. Environmental Protection Agency (EPA)
promulgated a regulation (EPA 1973) to control asbestos emissions at a
level designed to protect human health with an "ample margin of safety."
In the preamble to this standard the Agency stated that no numerical
-------
202
concentration or mass emission limit was practicable because it is
"impossible to estimate even roughly the quantitative relationship
between asbestos-caused illness and the doses that cause those illnesses."
Although we concluded that there are no levels known at which asbestos
does not involve risk, and that the effecto of inhaling asbestos are
cumulative, we did not ban the substance outright because to do so would
prohibit many extremely important activities. Accordingly the standard,
in major part, simply banned "visible emissions" of asbestos,
OSHA — Sec. 6 — Asbestos;
One week ago Thursday last, the Occupational Safety and Health
Administration (OSHA) of the Department of Labor proposed a new standard
(OSHA 1975) to regulate work place exposure to asbestos. The initial
OSHA asbestos standard (OSHA 1971a), promulgated in May of 1971, was 12
fibers, not longer than 12 micrometers, per cubic centimeter. In
December of 1971 (OSHA 1971b) this was reduced to five fibers, no longer
than five micrometers. The current standard (OSHA 1972), promulgated in
1972, established the level at five fibers, but added the provision that
this would automatically go down to two fibers on July 1, 1976. Now the
October 9 proposal would reduce the level, still further, to 0.5 fibers.
The proffered rationale is that (1) sufficient evidence has accumulated
to warrant designating asbestos a human carcinogen, (2) a "no effect"
level has not been demonstrated, and (3) in the absence of evidence to
establish a safe level, employee exposure must be reduced "as low as
feasible."
Thus both EPA in its promulgation under Sec. 112 of the Clean Air
Act and OSHA in its promulgation under the Occupational Safety and Health
Act concluded that there is no known exposure level for asbestos at which
adverse human health effects do not occur. Within that framework EPA
concluded that no numerical concentration limit is possible because the
dose-response relationship is unknown, while OSHA not only established a
numerical limitation, but has systematically reduced it over the years.
The significant point is that no data have been provided to the
decision-maker in either Agency regarding the dose-response relationship.
-------
203
Relative hazard assessment is therefore out of the question. In this
regard it is worth noting that, although the rationale for the OSHA
standard is to reduce exposure levels to as "low as feasible," no evi-
dence is provided to suggest that 0.5 fibers is either feasible or safe.
In point of fact the rationale and evidence used to justify the 1972
five fiber standard could just as well have been used to justify the
1975 0.5 fiber proposal, and vice versa. The evidence to establish a
"no effect" or "safe" level was just as absent when OSHA promulgated the
five fiber standard to protect against asbestosis as it was a week ago
Thursday last when it proposed the 0.5 fiber standard to protect against
mesothelioma — and this lack of evidence has nothing apparent to do with
any special attributes of carcinogens. In neither instance were "no
effect" levels or dose-response relationships established.
^
I now want to contrast the OSHA vinyl chloride standard with the
tolerance for aflatoxin proposed by the Food and Drug Administration
(FDA) .
OSHA - Sec. 6 - Vinyl Chloride:
In October of 1974 OSHA promulgated a standard (OSHA 1974) for
vinyl chloride based on (1) the fact that 31 vinyl chloride workers had
died of angiosarcoma of the liver, (2) Maltoni's experiments inducing
angiosarcoma in rats at 250 parts per million, and (3) Industrial Bio-
Test Laboratories' studies inducing angiosarcoma in rats and mice at
50 ppm. OSHA concluded in accordance with the 1970 Report of the
Surgeon-General's Ad Hoc Committee on the Evaluation of Low Levels of
Environmental Chemical Carcinogens that, on the basis of the demonstra-
tion of cancer in two animal species, vinyl chloride posed a carcinogenic
hazard to man. In further accordance with that report, OSHA took the
position that "safe exposure levels for carcinogenic substances can not
be scientifically determined" — a position supported at the Hearings
by both NIOSH and NCI. On the grounds that vinyl chloride is a carcinogen
and "safe" levels for carcinogens cannot be established, OSHA promulgated
a "no detectable level" standard as measured by methods sensitive to
one ppm.
-------
204
FD&CA — Sec. 406 — Aflatoxins;
The Food and Drug Administration's proposed December, 1974, toler-
ance for aflatoxins (FDA 1974) in shelled peanuts and peanut products,
provides an interesting contrast. The Federal Register NPRM notes that
25 rats, fed aflatoxins at 15 parts per billion, all developed liver
cancer, as did monkeys fed aflatoxin. In addition, epidemiological
studies in Southeast Asia and Africa indicated a correlation between the
incidence of liver cancer in humans and exposure to aflatoxins. FDA
concluded from this that human exposure should be held to the "lowest
level possible."
FDA survey data indicated that four percent of the U.S. peanut
butter exceeded the 20 ppb level, seven percent exceeded 15 ppb, 11 percent
exceeded 10 ppb, and 25 percent approximated one ppb. Thus, about four
percent of the U.S. production fell between 10 and 15 ppb, and three
percent between 15 and 20 ppb. FDA proposed a 15 rather than 10 ppb
tolerance to "avoid causing significantly increased losses of food."
Thus, in quite comparable situations —with substances demonstrated
to be carcinogenic in two animal species and with strong epidemiological
evidence implicating them as human carcinogens — OSHA promulgated a
"no detectable level" standard (for vinyl chloride) well below the lowest
levels at which any adverse effects have been found in any species, while
FDA proposed a tolerance (for aflatoxin) at the same level at which
25 of 25 mice developed liver cancer. The two Federal Register Notices
reflect the different philosophies. First the OSHA text regarding
vinyl chloride.
"There is little dispute that vinyl chloride is carcinogenic
to man and we so conclude. However, the precise level of
exposure which poses a hazard and the question of whether a
'safe1 exposure level exists cannot be definitively answered
on the record. Nor is it clear to what extent exposures can
be feasibly reduced. We cannot wait until indisputable
answers to these questions are available, because lives of
employees are at stake. Therefore, we have had to exercise
our best judgment on the basis of the best available evidence.
These judgments have required a balancing process in which
the overriding consideration has been the protection of
employees, even those who may have regular exposures to vinyl
chloride throughout their working lives."
-------
205
And the counterpart passage from the FDA proposal regarding aflatoxins.
"In addition, because there is no direct evidence that aflatoxins
cause cancer in man or of what may be the level of no effect, the
Commissioner cannot conclude that there is any tangible gain from
lowering the permissible level to either ten or five ppb. Such
uncertain benefit to the public health must be weighted against
the clear loss of food that would result."
Again the point is not who, if anyone, is right or wrong, but
rather that no data were provided to either decision-maker to permit a
reasoned analysis of the risks incurred at various possible exposure
levels. Without such information the levels can and will be set almost
anywhere from zero up to the levels that obtain without any regulation
at all. Any consonance between such permitted exposure levels, and a
balanced tradeoff between health and economic impact considerations, will
be purely fortuitous.
The inadequacies of these regulations warrant a moment's reflection.
Most remarkable is the fact that none makes any attempt to specify an
"acceptable" level of risk. Instead they offer analytically meaningless
platitudes about the need to reduce exposures "as much as possible" or
"feasible." There are no estimates of the extent of the adverse human
health effects presumed to be caused by these substances, much less of
the extent to which this incidence is expected to be reduced by the
regulation. Not only is it impossible to evaluate how well these regula-
tions achieve their objectives —we can't even define the objective. In
the most fundamental sense, then, it is impossible to assess their worth.
Even if "acceptable" levels of risk had been established, however,
there are still no health effects data available to indicate the exposure
levels at which those risk levels would be exceeded. In this situation
it is fatuous to speak of cost-risk-benefit analyses, judicious tradeoffs,
balancing competing factors, or any other phrase that connotes a reasoned
application of useful information.
One must ask how this came to pass? How is it we write regulations
in such an important area with so little comprehension of what we are
about? From whence comes the pressure to promulgate such regulations?
Why is the scientific documentation such a paltry product? What impels
-------
206
decision-makers to act upon such tenuous evidence? Aside from our
ignorance, what accounts for the striking inconsistencies found in these
regulations — not only in their stringency but, more basically, in their
underlying philosophy?
While there are obviously many reasons, I would like to briefly
mention two contributing factors. First I believe that we may sometimes
be moved to precipitate and unwarranted action in response to public
pressure. In this regard, incidentally, I have just read a perceptive
article in the Fall issue of The Public Interest (Nisbet 1975) which
reflects my concern precisely. I commend it to your attention and will
attempt to whet your appetite with the following quote.
"Of all the heresies afloat in modern democracy, none is greater,
more steeped in intellectual confusion, and potentially more
destructive of proper governmental function than that which
declares the legitimacy of government to be directly proportional
to its roots in public opinion — or, more accurately, in what
the daily polls and surveys assure us is public opinion. It is
this heresy that accounts for the constantly augmenting propaganda
that issues forth from all government agencies today — the
inevitable effort to shape the very opinion that is being so
assiduously courted — and for the frequent craven abdication of
the responsibilities of office in the face of some real or
imagined expression of opinion by the electorate."
This tendency to precipitate action in the face of uninformed but
aroused popular opinion is reinforced by the wide and increasing dis-
crepancy between our ability and disposition to detect the presence of
potential toxic substances and our ability to assess the degree of risk
they pose. EPA's recent and continuing concern with organics in drinking
water illustrates this phenomenon. Surveillance and analytical tech-
nology now yields impressively precise quantification of very low levels
of organics in water, while the state of the health effects assessment
art apparently permits only crude qualitative estimates of the human
health hazards posed at these levels. In 1969 the Federal Water Pollu-
tion Control Administration found chloroform, benzidine and bis-
chloromethyl ether in the New Orleans drinking water (EPA 1972). In
1974 EPA found 66 organic chemicals in the New Orleans drinking water
(EPA 1974). This past year we completed a survey of the water in 80
-------
207
cities and found at least one of the six organics for which we sampled —
and most particularly chloroform — in every location (EPA 1975a). At
various stages in the course of these events substantial pressure was
brought to bear on EPA to "do something." But what do we really know?
When asked to review the findings of the 80 city National Organics
Reconnaissance Survey, EPA's Science Advisory Board concluded that:
"... there may be some cancer risk associated with consumption
of chloroform in drinking water. The level of risk, estimated
from consideration of the worst case and for the expected cancer
site for chloroform (the liver) might be extrapolated to account
for up to 40% of the observed liver cancer incidence rate. A
more reasonable assumption, based upon current water quality data
which show much lower levels than the worst case in the majority
of U.S. drinking water supplies, would place the risk of hepatic
cancer much lower and possibly nil. Further, it is emphasized
that both the experimental carcj.nogenicity data and the mathe-
matical and biological extrapolation principles used to arrive
at the upper estimate of risk are extremely tenuous. Epidemic-
logic studies do not, thus far, support the conclusion of an
increased risk of liver cancer; although hypothesis formulating
studies in southern Louisiana suggest the possibility of an
association with contaminated water and overall high cancer
incidence" (EPA 1975b).
There is no obvious solution to this problem. It is clearly
important for a public agency to be responsive to the public; this not-
withstanding it is at least equally important to allocate resources and
conduct the public's business in an orderly and reasoned manner.
There is also apparently no immediate solution to our inability to
quantify the human health risks associated with low levels of chronic
toxicants. As I understand it, neither epidemiology nor test animal
experiments provide a really acceptable solution. The uncertainty
regarding exposure level, the long latency or exposure periods, and the
confusion created by multiple exposures, all diminish the utility of
epidemiology. On the other hand, the problem of translating from test
animal to human response and, perhaps more importantly, the high dose/low
dose extrapolation problem, seriously limit the utility of test animal
experiments. This problem is so serious that Messrs. Hoel, Gaylor,
Kirschstein, Saffiotti and Schneiderman, in a recent article in the first
edition of the Journal of Toxicology and Environmental Health (Hoel et al.
1975) flatly state that:
-------
208
"There is no adequate method for determining the best estimate of
risk for a given dose and the best estimate of dose for a given
risk. Because of model dependency there does not appear to be a
reliable method for obtaining such direct estimates and their
required confidence limits."
I have been given to understand that the National Center for
Toxicological Research was established, at least in part, precisely in
order to illuminate this problem by using very large numbers of test
animals to generate experimentally derived points much lower on the
dose-response curve. I certainly hope my understanding is correct —
and that NCTR is soon successful in this endeavor — for until we can
quantify the human health risks associated with very low levels of
serious chronic toxicants I see no hope of materially improving the
standard-setting process.
REFERENCES
Food and Drug Administration. 1974. Aflatoxins in shelled peanuts and
peanut products used as human foods, proposed tolerance. Federal
Register 39(236): 42748. December 6.
Hoel, David G., David W. Gaylor, Ruth L. Kirschstein, Umberto Saffiotti,
and Harbin A. Schneiderman. 1975. Estimation of risks of irrevers-
ible delayed toxicity. J. Toxicol. Environ. Health 1: 133-151.
National Academy of Sciences. 1975. Decision making for regulating
chemicals in the environment, Washington, D.C.
Nisbet, Robert. 1975. Public opinion versus popular opinion. The
Public Interest 41: 166.
Occupational Safety and Health Administration. 1971a. Occupational
safety and health standards, national concensus standards and
established federal standards. Federal Register 36(105): 10466.
May 29.
Occupational Safety and Health Administration. 1971b. Emergency standard
for exposure to asbestos dust. Federal Register 36(234): 23207.
December 7.
Occupational Safety and Health Administration. 1972. Standard for
exposure to asbestos dust. Federal Register 37(110): 11318.
June 7.
Occupational Safety and Health Administration. 1974. Standard for
exposure to vinyl chloride. Federal Register 39(194): 35890.
October 4.
-------
209
Occupational Safety and Health Administration. 1975. Occupational
exposure to asbestos, notice of proposed rulemaking. Federal
Register 40(197): 47652. October 9.
U.S. Environmental Protection Agency. 1972. Industrial pollution of
the lower Mississippi River in Louisiana. Region VI, Dallas, Texas.
April.
U.S. Environmental Protection Agency. 1973. National emissions standards
for hazardous air pollutants — asbestos, beryllium, and mercury.
Federal Register 38(66): 8820. April 6.
U.S. Environmental Protection Agency. 1974. New Orleans area water
supply study. Lower Mississippi River Facility, Slidell, Louisiana
and Region VI, Dallas, Texas. November.
U.S. Environmental Protection Agency. 1975a. National organics
reconnaissance survey. Office of Research and Development,
Cincinnati, Ohio. April.
U.S. Environmental Protection Agency. 1975b. Assessment of health risk
from organics in drinking water, p. ix. Science Advisory Board.
April.
-------
SESSION II. BIOMEDICAL EFFECTS OF CHLORO-ORGANICS
Robert B. Gumming, Session Chairman
Biology Division
Oak Ridge National Laboratory
Yesterday's sessions of this conference to a large degree dealt
with chlorination technology and the science involved understanding the
chemistry of the chlorination process. We perhaps gained some apprecia-
tion particularly in Dr. White's talk of the tremendous economic and
x"
social stake that this society has in chlorine and in chlorination. The
data are probably available to calculate in purely economic terms the
costs and benefits of water chlorination. If one neglects the difficult
areas of the impact on man himself and upon the environment in which
he must live, there is a clear benefit. The benefit from chlorine vastly
outweighs the cost.
Today, it is time to take a look at what we know and what we can
know about the other side of the ledger. With what precision can we add
the costs to human health and the costs of the potential environmental
degradation into the cost benefit calculation. It is clear that chlorina-
tion of water does add to the load of manmade chemicals which humans
encounter in their environment.
This morning we are going to focus on the very difficult area of the
biomedical effects. What we are really dealing with is exposure of a
large population to extremely low concentrations of a variety of poten-
tially biologically active compounds. The nature of the problem is such
that we can almost immediately eliminate from our consideration acute
toxicity. We do have to consider two areas of toxicity which are
particularly difficult to handle. One is carcinogenesis and the other
211
-------
212
is mutagenesis. This morning we have assembled a group of people who
can approach these toxicological endpoints from several different
angles. Two methods which will be explored, and perhaps they're not
the only two, are extrapolation of data from experimental animal models
and secondly the epidemiological approach.
-------
HALOGENATED ORGANICS IN TAP WATER:
A TOXICOLOGICAL EVALUATION
Robert G. Tardiff
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Gary P. Carlson
Department of Pharmacology and Toxicology
Purdue University
West Lafayette, .Indiana 47907
Vincent Simmon
Microbial Genetics Program
Stanford Research Institute
Menlo Park, California 94025
ABSTRACT
Evaluation of hazard to halogenated hydrocarbons in drinking water
requires the consideration of many factors including degree of exposure,
intrinsic toxicity of the agents, interactions among compounds and with
other environmental factors, and species sensitivity.
Halogenated hydrocarbons are ubiquitous in the public water supplies
of the United States. Of the compounds that have been identified in
the nation's drinking water, approximately 34% are halogenated. A recent
survey of five U.S. cities revealed that approximately 50% of the vola-
tiles in tap water are halogenated. Although the number of volatile
halogenated organics is relatively large, they constitute only a small
percent of the total organic concentration in drinking water. Generally,
chloroform is present in the highest concentration (approximately
100 ppb), and the maj ority of the other compounds are present in 1 ppb
or less.
Assessment of the toxicity of the organics in tap water is following
two lines of investigation. Mixtures of organics from tap water are
being bioassayed for mutagenic, carcinogenic, and teratologic activity.
Specific compounds present in tap water are being subjected to in-depth
213
-------
214
toxicity evaluations. Among the compounds being investigated are the
chloro-ethers, the chlorobenzenes, and the bromobenzenes. Compounds of
interest for future research include the halomethanes and the chloro-
ethanes. Toxicologic questions being addressed using these compounds
include questions of extrapolation from experimental animals to man
through investigations of comparative metabolism and predictability of
synergistic interactions through studies of alterations of basic meta-
bolic pathways.
Publicity of the past 12 months has focused attention on the pres-
ence in drinking water of some of the more toxic organic compounds. In
more recent months, research findings indicated that halogenated organic
compounds having carcinogenic properties were actually synthesized during
chlorine disinfection — a fact that has given way to claims that chlorina-
tion of drinking water may pose a cancer threat or hazard.
That experience reiterates the frequently observed misconception
that toxicity alone is equated with hazard or more specifically that a
toxic property such as experimental carcinogenesis is, of necessity,
identical with a threat to human health and a human cancer risk. Evalua-
tion of safety and hazard, the subject of this discussion, is the corner-
stone to understanding the impact on human health of chemicals in our
surroundings.
Evaulation of safety involves the conceptual integration and inter-
pretation of the physical, chemical, and biological properties or effects
of a particular compound or product for the purpose of assessing the
safety of that agent under conditions of its intended use.
The evaluation of safety and hazard of halogenated hydrocarbons in
drinking water requires consideration of many factors including the
degree of exposure (including both concentration and duration), intrinsic
toxicity of the agents, interactions among compounds as well as between
compounds and other environmental stimuli, and sensitivity of the exposed
population and its subsets.
Halogenated hydrocarbons are ubiquitous in the water supplies of the
United States (Symons et al. 1975). Of the compounds that have been
identified in the nation's drinking water, approximately 38% or 111 of
289 compounds are halogenated as noted in Table 1 (U.S. Environmental
Protection Agency 1976). A recent survey of five U.S. cities (U.S.
-------
215
Table 1. Occurrence of organic compounds in
tap water of the United States
Type
Halogenated
Nonhalogenated
Total
Aliphatic
Aromatic
Total
Number
111
178
289
190
99
289
Percent
38
62
66
44
Environmental Protection Agency 1975) revealed that approximately
50% of the volatile hydrophobic compounds in tap water were halogenated.
The data in Table 1 also categorized the organic compounds in drinking
water according to structure, thus it can be seen that 66% of the com-
pounds are aliphatic and the remainder aromatic in structure. Although
the number of volatile halogenated compounds is relatively large, these
agents constitute only a small fraction by weight of the total organic
concentration in tap water. The term "volatile" refers to those
compounds that are purged from water at 95°C with an inert gas (e.g.,
helium). Examples of these compounds include the halomethanes such as
chloroform and bromoform and the chloroethenes such as vinyl chloride
and vinylidine chloride. The "nonvolatile" fraction, by contrast, is
operationally defined as that mixture of compounds selectively extracted
from water by reverse osmosis (cellulose acetate and nylon membranes).
Generally, one volatile compound, chloroform, is present in the highest
concentration — approximately 100 ppb (pg/1) on the average — and the
majority of the other compounds are present at 1 ppb or less.
Recently, our laboratory initiated the analysis of the nonvolatile
fraction of one sample of organics from tap water. Preliminary data
indicate that the molecular weight of the compounds in this fraction
vary from 200 to 500 as contrasted to the predominantly lower molecular
weight species in the volatile fraction (i.e., less than 100 MW).
Approximately 200 compounds have been separated in the nonvolatile
-------
216
fraction, only a few of which have been identified. Several of the
approximately 200 compounds appear to contain from 4 to 10 halogen atoms
each, in contrast to the occurrence of between 1 and 6 halogen atoms in
the compounds of the volatile fraction. Work is proceeding to identify
the structures of the nonvolatile agents by mass spectrometry.
Since dosage is a primary factor in the evaluation of safety and
hazard, the concentrations to which man is, and can be, exposed requires
definition. Based on general parameters such as total organic carbon
(TOG), the concentrations of organics in tap water range from less than
1 mg/1 (i.e., parts per million) to at least 6 mg/1. Assuming that
carbon comprises on the average 50% of the total weight of the compound,
then the concentrations of the mixtures would range from 1 ppm to 12 ppm,
respectively. In view of the relatively large numbers of these compounds
in these mixtures, individual compounds are expected to be present in
microgram-per-liter (ppb) quantities. Quantitative analyses of volatiles
in tap water are in agreement with this hypothesis. Although those
doses are seemingly minute, exposure to these agents via drinking water
is chronic or continuous in nature; whereas, other forms of exposure are
believed to be intermittent. Because of the length of the half-life
of these chlorinated compounds (usually longer than 12 to 24 hours), it
is highly likely that continued exposure (not only on a daily basis but
throughout the day) may lead to a bioaccumulation of these compounds.
Consequently, the relationship between the equilibration concentrations
and the critical levels in soft tissues necessary to initiate toxic
responses must be determined. An additional modifying factor is the fact
that exposure occurs during certain life stages which may be more criti-
cal than others with regard to susceptibility to the toxic properties of
these chemical agents (i.e., increased sensitivity). One example of the
effect of dosage regimen was recently noted with chloroform, in which
the target organ changed when going from low to high exposures. Acute
toxicity was observed in the kidney after low dose exposure; whereas
higher dose exposure led initially to kidney pathology followed by
hepatic centrolobular necrosis (Hill et al. 1975).
The toxic properties of the compounds can be defined by employing
a battery of bioassays. The array of assays ranges from organ function
-------
217
test, to histopathologic evaluation, to tissue enzyme assays, to deter-
minations of molecular alterations of genetic material. No single
battery of tests is guaranteed of predicting all possible pathologic
lesions in the whole animal. Exhaustive testing with all conceivable
test systems would be economically prohibitive and wasteful. Neverthe-
less, there is a real need to obtain sound and relatively comprehensive
toxicity data on the compounds to which man is exposed (as for example
in his drinking water).
The evaluation of the toxicity of organic compounds, including
halogenated hydrocarbons in drinking water, follows two avenues of
investigation. First, an attempt is made to screen the toxic potential
of mixtures of organic compounds that are actually obtained from municipal
drinking water. Second, individual compounds are subjected to in-depth
toxicological evaluation in animal models that are predictive of human
responses. Although not a part of this discussion, note must be taken
of the fact that epidemiologic investigations are also pursued and that
both the toxicologic and epidemiologic work are complimentary to one
another.
For the purposes of toxicological assessment as well as chemical
characterization, samples of organic mixtures can be concentrated and
recovered from tap water through the use of reverse osmosis technology
(cellulose acetate and nylon membranes) (Kopfler et al. 1975). Such
procedures require the separation of inorganic salts from organic com-
pounds in order to avoid any additional injury due simply to high salt
concentrations. The utilization of solvent extraction and certain forms
of dialysis can be used effectively to generate samples of organic
mixtures. At present, recovery of organics from tap water is approxi-
mately 30 to 40% by utilizing these techniques.
A program has been developed to ascertain the types of water
supplies which most likely present some health hazard to the consumer.
The program requires the sampling and bioassay of six samples or organic
mixtures from water supplies representing each of the major sources of
raw water. Table 2 lists the cities that have been selected for this
study, the type of water supply utilized by each city, and the type of
-------
218
Table 2. Cities selected for bioassay of
organics in drinking water
Type of Type of
City water supply raw water
Cincinnati, OH Surface Industrial wastes
Miami, FL Ground Uncontaminateda
New Orleans, LA Surface Industrial wastes
Ottumwa, 10 Surface Agricultural runoff
Philadelphia, PA Surface Municipal wastes
Seattle, WA Surface Uncontaminateda
Refers to no known contamination from municipal, agricul-
tural, and industrial wastes — but contamination pre-
sumably from decomposition products of natural origin.
raw water used in generating drinking water. Consequently, a matrix of
both ground and surface sources as well as waters containing predomin-
antly industrial wastes, agricultural runoff, or municipal wastes are
also being separately sampled and assayed along with presumably uncon-
taminated ground and surface sources. One duplication is immediately
obvious: the inclusion of two cities both having surface sources
containing predominantly industrial wastes. Cincinnati was originally
selected as the prototype city, the one in which all of the analytical
developments and modifications would be made. Subsequently, the drinking
water from the City of New Orleans was selected to obtain additional
information for use with standardized methodology, as were Miami,
Ottumwa, Philadelphia and Seattle.
The screening of organic mixtures or concentrates from tap water
for toxicologic activity is summarized in Table 3. With the exception
of the range finding assay, the assays attempt to screen for irreversible
toxic phenomena. Such endpoints include mutagenesis, teratogenesis,
carcinogenesis and other forms of chronic intoxication. Measurement of
mutagenic potential is performed by the histidine reverse mutation system
in five strains of Salmonella typhimuriwn (TA1535, TA1537, TA1538, TA98,
and TA100) (Ames et al. 1973, McCane et al. 1975).
-------
219
Table 3. Toxicologic screens of organic concentrates
from tap water
Assay
Sample/city at 2 month intervals
12345 6
Range finding
(LD50 Mouse)
Mutagenesis
(Salmonella)
Mammalian, cell
trans format ion
In vivo
carcinogen bioassay
(Neonate)
Teratogen
assay (rat)
Chemical
characterization
(GC/MS)
x-f
x-f
x-f
x-f
x-i
x = Analysis of sample
x-f = Analysis of sample and subtractions
? • Possible analysis
Because of the relatively small amount of material required for the
in vitro assays, it is entirely possible that an active parent mixture
can be subfractionated with various solvents and other techniques to
yield fractions for further bioassay. Consequently, the active ingre-
dients can be narrowed down in specific fractions which can be analyzed
chemically at a later time.
Initial results using the aforementioned mixtures from the City of
Cincinnati's tap water have yielded data which indicate that the two
mixtures used with TA100 were both mutagenic (Simmon and Tardiff 1976).
Subfractionation of these parent mixtures sequentially with petroleum
ether, diethyl ether and acetone yielded results indicating that the
mutagenic components were located predominantly in the diethyl ether
fraction. Chemical characterization of this fraction is currently
ongoing.
-------
220
As a measure of possible carcinogenic activity, an in vitro mammalian
cellular transformation assay (3T3) is being utilized to determine the
ability of the mixtures to produce malignant transformations in culture
(DiPaolo et al. 1972). The malignancy of transformed cells is confirmed
by the development of carcinoma in whole animals injected with trans-
formed cells. Both the mutagenesis assay in bacteria and the transfor-
mation assay in mammalian cells are presumed to be indictors of a high
degree of suspicion of carcinogenesis. An additional carcinogen/chronic
toxicity bioassay is also utilized. Several studies have indicated that
the neonatal animal is far more susceptible to a chemical carcinogen
than is the adult animal (Kelly and O'Gara 1961, O'Gara et al. 1965,
Pietra et al. 1959). Consequently, such a model using a neonatal rat
is being employed, first, to determine whether the mixtures have carcino-
genic activity and, second, to ascertain whether other endpoints of
irreversible chronic intoxication with subchronic exposures can be
observed. The assay includes the oral administration of the test mixtures
to rats from day 1 to day 21 after birth with subsequent observations
throughout the majority of the animals1 lifespan and terminal histopatho-
logic examination. Additionally, the mixtures are assayed for teratogenic
potential by administration of the materials to pregnant rats during the
period of organogenesis with subsequent gross morphologic observations
of the progeny at birth.
If the outcome of any or all of the screening tests is positive, an
archival sample of the mixture is analyzed for its chemical composition
in order to obtain an indication of the compounds which may be responsible
for the observed effects. Suspect agents are then retested in the
appropriate bioscreens. Separation of the individual components is
performed with glass capillary columns in a gas chromatograph; and iden-
tifications are confirmed by mass spectrometry.
The determination that certain mixtures of organic compounds derived
from drinking water have the potential for irreversible toxicity inevit-
ably leads to a list of compounds for which in-depth evaluation of safety
and hazard must be conducted. In addition to the stated criteria, com-
pounds can be selected for safety/hazard evaluation on the basis of
-------
221
predominant occurrence in potable waters, as in the case of the haloforms.
The primary objectives of the toxicology program on halogenated organics
in drinking water are listed in Table 4. The identification of patho-
logic lesions and the identification of target organs in progressive
sensitivity leads to experimental designs aimed at defining the dose-
response relationships as well as to the definition of those pathologic
phenomena which are likely to occur in man after exposure. The elucida-
tion of the pathogenesis of subchronic and chronic intoxication may
yield valuable clues to enable epidemiologists to conduct informative
clinical and subclinical investigations of possible effects in the human
populations. Investigations of toxicokinetic properties of the agent
may lead to accurate predictions of accumulation and of possible toxic
metabolites and may yield valuable hints to determine possible inter-
actions with other environmental factors. Synergism or antagonism as
related to the exposure to a specific hydrocarbon or class of hydrocarbons
is approached from the point of view of comparative metabolism as well
as from the sytematic evaluation of various environmental factors
including host susceptibility.
Table 4. Primary objectives of toxicology program on
halogenated organics in drinking water for
extrapolation to man
1. Definition of pathologic lesions and target organs
2. Determination of dose-response relationships
3. Identification of pathogenesis
4. Elucidation of toxicokinetics
5. Investigation of factors influencing toxic manifestations
(i.e., interactions)
6. Elucidation of differential response and sensitivity
The classes of chlorinated compounds under active investigation for
their toxicologic activity are listed in Table 5. The three main classes
include the chloroethers which are chemical congeners of the human car-
cinogen bis-chloromethyl ether, the halobenzenes which are known to occur
in potable waters, and the haloforms which are synthetic byproducts of
disinfection of tap water with chlorine.
-------
222
Table 5. Classes of chlorinated compounds in drinking water
under toxicological investigation
I. Chloro-e thers
1. Bis(2-chloroethyl) ether
2. Bis(2-chloroisopropyl) ether
II. Halobenzenes
1. Chlorobenzenes
2. Bromob enzenes
3. Chloro—bromo-benzenes
III. Halof orms
1. Chloroform
2. Bromodichloromethane
3. Chlorodibromomethane
4. Bromoform
One of the greater concerns associated with the organic contaminants
in drinking water is the possibility that the population genome may be
effected such that adverse effects may be seen in future generations.
Consequently, it is of interest to determine which of the compounds in
tap water may be a mutagenic hazard to man. As indicated above almost
300 organic compounds have already been identified from tap waters from
the United States. Of this number, approximately one third have been
spot tested for mutagenic activity in Salmonella typhirrwiim TA100; and
twelve of these compounds have demonstrated mutagenic activity in this
strain which is a measure of base-pair substitutions with some overlap to
frameshift mutations. The compounds that demonstrated mutagenic activity
in this test system are listed in Table 6.
The extrapolation of experimental data from animal to man is one of
the most critical factors in the evaluation of the safety and hazard of
chemicals to man by the use of experimental animals. Three main param-
eters impact on this problem: (a) the biotransformation of the compound
with respect to both activation and detoxification of the chemical,
(b) the toxicokinetics of the agent, and (c) the sensitivity of the
target organism (man vs experimental animal). The biotransformation of
compounds has been under active investigation since it was originally
learned that organophosphate pesticides were initially activated to the
-------
223
Table 6. Compounds tested and found to be mutagenic
in S. typhimwiian TA100a
Bromoform Hexachloro-l,3-butadiene
Chlordane (technical grade)^ Isopropylbenzene
1,2-bis-Chloroethoxyethane Methylchloride
m-Chloronitrobenzene Methylbromide
1,2-Dichloroethane Nitroanisole
1,1-Dichloroethylene Vinyl chloride
Base-pair substitutions with overlap to frameshift
mutations.
Pure Chlordane is negative in same system.
toxic form prior to detoxification (Brooks 1972). Subsequent studies
demonstrated that the activity of chemical carcinogens often follows the
same route through oxidative pathways in which epoxides may be formed
(Arcos and Argus 1974). Consequently, if man metabolizes an agent or
several agents in such a fashion that the pathway is through the
activation-inactivation process, it is of practical interest to find a
species which metabolizes the compound in the same manner. Thus, the
active chemical form responsible for the pathologic lesions is present
in the experimental animals as well as in man. Not only is it important
to have the same metabolic pathways in the experimental animal as in man;
but the rates of reactions for activation and inactivation (detoxication)
should also be as similar as possible. One of our research programs is
currently focusing on this problem by attempting to study the metabolism
of specific chlorinated hydrocarbons in seven species of experimental
animals, both in vivo and in vitro, and in man in vitro. The test com-
pounds utilized in these investigations are principally the chlorinated
ethers; however, future plans involve the measurement of the metabolism
of the haloforms in experimental animals and possibly in man in vivo.
The species utilized in the comparative metabolism program are listed in
Table 7. The experimental species were selected on the basis of their
reasonable accessibility as well as on the basis of their extensive
utilization of toxicity experimentation. Confirmation that one or
-------
224
Table 7. Species employed in comparative
metabolism and toxicokinetics
Mouse (Carworth GDI)
Rat (Charles River CD)
Hamster (Syrian Golden)
Guinea Pig (Hartley)
Rabbit (NZW)
Dog (Beagle)
Monkey
Man
several species of experimental animals metabolizes the subject com-
pounds similarly to man will lead to formulation of protocols for
in-depth chronic toxicity investigations in the predictive animal models.
While interactions are being investigated indirectly in the studies
utilizing mixtures from organic materials taken from drinking water,
more extensive and systematic investigations with known compounds are
also being conducted. Of the many approaches that can be employed to
study toxic interactions, our laboratory has selected the mechanism by
which one compound alters the metabolism of other compounds or classes of
compounds. It has been demonstrated, for example, that compounds in the
organbphosphate class can have their toxicity potentiated by an altera-
tion of the biochemical processes that metabolize the organophosphates
(Murphy and DuBois 1957). Additional work has demonstrated that metabolic
alterations influencing toxicity are also observed with the chlorinated
hydrocarbon insecticides (Deichmann and Keplinger 1970) and the poly-
nuclear aromatic hydrocarbon contaminants (Uchiyama et al. 1974). For
this work, the homologous series of chlorobenzenes has been selected to
determine their influence upon the metabolism of various classes of
xenobiotics (Carlson and Tardiff, in press). Relatively little is known
about the simple halogenated benzenes despite their widespread use.
Chlorobenzene, p-dichlorobenzene, l-bromo-4-chlorobenzene, 1,2,4-
trichlorobenzene, and hexachlorobenzene were administered orally for
14 days to adult male rats. The parameters measured included hexabarbital
-------
225
sleeping time, cytochrome Q reductase, cytochrome P-450, EPN detoxication,
glucuronyltransferase, benzpyrene hydroxylase, and azoreductase. The
results indicated that all of the compounds with the exception of chloro-
benzene decreased hexabarbital sleeping time immediately following and/or
14 days after treatment, at doses of 600 to 2,000 mg/kgm/day. At levels
ranging from 10 to 40 mg/kgm/day, much lower doses than those necessary
to cause hepatotoxicity or mortality, all but chlorobenzene caused alter-
ations in the various parameters measured. Administration of p-dichloro-
benzene or 1,2,4-trichlorobenzene at these levels for 90 days resulted in
increases in EPN detoxification, benzpyrene hydroxylase, and azoreductase.
Despite cessation of administration of the compound for as much as 30
days, the EPN detoxification and benzpyrene hydroxylase were still altered
in animals having been exposed subchronically to either p-dichlorobenzene
or 1,2,4-trichlorobenzene. The more potent inducer of these metabolic
pathways and electron transfer components was judged to be 1,2,4-
trichlorobenzene. It is concluded that even simple chlorinated benzenes
can induce the metabolism of foreign organic compounds with effects
continuing substantially beyond the period of exposure. Thus, some
compounds in this class of chlorobenzenes may significantly compromise
the ability of the organism to properly respond to the ambient levels
of other exogenous compounds.
In summary, an attempt has been made to describe the program aimed
at the toxicologic definition of the chlorinated hydrocarbons in drinking
water. Two specific approaches were mentioned: (a) that dealing with
bioscreen of mixtures or organic compounds for mutagenesis, chronic
toxicity/carcinogenesis, and teratogenesis; and (b) that dealing with
specific chlorinated hydrocarbons or classes of these compounds with
specific emphasis on comparative metabolism for prediction of human
responses and on interactions for predictions of synergism and antagonism.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the technical assistance of
W. Emile Coleman and Judith L. Mullaney and the clerical assistance of
Shirley A. Tenhover.
-------
226
REFERENCES
Ames, B. N., F. D. Lee, and W. E. Durston. 1973. An improved bacterial
test system for the detection and classification of mutagens and
carcinogens. Proc. Nat. Acad. Sci. 70: 782-86.
Arcos, Joseph C., and Mary F. Argus. 1974. Chemical induction of cancer:
structural bases and biological mechanisms. Academic Press. New
York. P. 135-83.
Brooks, G. T. 1976. Pathways of enzymatic degradation of pesticides,
p. 106. In F. Coulston and F. Korte (ed.) Environmental quality
and safety: global aspects of chemistry, toxicology and technology
as applied to the environment. Volume 1. Academic Press. New York.
Carlson, Gary P., and Robert G. Tardiff. Effect of chlorinated benzenes
on the metabolism of foreign compounds. Toxicol. Appl. Pharmacol.
In Press.
Deichmann, W. B., and M. L. Keplinger. 1970. Protection against acute
effects of certain pesticides by pretreatment with aldrin, dieldrin
and DDT, p. 121-23. In Pesticides symposia. Halos and Associates,
Inc. Miami, Florida.
DiPaolo, J. A., K. Takano, and N. C. Popescu. 1972. Quantitation of
chemically induced neoplastic transformation of BALB/3T3 cloned
cell lines. Cancer Res. 32: 2686-95.
Hill, Richard N., Thomas L. Clemens, Dai Kee Liu, and Elliot S. Vesell.
1975. Genetic control of chloroform toxicity in mice. Science
190: 159-161.
Kelly, Margaret G., and O'Gara, Roger W. 1961. Induction of tumors in
newborn mice with dibenz(a,h)anthracene and 3-methyIcholanthrene.
J. Nat. Cancer Inst. 26(3): 651-73.
Kopfler, F. C., R. G. Melton, J. L. Mullaney, and R. G. Tardiff. 1975.
Human exposure to water pollutants. Presented at the Division of
Environmental Chemistry Meeting, American Chemical Society.
Philadelphia, Pennsylvania. April 6-11.
McCann, J., N. E. Spingarn, J. Kobori, and B. N. Ames. 1975. Detection
of carcinogens as mutagens: bacterial tester strains with R Factor
plasmids. Proc. Nat. Acad. Sci. 72: 979-83.
Murphy, Sheldon D., and Kenneth P. DuBois. 1957. Quantitative measure-
ment of inhibition of the enzymatic detoxification of malathion by
EPN. Proc. Soc. Exp. Biol. Med. 96: 813-18.
-------
227
O'Gara, Roger W., Margaret G. Kelly, Jewel Brown, and Nathan Mantel.
1965. Induction of tumors in mice given a minute single dose of
dibenz (a) anthracene or 3-methylcholanthrene as newborns. A dose-
response study. J. Nat. Cancer Inst. 35: 1027-42.
Pietra, Giuseppe, Kathyrne Spencer, and Philippe Shubik. 1959.
Response of newly born mice to a chemical carcinogen. Nature
183(4676): 1689.
Simmon, Vincent F., and Robert G. Tardiff. 1976. Mutagenic activity of
drinking water concentrates. Presented at the Annual Meeting of
the Environmental Mutagen Society. Atlanta, Georgia. March 11-15.
Symons, James M., Thomas A. Bellar, J. Keith Carswell, Jack DeMarco,
Kenneth L. Kropp, Gordon G. Robeck, Dennis R. Seeger, Clois J. Slocum,
Bradford L. Smith, and Alan A. Stevens. 1975. National organics
reconnaissance survey for halogenated organics in drinking water.
Water Supply Research Laboratory, U.S. Environmental Protection
Agency, Cincinnati, Ohio. April. (Prepublication Print).
Uchiyama, Mitsuri, Takako Chiba, and Kiichiro Noda. 1974. Cocarcinogenic
effects of DDT and PCB feedings on methylcholanthrene induced chem-
ical carcinogenesis. Bull. Environ. Contain. Toxicol. 12(6): 687-93.
U.S. Environmental Protection Agency. 1975. National organics recon-
naissance survey: analysis of tap water from five U.S. cities for
volatile organic compounds, a staff report. Health Effects Research
Laboratory, Cincinnati, Ohio.
U.S. Environmental Protection Agency. 1976. List of organic compounds
identified in drinking water in the United States, January 1.
Health Effects Research Laboratory, Cincinnati, Ohio.
DISCUSSION
Max Eisenberg, Maryland State Dept. of Health. Are you doing any
epidemiological studies? In particular, have you considered following
a specific fraction of the population at risk, such as pregnant women,
to observe any anatomical malformations?
Tardiff. I'll answer your question very briefly by saying that,
yes, our program is involved in epidemiologic studies. They will be
discussed by Dr. Cantor later in the program.
-------
ORIGIN, CLASSIFICATION AND DISTRIBUTION OF CHEMICALS IN DRINKING
WATER WITH AN ASSESSMENT OF THEIR CARCINOGENIC POTENTIAL
Herman F. Kraybill
National Cancer Institute
Bethesda, Maryland 20014
ABSTRACT
Of the wide array of chemical contaminants identified in potable
waters some have carcinogenic activity referenced to studies with
experimental animals. Some appear to be universally distributed both
nationally and internationally. Some carcinogenic chemicals may be
traced back to point source contamination while others may be formed
or magnified to levels above those in raw water supplies during the
chlorination process. Some carcinogenic chemicals fall into use classes
such as pesticides, industrial chemicals, drugs, and other categories.
Not all these chemicals classified as having carcinogenic potential or
activity can be assessed as to their equivalent hazard. Differentiation
is necessary to identify those that are well recogni-ed as classical
carcinogens and those that are of equivocal nature wuan referenced to
experimental animal studies and may thus be termed suspect carcinogens*
Some chemicals may be characterized as potential carcinogens on the
basis of structural relationships or ancillary studies on mutagenicity.
Many chemicals remain to be characterized for their carcinogenic activity,
The integrated insult from multiple carcinogens in the water supply may
have additive or inhibitory properties. This aspect of the problem
remains to be qualified or quantified in terms of human risk assessment.
INTRODUCTION
The interaction of man with his environment and the health conse-
quences therefrom are well recognized, but not fully comprehended.
Environmental stresses of biological, physical, and chemical nature may
entail both benefits and risks. More and more attention is being given
to assess these stresses in a benefit-risk equation. The spectrum of
biomedical responses to environmental stresses impose a cytotoxic effect
229
-------
230
which includes carcinogenic, mutagenic, or teratogenic activities. Our
concern in this presentation is oriented toward carcinogenic or neo-
plastic events.
Environmental cancer in its fullest context implies not the
examination of singular insults as a traditional approach, but rather
examination of the multiple insults that impact on man from air and
water pollutants, diet contaminants, drugs, and occupational exposures.
Water contaminants, in the form of inorganic or organic carcinogens,
add to the total carcinogenic environmental load encountered from all
stresses.
Relevant to hazard, carcinogenic or noncarcinogenic as components
of the overall toxic stress, there is not now and probably never will be
an ideal state of absolute safety or zero exposure. Cytotoxic agents
abound in nature from plants, bacteria, and geologic origin. These
environmental stressors on a global basis are omnipresent, as background
below the levels added by man through his technological developments.
Thus, conceptually, even if one could biologically and pragmatically
estimate a safe dose or threshold for a carcinogen, one should view this
problem in terms of a potentially "added risk;" that is, above the risk
that would prevail naturally. Boyland (1969) believes that 90% of all
cancers are due to chemicals, and Higginson (1969) maintains that 90%
of cancers are "theoretically preventable" because they may be environ-
mentally related. These views are shared by Epstein (1970).
With approximately four million chemicals already in the universe,
and the introduction of many new ones into the environment, it becomes
prohibitively expensive, if not logistically impracticable, to prove
out the safety or risk of each compound (Lewin 1974). Thus, one sees the
necessity to have new methods for rapid evaluation of the carcinogenic
potential of whole classes of compounds.
Experimental studies in animals are at best only presumptive
measures for assessment of a carcinogenic risk in man. Obviously,
epidemiologic pursuits are of high validity; but, it is socially,
legally, and medically unacceptable to subject man to tests unless he
is inadvertently receiving such challenges, as is the case with water
contaminants. However, animal studies are used for predictive purposes.
-------
231
The determination of carcinogenic and/or tumorigenic risk to chemicals
at ambient concentrations is a difficult task. One can never feel
secure as to whether or not a chemical is carcinogenic in man, since an
animal system may reveal false positives or false negatives. Indeed,
there is a variable response in man which leads one to ask "to what man
is this effect relevant?" Thus, the perfect model will probably never
be attained.
Beyond the considerations of species variation is the evaluation of
the potency of the chemical for tumor induction and the time frame for
tumor induction. The question arises whether very low levels, on a
dose-response relationship for a defined population, would require over
100, 200, or 300 years to induce tumors; which, of course, is beyond
man's life span.
Carcinogenic contaminants in the.drinking water supply may originate
from various sources. Since many municipal water supplies are derived
from rivers, lakes, streams, or ground waters, they carry pollution
from agricultural runoff, industrial effluents, and accidental or
deliberate dumping. Some organic chemicals are formed from treatment
processes such as chlorination. Many of the carcinogens are in tap
water not from treatment or processing alone but because of failure to
adequately remove them from raw water supplies. Carbon tetrachloride,
or chloroform, does not need to be formed by chlorination if the rivers
carry these chemicals originally.
For purposes of this presentation, the matter of carcinogens in
water will be referenced to those chemicals identified in drinking water.
An analysis of a list of 221 organic chemicals provided by the Water
Supply Research Laboratory of the Environmental Protection Agency in
Cincinnati, Ohio, on June 1, 1975, is used in the categorization of
carcinogens. Since this listing, a revised list with 14 more chemicals
was provided on September 1 (total 235). This listing was generated
from an exhaustive search of the literature and from reports on analysis
of some municipal water supplies. The listing is, therefore, not all-
inclusive and, on a quantitative basis, probably reflects only a few
percent of the total organics in the water (Tardiff et al. 1975).
-------
232
As a result of concern in 1974 over certain organic chemicals in
drinking water (halomethanes), the Environmental Protection Agency
instituted the National Organics Reconnaissance Survey. This survey
of eighty cities had three major objectives. One of these was to deal
with the trihalomethanes; the second was to ascertain the effect of
water treatment processes on the formation of these chemicals; and, the
third was to characterize and possibly quantify the organic content of
finished water (Symons 1975).
GENERAL CLASSIFICATION OF ORGANIC CHEMICALS IN DRINKING WATER:
REFERENCES TO TECHNOLOGICAL USE
The selection process of chemicals for carcinogenic activity assess-
ment by animal bioassay is based on the uses of that chemical, probable
exposures, structure activity relationships, and available toxicity
data. These criteria provide a priority scheme for investigating those
chemicals which may impose the maximum hazard to man in his environment.
The production estimate and extent of probable exposure are most signifi-
cant. Accordingly, classifying these chemicals according to their use
pattern is most important. In Table 1, 221 organic compounds identified
in drinking water as of June 1975, have been grouped by classes, and the
number in each class and the percentage are shown.
Table 1. Representation of organic chemicals in drinking water
by technological usea
Technological use
Industrial chemicals
Pesticides
Food chemicals — derivatives
Drugs — perfumes
Decomposition products
Natural product or toxin
Laboratory chemicals
Tobacco products
Miscellaneous or unknown
Total
Number of chemicals
in use class
100
28
16
13
10
9
3
1
41
"221
Percent of total
in class
45.2
12.6
7.3
5.8
4.5
4.2
1.3
0.4
18.5
100.0
aFor reference purposes, chemicals for bioassay origin by use patterns is
important. Chemicals could fall into several classes — major use
adopted (Furia and Bellanca 1971, Merck Index 1960).
-------
233
Because of uses in industry, agriculture and by consumers as food
chemicals, drugs, and for other applications, these chemicals may be
introduced into water supplies which are the sources of municipal tap
water.
CRITERIA USED IN CLASSIFICATION OF CARCINOGENS
Attempts have been made to categorize carcinogens under such terms
as potent, weak, recognized, suspect, or potential on the basis of
toxicological studies. Some prefer not to classify and simply indicate
that a carcinogen is a carcinogen, but such a rigid viewpoint provides
no basis for risk assessment when considering the host of environmental
agents which man encounters in his life. Potent carcinogens such as
aflatoxin (Kraybill and Shimkin 1964) and vinyl chloride (Maltoni and
Lefemine 1974) produce in a variety of experimental animals a high inci-
dence of cancers, even at low levels in a diet or in the environment.
Some of the potent experimentally-proven (animals) carcinogens, such as
benzidine, bischloromethyl ether, betanaphthylamine and vinyl chloride,
have been well established, epidemiologically, as occupational carcinogens
in man. A precedent has been established for classification of carcino-
gens by listing chemicals and mixtures that cause cancer in man by
direct observations on exposed populations (National Cancer Institute
1975a). There are 32 substances on that list.
Professor Maltoni (1973) in discussing occupational carcinogenesis
classified occupational carcinogens into three groups: (a) definite
carcinogens, (b) suspect carcinogens, and, (c) potential carcinogens.
Goldsmith (1975) also developed a taxonomic approach to environmental
carcinogenesis and provided, from epidemiological experience and analysis,
eight classes (including such descriptors as likely and possible) for
development of a decision scheme for the different classes that would
help in regulatory interpretation. Kraybill (1974) also developed a
classification scheme for carcinogens in water which, in 1975, was
adopted in a classification used by the Water Supply Research Laboratory
of the Environmental Protection Agency. This classification scheme
and criteria are presented in Table 2.
-------
234
Table 2. Criteria for development of classification
of aquatic carcinogens
I. Recognized and classical
A. From epidemiological observations
B. Response in many species, strains, and repeatedly confirmed
C. Tumors of rare type, tumor incidence increase in short time,
exposure route relevant, tumors transplantable
II. Suspect
A. Response only in one species, strain, or sex
B. Tumor incidence increase with exposure
C. Evidence from in vitro studies
III. Potential
A. Suggestive evidence, basis of structure activity
B. Mutagenicity data
IV. Promoters
A. Will increase tumor incidence of known carcinogens
V. Fragmentary data — inadequate tests
A. Inadequate number of test animals, observation time,
nonrelevant exposure, improper species and strain
B. Question of contaminants — cocarcinogens
C. Overt toxicity and metabolic overloading
VI. Noncarcinogenic
A. Negative in exhaustive texting
B. Physiological and cellular constituents
C. Negative findings — human observations — long time
D. Noncarcinogenic when contaminants are removed
Pragmatically, such a classification provides a working scheme for
decision-making processes as to relevant hazard. For example, one would
not develop the same type of guidelines for exposure situations relevant
to vinyl chloride as for DDT, bis(2-chloroethyl)ether, or other chemicals
where experimental evidence is tenuous.
-------
235
Although no formal agreement has been reached on criteria and a
taxonomic classification of the activity of chemicals proclaimed or
suspected as being carcinogenic, Hueper as early as 1961 recognized the
need for such a classification. He believed that the concept of "chemi-
cal group carcinogens" had restricted applicability and would be valid
for only two classes of carcinogens, that is, radioactive chemicals and
chemicals with estrogenic properties (Hueper 1961).
The first delineation of carcinogens by Hueper was by site of action
or target organ. His classification contained four groups: (a) primary
contact point, i.e., skin, lungs, etc.; (b) at site of excretion, i.e.,
bladder or skin; (c) at site of secondary retention and storage, i.e.,
radium, arsenic, thorotrast, etc. and effect on bones, bone marrow, liver,
and skin; and, (d) at site of special affinity, i.e., hormones, and
organotropic, goitrogenic, and estrogenic chemicals, etc.
Another classification was by species specificity. Here, Hueper
uses subclassifications such as recognized, suspect, potential, and
experimental. His explanation of varying biological activity is com-
parable to our previous assessment. In the experimental category, he
lists those chemicals which produce cancers in experimental animals but
which are not important contaminants in the environment yet may be of
scientific significance, for example, 2-acetylaminofluorene (2AAF).
Hueper's final classification is according to carcinogenic potency and
carcinogenic completeness. In the former case, he specifies the varia-
tion in activity in different species, strains, sexes, and specific test
«
conditions. In the latter class, he alludes to multistage mechanisms
for a chemical in the induction of cancer; he discusses the fact that
some chemicals are only "initiators" or "promoters;" and some chemicals
derive their activity at the cell level predominantly as cocarcinogens.
He also lists some chemicals that are "incomplete" carcinogens which are
conditioned by other factors for eventuating in neoplasia.
-------
236
CONSIDERATION AND ASSESSMENT OF THE CARCINOGENIC PROPERTIES
OF ORGANIC CHEMICALS IN WATER
Some lists on the status of organic compounds identified in drinking
water have been published, while others are unpublished. For purposes
of this presentation, we have confined our review mostly to the listing
of organic compounds identified in drinking water in the United States,
made available on June 1, 1975, by the Water Supply Research Laboratory
of the Environmental Protection Agency. This listing, over the months,
has been expanded when new organic chemicals were identified. As men-
tioned before, the June 1975 listing contained 221 organic chemicals;
235 chemicals were listed by September 1975.
The evaluation of carcinogenic activity, whether classified as
recognized, suspect, or potential, is based on reports in the literature
or recent reports on preliminary findings from the National Cancer
Institute Carcinogenesis Program. Many of the chemicals appearing on
lists of biorefractories in drinking water have been suggested for bio-
assay in the National Cancer Institute Bioassay Program. Some of the
chemicals appearing as biorefractories had been on bioassay at the
National Cancer Institute prior to the realization that they were water
contaminants. Data on these latter compounds have only recently become
available.
There are further studies planned on the haloethers, originally
reported by Van Duuren and coworkers (1972) who studied the structure
activity relationships of analogs of bis (chloromethyl) ether. One of
the haloethers appears to be a contaminant in many water supplies in this
country and in Europe (Piet et al. 1973). This haloether, the bis(2-
chloroisopropyl) ether, is currently under bioassay. Haloethers, products
of glycol synthesis in industrial processes, are introduced into water-
ways as industrial effluents. This class of compounds is interesting
in that the carcinogenic activity of the chemical is dependent on struc-
ture; namely, the location of the halogen atom in the alpha or beta
position. The alpha compound appears to be highly active, that is, for
such compounds as the chloroethyl ethers. As the chlorine atom is
positioned further away from the ether linkage, the chemical has less
-------
237
activity. Van Duuren et al. (1972) reported that chloromethyl methyl
ether (CMME), bis(chloromethyl) ether (BCME), and bis(a-chloroethyl)
ether have significant carcinogenic activity.
As previously indicated, the National Organics Reconnaissance Survey
of the Environmental Protect, on Agency had as one assignment the study
of six halogenated o ,anic compounds to determine their presence in
eighty water supplier surveyed. The six chemicals of concern are (1)
chloroform, (2) carbon tetrachloride, (3) bromodichloromethane, (4)
dibromochloromethane, (5) 1,2-dichloroethane, and (6) bromoform. Of
these six, chloroform, carbon tetrachloride and dichloroethane have
carcinogenic activity. The National Cancer Institute data base indicates
that carbon tetrachloride, showing a carcinogenic response in several
species and strains in several laboratories, would be a recognized car-
cinogen. There are some reports on carcinogenic activity in man (Tracey
and Sherlock 1968). Data on chloroform indicates that this chemical may
be suspect as a carcinogen (International Agency for Research on Cancer
1972, p. 61-65; Eschenbrenner and Miller 1945). Chloroform is obviously
of interest in that it appears in raw water supplies, and in all samples
from eighty locations surveyed. The concentration ranged from 0.1 to
311 pg/liter with 50 percent of the finished waters containing 25 ug/liter
of chloroform or less (Environmental Protection Agency 1975). While
chloroform appears in raw water supplies (range of 0.1 to 0.9 yg/liter),
its concentration was considerably higher in treated (chlorinated) or
finished water (Bellar and Lichtenberg 1974). Since chloroform is widely
distributed at significant levels, has carcinogenic activity (hepatoma
production), and occurs in other ingested products (drugs), its health
risk cannot be ignored. Further support for concern on chloroform may
be reinforced because of the ancillary data on carbon tetrachloride and
the potential additive effect of these two chemicals in the environment.
Out of the list of 221, or now 235, chemicals in drinking water,
only 21 were characterized as having carcinogenic activity. Four of
the chemicals listed are recognized carcinogens, the remaining are
classified as suspect. This is not to imply that there may not be more,
since some remain to be identified in future analytical works (Table 3).
-------
Table 3. Organic carcinogenic chemicals in drinking water
Chemical
Concentration**
(vig/liter)
Carcinogenesis reference
1. Aldrin
2. Benzene
3. Benzo(a)pyrene^
4. Bis(2-chloroethyl) ether
5. Bis(chloromethyl) ether^
6. BHC (Lindane)
7. Carbon tetrachloride^
8. Chlordane
9. Chloroform
10. 1,2-Dibromoethane (EDB)
11. 1,1-Dichloroethane (EDC)
12. Dieldrin
13. DDT
14. DDE
15. Endrin
16. Heptachlor
17. 1,1,2-Trlchloroethane
18. Trichloroethylene
19. Tetrachloroethane
20. Tetrachloroethylene
21. Vinyl chloride^
50
0.0002 - 0.002
0.07 - 0.16
0.1 - 311
0.05 - 0.09
0.004
0.35 - 0.45
10
0.4 - 0.5
Davis and Fitzhugh (1962)
Ishifaru et al. (1971)
IARC (1973a, p. 91-137)
Innes et al. (1969)
Thiess et al. (1973)
Thorpe and Walker (1973)
Tracey and Sherlock (1968)
NCI (1975b)
Eschenbrenner and Miller (1945)
Olson et al. (1973)
NCI (1975b)
Davis and Fitzhugh (1962)
Innes et al. (1969)
Tomatis et al. (1974)
Deichmann et al. (1970)
IARC (1974, p. 173-191)
NCI (1975b)
NCI (1975b)
NCI (1975b)
NCI (1975b)
Maltoni and Lefemine (1974)
NS
U>
00
a.
Determinations from EPA Water Quality Program
Recognized carcinogens
-------
239
There is also the possibility that those classified as potential carcino-
gens may prove to be positive carcinogens on further testing. Not to be
overlooked is the possibility that some chemicals may act as promoters or
cocarcinogens, giving an added insult to those present as recognized or
suspect carcinogens (Table 4). The characterization of inorganic and
organic carcinogens in water supplies (raw or untreated) deserves careful
scrutiny since they may not be completely removed in some municipal
water supplies. Six recognized carcinogens have been identified in
various water supplies (Table 5).
Table 4. Potential carcinogens and promoters
in drinking watera
1. Chloromethyl ethyl ether
2. Bis(2-chloroisopropyl) ether
3. Decane (promoter)
4. Dodecane (promoter)
5. Octadecane
Kraybill 1974
Kraybill 1974
Horton et al. 1965
Horton et al. 1965
Sice 1966
Horton et al. 1957
^Promoters or cocarcinogens demonstrated in tests with
recognized carcinogens. Potential on basis of struc-
tural analogs.
Table 5. Other carcinogenic chemicals in
raw water supplies
Chemical
Reference on activity
1. Arsenica
2. Asbestos^
3. 1,2-Benzanthracene
4. Cadmiuma
5. Benzidinea
6. Dibenz(a,h)anthracene
7. Ethylene thioureaa
8. Chromium (hexavalent)a
9. 3-Methylcholanthrene
10. Mirex
11. Polyurethane
12. Strobane
IARC (1973b, p. 48-73)
IARC (1973b, p. 17-47)
Public Health Service (1958-59)
IARC (1973b, p. 74-99)
IARC (1972, p. 80-86)
IARC (1973a, p. 178-196)
Innes et al. (1969)
IARC (1973b, p. 100-125)
Kelly and O'Gara (1973)
Innes et al. (1969)
Hueper (1961)
Innes et al. (1969)
aRecognized carcinogen
-------
240
DISCUSSION: RISK ASSESSMENT
Biorefractories in municipal water supplies represent a wide spec-
trum of organic and inorganic chemicals. The ability to detect the
presence of these organic compounds is exemplified by the ever-increasing
numbers identified in the surveillance program. Among the organic
chemicals, the chlorinated hydrocarbons and ethers, and a few polycyclic
aromatic hydrocarbons represent thus far the major concern because of
their carcinogenic properties. Some of the halogenated compounds are
ubiquitous, both in raw and treated waters, reflecting the extent of
pollution of waterways, the extent of formation through chlorination,
or the failure to remove them in filtration processes. Certainly, from
the standpoint of preventive health measures, a more comprehensive
understanding of this problem is needed.
The chlorination process may not only produce chlorinated hydro-
carbons or augment the levels already present, but also produce other
compounds, such as chlorinated phenols, which may have more toxic and
biological properties than their parent phenols. Some aromatic com-
pounds would fall in this class as well as the aliphatics. Of signifi-
cant interest is the fact that molecular species appear in drinking
water that were not previously identified in raw water. Ozonation, as
another choice for disinfecting water in the place of chlorine, could
also offer some problems since this process could yield such products as
ozonides, peroxides, epoxides, and aldehydes. Chemicals with some of
these structures have demonstrated carcinogenicity, but many of them
would have to be evaluated for their carcinogenic activity.
While major attention may be given to the organic chlorine com-
pounds, some more attention will also have to be given to inorganic
chemicals. Certainly, the asbestos in drinking water as a potential
hazard is a problem that remains to be assessed. Some studies have
been accomplished on the trace metal contaminants, such as arsenic,
beryllium, cadmium, chromium, cobalt, iron, lead, and nickel (Berg and
Burbank 1972) in an attempt to correlate these trace metal concentra-
tions in drinking water with cancer mortality. There appeared to be
some correlation between nickel concentrations and cancer death rates
-------
241
from mouth and intestinal cancer, and between arsenic and cancer of the
eye and larynx, and myeloid leukemia. Beryllium appeared to be corre-
lated with bone, breast, and uterine cancer; and lead was associated
with kidney, stomach, ovarian, and intestinal cancers. Cancers from
arsenic traced to water supplies in Taiwan and Argentina have been
reported (Tseng et al. 1968). Some trace metals are necessary for
physiological function. An absence of the metal can lead to nutritional
deficiencies. Beyond these low levels, however, an opposite effect may
be noted where there could be enhancement of toxicity through overloading
or interactions.
The relevance of epizootics in aquatic animals with respect to
neoplasia (fish and shellfish tumors) associated with water pollution,
as claimed by Brown and coworkers (1973), is a warning not only on
potential hazard but could also serve as clues to potential association
of point source contamination with increased incidence of human cancers
in certain geographical areas.
Some of the halogenated organic compounds found in water have car-
cinogenic activity and because of this there could be a cancer risk
associated with consumption of such compounds. This hypothesis, advanced
for a probable cause and effect relationship in the case of the New
Orleans water supply, might be confounded by other variables. Since
these compounds are predominantly hepatocarcinogenic (from experimental
studies), one might expect an increased liver cancer rate among certain
populations. Epidemiologic studies thus far have not revealed this to be
the case. Most of the attention has been focused on chloroform and
carbon tetrachloride, which could react additively in hepatoma induction.
Another additive effect could come from the haloethers and polynuclear
hydrocarbons, but their exposure levels to man are lower in drinking
water than those of the aforementioned halogenated methanes, ethanes,
and olefins.
The levels of the carcinogenic organic contaminants are in the
parts-per-trillion and parts-per-billion range. Collectively, they
could have a greater impact on a target organ through continuous expo-
sure than the insult from a single chemical exposure. From mathematical
-------
242
and biological extrapolation factors used to estimate an upper level of
risk, the evidence thus far does not support a conclusion of increased
risk. More critical and definitive data need to be set forth to test
the hypotheses currently advanced. Because of the low level exposures
to man, and in some locations an infrequent occurrence even at higher
concentrations, e.g., 300 ug/liter, it would appear that further studies
are needed. This would require more extensive experimental studies on
the carcinogenicity of lifetime exposures to provide clues on the
potential human health hazard. Beyond this, because of limited epidemic-
logic studies on these contaminants, more definitive monitoring and
surveillance data must be acquired delineating possible groups at risk
where there may be a continuous exposure at high levels from single
agents or multiple stresses. Until then no conclusive assessment of any
health risk can be made.
ACKNOWLEDGEMENT
The author wishes to acknowledge the technical assistance given by
Dr. Kirtland McCaleb and Mr. Arthur McGee, Stanford Research Institute,
Menlo Park, California; and Dr. Sidney Siegel, Carcinogenesis Program,
National Cancer Institute, Bethesda, Maryland.
REFERENCES
Bellar, T. A., and J. J. Lichtenberg. 1974. The occurrence of organo-
halides in chlorinated drinking waters. EPA-670/4-74-008. U.S.
Environmental Protection Agency, Washington, D.C.
Berg, J. W., and F. Burbank. 1972. Correlations between carcinogenesis
trace metals in water supplies and cancer mortality. Ann. N. Y.
Acad. Sci. 199: 249-264.
Boyland, E. 1969. The correlation of experimental carcinogenesis and
cancer in man. Prog. Exp. Tumor Res. 11: 222-234. Basel/New York:
S. Karger AG.
Brown, E. R., J. J. Hazdra, L. Keith, I. Greenspan, J. B. G. Kwapinski,
and P. Beamer. 1973. Frequency of fish tumors found in a polluted
watershed as compared to non-polluted Canadian waters. Cancer Res.
33: 189-198.
Davis K. J., and 0. G. Fitzhugh. 1962. Tumorigenic potential of aldrin
and dieldrin for mice. Toxicol. Appl. Pharmacol. 4: 189.
-------
243
Deichmann, W. B., W. E. MacDonald, E. Blum, M. Berilacqua, J. Radomski,
M. Keplinger, and M. Balkus. 1970. Tumorigenicity of aldrin,
dieldrin and endrin in the albino rat. Ind. Med. Surg. 39: 426.
Environmental Protection Agency. 1975. Report to the U.S. Congress on
suspect carcinogens in water supplies. U.S. Environmental Protec-
tion Agency, Washington, D.C. June 17.
Epstein, S. S. 1970. Control of chemical pollutants. Nature 228:
816-819.
Eschenbrenner, A. B., and E. Miller. 1945. Induction of hepatomas in
mice by repeated oral administration of chloroform with observa-
tions on sex differences. J. Nat. Cancer Inst. 5: 251.
Furia, T. E., and N. Bellanca. 1971. Fenaroli's handbook of flavor
ingredients. Chemical Rubber Co., Cleveland, Ohio.
Goldsmith, J. 1975. Proposed terminology and classification for
environmental cancer. National Cancer Institute, Division of
Cancer Cause and Prevention. Unpublished Report.
Higginson, J. 1969. Present trends in cancer epidemiology. Proc. Can.
Cancer Conf. 8: 40-75.
Horton, A. W., D. T. Denham, and R. P. Trosset. 1957. Carcinogenesis
of the skin. II. The accelerating properties of aliphatics and
related hydrocarbons. Cancer Res. 17: 758-766.
Horton, A. W., P. A. Van Dreal, and E. Bingham. 1965. Physiochemical
mechanism of acceleration of skin carcinogenesis. Adv. Biol. Skin
7: 165-181.
Hueper, W. C. 1961. Carcinogens in the human environment. Arch.
Pathol. 71: 237-267 and 355-380. March.
Innes, J. R. M., B. M. Ulland, M. G. Valerio, L. Petrucelli, L. Fishbein,
E. R. Hart, A. J. Pallota, R. R. Bates, H. L. Falk, J. J. Gart,
M. Klein, I. Mitchell, and J. Peters. 1961. Bioassay of pesticides
and industrial chemicals for tumorigenicity in mice: a preliminary
note. J. Nat. Cancer Inst. 72(6): 1101-1114.
International Agency for Research on Cancer. 1972. IARC monographs on
the evaluation of carcinogenic risk of chemicals to man. 1: 61-65,
80-86. Lyon, International Agency for Research on Cancer.
International Agency for Research on Cancer. 1973a. IARC monographs on
the evaluation of carcinogenic risk of chemicals to man: certain
polycyclic aromatic hydrocarbons and heterocyclic compounds. 3:
91-136, 178-196. Lyon, International Agency for Research on Cancer.
-------
244
International Agency for Research on Cancer. 1973b. IARC monographs on
the evaluation of carcinogenic risk of chemicals to man: asbestos,
inorganic and organometallic compounds. 2: 17-47, 48-73, 74-99,
100-125. Lyon, International Agency for Research on Cancer.
International Agency for Research on Cancer. 1974, IARC monographs on
the evaluation of carcinogenic risk of chemicals to man: some
organochlorine pesticides. 5: 173-191. Lyon, International Agency
for Research on Cancer.
Ishimaru, T., H. Okada, T. Tomiyasu, T. Tsuchimoto, T. Hoshino, and
M. Ichimaru. 1971. Occupational factors in the epidemiology of
leukemia in Hiroshima and Nagasaki. Am. J. Epidemiol. 93: 157-165.
Kelly, M., and R. O'Gara. 1973. Induction of tumors in newborn mice
with dibenz(a,h)anthracene and 3-methylcholantrene. J. Nat. Cancer
Inst. 26: 651-679.
Kraybill, H. F., and M. B. Shimkin. 1969. Carcinogenesis related to
foods contaminated by processing and fungal metabolites. Adv.
Cancer Res. 8: 191-248. Academic Press, New York.
Kraybill, H. F. 1974. The distribution of chemical carcinogens in
aquatic environments (UICC Symposium, Cork, Ireland, October 1974).
In Progress in experimental tumor research: a monograph. Basel:
S. Karger, AG. In press.
Lewin, R. 1974. Environmental search for the source of cancer. SR/
World — Science Section 1974: 50-51. April 20.
Maltoni, C., and G. Lefemine. 1974. Carcinogenicity bioassay of vinyl
chloride current results. Part II. Carcinogenesis associated
with vinyl chloride. Ann. N. Y. Acad. Sci. 246: 195-218.
Maltoni, C. 1973. Occupational carcinogenesis. In Advances in tumor
prevention, detection and characterization. Vol. 2. Detection and
characterization. Preprint of International Congress Series.
No. 322.
Merck Index of Chemicals and Drugs. 1960. P. Stecher, M. J. Finkel,
0. H. Siegmund, and B. M. Szafranski (Eds.). Merck and Company,
Rahway, New Jersey.
National Cancer Institute. 1975a. List of chemicals that cause cancer
in man, p. 85. House of Representatives Hearings, Subcommittee of
the Committee on Appropriations, March 1975. Department of Labor,
Health, Education and Welfare. U.S. Government Printing Office,
Washington, D.C.
-------
245
National Cancer Institute. 1975b. Preliminary findings on bioassay of
chemicals for carcinogenic activity. Unpublished data from Bioassay
Program.
Olson, W. A., R. T. Haberman, E. K. Weisburger, J. M. Ward, and
J. H. Weisburger. 1973. Brief communication: induction of
stomach cancer in rats and mice by halogenated aliphatic fumigants.
J. Nat. Cancer Inst. 51: 1993-1995.
Piet, G. J., B. C. T. Zoeteman, A. H. Nettenbreiger, and G. T. H. Ruijgrok.
1973. Bis(2-chloroisopropyl) ether in surface and drinking water in
the Netherlands. Orientation Report of Rijksinstituut voor Drink-
watervoorziening, S-Gravenhage, Parkweg 13, The Netherlands.
Public Health Service. 1958-59. Survey of compounds which have been
tested for carcinogenic activity. PHS-149, Series Entry 393.
DREW Publ. No. (NIH) 72-35. U.S. Government Printing Office,
Washington, D.C.
Sice, J. 1966. Tumor-producing activity of n-alkanes and 1-alkanols.
Toxicol. Appl. Pharmacol. 9: 70.
Symons, J. M. 1975. Suspect carcinogens in water supplies. Interim
Report to Congress. Appendix A. Character and extent of contamina-
tion. Section 1. National Organics Reconnaissance Survey. EPA
Progress Report. Cincinnati, Ohio. April.
Tardiff, R. G., G. F. Craun, L. J. McCabe, P. E. Bertozzi. 1975. Sus-
pect carcinogens in water supplies. Interim Report to Congress.
Appendix B. Health effects caused by exposure to contaminants.
EPA Progress Report. Cincinnati, Ohio. April.
Thiess, A. M., W. Hey, and H. Zeller. 1973. Zur Toxikologie von
Dichloromethylether — Verdacht auf kanzerogene Wirkung auch beim
Menschen. Zentralbl. Arbeitsmed. Arbeitschutz. 23: 97.
Thorpe, E., and A. I. T. Walker. 1973. The toxicology of dieldrin
(HEOD). II. Comparative long-term oral toxicity studies in mice
with dieldrin, DDT, phenobarbitone g-BHC and y-BHC. Food Cosmet.
Toxicol. 11: 433.
Tomatis, L., V. Turusov, R. T. Charles, and M. Boiocchi. 1974. The
effect of long-term exposure to l,l-dichloro-2,2-bis(p-chlorophenyl)
ethylene(p,p'-DDE), to l,l-dichloro-2,2-bis(p-chlorophenyl)ethane
(p,p'-DDD) and to the two chemicals combined on CF-1 mice. J. Nat.
Cancer Inst. 52. In press.
Tracey, J. P., and P. Sherlock. 1968. Hepatoma following carbon
tetrachloride poisoning. N. Y. State J. Med. 68: 2202.
-------
246
Tseng, W., H. Chu, S. How, J. Fong, C. Lin, and I. Shuyeh. 1968.
Prevalence of skin cancer in an endemic area of chronic arsenicism
in Taiwan. J. Nat. Cancer Inst. 40: 453-463.
Van Duuren, B. L., C. Katz, B. M. Goldschmidt, K. Frenkel, and A. Sivak.
1972. Carcinogenicity of haloethers. II. Structure activity -
relationship of analogs of bis(chloromethyl) ether. J. Nat. Cancer
Inst. 48: 1431-1439.
DISCUSSION
David Friedman, Food and Drug Administration. On your last slide,
Hueper has shown that polyurethane was in water and was also carcino-
genic. Could you indicate the type of urethane this was and what type
of carcinogenicity?
Kraybill. At this point we should correct some impressions gleaned
from identification of carcinogens (see Table 5) in water. This table
does not necessarily mean that these chemicals were tested in water and
then found to be carcinogenic. The reference to Hueper's work, for
example, on polyurethane, means that Hueper and others tested this
chemical separately and reported on the carcinogenicity. Most of these
polymeric materials have been tested by subcutaneous injection in which
instance they induced sarcomas (Hueper, J. Nat. Cancer Inst. 33: 1005,
1964). Other sites of tumor induction for this plastic and medicinal
agent is the cecum and connective tissue.
Friedman. Was this chemical shown to be actually in water?
Kraybill. This chemical appeared on lists of chemicals monitored
in water. Polyurethane (spandex) appeared on a list of organics in
the Kanawha River — a report made available (unpublished) by Dr. Edward
Light, Research Director of Campaign Clear Water, Box 567, Charleston,
West Virginia 25322.
-------
THE POTENTIAL FOR INCREASED MLJTAGENIC RISK TO THE HUMAN
POPULATION DUE TO THE PRODUCTS OF WATER CHLORINATION
Robert B. Gumming
Biology Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
The chlorination of water in which there are organic materials
produces stable chlorine-containing compounds which may have substantial
biological activity. Such compounds are released into surface waters by
chlorination processes at very low concentrations, and the assessment of
their potential for producing genetic damage in the human population
presents a formidable technological problem. The approach which seems
most productive is to select several from the array of chlorine-containing
compounds known to be produced by water chlorination and to study their
behavior in well-established genetic test systems including those which
have been used to estimate levels of genetic risk for other agents. If
risk estimates can be determined for several compounds and some factor
added for potential interaction then a genetic risk estimate can be made
for the entire water chlorination process. Such an estimate would always
be imperfect and subject to revision as additional data were obtained.
In the model studies described in this paper, 5-chlorouracil (5-C1U)
has been tested in several mammalian and submammalian genetic test systems.
The incorporation of 5-C1U into the DNA of mice which had been exposed to
this compound in their drinking water has been measured. Both a specific-
locus mutation test and a dominant-lethal mutation test have been performed
on mice which have been similarly exposed. The compound has been tested
for mutagenicity in several types of bacteria and has been found to be
highly mutagenic in E. ooli. The data from 5-C1U incorporation studies
into the DNA of mice together with the specific-locus mutation data allow
the calculation of the upper 95% confidence limit for mutations induced
in the human population at environmental exposure levels. This calculation
demonstrates that 5-C1U, by itself, does not pose a significant genetic
hazard to humans at current release levels. The more significant question
of whether, in sum, all of the chlorine-containing organic compounds pro-
duced by water chlorination pose a significant hazard is yet to be determined.
247
-------
248
INTRODUCTION
Our primary biomedical concern related to water chlorination and human
exposure to chlorinated organics has been the potential for increasing the
incidence of cancer among exposed individuals. Various aspects of this
problem are dealt with by the other three speakers in this session. I
will deal with a different but related problem, that is, the potential
for increasing the amount of induced heritable genetic damage in exposed
human populations.
There are several important differences in the techniques and the data
used for the estimation of mutagenic risk and carcinogenic risk. These
are two biomedical endpoints with very different implications for the
human population and very different problems associated with their assess-
ment. Although the medical consequences of genetic disease may be
altogether as devastating as the consequences of cancer, identifying the
environmental factors that lead to the damage is very much more difficult.
The expression of genetic disease is usually far removed in time from the
induction of the mutations which caused it and, thus, from the toxic agent
which lead to the mutations. The consequences of genetic damage would
not be expected to fall upon the individual in which the mutations were
induced but upon his progeny or descendants—perhaps several or many gen-
erations later. In carcinogenesis the target is the individual, and the
cancer statistics are compilations of many individual tragedies. The real
target for environmental mutagens is not the individual, or even the small
group, but the human gene pool, and the human gene pool is not only the
source of all our admirable and less than admirable attributes, it is also
the key to the continued existence of our species.
There are other differences between the assessment for mutagenicity
and carcinogenicity. A person who has cancer knows it. He is easy to
identify medically. He knows or can quickly estimate the consequences
of his disease. It is not very difficult for him or others to come up
with some estimate of the cost of his condition to himself and to the
society, and to the individual it is frequently catastrophic. One can
make cost estimates for induced cancer by simply adding up the number of
affected individuals and recognizing that for each individual we are
-------
249
dealing with a very serious problem. Genetic effects may be very much
more subtle and the estimation of the cost of genetic damage is very
much more difficult. Genetic disease ranges from things which are as
totally incapacitating as cancer—lethal conditions—to things which are
much less serious to the individual, but which in their sum total, since
they may affect very large populations, may have as great a cost to
society.
Another difference between cost assessments of mutagenicity and
carcinogenicity in human populations is Aat there is a straight forward
relationship between carcinogenicity and cancer incidence, whereas, the
relationship between increased mutation frequency and increased genetic
disease is not straight forward. The way in which increased mutation
frequency is related to increased medical costs to the society is
presently the subject of much controversy and debate. This makes the
assessment of mutagenic risks—risks of increased mutation frequency—
very much easier than estimating the increased medical costs to the
society.
These differences in the nature of the two major public health
problems which may be associated with exposure to chlorinated organics
demand different approaches to risk assessment. Ironically, simple
in vitro mutational tests may be more useful in dealing with the problem
of carcinogenesis than they are for predicting genetic hazzards (McCann
and Ames 1976).
What are the characteristics of the problem we are facing? The
problem is estimating the potential public health hazards from water
chlorination and from the products released therefrom. First, as has
been amply documented by the first two speakers in this session, we are
not dealing with a single compound, or even a few, but with many compounds
some of which are poorly characterized. Secondly, these compounds are
individually present in very low concentrations. Third, the exposure
would be expected to be continuous throughout the lifetime of the indi-
vidual or, more importantly for genetic damage, throughout the repro-
ductive lifetime of the individual which for humans is generally taken
to be thirty years. Fourth, very large populations would be exposed so
-------
250
that even, small effects might involve significant numbers of people. So
we have large populations exposed for long periods of time to low levels
of very many compounds.
Some of the compounds identified from this array would be expected
to have rather potent biological effects. For others we would not be
able to even guess what their effects might be. Some of the identified
compounds would be expected to induce, at least in some test systems, a
class of genetic damage which we can loosely refer to as point mutations,
though the induction of gross chromosomal abnormalities is not excluded.
This means that test systems should be provided which will measure point
imitations as well as chromosome aberrations.
What kinds of test systems and other means of obtaining data rele-
vant to this problem are available? The epidemiological approach may,
in certain selected instances, yield valuable information about carcino-
genicity. The next speaker will discuss this approach in greater detail.
The epidemiological approach is not likely to provide useful information
on potential genetic effects of chlorinated organics. This is true
because of the long time lag between exposure and observed effect, the
tremendous genetic diversity of the human population, the lack of dis-
crete large populations with identifiably different exposure conditions
over the required long periods of time, and several other factors.
Several in vitro tests for mutagenicity are available mostly
involving prokaryotes. A high correlation has been claimed (McCann and
Ames 1976) between tests for mutagenicity involving reversion to proto-
trophy of histidine auxotrophs of Salmonella and carcinogenicity in in vivo
tests in mammals. Thus, some simple bacterial tests may serve as indi-
cators of compounds with mutagenic or carcinogenic potential but these
tests do not provide data which can be used to estimate human risk. The
usefulness of bacterial tests as preliminary screens may be enhanced by
the addition of enzyme systems which mimic mammalian metabolism in the
activation of certain procarcindgens or promutagens, but in vitro meta-
bolic activation systems are complex and unpredictable. Therefore, these
tests may leave us with data which are difficult to interpret with
confidence.
-------
251
Historically there has not been good correlation of the results of
mutagenicity testing in whole mammals with the data derived from muta-
genicity tests in other systems (Russell 1972). For risk assessments,
data are needed from genetic tests involving intact mammals. In addition,
these tests must be able to measure both point mutations and transmitted
chromosomal effects. Clearly it is impossible to deal with the entire
range of chlorinated organics in a definitive way. Simple tests and
structural considerations must be used to identify potentially hazardous
compounds among the many present and we must concentrate initally on
these. There will be many compromises, but it is essential to start
gathering genetic data on a problem of such immense potential public
health importance as water chlorination.
STRATEGY FOR AN EXPERIMENTAL APPROACH TO MUTAGENIC RISK ASSESSMENT
We have adopted the following procedure as a strategy in investigating
the potential for genetic damage from chlorinated organics. We select
from the many compounds known to be produced by water chlorination
a few and try to understand the mutagenicity of these in mammalian
systems. We concentrate on basic mechanisms so that the information
derived from the studies may have some predictive value for a much larger
group of compounds. We also employ the selective use of nonmammalian
and in vitro test systems to clarify points about the mechanism of action
of our models. Only data from intact animals are used to quantify
probable risks for the compounds we study. Of course this is a big
order. We will never have all of the pieces of the puzzle in place, but
I think we will generate some data which will help to put the problem
into perspective.
I would like to illustrate this approach with a particular model
compound. This is a compound which we have worked on and which has also
received the attention of some of the environmental scientists in this
laboratory. It is a compound which has been identified from the chlorin-
ated sewage effluents in Oak Ridge and, therefore, has some local interest.
The compound is 5-chlorouracil.
-------
252
5-Chlorouracil has several attributes which commend it. First, on
theoretical grounds it would be expected to be mutagenic in some test
systems. If we look at the whole list of compounds known to be present
in chlorinated sewage effluents, from structural considerations we would
expect 5-chlorouracil to be one of the bad actors.
The mechanism of action of 5-chlorouracil (5-C1U), at least by
analogy with its big brother, 5-bromouracil (5-BrU), is thought to be
fairly well understood. Sometimes, however, when something is thought
to be well understood, and it is looked into deeply enough, certain
discrepancies become apparent. 5-Bromouracil is mutagenic in a number of
prokaryotic systems. It is mutagenic in phage (Litman and Pardee 1956;
Terzaghi, Streisinger, and Stahl 1962). It is mutagenic in E. eoli
(Witkin and Parisi 1974). It appears not to be mutagenic in certain
other bacterial systems, but we do not know enough about why it is not
mutagenic in those systems to say very much about it. We know that in
many of these systems it is incorporated into DNA where it replaces the
base thymine (Dunn and Smith 1954). Bacteria, and particularly thymine-
requiring bacteria, will take up this base analog and incorporate it into
their DNA, and incorporation into the DNA of the halogen-containing base
appears to be necessary for mutagenesis to occur.
The mechanism for mutagenesis by 5-bromouracil, and by analogy
5-chlorouracil, as mentioned earlier is thought to be well understood.
It results from what molecular geneticists call transitions. Transitions
are the substitution of one pyrimidine for another or one purine for
another in the DNA. This results in a change in the coded information
in the gene involved. Halogenated uracils could produce this type of
mutation by a mistake in pairing at DNA replication.
Since mutagenesis by these compounds require incorporation into the
DNA, we have a biochemical handle to work with in mammals and other
test organisms. We can measure the amount of incorporation of the com-
pound into DNA and compare that with the mutation frequency we observe.
The work with phage has demonstrated (Terzaghi, Streisinger, and Stahl
1962) that one gets mutations from 5-bromouracil after the compound is no
longer present in the environment of the organism as long as it has been
-------
253
previously incorporated into the DNA. This indicates that once it is in
the DNA, it can continue to make mistakes. We need to know if 5-chloro-
uracil, like 5-bromouracil, is incorporated into DNA in place of thymine,
if it can produce mutations in the same way, and what sort of effects we
can observe in mammals.
EXPERIMENTAL
The first question that needed to be answered was whether 5-chloro-
uracil like 5-bromouracil is mutagenic in bacteria. We used the same
bacterial strain, E. ool-i strain WP-1, used to study 5-bromouracil muta-
genesis by Witkin and Parisi (1974) and the same mutational test. Table
1 shows the results of a typical experiment. Mutations per survivor are
given for bacteria grown in the presence of the natural base thymine,
5-chlorouracil, or 5-bromouracil. The controls (thymine) show about
two mutants per 108 viable bacteria. There is about a 300-fold increase
in mutation frequency after the bacteria have grown in the presence of
either halogenated base at 50 ug/ml for one hour. The mutation frequency
drops for longer exposure times, and we think that at higher levels of
incorporation of the base analogs the bacteria lose their ability to
express the mutations. It is clear that 5-chlorouracil is mutagenic to
bacteria and that its mutagenic behavior is very similar to 5-bromouracil.
Table 1. Mutations induced in E. ooli by
halogenated pyrimidine base analogs
Time
(min)
0
30
60
90
120
150
Mutants /survivor x 10 ^
*
Thymine
3.22
3.00
1.89
0.93
0.46
1.77
5-C1U
2.01
628.92
635.13
414.93
13.19
6.56
5-BrU
5.78
318.88
641.12
337.12
60.12
24.31
Mean for all thymine points, 1.88 ± 1.09.
-------
254
Now we will look at 5-chlorouracil in mammals. The first question
is: how toxic is 5-chlorouracil to mice? The answer is that it isn't
toxic. Male mice, (101 x €3!!) F]_, were put on drinking water containing
5-chlorouracil at near the maximum soluble concentration (1 g/1). They
drank the water well and there were no adverse health effects noted for
a period of greater than one year. This concentration is more than one
million times the estimated human environmental level. The weight of
the animals was near that of controls kept on regular water. Reproduc-
tion was normal in every respect and the young appeared normal. The next
question is: do the mice incorporate 5-chlorouracil into their DNA? The
answer, reported in preliminary form elsewhere (Cumming, Pal, Walton,
and Russell 1975), is yes. DNA was extracted from the livers and testes
of mice which had been on water containing 5-chlorouracil for various
periods of time. This DNA was hydrolyzed by successive incubation with
DNase I, snake-venom phosphodiesterase, and alkaline phosphatase to
produce a mixture of deoxynucleosides. 5-Chlorodeoxyuridine was iden-
tified and quantified by chromatography of the DNA hydrolysate on Aminex
A-6 (cation exchange resin). It was absent from the DNA of animals which
had not been exposed to 5-chlorouracil. About 1.4% of the thymine
residues were replaced by 5-chlorouracil in those animals which had been
exposed to this compound at a level of 1 g/1 in their drinking water.
This amounts to about one 5-C1U for every 250 nucleotides (averaged
throughout the genome) or about 2.2 x 107 5-ClU's per genome or about
four 5-ClU's per structural gene. At this level of incorporation no
physical effects on the mice were noted. It is clearly necessary to look
for genetic effects in mammals. The meaning of these experiments appears
to be that if mammals, presumably including humans, drink water con-
taining 5-chlorouracil, some of it goes into their DNA.
A small dominant lethal mutation experiment was performed to check
the possibility that the incorporated 5-chlorouracil was leading to
increased chromosome breakage. The procedure for doing the dominant
lethal test is essentially as described by Ehling (Ehling, Cumming, and
Mailing 1968) except that a modification was made to allow for continuous
exposure of the males to water containing 5-chlorouracil while not
-------
255
exposing the females. The way this test was performed all male germ cell
stages were exposed. Females were checked each morning for vaginal plugs
as an indication of mating, and mated females were removed and replaced.
At 13 to 15 days of gestation the females were sacrificed and the uterine
contents were scored. The results are shown in Table 2.
Table 2. Dominant lethal mutation experiment with 5-chlorouracil
Control
5-Chlorouracil
Number
of
females
36
42
Total
implants
9
7.53
6.90
Live
embryos
?
6.67
6.33
Dead
embryos
(%)
11.44
8.27
Dominant
lethal
effect
—
5%
The 5% dominant lethal frequency shown in the table is not signifi-
cant and thus, within the limits of resolution of this test, no dominant
lethal effect is demonstrated at about one million times the maximum
human exposure level. But these data still cannot be used to calculate
a maximum level of human genetic risk.
There are two general ways of estimating genetic risk to a popula-
tion exposed to a particular insult.
1. To estimate the overall damage. This is done by estimating the
total number of new mutations which will be induced by the
agent at a given exposure level and the medical consequences
of these mutations. This would be a good way to estimate the
effect of mutagens exce^-i. that there is presently no way to do
it very effectively in mammals.
2. To express the increased mutation frequency at particular loci
in treated animals in terms of the spontaneous frequency at the
same loci. The assumption is made that man has a spontaneous
mutation frequency with which he can live, and that if we do not
alter it much the compound will be relatively safe.
For the purpose of estimating risk by the second method, we have
used the specific-locus test of Russell (Russell 1951, Russell and
Gumming 1975) .
-------
256
The specific-locus test measures recessive visible mutations at
seven marked loci in the mouse. The test is highly quantitative and
simple to perform. It has been very important historically as the source
of data which have been used to set maximum allowable radiation exposure
levels for man. The test may be very large and has gained a reputation
for being cumbersome and expensive. However, as I shall show, under
certain conditions useful information can be obtained with a very small
test. The specific-locus test has the advantages of producing highly
quantitative data on a very relevant genetic endpoint.
The specific-locus test has been used on a very small scale to
estimate the upper limit of genetic risk that might occur at the human
exposure levels of 5-chlorouracil (Russell and Gumming 1975). We did
a mini-specific-locus test. Eleven males were continuously exposed to
1 g/1 5-chlorouracil and mated to T-stock females. No mutations were
observed in 314 offspring. The mice had been exposed to a concentration
of 5-chlorouracil in the drinking water at least one million times
higher than the human exposure level. Their offspring were conceived,
on the average, three months after a steady-state value for incorpora-
tion of 5-chlorouracil had been reached. This is 1/120 of the 30-year
generation time exposure in man, but the mouse still gets about 8000
times the human exposure (106/120). With this exposure factor and the
observation of zero mutations in 314 offspring we can calculate an
estimate of the upper level of genetic risk that might occur at the
human exposure level. Taking 3.3 as the upper 95% confidence limit of
the observed zero mutation frequency in the 314 offspring, subtracting
the unknown spontaneous frequency of 28 mutations in 531,500 offspring
and dividing by the exposure factor of 106/120, we come out with an esti-
mated induced mutation rate, at the human exposure level, that is only
2% of the spontaneous frequency:
f3. 3/314 - 28/531.500\ 7/28
106/120 // \531,500j
-------
257
We can see from this calculation that 5-chlorouracil by itself
does not seem to be a particular genetic hazard to man even though it
is mutagenic in bacteria and other lower organisms. We are 95% confi-
dent that at human exposure levels we have not increased the mutation
frequency by more than 2% of the spontaneous. That is a small number.
There are few compounds in our environment for which we have that kind
of information.
What is an acceptable increase in the mutation frequency is a
political and not a scientific question. We really do not know what a
given increase in mutations means in terms of increased human suffering
and economic expense. We assume that any increase in the human muta-
tion frequency will not be without its costs.
We have looked at one compound from the many present and find that
by itself it is relatively safe. But we have not addressed the more
difficult problem of additive effects of many compounds and interactions.
Thus we have just started to scratch the surface of a large area of
legitimate concern.
ACKNOWLEDGMENTS
I would like to acknowledge my collaborators on the various
projects mentioned in this paper. They are: W. L. Russell, Bimal Pal,
Marva F. Walton, Barbara J. Elmhorst, Donna L. George, and David L.
Sultzer. This work is sponsored by the United States Energy Research
and Development Administration under contract with Union Carbide
Corporation.
REFERENCES
Gumming, R. B., and B. J. Elmhorst. 1975. A test for 5-chlorouracil-
induced dominant-lethal mutations in the mouse. Biol. Div. Annu.
Prog. Rep., ORNL-5072, p. 135.
Gumming, R. B., D. L. George, M. F. Walton, and B. J. Elmhorst. 1975.
Mutations produced by 5-chlorouracil and 5-bromouracil in
Escherich-ia eoli. Biol. Div. Annu. Prog. Rep., ORNL-5072, p.
135-136.
Cumming, R. B., B. C. Pal, M. F. Walton, and W. L. Russell. 1975.
Studies on potential genetic effects of 5-chlorouracil in mammals.
Biol. Div. Annu. Prog. Rep., ORNL-5072, p. 134-135.
-------
258
Dunn, D. B., and J. D. Smith. 1957. Effects of 5-halogenated uracils
on the growth of Eseherichia aoH and their incorporation into
deoxyribonucleic acids. Biochem. J. 67: 494-506.
Ehling, U. H., R. B. Gumming, and H. V. Mailing. 1968. Induction of
dominant lethal mutations by alkylating agents in male mice.
Mutation Res. 5: 417-428.
McCann, J., and B. N. Ames. 1976. Detection of carcinogens as mutagens
in the Salmonella/microsome test: Assay of 300 chemicals: Dis-
cussion. Proc. Nat. Acad. Sci. U.S.A. 23: 950-954.
Russell, W. L. 1951. X-ray-induced mutations in mice. Cold Spring
Harbor Symposia on Quant. Biol. 16: 327-336.
Russell, W. L. 1972. Radiation and chemical mutagenesis and repair in
mice, p. 239-247. In Roland F. Beers, Jr., Roger M. Herriott,
and R. Carmichael Tilghman (eds.) Proceedings of Miles Fifth
International Symposium on Molecular Biology: Molecular and
Cellular Repair Processes, Johns Hopkins University, Baltimore,
Maryland, June 3-4, 1971.
Russell, W. L., and R. B. Gumming. 1975. An example of conditions that
make the mouse specific-locus test highly efficient at low expense.
Biol. Div. Annu. Prog. Rep., ORNL-5072, p. 126-127.
Terzaghi, B. E., G. S. Streisinger, and F. W. Stahl. 1962. The
mechanism of 5-bromouracil mutagenesis in the bacteriophage T-4.
Proc. Natl. Acad. Sci. U.S.A. 48: 1519-1524.
Witkin, E. M., and E. C. Parisi. 1974. Bromouracil mutagenesis:
mispairing or misrepair? Mutation Res. 25: 407-409.
DISCUSSION
A. D. Venosa, U.S. Environmental Protection Agency. Since uracil is
a base in RNA, would you expect 5-chlorouracil to be incorporated into
the RNA as opposed to DNA and thereby elicit a translation effect rather
than transcriptional effect on protein synthesis?
Gumming. I would have no a priori expectation, but in fact, when
you measure it, you find that 5-chlorouracil goes only into DNA and
not into RNA. The 5-chlorouracil molecule is shaped more like thymine
than uracil since the chlorine atom is in the same position as a methyl
group on thymine. Apparently the enzymatic machinery which assembles
nucleic acids recognizes 5-chlorouracil and 5-bromouracil only as
thymine. 5-Fluorouracil is, however, incorporated into RNA.
-------
THE EPIDEMIOLOGIC APPROACH TO THE EVALUATION OF
WATER-BORNE CARCINOGENS
Kenneth P. Cantor
U.S. Environmental Protection Agency
Washington, D.C. 20460
and
National Cancer Institute
Bethesda, Maryland 20014
ABSTRACT
The classic mid-19th Century London study of John Snow, in which
cholera was linked to water-borne contamination, is reviewed to provide
a context for discussing the epidemiologic approach to evaluation of
carcinogens in drinking water. The value of epidemiologic studies arises
from the fact that observations are made directly on human populations
so that extrapolation from animal models and/or unrealistically high
doses is not necessary to predict effects in humans. Limitations of
the approach include the long latent period for most cancers, difficul-
ties in estimating dose, the definition of at-risk populations, and the
relatively low exposure levels to carcinogenic agents. Recent studies
of a preliminary nature are reviewed,
INTRODUCTION
Previous sessions of this conference have raised the issue of how
one should weigh the findings of various scientific disciplines which
help shape environmental regulatory policy. In this discussion, I will
focus on one of these fields — epidemiology — and dwell on problems of
methodology central to epidemiologic investigation of chronic diseases
as related to drinking water contaminants. An understanding of method-
ology leads to an appreciation of the utility of epidemiologic studies
as well as knowledge of their limitations.
259
-------
260
The concern for methodology is related to the philosophical
problem of what we accept as evidence of a cause and effect relationship
in disease etiology. When, for example, should we demand more than a
high degree of statistical association between presumed cause and effect
before taking action on the findings of epidemiologic investigations?
In addition, an understanding of methodology is crucial in deciding how
to approach the study of possible carcinogenic effects in human popula-
tions exposed to very low levels of substances which are known to cause
cancer in laboratory animals, induce mutations in microorganisms, or
are carcinogenic to persons who were unknowingly exposed in occupational
settings to far higher levels.
Critical examination of the methodology used in epidemiologic
investigations is of paramount importance when we are faced with acting
on the results of such studies. To put these problems in context, I
will review one of the first and most definitive epidemiologic investi-
gations into health aspects of drinking water conducted in the mid-19th
century by Dr. John Snow. Snow's study of cholera in London — well known
to students of epidemiology and biostatistics — will serve as a basis
from which we can examine aspects of today's epidemiologic approach to
water-borne carcinogens.
CHOLERA AND DRINiCING WATER
In 1855, Snow, an English physician, published the second edition
of "On the Mode of Communication of Cholera" (Snow 1855). This work
was the culmination of seven years of intensive investigation in which
Snow personally looked into conditions surrounding thousands of cholera
deaths. The conclusions he drew regarding the cause and spread of this
disease are recognized today as being essentially correct.
Snow commenced his investigation in 1848 with the hypothesis that
the disease is caused by some characteristic of particular drinking
waters — in his words, some "poison." His reasoning grew from the
observation that in cholera, initial symptoms localize in the alimentary
canal and then proceed to more general systemic involvement. With
other diseases studied at that time, symptoms such as headaches, fever,
-------
261
and a higher pulse rate appeared first. Snow argued that this pattern
of symptom development indicated that the causative agent in cholera
must first come into contact with the alimentary mucosa before reaching
other organs. In addition to the observation of disease progression,
there were also a large number of anecdotal accounts of the occurrence
and local spread of cholera which tended to support the water-borne
"poison" hypothesis. Drinking water was therefore implicated as a prime
suspect as the carrier of the "morbid material."
Snow was able to test his hypothesis owing to the unique arrangement
of London's water supply system. The system was operated by two private
companies, each of which had intakes in the Thames River as well as its
own distribution system within the city. London suffered a series of
cholera epidemics in the 1840*s and 50's, two major ones occurring in 1849
and 1854. In the interval between these outbreaks, in 1852, one of the
suppliers (the Lambeth Company) moved its intake from the highly polluted
part of the Thames, which was serving both companies, to a less contam-
inated stretch of the river. Intakes of the other utility (the Southwark
and Vauxhall Company) remained in place.
In examining mortality rates from cholera, Snow observed that areas
served by the Lambeth Company, with its new water supply, fared much
better in the epidemic of 1854 than in the 1849 episode. Snow's compil-
ation of deaths and rates is shown in Table 1. Cholera death rates in
districts served by the Lambeth Company in 1849, before the intakes
were moved, were comparable to those in districts served by the Southwark
and Vauxhall Company, the death rate in Southwark and Vauxhall districts
being 1.2 times that in Lambeth Company districts. In the 1854 epidemic,
after the change in supply for the Lambeth Company, Southwark and
Vauxhall districts suffered death rates 8.6 times those of Lambeth
Company service areas. The part of the city served by both companies
had intermediate rates. The data necessary to make these observations
consisted of group statistics - cholera mortality rates for different
districts of the city and knowledge of the water distribution patterns
for these same districts. Snow used no information regarding the
personal backgrounds, demographic characteristics, economic class, and
-------
Table 1. Cholera deaths in London, 1849 and 1854.a in districts
served by different water companies*
Cholera deaths
Risk relative to the
Water supplier
of district
Southwark and
Vauxhall Co.
Both Companies
Lambeth Company
Population
(1851)
167,654
300,149
14,632
Number
1849
2,226
3,905
162
1854
2,458
2,547
25
Rate/ 1000
1849
13.5
13.0
11.1
1854
14.7
8.5
1.7
Lambeth Co
1849
1.2
1.2
1.0
. district
1954
8.6
5.0
1.0
In 1849 data was collected for the 13-week period ending Nov. 19. In 1854 data was collected
for the 14-week period ending Oct. 14.
From John Snow. 1855. On the mode of communication of cholera, 2nd edition.
-------
263
so on, of the populations he studied to make his observations. While
there was a clearly observable difference in death rates in districts
served by the water companies, Snow was concerned that the fact of
association of cholera deaths and water supply was not sufficient to
make the causal argument incontrovertible. After all, one could argue
that differences in geographically separated populations other than water
supply — what we today would call confounding causes — could have led
to his observation. He therefore carried out a detailed house-to-house
survey of cholera deaths in the 1854 epidemic in districts served by
both companies. In his words:
". . . the intermixing of the water supply of the Southwark and
Vauxhall Companies, over an extensive part of London, admitted
of the subject being sifted in such a way as to yield the most
incontrovertible proof on one side or the other . . . The
pipes of each company go down all the streets, and into nearly
all the courts and alleys .... In many cases a single house
has a supply different from that on either side. Each Company
supplies both rich and poor, both large houses and small;
there is no difference either in the condition or occupation
of the persons receiving the water of the different companies."
Snow thus had available for observation exposed and nonexposed
populations which were almost perfectly matched with respect to all
other possible differences which might have had a bearing on the out-
come. Table 2 shows the result of his investigation. For the first
eight weeks of the 1854 epidemic, houses served by the Southwark and
Vauxhall Company experienced a cholera death rate of 4.76 per 1000
population, over 7 times the rate of 0.67 per 1000 in houses supplied
by the Lambeth Company. The rates from Table 2 are not directly com-
parable with those on Table 1 since cholera deaths for different lengths
of time were counted in each case.
Snow's successful investigation was performed before the germ
theory of disease had been formally proposed, and thus modern methods
of bacteriology were not available to him. His achievement, however,
was aided by the characteristics of the disease itself. Cholera is
highly specific in that it is caused only by one offending agent —
Vibrio cholera — which enters the body via an oral route. The disease,
-------
Table 2. Cholera deaths in London houses served by
two water companies,
Cholera deaths
Water supply for
individual houses
Southwark and
Vauxhall Co.
Lambeth Co.
Total
Population (est.)
of houses served
106,309
139,472
245,781
Number
506
94
600
Rate/1000
4.76
0.67
2.44
Risk relative
houses served
the Lambeth
7.1
1.0
to
by
Co.
a,
Data for the 8-week period through August 26.
?John Snow. 1855. On the mode of communication of cholera, 2nd edition.
NJ
-------
265
moreover, has a brief incubation period. A cholera patient shows
symptoms within a few days of exposure and occasionally within hours.
A close relationship between cause and effect, both in specificity and
in elapsed time between one and the other, aided his analysis.
CHRONIC DISEASE EPIDEMIOLOGY
The incompletely understood causative links in most chronic disease
development stand out in contrast to cholera. Rather than a few hours
or days, chronic diseases such as heart disease, cancer, and chronic
respiratory disease develop over a period of years or decades. While
many factors would seem to influence whether or not an individual con-
tracts and dies of cholera, the presence of the infective agent is a
necessary precondition for development of the disease. In the case of
many chronic diseases, any of a number of environmental, genetic,
metabolic, or other factors, acting alone or in concert can give rise
to similar disease outcomes in different individuals.
There are a few exceptions to this general rule. Mesothelioma, for
example, a cancer of the pleural or peritoneal lining, has been closely
linked to exposure to asbestos fibers (Newhouse and Thompson 1965).
Angiosarcoma of the liver has been associated with occupational expo-
sure to vinyl chloride (Selikoff and Hammond 1975a) but may also be
caused by a drug, thorotrast, and probably by certain arsenicals. In
choosing populations for study, one must make an effort to obtain
information on every possible known variable which might affect the
outcome as expressed by morbidity or mortality rates. In practice,
this can be difficult or impossible to accomplish.
An epidemiologic determination of possible carcinogenic effects of
water-borne chlorinated organics will, no doubt, be performed under
conditions somewhat less ideal than those enjoyed by Snow in his cholera
study. The great value of the epidemiologic approach for Snow and for
us resides in the fact that human beings are the subjects of the study.
To evaluate results, there is no need to extrapolate from microbe or
laboratory animal to man, from high to low level exposure, or from both.
In a large number of cases, of course, when suggestive data from other
-------
266
lines of inquiry are available, it would be unethical to await the
results of an epidemiologic study before taking preventive action. But
if effects are observed in human populations, such action can proceed
with great confidence.
Aspects of carcinogenesis which must be carefully weighed when
designing epidemiologic studies include the long latent period for
development of cancer, influence of the age at first exposure to the
presumed carcinogen, the magnitude of risk posed by the carcinogenic
agent, and the presence of confounding factors.
Table 3 lists a number of substances which have been identified as
carcinogens and strongly linked to cancer of specific sites. Indicated
on the table are the so-called "latent" periods for most of the listed
carcinogens as well as the relative risk where it has been estimated.
The concept of "latent period" is used in different ways. In some
studies, it refers to the time after a single exposure or after the
start of a continuous exposure when a peak in the mortality rate is
observed. In other investigations, for example the ongoing studies in
Japan by the Atomic Bomb Casualty Commission, the term refers to the
average time between a single exposure and appearance of cancer in
affected members of the population. In cases where a specific carcino-
genic effect has been demonstrated, the latent period before a tumor
is observed is on the order of decades, not days or weeks. In most
epidemiologic investigations of carcinogenic effects of environmental
contaminants, estimation of dose — even of relative dose — is difficult,
as it usually must be made many years after the offending exposure.
In both animal studies and the few epidemiologic investigations
where the question has been raised, it appears that the length of the
latent period is inversely related to dose. In the study of Japanese
atomic bomb survivors, for example, Bizzozero and co-workers (1966)
observed that acute leukemia patients under 15 years of age who were
within 1,500 meters of the hypocenter, developed the disease 8.6 years,
on the average, after exposure; while those who were between 1,500 and
10,000 meters away from the blast center experienced a latent period
of 11.6 years. The data base is not yet extensive enough to feel fully
-------
Table 3. "Latent" period and relative risks for cancers
of known etiology
Carcinogen
Anatomical site
or type
"Latent" period
(years)
Risk relative to
unexposed populations
Reference
Mustard gas
Radiation
Respiratory tract
(various sites)
Leukemia
24
8-9
Wada et al. (1968)
Bizzozero et al.
Vinyl chloride
monomer
B-Naph thylamine
Benzidine
Asbestos and
smoking
Asbestos
Cigarettes
Liver angiosarcoma
Bladder
Bladder
Lung
Pleural mesothelioma
Lung
17-19
18
18
20
20
400-3000
90
250
5-20
(1966)
Selikoff and Hammond
(1975a)
Case et al. (1954)
Case et al. (1954)
Selikoff and Hammond
(1975b)
Rail et al. (1973)
U.S. Public Health
Service (1964)
to
-------
268
confident that the latent period is universally an inverse function of
dose, but most evidence is pointing in this direction.
The sensitivity of the human organism to carcinogens is a complex
function of age, with the age at exposure often an important determinant
of cancer development many decades later. Fraumeni and Hoover (in press),
in reviewing the epidemiologic literature on age-at-first exposure and
environmental carcinogenesis, report that the risk for some cancers
appears to be elevated among persons exposed at older ages (bladder
cancer in dye workers, nasal sinus cancer in nickel refiners, lung cancer
in asbestos workers) while in other situations risk is greatest in persons
exposed at younger ages (various cancers in atomic bomb survivors, lung
cancer in cigarette smokers)*
In designing studies to examine the effects of chlorinated organics
in water, the long latent periods for carcinogenesis and their possible
dependence on dose impose major boundary conditions. How should we
estimate dose, when the exposures of greatest significance to today's
disease experience took place 20, 30, or 40 years ago? To what extent
are present environmental measures of value? Are there reliable methods
for extrapolating present analyses back in time, given that we know
something about the industrial mix and pollution potential in years
gone by? Can we assume that the relative contamination of water between
regions in the past was similar to today's situation and that only the
absolute values have changed?
In addition to the changes in water supply which might have
occurred, migration of the observed populations must be carefully evalu-
ated and statistically controlled for. People who die of cancer were
not necessarily exposed to effective doses of carcinogens at their
place of death but may have moved from the area of crucial exposure
many years before. The cancer mortality patterns in Los Angeles, for
example, resemble those of the Midwest. The rates for many different
cancers in Miami, Florida, are similar to those in the New York/New
Jersey area. In indirect studies, where the primary unit of observa-
tion is the group, the migration statistic must be considered. In
direct studies, where individuals are followed, residential histories
are of prime importance.
-------
269
The problems inherent In dealing with diseases having long latent
periods are illustrated by a study of cancer mortality in Duluth. Mason
and co-workers (1974) did not observe cancer mortality rates consistent
with the hypothesis that water-borne asbestos fibers, as they are found
in the Duluth water supply, cause cancer. The asbestos fibers in Duluth*s
drinking water originate from taconite-ore tailings which have been
emptied into Lake Superior since 1955. Mason examined the cancer mor-
tality data through 1969, 14 years after exposure to water-borne asbestos
had started. He concludes: "The period of observation is short relative
to the latent period for occupationally induced carcinogenesis from
asbestos. A longer period of follow-up than was possible in this study
will be necessary before one can conclude that there is no cancer
hazard related to the drinking water supplies of Duluth and neighboring
communities." In this case the limits of information from an epidemio-
logic study preclude a definitive result. A positive result from an
epidemiologic study, in a case such as this, usually means that we have
waited too long. In the case of chlorinated organics in drinking water,
it is possible that adverse health effects will be shown to have
occurred in human populations, but if unambiguous epidemiologic evidence
is not forthcoming, a negative result can not be interpreted as proof
of no effect.
Table 3 lists the cancer risk of exposed persons relative to the
risk in the general population. In general, a high relative risk makes
the work of the epidemiologist relatively easy, since a high relative
risk implies a strong cause-effect relationship and it is unlikely that
other differences between groups (confounding factors) are responsible
for the observation. Recall that John Snow observed relative risks of
7 or 8 for cholera. As indicated in Table 3, workers exposed to high
levels of vinyl chloride monomer have a risk of contracting liver
angiosarcoma much greater than 400 times the general population — perhaps
about 3,000 times as great (Selikoff and Hammond 1975a). This carcinogen
first became suspect when only a few workers in a single vinyl chloride
plant developed this rare disease and the suspicions of the plant
physician were thereby aroused. Pleural mesothelioma due to asbestos
-------
270
exposure, with a relative risk well in excess of 250, belongs in the
same class (Rail et al. 1973). Lung cancer related to asbestos exposures
or to cigarette smoking have much lower relative risks. This is not an
indication of the relative public health significance of these diseases,
since this is determined not only by the relative risks but also by the
frequency of the disease and exposure of the general population. Lung
cancer, for example, is not an uncommon disease, and the relative risk
for smokers of 8 or 10 is a reflection of tens of thousands of premature
deaths.
Developing evidence for a cause-effect relationship, especially
with chronic diseases having multifactorial etiologies, becomes diffi-
cult when the relative risk is lower than 2.0. When this is the case,
the need for specific data on exposure and disease incidence in large
populations becomes ever greater, and obtaining the needed information
becomes exceedingly difficult. At this time, we do not know the relative
risk of cancer in populations exposed to chlorinated or other organics
found in drinking water, but if it is much below 2.0, epidemiologic
answers to the questions posed by this conference may not be forthcoming.
Let us now turn briefly to a few studies which have examined the
possible relationship between organics in drinking water and cancer
rates. The studies cited below use the descriptive approach. They
examine characteristics of populations with regard to their exposure to
suspect carcinogens and their cancer mortality experience. The best
known are the investigations by Harris and co-workers of the Environ-
mental Defense Fund, in which multiple regression techniques were used
to examine this relationship (Harris 1974; Page, Harris, and Epstein
1975).
In a short paper which has not been widely circulated, Mr. Lee
McCabe of the EPA Health Effects Research Laboratory at Cincinnati, Ohio,
performed a simple linear regression analysis of age-adjusted cancer
mortality rates against chloroform concentration in 50 U.S. cities which
were included as part of the EPA 80-City National Organics Reconnaissance
Survey (McCabe 1975, U.S. Environmental Protection Agency 1975). Cancer
death rates were not available for the cities not included in this
-------
271
analysis. McCabe recognized several problems with this study and con-
cluded that "these preliminary analyses only provide enough in the way
of results to stimulate more study of the problem." A third indirect
study was conducted by Dr. Thomas Mason at the Epidemiology Branch of
the National Cancer Institute (Mason 1975). Mason used county cancer
mortality data for the 10 counties which are coterminous with cities for
which data on the chloroform content of drinking water was available.
Cancer mortality rates were calculated for these counties and compared
with standard rates adjusted for several potentially confounding
variables.
These studies are valuable in that they represent the first steps
taken to evaluate the possible role of low-level carcinogens in drinking
water. All of the studies are lacking with respect to the pollution
indicator that was used. The Harris study employed a variable indi-
/•
eating the percentage of each Louisiana parish supplied with drinking
water from the Mississippi River, a source of presumably highly con-
taminated water. The McCabe and Mason studies used the amount of
chloroform in the drinking water of the cities they analyzed, based on
measurements made on a one-time grab sample from an EPA nationwide
survey of the drinking water of 80 cities. In one of the studies, can-
cers of all anatomical sites were grouped together, and there was no
consideration of other possible factors contributing to the observed
cancer rates (McCabe 1975). In another, possible confounding of the
results occurred because populations with large exposure differences
also differed ethnically and religiously (Harris 1974). The remaining
study used a small number of cities and only an extremely large effect
would have been observed as giving a statistically significant result.
The authors of these studies recognize several of these shortcomings and
point out the need for better exposure information and data on the
health status of exposed populations.
We should keep in mind that epidemiologic studies are comparative
studies. An initial task of the epidemiologist is to define populations
for comparison, either on the basis of differential mortality or mor-
bidity rates, or of gradients in exposure to known or suspected toxic
-------
272
substances. Both approaches require information of three types:
measures of response — that is, disease incidence and prevalence data;
estimates of dose; and information on as many other demographic and
exposure variables as may reasonably be thought to influence the develop-
ment of the disease or diseases of concern. Of these areas, we are
probably on weakest ground in the availability of good dose estimates.
EPA is currently expanding the 80-City survey of organics in drinking
water by including additional supplies for analysis and performing
year-round measurements to determine the influence of temperature, flow
fluctuations, and other variables on the concentration of these chemicals.
This effort will enable us to develop somewhat better estimates of dose
to populations presently exposed to the measured drinking water supplies
and hopefully provide information to allow assessment of historical
exposures.
In the future, it will be necessary to further coordinate activities
between epidemiologists, demographers, and water supply experts so that
dose estimates over time, measures of disease frequency, and potential
confounding variables are collected for the same groups of people. Only
with such cooperation can we develop the necessary data base to conduct
meaningful epidemiologic studies of the health effects of low levels of
chlorinated organics in drinking water.
ACKNOWLEDGMENTS
This work was performed while the author was detailed as a visiting
scientist to the National Cancer Institute from the U.S. Environmental
Protection Agency. Thanks are due to all in EPA and NCI who helped make
this arrangement possible. The author is especially grateful to
Dr. Robert Hoover of the National Cancer Institute, who provided several
helpful suggestions during development of the presentation and in
clarification of ambiguities in the manuscript.
REFERENCES
Bizzozero, 0. J., Jr., K. G. Johnson, and A. Ciocco. 1966. Radiation
related leukemia in Hiroshima and Nagasaki, 1946-1964. I. Distri-
bution, incidence, and appearance time. New Eng. J. Med. 274:
1095-1101.
-------
273
Case, R. A. M., M. E. Hosker, D. B. McDonald, and J. T. Pearson. 1954.
Tumours of the urinary bladder in workmen engaged in the manufac-
ture and use of certain dyestuff intermediates in the British
chemical industry. Brit. J. Industr. Med. 11: 75-104.
Fraumeni, J. F., and R. Hoover. Immunosurveillance and cancer: epidemi-
ologic observations. In Proceedings of the Symposium on Epidemiology
and Cancer Registries in the Pacific Basin. In press.
Harris, R. H. 1974. The implications of cancer-causing substances in
Mississippi River water. Environmental Defense Fund, Washington,
D.C.
Mason, T. 1975. Personal communication.
Mason, T., F. W. McKay, R. W. Miller. 1974. Asbestos-like fibers in
Duluth water supply. J. Am. Med. Assoc. 228: 1019-1020.
McCabe, L. J. 1975. Association between trihalomethanes in drinking
water and mortality. U.S. Environmental Protection Agency.
Unpublished paper.
Newhouse, M. L., and H. Thompson. 1965. Mesothelioma of pleura and
peritoneum following exposure to asbestos in the London area.
Brit. J. Industr. Med. 22: 261.
Page, T., R. H. Harris, and S. S. Epstein. 1975. Relation between
cancer mortality and drinking water in Louisiana. Unpublished
paper.
Rail, D., J. Churg, E. C. Hammond, A. M. Langer, W. J. Nicholson,
I. J. Selikoff, and Y. Suzuki. 1973. Proceedings of Conference
on Biological Effects of Asbestos. National Institutes of Health.
February 1.
Selikoff, I. J., and E. C. Hammond (eds.). 1975a. Toxicity of vinyl
chloride — polyvinyl chloride. Ann. N.Y. Acad. Sci. 246: 1-337.
Selikoff, I. J., and E. C. Hammond. 1975b. Multiple risk factors in
environmental cancer. In J. F. Fraumeni, Jr. (ed.). Persons at
high risk of cancer. Academic Press.
Snow, J. 1855. On the mode of communication of cholera, 2nd ed.
Churchill, London. Reproduced in Snow on cholera. 1936. Common-
wealth Fund. New York. Reprinted by Hafner. 1965. New York.
U.S. Environmental Protection Agency. 1975. Preliminary assessment
of suspected carcinogens in drinking water. April.
U.S. Public Health Service. 1964. Smoking and health: report of the
Surgeon General of the Public Health Service. Publication No. 1103.
-------
274
Wada, S., M. Michihiro, Y. Nishimoto, S. Kambe, and R. W. Miller. 1968,
Mustard gas as a cause of respiratory neoplasia in man. Lancet
1968: 1161-1163.
DISCUSSION
Nancy Stroup, Environmental Defense Fund. I want to know if EPA
or anyone else has been trying to develop retrospective studies on
chloro-organic dosages in drinking water.
Cantor. Retrospective in the sense that we are trying to evaluate
historical doses? Gordon Robeck might be doing this. Do you want to
answer?
Gordon G. Robeck, U.S. Environmental Protection Agency. We are
not doing this specifically. We are trying to get grantees to make
proposals, however.
Jerome R. McKersie, Wisconsin Department of Natural Resources.
You asked if there was any city that chlorinated part of their water
system and not other parts. I might suggest that for years Minneapolis,
Minnesota, had a ground water supply while its adjacent city, St. Paul,
obtained their drinking water from the Mississippi River. It might
be interesting to compare health statistics for these two cities. My
question is has there been any studies of bioaccumulation of chloro-
organic compounds in human fat between those who have cancer versus
noncancer deaths?
Cantor. To my knowledge there has been no such study to date. At
EPA we are looking into the possibility of correlating organic residues
in human fat with organic constituents in drinking water supplies.
-------
SESSION III. ENVIRONMENTAL TRANSPORT AND EFFECTS
William A. Brungs, Session Chairman
Environmental Research Laboratory
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
I saw and/or met many of you about two weeks ago at the Water
Pollution Control Federation Meeting in Miami Beach and at that time I
stewed all week because I had to present something. This week I have
really enjoyed myself because I know I don't have to say anything. At
least nothing that takes preparation. So I'm going to get right into
the program. The papers this afternoon are, of necessity, general in
nature because we're not trying to give you a thorough understanding of
the toxicity of residual chlorine to aquatic life but to give you a
summary of it and some of the detailed research that is being undertaken.
275
-------
THE TOXICITY OF CHLORINE TO FRESHWATER ORGANISMS
UNDER VARYING ENVIRONMENTAL CONDITIONS
Arthur S. Brooks and Gregory L. Seegert
Center for Great Lakes Studies
The University of Wisconsin-Milwaukee
Milwaukee, Wisconsin 53201
ABSTRACT
Chlorine enters freshwater systems from many sources. It enters
waters of variable chemical quality in a variety of chemical forms over
wide ranges of temperature and for varying periods of time. Each of
these factors is important in determining the toxicity of chlorine to
aquatic life.
This paper reviews studies which have been conducted under a myriad
of environmental conditions in attempts to quantify the toxicity of
chlorine to freshwater organisms. Included in the review are experiments
run in relatively clean waters and sewage effluents. Studies are also
included which involved continuous and intermittent chlorine applications
and tests conducted under a wide range of temperatures. Data from these
studies are reviewed in light of the conditions under which the tests
were conducted. A synthesized view is presented of the toxicity of
chlorine to freshwater biota in terms of the level and duration of expo-
sure, temperature, and the chemical nature of the water in which the
experiments were conducted.
INTRODUCTION
Chlorine is used in numerous industrial and water treatment processes
as a biocide for undesirable organisms. Chlorine is a very effective
biocide, not only for undesirable organisms, but for other forms of
aquatic life as well. Because of this lack of specificity great concern
has been expressed over the impact of chlorine on freshwater ecosystems.
Chlorine enters freshwater systems under a wide variety of environ-
mental conditions. These conditions include wide ranges of temperature,
277
-------
278
water quality, and variable application times and dose rates. Each of
these variables must be carefully evaluated in assessing the toxicity
of chlorine to freshwater organisms. This paper reviews the literature
in an attempt to evaluate the influence of the above-mentioned environ-
mental variables with respect to the toxicity of chlorine to freshwater
organisms. The review is not intended to be a complete compilation of
the chlorine literature. It draws only on studies that have presented
sufficient data to permit meaningful comparisons. Brungs (1973, 1975)
has compiled two excellent literature reviews to which the reader may
refer for more complete coverage.
FRESHWATER FISH
Species-related factors
Species composition. One of the most important factors in deter-
mining the impact of chlorine on a fish community is the species composi-
tion of that community. Cold water species (salmonids) have generally
been considered more sensitive to chlorine than are warm water species
(Brungs 1973, Basch and Truchan 1974). Recent evidence, however, indi-
cates that this distinction may not always be justified. Several species
of minnows have been found to have LC50 values approximating those of
salmonids (Ward 1975, as cited in Brungs 1975). This new information
suggests that in order to accurately determine the impact of chlorine
on a particular aquatic environment, representatives from the important
families making up the local community should be tested.
Size. There exists a general lack of continuity in the literature
on the effects of size on chlorine toxicity. Several authors (Fobes
1971, Wolf et al. 1975, Eren and Langer 1973) using a variety of species
have reported that smaller fish are more sensitive to chlorine than are
larger individuals of the same species. Rosenberger (1971), however,
found the opposite relationship in coho salmon, Oneorhynchus kitttsch.
Others (Wolf et al. 1975, Warren 1975) observed no effect resulting
from size. This lack of agreement reflects species differences and
probably differences in analytical techniques between authors.
-------
279
Effects on eggs, larvae, and reproductive behavior. Perhaps because
of the inherent difficulties few authors have attempted to determine the
effects of chlorine on fish eggs, larvae, or reproductive behavior.
Arthur and Eaton (1971) found that spawning in fathead minnows, Pimephales
pvomelas, was practically eliminated at 0.085 mg/1 of chlorine and that
the number of spawnings per female was significantly reduced at 0.043 mg/1.
The lowest concentration that had no effect in long-term tests was
0.016 mg/1. Carlson (1975, as cited in Brungs 1975) found that three-spine
sticklebacks, Gastevosteus asuleatus3 exposed for three and half months
to chloramine concentrations up to 0.0114 mg/1 exhibited no change in
reproductive behavior. Hughes (1973, as cited in Brungs 1973) found that
larval striped bass, Morone eaxat-i-lisj were considerably more sensitive
to chlorine than were fingerlings. Gehrs et al. (1974) studied the
effects of 4-chlororesorcinol and 5-chlorouracil on the hatching success
of carp, Cyprinus oarpio^ eggs. A significant inhibition of hatching
success was seen at most of the concentrations tested between 0.001 mg/1
and 10 mg/1. It has been found for some marine species that the egg
membrane affords considerable protection against chlorine (Alderson
et al. 1975). This has yet to be demonstrated in freshwater fish.
Avoidance. Avoidance to chlorine by fish has been observed in both
lab and field situations. Sprague and Drury (1969) observed that
rainbow trout, Salmo gairdnem,_, placed in a test tank offering choices
between unchlorinated and chlorinated water clearly avoided total residual
chlorine (TRC) concentrations of 0.001, 0.01, and 1.0 mg/1. Unexplicably
they showed a preference for the chlorinated side at 0.1 mg/1. In
another lab study Borgardus (1975) found that mimic shiners, Notropis
vo'Luaeltuss river shiners, N. Blenni-us3 and bullhead minnows, P-tmephales
vigilaXj when maintained in a narrow rectangular tank having a laminar
flow of monochloramine down one side actively avoided that side. The
lowest concentrations avoided were 0.005, 0.150, and 0.05 mg/1 for the
mimic shiner, river shiner, and bullhead minnow, respectively. Meldrin
et al. (1975, as cited in Brungs 1975) found that several species of
estuarine fish could avoid chlorine. The avoidance concentrations of
these fish were generally inversely related to temperature and light
levels.
-------
280
Several field studies have shown that fish can detect and sometimes
avoid chlorinated effluents. Fava and Tsai (1973, as cited in Tsai 1975)
determined that blacknose dace, Rh-inehthys atratulus3 could discriminate
and avoid chlorinated sewage effluents containing TRC levels as low as
0.01 mg/1. Several authors (Massey 1972, Basch 1971) have reported
observing fish leaving discharge channels during chlorination periods
at power plants. Conversely, Dickson and Stauffer (1974) reported that
most of the resident fish population in the power plant discharge they
surveyed remained in the channel during the thrice-daily chlorination
periods. They hypothesized that the TRC chlorine levels in the channel
were not high enough to evoke an avoidance reaction. The lack of
species numbers and diversity repeatedly observed in rivers below waste
treatment plants (Tsai 1968, 1970) probably reflects an avoidance of
these areas by fish.
It is apparent that when given a clear choice (as in lab studies)
or when chlorine is continuously discharged (as with waste treatment
plants) fish can actively avoid low levels of chlorine. However, whether
fish can safely avoid intermittently discharged effluents has yet to be
demonstrated. Data compiled by Truchan (1975) showed that fish kills
have occurred as a result of the intermittent discharge of chlorine.
These incidences suggest that either the intermittency of chlorination
or a temporary misorientation by the fish can override any avoidance
response which they would normally exhibit.
Exposure time
One of the most important factors determining the toxicity of a
given level of chlorine is the exposure time. A comparison of several
studies on rainbow trout illustrate this importance. Safe concentrations
for continuous exposure are generally considered to be from 0.002 to
0.005 mg/1 (Brungs 1973, Basch and Truchan 1974, Merkins 1958). Basch
et al. (1971), Esvelt et al. (1971), Sprague and Drury (1969), and
Wolf et al. (1975) cite 96-hour LC50 values near 0.1 mg/1. One mg/1
was required to kill rainbow trout exposed only four hours (Sprague and
Drury 1969) and in one-half hour tests at 10°C an LC50 of 2.0 mg/1 was
found (Seegert et al. 1975).
-------
281
Blacknose dace exposed for one-half hour to 0.74 mg/1 of free
chlorine suffered only 4 percent mortality but suffered 72 percent
mortality after a two-hour exposure (Tsai and Tompkins 1974). When the
dace were exposed to 0.15 mg/1 of free chlorine for six hours only
10 percent mortality occurred, whereas doubling the exposure time to
12 hours increased the mortality approximately eightfold to 83 percent.
Similar results occurred for dace exposed to chloramines. This study
by Tsai and Tompkins (1974) also illustrates the interrelationship
between exposure time and chlorine levels. At 0.15 mg/1 of free chlorine,
a 12-hour exposure caused 83 percent mortality; while at 1.38 mg/1, a
40-minute exposure caused 65 percent mortality; and at 6.6 mg/1, only
eight minutes were needed to kill all the dace. Again, similar results
were found when chloramines were tested.
Stober and Hanson (1974) tested chinook, Oncovhynckus tshcaoytsoha.,
and pink salmon, 0. gorbuscha, at exposure times ranging from 7.5 minutes
to 60 minutes. Chinook salmon exposed to chlorine for 7.5 and 15 minutes
had LC50 values of 0.5 mg/1 with the LC50 values decreasing to 0.25 mg/1
for the 30- and 60-minute exposures. The LC50 value for pink salmon
exposed for 7.5 minutes was 0.5 mg/1 and dropped to 0.25 mg/1 for the
15-, 30-, and 60-minute exposures. Both species had an LC50 of 0.045
mg/1 after exposure to fluctuating chlorine concentrations for two hours.
When both species were subjected to a variety of temperature shocks in
addition to the chlorine the LT50 (median lethal time) generally
decreased with increasing exposure time at each test temperature. In
another study Eren and Langer (1973) found that when the exposure time
of Tilapia ccurea was shortened from 18 to 4 hours the lethal total
residual chlorine concentrations decreased about 20 percent.
Temperature
Cairns et al. (1975) concluded that there was little information
in the literature regarding the effects of temperature on chlorine
toxicity. Recently, however, several papers have appeared indicating
a general trend of increasing sensitivity to chlorine of fish at higher
temperatures. Two types of studies have emerged. The first type tests
-------
282
a species over a wide range of acclimation temperatures. Seegert et al.
(1975) found that the 30-minute LC50 values for the yellow perch,
Perea flavesoens3 decreased from 7.7 mg/1 to 1.0 mg/1 at acclimation
temperatures of 10 and 25°C, respectively. Similarly, Eren and Langer
(1973) found that Tilapia aurea was more sensitive to chlorine at higher
acclimation temperatures. In contrast, Bass and Heath (1975a) observed
that the 96-hour LC50 for bluegills, Lepomis maavoeh-Lrus , remained
constant at acclimation temperatures from 6 to 32°C. They did find,
however, that at the higher temperatures the times to death were shorter.
Heath (1974) also found that temperature had little effect on the
lethal concentrations of chlorine to bluegills.
The second type of temperature-chlorine study involves exposing
the fish to a variety of temperature shocks over a range of chlorine
concentrations. Stober and Hanson (1974) found that pink and chinook
salmon when subjected to a variety of heat shocks of 3 to 10°C above
ambient demonstrated a decreased resistance to chlorine. They concluded,
however, that the chlorine level and not the addition of heat had acted
as the principal lethal agent. Thatcher et al. (1975) found that there
was no difference in the 96-hour LC50 values of brook trout, Salvel-inus
font-indiesy acclimated to 7, 10, 15, and 20°C, when they were tested at
10 and 15°C. However, the LC50 dropped significantly when these fish
were tested at 20°C. This indicates that it is the test temperature and
not the acclimation temperature that determines the toxicity of the
chlorine. Wolf et al. (1975) found that rainbow trout must experience
thermal stress greater than 10°C before a combined effect with chlorine
could be detected.
The above results suggest that fish have some temperature range
within which temperature has little effect on chlorine toxicity. How-
ever, outside this range, which is undoubtedly species-dependent, test
temperatures and/or shock temperatures can increase their sensitivity
to chlorine.
-------
283
Water quality and chemistry
The quality of the water in which an organism is exposed to chlorine
is very important in determining the response of that organism to chlo-
rine. The pH of the water and the concentrations of ammonia and organic
materials are responsible for determining the specific chlorine compounds
that will be present.
PH. Although pH plays an important role in the equilibria constants
between the various chlorine species (White 1972, Draley 1972) little
information exists on the effect of pH on chlorine toxicity. Tsai and
Tompkins (1974) working over a pH range of 6.8 to 7.6 found that if
there was any effect by pH it was masked by much larger effects caused
by varying chlorine levels and exposure times. Similarly, Warren
(1975, as cited in Brungs 1975) observed no change in the LC50 values of
coho salmon and cutthroat trout, Salmo elarkiy exposed to chlorine over
a pH range of 7.5 to 8.1. Merkins (1958) reported that rainbow trout
were somewhat more resistant to chlorine at pH 6.3 than at pH 7.0.
However, he attributed this difference not to pH per se but to the
relative proportions of free chlorine and chloramines which existed at
the two pH values. The primary role of pH in chlorine toxicity is
probably related to how it affects the relative proportions of various
chlorine species and not to a direct effect on the organisms.
Chlorinated organic compounds. Chlorinated organics are found
wherever chlorine is added to waters with relatively high concentrations
of organics such as in sewage treatment effluents (Jolley 1973, 1974;
Glaze et al. 1973). Jolley et al. (1975) have also found chlorinated
organics in the discharge waters of power plants. Despite this apparent
ubiquitous occurrence, the study of the toxicity of chloro-organics is
still in its infancy.
Recently several authors have attempted to determine the toxicities
of various chloro-organic compounds to aquatic life (Gehrs and Jolley
1975, Leach and Thakore 1975). A study by the Manufacturing Chemists
Association (1972, as cited in Brungs 1975) tested several chloro-
organic compounds and found 96-hour LC50 values for fathead minnows
-------
284
ranging from 0.01 to 10 mg/1. As previously discussed, Gehrs et al.
(1974) determined that 4-chlororesorcinol and 5-chlorouracil signifi-
cantly affected the hatching success of carp eggs.
In view of the lack of specific information on the toxicities of
chloro-organic compounds further bioassay work in this area seems
justified.
Synergistic effects with chemical pollutants. Synergistic effects
with other chemical pollutants have been largely ignored. Allen (1946)
reported that chlorine combined with thiocyanate in industrial wastes
to form cyanogen chloride which was highly toxic to rainbow trout. A
similar reaction with potassium thiocyanate was reported by Schaut
(1939). Hoss et al. (1974, as cited by Brungs 1975) reported on the
effects of copper and chlorine. As with the case of chlorinated organic
compounds, additional study is required in this area, especially in a
situation where other toxic substances are known to occur simultaneously
with chlorine.
Free versus combined chlorine. While the toxicities of free and
combined chlorine are considered to be of the same order of magnitude
(Brungs 1973, Merkins 1958) there is some disagreement regarding the
specific toxicities of the two fractions. Most authors (Eren and Langer
1973, Merkins 1958, and Rosenberger 1971) consider free chlorine to be
more toxic and swifter acting. However, others (Westfall 1950, Holland
et al. 1960) consider chloramines to be more toxic. Holland et al.
(1960) found that dichloramine in particular was considerably more toxic
than free chlorine to the coho salmon. However, some questions on
the analytical methods they used and the unlikelihood of a solution
remaining 100 percent dichloramine at a pH of 7.6 raise some questions
about the exact levels that were toxic. Tsai and Tompkins (1974) found
that in the blacknose dace the relative toxicities of free chlorine and
chloramines were dependent on the chlorine concentrations being tested.
At concentrations greater than about 0.5 mg/1 they found chloramines to
be more toxic while at concentrations less than about 0.5 mg/1 free
chlorine was more toxic.
-------
285
Mechanisms of toxic action
Many authors including Cairns (1975), Dandy (1972), Rosenberger
(1971), and Bass and Heath (1975b) have concluded that the gills are
the primary site of chlorine toxicity. Bass and Heath (1975b) found
that rainbow trout exposed thrice-daily to free chlorine exhibited
increased mucous production and damage to respiratory epithelium which
was apparent in histological sections of gill tissue. Measurement of
several physiological parameters (blood pO£ and pH, heart and breathing
rates) indicated that the changes seen in these parameters during
chlorination were only partially restored to normal levels between
chlorinations. With each succeeding chlorination pulse the amount of
recovery became less until death finally ensued. They concluded that
the primary mode of chlorine toxicity is gill damage resulting in asphyx-
iation. Conversely, Fobes (1971) found no change in the respiration
rate of gill tissue excised from white suckers, Catastomus oormevQonni3
subsequent to the exposure of the suckers to a lethal dose of chlorine.
He concluded that death was not from gill damage and that the gills
were not the primary site of chlorine toxicity. He hypothesized that
chlorine enters through the gills and then somehow affects the nervous
system of fishes. Wolf et al. (1975) also concluded that the gills
were not the primary site of chlorine toxicity. In contrast to the
findings of Bass and Heath reviewed above which had indicated a summation
effect with regard to chlorine toxicity, Wolf and co-workers hypothesized
that chlorine toxicity is an all or none effect.
The answer to the apparent disparity in findings between authors
may be that chlorine toxicity is not simply acting in one mode or at one
site. Rosenberger (1971) has concluded that free chlorine and mono-
chloramine have different modes of action. Holland et al. (1960) also
observed differences in toxic patterns between the various chlorine
species. Eaton (1973) found that monochloramine caused severe methemo-
globinemia in fathead minnows. Free chlorine, however, did not have
this effect. Similarly, bluegills exposed to free chlorine had normal
levels of methemoglobin (Bass and Heath 1975a). Lending further support
to the multimode or site theory are the previously discussed findings
-------
286
of Tsai and Tompkins (1974) who found that the relative toxicities of
free chlorine and chloramines varied with the chlorine concentrations.
FRESHWATER INVERTEBRATES
The literature concerning the effects of chlorine on freshwater
invertebrates is not nearly as extensive as is that for fishes. Several
recent unpublished reports and theses reviewed by Brungs (1975) have
increased our knowledge considerably within the past year. Several of
the most recent works were not available to the authors for inclusion
here. Even with this new information a comparison of experimental
results relating to a given species under varying environmental condi-
tions is difficult.
Exposure level
In general, the range of 96-hour LC50 values reported in the litera-
ture for invertebrates is much more narrow than those reported for fish.
With few exceptions, the 96-hour LC50 for invertebrates is in the range
of about 0.10 mg/1. Among the more sensitive invertebrates tested to
date are the rotifers with 1-, 4-, and 24-hour LC50 values of 0.032,
0.027, and 0.013, respectively (Grossnickle 1974). Gregg (1974, as
cited by Brungs 1975) investigated several invertebrate species over a
range of chlorine concentrations from 0.01 to 0.1 mg/1. He observed
that mayflies were the most sensitive group tested while beetle larvae
were the most resistant. Other studies have indicated that crayfish
are quite resistant to chlorine. Schneider (1975, as cited by Brungs
1975) observed 96-hour LC50 values for crayfish of 0.96 mg/1. Arthur
(1971, as cited by Brungs 1973) also indicated that crayfish, Oroaneotes
wLrilis3 were the least sensitive to chlorine among the organisms tested
with 7-day TL50 values greater than 0.78 mg/1.
Exposure time
Chronic exposures of invertebrates to chlorine have been conducted
for a number of species. Arthur et al. (1975, as cited by Brungs 1975)
-------
287
reported that the lowest mean total residual chlorine (TRC) concentra-
tion having measurable adverse chronic effects were 0.19 mg/1 for an
amphipod and 0.010 mg/1 for Daphnia rnagna. The same study also cites
the highest mean TRC concentrations having no measurable effect. These
values were 0.021 mg/1 and 0.002 to 0.004 mg/1 for the amphipod and
D. magnay respectively. Carlson (1975, as cited by Brungs 1975) observed
no effect on the standing crop of amphipods following a three and a half
month exposure to chloramine at concentrations of 0.001, 0.0043, and
0.0114 mg/1.
A comparison of LC50 values which were determined following expo-
sures to chlorine for varying lengths of time demonstrates the importance
of the duration of exposure. Grossnickle (1975) exposed the rotifer
Keratella cochlearis to mono chloramine for periods of 1, 4, and 24 hours
at 15°C. He observed decreasing LC50 values of 0.032, 0.027, and 0.0135
mg/1 as exposure times increased. Beeton et al. (1975) reports a 96-hour
LC50 value for the copepod Cyclops biauspidatus thomasi of 0.069 mg/1
for a mixture of free chlorine and monochloramine at 15°C. Research
conducted in the same laboratory under similar conditions determined
that the 30-minute LC50 value for C. b. thomasi was 15.61 mg/1 at 15°C
(Latimer et al. 1975). Gregg (1974, as cited by Brungs 1975) observed
that the 4- to 7-day LC50 values determined for mayflies, stoneflies,
sowbugs, amphipods, caddisflies, water pennies, and snails were inter-
mediate between the LC50 values for intermittent exposures based on
maximum or peak concentrations and those determined based on mean
concentrations.
Effects of temperature
Gregg (1974, as cited by Brungs 1975) reported that temperature
exerted an influence on chlorine toxicity although the degree of effect
varied for different species and for some there was no observed tem-
perature effect. Latimer et al. (1975) reported that the 30-minute
LC50 value for LimnoQalanus maarwnis was 1.54 mg/1 at both 5 and 10°C.
Cyclops bicuspidatus thomasi was tested at 10, 15, and 20°C by the same
authors and was found to have 30-minute LC50 values of 14.68, 15.61, and
-------
288
5.76 mg/1, respectively, indicating a definite temperature effect
between 15 and 20°C.
Water quality
The influence of water quality on the toxiclty of chlorine to
freshwater invertebrates has not been extensively studied. Beeton et al.
(1975) reported a 96-hour LC50 value for C» 2>. thomasi, exposed to mono-
chloramine of 0.089 mg/1 while the LC50 was 0.069 mg/1 for a solution
which was predominantly free chlorine.
Gehrs and Jolley (1975) have investigated the effect of 5-chlorouracil
and 4-chlororesorcinol, two stable compounds found in relatively high
concentrations in chlorinated sewage effluents, on survivorship and
maturation of Daphnia magna. Only a slight decrease in survivorship
occurred for 5-chlorouracil over a concentration range between 0.01 and
10.0 mg/1. A definite dose-response relationship was noted between
survivorship and the concentration of 4-chlororesorcinol. The greatest
increase in mortality was observed between 0.1 and 1.0 mg/1. The authors
also report a general delay in the onset of reproduction with 5-chlo-
rouracil and a decrease or total elimination of reproduction with
4-chlororesorcinol.
FRESHWATER ALGAE
Research on the toxicity of chlorine to freshwater algae is sorely
lacking. Brook and Baker (1972) conducted studies on the phytoplankton
of the St. Croix River, Minnesota/Wisconsin. They noted that photo-
synthetic rate and respiration were depressed to 50 percent of control
values at a chlorine concentration calculated by serial dilution to be
0.320 mg/1. At the highest concentration actually measured, 2.7 mg/1,
both photosynthetic rate and respiration were reduced to zero. Tempera-
tures ranged from 32 to 36°C at the time of chlorination to 23 to 25°C
during sample incubation. Although it is not specifically stated in the
paper, the exposure time to the chlorine is assumed to be slightly in
excess of the 2- to 4-hour incubation period.
-------
289
Research currently underway by the authors of this report indicate
that the photosynthetic rate of Lake Michigan phytoplankton is reduced
25 to 100 percent following 30-minute exposures to chlorine ranging in
concentration between 0.01 and 1.375 mg/1. The most significant reduc-
tions occurred at concentrations above 0.5 mg/1. No recovery was
observed up to 24 hours after the initial chlorine exposure. Reductions
in chlorophyll a concentrations following the 30-minute chlorine expo-
sures increased with increasing chlorine concentrations. As in the case
of photosynthetic rates, the greatest chlorophyll a reductions were
observed at chlorine concentrations above 0.5 mg/1. Judging from the
fact that both chlorophyll a concentrations and photosynthetic rate
are reduced with increasing chlorine concentrations and that no recovery
in either parameter was noted, it seems quite possible that chlorophyll
a. destruction is the principal cause of the reduced photosynthetic rates.
Other studies on the effects of chlorine on marine and estuarine
phytoplankton have shown photosynthetic rate reductions similar to those
discussed above (Hamilton et al. 1970; Hirayama and Hirano 1970;
Carpenter et al. 1972; Brooks et al. 1974; Gentile 1972, as cited by
Brungs 1973; and Fox and Moyer 1975). A specific discussion of these
reports is not within the scope of this freshwater review, however the
results obtained from these studies are helpful in evaluating the overall
algae-chlorine picture. It appears that algae are extremely sensitive to
chlorine especially at concentrations exceeding 0.5 mg/1. Evidence
further suggests that exposure times on the order of a few minutes may
be all that is required to effect irreversible damage.
CONCLUSIONS
The importance of environmental variables such as temperature,
water quality, and exposure time and dose rate have been demonstrated
for a number of organisms. Unfortunately, many studies have been con-
ducted and reported without sufficient detail regarding experimental
conditions to permit complete evaluation. Researchers should be
encouraged to include as much pertinent environmental information as
possible in research reports so that these important variables may be
considered in evaluating the responses of organisms to chlorine.
-------
290
Based on the information presented in this review, the following
may be concluded:
1. The response of freshwater organisms to chlorine is species-
dependent. Sweeping statements concerning the response of
major taxonomic groups to chlorine should be avoided unless
specific information exists for important families and genera
of the major taxon.
2. The life stage and size of an organism must be considered in
evaluating chlorine toxicity. Due to conflicting information
in the literature, no assumptions should be made relative to
the responses of different-sized organisms without specific
data.
3. The period of time that an organism is exposed to chlorine
and the concentration of chlorine are critical in determining
the final response of that organism. Exposure times should be
carefully noted and chlorine measurements determined by an
accepted standard procedure.
4. The effect of temperature on the response of freshwater organisms
to chlorine appears to be species-dependent and also dependent
on the specific range of temperature concerned. At lower tem-
peratures there appears to be little temperature effect, while
at higher temperatures the influence of temperature appears
to be increased.
5. The role of chemical water quality in determining chlorine
toxicity is important in determining what chlorine species will
be present. No definitive statement can be made with regard
to which chlorine species are more toxic. Some evidence suggests
that different forms of chlorine may affect organisms by
different modes of action depending on the concentration range.
6. The avoidance of chlorine by fish has been demonstrated in
laboratory tests. While this is also the case in field situa-
tions where the chlorine is discharged continuously, the
avoidance of fish to chlorine intermittently discharged is
poorly defined and needs further study. Consequently, the
-------
291
mode of chlorine application and the physical constraint of
the discharge area should be considered when evaluating the
impact of any chlorinated discharge.
REFERENCES
Alderson, R. 1970. Effects of low concentrations of free chlorine on
eggs and larvae of plaice, Pleuraneetes platessa L. Report No. FIR:
MP/70/E-3. Technical Conferences on Marine Pollution and Its Effects
on Living Resources and Fishing. Food and Agriculture Organization
of the United Nations. Rome, Italy. December 9-18, 1970.
Allen, L. A., N. Blezard, and A. B. Wheatland. 1946. Toxicity to fish
of chlorinated sewage effluents. Surveyor 105: 298.
Arthur, J. W. 1971. Progress Report National Water Quality Laboratory,
U.S. Environmental Protection Agency. Duluth, Minnesota.
Arthur, J. W., and J. G. Eaton. 1971. Chloramine toxicity to the
amphipod, Gammarus pseudoliimaeus 3 and the fathead minnow, Pimephales
promelas. J. Fish. Res. Bd. Can. 28: 1841-1845.
Arthur, J. W., et al. 1975. Comparative toxicity of sewage-effluent dis-
infection to freshwater aquatic life. Ecological Research Series.
U.S. Environmental Protection Agency. Washington, D.C. In press.
Basch, R. E. 1971. Memo on Campbell Power Plant chlorine measurement.
Michigan Water Resources Comm. Unpublished.
Basch, R. W., M. E. Newton, J. G. Truchan, and C. M. Fetterolf. 1971.
Chlorinated municipal waste toxicities to rainbow trout and fathead
minnows. Water Pollut. Control Res. Ser. No. 18050GZZ. U.S. Envi-
ronmental Protection Agency.
Basch, R. E., and J. G. Truchan. 1974. Calculated residual chlorine
concentrations safe for fish. Tech. Bull. 74-2. Michigan Water
Resources Comm., Dept. Natural Resources. 29 p.
Bass, M. L., and A. G. Heath. 1975a. Toxicity of intermittent chlorine
exposure to bluegill sunfish, Lepomis maaroahirus: interaction
with temperature. Assoc. Southeast Biol. Bull. 22: 40..
Bass, M. L., and A. G. Heath. 1975b. Physiological effects of inter-
mittent chlorination on fish. Am. Zool. 15: 818.
Beeton, A. M., P. K. Kovacic, and A. S. Brooks. 1975. Effects of
residual chlorine and sulfite reduction on Lake Michigan inverte-
brates. Ecological Research Series. U.S. Environmental Protection
Agency, Washington, D.C. In press.
-------
292
Bogardus, R. B., D. B. Boies, and D. A. Etnier. 1975. The avoidance
responses of selected Wabash River fishes to monochloramine. J.
Water Pollut. Control Fed. In press.
Brook, A. J., and A. L. Baker. 1972. Chlorination at power plants:
impact on phytoplankton productivity. Science 176: 1414-1415.
Brooks, A. S., R. A. Smith, and L. D. Jensen. 1974. Phytoplankton and
primary productivity. In L. D. Jensen (ed.) Environmental responses
to thermal discharges from the Indian River Station, Indian River,
Delaware. Electric Power Research Institute Pub. No. 74-049-00-3.
205 p.
Brungs, W. A. 1973. Effects of residual chlorine on aquatic life.
J. Water Pollut. Control Fed. 45: 2180-2193.
Brungs, W. A. 1975. Effects of wastewater chlorination on freshwater
aquatic life. Prepublication Report. Environmental Research Labor-
atory, U.S. Environmental Protection Agency. Duluth, Minnesota.
43 p.
Cairns, J., A. G. Heath, and B. C. Parker. 1975. Temperature influence
on chemical toxicity to aquatic organisms. J. Water Pollut.
Control Fed. 47: 267-280.
Carlson, A. R. 1975. Reproductive behavior to the threespine stickle-
back exposed to chloramines. M. S. thesis. Oregon State University.
Corvallis.
Carpenter, E. J., B. B. Peck, and S. J. Anderson. 1972. Cooling water
chlorination and productivity of entrained phytoplankton. Mar.
Biol. 16: 37-40.
Dandy, J. W. T. 1972. Activity response to chlorine in the brook trout,
Salvelinus fontinaUs (Mitchill). Can. J. Zool. 50: 405-410.
Dickson, K. L., and J. Stauffer. 1974. Observations on the effects of
chlorination on fish at Appalachian Power Company's Glen Lyn,
Virginia, Plant, January 10, 1974. Unpublished Memo. Virginia
Polytechnic Institute and State University. Blacksburg, Virginia.
Draley, J. E. 1972. The treatment of cooling waters with chlorine.
ANL/ES-12. Argonne National Laboratory, Argonne, Illinois. 11 p.
Eaton, J. W., C. F. Kolpin, C. W. Kiellstrand, and H. S. Jacob. 1973.
Chlorinated urban water: a cause of dialysis-induced hemolytic
anemia. Science 181: 463-464.
Eren, Y., and Y. Langer. 1973. The effect of chlorination on tilapia
fish. Bamidgeh 25: 56-60.
-------
293
Esvelt, L. A., W. J. Kaufman, and R. E. Selleck. 1971. Toxicity
removal from municipal wastewaters. Vol. IV. A study of toxicity
and biostimulation in San Francisco Bay-Delta waters. SERL Report
No. 71-7. Sanitary Engineering Research Laboratory, University
of California. Berkeley, California. 224 p.
Fava, J. A., and Chu-fa Tsai. 1973. Fish avoidance of chlorinated
effluents. Program Report 1973. Water Resources Research Center,
University of Maryland. College Park, Maryland.
Fobes, R. L. 1971. Chlorine toxicity and its effect on gill tissue
respiration of the white sucker, Catostomus commeTSoni (Lacepede).
Thesis. Michigan State University. East Lansing, Michigan.
Fox, J. L., and M. S. Moyer. 1975. Effect of power plant chlorination
on estuarine productivity. Chesapeake Sci. 16: 66.
Gehrs, C. W., and R. L. Jolley. 1975. Chlorine-containing stable
organics: new compounds of environmental concern. Verh. Internat.
Verein. Limnol. 19: 2185-2188.
Gehrs, G. W., L. D. Eyman, R. L. Jolley, and J. E. Thompson. 1974.
Effects of stable chlorine-containing organics on aquatic environ-
ments. Nature 249: 675-676.
Gentile, J. H., et al. 1972. Unpublished data. National Marine Water
Quality Laboratory, U.S. Environmental Protection Agency. West
Kingston, Rhode Island.
Glaze, W. H., J. E. Henderson, IV, J. E. Bell, and V. A. Wheeler. 1973.
Analysis of organic materials in wastewater effluents after chlo-
rination. J. Chromatogr. Sci. 11: 580-584.
Gregg, G. C. 1974. The effects of chlorine and heat on selected stream
invertebrates. Ph.D. thesis. Virginia Polytechnic Institute
and State University. Blacksburg, Virginia. 309 p.
Grossnickle, N. E. 1974. The acute toxicity of residual chloramine to
the rotifer, Keratella aoahlearis (Goose) and the effect of
dechlorination with sodium sulfite. M.S. thesis. University of
Wisconsin-Milwaukee. Milwaukee, Wisconsin. 39 p.
Hamilton, D. H., Jr., D. A. Flemer, C. W. Keefe, and J. A. Mihursky.
1970. Effects of chlorination on estuarine primary production.
Science 169: 197-198.
Heath, A. G. 1974. A preliminary investigation of chlorine toxicity
to fish and macroinvertebrates: interactions with temperature.
Final Report on Grant with American Electricity Production Service
Corp. Virginia Polytechnic Institute and State University.
Blacksburg, Virginia.
-------
294
Hirayama, K., and R. Hirano. 1970. Influence of high temperature and
residual chlorine on marine phytoplankton. Mar. Biol. 7: 205-213.
Holland, G. A., J. E. Lasater, E. D. Newmann, and W. E. Eldrldge. 1960.
Toxic effects of organic and inorganic pollutants on young salmon
and trout, p. 198-214. In Res. Bull. No. 5. Dept. of Fish. State
of Washington.
Hoss, D. E., L. C. Coston, J. P. Baptist, and D. W. Engel. 1974. Effects
of temperature, copper and chlorine on fish during simulated
entrainment in power-simulated plant condenser cooling systems.
Symposium on Physical and Biological Effects on the Environment of
Cooling Systems and Thermal Discharges at Nuclear Power Stations.
Oslo, Norway. August 26-30.
Hughes, J. S. 1973. Acute toxicity of thirty chemicals to striped bass
(Morane saxatilis). Presentation at the Western Assoc. of State
Game and Fish Comm. Salt Lake City, Utah.
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants. ORNL/TM-
4290. Oak Ridge National Laboratory. Oak Ridge, Tennessee. 342 p.
Jolley, R. L. 1974. Determination of chlorine-containing organics in
chlorinated sewage effluents by coupled 3^C1 tracer—high-resolution
chromatography. Environ. Lett. 7: 321-340.
Jolley, R. L., C. W. Gehrs, and W. W. Pitt, Jr. 1975. Chlorination of
cooling water: a source of chlorine-containing organic compounds
with possible environmental significance. Proc. Fourth National
Symposium on Radioecology. Corvallis, Oregon. May 12-14. In press.
Latimer, D. L., A. S. Brooks, and A. M. Beeton. 1975. The toxicity of
30-minute exposures of residual chlorine to the copepods Limnoaaianus
maorurus and Cyclops biauspidatus thomasi. J. Fish. Res. Bd. Can.
32. In press.
Leach, J. M., and A. N. Thakore. 1975. Isolation and identification of
constituents toxic to juvenile rainbow trout (Salmo gairdneri) in
caustic extraction effluents from Kraft pulpmill bleach plants.
J. Fish. Res. Bd. Can. 32: 1249.
Manufacturing Chemists Association. 1972. The effect of Chlorination
on selected organic chemicals. Wat. Poll. Contr. Res. Series
12020 EXG 03/72. U.S. Environmental Protection Agency. Washington,
D.C.
Massey, A. 1972. A survey of chlorine concentr.ations in Consumer's
Power Company's Big Rock Point power plant discharge channel.
Michigan Water Research Commission Report. Unpublished.
-------
295
Meldrin, J. W., J. J. Gift, and B. R. Petrosky. 1974. The effect of
temperature and chemical pollutants on the behavior of several
estuarine organisms. National Technical Information Service,
PB-239, 347. 129 p.
Merkens, J. C. 1958. Studies on the toxicity of chlorine and chlora-
mines to the rainbow trout. Water Waste Treat. J. 7: 150-151.
Rosenberger, D. R. 1971. The calculation of acute toxicity of free
chlorine and chloramines to Coho salmon by multiple regression
analysis. Thesis. Michigan State University. East Lansing,
Michigan. 33 p.
Schaut, G. G. 1939. Fish catastrophes during droughts. J. Am. Water
Works Assoc. 31: 771.
Schneider, M. J., et al. 1975. Aquatic physiology of thermal and
chemical discharges. In Environmental effects of cooling systems
at nuclear power plants. Internat. Atomic Energy Agency, Vienna.
IAEA-SM-187/15. Battelle-Pacific Northwest Laboratory. Richland,
Washington.
Seegert, G. L., A. S. Brooks, and D. L. Latimer. 1975. The effects of
a 30-minute exposure of selected Lake Michigan fishes and inverte-
brates to residual chlorine. Proc. Workshop on an Assessment of
Technology and Ecological Effects of Biofouling Control Procedures
at Thermal Power Plant Cooling Water Systems. Johns Hopkins
University. Baltimore, Maryland. June 16-17. In press.
Sprague, J. B., and D. E. Drury. 1969. Avoidance reactions of salmonid
fish to representative pollutants, p. 169. In Advances in water
pollution research. Proc. 4th Internat. Conf. Water Pollut. Res.
Pergamon Press. London, England.
Stober, Q. J., and C. H. Hanson. 1974. Toxicity of chlorine and heat
to pink (Onoorhynohus gorbuscha) and chinook salmon (0. tshccaytsoha).
Trans. Am. Fish. Soc. 103: 569-576.
Thatcher, T. 0., M. J. Schneider, and E. G. Wolf. 1975. Bioassays on
the combined effects of chlorine, heavy metals, and temperature on
fishes and fish food organisms. Bull. Environ. Contam. Toxicol.
In press.
Truchan, J. G. 1975. Toxicity of residual chlorine to freshwater fish:
Michigan's experience. Proc. Workshop on an Assessment of Tech-
nology and Ecological Effects of Biofouling Control Procedures at
Thermal Power Plant Cooling Water Systems. Johns Hopkins University,
Baltimore, Maryland. June 16-17. In press.
-------
296
Tsai, C. 1968. Effects of chlorinated sewage effluents on fishes in
Upper Patuxent River, Maryland. Chesapeake Sci. 9: 83-93.
Tsai, C. 1970. Changes in fish populations and migration in relation
to increased sewage pollution in Little Patuxent River, Maryland.
Chesapeake Sci. 11: 34-41.
Tsai, C. 1975. Effects of sewage treatment plant effluents on fish:
a review of literature. Chesapeake Research Consortium, Inc.,
CRC Pub. No. 36. Center for Environmental and Estuarine Studies.
University of Maryland, College Park, Maryland. 229 p.
Tsai, C., and J. A. Tomkins. 1974. Survival time and lethal exposure
time for the blacknose dace exposed to free chlorine and chloramine
solutions. Tech. Report No. 30. Center for Environmental and
Estuarine Studies. University of Maryland, College Park, Maryland.
22 p.
Ward, R. W., et al. 1975. Comparative disinfection and residual
toxicity studies with several disinfection agents. Ecological
Research Series, U.S. Environmental Protection Agency. Washington,
D.C. In press.
Warren, C. E. 1974. Laboratory determination of chloramine concentra-
tions safe for aquatic life. Quarterly Progress Report, EPA Grant
No. R-802286. Dept. Fish, and Wildlife, Oregon State University.
Corvallis, Oregon. October 1-December 31.
Westfall, B. A. 1946. Stream pollution hazards of wood pulp mill
effluents. Fishery Leaflet 174. Fish and Wildlife Service, U.S.
Department of the Interior.
White, G. C. 1972. Handbook of chlorination. Van Nostrand Reinhold
Co. New York. 744 p.
Wolf, E. G., M. J. Schneider, K. 0. Schwarzmiller, and T. 0. Thatcher.
1975. Toxicity tests on the combined effects of chlorine and
temperature on rainbow (Salmo gairdnevi,) and brook (Salvelinus
font-inalUs) trout. Proc. 2nd Thermal Ecology Symp., Augusta,
Georgia. April 2-5. In press.
DISCUSSION
George C. White, Consulting Engineer. Were the chlorine residuals
in the 0.001 to 0.005 part per million range actually measured by
amperometric titration or were these numbers extrapolated? From a
practical standpoint the equipment operator will not be able to get
reproducibility of anything less than 0.05, and it would be more believ-
able to me if it were 0.1. So, somewhere, it ought to be said whether
the numbers are extrapolated. If they are extrapolated, anything below
-------
297
0.05 should be considered zero because that is the way we are operating
our dechlorination systems.
Brooks. I think that most of the values cited for the chronic
exposures were done in Duluth, and I think the sophisticated ampero-
metric titrator can get down to between 1 and 10 parts per billion.
William A. Brungs, U.S. Environmental Protection Agency. Plus
or minus 100%.
Brooks. I am not sure about the plus or minus 100%, and I am not
sure exactly what is being measured.
Brungs. In many cases the numbers are also based on the known
dilution. You can pick up additional confidence that way. For example,
if 0.01 were diluted by half and they get a measurement around 0.005.
This would be confirmation.
Brooks. I think most of these studies with the chronic exposure
were done with a serial dilution apparatus,
White. We need to adhere to the recognition of significant figures
and the ability of control equipment and operating personnel to measure
these residuals. From a practical standpoint this is about 0.1 mg/liter
total chlorine. Therefore, somewhere in your reporting you have got to
say something about significant figures that will help the person
operating the equipment.
Brooks. In our plea for additional information in these reports,
I think that is a very important thing that should be included.
Brungs. We don't want to spend all our time on chemistry only.
Let us move to another subject.
Alan G. Heath, Virginia Polytechnic Institute and State University.
I wonder if you or anyone in the audience has any information on the
formation or possible formation of organochlorine compounds in the blood
of fish?
Brooks. No, we don't. But I was talking with either Bob Jolley
or Carl Gehrs about trying to get some chlorine-36 to see if we can find
any taken up by the fish.
Michael L. Bass, Mary Washington College. Do you or anyone at this
meeting have any information on the relationship of water hardness and
chlorine toxicity?
Brooks. Some of Warren's work cites alkalinity values over quite
a rangedI believe they determined there was no effect. I don't recall
the exact numbers they cited, but it was something like from 100 to 400.
-------
298
Brungs_. Dr. Ingols is in the audience and keeps asking me about
the calcium concentration. Would you like to make the point?
Robert S. Ingols, Georgia Institute of Technology. A report has
been published on the TLm against minnows of halogenated phenols. The
toxicity is dependent upon the number of halogens on the phenol. The
toxicity increases with number. Chlorine compounds are more toxic than
brominated ones which are more toxic than iodinated ones. An examina-
tion of the swimming pool literature may give some information on the
chlorination of algae. Bluegreen algae may live in a pool for months
in spite of shock doses and continuous low levels of chlorine. My
personal observation indicates that the presence of calcium (hardness)
protects the algae.
Brooks. There is some work by Japanese investigators that show
that some marine species can tolerate very high chlorine levels. We
looked through the biological literature and even, as a biologist, reached
out into the engineering literature. But I didn't get into the swimming
pool literature.
-------
A REVIEW OF THE IMPACT OF CHLORINATION PROCESSES
UPON MARINE ECOSYSTEMS
William P. Davis and Douglas P. Middaugh
Gulf Breeze Environmental Research Laboratory
Bears Bluff Field Station
Wadmalaw Island, South Carolina 29487
ABSTRACT
For over 175 years chlorine gas has been used in industrial, bio-
cidal, and disinfection applications. The chemistry of chlorine in
freshwater is relatively well known, but long-range effects on the
organisms and the ecological communities of marine waters have barely
been studied. Until recently, the so-called "chlorine demand" of
treated or receiving waters has been considered a desirable feature
which assured degradation of actively oxidizing states of chlorine to a
nontoxic state. With continuing and increased use of chlorine as an
antifouling biocide in powerplants, and as a disinfectant of municipal
wastes, concern has arisen that resulting byproducts, such as induced
halogenated hydrocarbons, could potentially reach environmentally harm-
ful levels. For example, in the State of Maryland the quantity of
chlorine used, which subsequently reaches the Chesapeake Bay, would have
sterilized that body of water were not chemical/biological degradation
processes in effect. But, what are the limits of natural degradation
systems? What, for one example, are the known environmental costs of
our present rates of chlorine applications on renewable fishery re-
sources? What kind and at what rate are persistent halogenated com-
pounds being produced? Where do these go in natural systems? From
partial or complete answers to these questions will come meaningful
environmental management criteria.
This paper presents a theoretical degradation model of chlorine
added to marine waters. Additionally it summarizes literature reporting
laboratory or ecological effects of chlorine.
299
-------
300
INTRODUCTION
Chlorine gas has been used in industry as a bleaching agent since
1800 and has become one of the most versatile chemicals known. In
freshwater it is well known as a disinfectant in drinking and recre-
ational water, a biocide for slime and fouling control, for treatment of
municipal wastes for pathogen control, and additional great amounts are
used as bleach in the pulp and paper industry. From all these applica-
tions where vast total quantities of chlorine are used, byproducts find
their way through society's effluents to natural ecosystems. The tox-
icity desired in disinfection biocide applications can continue on with
nondesirable effects to wildlife and ecosystems. Recent findings of
halogenated organics traceable in drinking water in 80 cities, under-
scores the need for responsible assessment of the management and effects
of our chlorination processes, and the environmental costs incurred.
The State of Maryland is often used by planners as a mini-model for
the United States. In the case of rate of chlorine use, some of the
most accurate statistics exist for Maryland. Furthermore, relative to
this compilation of impacts upon marine ecosystems, resulting chlorina—
tion constituents from Maryland mostly drain into the Chesapeake Bay.
An inventory of chlorine discharge from the Maryland contribution alone
into Chesapeake Bay, assuming no degradation, reveals discharges of
27 million pounds per year of chlorine from municipal treatment plants
and 2.2 million pounds per year from power generation facilities. It
would appear to the casual observer that without the action of degrada-
tion processes, these amounts would soon sterilize the Bay. Further-
more, it has been calculated that 3% or more of these amounts might
produce halogenated organic compounds which potentially persist in
marine ecosystems (Jolley, in this symposium). Twenty-four other states
border marine waters where various forms of chlorination discharges
enter marine ecosystems under various physical-chemical and regional
conditions.
The initial purpose of this paper is to compile the scarce data
currently available on chlorine effects upon biota of estuarine and
marine ecosystems. The chemistry of chlorine is adequately reviewed in
-------
301
other parts of this symposium. Secondly, implications of chemical
detection and marine ecosystem effects reported in the literature are
discussed to emphasize needs for future investigations.
DEGRADATION PATHWAYS OF CHLORINE IN
MARINE SYSTEMS
Figure 1 shows a theoretical degradation model prepared by Block
and Helz (1975) to summarize the suspected pathways of chemical reac-
tions resulting from chlorination of natural marine waters. Obviously
missing are any coefficients, rates, or the many effects and influences
from physical-chemical conditions and factors. Although some data exist
on effects of residual chlorine and a limited number of byproducts upon
specific organisms, there is virtually no information on transport
processes, persistence, bioaccumulation, and fate of the halogenated
compounds resulting from chlorination processes.
The reaction occurring between levels I and II is a result of
chlorine decay from a diatomic gas to hypochlorous acid, hypochlorite
ions, and sodium hypochlorite. As pointed out by Moore (1951) and Lewis
(1966), this reaction occurs rapidly and goes to completion within
seconds after the addition of chlorine. Inclusion of sodium hypochlo-
rite within level II is based on work by Sugam and Helz (1975).
The chemical composition and abundance of products formed from
level II to level III is a function of physical and chemical parameters
of the water, including, but not limited to, temperature, pH, ammonia,
and bromine available as reaction components. In seawater it is possi-
ble that the predominant species would be bromamines, especially if NH^
ions are less abundant than Br ions.
Level IV includes halogenated organic constituents which may be
formed by level II or level III species, including chloramines, hypo-
bromite, and bromamines. The stable end products in level V occur
through a diverse group of mechanisms operative in steps I through IV.
Charge balance results in one atom of Cl passing from level I to
level V to each atom passing from level I to level II. Reduction of
hypochlorite by Br~ or Fe2+ and Mn2+ may release Cl from level II to
level V. Movement of Cl~ from level III to level V can also occur
-------
302
II
III
IV
CL-
HOCL,OCi_~, NAOCt
NH2Cu NHCi_2, NH2
NHBR2, BRO-, HBRO
HALOGENATED ORGANIC
CONSTITUENTS
, BR'
Fig. 1. Degradation processes for chlorine in saline waters,
(Modified from Block and Helz, in preparation).
-------
303
in a number of ways, the most obvious, suggested by Laubusch (1971),
involves destruction of chloramines when the OCl~/NHit+ ratio is large.
Some of the chlorinated organics identified by Jolley (1973) are
persistent and the decay from level IV to level V is probably a slower
process than decay from levels I through III to level V.
TOXICITY OF CHLORINE IN ESTUARINE ENVIRONMENTS
The relative toxicity of chlorine in water is considered to be
related to the amount and proportions of free and residual chlorine.
Some authors report that free chlorine is generally more toxic to fresh-
water organisms than chloramines (Doudoroff and Katz 1950, Merkens 1958)
even though the toxicity of various forms of chlorine were of the same
order of magnitude. Rosenberger (1971) and Basch and Truchan (1973)
found that dichloramine was more toxic than monochloramine in fresh-
water. A comprehensive review paper by Brungs (1973) summarizes the
toxic effects of residual chlorine on freshwater aquatic organisms.
Additionally, Whitehouse (1975) in his review for Central Electricity
Research Laboratories of Great Britain covered references from both
fresh and marine waters.
In seawater, Holland et al. (1960) determined that dichloramine is
apparently more toxic than monochloramine and reported chloramines are
more toxic than free chlorine. These findings may reflect the complex
chlorine-bromine reaction kinetics suggested by Johanesson (1958, 1960),
Lewis (1966), and Carpenter (this symposium).
CHLORINE TOXICITY TO MARINE PHYTOPLANKTON
The effects of chlorination and thermal pollution on phytoplankton
productivity has been investigated in some detail, Table 1. Carpenter
et al. (1972) observed an 83% decrease in the productivity of phyto-
plankton cultured in waters which had passed through the cooling system
of a nuclear generating plant on Long Island Sound.
Intake water was chlorinated at a rate of 1.2 mg/1 with a residual
of 0.4 mg/1 measured at the discharge. Addition of 0.1 mg/1 chlorine at
-------
Table 1. Summary of toxic effects of chlorinated wastes and water on marine phytoplankon
Species
Phytoplankton
Chicory domonas sp.
Skeletonema
oostatim
Phytoplankton
Phytoplankton
Toxicant
C12 injection
sodium
hypochlorite
solution
sodium
hypochlorite
solution
C12 injection
Reported
residual
chlorine
(mg/1)
0.05-0.40
0.69-12.9
0.18-2.4
0.32
0.01
0.075-0.25
— —
Duration
of
test
12 hr +
4 hr incu-
bation
5 min
5 min
2 min
45 min
24 min
15 min
Effect
50-98% loss of
productivity
reduced
growth rate
none up to
0.29 rag/1;
greater amts .
inhibited
growth
55% decrease
in ATP
77% decrease
in ATP
55% decrease
in growth
91% reduction
in photosynthe-
sis
Reference
Carpenter
et al.
(1972)
Hirayama
and
Hirano
(1970)
Gentile
et al.
(1972,
1973)
Hamilton
et al.
(1970)
u>
o
-------
305
the intake with nondetectable residuals at the outfall decreased produc-
tivity by 79 percent. Essentially no decreases in productivity were
observed when phytoplankton passed through the cooling system under the
same thermal conditions, but without addition of chlorine. Hirayama and
Hirano (1970) measured the effect of chlorination on the photosynthetic
activity of Skeletonema oostatwn and found that cells were killed when
subjected to 1.5 to 2.3 mg/1 chlorine for 5 and 10 minutes.
Gentile et al. (1972, 1973) observed a 55 percent decrease in ATP
content of marine phytoplankton exposed for two minutes to 0.32 mg/1
residual chlorine and a 77 percent decrease after 45 minutes of exposure
to chlorine concentrations as low as 0.01 mg/1. A 50 percent reduction
in growth rates was observed for 10 species of marine phytoplankton
exposed for 24 hours to chlorine concentrations that ranged from 0.075
to 0.25 mg/1.
Morgan and Stress (1969) used photosynthetic rates to evaluate
response of estuarine phytoplankton passed through the cooling system of
a steam electric power station on the Patuxent River, Maryland. The
photosynthesis rate increased with an 8°C rise in temperature when
ambient water temperatures were 16°C or less, but inhibition occurred
when ambient temperatures were above 20°C. In a related study conducted
at the same site, Hamilton et al. (1970) measured a 91 percent decrease
in primary productivity during intermittent chlorination.
CHLORINE TOXICITY TO INVERTEBRATES
Muchmore and Epel (1973) investigated the effects of chlorination
of wastewater on fertilization in marine invertebrates, Table 2. Un-
chlorinated sewage from the Pacific Grove, California, Municipal Treat-
ment Works was a weak inhibitor of fertilization in the sea urchin,
Strongylocentrotus purpuratus. Exposure of gametes to a 10 percent
unchlorinated sewage-seawater mixture typically reduced fertilization
success by 20 percent. A 0.5 percent dilution of moderately chlorinated
sewage (11 mg/1 "total residual chlorine"* undiluted), significantly
*Total residual chlorine (TRC) in this case is being reported for marine
waters where residual chlorine per se is not likely to be found.
-------
306
reduced fertilization. Chlorination had more effect on sperm cells than
on eggs. Eggs incubated for 5 minutes in 0.77 mg/1 hypochlorite solu-
tion and subsequently washed to remove the hypochlorite showed no
reduction in fertility. Whereas incubation of sperm at a 0.07 mg/1
hypochlorite caused loss of fertilization ability. This loss was
attributed to a loss of sperm motility which was not restored by washing
to remove the hypochlorite. Gametes of the echiuroid, Vvedhis oaupo,
and sperm of the annelid worm, Phragmatopoma californiea, were not as
sensitive to chlorination.
A number of power plant-related studies have been conducted to
determine the effect of chlorination of seawater on fouling organisms.
Waugh (1964) observed no significant difference in the mortality of
unexposed oyster larvae, Ostrea edulis, or those exposed to 5 mg/1
chlorine for 3 minutes at ambient temperature.
Exposure of larvae to thermal stress (10°C above ambient) and
10 mg/1 chlorine for 6 to 48 minutes also had no significant effect on
survival 64 hours after treatment. Nauplii of the barnacle, Elmi-nius
modestus, was more acutely sensitive to chlorine. Residual chlorine
concentrations in excess of 0.5 mg/1 caused heavy mortality and reduced
growth of survivors.
McLean (1973) simulated the conditions encountered by marine organ-
isms passing through a power plant on the Patuxent River, Maryland.
Intake chlorination to 2.5 mg/1 residual, entrainment for approximately
3 minutes, and sustained exposure to elevated temperatures for up to
3 hours were used as experimental parameters. While barnacle larvae,
Balanus sp., and copepods, Acartia tonsi, were not affected by a 3-hour
temperature stress of 5.5 and 11°C above ambient; exposure to 2.5 mg/1
residual chlorine for 5 minutes at ambient temperatures caused respec-
tive mortality rates of 80 and 90 percent. The amphipod, Mel-ita
nitida, and the grass shrimp, Pataemonetes pugio, showed a delayed death
response after exposure to 2.5 mg/1 TRC for 5 minutes. Nearly 100 percent
mortality was observed for both species 96 hours after exposure to the
chlorine residual. McLean (1972) showed that established colonies of
the euryhaline colonial hydroid, Bimeria franciscana, were not greatly
affected by 1 and 3 hours of exposure to 4.5 mg/1 TRC.
-------
Table 2. Summary of toxic effects of chlorinated wastes and water on marine invertebrates
Species
Strongylaaentpotus
puzpupatus
(gametes)
Ufsahie aaupo
(gametes)
Phragma topoma
calfforniaa
(sperm)
Ostrea edulis
Elminius modes tus
Balanus sp.
Acart-ia tonei
Melita nitida
Palaemcmetes pugio
Bi-rrwia fTO.ncieaa.-na
Anemones
Mussels
Barnacles
Mytilus edulis
Reported
T°^- chlorine
(mg/1)
chlorinated 0.02
sewage 0.11
effluents
0.2
1.0
0.2
1.0
residual 10.0
chlorine
2.0
5.0
C12 injection 2.5
2.5
2.5
2.5
4.5
residual 10.0
chlorine
2.5
1.0
10.0
2.5
1.0
10.0
2.5
1.0
C12 injection 0.02-
0.05
Duration
of
test
5 din
5 din
5 min
5 min
5 min
5 min
48 rain + 10°C
10 min
3 Bin
5 win
5 min
5 min
5 min
4 days
1,2,4,8 hr/day
for 10 days
8 days
15 days
1,2,4,8 hr/day
for 10 days
5 days
15 days
1,2,4,8 hr/day
for 10 days
4 days
7 days
a few hours
Effect
none
1002 inhibition of
fertilization
22% inhibition of
fertilization
100% inhibition of
fertilization
22% loss of motility
86% loss of motility
none
death and inhibited
growth
none
80% mortality
90% mortality
near 100% mortality
96 hr after exposure
near 100% mortality
96 hr after exposure
none
none
100% mortality
100% mortality
none
1007. mortality
100% mortality
95-100% mortality
100% mortality
100% mortality
detachment and
migration
Reference
Muchmore
and Epel
(1973)
Waugh
(1964)
McLean
(1972,
i m "i^
1973)
Turner
et al.
(1948)
James
(1967)
u>
o
-------
308
Turner et al. (1948) determined that continuous treatment of sea-
water conduits with 0.25 mg/1 chlorine prevented fouling during a 90-day
interval when the flow velocity was 52 cm/second or less. Intermittent
treatment with 10 mg/1 residual chlorine for 8 hours a day was ineffec-
tive in preventing fouling by anemones, mussels, and barnacles.
James (1967), working in Great Britain, observed that residual
chlorine concentrations of 0.02 and 0.05 mg/1 caused detachment and
movement of mussels in the direction of water flow through an aquarium
with eventual elimination of the mussels. He concluded that the most
effective way to prevent fouling by mussels was not to kill but to
discourage settling in cooling water systems by continuous low level
chlorination.
Markowski (1959, 1960) compared the occurrence of marine organisms
on concrete slabs placed in the intake and outfall canals of an electric
generating plant. Chlorine was injected into cooling condensers for two
hours a day at a concentration between 1 and 2.5 mg/1. No vegetation
was observed to grow in the intake canal where dense animal populations
occurred (predominantly invertebrates, Coelenterata, and Polyzoa). The
outfall canal contained a prolific growth of algae, Enteromorpha sp.,
but fewer invertebrates. The barnacle, Balanus improvises, collected
with some regularity from the intake canal, was never observed in the
outfall canal. The mollusk, Eubpanahus sp., was more abundant on the
intake slabs than in the outfall.
Gibson et al. (in press) report critical thermal maxima (CTM) for
coon stripe shrimp (Pandalus danae) and LCso values for chlorine and
copper, both singly and in combination. Pandalus danae were more re-
sistant to chlorine when acclimated and exposed at 8 to 10°C than when
acclimated at 8°C and exposed at 15 or 20°C, or when acclimated and
exposed at 15°C which is near their optimum short-term growth temper-
ature (16°C). Chlorine at a reported 0.18 mg/1 concentration is lethal
to 1 to 2 gm coon stripe shrimp at 16°C and reduced growth was observed
in shrimp exposed to 0.08 mg/1 for one month.
-------
309
CHLORINE TOXICITY TO ESTUARINE FISH
Tsai (1968, 1970, 1975) observed decreases in abundance and occur-
rence of brackish water fish species in certain areas of the Upper and
Little Patuxent Rivers that received chlorinated sewage effluent. Tsai
suggests that chlorinated sewage effluent may also block the upstream
migration of such semianadromous species as white catfish and white
perch. He attributed the "blocking effect" to chlorination products
rather than to reduced dissolved oxygen or pH resulting from organic
decomposition of the effluent (Table 3).
Tsai (1973) measured the diversity index of fishes upstream and
downstream from 98 sewage treatment plants in Virginia, Maryland, and
Pennsylvania. Sewage treatment plants were categorized as Type I
engineering facilities (sludge activation, aeration, sedimentation, and
filtration) with effluent chlorination; Type II engineering facilities
with chlorination and effluent holding lagoon; and Type III engineering
facilities with lagoon and effluent chlorination at the lagoon outlet.
Reductions in number of individual fish, number of species, and the
species diversity index were significant downstream from Type I and III
plants. These reductions were attributed to total residual chlorine
levels and turbidity. No significant changes in diversity indices were
found downstream associated with Type II plants.
A study of the effect of chlorinated sewage effluents on sockeye
salmon, Onchorhynchus nerka, and pink salmon, 0. gorbuscha, has been
conducted by Servizi and Martens (1974). They used three study sites to
conduct cage bioassays. The first, Site I, was adjacent to a primary
treatment plant with effluents chlorinated following settling and dis-
charged through a 600-ft pipe line directly into the receiving stream.
Site II was on a stream receiving wastes from an activated sludge plant
in which chlorinated effluents were discharged into a large effluent
holding lagoon and retained for 30 to 60 days. Site III was located on
a stream receiving effluents which were chlorinated as they left a
nonaerated lagoon.
-------
Table 3. Summary of toxic effects of chlorinated wastes and water on marine and freshwater fishes
Species
Freshwater and
Brackish Fishes
Leiostomu.8 xanthuma
Morone sp .
Pomatoima aoltotvix
Cynoecion regalia
Brevoortia tyrannue
L. xanthurue
Oncorhynahus nerka
0. gorbusoha
(freshwater)
0. gorbueaha
0. gorbueaha
Morone ameviaana
Menidia menidia
Fundulus heteroalitus
Trineates maaulatua
Pleuponeotee plateeaa
(eggs)
(Larvae)
Solea aolea
(Larvae)
Cyprinus aarpio
Toxicant
chlorinated sewage
effluents
chlorinated sewage
effluents
sodium hypochlorite
chlorinated sewage
effluents
residual chlorine
residual chlorine
residual chlorine
residual chlorine
4-chlororesorcinol
5-chlorouracil
(0.001 mg/1)
Reported
residual
chlorine
(fflg/D
0.6-2.0
0.07-0.28
0.09
0.14
0.28
0.02-0.26
0.16
0.5
0.5
0.08
0.08
0.03
0.03
0.04-0.08
0.62
0.10
0.62
0.034
0.03-0.06
Duration
of
test
long term
May- June 1973
96 hr
24 hr
6 hr
24 hr
72 hr
80 min + 10°C
10 min + 10°C
10 min
10 min
10 min
10 min
8 days
72 hr
96 hr
48 hr
96 hr
48 hr
3-7 days
Effect
decreased population
size and diversity
probable kill
5-10 million fish
50% mortality
50% mortality
50% mortality
100% mortality
100% mortality
50% mortality
50% mortality
avoidance
avoidance
avoidance
avoidance
none
50% mortality
50% mortality
50% mortality
50% mortality
50% mortality
reduced hatch
Reference
Tsai (1968, 1970,
1973)
Virginia State
Water Control
Board (1974)
Servizi and
Martens (1974)
Stober and Hanson
(1974)
Me Id rim et al.
(1974)
Alderson (1972,
1974)
Gehrs et al.
(1974)
-------
311
Measured chlorine residuals in the receiving stream at Site I
ranged from 0.02 to 0.26 mg/1. These concentrations resulted in 100
percent mortality of caged sockeye fingerlings placed 30, 60, and 250
feet below the effluent discharge point. Additional tests indicated
that the primary effluent without chlorination was also toxic. However,
fish exposed to the unchlorinated effluent lived ten times longer than
ones exposed when effluents were being chlorinated. Toxicity of the
unchlorinated effluents was attributed to MBAS and ammonia.
Tests at Site II indicated that chlorinated effluents retained for
30 to 60 days were not toxic to sockeye fingerlings and alevins and pink
salmon alevins after 26 days of exposure.
In tests at Site III, with fingerling sockeye salmon, chlorinated
sewage effluents (measured TRC, 0.85 mg/1) resulted in 50 percent mor-
tality after 48 minutes. Fifty percent mortality occurred after 13
hours of exposure to the unchlorinated effluents. Sublethal exposures
of fingerling sockeye salmon to the effluents from Site III (1 to
3 hours of exposure to 0.22 mg/1 TRC) resulted in gill damage including
hyperplasia, swollen epithelial cells, and separation of epithelium from
pillar cells.
The toxicity of chlorine and heat to pink salmon, Oncophynchus
gorbuscha, and chinook salmon, 0. tshawytseha, has been determined by
Stober and Hanson (1974). Juveniles of each species were tested in
seawater at five residual chlorine concentrations ranging from 0.05 to
1.0 mg/1 and four temperatures from At 0 to 10°C where the LTso (lethal
time for 50 percent mortality) ranged from approximately 10 minutes at
0.5 mg/1 TRC for chinooks to 80 minutes for pinks.
Meldrim et al. (1974) in flowing water bioassays studied the effect
of chemical pollutants on estuarine organisms. They found that white
perch, Morone americana, consistently avoided TRC levels as low as
0.08 mg/1 at temperatures from 7 to 17°C. Silversides, Menidia menidia,
also avoided 0.08 mg/1 TRC at temperatures from 8 to 28°C but showed a
preference for 0.08 mg/1 TRC when fish acclimated to 7°C were exposed at
12°C. Mummichogs, Fundulus heteroelitus, and hog chokers, Trineetes
maculatus, avoided TRC levels as low as 0.03 mg/1.
-------
312
Alderson (1972, 1974) found that the 48- and 96-hr Tl of free
m
chlorine for plaice larvae, Plenroneetes platessa, was 0.062 and 0.34
mg/1, respectively. Eggs were not affected when exposed to 0.075 and
0.04 mg/1 residual chlorine for 8 days indicating that the egg membrane
gives considerable protection over long periods. The 72- and 192-hr
Tl for eggs was 0.7 and 0.12 mg/1 TRC, respectively. Larval Dover
sole, Solea solea, were most sensitive to chlorine immediately after
hatching. The 48-hr LC$Q was 0.03 mg/1 for stage Ic to Id larvae and
0.06 mg/1 residual chlorine for stage 4b larvae.
Gehrs et al. (1974) tested the sensitivity of carp eggs, Cyprinus
carpi-a, to two of the compounds identified by Jolley, 4-chlororesorcinol
and 5-chlorouracil. Significant reductions in the hatchability of non-
water-hardened carp eggs were observed in concentrations of each com-
pound as low as 0.001 mg/1.
Massive fish kills occurred on the James River, Virginia, during May
and June, 1973 (Virginia State Water Control Board 1974). Species
affected by the kill included spot, Lei-ostomus xanthnrus; white perch,
Movone americana; bluefish, Pomatomus saltatrix; grey seatrout,
Cynoscion vegali-s; and menhaden, Brevoortia tyrannus. Most of the fish
kill in the James River occurred adjacent to sewage treatment plants.
Total residual chlorine (TRC) levels as high as 0.7 mg/1 were observed
in the James River while effluents from both plants showed more than
3.0 mg/1.
Distress symptoms of dying fish included spiral swimming patterns,
broken vertebral columns, listless floating, inverted swimming, disten-
sion of the air bladder, loose body scales, mucous on the skin, and
hemorrhaging along fins and body surface.
Live box tests conducted adjacent to the James River Municipal
Treatment Works (MTW) demonstrated a correlation between rates of
effluent chlorination and mortality of juvenile spot and croaker. At
an average daily chlorine feed of 1200 pounds (total flow of water was
approximately 10 mgd during tests) and a measured residual chlorine
concentration of 3.0 mg/1, 100 percent of caged fish died within 20
hours. When chlorine feed rate was reduced to approximately 400 pounds
per day, only 20 percent mortality was observed after 20 hours.
-------
313
On-site aquaria tests complimented results of the cage tests.
Water from an area adjacent to the outfall of the James River MTW was
pumped through aquaria containing juvenile spot. Mortalities ranged
from 91 to 100 percent after 40 to 85 minutes of exposure prior to
reduced chlorination. After chlorination rates were reduced, mortal-
ities were 0 to 26 percent after 120 minutes of exposure.
Continuous flow laboratory bioassays were also conducted. The
estimated 96-hr LC50 for juvenile spot was 0.09 mg/1 TRC. The estimated
24-hr LC50 was 0.14 mg/1 and the 6-hr LC50 was 0.28 mg/1 TRC.
Separate field studies on the spot, Leiostomus xantkurus, revealed
that up to 40 percent of juveniles from the 1973 year class exhibited
vertebral column deformities. The abnormal "bent back" forms have
become identifiable as "tags" to distinguish the 1973 year class in 1975
population samples from the James River (Labbish Chao and John V. Merriner,
Virginia Institute of Marine Science, personal communication).
In California, Young (1964) observed tumor-like sores around the
mouths of white croaker, Genyonemus lineatus, collected near the Hype-
rion sewage outfall in Santa Monica Bay. Although no direct evidence
was discovered to link occurrence of lesions with chlorinated sewage
effluents, there was a general decline in fitness of croakers and other
species found in close proximity to the outfall area.
Relevant to the decline of fitness of marine fishes exposed to
chlorinated effluents is the report by Grothe and Eaton (1975) where
freshwater fishes were killed by anoxia due to chlorine oxidation of
hemoglobin to methemoglobin. Fish exposed to freshwater containing
chloramines exhibited striking elevations in blood concentrations of
methemoglobin, up to 29 to 32%. This results in decreased tissue oxygen
delivery of 30% in the test conditions and provides a logical basis for
the malaise observed near MTW outfalls.
IMPLICATIONS OF ECOSYSTEM IMPACT OF CHLORINATION IN
MARINE WATERS
The foregoing information summarizes uncritically the available
literature on effects of chlorination upon estuarine and marine organisms.
-------
314
It is appropriate to focus on the few available implications of pub-
lished and unpublished research on effects of chlorine upon marine
organisms.
As we have seen during this symposium, chlorine reacts rapidly, and
in yet to be defined ways, in marine ecosystems. Differences in natural
systems, including the presence of organic and inorganic compounds,
interreactions involving photolysis, specific reactions favored by
specific temperatures, and suspended particles, all combine to give us a
highly dynamic set of events. Couple these physical-chemical parameters
with the uncertainty of organism responses and bioassays and we have a
truly difficult task to produce relevant or meaningful standards.
Furthermore, much of the current available literature has accepted
the amperometric titrator as actual representation of true residual
chlorine concentrations. We know now from the discussions of marine
chemistry during this symposium that these readings actually represent
abstractions of chemical processes yet to be described in the case of
marine waters. Therefore, without clear definitions of the physical-
chemical parameters much of the reported data reviewed in this paper
cannot be meaningfully compared. Finally, "residual chlorine" readings
in fresh water, or marine waters of different salinities or components,
can represent entirely different chemical species. No calibration or
critical comparability is now appropriate. If marine biologists change
nothing else, they must ascertain that any reported effects data be
accompanied with synoptic measurements of temperature, salinity, pH,
light intensity, suspended particles, and, wherever possible, dissolved
and suspended organics. These reference points will insure both compa-
rability and relevance for future examination at a time when chemistry
of chlorine in marine waters is defined and better understood.
CURRENT RESEARCH ON EFFECTS OF CHLORINATION
UPON MARINE ORGANISMS AT EPA
Mixed species experiments at Bears Bluff Field Station, South
Carolina, consist of an array of ninety-six, 38-liter flowing water
units. Each receives 40 liters/hr of unfiltered marine water from the
tidal North Edisto River that contains entrained natural planktonic
-------
315
organisms and suspended particles. Thirty-two of these units are "har-
vested" each 30 days, another set (32) each 60 days, and a third set
(32) each 120 days. Within each harvest set three levels of continual
chlorination are applied using sodium hypochlorite adjusted to calcu-
lated levels of 0.0 (control), 0.125, 0.250, and 0.500 ppm at the de-
livery head. The control and each of the chlorination levels is
triplicated for three sets of eight test tanks. Chlorination delivery
rates are maintained by the use of syringe pumps. Amperometric titra-
tion can detect only traces of total residual chlorine at the highest
level cited.
Harvests are followed by identification, enumeration, and recording
size of species. The harvest effort requires 5 to 8 people to sort and
identify. The principal question is: How will continually introduced
low levels of sodium hypochlorite affect the settling and development
of a benthic estuarine community?
This experimental design fulfills the need for a bioassessment
technique which correlates data from laboratory single-species bioassays
with data from field experiments. Furthermore, there is need for data
representing summation of all the potential ecological effects of chlo-
rine introduced into marine waters, regardless of the state of the art
of the chemical detection.
An important feature of the design is that additional individual
trays of oysters (well known bioaccumulators) are placed in the water
collector systems downstream from the community test tanks. These
oysters will be analyzed for body burdens of persistent halogenated
organics. There are no results to report at this time from these
endeavors.
Simultaneously, single-species bioassays are being conducted on
such fishes as spot (Le-iostomus xanthwcus), mummichog (Fundulus
heteroclitus), silversides (Menidia menidia), and striped bass (Morone
saxi.ti.1is).
Some of the results at this time from fish bioassays are discussed
in the following paragraphs.
-------
316
Short-term studies with juvenile spot (10 to 15 mm TL) during
February and March, 1975, showed that they were sensitive to chlorina-
tion of flowing seawater. "Classic freshwater total residual chlorine"
in these bioassays was determined by amperometric titrator. The esti-
mated incipient LC50 at 10°C was 0.12 mg/1 and at 15°C was 0.06 mg/1.
Spot also showed temperature dependent avoidance responses to chlorina-
tion in experiments run simultaneously to the LCso exposures. Concen-
trations avoided were in general similar to the estimated incipient
LCso's at respective test temperatures of 10 and 15°C (Middaugh et al.,
in press).
Histological monitoring of the juvenile spot demonstrate damage to
gill tissue in specimens exposed to residual chlorine (Fig. 2). Figure
2a is an example of normal (control) pseudobranch structure illustrating
plump epithelial cells surrounding lamellar capillaries; Fig. 2b shows
separated epithelial membranes in specimens exposed to amperometric
reading levels of 1.5 mg/1 for 95 minutes. As can be seen, separated
epithelial membranes exhibit a scalloped border which is not an example
of normal artifact.
In work with early life history stages of the striped bass, we
found that the eggs were less sensitive than newly hatched prolarvae and
that the sensitivity of prolarvae and larvae to chlorination level
decreased with increasing age.
Additionally, through an extramural grant to Dr. Morris Roberts at
the Virginia Institute of Marine Science we are supporting bioassays on
selected invertebrate species including the oyster crab (Penopeus
herbstii), hermit crab (Pagwms sp.), and blue crab (Callinectes
sapidus), and coot clam (Mulinia lateral-is).
In a "mid-stream" research summary such as this, conclusions are
neither necessarily relevant or appropriate. However, one point we wish
to make is the tantamount need for continual information exchange among
chlorine effects researchers.
-------
317
2b
Fig. 2. Sections of pseudobranch from juvenile spot. 2a is a
section of control fish not exposed to residual chlorine in flowing
seawater. 2b illustrates tissue damage when exposed to levels of sodium
hypochlorite. The scalloped appearance indicates uplifted epithelial
tissue from grossly inflated lymphatic sinuses.
-------
318
ACKNOWLEDGMENTS
Our sincere thanks for much of the labor of obtaining literature
goes to Ms. Ruth Yoakum; and for preparation of summary tables to
Ms. Genie Floyd. Dr. George S. Helz kindly consented to the use of his
model of chlorine degradation in marine waters and provided data on
chlorine usage patterns for Chesapeake Bay. Dr. Ronald Block and
Dr. James Carpenter offered valuable comments on drafts of the paper.
Dr. John Couch examined and interpreted histopathological slides of
juvenile spot. Dr. Christopher Koenig made constructive criticisms
during drafting of the manuscript. Dr. Nelson R. Cooley edited in
detail the final draft of the manuscript.
REFERENCES
Alderson, R. 1972. Effects of low concentrations of free chlorine on
eggs and larvae of plaice, Pleuponeetes platessa L., p. 312-315.
In Marine pollution and sea life. Fishing News, Ltd. London.
Alderson, R. 1974. Seawater chlorination and the survival and growth of
the early developmental stages of plaice, Pleuroneetes platessa L.,
and Dover sole, Solea solea (L). Aquaculture 4: 41-53.
Basch, R. E., and J. G. Truchan. 1973. Calculated residual chlorine
concentrations safe for fish. Interim report. Michigan Water Re-
sources Commission, Bureau of Water Management. Lansing, Michigan.
Block, R. M., and G. R. Helz. 1975. Biological and chemical implica-
tions of chlorine in estuarine and marine systems. Chesapeake Sci.
In press.
Brungs, W. A. 1973. Effects of residual chlorine on aquatic life.
J. Water Pollut. Control Fed. 45(10): 2180-2192.
*Brungs, W. A. General considerations concerning the toxicity to aquatic
life of chlorinated condenser effluent. Workshop for assessment of
technology and ecological effects of biofouling control procedures
at thermal power plant cooling water systems. Johns Hopkins
University. Baltimore, Maryland. June 16-17, 1975. In press.
Carpenter, E. J., B. B. Peck, and S. J. Anderson. 1972. Cooling water
chlorination and productivity of entrained phytoplankton. Mar.
Biol. 16: 37-40.
Doudoroff, P., and M. Katz. 1950. Critical review of literature on
the toxicity of industrial wastes and their components to fish.
Sewage Ind. Wastes 22(11): 1432-1458.
Literature not cited.
-------
319
*Fair, G. M., J. C. Morris, S. L. Chang, I. Weil, and R. P. Burden.
1948. Chlorine as a water disinfectant. J. Am. Water Works Assoc.
40: 1051-1061.
Gehrs, C. W., L. D. Eyman, R. L. Jolley, and J. E. Thompson. 1974.
Effects of stable chlorine-containing organics on aquatic environ-
ments. Nature 249: 675-676.
Gentile, J. H., J. Cardin, M. Johnson, and S. Sosnowski. 1972. The
effects of chlorine on the growth and survival of selected species
of estuarine phytoplankton and zooplankton. Unpublished manuscript.
Environmental Research Laboratory, U.S. Environmental Protection
Agency. Narragansett, Rhode Island.
Gentile, J. H., S. Cheer, and N. Lackie. 1973. The use of ATP in the
evaluation of entrainment. Unpublished data. Environmental
Research Laboratory, U.S. Environmental Protection Agency.
Narragansett, Rhode Island.
Gibson, Charles I., Thomas 0. Thatcher, Charles W. Apts. Some effects
of temperature, chlorine and copper on the survival and growth of
the coon stripe shrimp, Pandalus danae. Proc. Thermal Ecol. Symp.
Augusta, Georgia. April 1975. In press.
Grothe, D. R., and J. W. Eaton. 1975. Chlorine-induced mortality in
fish. Trans. Am. Fish. 104(4): 800-802.
Hamilton, D. H., D. A. Flemer, C. V. Keefe, and J. A. Mihursky. 1970.
Power plants: effects of chlorination on estuarine primary pro-
ductivity. Science 169: 197-198.
Hirayama, K., and R. Hirano. 1970. Influence of high temperature and
residual chlorine on marine phytoplankton. Mar. Biol. 7: 205-213.
Holland, G. A., J. E. Lasater, E. D. Neumann, and W. E. Eldridge. 1964.
Toxic effects of organic and inorganic pollutants on young salmon
and trout. Dep. Fish. Res. Bull. No. 5. State of Washington.
264 p.
*Houghton, G. U. 1946. The bromide content of underground waters.
Part II: Observations on the chlorination of water containing free
ammonia and naturally occurring bromide. J. Soc. Chem. Ind.
65: 324-328.
*Ingols, R. S., H. A. Wyckoff, T. W. Kethley, H. W. Hodgden, E. L. Fincher,
J. C. Hildebrand, and J. E. Mandel. 1953. Bactericidal studies of
chlorine. Ind. Eng. Chem. 45: 995-1000.
James, W. G. 1967. Mussel fouling and use of exomotive chlorination.
Chem. Ind. (Lond.) 24: 994-996.
*
Literature not cited.
-------
320
Johanneson, J. K. 1958. The determination of monobromamine and
monochloramine in water. Analyst 83: 155-159.
Johanneson, J. K. 1960. Bromiriation of swimming pools. Am. J. Public
Health 50: 1731.
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants. Ph.D.
Dissertation. University of Tennessee. 339 p.
Jolley, R. L., C. W. Gehrs, and W. W. Pitt. Chlorination of cooling
water: a source of chlorine-containing organic compounds with
possible environmental significance. Proc. 4th National Symposium
Radioecology. Corvallis, Oregon. May 12-14. In press.
Laubusch, E. J. 1971. Chlorination and other disinfection processes.
In Water quality and treatment. Am. Water Works Assoc. 654 p.
Lewis, B. G. 1966. Chlorination and muscle control. I. The chemistry
of chlorinated seawater. A review of the literature. Lab. Note No.
RD/L/N/106/66. Central Electric Research Laboratory.
Markowski, S. 1959. The cooling water of power stations: a new factor
in the environment of marine and freshwater invertebrates. J. Animal
Ecol. 28: 243-258.
Markowski, S. 1960. Observations on the response of some benthonic
organisms to power station cooling water. J. Animal Ecol.
29: 349-357.
*McKee, J. E., and H. W. Wolf. 1963. Water quality criteria. 2nd Ed.
Publication 3A. California State Water Quality Control Board.
Sacramento. 548 p.
McLean, R. I. 1972. Chlorine tolerance of the colonial hydroid,
Bimeria franciscana. Chesapeake Sci. 13: 229-230.
McLean, R. I. 1973. Chlorine and temperature stress in estuarine
invertebrates. J. Water Pollut. Control Fed. 45: 837-841.
Meldrim, J. W., J. J. Gift, and B. R. Petrosky. 1974. The effect of
temperature and chemical pollutants on the behavior of several
estuarine organisms. Ichthyological Associates, Inc., Bulletin
11: 1-129.
Merkens, J. C. 1958. Studies on the toxicity of chlorine and chlora-
mines to the rainbow trout. Water Waste Treat. J. 7: 150-151.
Literature not cited.
-------
321
Middaugh, Douglas P., Allan M. Crane, and John N. Couch. Toxicity of
chlorine to juvenile spot, Leiostomus xanthurus. In press.
Moore, E. W. 1951. Fundamentals of chlorination of sewage and wastes.
Water Sewage Works 98: 130-136.
Morgan, R. P., and R. G. Stross. 1969. Destruction of phytoplankton
in the cooling water supply of a stream electric station. Chesapeake
Sci. 10: 165-171.
Muchmore, D., and D. Epel. 1973. The effects of chlorination of waste-
water on fertilization in some marine invertebrates. Mar. Biol.
19: 93-95.
Rosenberger, D. R. 1971. The calculation of acute toxicity of free
chlorine and chloramine to coho salmon by multiple regression
analysis. Thesis. Michigan State University. East Lansing,
Michigan.
*Sawyers, C. N., and P. L. McCarty. 1969. Chemistry for sanitary
engineers. 2nd Ed. McGraw-Hill. New York.
Servizi, J. A., and D. W. Martens. 1974. Preliminary survey of toxicity
of chlorinated sewage to sockeye and pink salmon. Int. Pac. Salmon
Fish. Comm. Prog. Rep. 30: 1-42.
Stober, Q. J., and C. H. Hanson. 1974. Toxicity of chlorine and heat to
pink, Onahrhynchus gorbuscha, and chinook salmon, 0. tshcajytsaha.
Trans. Am. Fish. Soc. 103(3): 569-576.
Sugam, R., and G. R. Helz. 1975. Apparent ionization constant of
hypochlorous acid in seawater. Unpublished manuscript. University
of Maryland. College Park, Maryland.
Tsai, C. 1968. Effects of chlorinated sewage effluents on fishes in
upper Patuxent River, Maryland. Chesapeake Sci. 9(2): 83-93.
Tsai, C. 1970. Changes in fish populations and migrations in relation
to increased sewage pollution in Little Patuxent River, Maryland.
Chesapeake Sci. 11(1): 34-41.
Tsai, C. 1973. Water quality and fish life below sewage outfalls.
Trans. Am. Fish. Soc. 102(2): 281-292.
Tsai, C. 1975. Effects of sewage treatment plant effluents on fish:
a review of the literature. Chesapeake Research Consortium
Publication 36: 1-229.
Literature not cited.
-------
322
Turner, H. J., D. M. Reynolds, and A. C. Redfield. 1948. Chlorine and
sodium pentachlorophenate as fouling preventatives in seawater
conduits. Ind. Eng. Chem. 40: 450-453.
Virginia State Water Control Board. 1974. James River fish kill 73-025.
Bureau of Surveillance and Field Studies, Division of Ecological
Studies. 61 p.
Waugh, G. D. 1964. Observations on the effects of chlorine on the larvae
of oysters, Ostrea edulis L., and barnacles, Eliminius modestus
Darwin. Ann. Appl. Biol. 54: 423-440.
*White, G. C. 1972. Handbook of chlorination. Van Nostrand Rheinhold
Co. New York. 744 p.
*White, G. C. 1973. Disinfection practices in the San Francisco Bay
area. J. Water Pollut. Control Fed. 46: 89-101.
Whitehouse, J. W. 1975. Chlorination of cooling water: a review of
literature on the effects of chlorine on aquatic organisms. Job
VJ 440-RD/L/M496;l-22. Central Electricity Research Laboratory.
Young, P. H. 1964. Some effects of sewage effluents on marine life.
California Fish and Game 50(1): 33-41.
Literature not cited.
DISCUSSION
James H. Carpenter, University of Miami. Are there any studies of
behavioral response of marine fish?
Davis. Yes. The ones we are operating and the ones John Meldrim
is operating are simplified types of avoidance responses in juvenile
fish. In both cases the fish were successful and the spot were success-
ful in avoiding lethal concentrations, that is, based on the LDso's at
10°C and above. But in the case of the spot, they were unsuccessful in
avoiding lethal concentrations below 10°C. There are, of course, the
behavioral papers referred to in the previous presentation.
Arthur Brooks, University of Wisconsin-Milwaukee. In Bill Brungs'
recent review article there are a number of estuarine studies made in
the last ten months referring to chlorine and invertebrates.
-------
323
Davis. The questions and statements refer basically to the review
of Brungs and, I might add another review by Whitehouse in 1975, which
both deal with estuarine invertebrates. I have avoided discussing these
in great detail in the oral presentation. We did summarize them in the
version that will be printed. We avoided this from the simple point of
view that we can't tell from the studies in most cases just what technique
they are using for residual chlorine measurements and, therefore, I don't
find them intercomparable.
Donald J. Modell, Electrode Corporation. Due to papers discussing
the measurement of HOC1 and total chlorine residual by Dr. Johnson, how
are you planning to separate the various components of total chlorine
residual via measurement methods to ensure that your results on HOC1 and
chlorinated organics can be directly correlated to marine ecosystem
effects? And what measurement methods can or are you planning to use to
obtain meaningful results?
Davis. I would like to challenge that meaningful characteristic,
right there.
Modell. Meaningful to Dr. Johnson.
Davis. Meaningful to Dr. Johnson may not necessarily be meaningful
to the environment. What we are looking for are ecological effects. So
in essence we are looking for summation effects. Now, just in a very
general way, if you are going to run LC50 tests side by side you need
something to measure for comparison. We do our calculations, our
dilutions, and use an amperometric titrator to give us a number. Now
if we were to compare at a given moment the reactions of spot, a small
juvenile fish, to the reactions of a typical estuarine decapod Poleomynetes
we would get vastly different responses. Now at this point we ask our-
selves your question, "What is meaningful?" We have a number to index
ourselves to and we have references in terms of LC50 of what the organism
response is. In general terms the response of the fish is much more
sensitive than for this particular vertebrate. Therefore, we have to
ask the question, "Why?" We can make a suggestion, epithelium. Maybe
this is reinforced by the observation made from the swimming pool study
where an algae with a siliceous shell and other particular adaptive
mechanisms became relatively immune to chlorine effects compared to a
fish which has a weak point at the gill surface. Now, did I approach
your question or did 1 avoid it?
Modell. As I understand it you will continue to use present measure-
ment methods.
Davis. We will use standard methods because what else do we have
available. In the meantime, we are asking ourselves, "What else is going
to become available?"
-------
324
T. 0. Thatcher, Batelle-Northwest Marine Research Laboratory.
There is going to be quite a bit of information appearing in the
literature concerning organisms exposed to chlorine in salt water.
In reporting LC50 values, what physical and chemical parameters of the
test water should also be reported with the total residual oxidant data
to make the information more useful?
Davis. Well, I can't really comment too much on the bromine levels
right now until we have the analytical technique. Other things? Of
course it is standard to say temperature. In marine waters I think you
will have to define very carefully the site from which you get your data.
If you wanted to transfer the data we are getting at Bears Bluff to the
northwest — I would be anxious about that. For example, we have a very
high sediment load in our estuaries. One part of our strategy for next
year is to actually work more with the influence of entrained particles
and the effects on what is called chlorine demand. Other factors are
sunlight, oxygen, and contained organics.
William A. Brungs, U.S. Environmental Protection Agency. I think
it's a good question because I'm not a chemist, but if there is any
possibility, additional information should be included with the pub-
lished data on LC50. Maybe a few years from now somebody can go back
and make some back calculations and perhaps use those data. Whereas,
right now, it would be very difficult because most people do not explain
the test conditions adequately to make that transposition.
George C. White, Consulting Engineer. Because the residual chemistry
in seawater is so complex you do not know what is reacting. Maybe a model
could be developed with part of the model based on the chlorine consumption
of the test water. You can measure dosage. That is, you know what you are
putting in and you can measure the chlorine consumption. Thus you could
make a model for each site and work back from that. But I would not want
to trust the residuals.
Davis. No. I think the residual measurements have too much weight
put on them.
Robert J. Huggett, Virginia Institute of Marine Science. Bioassay
results are dependent on many things but one which should be noted for
most anadromous fish in the Chesapeake Bay is that they appear to be
more sensitive to chlorine derivatives in the spring soon after returning
from the ocean.
Also one point that must not be forgotten when we work on chlorine
in the environment is that even though we may not know what we are
measuring when we analyze for chlorine-chloramines, we do get a meter
reading. This meter reading seems to be proportional to the chlorine
related toxicity of the water and can be used to predict acute biological
effect after bioassays have been performed. It is essential, therefore,
that we not lose sight of the fact that some management decisions can be
made based on "meter readings" and that what is killing the animals
(bromine, chlorine, hypochlorous acid, monochloramine, or so forth) is
proportional to the chlorine dose.
-------
325
Davis. I think another point that you have just illustrated is the
need for going ahead with field management and not waiting for bioassays.
According to T. 0. Thatcher, the following points were made during
the intermission which are appropriate to include in the context of the
discussion because of their relevance to data which should be included
in reports:
George R. Helz, University of Maryland, suggested that the pH value
of the test water, the phenol content, and total proteinaceous nitrogen
should be included.
Edward G. Wolf, Batelle-Pacific Northwest Laboratories, and Ronald
M. Block, University of Maryland, suggested that total organic carbon,
iron, manganese, turbidity, total nitrogen, total organic nitrogen,
nitrate, nitrite, ammonia, and salinity should be included.
-------
CHLORINATED COMPOUNDS FOUND IN WASTE-TREATMENT EFFLUENTS
AND THEIR CAPACITY TO BIOACCUMULATE
Herbert L. Kopperman
Department of Chemistry
University of Minnesota-Duluth
Duluth, Minnesota 55812
Douglas W. Kuehl, and Gary E. Glass
Environmental Research Laboratory-Duluth
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
ABSTRACT
As part of an ongoing research program to assess possible long-term
environmental effects due to the formation of stable reaction products
during disinfection processes, fish (fathead minnows, Pimephales promelas)
and water from the 9-month chronic toxicity tests at two wastewater
treatment plants in Michigan are being analyzed for chemical residues at
this laboratory (formerly named the National Water Quality Laboratory).
Gel permeation chromatography was used for sample clean-up and gas
chromatography/mass spectrometry was used for sample analysis. Di- and
trichlorophenols, di- and trichlorobenzenes, and trichloroanisoles were
either not detected or detected at lower levels in the fish from nondis-
infected effluent exposures compared to fish exposed to chlorinated
effluent. Tetra- and pentachlorophenols, PCB's, DDT's, toxaphene com-
ponents, chlordane, and nonachlor were found in all fish raised in the
sewage effluent. Tribromoanisole was tentatively identified in fish
that lived in BrCl-treated wastewater.
Reports appear to be conclusive in support of the argument that
even the most gentle chlorination conditions will cause chlorine to be
incorporated into organic molecules. The incorporation of chlorine into
an organic molecule increases its lipophilic character and at the same
time normally causes an increase in the observed toxicity or bioaccumula-
tion, or both.
The persistence of these compounds is now becoming a concern. Not
all organochlorine compounds bioaccumulate to high levels. The data
327
-------
328
suggests that polar compounds are more easily biodegraded, and the
nonpolar (highly lipophllic) compounds accumulate. Some investigators
have been able to demonstrate positive correlation between the n-octanol/
water partition coefficients for given compounds and their ability to
bioaccumulate in various species of fish.
Chlorination has been used extensively for disinfection of waste
effluents by both industry and municipalities. In addition, it was
assumed that chlorine was oxidizing the organic compounds to innocuous
substances (C02» H£0, and easily degradable organics) and thereby lower
the BOD of the effluent. It has now been shown that although chlorine
does oxidize many compounds, it also creates many organochlorine com-
pounds which are more toxic. Carlson (1975) has shown that chlorinated
phenols lower the BOD by destroying the bacteria. Carlson et al. (1975)
have also shown that phenol incorporates chlorine very readily over a
wide range of pH, while under conditions similar to disinfection
(Table 1).
Table 1. Percentage of available chlorine incorporated by
various compounds treated with 7 x 10~4 M aqueous
chlorine at three pH levels (3, 7, and 10)
for 20 minutes, at 25°Ca
Compound^
Phenol
Anisole
Acetanilide
Toluene
Benzyl alcohol
Benzonitrile
Nitrobenzene
Chlorobenzene
Kethylbenzoate
Benzene
3
97.8
80.7
55.3
11.1
2.3
2.1
1.8
1.8
1.8
1.5
pH
7 10
97.6 97.6
11.4 2.8
3.4
2.9
- -
- -
- -
- -
- -
- -
aR. M. Carlson et al. (1975).
Aqueous concentration = 9.5 ± 0.6 x 10~4 M.
-------
329
Chlorination of water supplies has received considerable attention
(Jolley 1973, 1975, Jolley et al. 1975, Glaze and Henderson 1975, Glaze
et al. 1973, Bellar et al. 1974) in the last few years (Tables 2, 3,
and 4). People are becoming concerned not only with the production of
toxic materials but also with the production of possible carcinogens.
Table 2. Tentative identification and concentration of chlorine-
containing constituents in chlorinated effluents
(Oak Ridge, Tennessee) by liquid chromatographic techniques'2
Organic compound Concentration (ug/liter)
5-chlorouracil 4.3
5-chlorouridine 1.7
8-chlorocaffeine 1.7
6-chloroguanine 0.9
8-chloroxanthine 1.5
2-chlorobenzoic acid 0.26
5-chlorosalicylic acid 0.24
4-chloromandelic acid 1.1
2-chlorophenol 1.7
4-chlorophenylacetic acid 0.38
4-chlorobenzoic acid 1.1
4-chlorophenol 0.69
3-chlorobenzoic acid 0.62
3-chlorophenol 0.51
4-chlororesorcinol 1.2
3-chloro-4-hydroxybenzoic acid 1.3
4-chloro-3-methylphenol 1.5
aJolley (1975).
To answer some of the questions being raised a program was initiated
in Grandville and Wyoming, Michigan, to observe the effect of various
disinfection processes upon the chemistry of organic molecules in munic-
ipal waste effluent. The four treatments examined were (1) Chlorination,
(2) Chlorination followed by dechlorination with S02 treatment, (3)
ozonation, and (4) chlorobromination (Fig. 1).
-------
330
Table 3. Chlorinated organlcs in wastewater effluent (Denton, Texas)
identified by gas chromatography/mass spectrometry methods*
Compound0 Concentration^1 (yg/1)
Chloroform"3 —
Dibromochloromethanec —
Dichlorobutane (-)" 27
3-chloro-2-methylbut-l-enec 285
Chlorocyclohexane (118)^ 20
Chloroalkyl acetate (-)<* —
0-dichlorobenzenec 10
Tetrachloroacetone6 11
p-dichlorobenzene^ 10
Chloroethylbenzenee 21
Pentachloroacetone"5 30
Hexachloroacetonec 30
Trichlorobenzenec —
Dichloroethylbenzene0 20
Chlorocumene (154)"
N-methyl-trichloroaniline (209)^ 10
Dichlorotoluene6 —
Trichlorophenol6 —
Chloro-a-methyl benzyl alcohol6 —
Dichloromethoxytoluene6 32
Trichloromethylstyrene (220)^ 10
Trichloroethyl benzene (208)" 12
Dichloro-a-methyl benzyl alcohol (190)^ 10
Dichloro-bis(ethoxy)benzene (220)" 30
Dichloro-a-methyl benzyl alcohol (190)^ —
Trichloro-a-methyl benzyl alcohol6 25
Tetrachlorophenol0 30
Trichloro-a-methyl benzyl alcohol6 50
Trichlorocumene (222)^ —
Tetrachloroethylstyrene (268)" ,
Trichlorodimethoxybenzene (240) —
Tetrachloromethoxy toluene (258)" 40
Dichloroaniline derivative (205)^ 13
Dichloroaromatic derivative (249)-^ 15
Dichloroacetate derivative (203)J 20
Trichlorophthalate derivative (296)-^
Tetrachlorophthalate derivative (340)-'
_
Glaze and Henderson 1975.
Compounds may be listed more than once if gas chromatographic retention
times indicate distinct positional isomers.
Quantitative values should only be considered as estimates because explicit
recovery data are not available for the authors1 extraction system.
^j
Completed identification based on mass spectrometric interpretation and
confirmed by comparison with a reference spectrum.
Fragmentation pattern tentatively suggests proposed compound; probable
molecular weight indicated in parentheses.
Probable identification based on mass spectral interpretation.
Mass spectral information too incomplete
molecular weight indicated in parentheses.
f
Mass spectral information too incomplete to propose a structure; probable
-------
331
Table 4. Concentration of organochlorine compounds (pg/liter)
in water from sewage treatment plants (several cities)
measured by gas chromatography/mass spectrometry techniques'2
Compound
Methylene chloride
Chloroform
1, 1, 1- trichloroe thane
1,1, 2-trichloroethylene
1, 1, 2, 2-tetrachloroethy lene
2 Dichlorobenzenes
£ Trichlorobenzenes
Influent
before
treatment
8.2
9.3
16.5
40.4
6.2
10.6
66.9
Effluent
before
chlorination
2.9
7.1
9.0
8.6
3.9
5.6
56.7
Effluent
after
chlorination
3.4
12.1
8.5
9.8
4.2
6.3
56.9
a
:T. A. Bellar et al. (1974).
One of the main objectives of this study was to investigate the
chlorinated effluent for newly formed chloro compounds. It has been
shown that phenols will incorporate chlorine under the conditions nor-
mally encountered in water treatment plants (Carlson et al. 1975). The
number of chlorines incorporated depends on the reactivity of the
particular phenol toward electrophilic attack. Phenol itself was not
identified in either the water or the fish; however, di'chloro- through
pentachlorophenols were identified. An attempt was made to identify as
many organics as possible by using the GC/MS (gas chromatography/mass
spectrometry) system. The usual chlorinated pesticides and PCB's were
identified in the fish as well as many aromatic hydrocarbons (Table 5).
Di- and trichlorobenzene were only observed in the fish reared in the
chlorinated effluents (Fig. 2).
An unreported compound type, anisole, was identified in both the
water and the fish. Analysis of the effluent indicated that the parent
phenols were present in larger quantities than the corresponding
anisoles, and the fish analysis showed that the anisoles had become the
more abundant of the two. Figure 3 shows a representative sample of
anisole data illustrated by one of these compounds, pentachloroinisole.
-------
332
NON-DISINFECTED
ACTIVATED SLUDGE
EFFLUENT
CI2
11
FISH
DILUTION
WATER
r
FISH
FISH
Fig, 1. Schematic diagram indicating the experimental disinfection
processes used in the chronic bioassay tests at the Environmental Pro-
tection Agency's Grandville, Michigan, project.
-------
333
Table 5. Chlorinated organic compounds identified by gas
chromatography/mass spectrometry in fish raised in
disinfected municipal effluent (Grandville, Michigan)
Compound Identification01
Dichlorobenzene A
Trichlorobenzene A
Trichlorophenol B
Trichloroanisole B
Tetrachlorophenol B
Tetrachloroanisole (2) A/B
Pentachlorophenol B
Pentachloroanisole B
Trichloropropylmethoxynaphthalene C
Tetrachloropropylmethoxynaphthalene C
<3is-chlordane B
trans-chlordane B
-------
334
DICHLOROBENZENE
100
cn
LU
I-
UJ
5O
IOO
MASS
ISO
2OO
30,000
>-
h;
z
20.OOO
10,000
6.7%
CI
50%
BrCI
IOO%
Non-Dis.
IOO%
IO 2O 0 10 2O O 10 2O O
SPECTRUM NUMBER
50%
10 20
10 2O
Fig. 2. Mass spectrum and reconstructed mass chromatograms (M* 146)
of dichlorobenzene found in fish raised in chlorinated sewage effluent.
The disinfectant and percent effluent concentration are given in the
lower figure. The spectrum number is a guide to the gas chromatographic
retention time.
-------
335
PENTACHLOROANISOLE
IOO
UJ
§
b
50
IOO
150 2OO
MASS
250
300
30,000
20,000
(O
UJ
10,000
6.7%
CI2
50%
BrCI
100%
i-Di&
100%
10 0 10 O 10 20 0 10 20
SPECTRUM NUMBER
50%
CI2/S02
10 20
Fig. 3. Mass spectrum and reconstructed mass chromatogram (M 278)
of pentachloroanisole found in fish raised in sewage effluent. The
disinfectant and percent effluent concentration are given in the lower
figure. The spectrum number is a guide to the gas chromatographic
retention time.
-------
336
It is apparent that some compounds have a greater tendency to
accumulate in the tissue than others. This phenomenon has been investi-
gated by Macek et al. (1975) and Neely et al. (1974). They have shown
that bioaccumulation is directly related to the partition coefficient
(P) of the compound.
The use of the n-octanol/water partitioning system in correlating
observed responses to compounds is not new. In 1899 Overton and Meyer
correlated the solubility characteristics (n-octanol/water partition
coefficient, P) of anesthetics to their observed biological response
(log 1/C) (Overton 1899, Meyer 1899, Meyer and Hemmi 1935, Meyer 1937):
log 1/C = a log P + b . (1)
In this equation C is the molar concentration of a given compound
required for a particular response (narcosis, LC50, LD50). This original
work dealt with isonarcosis in tadpoles, and the chemicals studied were
alcohols, ketones, esters, and aromatic hydrocarbons. Ferguson (1939)
placed the Overton-Meyer theory on a quantitative basis by defining the
free energy change, AG, in terms of the chemical potential, y.
In the 1960's Hansch and coworkers (1969) began applying these
theories and were able to correlate many of the previously reported data
by using the n-octanol/water system as the basic criterion. Electronic
and steric factors have since been included to improve the correlation.
Hansch introduced the need of a parabolic relationship in addition to a
straight line relationship because as a compound became more lipophilic
(higher partition coefficient, P) it tended to remain in the membranes
and not be transferred to the active site and therefore caused a decrease
in the observed activity. A parabola would be the best empirical
representation for this type of observed biological response.
The onset of bioaccumulation is observed when the compounds start
to remain in the cell membranes. It has been shown that the most
likely place for the accumulation of lipophilic compounds is in the
lipids of the animal, fish, or bird. At this point these compounds are
transferred from water to birds throughout the environment by the food-
chain mechanism.
-------
337
Kapoor et al. (1973) studied food-chain mechanisms extensively and
have shown how various chemicals are transported by this means.
Bemelmans and ten Noever de Brauw (1974) observed this process in
chickens that had been fed contaminated food. The contamination involved
chloroanisoles of the same type as those isolated from the fish raised
in the chlorine-treated effluent from Grandville, Michigan. Chlorinated
anisoles have been reported (Engel et al. 1966, Curtis et al. 1972,
Cserjesi and Johnson 1972) to impart a distinct musty taint to both eggs
(yolks) and flesh of chickens. One source of the contamination was the
wood shavings in the poultry houses, which were produced from lumber
treated with technical grade pentachlorophenol (Impurities constitute
mainly isomeric tetrachlorophenols). The bacteria in the poultry litter
were shown to be capable of methylating the chlorophenols on the surface
of the wood shavings. It was also shown that some bacteria were capable
of dechlorinating the higher chlorinated phenols. The problem persisted
even when wood shavings were not used. This resulted in the discovery
that the poultry feed was contaminated with trichloroanisole, which has
a very low taste threshold (Table 6).
Table 6. Taste detection threshold of compounds
in aqueous solution^
Compound Concentration (yg/g)
^ Q
Pentachloroanisole 4 x 10
2,3,4,6-tetrachloroanisole 4 x 10"6
2,4,6-trichloroanisole 3 x 10~8
2,3,6-trichloroanisole 3 x 10~10
aCurtis et al. 1972.
Bemelmans and ten Noever de Brauw (1974) were unable to ascertain
the cause of the tainted feed. Investigation has determined that fish
products have been used in various formulations of chicken feed and in
light of the contaminated Grandville fish it is highly possible that
contaminated fish could have been involved. Thus aqueous chlorination
may create a problem as large as the initial one that required its use
as a disinfectant.
-------
338
Anisoles are more persistent as compared to the parent phenols.
The methylation of phenol increases the partition coefficient from 1.48
to 2.04. Chlorination also increases the partition coefficient (0.7
units for aromatics and 0.4 units for aliphatics). Both of these increase
the bioconcentration factor.
Neely et al. (1974) have shown with trout (Salmo gairdneri
Richardson) that the bioconcentration factors of organic compounds when
correlated to their n-octanol/water partition coefficients (P) resulted
in a straight line (equation 2):
log (bioconc. factor) = 0.542P + 0.124 r = 0.948 s = 0.342 (2)
In their study a series was chosen so that the partition coefficients
were spread over a wide range. The bioconcentration factor was deter-
mined by the ratio of the rate of chemical uptake (k^) in the test
solution to the rate of chemical clearance (kc) in fresh water. Figure
4 gives a graphical representation of their results. Table 7 illustrates
the predictive application of equation 2. Mosquitofish (Gcaribus-ia affinis)
were used for these bioassays, which demonstrate the importance of the
use of partition coefficients as a predictive tool.
A standard approach to determining the bioconcentration factor
should be defined so that data can be reported in a standard way and
more important can be more easily compared to previously reported data.
A reasonable standard expression would be the ratio of the initial uptake
rate (k-j^) in the test solution to the clearance rate (kc) in fresh water.
With this in mind three basic patterns have been observed during a
28-day exposure: (1) k^ > kc, continuous increase in residue concentra-
tion during exposure; (2) ^ - kc, initial residue concentration increase,
until an unchanging level is observed; and (3) k.^ < kc, initial residue
concentration increase followed by a continuing decrease in concentra-
tion with exposure time. The initial rate constant will be dependent
on the partition coefficient (P) and independent of kc because the com-
pound has not reached the site of action in large enough quantities to
be of concern under the conditions of the bioassay. In order that kc
-------
339
o
£
O
O
C 3
• Confidence
Region
2 3456789
Log Partition Coefficient
Fig. 4. Linear regression between logarithms of n-octanol-water
partition coefficient and bioconcentration of chemicals in trout muscle
(Neely et al. 1974). Compounds plotted are: (1) 1,1,2,2-tetrachloro-
ethylene; (2) carbon tetrachloride; (3) p-dichlorobenzene; (4) diphenyl-
ether; (5) diphenyl; (6) 2-biphenylphenylether; (7) hexachlorobenzene;
and (8) 2,2',4,4f-tetrachlorodiphenylether.
-------
340
Table 7. The use of regression equation 2 for
predicting the bio concentration factor'2
Chemical
Endrin
Chlorpyrifos^
3,5»6-trichloro
pyridinol
Log partition
coefficient
5.6
4.82
1.35
Log bioconcentration
factor
Calculated
3.47 ± 0.989
2.87 ± 0.963
0.88 ± 1.139
Experimental
3.17
2.67
0.49
^Neely et al. 1974.
0,0-diethyl-0-(3,5,6-trichloro-2-pyridyl)phosphorothioate,
be independent of k^ in the second step, the fish have to be removed
from the test solution and placed in fresh water. In this way one should
be able to observe only the clearance mechanism. There may be several
steps in each of these two general mechanisms, but the rate should be
dependent only on the slowest step.
These examples illustrate some of the possible ramifications of
the use of partition coefficients in an attempt to predict not only
toxicity but bioaccumulation as well.
SUMMARY
A method is needed to rapidly screen the large number of organic
compounds which if allowed to be discharged will contaminate our envi-
ronment. The pharmaceutical industry has used water-solvent partition
coefficients with great success in determining which compounds might
be most biologically active. These coefficients can also be used as a
predictive tool in the environmental world to study both toxicity and
bioconcentration of organic contaminants. Unfortunately, the problems
that arise with bioaccumulation of organic compounds which have been
discharged into the environment are not immediately apparent. By the
time a problem is identified it is usually too late and a major segment
of the ecosystem has been contaminated. New products should be
-------
341
manufactured with this in mind, and a complete data base of bioaccumula-
tion data needs to be developed to set reasonable guidelines which
industry can use.
The incorporation of chlorine into compounds during the disinfection
of waste effluents with chlorine is an undesirable end result of efflu-
ent treatment in that compounds become more persistent and bioaccumulate
to a greater extent.
EXPERIMENTAL
Compound library generation. A large number of compounds which
were predicted to be found in the chlorinated effluent were purchased
or prepared by one of the following two procedures:
Procedure A. Five hundred milligrams of the starting organic
compound was dissolved in 20 ml CHCla. Chlorine gas was bubbled into
the solution at room temperature for 30 min. The excess chlorine was
then stripped off with CHC13
Procedure B. Five hundred milligrams of the starting organic
compound was dissolved in 20 ml CCl^ Chlorine gas was bubbled into the
solution at room temperature for 30 min. The solution was then irradiated
with UV light for 10 min at 254 nm. The excess chlorine was then stripped
off with CCli,.
GC retention time and mass spectral data were then obtained on each
product. The mass spectra were loaded onto magnetic tape for library
searches during sample investigations.
Water Analysis. The three water samples taken for examination were:
(1) 8.0 gal of nondisinfected effluent, (2) 9.0 gal of chlorinated
effluent, and (3) 4.5 gal of chlorinated effluent adjusted to pH 2.0.
Samples were obtained by running the desired effluent through a glass
column (2.5 x 20 cm) filled with a 50-ml bed volume of XAD-2 resin
(Rohm and Haas). Each column was washed with 200 ml of 0.1 M NaOH,
200 ml 0.1 M HC1, and finally 200 ml of methanol. The acid and base
washes were neutralized and extracted with three 50-ml portions of
ether. The extracts were concentrated to 10 ml and the base extract
methylated with diazomethane for GC and GC/MS analysis.
-------
342
Fish Analysis;. Whole fish (30 to 50 g) were blended with enough
anhydrous sodium sulfate to make a dry powder and extracted with four
100-ml portions of hexane/ether (1:1). The extracts were concentrated,
and the resulting oil residue was diluted with cyclohexane to 1 g/5 ml.
The oil residue solution was placed on a 2.5-cm x 30-cm SX-2 (Bio-Rad
Laboratory) gel permeation column and eluted with cyclohexane at 3.5 ml/
min. The first 150 ml was vented, and the next two 60-ml fractions were
collected. The collected fractions were combined and concentrated by
using a Kuderna-Danish flask fitted with a three-ball Snyder column.
Volumes were adjusted with hexane for GC and GC/MS analysis. Fish were
obtained from the following bioassay tanks: (1) 6.7% Cl2, (2) 50%
C12/S02, (3) 50% BrCl, (4) 100% 03, and (5) 100% nondisinfected.
Instrumental Conditions. All gas chromatograms were run on a Varian
Areograph 1700 equipped with a 6-ft x 1/8-inch glass column packed with
3% OV-101 on 80/100 mesh gas chrom Q. All GC effluent was monitored by
electron capture-flame ionization detectors with a 1:50 splitter. Injec-
tion port and detector block temperatures were maintained at 230°C.
The mass spectrometer system was a Varian MAT CH-5 equipped with a
Varian MAT mass spectrosystem 100 MS data system. All mass spectra
were obtained at 70 eV with a 5-sec-per-mass decade scan.
ACKNOWLEDGMENTS
The authors acknowledge the help and assistance of M. DeGraeve and
R. Ward, and others, who conducted the bioassays, and to G. Veith for
collecting samples and providing useful advice concerning the analysis.
REFERENCES
Bellar, T. A., J. T. Lichtenberg, and R. C. Kroner. 1974. EPA-670/4-
74-008. U.S. Environmental Protection Agency. November.
Bemelmans, J. M. H., and M. C. ten Noever de Brauw. 1974. J. Agr. Food
Chem. 22: 1137.
Carlson, R. M. 1975. Personal communication. See also R. M. Carlson
and R. Caple, Organo-chemical implications of water chlorination
(this conference).
-------
343
Carlson, R. M., R. E. Carlson, H. L. Kopperman, and R. Caple. 1975.
Environ. Sci. Technol. 9' 674. (And references cited therein.)
Cserjesi, A. J., and E. L. Johnson. 1972. Can. J. Microbiology 18: 45.
Curtis, R. F., D. G. Land, N. M. Griffiths, M. Gee, D. Robinson,
J. L. Peel, C. Dennis, and J. M. Gee. 1972. Nature 235: 223.
Engel, C., A. P. de Groot, and C. Weurman. 1966. Science 154: 270.
Ferguson, J. 1939. Proc. Roy. Soc. (London) Ser. B, 127: 387.
Glaze, W. H., and J. E. Henderson, IV. 1975. J. Water Pollut. Control
Fed. 47: 2511.
Glaze, W. H., J. E. Henderson, IV, J. E. Bell, and V. A. Wheeler. 1973.
J. Chrom. Sci. 11: 580.
Hansch, C. 1969. Ace. Chem. Res. 2: 232. (And references cited
therein.)
Jolley, R. L. 1973. Ph.D. Thesis. University of Tennessee, Knoxville.
Jolley, R. L. 1975. J. Water Pollut. Control Fed. 47: 601. (And
references cited therein.)
Jolley, R. L., S. Katz, J. E. Mrochek, W. W. Pitt, Jr., and W. T. Rainey.
1975. Chem. Technol. 5: 312. (And references cited therein.)
Kapoor, I. P., R. L. Metcalf, A. S. Hirwe, J. R. Coats, and M. S. Khalsa.
1973. J. Agr. Food Chem. 21: 310.
Macek, K. J., M. E. Barrow, and R. F. Frasny. 1975. International
Joint Commission Symposium on Structure-Activity Correlations in
Studies of Toxicity and Bioconcentration with Aquatic Organisms,
p. 119. (And references cited therein.)
Meyer, H. 1899. Arch. Exptl. Pathol. Pharmakol. 42: 109. As cited,
p. 134 in A. Burger (ed.) Medicinal chemistry, 3rd ed. 1970.
Wiley-Interscience, New York.
Meyer, K. H. 1937. Trans. Faraday Soc. 33: 1063. As cited, p. 134 in
A. Burger (ed.) Medicinal chemistry, 3rd ed. 1970. Wiley-
Interscience, New York.
Meyer, K. H., and H. Hemmi. 1935. Biochem. Z. 277: 29; 282: 444, 447.
As cited, p. 134 in A. Burger (ed.) Medicinal chemistry, 3rd ed.
1970. Wiley-Interscience, New York.
Neely, W. B., D. R. Branson, and G. E. Blau. 1974. Environ. Sci.
Technol. 8: 1113.
-------
344
Overton, E. 1899. Viertljahrsschr. Naturforsch. Ges. Zurich 44: 88.
As cited, p. 134 in A. Burger (ed.) Medicinal chemistry, 3rd ed.
1970. Wiley-Interscience, New York.
DISCUSSION
Alan H. Molof, Polytechnic Institute of New York. There are two
parts to my question. First, in your testing protocol, did you use
a clean tap water with the disinfectant you used for the activated sludge
effluent? This would be valuable since it would relate to the effect of
the disinfectants themselves, such as the effect of chlorine itself on
fish. Second, have you or anyone else looked into the effects of dis-
infectants themselves in drinking water on man? Even though the con-
ference title is "Environmental Impact of Water Chlorination," there
does not seem to be any coverage of this subject. The importance of
this area comes from the use of chlorine residuals in our tap water.
Kopperman. Well, again it comes to grant money. I am assuming
such studies are being done in other places such as at the EPA Laboratory
in Cincinnati.
Herman F. Kraybillt National Cancer Institute. Bioaccumulation
and partition coefficients cause a problem in bioassay for carcino-
genicity of organochlorine compounds. In setting an MTD (maximum tol-
erated dose) one proceeds for 6 to 7 months in the test and then the
rodents being tested are killed due to the concentration increase in
the adiposity. Even at 0.5 MTD, lethalities occur several months later
and one is not able to pursue the bioassay. One has to start over at
much lower doses where the bioaccumulation will not overload the
metabolism.
Kopperman. Actually all of us are a little overweight. We have
probably stored more than our share of some of the chlorinated compounds.
All sorts of things might happen if we were to go on crash diets. These
compounds have to go somewhere when you start getting rid of your fat
tissue. They go right into your system.
Jerry J. Nelson, U.S. Energy Research and Development Administration.
You have what appears to be classic linear first order kinetics. Have
you made any attempt to build a model for the uptake and movement such
as a Van Dyne type descriptive model?
Kopperman. No. I have not because as it turns out I am no longer
associated with EPA. I am not presently funded to do this type
research.
-------
345
Nelson. In addition, the slide showing initial uptake rates which
overshoot and decay may involve two processes such as sorption and
intake with subsequent desorption. Or do you think this perhaps is
decay in addition to clearance due to metabolic decomposition or some
such mechanism?
Kopperman. Yes, those are possible mechanisms.
-------
INVESTIGATING THE EFFECTS OF CHLORINATED ORGANICS
Carl W. Gehrs and George R. Southworth
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
The recent identification of stable chlorine-containing organics
arising from the chlorination of natural waters has revealed a group
of reaction products whose toxicities to aquatic organisms are unknown.
In this paper we present information on the toxicity of two chlorinated
compounds (5-chlorouracil and 4-chlororesorcinol) and a mixture of
identified chlorinated organics to zooplankton and fish. We compare
data emphasizing differences in relative toxicity depending on the
response parameter used. Problems associated with studying individual
compounds and complex mixtures are discussed and a systematic approach
for overcoming the identified shortcomings presented.
INTRODUCTION
Research designed for evaluating the environmental implications
of chemical releases has as its goal the development of the data
necessary for determining the probability of, and kinds of, adverse
effects that may impinge on man or his environment. Critical to
attaining this goal is an understanding of the kinetics, transformation
products, and ultimate fate in the environment of the chemical of
concern.
Chlorine has been used as a disinfectant for many years, with an
estimated 60,000 tons being added to effluents from sewage treatment
plants in the United States during 1962 (Laubusch 1962). Although
the fate of reactive chlorine residuals in aquatic environments is
very well documented (Palin 1950, Draley 1972), little is known about
either the fate or effects of the recently identified stable-chlorine
347
-------
348
containing organics (Jolley 1973). We have decided to emphasize the
stable chlorine-containing organics in our effects research because they
are more persistent and, consequently, have a greater potential for
adversely impacting the aquatic environment. The purpose of this
paper is to present information on the effects of chlorinated organics
and to outline a method for organizing a research program aimed at
evaluating the potential problem arising from chlorinated organics.
METHODS AND MATERIALS
In the initial investigations into the potential formation of
stable chlorine-containing organics, Jolley (1973 and 1975) employed
high-resolution liquid chromatography to analyze secondary sewage
o/r
effluents treated to 1 mg/1 chlorine residual with Cl-tagged chlorine
gas or hypochlorite solution. He found more than 60 specific peaks
of which 17 individual compounds have been tentatively identified at
microgram-per-liter concentrations. The compounds identified are of
three classes: phenols, purines and pyrimidines, and aromatic
acids. Two compounds were chosen for initial investigation: the
pyrimidine, 5-chlorouracil (because of its potential for incorporation
into DNA as a base analog) and a phenol, 4-chlororesorcinol (because
of the toxicity of phenols and the relatively high concentration, 1.2
yg/1, found by Jolley). Later it was decided to produce and test a
synthetic mixture made up into a stock solution of 1 g/1 total chlorinated
organics in the proportions in which they were found in the treated
secondary effluent (Table 1).
Toxicity studies were conducted using two species of aquatic
organisms, the zooplankter Daphnia magna and the fish Cypvinus Garp-Lo
(carp). The parameters used were mortality and maturation in the
zooplankter and hatching success (mortality) and malformation of fry
for the fish.
Zooplankton studies. Zooplankton studies were conducted in 100-
ml beakers containing 80 ml of appropriate concentration of the test
solutions in spring water. Ten replicates were used at each concentra-
tion with two individuals (12 + 12 hr old) added to each beaker to
-------
349
Table 1. Composition of synthetic chlorinated effluent
Percent
5-chlorouracilb 20.8
5-chlorouridine 8.2
8-chlorocaffeine 8. 2
6-chloroguanine 4.3
8-chloroxanthine 7.2
5-chlorosalicylic acid 1.2
4-chloromandelic acid 5.3
2-chlorophenol 8.2
3-chlorophenol 2.5
4-chlorophenol 3.5
4-chloro-3-methylphenol 8.2
4-chlorophenylacetic acid 1.8
4-chlorobenzoic acid 5.3
2-chlorobenzoic acid 1.3
3-chloro-4-hydroxybenzoic acid 6.3
3-chlorobenzoic acid 3.0
rt
4-chlororesorcinol 5.9
Composition was determined using the data of Jolley (1973).
b4.3 yg/1.
°1.2 pg/1.
-------
350
initiate a test. Five concentrations of each test material plus a
control (0.00, 10~2, lO"1, 10°, 101, and 102 mg/1) were used. The
initial water temperature was 21 C. Beakers were kept at 21 + 1 C in
a constant temperature chamber having a 12-hr light - 12-hr dark
cycle during the testing period. Animals were censused daily (Gehrs
and Jolley 1975) for mortality and production of young, with dead
animals immediately removed. Animals were fed 0.1 ml of a trout chow
and water mixture twice weekly (Gehrs 1972). At seven-day intervals,
animals were transferred to fresh solution, with young counted and
removed. An LC^ compilation was used to evaluate mortality effects.
Maturation was determined by counting the number of beakers at each
concentration of each chemical which contained young after the initial
seven-day period.
Fish egg hatching studies. The effects of the chemicals on
hatching success of fish eggs were determined following modified
procedures of Blaylock and Griffith (1971). Gravid carp were collected
and spawned artificially in the laboratory. Eggs were fertilized
immediately in the test solutions as would be expected in the natural
environment so that they were subjected to the toxicant during the
hardening procedure. After 30 min., the test solutions were decanted
and fresh solutions of the appropriate concentration added. Two
replicates of 200-400 eggs were used for each concentration. For the
duration of the test (about 3 days) eggs were held in an environmental
chamber at 26°C + 1°C with a 12-hr light - 12-hr dark photoperiod.
Twice daily during this period solutions were changed and dead eggs
recorded and removed. Six concentrations plus controls were used for
the tests (0.00, 10 , 10~2, lO"1, 10°, 101 and 10 mg/1). Concentrations
were based on serial dilutions prepared in the laboratory rather than
on in vivo measurements. While the latter is preferable, the capabilities
for such measurements have not yet been realized.
-------
351
RESULTS
Zooplankton studies
When mortality was used as the response variable (Fig. 1), the
three test solutions (synthetic mixture, 5-chlorouracil, and 4-
chlororesorcinol) were dissimilar in their level of activity. 5-
chlorouracil caused no change in median survival times (when compared
to controls) at concentration up to 1 mg/1 (three orders of magnitude
above effluent concentrations; 4.3 yg/1, Jolley 1973). Both 4-
chlororesorcinol and the synthetic mixture, however, produced discernible
changes in median survival times over the ranges tested. 4-chlororesorcinol
reduced median survival time approximately 20% (from 30 to 24 days)
_2
at the lowest concentration (10 mg/1) used, with higher concentrations
causing further reductions until the 10 mg/1 level resulted in a
median survival time of slightly more than one day. The synthetic
0
mixture originated a different response with all concentrations j£ 10
mg/1 increasing the median survival time. The 10 mg/1 concentration,
for example, caused an almost doubling of the median survival time as
compared to controls C^58 days in 10 mg/1 concentration, 30 days in
controls). A major decrease in survival time at concentrations _> 10
mg/1 was seen for the synthetic mixture, suggesting a threshold
toxicity level at concentrations somewhere between 10 and 10 mg/1.
Comparison of the responses of this variable (median survival time)
in D. magna to the three test substances reveals an interesting
point. One compound, 5-chlorouracil, caused no apparent effect over
the entire range of concentrations tested. The second, 4-chlororesorcinol,
initiated a decrease in median survival times at even the lowest
_9
concentration (10 mg/l) while the synthetic mixture effected an
increase in median survival time until >_ 10 mg/1 was reached, after
which a negative effect was noted.
While mortality data are useful for comparing relative toxicities
of substances, quite often it is the more subtle effects that might
impinge upon large segments of the population that are more ecologically
important. For zooplankton this includes parameters such as production
-------
352
/\A
o
m
o
<
50
g
20
10
5
2
1
0.5
0.2
0.1
^
o
A
V
AA
' 9
i
c
k
L j
i
k
<
ETIC MIXT
OROURACIL
ORORESOR(
k
(
t
LIRE
I
i
A
-
* * ~-? «.-! .^O ,f\\
10
io
10'
CONCENTRATION (mg/liter)
Figure 1. Days to LC^Q for Daphnia magna in various concentrations
of a synthetic mixture of stable chlorine-containing organics, 5-
chlorouracil, and 4-chlororesorcinol. Data on 5-chlorouracil and 4-
chlororesorcinol from Gehrs and Jolley 1975.
-------
353
of young. In a study on the effects of 5-chlorouracil and 4-chloro-
resorcinol on D. magna using this response variable, all concentrations
—3 1
(10 mg/1 to 10 mg/l) of 4-chlororesorcinol and all but the lowest
_2
concentration (10 mg/1) of 5-chlorouracil caused at least a 50%
decrease in the number of young produced during the first seven days
of free life of the zooplankton. When these data were manipulated to
provide specific birth rates (number of young/adult/day), all but the
_2
10 mg/1 concentration of 5-chlorouracil revealed a statistically
significant (P £ 0.05) decrease (Gehrs and Jolley 1975).
No young were produced in any concentrations (controls to 10
mg/1) of the synthetic mixture during the first seven days of a study
designed to investigate the response of this parameter. Specific
birth rates in this investigation were consequently determined after
a two-week period, and the research was not an exact replication
(experimentally the time frame was different) of that conducted on
the individual compounds. The data from the two studies were therefore
normalized to show percent of specific birth rate of controls (Table
2). Whereas, both 4-chlororesorcinol and 5-chlorouracil caused
decreases in specific birth rates, the synthetic mixture (containing
similar concentrations of each of the two compounds, as well as 15
other compounds) caused specific birth rates greater than controls.
Apparently, some antagonistic interaction took place in the synthetic
effluent, resulting in a specific birth rate response dissimilar to
that produced by either individual compound.
Fish hatching studies
Both 5-chlorouracil and 4-chlororesorcinol have been found to
significantly lower (P <_ 0.05) hatching success of carp eggs at
_0
concentrations as low as 10 mg/1 (approximately that measured
in the effluent) (Gehrs et al. 1974). The synthetic mixture gave
drastically different results (Fig. 2). Not only was there no lowering
_3
of hatching success at levels similar to the 10 mg/1, but there
was no apparent effect over the next three higher orders of magnitude
of concentrations. The breakpoint in toxicity (hatching success
-------
354
o
Table 2. Specific birth rate (number of young/adult/day) for
zooplankton in different concentrations of 5-chlorouracil,
4-chlororesorcinol, and the synthetic mixture
Concentration
(mg/D
io-4
10~3
10~2
10"1
10°
io1
5-chloro-
uracil
1.50
0.81
0.64
0.73
4-chloro-
resorcinol
0.57
0.37
0.36
0.43
0.0
0.0
Synthetic
mixture
1.89
1.55
1.61
0.86
1.30
data have been normalized as percent of specific birth rate of
controls (1.00).
-------
355
100
80
I 60
o
h-
I
h-
LU
O
tr
SL 40
20
0
vr
• s
A 4
0 5
AA
<
j
)
YNTHET
CHLORC
CHLORC
?
k
>
4
•
1C MIXT
)RESORC
URACIL
i
>
)
k
URE
INOL
>
)
i.
t
•
•>
k
(
•3 rt /^ i
0
CONCENTRATION (mg/liter)
Figure 2. Percent hatch of CypvinttS eapp'io in various concentrations
of a synthetic mixture of stable chlorine-containing organics, 5-
chlorouracil, and 4-chlororesorcinol. Data on 5-chlorouracil and
4-chlororesorcinol from Gehrs et al. 1974.
-------
356
< 10%) of the synthetic mixture was approximately 30 mg/1 (geometric
mean). Concentrations this high were not investigated in the 5-
chlorouracil and 4-chlororesorcinol studies and, consequently, no
real comparison of data at this level can be made. If, however, one
extrapolates the higher concentration (10 and 10 mg/1) data from 4-
chlororesorcinol, a response would be seen similar to that generated
by the synthetic mixture.
The percent of hatched embryos that were malformed is an important
parameter that should be included in analyses designed to assess
effects of contaminants on exposed fish populations, because malformed
individuals would not be expected to survive and reproduce. Malformation
augments the negative response of nonhatching by effectively removing
a discrete subset of the hatched embryos from the population. Eyman
et al. (1975) observed a positive correlation between concentration
of 5-chlorouracil and percent malformation in carp with a linear
dose-response relationship originating at 0.5 mg/1. No similar response
was found for either 4-chlororesorcinol or the synthetic mixture.
Unfortunately, at the acute toxicity threshold concentration for the
synthetic mixture, the concentration of 5-chlorouracil was less than
0.5 mg/1. Consequently, nothing can be said concerning possible
interactions of the chemicals in the mixture relative to this parameter.
DISCUSSION
The research results presented above call attention to two types
of problems faced in assessing the potential environmental impact
associated with any group of chemicals, in this case chlorinated
organics. The first problem is the differential response seen both
within a species (when different parameters are used as variables)
and between species. The second problem is that of interpreting and
quantifying the potential toxic interactions between chemicals in a
complex mixture.
The data on zooplankton exemplify the differences in response,
depending on the parameter studied, and show the necessity of utilizing
more than one parameter when attempting to evaluate the potential
-------
357
ecological effects of a chemical on a population. Although no effect
on median survival times results from 5-chlorouracil, decreases in
specific birth rate were observed. If only the former parameter
(median survival times) were used there would be no potential for 5-
chlorouracil adversely effecting a zooplankton population identified.
Zooplankton populations are cropped very heavily, however, with as
much as 20% of the population removed each day (Hall 1964). While
the levels of 5-chlorouracil causing a decrease in specific birth
rate are above the concentrations identified in effluents, and consequently
not expected to adversely effect zooplankton, a decrease in the
production of young could have detrimental effects on zooplankton
populations.
If the purpose of toxicity research is to evaluate the potential
effects on an ecosystem or the environment, then it is necessary to
determine the response of several components of the ecosystem. It is
only necessary to look at the compilation tables of Becker and Thatcher
(1973) to verify this statement. For example, the 96-hr TL is
shown as 3.4 and 90.0 mg/1, respectively, for Lepomis sp. and Physo. sp.
The difficulty involved in evaluating the potential effects of a
mixture of chemicals (i.e., an effluent) is exemplified by the differing
responses observed for the individual compounds as compared to those of
the synthetic mixture. An obvious question is how does one design
research to evaluate the potential effects of mixtures of chemicals
(whether they are chlorinated organics or anything else)? The approach
taken is dependent on the type of information desired. If the concern
is simply the immediate area into which an effluent is released, then
testing the composite mixture by using several different parameters
for several components of the ecosystem (as discussed above) is
sufficient. If the concern is broader, and includes the potential
effect on the environment downstream from the release, then the
necessary research is more difficult to design. The environment
(both abiotic and biotic components) will modify the effluent (chemical
mixture), altering the composition of the materials that will be
confronted at different distances from the release. All compounds
-------
358
will not behave similarly with simple dilution of the effluent being
the only variable necessary to evaluate potential effects. Factors
such as sedimentation, physical degradation, and microbial degradation
are very real activities functioning to alter the effluent. As an
example, we recently found differences in photolysis and microbial
degradation of two of the components of chlorinated effluents, 5-
chlorouracil and 4-chlororesorcinol (Southworth and Gehrs 1976).
Furthermore, if research were conducted solely on the composite
mixture, then effects of individual compounds may be overlooked, as
would be the case for effects of both 5-chlorouracil and 4-chloro-
resorcinol on fish eggs.
While the above statements suggest problems with the total
effluent, evaluation of the toxic effects of each of the individual
compounds also has limited value for predicting potential effects of
complex mixtures on natural systems. This is the result of the
numerous types of interactions that can occur between components of a
complex mixture that substantially alter the relationship between
predicted toxicity (based on data from individual compounds) and
actual response resulting from the mixtures.
Interactions can range from antagonism, where the response to
the mixture is less than would be predicted from the individual
compounds, to synergism, where the response is greater than anticipated
from the toxicity levels of the individual compounds. Only in the
case of an additive interaction, where the observed effect to the
mixture resembles that predicted from the relative toxicities of the
individual compounds, would predictions from individual compounds
approximate the response caused by the mixture. None of these interactions
occurring between compounds would be discovered if only individual
compounds were tested. The number of mixtures of compounds necessary
for evaluation would be extremely large, being given by the general
expression:
,
c = n!
n r (n-r).'r!
-------
359
where
nCr = total number of possible interactions,
n = number of compounds, and
r = number of compounds to be examined for interactions in a
specific subset.
Even in a simple situation of five individual compounds, the number
of possible interactions for evaluation would be thirty-one. Such an
approach would be both extremely time consuming and expensive.
The approach we believe appropriate combines both of the previously
mentioned systems (composite mixtures and individual compounds). It
is designed to determine the potential for interactions between
classes or groups of compounds in the mixture (effluent) rather than
attempt to look at all possible interactions initially (which we have
shown above to be an astronomical undertaking). In this system,
representative compounds from the various classes present in the
mixture would be used to provide an early estimate of potential
interaction. Selection of the model compounds for use can be based
on several criteria (e.g., amount of data currently existing, concentration
in effluent, structure, etc.). For the stable chlorine-containing
organics, initial evaluation would be concerned with three classes of
compounds (phenols, pyrimidines and purines, and aromatic acids)
rather than the seventeen (and there are certainly more) individual
compounds identified. Depending on data derived from these investigations,
evaluation of other inter- and intra-class interactions would be
undertaken. For example, if a pyrimidine and phenol showed no interaction
while a pyrimidine and aromatic acid produced an additive response,
we would next pursue the interactions that might arise from the
pyrimidine and aromatic acid rather than the pyrimidine and phenol.
We do not mean to imply that this approach should be taken with
the exclusion of investigating either composite effluents or individual
compounds. On the contrary, all three approaches should be used.
What we suggest, however, is a method that will aid in providing the
most timely data (including future research needs) for the effort and
money expended.
-------
360
To conclude* what is necessary for evaluating the environmental
impact of water chlorination is a systematic approach. This paper
has been designed to address only the effects segment of this research.
To attain a full understanding of, and consequently the capability of
predicting effects of, chlorination requires research on chlorine
kinetics, transformation, metabolism, and bioconcentration as well.
ACKNOWLEDGMENTS
Research sponsored by Energy Research and Development Administration
under contract with Union Carbide Corporation. Publication No. 864,
Environmental Sciences Division, Oak Ridge National Laboratory.
REFERENCES
Becker, C. D., and T. 0. Thatcher. 1973. Toxicity of power plant
chemicals to aquatic life. WASH-1249. Battelle Pacific-Northwest
Laboratories. Richland, Washington. June.
Blaylock, B. G., and N. A. Griffith. 1971. A laboratory technique
for spawning carp. Prog. Fish. Cult. 33: 48-50.
Draley, J. E. 1972. The treatment of cooling water with chlorine.
ANL/ES-12. Argonne National Lab. Chicago, Illinois.
Eyman, L. D., C. W. Gehrs, and J. J. Beauchamp. 1975. Sublethal effects
of 5-chlorouracil on carp (Cyprinus carpio) larvae. J. Fish. Res.
Bd. Can. 32: 2227-2229.
Gehrs, C. W. 1972. Aspects of the population dynamics of the calanoid
copepod, Diaptomus alavipes. Ph.D. Thesis, Univ. of Oklahoma,
Norman.
Gehrs, C. W., L. D. Eyman, R. L. Jolley, and J. E. Thompson. 1974.
Effects of stable chlorine-containing organics on aquatic
environments. Nature 249: 675-676.
Gehrs, C. W., and R. L. Jolley. 1975. Chlorine-containing stable organics:
new compounds of environmental concern. Verh. Intemat. Verein.
Limnol. 19: 2185-2188.
Hall, D. J. 1964. An experimental approach to the dynamics of a
natural population of Daphnia galeata mendotae. Ecology 45:
94-112.
-------
361
Jolley, R. L. 1973. Chlorination effects on organic constituents in
effluents from domestic sanitary sewage treatment plants. ORNL/TM-
4290. Oak Ridge National Laboratory. Oak Ridge, Tennessee.
Jolley, R. L. 1975. Chlorine-containing organic constituents in
chlorinated effluents from sewage treatment plants. J. Water Pollut.
Control Fed. 47: 601-618.
Laubusch, E. J. 1962. In J. S. Sconce (ed.). Chlorine, its manufacture
and use. Am. Chem. Soc. Monogr. Series No. 154. Reinhold
Publishing Corp., New York.
Palin, A. T. 1950. A study of chloro derivatives of ammonia and
related compounds, with special reference to their formation in the
Chlorination of natural and polluted waters. Water Water Eng. 54:
248-256.
DISCUSSION
Mathilde J. Kland, Lawrence Berkeley Laboratory. What was the
dark reaction control in the 5-chlorouracil reaction?
Gehrs. I am going to call on George Southworth to answer that
question.
George Southworth, Oak Ridge National Laboratory. The controls
were maintained under the same conditions, that is, temperature and pH,
but with no light.
Kland. Did you identify the product or products of the light reaction?
Southworth. No. We don't know specifically, but it is degraded to
the point where it has little uv absorption.
Kland. Under the influence of uv radiation, uracil and methyluracil
both add water across the double bond.
Southworth. We have postulated that it is a hydrolysis mechanism
as has been demonstrated for halouracils by Garrett and co-workers at
the University of Florida. The reaction involves dehalogenation and loss
of chromophoric properties.
Kland. Did you examine the behavior of the solution obtained after
light exposure in the presence of acid and base? The hydrate undergoes
ring opening to the corresponding aldehyde, dehydration, or reconstitution,
for example to the original uracil, depending on pH.
Southworth. Afterwards? Yes. You are talking about reversibility
as in the hydration of uracil. There is no evidence of rhat in our
experiments.
-------
362
Greg L. Seegert, University of Wisconsin at Milwaukee. Your graphs
on the effects of 5-chlorouracil and 4-chlororesorcinol on the hatching
success of carp eggs indicated a significant depression in hatching
success over a range of concentrations from 0.001 to 10 mg/liter.
However, this effect was not seen at comparable levels when these chemicals
were tested in combination with a series of chloro-organic compounds.
Do you have an explanation for this?
Gehrs. I am glad you raised the point. You are fortifying the
concern we are raising with respect to mixtures and complex effluents.
We can't postulate what the interactive activity might be in the
synthetic effluent. I believe our results to be quite valuable, however,
in showing the complexity of the problem. While we can look at form,
or structure, and at least calculate or predict effects from these data,
what happens when we move into the natural system and find all of the
materials together?
-------
SESSION IV. MODELING AND PREDICTION
Carl W. Gehrs, Session Chairman
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
The first two days of our discussions have been aimed at presenting
information raising questions with respect to what types of research
are still needed. After our initial planning for this conference, we
came to the realization that one of the things that we did not have
included was a time when individuals, who are involved in assimilating
data and coming up with generalizations from these data, could present
the information. And that is the purpose of this section this morning.
It's to allow individuals who have taken a look at the data which are
available on chlorine to present to us their concepts and their ideas
with respect to the question, "What can we do with the data so that
we can use it?"
363
-------
MODELING RESIDUAL CHLORINE LEVELS:
CLOSED-CYCLE COOLING SYSTEMS
Guy R. Nelson
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
A mathematical model which predicts residual chlorine levels in
cooling tower blowdown streams at any time during the chlorination cycle
is discussed in this paper. To quantify the absence or presence of
residual chlorine in the blowdown, the model interprets residual chlorine
as negative chlorine demand.
The general model has eight variations applying to specific chlo-
rination program characteristics. The program characteristics affecting
the general model are:
a. Split stream vs no split stream chlorination (the fraction
of the recirculating water chlorinated).
b. Residual data feedback vs no residual data feedback (the
type of chlorine feed equipment used).
c. Positive vs negative demand at the end of the chlorine
feed period (the time length of the chlorine feed period).
The variations to the model are useful not only in predicting residual
chlorine levels in the blowdown, but also in making alterations in
existing chlorination programs which minimize chlorine waste, provide
more disinfecting efficiency, and reduce residual chlorine levels in
the blowdown.
INTRODUCTION
Many cooling tower systems use chlorine or hypochlorites to control
bacteria and slime growth in the condenser tubes carrying cooling water
in the plant process. The biological growth, if left uncontrolled,
365
-------
366
causes excessive tube blockages, poor heat transfer, and accelerated
system corrosion—all of which reduce plant efficiency. For any cooling
tower system the length of time of the chlorine feed period and the
number of chlorine feed periods per day, week, or month change as the
biological growth problem changes. In most cooling tower systems, the
chlorine is added at or near the condenser inlet in enough quantity to
produce a free residual chlorine level of 0.1 to 0.6 mg/1 in the water
leaving the condenser. The amount of chlorine added to maintain the
free residual chlorine depends upon the amount of chlorine demand agents
and ammonia in the water.
Chlorine and ammonia react to form chloramines. These chloramines
constitute the combined residual chlorine of the water. This combined
residual chlorine is less efficient and slower in providing biological
control than free residual chlorine (AIChE 1972, Draley 1972a, White
1972, Puckorius 1975).
Although chlorination is effective for slime control in the con-
denser tubes of cooling tower systems, its application may result in
residual chlorine in the blowdown discharged to the receiving water.
The effects of residual chlorine on aquatic life are of great concern
(Hamilton et al. 1970, Brooks and Baker 1972, Draley 1972a, Brungs 1973).
Data on the toxicity of residual chlorine to aquatic organisms is
available. W. A. Brungs1 publication, "Effects of Residual Chlorine
on Aquatic Life: Literature Review," recommends criteria for maximum
residual chlorine concentrations in receiving waters (Brungs 1973). For
the intermittent presence of residual chlorine, not to exceed two hours
per day, the criteria indicate a ma-x-timnn tolerable concentration of
0.2 mg/1. For the continuous presence of residual chlorine, the maximum
tolerable concentration is 0.01 mg/1. These concentrations would not
protect trout, salmon, and some important food organisms and are poten-
tially lethal to sensitive life stages of sensitive fish species. Brungs
recommends a lower criteria to protect trout and salmon. These lower
criteria are 0.04 mg/1 for the intermittent presence (2 hr/day) of
residual chlorine and 0.002 mg/1 for the continuous presence of residual
chlorine.
-------
367
The potential effects of residual chlorine on aquatic organisms
require the measurement, prediction, and control of residual chlorine
in effluent discharges to the aquatic environment. The purpose of this
paper is to discuss a model which predicts levels of residual chlorine
in the recirculation systems and blowdown streams of cooling towers.
The model can be used to improve chlorination programs resulting in
less chlorine waste, more disinfecting efficiency and reduced environ-
mental impact. Although the paper is directed toward the application of
chlorine in condenser cooling systems of thermal electric power plants,
the model presented can be used in other industrial chlorination programs
where the conditions are similar.
MODEL ANALYSIS
In an Environmental Protection Series report, U.S. Environmental
Protection Agency (EPA), I develop and analyze a model which calculates
the residual chlorine level in cooling tower blowdown (Nelson 1973).
The vocabulary, concepts, and notations that apply to the model are
defined in the Appendix. The mathematical expression of the model itself
is contained in the appendix also. To illustrate the model's utility,
I apply it to studies of three (3) hypothetical chlorination programs
on two cooling tower systems. One system is a mechanical draft cooling
tower; the other is a natural draft cooling tower.
Six cooling tower characteristics affect the model's results in both
cooling tower systems :
(a) The cooling system volume (V) .
(b) The recirculation rate (QR) .
(c) The initial chlorine demand in the blowdown (CBO)«
(d) The flashing rate of residual chlorine in the atmosphere (F) .
(e) The initial residual chlorine level in the condenser effluent
(f) The blowdown flow rate (Qfi).
These characteristics manifest themselves in the predictive model through
the terms discussed below.
-------
368
The V/QR term (expressed in minutes) is the ratio of the water
volume of the entire cooling system to the recirculating water flow
rate. This term indicates the length of time for a given water parcel
to make one pass through the system. With conventional system design,
this value is 10 to 15 minutes for mechanical draft towers and 20 to 22
minutes for natural draft towers.
The dimensionless term Q-/Q-, is the ratio of the blowdown water
D K
flow rate to the recirculating water flow rate. Its value, along with
the V/Q term, expresses the length of time that a conservative chemical
remains in the system.
The dimensionless term C^^/C,,- is the ratio of the initial residual
KO i>O
chlorine level in the recirculating water (CD_) to the initial chlorine
KU
demand in the blowdown (C^.).
o(J
The dimensionless term F is the fraction of the residual chlorine
which flashes or decomposes as it passes through the cooling tower fill.
The theoretical range of the value of F can be 0.0 to 1.0. Probably F
is related in some yet unquantified way to tower inlet temperature and
to cooling range and/or the water to air ratio; the chemical form of the
chlorine residual may also be a factor. The concept of F is documented
(Draley 1972a, 1972b; White 1972). In the draft Environmental Impact
Statement for the Davis Besse steam electric generating plant, Draley
suggests an F value of 0.5 for combined residual chlorine in the
natural draft tower system (Draley 1972b).
Table 1 lists the values of cooling tower and chlorination program
characteristics found in typical cooling tower systems. The table also
lists the values which were applied to my hypothetical cooling tower
systems. Table 2 and Figs. 1, 2, and 3 summarize the results of the
model's application on the cooling systems.
Program study one
In this first study, the chlorination program characteristics do
not include residual data feedback or split stream chlorination. The
expression for TSA of models NNN and MNP is the following (see Appendix);
-------
369
Table 1. Cooling tower and chlorination program characteristics
(Nelson 1973)
Characteristic
Typical values
Minimum Maximum
Program study values
A. Chlorination program:
Chlorination cycle
Chlorine feed period
Split stream C12
Residual feedback
B. Cooling tower:
8 hr 7 days
10 min 30 min
Optional
Optional
Mechanical draft
Natural draft
24 hr
15 min
Program study 3
Program study 2
QB/QR
F
CRO
CBO
CRO/CBO
V/QRa
T/V
0.008
0.3
-0.1
0.5
-0.1
10
20
0.015
0.6
-1.0
3.0
-1.2
15
22
0.01
0.4
0.4
0.667
-0.6
10
20
Table 2. Program study results
Program study
Time chlorine
in blowdown (min)
Maximum concentration
(mg/1)
Number 1
a. Natural draft
b. Mechanical draft
Number 2
Number 3
0
83
15
0
0
0.349
0.033
0
-------
370
1.0 H
RATIO
-1.0-
NATURAL DRAFT TOWER^
TIME (MIN.)
MECHANICAL DRAFT TOWER
CHLORINATION CYCLE
Fig. 1. Program study one.
1.01
RATIO
-1.0
\TFEEDBACK
FEEDBACK
CHLORINATION CYCLE
Fig. 2. Program study two.
TIME
1.0n
RATIO
0
-1.0-
SPLIT STREAM
VNO SPLIT STREAM TIME
CHLORINATION CYCLE
Fig. 3. Program study three.
-------
371
RATIO = X1(l-e~Ylt) + e~Ylt .
In the natural draft program, RATIO remains positive throughout the
chlorine feed period. This indicates that there is no residual chlorine
in the blowdown. In the mechanical draft program, RATIO becomes nega-
tive during TSA. The boundary condition, t = 15 min, determines its
value at the end of TSA. The first appearance of residual chlorine in
the blowdown occurs when there is zero chlorine demand in the sump, or
RATIO = 0. The time during the chlorine feed period when this occurs
is 8.75 min. Therefore, residual chlorine is present in the blowdown
for 6.25 min during TSA.
A residual first appears at 8.75 min after the start of the chlorine
feed period. This residual increases in value until the end of the
feed period, at which time it is at its maximum value. Since the model
defines residual chlorine as negative chlorine demand, RATIO is at its
minimum value at this time.
The TSB expression does not apply to the natural draft program,
because its terminal demand value is not negative. The TSB expression
for the mechanical draft program is the following:
RATIO = X2(l-e~Y2t) + (CB/CBO)B e"Y2t .
The value of RATIO at the start of TSB is -0.523. RATIO increases in
value through TSB until it reaches the value of zero. The boundary
condition, RATIO = 0, quantifies the time of TSB. For this case it is
76 min. Figure 1 illustrates the value of RATIO through the complete
chlorination cycle for both programs under program study one.
To minimize repetition, the other two program studies concentrate
on alterations to the mechanical draft program. The concepts in program
studies two and three are applicable to natural draft programs.
Program study two
This program study analyzes the effect of residual feedback on the
value of RATIO for the mechanical cooling tower draft. Model NRN applies
-------
372
in this case—because the value of RATIO is negative at the end of TSA.
The comparison of the two studies shows a marked reduction in the con-
centration of residual chlorine in the blowdown of the cooling tower in
study two. The use of residual data feedback in the mechanical draft
program not only reduces the concentration of residual chlorine in the
blowdown, but also reduces the length of time during the chlorination
cycle in which the blowdown contains residual chlorine.
Program study three
In this program study, the chlorination program for the mechanical
draft tower includes split stream chlorination and excludes residual
data feedback. In this case, Model SNP applies—because the value of
RATIO at the end of TSA is positive.
There are two substeps to TSA in model SNP. The first substep
applies until there is residual chlorine in the recirculating water after
the split stream is remixed with the remaining streams. The flashing
term (F) does not appear in the substep 1 expression because the chlo-
rine demand is positive (no residual chlorine). Negative demand appears
in the recirculating water when RATIO = S(l - C^./C^,.). Once the
1U D(J
condition is reached, then the substep 2 expression applies for the
rest of TSA.
The positive demand value at the end of TSA indicates that there
is no residual chlorine in the blowdown. Therefore, TSB does not apply.
RATIO regenerates back to the value of positive unity before the start
of the next chlorine feed period.
This study shows the value of split stream chlorination in mini-
mizing or eliminating residual chlorine in the blowdown.
CONCLUSIONS
The model discussed in this paper predicts the value of RATIO
during the chlorination cycle of a cooling tower system. RATIO is
defined as the ratio of the blowdown chlorine demand value at any time
during the chlorination cycle to its value at the start of the chlorina-
tion cycle. A positive value of RATIO indicates that there is no
-------
373
residual chlorine present in the blowdown. A negative value of RATIO
indicates that residual chlorine is in the blowdown. RATIO'S rate of
decrease and increase during the chlorination cycle, and its minimum
value at the end of the chlorine feed period, are all a function of the
cooling tower and chlorination program characteristics of specific
systems.
The analysis of the models shows potential methods of reducing or
eliminating residual chlorine levels in the blowdown by taking advantage
of some of the optional program characteristics. These potential
methods include one or more of the following alternatives:
(a) Installing residual data feedback equipment into the
chlorine feed system.
(b) Practicing split stream chlorination.
(c) Reducing the chlorine feed period, if possible.
(d) Reducing the initial residual chlorine level in the
condenser effluent.
(e) Increasing the water volume of the cooling tower. This
alternative may not apply to existing cooling towers
because it involves the system design. The alternative
can apply to systems on the engineering drawing boards.
This alternative may have other advantages—such as an
extra supply of water for fire protection.
(f) Cutting off the blowdown when residual chlorine appears
in the sump. The blowdown flow can resume after the
residual is dissipated by the flashing effect and the
makeup water chlorine demand. The length of time
during which the blowdown can be eliminated is a func-
tion of the system's upper limit on dissolved solids.
(g) Mixing the blowdown with another stream which has a
high chlorine demand.
In cases where limited field analyses have been performed trends
are identified which conform to model predictions (Baker 1973, Draley
1973). For field verification it is necessary to use the amperometric
procedure to measure low levels of residual chlorine (Brungs 1973).
-------
374
The Standard Methods amperometric procedure (American Public Health
Association 1965) for determining residual chlorine in aqueous solutions
is not applicable to all cooling tower waters. Some cooling tower
waters contain copper, turbidity, natural buffering, and water treatment
chemicals. These constituents may produce interferences in the analyt-
ical procedure which result in erroneous residual chlorine readings.
A report by Manabe (1972) which describes a modification of the procedure
is available. The modification increases the efficiency of both the
sampling and the titrating procedures.
Even without substantial field verification, the models have utility,
They can be used to modify existing and proposed chlorination programs
in cooling tower systems. The modifications can provide increased
chlorination efficiency and reduced residual chlorine levels in the
blowdown.
REFERENCES
American Institute of Chemical Engineers (AIChE). 1972. Cooling towers,
p. 79-80.
American Public Health Association. 1965. Standard Methods for the
examination of water and wastewater, 12th Ed. New York. P. 103.
Baker, R. J. 1973. Fate and disposal characteristics of residual
chlorine discharged to receiving streams. Wallace and Tiernan.
Brook, A. J., and A. L. Baker. 1972. Chlorination at power plants:
impact on phytoplankton productivity. Science 176: 1414-5.
Brungs, W. 1973. Effects of residual chlorine on aquatic life:
literature review. J. Water Pollut. Control Fed. 45: 2180-2193.
Draley, J. E. 1972a. The treatment of cooling waters with chlorine.
ANL/ES-12. Argonne National Laboratories, Argonne, 111. 11 P.
Draley, J. E. 1972b. Davis Besse Nuclear Power Station draft environ-
mental impact statement, Appendix B. U.S. Atomic Energy
Commission. November. (Final statement issued March 1973.)
Draley, J. E. 1973. C12 experiments at the John E. Amos Plant.
ANL/ES-23. Argonne National Laboratory, Argonne, 111.
Hamilton, D. H., Jr., D. A. Flemer, C. W. Keefe, and J. A. Mihursky.
1970. Effects of chlorination on an estuarine primary production.
Science 169: 197-8.
-------
375
Manabe, R. 1972. Measuring residual chlorine levels in cooling water/
amperometric method. EPA 660/2-73-039. U.S. Environmental
Protection Agency.
Nelson, G. R. 1973. Predicting and controlling residual chlorine in
cooling tower blowdown. EPA-R2-73-273. U.S. Environmental Pro-
tion Agency.
Puckorius, P. 1975. Personal communication. Zimmite Corp., West
Lake, Ohio.
White, G. C. 1972. Handbook of chlorination. Van Nostrand Reinhold
Company, New York.
APPENDIX A
ALTERNATE MODEL EXPRESSIONS FOR SPECIFIC CHLORINATION PROGRAMS
The models for the specific chlorination programs are given in
Tables 3a and 3b.
Table 3a. Programs without split
stream chlorination
Equations
Model
TSA TSB
NNN 1A 2A
NNP 1A
NRN 5A 2A
NRP 5A
TSC
4A
3A
4A
3A
Table 3b . Programs with split
stream chlorination
Equations
TSA TSB
Model
Sub 1 Sub 2
SNN 6A 7A 2A
SNP 6A 7A
SRN 8A 9A 2A
SRP 8A 9A
TSC
4A
3A
4A
3A
-------
376
Equations referred to in Tables 3a and 3b
(1A) C
LBO
where
and
R
/CB \
IF— K
V BO/1
(2A) CR -Y t f-R I -Y f
O ir /l *2tx I I " 1 ~IOt
— = X2(l-e ^ ) +1^—1 ° ^
CBO IC_|B
QR
where Xo = —
0 + F
V_ f
^
and l-z=-L = RATIO'S value at the start of TSB
-------
377
(.JAJ ^^ f"o % _v t-
,) e Y3t
^-= i-a-(^-V
CBO \CBO/C
QB
where Y3 = — ,
and [-—L = RATIO'S value at the start of TSC
"B0>
**********
(4A) C
r
LBO
QB
where ¥ = —
**********
(5A)
4- e
CBO
Q
R BO
where X5 = - - - ,
QR
and Y5 = v
Q
R
**********
-------
378
(6A) CR
^
CBO
QB
C_, = Chlorine demand in chlorinated condenser
effluent at the start of ISA,
and S = Chlorinated fraction of total recirculating
water flow.
**********
(7A) CR
?^-
LBO
where
C = Chlorine demand in chlorinated condenser
effluent at the start of ISA,
and S = Chlorinated fraction of total recirculating
water flow.
-------
379
(8A) C
^-=X8(l-
UBO
where X8 =
V S
Q C
'.-
R
^TO ~ Chlorine demand in chlorinated condenser
effluent at the start of TSA,
and S = Chlorinated fraction of total recirculating
water flow.
**********
^ * —15 \r *_ fnif^'m.n.i -tr .._
-Y9t
CBO
fe)
where X9
-^ + S+F-FS
— 4- S + F - FS
-9=^ E
\
C = Chlorine demand in chlorinated condenser
effluent at the start of TSA,
and S = Chlorinated fraction of total recirculating
water flow.
-------
380
APPENDIX B
GLOSSARY
Vocabulary
Free residual chlorine is that portion of the total residual chlo-
rine which will react chemically and biologically as hypochlorus acid
or hypochlorite ion.
Combined residual chlorine is that portion of the total residual
chlorine which will react chemically and biologically as chloramines.
Total residual chlorine is the sum of the free and combined resid-
uals. Unless otherwise specified, throughout this paper residual
chlorine refers to the total residual chlorine.
Chlorine demand is the amount of chlorine (mg/1) required to be
added to a water (sample) before any stable residual chlorine is formed.
Organics and reducing agents in the water cause this demand. These
materials have varying reaction rates with chlorine. The reaction rates
cause the chlorine demand value to be time dependent. For the purpose
of this paper, the chlorine demand is that demand which reacts with
chlorine within five minutes of exposure.
The chlorination program describes the manner in which chlorine is
fed and controlled in the cooling tower system.
The chlorination cycle is the length of time between the start of
two sequential chlorine feed periods.
Split stream chlorination is an alternate method of chlorine addi-
tion. It is the practice of splitting the total recirculation flow
through the condenser into a number of separate streams. One of these
streams is chlorinated at a time. The chlorinated stream is then mixed
with the remaining streams. The presence or absence of split stream
chlorination is a chlorination program characteristic.
Residual feedback describes a function performed by chlorine feed
equipment. If the control system in the equipment is capable of adjusting
the flow of chlorine to produce a constant residual in the recirculation
water out of the condenser, the system has residual feedback. The
presence or absence of residual feedback is a chlorination program
characteris tic.
-------
381
Concepts
Residual chlorine vs chlorine demand. The model expresses the
change in the chlorine demand of the blowdown during the chlorination
cycle. To quantify the absence or presence of residual chlorine in
the blowdown, the model interprets residual chlorine as negative chlorine
demand. By conceptual definition, residual chlorine is not present in
the blowdown unless the chlorine demand is satisfied.
C^. In the model development, the term CD represents the chlorine
Ji a
demand in the blowdown at any time during the chlorination cycle.
. The term (]„_ represents the chlorine demand in the blowdown
,,-.
BO
„_
BO
at the beginning of the chlorination cycle.
RATIO. The term RATIO represents the ratio of the chlorine demand
in the blowdown at any time during the chlorination cycle to its initial
value at the beginning of the cycle (i.e., RATIO = CB/CBO) •
Time steps. In order to predict the chlorination cycle, the model
breaks down the cycle into time steps. Figure 4 illustrates the time
step concept.
1.0H
RATIO
-1.0-
CHLORINATION CYCLE
TIME (MIN.)
TSC
Fig. 4. Time steps of the chlorination cycle
-------
382
Time Step A (TSA). TSA is the length of time during which chlorine
is added to the system (the chlorine feed period).
Time Step B (TSB). TSB is the length of time after TSA in which
the value of RATIO is negative.
Time Step C (TSC). TSC is the length of time after TSA of a chlo-
rination cycle in which the value of RATIO is zero or positive. TSC ends
at the start of the next chlorination cycle.
Although equations can be (and have been) written for TSC the
numerical output is rather spurious and of no practical value in either
environmental protection or chlorination program design. This is true
primarly because the length of TSC is dictated by the biocidal require-
ments of the cooling system rather than any level or function of chlorine
demand or residual obtainable from the equations. Also, extraneous
factors such as dust washout or changes in makeup water characteristics
are not accounted for. Finally, optimal use of the models for environ-
mental protection and chlorine conservation require chlorine demand data
at start of the cycle (C ).
o(j
Terminal Demand Value (TDV). TDV is the value of RATIO at the end
of TSA. It is the lowest value of RATIO during the chlorination cycle.
Positive vs negative TDV. TDV can be either positive or negative
at the end of TSA. If TDV is negative as shown in Fig. 4, then the
cycle contains three time steps (TSA, TSB, and TSC). If TDV is positive
or zero, then the cycle contains only two time steps (TSA and TSC). The
positive vs negative TDV is a chlorination program characteristic.
Model notation
There are eight specific models based upon the general model dis-
cussed in this paper. Each specific model applies to a set of chlorina-
tion program characteristics. Table 4 is a matrix which defines the
shorthand notation used to describe each model. For easy reference, the
first letter in the shorthand notation refers to sidestream filtration
or no sidestream filtration (S or N). The second letter refers to
residual feedback (R or N). The third letter refers to positive TDV
or negative TDV (P or N).
-------
383
Table 4. Model notation
Residual feedback
No residual feedback
(-) TDV
(+) TDV
(-) TDV
(+) TDV
Split stream
chlorination
No split stream
chlorination
SRN
NRN
SRP
NRP
SNN
NNN
SNP
NNP
For example, model NNN applies to a chlorination program which
(a) does not have split stream chlorination, (b) does not have residual
feedback, and (c) has a negative chlorine demand at the end of the chlo-
rine feed period. Model SRP applies to a program which (a) has split
stream chlorination, (b) has residual feedback, and (c) has a positive
chlorine demand at the end of the chlorine feed period.
Cooling system notation
Figure 5 is a schematic flow diagram of a typical cooling tower
system;
QBCB
Fig. 5. Cooling tower system.
-------
384
Where:
Qfi - Blowdown flow rate in m3/min (gpm);
Q.J. = Water flow rate into the condenser in m3/min (gpm);
QL = Chlorine flow rate into the condenser in m3/min (gpm);
QR = Water flow rate returning to tower from the condenser in
m3/min (gpm);
Q£ = Water evaporation rate leaving tower stack in m3/min (gpm);
Qs = Water flow rate to the sump in m3/min (gpm);
QM = Makeup water flow rate into the tower sump in m3/min (gpm);
V = Cooling tower system volume in m3 (gallons);
Cfi = Chlorine demand in the blowdown, (mg/1);
C = Chlorine demand in the blowdown at the start of the
chlorine feed period (mg/1);
C = Chlorine demand in the chlorine feed stream (mg/1);
C_ = Chlorine demand in water returning to the tower from
the condenser (mg/1);
C = Chlorine demand in the water returning to the tower
from the condenser at the start of the chlorine feed
period (mg/1);
CE = Chlorine demand in the evaporated water leaving the
tower (mg/1);
C = Chlorine demand in the recirculating water entering
O
the sump (mg/1);
CL. = Chlorine demand in the makeup water entering the sump.
DISCUSSION
John E. Butts, Envirosphere Company. You mentioned that cooling
tower blowdown could be held up to allow residual chlorine decay. In
your experience what are the time ranges required to minimize the dis-
charge of residual chlorine and what would be the approximate chlorine
concentrations ?
Nelson. It is highly specific to the type of cooling system that
you are operating, and the type of chlorination program you are applying
to the system. I would say, in general, about two hours. And, in that
time, you would probably notice about a 5% increase in total solids.
-------
385
George C. White. Consulting Engineer. Plants are getting so large
some blowdown lines are as large as 72 inches in diameter. How is it
practical to shut down such a large pipe two or three times every
24 hours?
Nelson. In other words, you can't shut it off. You bring out an
interesting point. For some of these already on-board systems, you can
not apply all the options 1 have listed. Hopefully, we can consider
these kinds of options on the drawing board, so that we can try to solve
some of the problems.
Robert J. Baker, Wallace and Tiernan Division of Pennwalt Corporation,
An additional method of reducing or eliminating residual chlorine in
tower blowdown is to chemically dechlorinate.
Nelson. Very good, Thank you, Bob. That should be added to the
methods. An interesting thing about the model is that you can use the
model to determine how much feed rate of sulfite or bisulfite to add to
reduce the residual chlorine to the desired level.
William A. Brungs, U.S. Environmental Protection Agency. Cooling
tower blowdown contains a much higher concentration of dissolved solids
due to evaporation and other dissolved materials introduced intentionally.
Did the presence of these materials affect the accuracy of the ampero-
metric method?
Nelson. Yes. We have found that copper and iron in the cuprous
and ferrous state did interfere with amperometric measurements in the
field to the point where we were actually measuring a chlorine residual
when none was present. We have since then solved the problem and can now
get accurate amperometric measurements by adding a chelating agent to
the water sample before we titrate. The reference is Manabe (1972). He
describes not only the procedure for eliminating interferences, but also
a good method for getting field data on cooling system blowdown. I
don't know how many of you have tried but it is really an experience to
get out in the field. At a desk we can draw all these curves that we
want, but out in the field, cold and wet, with cooling tower drift all
over you, it is different. Manabe takes the practical experience that he
has gained and has developed a good procedure for field verification.
Thomas A. Miskimen, American Electric Power Service. The following
represents a written comment submitted after the preceding discussion:
The management of chlorine residual in cooling tower blowdown should
not be isolated from the management of all other wastewater discharges
from a power plant. For coal burning plants, there are advantages to
using tower blowdown to transport fly ash. For any plant, the TSS of
tower blowdown may prove unacceptable to the state water pollution con-
trol agency unless (a) either makeup water is treated for TSS removal
or (b) blowdown is treated for TSS removal, such as being used for ash
transport. The second choice is widely used.
-------
A KINETIC MODEL FOR PREDICTING THE COMPOSITION OF CHLORINATED
WATER DISCHARGED FROM POWER PLANT COOLING SYSTEMS
Milton H. Lietzke
Chemistry Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
We are in the process of developing a kinetic model for predicting
the composition of chlorinated water discharged from power plant
cooling systems. As a start this model will contain three rate equations:
the reaction of hypochlorous acid with ammonia, the reaction of hypo-
chlorous acid with an organic amine, and the further reaction of hypo-
chlorous acid with monochloramine. The simultaneous differential
equations will be solved numerically to give the composition of the water
as a function of time. Other rate equations will later be added to the
model to account for other reactions that are known to take place.
Eventually the model will be incorporated into a large unified transport
program.
Kinetic and thermodynamic data for the most important chemical
reactions that may occur during the chlorination of cooling waters ar->
presented.
Natural waters used for cooling purposes at power stations vary
widely in composition. Depending on the source, such waters contain both
organic and inorganic impurities. In addition to bacteria, algae, spores,
and viruses there will usually be traces of organic amines, other
organic compounds, ammonia (or ammonium ion), traces of heavy metals, and
various anions. The analyses of two typical natural waters are shown in
Table 1.
387
-------
388
Table 1. Analyses of two typical natural waters.
Sample date
pH
Chloride, mg/1
Organic carbon, mg/1
Inorganic carbon, mg/1
Organic nitrogen, mg/1
Ammonia (as N) , mg/1
Watts Bar Lake
July 1973
7.5
0.5
2.7
0.04
2.7
0.5
Mississippi River
July 1974
7.3
7.7
9.0
16.3
<0.05
0.15
In order to prevent slime formation in the cooling towers chlorine
is commonly added to the water. The chlorine is rapidly hydrolyzed to
yield equimolar quantities of hypochlorous acid and hydrochloric acid.
The hypochlorous acid dissociates into hydrogen ions and hypochlorite
ions, the extent of dissociation being a function of both pH and
temperature. Both the hypochlorous acid and the hypochlorite ion are
powerful chlorinating agents. They will react rapidly, for example, with
ammonia or ammonium ions to produce chloramines and with organic amines
to produce N-chlorinated amines. The rates of these reactions are also
functions of both pH and temperature. The treated cooling water
containing the chloramines and unreacted chlorine is eventually returned
to the biosphere.
Although the toxic nature of chloramines has long been recognized,
it has only been within the last few years that national attention has
been focused on the problem. For example, a massive fish-kill in the
cooling waters of a large generating station on the shore of Saginaw Bay
has been attributed to a lethal concentration of residual chlorine in the
water. Whether the chloramines or other chlorinated compounds will have
a long-term toxic effect on man is at present unknown. In any event it
is now necessary to take cognizance of this problem in assessing the
impact of power generating stations on the environment.
-------
389
In developing a model for predicting the composition of chlorinated
water discharged from power plant cooling systems it is first necessary
to have kinetic and thermodynamic data on the various chemical reactions
that may occur. Unfortunately these data are scattered in the literature
and in many cases difficult to find. Hence a literature search has been
made and the available data on the most important reactions to be
expected have been compiled and are presented in the appendix of this
paper. In each case reference is made to the source of the data.
Examination of the appendix reveals that in many cases two rate
constants are given for a particular reaction: a theoretical rate
constant and an observed rate constant. The theoretical rate constant
refers to the chemical reaction exactly as written. In practical
calculations the theoretical rate constant is inconvenient to use since
the reacting species may themselves participate in equilibria which are
functions of pH. These equilibria are taken into account in the case of
the observed rate constants. For example, consider the reaction between
ammonia and hypochlorous acid (equation 5, appendix). When the observed
rate constant for this reaction is used it is only necessary to specify
the total concentration of nitrogen as ammonia and ammonium ion and the
total concentration of chlorine as hypochlorous acid and hypochlorite
ion. Similar considerations apply to many of the other chemical reactions
tabulated.
In order to predict the composition of chlorinated natural water
returned to the environment from a cooling tower we are developing a
kinetic model under the unified transport approach. At present the model
considers three chemical reactions: the reaction of hypochlorous acid
with ammonia, the reaction of hypochlorous acid with a.composite organic
amine, and the further reaction of hypochlorous acid with the
monochloramine formed in the first reaction. These reactions are the
following:
NH + HOC1 N NH Cl + H20,
RNH2 + HOC1 * RNHC1 + H20,
NH Cl + HOC1 ——^ NHC12 + H20.
-------
390
The three differential equations representing the rates of these
reactions in terms of the appropriate observed rate constants and the
concentrations of the several species involved are solved simultaneously
using the Runge-Kutta technique. Additional chemical reactions will be
added to the model as seem warranted.
In addition to the kinetic model an equilibrium model has also been
developed for the same chemical reactions. Since under some conditions
of pH and temperature prevailing in natural waters the chlorination
reactions are extremely rapid (a few seconds to a few minutes for 99%
reaction), it is only necessary in these cases to perform an equilibrium
calculation to predict the composition of the chlorinated water returned
to the environment.
At the present time both of the programs are designed for calculations
at a fixed temperature only. In the immediate future temperature
dependent expressions for the various rate and equilibrium constants will
be incorporated into the models.
ACKNOWLEDGEMENT
The author would like to express his sincere thanks to Dr. R. M. Rush
for presenting this paper at the Conference while the author was out of
the country.
APPENDIX
Thermodynamic and kinetic data on the following reactions are
summarized:
1. NH0 + H.O ^ NH.+ + OH~
3 2 ^ 4
2. NH4+ (H20 * NH3 + H+
3. C12 + H20 ^ HOC1 + H+ + Cl~
4. HOC1 * H+ + OC1~
5. NH + HOC1 N NH Cl + HO
6. CH-NH. + HOC1 N CH.NHCl + H00
3 2 N: 3 2
-------
391
7.
OH
8.
OH
9.
OH
10. NH2C1 + HOC1
11. 2NH Cl + H
12.
+
NHC12 +
NH + NHC1
NHC1
13. NH2C1 +
14.
NH + HOC1
HOC1
15. CH3NHC1 + HOC1
16. 2CH NHC1
CH3NC12 + H2°
CH.NC1 + CH NH
j 2. j 2.
17. CH CONHCH,. + HOC1
J J
18. CH0CONHCH0COO~ + HOC1
J 2.
CH0CONC1CH. + H00
j j Z
CH0CONC1CH0COO
j 2
19. NH CONH + HOC1
^ NH CONHC1 + HO
20. Rate constants for N-chlorination of a number of additional
compounds relative to ammonia.
1.
NH + H 0
3 2
Kl -
t°c
0
10
15
* NH + OH
+ _
[NH ] [OH ]
4
[NH ]
K. x 105 t°C K- x 105
1 1
1.51 25 1.81
1.62 40 2.00
1.70 50 1.95
Reference: H. Lunden, J. chim. phys. 5, 574 (1907).
-------
392
2. NH4+ (H20) * NH3 + H+
K2 =
t°C K2 x 1010
0 (0.83)
5 1.25
10 1.86
15 2.73
20 3.98
25 5.68
Reference: R. G. Bates and G. D. Pinching, J. Res. Nat. Bur. Stand.
42^ 419 (1949).
3. Cl + HO ^ HOC1 + H+ + Cl~
t°c
30
35
40
45
50
K2 x 1010
8.06
11.3
15.7
21.4
28.9
[HOC1]
K3 =
[ci2]
t°c
0
15
25
35
45
K x 104
1.46
2.81
3.94
5.10
6.05
Reference: R. E. Connick and Yuan - tsan Chia, J. Am. Chem. Soc.
81, 1280 (1959).
Rate constants:
k_ = 5.60 sec ±0.45 if the reaction is monomolecular, and
-------
393
3. (cont'd)
k =0.1 I/mole - sec if the reaction is bimolecular (the
reaction mechanism has not been resolved).
Reference: A. Lifshitz and B. Perlmutter - Hayman, J. phys,
chem. 64, 1663 (1960).
4. HOC1
| H+ + OC1
[HOC1]
t°C
0
5
10
15
20
25
K4X
2
2
2
3
3
3
io8
.0
.3
.6
.0
.3
.7
Reference: G. M. Fair, J. C. Morris, S. L. Chang, I. Weil, and
R. P. Burden, J, Am. Water Works Assoc. 40, 1051 (1948)
5. NH, + HOC1 > NH-C1 + HO
J \ z z
[NH2C11
5 [NH ] [HOC1]
K = 3.6 x IO9 at 25°C
Reference: J. E. Draley, ANL/ES - 12 (1972), Argonne National
Laboratory, Argonne, 1L.
-------
394
5, (cont'd)
Rate constant; Over the temperature range 5 - 35°C the
activation energy for the reaction is 3 kcal.
The theoretical rate constant is given by
k5 = 9.7 x 108 exp (-3000/RT) I/mole - sec.
Reference: J. C. Morris in Principles and Applications of Water
Chemistry, S. D. Faust and J. V. Hunter (ed.),
John Wiley and Sons, Inc., New York (1967), p. 27.
k (theoretical) is related to k (obs.) by
k_ (theoret)
k (obs) =
K K K
K [H ] K
w J w
where K = K ; 1C = K ; and K is given by K = [H+] [OH~]
for the dissociation of water: HO ^ H + OH~.
Reference: I. Weil and J. C. Morris, J. Am. Chem. Soc. 71, 1664
(1949).
The temperature dependance of K (at saturation vapor pressure)
w
is given by
P P
log K = -~ + P0lnT + P T + -- + P.,
w 1 i J 2 5
where P, = 3.12860 x 10A P. = -2.17087 x 106
1 4
P2 = 94.9734 P = -6.06522 x 102
P3 = -0.097611 T = T°K
Reference: F. H. Sweeton, R. E. Mesmer, and C. F. Baes, Jr., J.
Solution Chem. 3, 191 (1974).
-------
395
6. CH3NH + HOC1 N CH3NHC1 + HO
Rate constant; over the temperature range 5 - 25°C the
activation energy for the reaction is 1.9 kcal.
The theoretical rate constant is given by
k = 7.8 x 109 exp (-1900/RT) I/mole - sec.
D
Reference: J. C. Morris (1967), loc. cit., p. 32.
k, (theoretical) is related to k (obs.) by
b b
K (theoret)
k, (obs) =
6 v ' K 1C K
a b . 3
1 Kw
where K = K. ; K, = K.,: and K is rep. as in 5 (above)
a 4 b / w
Reference: I. Weil and J. C. Morris, loc. cit.
mo+] [OH~]
t°c
0
10
20
30
40
50
-log K?
3.449
3.405
3.380
3.367
3.375
3.386
Reference: D. H. Everett and W. F. K. Wynne - Jones, Proc. Roy.
Soc (London), 177A, 499 (1941).
-------
396
8. (CH3)2NH + H20 ——^ (CH0)0NH0+ + OH"
[(CH ) NH ] [OH ]
•tr ___ ~J £* £•
K8 " ——— -
[(CH3)2NH]
t°C -log Kg
0 3.392
10 3.307
20 3.245
30 3.203
40 3.183
50 3.175
Reference: D. H. Everett and W. F. K. Wynne - Jones, loc. cit,
9. (CH3)3N + H20 - - *• (CH3)3NH+ + OH
[(CH ) NH+] [OH"]
t°C -log Kg
0 4.591
10 4.407
20 4.261
30 4.141
40 4.059
50 3.992
Reference: D. H. Everett and W. F. K. Wynne - Jones, loc. cit,
-------
397
10. NH0C1 + HOC1 N NHC10 + H00
2
-------
398
12. 2 NH Cl > NHC1 + NH
The activation energy for this reaction in the temperature range
7 - 49°C is 4.3 kcal.
Rate constant:
k12 = 80 exp (-4300/RT) I/mole - sec.
Reference: M. L. Granstram, PhD Thesis, Harvard University, (1954)
13. NH.C1 + H.O * HOC1 + NH
L t.
[HOC1]
K13= -
[NH2C1]
K = 9.0 x 10~2 exp (-14,000/RT) moles/1.
Rate constant; as the first order process (in the absence of
competing NH.) the specific rate is given by
k = 8.7 x 107 exp (-17,000/RT) sec'1.
Reference: M. L. Granstrom, loc. cit.
14. (CH0)-NH + HOC1 ^ (CH.)0NC1 + H.O
J Z. \^ J i. £-
Rate constant: over the temperature range 5 - 38°C (no change
with temperature) the theoretical rate constant is given by
o
kn. = 3.3 x 10 I/mole - sec.
14
Reference: J. C. Morris (1967), loc. cit., p. 34.
k , (theoretical) is related to k , (obs) by
-------
399
14. (cont'd)
k
14 (theoret)
ki4 (obs) = TK^
K [H ] K
w w
where K = K.; K, = K0; and K is rep. as in 5 (above)
a 4 b o w
Reference: I. Weil and J. C. Morris, loc. cit.
15. CH NHC1 + HOC1 ^ CH NC12 + H20
2
Rate constant: k = 1.1 x 10 I/mole - sec at 25°C,
The temperature coefficient has not been determined.
Reference: J. C. Morris (1967), loc. cit., p. 32.
16. 2CH3NHC1 ; *• CH3
_2
Rate constant: k, , = 4.7 x 10 1/raole - sec at 25°C.
ID
The temperature coefficient has not been determined.
Reference: J. C. Morris (1967), loc. cit.
17. CH3CONHCH3 + HOC1
Rate constant:
k = 1.4 x 10~3 I/mole - sec at 25°C.
Reference: J. C. Morris (1967), loc. cit.
-------
400
18. CH3CONHCH2COO + HOC1 * CH3 CONC1CH2COO +
Rate constant:
k. = 5 x 10 I/mole - sec at 25°C.
J.O
Reference: J. C. Morris (1967), loc. cit.
19. NH- CONH. + HOC1 ^ NH-CONHC1 + H.O
2. L ^~ 2. L
Rate constant:
k - 7.5 x 10~2 I/mole - sec at 25°C.
Reference: J. C. Morris (1967). loc. cit.
20. Summary of rate constants for N - chlorination of a number of
nitrogenous compounds relative to NH,. F0 is the ratio of the rate
constant for the particular reaction relative to that for the
reaction with NH .
Table 2. Rate constants for N-chlorination relative to NH3
Nitrogenous
compound
Methylamine
Dime thy lamine
Diethylamine
Morpholine
Diethanolamine
E thy 1 amino ace tate
Glycine
Alanine
Leucine
(-) Alanine
Serine
Glycylglycylgylcine
Ch lor amide
N- chlorine thy lamine
PKo(25°C)
3.376
3.226
3.067
5.30
5.12
6.27
4.221
4.133
4.256
3.765
4.795
6.09
ca. 15
ca. 13.8
F =k /k
our
«£
54°
23
9
9.3
2.0
22
19
14
33
6.7
2.3
5.5 x 10-5
1.8 x lO"1*
log Fo
1.78
1.73
1.36
0.95
0.97
0.30
1.34
1.28
1.15
1.52
0.83
0.36
-4.26
-3.74
?Based on reaction between HOC1 and basic form of N compound.
Computer from individual rate constants.
Reference: J. C. Morris (1967), loc. cit., p. 37.
-------
401
DISCUSSION
Arthur S. Brooks, University of Wisconsin at Milwaukee. Have you
used light as a factor in your model?
Richard M. Rush, Oak Ridge National Laboratory. I do not think it
is planned although I gather from what I have heard in the last day or
two that it might be useful to do so.
Brooks. Our work in Lake Michigan indicates that light is a very
important factor in determining the rate of chlorine dissipation in
natural waters. Perhaps a light extinction coefficient could be incor-
porated in the model to account for various light levels in an effluent
plume.
Rush. Very good.
J. Donald Johnson, University of North Carolina at Chapel Hill.
Have you found any of the methylamine compounds in cooling tower systems?
If so, how much of these compounds did you find?
Rush. We are not doing experimental work for this program. We
will simply utilize kinetic and thermodynamic data on reactions for
which that information is available. Obviously, we are looking at the
first two very simple reactions. We do not have any results from this
initial computation. We just got the program operational about the time
that Dr. Lietzke left for Europe and so things are in a bit of a stand-
still right at the moment.
Robert L. Jolley, Oak Ridge National Laboratory. The type of pro-
gram we are developing at this time is really a first approximation, and
it will be more fully developed to represent the real situation as new
program components are added. Because the kinetics of methylamine and
HOC1 reactions are known it was assumed, in order to start developing a
working model, that the organic nitrogen which we have measured in cool-
ing water samples was indeed methylamine. The program will be refined
later.
-------
ASSESSING TOXIC EFFECTS OF CHLORINATED EFFLUENTS ON AQUATIC
ORGANISMS: A PREDICTIVE TOOL
Jack S. Mattice
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
ABSTRACT
Increased chlorination of surface waters has necessitated development
of an analytical tool for assessing the toxicity of a chlorinated effluent
so that environmental damage may be minimized. The tool proposed includes
length of exposure as well as concentration as factors of importance and
allows for protective limitations to be applied to releases on a site spe-
cific basis. Cornerstones of the assessment tool are acute and chronic
mortality thresholds. The thresholds are derived by (1) summarizing ex-
tant chlorine toxicity data (mostly as median tolerance limits) in log
concentration-log duration plots, (2) bounding these data points from be-
low with straight lines to estimated acute and chronic median lethal thres-
holds, and (3) shifting the acute "median lethal threshold" to estimate a
true mortality threshold (zero mortality) using an empirically derived
relationship between exposure time necessary to yield fifty and zero per-
cent mortality. Dose-time exposures for organisms entrained through the
plant and into the discharge plume are compared with these thresholds to
derive yes or no decisions regarding mortality. Limits may then be set
to minimize environmental harm. Accurate plume dilution data are necessary
for optimal application of this procedure, but parameters which increase
discharge dilution rate are taken into consideration.
INTRODUCTION
Prediction of the magnitude of environmental perturbation resulting
from release of chlorinated surface waters depends upon generic information
dealing with (1) fate of the chlorine as regards both chemical form and
-------
404
concentration, (2) effect of these resultant forms on the individual organ-
isms exposed, and (3) translation of the effects on individuals to those
on populations, communities and the ecosystem. Reaction products derived
from chlorinated natural waters are determined by initial chlorine concen-
tration, local water quality (especially relative concentrations of poten-
tial reactants), time of reaction and such physical conditions as pH,
temperature, etc. Toxicity of chlorine compounds varies over a wide range;
for example, some of the chloroorganic compounds are effective toxicants
at parts per billion (ppb) concentrations (Allen et al. 1946 and 1948;
Parker 1935; Hopkins and Bean 1966; Gehrs et al. 1974). In some instances,
even the less chemically complex chloramines differ in toxicity from free
chlorine (Butterfield 1948; Douderoff and Katz 1950; Moore 1951; Merkens
1958; Rosenberger 1971; Holland et al. 1960). In addition, toxicity has
been found to vary with temperature (Scheuring and Stetter 1950; Dressel
1971; Boss et al. 1974; Stober and Hanson 1974; Gregg 1974), heavy metal
exposure (Hoss et al. 1974; Wolf et al. 1975), species of organism, con-
centration, and time of exposure, even for the same form of chlorine.
Preliminary evidence also indicates that intermittent exposure to chlorine
may be more efficacious than continuous exposure at the same mean concen-
tration. Problems encountered in extrapolating individual effects to re-
sponses at higher levels of organization are as complex as those involved
at the other generic information categories. The above complexity is con-
founded by the limited data base available from which to make valid gener-
alizations for predictive purposes.
Despite the problems involved in predicting the impacts of chlorina-
tion, pragmatic considerations require the development of this capability.
Water use by the power industry has been rapidly increasing (Krenkel, un-
dated; Science and Public Policy Program, Univ. of Oklahoma 1975; Federal
Power Commission 1971). Since chlorine is the biocide specified for most
power plants (Motley and Hoppe 1970; Lee and Stratton 1972; Becker and
Thatcher 1973), use of chlorine has paralleled this rapid increase. Ex-
pansion of other industries which discharge chlorinated effluents has also
increased the potential for environmental damage. The potential problems
-------
405
revolving around use of chlorine would appear to indicate an obvious con-
comitant solution—elimination of chlorine from the list of acceptable
biocides—except for the many advantages over currently available alter-
natives (Baker 1959). The only rational policy would be to maintain func-
tional capability of chlorine within the facility (e.g., power plant) while
controlling discharge levels to eliminate (or limit to acceptable levels)
environmental effects. Reasonable limits on effluent releases must be
based on the ability to predict the impact of various discharge levels.
In this paper, I will present in summary form, a method or tool [pre-
sented in full in Mattice and Zittel (1976) and summarized for another
purpose in Mattice (1976)] which can be used to predict the mortality re-
sulting from a specified chlorination schedule. This tool, which is aimed
at filling the second of the generic information gaps described above, is
based on comparison of chlorine dose-times (1) within an electric gener-
ating facility and (2) within its discharge area, with toxicity thresholds
derived from the limited data base available. The tool is applicable in
both marine and freshwater environs, but this presentation is confined to
freshwater analyses.
Development of this tool was stimulated by the need for the Nuclear
Regulatory Commission (formerly the U.S. Atomic Energy Commission) to
respond to the requirements of the National Environmental Policy Act (NEPA
1969). Early development of this tool by the Impact Assessments staff at
Oak Ridge National Laboratory proceeded through the work of Coutant and
coworkers (U.S.A.E.G. 1972; Coutant 1974; Goodyear et al. 1974). Subse-
quent modifications have been primarily due to the work of Mattice
(U.S.A.E.C. 1973; Goodyear and Mattice 1973; U.S.A.E.G. 1974; Mattice and
Zittel 1976; Mattice 1976), although other members of the staff have also
contributed significantly.
MORTALITY THRESHOLDS
In order to evaluate the effect of a proposed chlorination, it is
necessary to know the vulnerability of fresh water organisms to chlorine
exposures. Existing data regarding chlorine toxicity to freshwater or-
ganisms are presented as a plot of log chlorine concentration versus log
-------
406
exposure time (Fig. 1), a form suggested by Coutant (1974). Concentrations
are presented as total residual chlorine without regard to the type (free
or combined) which predominated during the experiment. This appears rea-
sonable as a first approximation (Mattice and Zittel 1975). Data shown
were derived using appropriate biological methodology and either accurate
or conservative (i.e., chemical analytical procedures which would tend to
underestimate the concentrations of chlorine tolerated by the experimental
organisms) chemical methods. Experimental conditions other than chlorine
concentration were largely ignored in compiling and interpreting the data.
Further information concerning species involved, investigators, and meth-
odology may be found in Mattice and Zittel (1976). A bibliography of
abstracts of the papers from which data have been obtained is also avail-
able (Mattice and Pfuderer 1976). Most of the data points (Fig. 1) are
dose-times yielding median (50%) mortality. The data shown represent 64
species, of which 8 are algae, 27 are invertebrates, 28 are fish, and 1
is an amphibian.
Environmentally protective mortality thresholds were based on the
most sensitive organisms for which data are available. This type of
treatment is appropriate in cases in which it is necessary to make deci-
sions based on limited data such as is true for chlorine toxicity. The
form of the threshold lines (acute and chronic) was based on general
toxicological principles. At concentrations of toxicant above some low
value .(the chronic threshold), increase in exposure time increases mor-
tality at a given concentration. Below the chronic threshold, exposure
time has no effect on mortality. On a plot of log concentration versus
log exposure time the chronic threshold line would thus be parallel to
the abscissa, and the acute threshold line would have a negative slope
and end at the chronic threshold line. Preliminary threshold lines were
drawn by enclosing the median mortality data with two straight lines —
the chronic and acute median mortality thresholds. Once these straight
lines were drawn, it was necessary to shift the preliminary acute mor-
tality threshold to estimate the true threshold (zero mortality). Mattice
and Zittel (1976) found for fourteen species of organisms that for a given
concentration of chlorine the maximum exposure time which would not result
-------
407
10
5
2
sITRATION (ppm)
0 0
no m —
UJ
^ 0.1
0
o
uj 0.05
z
3 0.02
X
0
0.01
0.005
0.002
0.001
1 1 1 1 1 1 1
"22 2
H-2
*v
>v •
I \>
—
—
—
* I I I I I I II
•
• •
*. * i
.
,
U^N/
*^
1 1 1 ' ' II
•
•6
..•" 2'3
"•2 . .
1
i
-2
•
•
>^ •
^N^
1 M 1 1 Ml
".2
•2
,2
1
•
—
—
—
==
—
—
—
CHRONIC TOXICITY THRESHOLD
O1 2 5 102 2 5 103 2 5 104 2 5 105 \
DURATION OF EXPOSURE (min)
Figure 1. A summary of data on the toxicity of chlorine to aquatic
life. Most points are for dose-times yielding median mortality. Acute
and chronic toxicity thresholds (estimates of the 07, effect levels) are
also shown (modified from Mattice and Zittel 1976).
-------
408
in any mortality was 37% of that which would cause death of 50% of the
exposed organisms. The acute mortality threshold line shown in Fig. 1
is the result of shifting the preliminary line to the left to exposure
times only 37% as long. The threshold lines drawn indicate that for
concentrations greater than 0.0015 mg/1, an increase in time of exposure
generally increases the toxicity of chlorine. For concentrations less
than 0.0015 mg/1, however, toxicity is no longer time-dependent and ex-
posure for infinite time would not result in death. This concentration
was thus designated the chronic mortality threshold.
Evaluation of a given exposure to chlorine can be made by comparing
the dose-time combination with the mortality thresholds (Fig. 1). If the
dose-time combination is above the chronic mortality threshold and to the
right of the acute mortality threshold, organisms exposed are killed; if
not, they are not affected at all (i.e., there are no sublethal effects).
The latter results from the fact that a preponderance of the data involved
mortality. Based on this procedure, exposure to 0.1 mg/1 total residual
chlorine for ten minutes would not be lethal, while exposure to the same
concentration for about 30 minutes would be lethal. In some cases (for
example, if an area were populated entirely by highly tolerant organisms),
this method could result in prediction of mortality when none would actually
occur. However, because chlorine mortality thresholds are known for so few
species, this type of conservative approach is necessary to ensure environ-
mental protection. On the other hand, the procedure may underestimate
sublethal effects. Chance of this latter, however, is reduced by inclusion
of some sublethal effects as data points (see Mattice and Zittel 1976).
Setting the acute and chronic toxicity thresholds as shown above (Fig.
1) satisfies the criteria most important to preoperational evaluation of
a chlorination. Any method for analysis must be (1) predictive, which im-
plies a basis in principles of general applicability, and (2) reliable,
which implies a certain degree of conservatism. Although based on data
from a limited number of species, the relationships between time and con-
centration and chlorine toxicity are similar to those found for other
toxicants (Warren 1971), indicating that the form used here is probably
realistic. The predictions of mortality are conservative because the
-------
409
thresholds are set based on the most sensitive species at levels expected
to be below lethal levels. Demonstration that a relationship exists be-
tween general species groups and relative toxicity, such as that postulated
by Brungs (1973) for freshwater fish, may allow less conservative thresh-
olds for some environments.
DEMONSTRATION OF THE ANALYTICAL TOOL
Three hypothetical examples will be used to demonstrate the procedures
involved in using the proposed analytical tool as well as to indicate the
efficient way in which the important time variable is treated (Fig. 2). In
each of the cases, plant specifications regarding chlorination are the
same: chlorination proceeds at a rate sufficient to yield 2 mg/1 initially;
residence time in the facility following chlorination is 20 minutes; and
discharge concentration is 1 mg/1 total residual chlorine. Dilution rates
following discharge differ in each case, with dilution being most rapid in
Case III and least rapid in Case I. A given concentration isopleth (e.g.,
0.62) is shown at the same spatial location for each plume to indicate the
equal volumes of water needed for the dilution, but the times differ in
each case. None of the examples necessarily represent actual cases, but
were chosen for purposes of demonstration.
Toxicity of both plant and plume entrainment may be assessed using
plots of weighted concentration versus cumulative time since discharge.
In each of the cases given, organisms entrained through the facility would
be killed because the dose-time combination (See Plant Entrainment, Fig.
3) lies to the right of the acute toxicity threshold and above the chronic
toxicity threshold. Therefore, in this case, because further analysis need
only concern organisms not passing through the plant, the times shown (Fig.
3) refer to the interval since discharge. Such would also be the case for
many industrial or waste treatment discharges. An organism entrained into
a certain area of the discharge plume during dilution is exposed to a con-
centration bounded by those of the two surrounding isopleths for a time
bounded by the total time for water movement between the isopleths. In
order to determine mortality, it is necessary to approximate both the
dose-times of exposure over the periods between each concentration
-------
410
CASE
90 170 200 240
0.62 XX38X0.26X0.13 ^0.09 ^-0.04 ^0.03^0.02
CONCENTRATION ISOPLETHS (mg/liter)
\
\
\
\
\
\
CASE ^
in
DISCHARGE PLUME
Figure 2. Comparison of the chlorine concentration in the discharge
plumes of three hypothetical power plants showing three different rates
of dilution. Concentration isopleths are identical for all the plumes,
but the times to reach a given isopleth differ. Chlorine concentration at
the point of discharge was given as 1 mg/1.
-------
2.0
1.0
—
b.
0)
0.5
0.2
0.1
0.05
0.02
0.01
en
E
o:
h-
2
UJ
o
/CASE H
I ' ^
* PLANT ENTRAINMENT
ACUTE TOXICITY
THRESHOLD-^
\
s
•CASE I
\
10 20
50 100 200 500 1000
TIME (min)
Figure 3. Chlorine dose-time combinations for organisms entrained
into each of the plumes (Cases I, II and III) (Fig. 2) prior to the 0.62
mg/1 isopleth shown in relation to the acute toxicity threshold (Fig. 1).
Time shown is that from point of discharge. The dose-time for organisms
entrained through the plant is also shown.
-------
412
isopleth and the total dose-time of exposure in the plume. The concen-
tration equal to the mean of two consecutive isopleths and applied over
the total time between them was chosen as a reasonable estimate of the
true exposure for that interval. Total exposure concentration for or-
ganisms entering the plume in a given area was estimated by computing
the weighted average concentration for the total exposure period. Weighting
was based on the exposure time at a given concentration relative to the
total exposure duration. This method of computation assumes that chlorine
doses are additive. Effects of entrainment in the plume differ for each
of the "Cases." This is indicated clearly by the curves relating weighted
concentration and cumulative time since discharge (Fig. 3). For compara-
tive purposes, only the concentration path for the organisms entrained
into the plume prior to the first isopleth is shown for each case. In
reality, a series of such plots would need to be made for organisms en-
trained into each area of the plume for each "Case." Mortality would
result for organisms in Cases I and II, but not for Case III, because in
the former cases one or more of the dose-times is enclosed by the acute
and chronic toxicity thresholds; in the latter case none of the points is
to the right of the acute threshold and above the chronic threshold. Be-
cause organisms entrained into the initial sector of the plume receive
maximum exposures, no plume mortality would result at all in Case III, and
it has not been considered further.
Further analyses of Cases I and II indicate that a larger area of the
plume is impacted when dilution rate is less rapid (Tables 1 and 2). In
order to determine if organisms entrained into an area of the plume will
be killed, the dose-time of exposure must be estimated. For example, or-
ganisms entrained prior to the sixth isopleth (0.04 mg/1, Fig. 2 and Table
1) will be exposed to 0.065 mg/1 for 144 min in Case I or to the same con-
centration for 30 min in Case II. Comparison of these dose-times with the
acute toxicity threshold (Fig. 1) indicates that mortality will result from
the former but not the latter case. In a full analysis, the full series
of weighted mean dose-times might need to be compared with the thresholds.
If any dose-time combination is above the thresholds, death is predicted
for any organisms initially entrained in the area under consideration. By
-------
Table 1. Concentration and time of exposure for organisms entrained into that area of the plume enclosed by a given Isopleth and continuing In the
plume during subsequent dilution for Case I. By following down a given column it is possible to observe the cumulative time course of
exposure of organisms entrained into various areas of the plume as they reach each subsequent isopleth. Times shown are cumulative using
discharge as time zero and concentrations are weighted according to percent of the total time since discharge. Lethal dose times for
each case are enclosed.
Concentration
Isopleth
Dose-Time Experienced for Each Initial Isopleth Encountered Upon Entrainment'
0.62
0.38
0.26
0.13
0.09
0.04
0.03
0.02
0.01
Dose Time
Dose Time
Dose Time
Dose Time
Dose Time
Dose Time
Dose Time
Dose Time
Dose Time
0.62
0.38
0.26
0.13
0.09
0.04
0.03
0.02
0.01
0.81 4
0.60 12
0.45 26
0.30 62
0.24 90
0.16 170
0.14 200
0.12 240
0.09 320
_.
0.50 8
0.39 22
0.23 58
0.19 86
0.13 166
0.11 196
0.10 236
0.08 316
--
--
0.32
0.23
0.19
0.13
0.11
0.10
0.07
--
—
14
50
78
158
188
228
308
._
—
..
0.20 36
0.16 64
0.11 144
0.09 174
0.08 214
0.06 294
—
—
—
—
0.11
0.08
0.07
0.06
0.04
—
--
—
--
28
108
138
178
256
.-
--
—
--
—
0.07 80
o.oe no
0.05 150
0,04 230
--
—
—
—
—
.
0.04
0.03
0.02
.-
--
--
--
--
--
30
70
150
—
..
„
..
—
..
..
0.03 40
0.02 UO 0.02 80
Organisms entrained into the area between isopleths 0.62 and 0.38 are listed in column 0.38. Refer to Figure 2.
-------
Table 2. Concentration and time of exposure for organisms entrained into that area of the plume enclosed by a giver Isopleth and continuing in the
plume during subsequent dilution for Case II. By following down a given column It Is possible to observe the cumulative time course of
exposure of organisms entrained Into various areas of the plume as they reach each subsequent Isopleth. Times shown are cumulative using
discharge as time zero and concentrations are weighted according to percent of the total time since discharge. Lethal dose times for
each case are enclosed.
Concentration
Isopleth
0.62
0.38
0.26
0.13
0.09
0.04
0.03
0.02
0.01
0.62
Dose Time
0.81 1
0.56 5
0.44 10
0.32 20
0.25 30
0.20 40
0.17 50
0.15 60
0.08 120
Dose-Time
0.38
Dose Time
..
0.50 4
0.40 9
0.29 19
0.23 29
0.19 39
0.16 49
0.13 59
0.07 119
Experienced
0.26
Dose Time
--
-.
0.32 5
0.24 15
0.19 25
0.15 35
0.13 45
0.11 55
0.06 115
for Each
0
Dose
*.
—
..
0.20
0.15
0.12
0.10
0.09
0.05
Initial Isopleth
.13 0.
Time Dose
—
—
10
20
30
40
50
110
—
..
--
o.n
0.09
0.07
o.ne
0.03
Encountered Upon
09
Time
--
--
--
10
20
30
40
100
Entrapment
0,04
Dose
--
--
--
--
0.07
0.05
0.04
0.02
Time
--
--
--
—
10
20
30
90
0.03
Dose Time
_-
—
__
_.
—
0.04 10
0.03 20
0.02 80
0.02 0.01
Dose Time Dose Time
-.
—
-.
--
_.
-.
0.03 10
0,02 70 0.02 70
Organisms entrained Into the area between isopleths 0.62 and 0.38 are listed in column 0.38. Refer to Figure 2.
-------
415
following a horizontal row, it is possible to find the area of entry into
the plume which does not yield any lethal dose-times. The areas enclosed
in the boxes (Tables 1 and 2) define the plume areas in which organisms
would be killed for Cases I and II. The mean initial concentrations en-
countered by organisms entrained near the periphery of these areas are 0.07
for Case I and 0.20 for Case II. Thus, organisms in about 14 and 5 times
the volume of the plant flow would be killed in Cases I and II, respec-
tively, because dose-time combinations of organisms exposed in those vol-
umes would be enclosed by the threshold lines (Fig. 1). Of course, the
ecosystem impact on the areas involved would have to be assessed on the
basis of the relative importance of that volume and of the organisms in
it to that of the rest of the ecosystem which is not directly impinged
upon by the chlorine-caused mortality. Assessment at this level would
involve population or ecosystem dynamics including compensatory mechanisms
and population interactions. Pertinent information at this level is
sparse and limits real statements regarding impacts of any chlorine-caused
mortality.
DISCUSSION
The proposed analytical tool for evaluating impacts associated with
chlorination has a number of advantages: (1) application is relatively
simple, (2) bias incorporated into the procedure favors environmental pro-
tection, (3) site-specific information regarding such factors as hydrology
and facility design is incorporated, and (4) flexibility exists, allowing
inclusion of multi-plant considerations and modification following appro-
priation of new data. The flexibility of this analytical tool is one of
its greatest assets, because incorporation of topical information will
probably not result in major modification of the tool itself, i.e., the
concepts and procedures involved in the analysis will remain constant.
This is important since major efforts are being made in the areas of mod-
eling the fate of chlorine following discharge, of determination of the
toxicity of free and combined chlorine to aquatic organisms, and of popu-
lation modeling under conditions of increased mortality. Other advantages
are discussed further in Mattice and Zittel (1976).
-------
416
As the tool evolves, the general trend of analyses will be from the
more to the less conservative. It is obvious from the preceding dis-
cussion that the proposed method of estimating mortality (and chlorine
limits based on them) is conservative at each step: setting the thresh-
olds, estimating plume chlorine concentrations, and estimating the toxi-
city of each dose-time combination. Unless compensatory population
mechanisms are taken into account, it is also conservative in extrapolating
the resultant mortality to population effects. If environmental protection
is the goal, each of these steps must be treated conservatively because of
deficiencies in knowledge. As the information base grows in both depth
and breadth, many of the conservative assumptions can be replaced by en-
vironmental data.
As with all summarizations of data, disadvantages of this method are
based in the generalizations or assumptions made to facilitate collection.
The most important of these is the assumption of equal toxicity of the
reaction products (other than chlorine demand) resulting from chlorination.
Some of the chloroorganics such as the pyrimidine analogue 5-chlorouracil
(Gehrs et al. 1974) may by their function in translation of the genetic
material have relatively large effects at low concentrations. On the other
hand, toxicity levels found by Zillich (1969), Basch et al. (1971), and
Arthur et al. (1974) using chlorinated sewage effluents were in the same
general range as those found following chlorinatio.n of less organically
enriched waters (Mattice and Zittel 1976). Most of the other assumptions
of this method would be found in any method of toxicity assessment.
Although this method has advantages over other possible analytical
methodologies, as discussed here and by Mattice and Zittel (1976), perhaps
its greatest strength lies in its potential for use in setting realistic
limits on discharges of chlorine. Toxicity of chlorine is influenced
most strongly by concentration and time of exposure (Coutant 1974; Goodyear
et al. 1974). The importance of the time factor has played a role in
development of complex discharge structures designed to dilute the effluent
as rapidly as possible with the receiving water, but the time factor has
largely been ignored in limits proposed by previous authors (Brungs 1973;
Collins and Deaner 1973; Baker and Cole 1974; Dawson 1974; EPA 1974;
-------
417
Gentile et al. 1974). The examples in Fig. 2 and Table 1 indicate that,
in using the proposed tool, time of exposure is a critical factor in de-
termining toxicity of chlorine above the chronic mortality threshold.
Limits may thus be adjusted to include consideration of site-specific
dilution patterns and rates following release from the facility. These
limits would thus have a realistic basis in the data available.
The tool proposed is not intended to represent the ultimate procedure
for analyzing for impact of chlorine or for setting safe limits on releases.
The information required for this does not exist. However, in being rela-
tively conservative with respect to environmental protection and in having
the broadest data base available, the tool probably represents the optimum
given present knowledge and does allow interim decisions to be made regard-
ing proposed chlorine utilization.
ACKNOWLEDGEMENTS
I wish to thank Dr. Charles C. Coutant for presenting this paper to
the Conference.
This research was supported by the Energy Research and Development
Administration under contract with the Union Carbide Corporation. Pub-
lication No. 845, Environmental Sciences Division, ORNL.
-------
418
REFERENCES
Allen, L. A., N. Blezard, and A. B. Wheatland. 1946. Toxicity to fish
of chlorinated sewage effluents. Surveyor (England) 105: 298.
Allen, L. A., N. Blezard, and A. B. Wheatland. 1948. Formation of cyan-
ogen chloride during chlorination of certain liquids: toxicity of
such liquids to fish. J. Hyg. 46: 184-195.
Arthur, J. W., R. W. Andrew, V. R. Mattson, D. T. Olson, G. E. Glass,
B. J. Halligan, C. T. Walbridge. 1974. Comparative toxicity of
sewage-effluent disinfection to freshwater aquatic life. National
Water Quality Laboratory, U.S. Environmental Protection Agency,
Duluth, Minn. 62 p. (Unpublished manuscript).
Baker, R. J. 1959. Types and significance of chlorine residuals. J. Am.
Water Works Assoc. 51: 1185-1190.
Baker, R., and S. Cole. 1974. Residual chlorine: something new to worry
about. Ind. Water Eng. (March/April) 10-18.
Basch, R. E., M. B. Newton, J. G. Truchan, and C. M. Fetterolf. 1971.
Chlorinated municipal waste toxicities to rainbow trout and fathead
minnows. Water Pollution Control Research Series No. 18050 GZZ 10/71,
49 pp.
Becker, C. D., and T. 0. Thatcher. 1973. Toxicity of power plant chemi-
cals to aquatic life. WASH-1249, UC-11. Battelle Pacific Northwest
Laboratories, Richland, Washington.
Brungs, W. S. 1973. Effects of residual chlorine on aquatic life. J.
Water Pollut. Control Fed. 45: 2180-2193.
Butterfield, C. T. 1948. Bactericidal properties of free and combined
available chlorine. J. Am. Water Works Assoc. 40: 1305-1312.
Collins, H. F., and D. G. Deaner. 1973. Sewage chlorination versus
toxicity—a dilema? J. Environ! Eng. Div. ASCE, 99(EE):761-772, Proc.
Paper No. 19190.
Coutant, C. C. 1974. Evaluation of entrainment effect, p. 1-11. In
L. D. Jensen (ed.). Proceedings of the second entrainment and intake
screening workshop. The John Hopkins University Cooling Water Re-
search Project, Report No. 15.
Dawson, G. W. 1974. The chemical toxicity of elements. BNWL-1815, UC-
70. Battelle Northwest Laboratory, Richland, Washington. 23 pp.
Doudoroff, P., and M. Katz. 1950. Critical review of literature on the
toxicity of industrial wastes and their components to fish. Sew.
Ind. Wastes 22: 1432-1458.
-------
419
Dressel, D. M. 1971. The effects of thermal shock and chlorine on the
estuarine copepod (Acartia tonsa). M.S. Thesis. University of
Virginia. 58 p.
Environmental Protection Agency. 1974. Steam electric power generating
point source category - proposed effluent limitations guidelines and
standards. Federal Register 39: 8294-8307.
Federal Power Commission. 1971. The 1970 national power survey: guide-
lines for growth of the electric power industry. Part I. Chapter
10. Disposal of waste heat from steam-electric plants. 20 p.
Gehrs, C. W., L. D. Eyman, R. L. Jolley, and J. E. Thompson. 1974.
Effects of stable chlorine-containing organics on aquatic environ-
ments. Nature 249: 675-676.
Gentile, J. H. 1974. Toxicity to marine organisms of free chlorine and
chlorinated compounds in seawater. Environmental Protection Agency,
Research development Report. 27 pp. (Unpublished manuscript).
Goodyear, C. P., and J. S. Mattice. 1973. Testimony on effect of Indian
Point unit 2 chlorination on the aquatic biology of the Hudson River.
Indian Point unit. No. 2, ASLB Hearings, Docket No. 50-247, March
1. 13 pp.
Goodyear, C. P., C. C. Coutant, and J. R. Trabalka. 1974. Sources of
potential biological damage from once-through cooling systems of
Nuclear Power Plants. ORNL/TM-4180. Oak Ridge National Laboratory,
Oak Ridge, Tenn. 41 pp.
Gregg, B. C. 1974. The effects of chlorine and heat on selected stream
invertebrates. Ph.D. Thesis, Virginia Polytechnic Institute and
State University. 311 pp.
Holland, E. A., J. E. Laster, E. D. Neumann, and W. E. Eldridge. 1960.
Chlorine and chloramine experiments. Toxic effects of organic and
inorganic pollutants on young salmon. Washington Department of
Fisheries Research Bulletin No. 5, 188, 264 pp.
Hopkins, E. S., and E. L. Bean. 1966. Water purification control. The
Williams and Winkins Co., Baltimore, Maryland.
Boss, D. E., L. C. Coston, J. P. Baptist, and D. W. Engel. 1975. Effects
of temperature, copper, and chlorine on fish during simulated entrain-
ment in power-plant condenser cooling systems, p. 519-527. In
Proceedings of Symposium, Oslo, International Atomic Energy Agency,
Vienna, August 26-30, 1974.
Lee, G. F., and C. L. Stratton. 1972. Effect of cooling tower blowdown
water on receiving water quality - a literature review. Presented
to the Lake Michigan Enforcement Conference, Chicago, Illinois.
September. 53 p.
-------
420
Mattice, J. S. 1976. A method for estimating the toxicity of chlorinated
discharges. Presented at Workshop on Impact of Power Plants on Aqua-
tic Systems, Pacific Grove, California, sponsored by Electric Power
Research Institute. Sept. 28-Oct. 2, 1975. 12 p. In press.
Mattice, J. S., and H. A. Pfuderer. 1976. Chemistry and effects of
chlorine in aquatic systems: an annotated bibliography. ORNL/EIS-82.
Oak Ridge National Laboratory, Oak Ridge, Tenn. 69 pp.
Mattice, J. S., and H. E. Zittel. 1976. Site specific evaluation of power
plant chlorination: a proposal. J. Water Pollut. Control Fed. In
press.
Merkens, J. C. 1958. Studies on the toxicity of chlorine and chloramines
to the rainbow trout. Water Waste Treat. J. 7: 150-151.
Moore, E. W. 1951. Fundamentals of chlorination of sewage and waste.
Water Sewage Works 98: 130-136.
Motley, F. W., and T. C. Hoppe. 1970. Cooling tower design criteria
and water treatment. Presented at Cooling Tower Institute Meeting,
Aspen, Colorado. June.
National Environmental Policy Act of 1969 (NEPA). 1970. P. L. 91-190,
83 Stat. 852. January 1.
Parker, A. 1935. Some problems of water supply. J. Soc. Chem. Ind.
34: 49.
Rosenberger, D. R. 1971. The calculation of acute toxicity of free
chlorine and chloramines to coho salmon by multiple regression
analysis. Thesis. Michigan State University, East Lansing. 72 p.
Scheuring, L., and H. Stetter. 1950. Experiments on the effects of
chlorine on fish and other aquatic organisms. Vom Wasser 18: 101-
140.
Stober, Q. J., and C. H. Hanson. 1974. Toxicity of chlorine and heat
to pink (Oncorhynchus gorbuscha) and chinook salmon ((£. tshawytscha).
Trans. Am. Fish. Soc. 103: 569-576.
U. S. Atomic Energy Commission. 1972. Final environmental statement
related to operation of Shoreham Atomic Power Station. Docket No.
50-322, September.
U. S. Atomic Energy Commission. 1973. Draft environmental statement
related to the proposed Newbold Island Nuclear Generation Station
units 1 and 2. Docket Nos. 50-354 and 50-355. July.
U. S. Atomic Energy Commission. 1974. Final environmental statement
related to the proposed Seabrook Stations units 1 and 2. Docket
Nos. 50-443 and 50-444. December.
-------
421
University of Oklahoma, The Science and Public Policy Program. 1975.
Chapter 13, Energy consumption. In Energy alternatives: a com-
parative analysis. University of Oklahoma, Norman.
Warren, C. E. 1971. Biology and water pollution control. W. B. Saunders,
Philadelphia. 434 p.
Water Resources Council. 1968. The nation's water resources. U. S.
Government Printing Office, Washington, D. C.
Wolf, E. G., M. J. Schneider, and T. 0. Thatcher. Bioassays on the com-
bined effects of chlorine, heavy metals, and temperature on fish
and fish food organisms. Battelle-Northwest Research Laboratory.
Unpublished manuscript.
Zillich, J. S. 1969. The toxic effects of the Grandville Wastewater
Treatment Plant effluent to the fathead minnow, Pimephales promelas.
Michigan Water Resources Commission, Dept. of Water Management Report.
9 P.
DISCUSSION
Joseph E. Draley, Argonne National Laboratory. For large bodies of
water, dilution is the most rapid method of reducing chlorine concentra-
tion, except when there is fast-acting chlorine demand in the admixed
water (especially for the case in which chlorine is discharged). Thus
the isopleth method shown, corrected for currents, provides a reasonable
approximation for some cases only, and it is important to identify those
cases to which it can be applied. The other cases should be addressed
too, such as (1) flowing rivers, where slugs of chlorine and derivatives
flow downstream, and (2) reactions between substances in the receiving
water and the chlorine and chlorine derivatives that remove the toxicant.
Charles C. Coutant, Oak Ridge National Laboratory. I agree that the
examples shown are simplistic. However, if we assume that dilution is
the only cause for decline in toxic chlorine concentration, the same meth-
od can be applied to any known series of isopleths. The actual case will
probably never be a nice simple series of ellipses like I showed in the
slides. Time dependent chlorine demand could be factored into such a
model if it were known.
Guy R. Nelson, U. S. Environmental Protection Agency. Can the model
predict residual chlorine levels in the plume based on an increasing and/
or decreasing amount of residual chlorine in the discharge?
Coutant. This method is based on given residual chlorine levels and
cannot predict them. The method is probably not applicable in situations
in which organisms are exposed to an increasing residual chlorine concen-
tration because the time-dose would probably be underestimated.
-------
422
A. D. Parsons, Wisconsin Electric Power Company. Is the model devel-
oped by Dr. Mattice adaptable to fixed biological populations such as the
benthos and fish positioned within the discharge, as well as moving plank-
tonic populations?
Coutant. Yes. That's a very good question and one that we just
didn't get into because of time. It bears on Joe's comment on the plume.
In most situations the plume is going to move back and forth. The wind
is going to change. It's going to run this direction one time and another
direction another time due to tides, winds, and other situations. And if
you picture a stationary organism at one point out there in the overall
potentially affected area, the chances are he's never going to be exposed
to a constant concentration. He's going to have ups and downs of con-
centration. He'll be exposed at some times and not others. For the same
type of analysis, you just sort of turn the tables and say we've got a
fixed spot out here with a plume waving back and forth and the organism
is exposed to a time series of concentrations and you do the same sort of
analysis. An important question comes up there though, that may come up
in entrainment perhaps to a lesser extent, and that is repeated exposure.
Are repeated exposures additive? Or, if an organism survives one cycle,
is it sort of back to normal, and you start from zero again on the next
cycle? Those questions need to be resolved.
Robert A. Goldstein, Electric Power Research Institute. Written
question submitted after the discussion ended: How do you sum exposures
to varying concentrations? For instance, if an organism is exposed to 10
parts-per-million for 10 minutes and then immediately following to 5 parts-
per-million for 10 minutes, how do you relate the total exposure to your
threshold curve which is based on exposure to a constant concentration for
a given period of time?
Jack S. Mattice, Oak Ridge National Laboratory. The dose-time esti-
mated by this method is the weighted mean of the two exposures for the
total time, that is, 7.5 parts-per-million for twenty minutes. In the
case specified, the method might not be particularly accurate, but organ-
isms in a plume would not be likely to encounter this step-function ex-
posure.
-------
SESSION V. ROUNDTABLE DISCUSSION
ROUNDTABLE PARTICIPANTS
Carl W. Gehrs, Moderator
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
William A. Brungs
National Water Quality Laboratory
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
Joseph E. Draley
Assistant Director
Argonne National Laboratory
Argonne, Illinois 60439
J. Carrell Morris
Division of Engineering
and Applied Physics
Harvard University
Cambridge, Massachusetts 02138
Robert B. Gumming
Biology Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
D. Heyward Hamilton
Division of Biomedical and
Environmental Research
Energy Research and
Development Administration
Washington, D.C. 20545
George Clifford White
Consulting Engineer
San Francisco, California 94118
DISCUSSION
Carl W, Gehrs, Oak Ridge National Laboratory. One of the major
purposes of this conference was to bring together scientists of diverse
background in order to interact and develop, if you will, some hybrid
vitality or viability. In this last session, it is our goal to present
and to allow a forum for a semiformal interaction. Somewhat more formal
than we had at the dinner on Wednesday night and last night at the
social hour; more formal in the sense that the interaction will be
recorded and will consequently be included in the proceedings. As Bob
mentioned at the beginning of the session today, those of you who will
be presenting a question or comment, kindly identify yourself and write
it out, and give them either to Bob or myself or send them to us so
that we can include them in this discussion. We will give each of our
423
-------
424
roundtable participants approximately three to five minutes to summarize
and to present their feelings now with respect to where we are and where
we need to go. Following this we will then open it up to other partici-
pants for questions and comments.
William A. jrungs, U.S. Environmental Protection Agency. As you
all have learned yesterday, quite thoroughly I think and I hope we
all benefit from it, the marine toxicity data and the marine chemistry
of chlorination information is extremely difficult to interpret at this
time. I hope we will benefit from this in the sense that maybe by
including as much information as possible in reports on toxicity tests,
it will be useful later on in interpreting really what levels of residual
chlorine you might have had. Somebody brought the point up yesterday,
and I think it might be quite useful.
There was a little discussion yesterday on the other types of chloro-
organics that are formed when you chlorinate various types of wastewaters.
I think the fact that we had only two papers or so that discussed this
is an indication that it's just a very new concept. It's something that
is so new, we really have no idea or way of interpreting its potential
impact. We know they are formed. We know now they bioaccumulate in
fish. Nobody has even an inkling of what that means. So 1 would hope
over the next few years a few people who have unlimited funding will be
able to answer all these questions for us and maybe even the utilities
could help out.
I think, also, as you heard people talk, especially if you were
somewhat familiar with what they were going to say, you could read between
their lines and sort of keep picking out what might be considered to be
a series of research needs. I'm only going to mention a few because
sometimes we have people in attendance who are just going to begin to do
things. I just spoke to some gentleman who is going to do chronic bio-
assays with copepods in marine water. Maybe between some comments I can
make and other ideas he may have gained this week, he'll be able to do
something that can be even more fruitful than he had originally
anticipated.
I think it's apparent that, from Art Brooks, Will Davis, and others,
that various factors can influence a toxicity residual chlorine, that
is, things such as temperature and water quality. We keep getting incon-
sistent results. I think one reason we get inconsistent results is that
there may be slight differences small enough that some people, because
of experimental design, cannot see it. For example, using another type
of method for sequential temperatures exposure might show differences in
toxicity. So a little more effort in that area would be well worthwhile.
Also the idea of side-by-side comparison of free versus combined
chlorine in fresh water would be extremely useful. It's difficult to
compare somebody's study with free chlorine in the Columbia River with
somebody else's combined chlorine data in the Ohio River. Much of my
comments, I must emphasize, are for fresh water. I am not about to get
into the same problem that Will Davis has right now.
-------
425
I think also, as we heard yesterday and has been discussed and
presented for the last several years, there's a concept of relative
sensitivity of freshwater fish that needs a little more clarification.
We had in the past talked about some monads being more sensitive than
warm water fish. And, as indicated yesterday again by Art, there are
some warm water forage fish of great importance that are extremely sen-
sitive, the same as some monads. So I think we need to also look at
some other important fish families that have not been looked at in the
past. I don't know of any really good data on the catfish family in
fresh water. They could be extremely sensitive or they might be resis-
tant. Without checking a few more of these I think we may be making
some erroneous estimates.
A key point I want to discuss is in Jack Mattice's paper which was
very well given by Chuck Coutant. I hope he's around to answer questions
if it is necessary. I want to bring up a critical point here. The
line on the slide, as you noticed, was not a line that would estimate
a 50% lethal concentration. He mentioned it was an adjustment down
by a factor of 0.37. In other words, if you multiply the LC50 by 0.37,
you obtain an estimated concentration which allows 100% survival. In
playing around with criteria in this review paper I'm finishing up now,
I started looking at the old concept of application factors, whereby
you compare acute and chronic toxicity results to get a ratio which is
called an application factor. Using that concept in this criteria evalu-
ation, you come up with an application factor of 1/8. Which means that
you take 1/8 of the LC50 and you get a chronically safe concentration
in fresh water. If you take 1/3 of that same LC50 value, you're getting
something that would just keep them from dying. So the difference
between 1/3 and 1/8, 1/3 being no mortality and 1/8 being no chronic
effects at all, is extremely narrow. So criteria for short-term exposure
and criteria for continuous chronic exposure are not going to be tremen-
dously different. I think I'll stop with that point. Maybe I've planted
a few seeds that will carry us through the rest of the morning, along
with all the other interesting points to come up.
Robert B. Dimming, Oak Ridge National Laboratory. The job of the
biomedical scientist in dealing with the kind of problems that we have
been discussing for the last several days is to provide information on
compounds which might have an adverse affect on human health. At the
present time that's about where we stand. I would like to suggest that
many of the biomedical scientists aspire to be able to do a little more
than that. That is, to provide information, quantitative information,
that may be useful in making risk-benefit calculations.
I'd like to talk a little about risk and make a few statements
which are largely my own opinion. Carl has said we want to stimulate
discussion, and I think that if I say a few things that a substantial
number of people disagree with that will tend to stimulate discussion.
The first is that all human activity, regardless of what it is, carries
with it risk. Risk cannot be completely avoided. It can only be min-
imized or controlled—understood. It's not realistic to say that no
-------
426
risk is acceptable. And that includes cancer risk. We can't realisti-
cally expect to live in a society with zero risk. One or two comments
creep in every now and then about risk relative to whether or not it is
a voluntary risk or an involuntary risk. I would like to suggest that
most of the risks that we accept in this society are really involuntary.
To say that it's fine to accept a large risk from cigarette smoking,
because that is a voluntary risk, but we cannot accept a small risk from
something else, because it is involuntary, is not an altogether realistic
approach to the problem. Most so-called voluntary risks in this society
are really not voluntary. There is some question, for example, whether
people voluntarily expose themselves to risks from smoking. But if they
do, the person who is sitting next to them, and breathing the same smoke
is undergoing an involuntary risk which is part of being a member of
this society. The same thing may be said of automobiles. People cheer-
fully accept the risk, the large risks involved, on the basis that this
is voluntary. But many, many pedestrians are killed every year involun-
tarily, and I would like to say that, in a city like Oak Ridge with no
public transportation system, driving an automobile is not an altogether
voluntary activity.
So we have risks. What do we wish to do with them. We want to
quantify them so that we can deal with them consciously, if not volun-
tarily—so we can handle these risks and factor them into our risk-
benefit estimates. There are two hazards, which we can identify, which
are associated with the compounds that we have been dealing with in this
conference. These are carcinogenic risks and mutagenic risks. Of those
two, carcinogenesis seems to be, the way we read it now, the most impor-
tant. For qualitative investigations of these kinds of hazards, we have
available to us a battery of submammalian tests which really are good
for alerting us to problems. But they don't tell us anything about
risks. And they don't really tell us, in many cases, whether that partic-
ular compound is really a hazard to humans. To determine that we have
to have other kinds of information, other kinds of- tests.
There are really only two kinds of information which give us any
kind of quantitative information — Information that might be useful
ultimately for risk assessment. Thesfe two are extrapolation of data
obtained from studies on intact mammals and, secondly, epidemiology.
Where it can be applied successfully, epidemiology provides the best
information. This is information that is directly applicable to human
populations. But the problem is that very frequently there are problems
which limit the use of epidemiology. There are confounding factors
which make the interpretation impossible or very difficult. So, while
epidemiology from some points of view can provide information, good
information, frequently it is not available to us. Epidemiology is not
likely, for example, to be useful in assessing mutagenic risks to the
human population. I do not see at any time in the future when it will
be a useful technique for mutagenicity. But it is or could be useful,
could be crucially valuable, in certain cases for evaluating carcino-
genicity. For dealing with mutagenicity, all that we have is quantitative
extrapolation from mammalian test models, and this also provides a use-
ful technique in evaluating carcinogenic problems.
-------
427
Josep_h_E. Draley, Argonne National Laboratory. I am glad I wasn't
last because I kept thinking I might have to change what I wanted to
say because somebody else said it in front of me. What I wanted to do
was to address the "chlorine problem." To address anything that broad
it will be necessary to restrict what you want to say. I have put
three headings down and I'm going to spend a few minutes going into a
little detail under these. One is to identify this problem. The second
one is to determine what to do. And the third one is to regulate and
develop standards or to apply standards.
To know what the problem is, to add some detail to the problem so
you know if there is a problem or if there is a significant problem or
a bad one or a very small one, you need to know some things — to be
able to answer some questions. First of all, how necessary or important
is it to use chlorine of the type that would lead to any discharge?
Secondly, what is released? You can measure releases and you can calcu-
late releases. I suggest to you that it is not now possible to measure
everything you release, and I expect it will not become possible to
measure everything you release. Some of this problem is related to low
concentration, and some of it is related to the almost myriad composition
of your discharge. So there's going to have to be some uncertainty,
and some calculation estimation involved. The third question is what
harmful affects do these have or will they have, as we build more plants
or whatever kind of plants that we're talking about? That problem has
to do with the effect on the environment and the effect directly on man.
Again, you can measure some of them and you are going to have to estimate
some. I think that we've just heard some comment on what can be measured
in one area, and similar kinds of comments will turn out to be made in
other areas.
Under those circumstances, I have, in the day or so that I've been
aware I had to give this little talk, wondered what one ought to do
about it. I must say that it is a credit to the gentleman who thought
up the conference and also organized it, etc. In my own opinion, an
interdisciplinary and interorganizational meeting is one of the very
best things to do to lead to the development of appropriate activities.
It turns out that the communication is often a major part of the problem
of getting things done or taken care of. This kind of meeting, it seems
to me, ought to go a long way over some period of time in helping to
remedy a situation that I think started out with quite bad communication
between the kinds of people that need to communicate. It still isn't
great, but it's quite a bit better. I think this meeting helps it a
lot.
Secondly, I would like to suggest for your consideration in the
group here today, whether a panel could be constituted to make recommen-
dations to appropriate government boards about what kinds of activities
ought to be started or continued. I guess I mean funded — because I
think that is really what it boils down to. What kinds of activities
ought to be funded? That is not an easy question to answer — vfry
difficult. You really are going to have to take all of the points of
-------
428
view that you see here, and that you hear, and you'll presumably hear
more after we're through talking, to get a decent answer to that. I am
not sure that a panel is of any more than a limited value. But I think
we need something that will help, and a panel is the first thing that
occurs to me that sounds like it might be beneficial. I will leave
that.
Finally, with respect to regulations and standards, we cannot
realistically expect that there will be no regulations or standards
until we know exactly how they ought to be constituted and operated.
I think we are going to have to make some kinds of regulations and
standards earlier than that. As you know that has been done for some
of the things. I think there will probably have to be other standards
too. So, the activity of getting optimum regulation and standards needs
to go on. What I mean by that is, although I think it is inevitable
and not undesirable that there be standards before we know how to write
them perfectly or near perfectly, we need to provide a mechanism to
have a review system and improve on the standards as time goes on. As
has been mentioned, the risk-benefit comparison has to be made. I don't
think I need to pursue that. And, in establishment of standards you
ought to have — and this is not a new thought at all, none of these are
really — you ought to have input from the varied kinds of people who
should have input for standards and regulations.
D. Heyward Hamilton, U.S. Energy Research and Development Adminis-
tration. I have some skepticism about the possibility of someone in an
administrative post, like myself, bringing appropriate perspective to the
question that has been put to us, which is to try to suggest something
that might be provocative in an interdisciplinary sense with respect to
cooling problems and chlorination problems. I'd like to run through a
couple of thoughts that I've had in advance of the meeting itself and in
listening to the discussions that have gone on since I've been here.
There is a good deal of work which has only very briefly been referred
to.
First, let me back up and say somebody indicated this morning that
they thought this would be a very difficult act to follow. The question
has been brought up and I'm sure several people have talked about it,
is there a need for another conference of this kind? It seems to me
that it is probable that something like this again would be useful, in
the not too distant future, when we have a little more information in
hand about these problems in estuarine and in marine systems.
I do have some feeling that on the question of "chlorine problems"
as in other similar situations, the industrial growth and the develop-
ment that is going on is somewhat ahead of us. The number of power
plants being sited and likely to be sited in estuarine and marine situ-
ations is growing and increasing rapidly, and we really don't have, as
far as I can tell, a great deal of information about what to expect in
these situations. For example, the importance of pH in controlling what
-------
429
kinds of residuals, if any, are going to be formed. Whether it is or
is not advantageous to dilute the chlorinated effluent stream from a
condenser system with nonchlorinated effluent streams. What kind of
pH changes do you induce in that?
There is at least one person here that feels very strongly that we
should not use chlorine gas or hypochlorite in estuaries. There is a
modest amount of work, that I am aware of at this time, going on in
marine environments. There is a program at Woods Hole under Ryther and
Goldman and there are two representatives of that program here. That
is a toxicity study. We've heard from Dr. Huggett at VIMS and the
potentially frightening issues that he's raised with respect to the impact
on oyster spat settlement, and the problems related to that in measuring
residuals and understanding what is present in an estuarine environment.
Oysters are an important crop in many locations. Dr. Will Davis, Bears
Bluff Field Station; Dr. Carpenter, University of Miami; Dr. Helz,
University of Maryland; Dr. Tom Thatcher at Battelle-Northwest; Dr. Donald
Hoss at Beaufort, North Carolina — those are some of the people, some
supported by EPA, some by NSF, some by us (ERDA) that I am aware of who
are presently carrying out work related to the potential for "chlorine
problems" in marine and estuarine environments. It doesn't amount to a
very great deal, and I think most of it is in waters of full ocean
salinity rather than in estuaries. So it seems to me, there is somewhat
of a gap there that we are going to need to be working on.
For those of you who may be particularly interested in the marine
aspect, let me — you may not be aware of it — let me come back to some-
thing I started to mention before, and then I'm going to stop. The
current work at the food chain research group at Scripps includes three
or four papers on the impact of the San Onofre plant on that coastal
environment, and also on their efforts to screen some marine organisms
for the presence of chlorinated hydrocarbons or halogenated hydrocarbons,
rather, that may be produced in the chlorination of that plant. Some of
these have been submitted, some are in press, and I think I've actually
sent copies to a number of people recently that I thought might be
interested. They are finding some very interesting things. They are
also doing some interesting experimentation. For example, they are
looking at the question of whether or not the toxic residuals formed on
chlorination of sea water may interfere with the deposition of silicon
in diatoms. This kind of second and third order problem that may exist,
I think, is illustrative of the sort of thing we need to worry about
in the marine and estuarine environment.
J. Carrell Morris. Harvard University. It's easy to get a reputa-
tion, I think, for being wise sometimes, by repeating obvious things
that most everyone knows in a very firm voice often enough. And so much
of what I have to say this morning will fall into that category of
repeating for you the multiplication tables.
It seems to me the information we have been dealing with here comes
on two levels. These are not distinct and one can find instances of
-------
430
considerable overlapping and ambiguity, but nonetheless, I do think we
ought to think in terms of information at the pragmatic level and
information at the fundamental level. We do have to have in order to
meet immediate situations and to deal with regulation and control of
immediate factors, a lot of pragmatic research that simply says when you
put so much chlorine into it you get thus and thus toxic effect. The
difficulty is that information like this is restricted more or less to
the particular situation. For example, you can not move from the
Chesapeake Bay to the Pacific Ocean with this kind of information and
translate it directly. In order to do that, in order to universalize
the applicability of your data, in order to understand the factors that
are operating, you do have to have investigations at the fundamental
level.
I would make a plea here that in attempting to meet the problems
of the immediate situation we not overlook the need, also, for long-term
information that leads to long-term predictability and understanding
of the factors that are going on. Otherwise, we will be faced with a
continual need for repeated investigation each time we come up with a
new situation. We will be faced with these continuing inconsistencies
in the data.
I don't think that we ought to blame the biologists because the
chemists have been unable to come up with adequate analytical techniques
for distinguishing exactly the forms and species of chlorine that we have
available, in some of these situations, particularly the estuarine and
marine situations. I think this is a challenge to the analytical and
the physical chemists to find out a little more fundamental information
about these kinds of situations, so that they can inform the biologists
as to what is really going on, so that the biologists are able to do
their fundamental work more adequately.
I do think that in some instances the biologists have been a little
bit careless informing themselves of the extent of information that has
been available, and in setting up their experiments so that they can
know the kinds of things that they are dealing with. It does not, in
many instances, take much adaptation of the experimental regime in order
to have a clear cut, rather than an ambiguous situation.
At the same time I also think that the toxicological and biological
people should always be aware and should look with some suspicion on the
claims of analytical methods or even of the claims of physical chemists,
although they are a noble breed, because they do carry out their work
in relatively clean systems. When Don Johnson tells you that he is able
to differentiate between thus and so, and thus and so, he has picked a
clean system in which to do this. And while he has tried to imagine all
sorts of interferences that might occur, and to look at these, when one
gets into a real situation there are likely to be unsuspected inter-
ferences. So one must not always take the claims, particularly claims
of differentiation, at their face value when they are applied to these
real situations. There should be some investigation of the reliability
of analytical methods under the situations, either by spiking techniques
or things like that.
-------
431
George C. White, Consulting Engineer. The question that was put
to me about this discussion was what was the message that you got after
being here for three days? And since I'm a practitioner of chlorination,
the message comes through to me like this, and it's loud and clear. We
have fouled our own nest, and it's about time we recognize it and we
should clean it up. I'm going to take this in three steps: potable
water, wastewater, and cooling water.
But first I'm going to say, doesn't it seem ridiculous when you
look at the bathtub and I'll site a local bathtub, which is San Francisco
Bay — when you can just reach down anywhere in San Francisco Bay, near
the ferry building, and grab a gallon jug of water and go run a couple
of amperometric residuals, do some chlorine demands, and the biggest
chlorine demand you can get is only a part and a half per million. That's
free chlorine. Yet you take a jug out of Passaic River and you've got
to put in 16 mg/1 to get a proper water. To me this is ridiculous. So,
let's take them one at a time.
Cooling water — the problem that has been presented with the resid-
ual in cooling water — I can not get myself excited about. And dechlo-
rination, to me, is utter desperation. It seems to me that with the
engineering brains we have that cooling waters can be properly diluted
and also the systems- can be optimized. One of the good things that has
come out of all this scare on the chloro-organics is that people are
looking at what they are really doing with chlorine. As you know, human
nature is so "want to do" that if a little bit is good, a whole lot more
is much better, and that's what we've been doing all along. I can't get
excited about the poor oysters, if we've got to trade off something.
Because we in San Francisco lost our whole oyster crop, which was con-
siderable, simply because we had labor problems. We couldn't get people
out there to build the necessary devices to grow the oysters. You can
call that an involuntary risk but that is part of the whole problem.
Also I can't get too excited about the effluent of the San Onofre plant,
when the anchovies in the ocean out off of Los Angeles coast put in more
waste than do the outfalls from L.A. County, and Orange County Sanitation
Districts. So, we've got all these things to look at.
But I am concerned about potable water and wastewater. If you
recall in my opening remarks, I said that a good clean water can be
disinfected with 1 mg/1 and that the range of chlorination dosage in
potable waters in the USA was 1 to 16. Now, there we have a factor of
16 to 1 that we can reduce our chlorine dosages if we clean up our
environment. You've got to first start cleaning up the environment with
the wastewater. Now 16 mg/1 in wastewater, and this is the part that is
ridiculous, can achieve good disinfection. Even a primary effluent, if
you do it right. So just think of the chlorine we could save if we go
to better wastewater traatment.
I don't know how we're going to get it, or whether it's too late or
not, but with the new clean water act the cities can now bear down on
the industries, and get the industries to clean up and to conserve. I
-------
432
think we are going in the right direction. We are just like the race
cars at Indianapolis — they only turn left. We can't turn right. We
are still going left and we are going to stay going that way.
To get back to just one other thing that bothers me. I'll let some-
body else worry about cleaning up the environment, but you can take
a good secondary — let me just back up a little bit. I am a great
believer in letting nature do the work. So I am all in favor of bio-
logical systems for cleaning things up. You can take a good secondary
effluent, nitrify it, get a free chlorine residual with very low dosages,
and you can get virus kills. That is really cleaning up the environment.
I don't know whether we need to kill viruses in sewage effluents, but if
we clean up the wastewater then we're going to clean up our raw potable
water. Maybe this requires storage, but I think it is absolutely ridic-
ulous to subject the people in Passaic Valley to that kind of water.
That's like direct water reuse which is something aesthetically not
acceptable.
Now I'll just close by saying that I agree with Joe Draley that
communication is necessary. To me, it is very frustrating to hear a
paper at one of these conferences and you know that if it is an excellent
paper, you have to wait a year before it is in print. There ought to be
a clearing house for all these presentations so that we don't have to
wait a year for them. I am impressed with this conference because I
never realized so many people could gather in one place and only talk
about chlorine. That is all I want to say. That ought to start something.
AUDIENCE PARTICIPATION IN DISCUSSION
C. Sengupta, Public Service Electric and Gas Company. What I am
going to say is a comment. My choices are between either chlorinating
and running the plant or not chlorinating and not running the plant.
So until someone gives me an alternative to chlorine, it doesn't matter
whether the fish are dying or not. I can go only down to a certain
level. We will try our best to go down to that, but beyond that point
we need an alternate. We have heard so many people doing research here,
so many millions of dollars being spent, but very little on alternates.
What about alternatives to chlorination?
Draley. I made a very brief reference to the subject by making my
number one item in the "chlorine problem" to identify how necessary or
important it is to use chlorine. It becomes less necessary as you pro-
vide alternatives, which is what your really asking. There are some
things that are being done at the moment, but White's statement was that
he didnft think there was any alternative now. Is that right?
White. I just can not get worried about the power plant problem
because with all the things that we have listened to here, the message
is coming through clear to me that the people that are designing the
plants now — we may have some problems with some existing plants that
you can not change — but with the new ones that are coming on the lines,
-------
433
I am sure that there is some way that dilution will allow you to operate
below that acute lethal line.
Brungs. I agree entirely that in many new plants they can do things.
In terms of a lot of the older ones, I don't know how broadly this is
being done. But at least in the state of Michigan, the state in coopera-
tion with Region 5, EPA, they have been doing a lot of work on what they
call minimization programs at generating stations. I will pick the
most extreme one, a plant on the Chicago ship channel or canal, which was
chlorinating fantastically, because it's principally a sanitary waste
effluent. So they had a tremendous slime problem, etc. They dosed a
long time. So they kept cutting back and cutting back on their chlorine
use until they started having back pressure problems, I believe. I'm a
biologist, not an engineer. And then they had to kick it back up a
little bit. But they ended up with, I think, a single dose for 10 minutes
a day from something that had been multiple hours per day. Ignoring
the savings in chlorine, I imagine they saved a few aquatic organisms.
That is something that can be done with older plants. It may or may not
meet the regulations or the criteria or standards. I'm quite sure, also
that some plants may never be able to do that in existing situations
where they are at low flow for 100% of every flow. I think somebody
will have to turn their face on those until they are phased out.
Sengupta. If the power investor does go back to the board and
minimizes the amount of chlorine that is necessary, is that going to be
acceptable? Or do we still have to look and see where the fish are
dying? As I saw it, the level at which we don't have any chronic effect
is 0.0015 ppm and I don't think that we are going to go to that level.
Brungs. The 0.0015 ppm is a chronic value that gets out beyond
96 hours, so we don't have that concern there. I think, as I mentioned,
sometimes somebody in a regulatory agency is going to turn their face
and ignore existing problems because nothing can be done.
J. Donald Johnson, University of North Carolina at Chapel Hill.
I wanted to ask a question of the panel as to what they thought of the
recent proposed elimination of the coliform standards for waters which
are being polluted by wastewater, but which are not subject to direct
reuse or swimming? I wonder if we are throwing out the baby with the
bath water with that kind of approach. Perhaps we should go back and
better control our wastewater disinfection process, as we are suggesting
that we should do for our cooling waters — to minimize the amount of ___
chlorine being used.
White. I don't want to waste time going into philosophy of disin-
fection of wastewater. It's here. I'm sure it's here to stay. But
the important thing about the disinfection of wastewater is that you
clean up the operation. We're going to have to have some patience.
There is going to be some risk in the mean time. But as we go to
secondary effluents, and as the various states set these requirements
for the receiving waters, we are going to stick with the coliform.
-------
434
Johnson. They are not. They are proposing to throw it out.
White. That's not a good idea. We have got to stick with the
coliform count for two very good reasons. It serves us well in iden-
tifying a safe potable water and, if you will recall in ray opening
remarks at the beginning of this session, there is an implication in
the coliform regulation of the type of effluent that you should have to
meet the regulation. Now, a 2.2 effluent implies tertiary treatment.
A 23.2 implies a very good consistent secondary effluent, and this is
why we have got to keep the coliform count in there, because it is
directly related. Now, you can have all the other things in there you
want, you can have BOD, you can have suspended solids, that's great,
and grease and so on. That keeps the operator on his toes. It's some-
thing the operator can see. It's something we can work from. And you've
got to give him the coliform in order to maintain that quality.
Johnson. We are not talking about California.
White. No.
Johnson. These standards they are proposing to throw out are
200 per 100 ml.
White. That is pretty sad.
Johnson. They are talking about eliminating that altogether so
there will not be any coliform standard for nondirect use.
White. That would be a very bad mistake.
Guy R. Nelson, U.S. Environmental Protection Agency. I'd like to
direct my question to Bill Brungs. I want to do it very carefully
because I am an engineer asking the best I can a biological question to
a biologist. In most of the criteria for chlorine levels and the effects
on various animals and plants, there are two parameters: time and
maximum concentration. Is that the best way that we can present them?
For example, I am thinking of one of your famous criteria, that is,
0.2 mg/1 for two hours — that being a maximum concentration. Now, to me
it is more lethal to any organism to have 0.2 mg/1 throughout that
two-hour span versus having it maximum maybe for two minutes during that
two hours.
Brungs. The criteria referred to, I assume, would include those
such as Chuck Coutant showed today for Jack Mattice. Those to me are
criteria. The point you bring up is somewhat of an academic one in the
sense that the data that one has available to him when he is working up
a criterion does not give you the precision that would let you make a
comment on that. The assumption inherent in the criteria is that
theoretically somebody can meet them. If you say, the residual shall
not exceed 0.2 ppm for two hours, somebody can do it for two hours and
maintain a constant 0.2 ppm. That's sort of the assumption one makes.
-------
435
It can be done, and therefore the implication is that when you come
up with a number it should be protective.
I like what Jack has worked up on the time-concentration relation-
ship. I've mentioned, and other people have mentioned in the past
quite frequently, that this is the ultimate for all kinds of criteria,
not just chlorine. It is time-concentration dependent. I think a system
like this will be extremely useful. I know Jack didn't do it intention-
ally, but if you take some of the older criteria from a few years ago
and compare it to his line, the similarity for that one point, at least
in relation to that line, is unusually good considering that a biologist
did the playing around with the data. A long answer that didn't provide
you with much information.
Johnson. Along that same line, I wanted to ask a question about
transients and spikes of concentration that might be missed looking at
toxicity data. I look at the data that is coming out of the studies
and I see median values and relations of the toxicity to those median
values. But when you look back at the data, you find some very large
spikes within those medians. Might not the toxicity really be due to
the sudden surge that you don't see in the medians.
Brungs. The more recent studies, that really have not even produced
manuscripts yet, are utilizing curves such as Guy showed this morning
on the cooling tower blowdown, where they actually are trying to design
their intermittent exposures dose systems so that you get a rapid rise
and then a gradual decline. The difficulty that all of these people
have had is there is no presently acceptable biological concept like
LC50 or TLM or anything like that, that provides an answer to a variable
exposure condition. In one of Jim Truchan's papers, he has introduced
a term called intermittent lethal concentration, and you have to put
in the number of exposures, the maximum concentration and all these.
You can no longer get a median value. It's meaningless. You can not
just give the extremes either. You almost have to plot a curve and
say that these are the experimental conditions and these are my results.
Ideally, we will come up with a term that will be so many milligrams per
liter per minute — a chlorine-minute concept.
Robert C. Paladino, Edison Electric Institute. If utilities min-
imize their use of chlorination and, in the face of no alternatives,
must chlorinate to protect their plants, on what basis should regulations
be written? Protection of fish? Protection of plant? Or both?
White. * thought that when I came here that we were worried about
the chloro-organics in our drinking water. Now it seems that we are
more worried about what we are doing to fish with the cooling waters.
Maybe I have missed a point, but along that line I would like to inter-
ject this when you start talking about regulations. I think one of the
regulations that the government ought to look at very closely, and
I'm thinking of drinking water, that there should be a limit placed on
the amount of chlorine required in a raw water to produce, just to toss
-------
436
a number out, to produce a 1.0 mg/1 free chlorine at the end of 30
minutes. That would stop difficulties with various kinds of lousy
water. Now there are a lot of ways to reduce the chlorine demand of
water. You can do as they do in England. They store water for long
periods of time. It can be aerated. The chlorine demand can be reduced
drastically with a little bit of natural pretreatment. I think that
anything that we can do to get some natural pretreatment we should look
at. I am more concerned about the potable water than I am about the
cooling water. I think that is a solvable problem.
Draley. Did that answer your question?
Paladino. No.
Draley. My answer would be that you can write regulations and
standards with times of applicability, that is, it will become applicable
to plants at different times. And, some plants can be exempted for the
lifetime of the plant. When you make those exemptions, you have to do
it with your eyes open. That is, you can make your choice. You can
either make an exemption and take the consequences or you can shut the
plant down, if it's bad enough, which nobody likes to do. That's the
kind of thing you have to face. You have no alternatives to facing
it that way. As you know, a lot of the regulations that are being
developed do have times on them, and you can find an occasional one
that has exemptions too.
William J. Ross, Nuclear Regulatory Commission, I came to the
conference to get a status report. Dr. White may have already given me
my report, because we're more interested in the cooling water, of course,
than the actual drinking water. But I am going back with three or four
very serious questions. I would like to throw them out and then give
you a very brief reason for them. One, are we in the power plant busi-
ness over regulating right now? Because we do have regulations that are
based on the EPA regulations and I'm not infringing upon their preroga-
tives. Second, speaking as one who has been an analytical chemist for
many more years than a regulator, I was very concerned yesterday to hear
that in our special methods for measuring chlorine, we are not quite
sure what we are measuring. We are now talking about a meter reading.
There was quite a bit of debate about the limit of detection, and yet,
the biologists continue to talk about very low concentrations of chlorine.
That question is when will this situation be improved? Third, can we
(NRC) or EPA do something by making our regulations more site specific
for the near term? I am very interested in this because we are begin-
ning now to talk more and more of marine siting on both the Pacific and
the Atlantic. We run into problems immediately, on the scope, on the
chemistry, on everything. But I guess my main problem is we are trying
to protect the public — and I donft want to get into our banquet speech
again — but are you able to give us what we need right now? Dr. Draley's
comment a while ago was apropos. Can we do something to speed it up?
-------
437
White. Those are good questions.
Draley. Your questions are duely recorded.
Johnson. I wish to speak to the question. I think the first
question is what kind of analytical procedures should we use? First of
all, I can tell you what not to use. This doesn't help much, does it?
But yet you find it in many regulations still today — that is the acid
orthotolidine test. Let's throw those out of the regulations. This
method is coming out of Standard Methods, and it sure ought to come out
of the regulations, state or otherwise. Don't use residuals of acid
orthotolidine to regulate with. The next questions is what can we use
today? My answer is that the amperometric titration is a good laboratory
method. It is the method of choice for analyzing the chlorine species.
I realize it is not a very good field method, and it also lacks some
sensitivity. There are modifications you can make in the amperometric
titration to make it sensitive enough, but you don't find those in the
commercial units. So there is a problem of sensitivity and selectivity.
I don't want to brag but I talked about an electrode we have been
working on and it is commercially available, if anybody wants to buy it.
I don't get a nickel from every unit sold so I'm not selling it. But
that is one method. But it is not sensitive enough. So it is a good
method for controlling in the plant, but it is not a good method for the
field. How do we determine Bill Brungs's levels? 2 ppb? There is a
new method on the horizon, the flux monitor. That is the only method I
know of that is able to measure that kind of level of chlorine out in the
environment. So that is my answer.
White. I would like to make a comment on that. First of all, you
have to accept some kind of an intermediate philosophy and take off from
there. I am not a zero discharge man by any means. This is a sort of
dilemma we are in now. These marine biologists keep talking about 0.000
something or other and they have modified titration procedures that you
have to walk on your tiptoes around — you have to ground the jar with
aluminum foil to keep the noise out and all this sort of thing. First,
if you have a good control system, which very few of these power plants
have, you can control those plumes going out of the condenser at some
very low chlorine residuals. If a regulatory agency would come along
and say 0.1 mg/1 in the plume going into the receiving water is acceptable
under all conditions, then equipment manufacturers would sit down and
they would make control equipment that could do that. But right now,
let's take California for example, they have a 0.02 or 0.05 chlorine
residual in the wastewater effluent. They do not say how you measure
it. I can calibrate a chlorine residual analyzer to put out a 0.05
orthotolodine residual and yet it will be putting out a 1.5 ampero-
metric residual. That is the status of the situation right now. So
what do we do? We say the best way is to go to zero residual and so we
feed forward all of our dechlorination equipment. We are going to
excess S02 which fortunately, so far, is not indictable for fouling the
environment. So we go to about a 0.5 S02 residual, but there's no way
to measure S02 residual practically. You can take probe, and you can
-------
438
fiddle around in the lab, and you can sort of guess at it, but you can't
automate equipment with it. So that is all we have, feed forward con-
trol. You have got to have a primary meter that is controlling the
chlorinator. You have got to have an analyzer controlling the chlorinator.
You have got to have an analyzer at the end of the contact time that is
controlling the sulfonator. You feed a closely coupled meter signal with
the residual signal into the sulfonator, and then you just hope that
everything is all right. Then you monitor it intermittently. The trouble
with monitoring it intermittently is that a chlorine residual analyzer
will go out of calibration when it doesn't see the halogen oxidant over a
long period of time. So we are in trouble. But if somebody would say
we will allow 0.1 mg/1 residual, then the equipment manufacturers could
put in closed-loop control on dechlorination. I want to get that across
because I think that is very important.
Carol H. Tate, James M. Montgomery Consulting Engineers, Inc. I
would like to go back to the drinking water question and throw the
question to the panel as to whether they foresee, or whether they would
advocate, standards for any of these 200 organic compounds in drinking
water?
White. 1 can't answer the question, but I would make the comment
that Russia has standards on 400. Did you mention that, I was talking
when you started, I don't know whether you mentioned that or not. It
makes us look pretty sad. What do we have, 35 or something like that?
32?
Tate. We have a CCE.
White. No. I mean the total number of standards for all constitu-
ents of water. It is very low. There is much talk about other standards.
Tate. How about 5-chlorouracil?
Gumming. I think that the compounds that we know about are present
in such small levels that standards are not really meaningful. I pre-
sume that one could set standards. But at this point we know so little
about the distribution of these compounds in surface waters that it is
not clear to me what would be achieved by attempting to set standards
at this time. I think more information is necessary.
Albert Dietz, Jones Chemical Company. I wanted to follow up on
the question of alternatives for chlorine. It has been suggested that
chlorination may be eliminated in some instances of sanitation in favor
of an alternative sanitizing chemical. What are the alternative chem-
icals proposed for disinfecting raw potable water, municipal wastewater,
cooling water, etc., and what is the availability of these alternative
chemicals?
-------
439
White. I don't think there is any alternative to chlorine. I
know that is a cop-out. I am just a believer in cleaning up the envi-
ronment, optimizing these systems, cutting the dosage down to 1 to
2 mg/1, straightening out our thinking a little on disinfecting waste-
water, and cleaning up the industrial discharge. This will do more to
turn this situation around than anything else. We are just going to
have to live with some of these risks. We just can't make a 180 degree
turn now with the commitment, the investment, the education. Do you
realize that even in some of the most sophisticated plants, you can't
get operators to really appreciate the difference in the species of
chlorine residual. It gets so frustrating, you will have an "umpteen"
million dollar plant and here you see a guy running around with an
orthotolodine kit. You just sort of want to throw up your hands. Not
only that, they will get in trouble with the regulatory agency, and
they will say you are not disinfecting and that you have 2300/100 coloform.
So what does the operator do, he cranks up the chlorinator. He goes out
to the end of the outfall. He takes an orthotolodine residual. He has
a 0 to 10 mg/1 disc and he looks at it and says it's 10 mg/1. But if
you put a titrator on it, it might be 20. He can't measure orthotolodine
higher than 10 mg/1, and he won't take the trouble to go through a dilu-
tion. So these are the things that we are up against. If we are going
to turn around 180 degrees and do something else, we just could never
educate enough people.
Morris. I think George White is an extremist. There are a number
of possible alternatives to chlorine for disinfection purposes. George
has pointed out that we use chlorine for a variety of purposes — not just
for disinfection alone. Moreover, as an introduction here let me also
say that in thinking of alternatives, one should think not only of
alternative chemical substances, but also of alternative ways of applying
chlorine. We have done very little investigation of possible effects
on some of these undesirable side reactions — of what might happen for,
say, a very short-time high dosage of chlorine for disinfection followed
by all most immediate dechlorination before any of the chemical reactions
occur. We have done very little in terms of seeing what might happen
with small multiple doses of chlorine at different points. None of this
has been dealt with systematically and might enable us to minimize a
great many of these problems.
But beside this I think that one should give some credit to the
fact that ozone is able to do a good many of the things that chlorine
can do. If, as in George White's case, one is interested in having a
water with just a little bit of chlorine demand when you add chlorine to
it, one of the ways to clean up the water to this point is to make use of
ozone as a pretreatment. One does not really even need chlorine to go
out into the distribution system. It is possible, as they are doing in
Zurich, to use chlorine dioxide as a substance going out into the dis-
tribution system. While these substances have many of the properties
of chlorine, one of the things that we do know is that they do not cause
substitution of chlorine into the organic molecule in the same way that
chlorine itself does. So that one must keep these in one's eye and
posibly the other new things that still may be coming along and give
-------
440
them a reasonable chance and not, simply because of our long experience
with chlorine, just to say, "Oh, it's the traditional thing. We can't
do anything else."
White. Yes. If you recall in my opening remarks this is one of
the things that I want to see funded. I think we have glossed over the
benefits of chlorine dioxide. We have not investigated ozone. We don't
know enough about the chemistry of the combinations. I totally agree
with Dr. Morris that we need to do this investigation. I was trying to
emphasize cleaning up the environment first, because we can do that at
the same time. In other words, the wastewater people can be working with
that part of the environment while we are working with these different
chemicals. We need to look into water storage before treatment. We need
to look into natural means of pretreatment. We need to look into ozone
first, then chlorine, dechlorination, aeration, rechlorination, and also
preforming of chloramines. There are a whole lot of things that we
haven't done anything about. No real serious chlorination chemistry on
combinations has been done, not to speak of, ever since Dr. Morris did
it in the 40*s and 50's at Harvard. So we have just sort of been
limping along.
Draley. I think we are a little unfair to your question. We have
not addressed it in an "honest-to-God" serious way about practical alter-
natives. The first thing I would say is that it is site specific. There
is not any across-the-board generality about whether you need to use
chlorine or not. For example, the Point Beach Nuclear Plant has never
done anything to defoul, and they've been running for some three years.
They do not have evidence of any problem. Incidentally, the people's
thought about that is, primarily, that there is just enough of the right
kind of fine silt in the water (the plant is located on Lake Michigan)
that the scrubbing is helping them as well as the low contamination level.
That's one thing to say. There are other things. Most everybody here
that has something to do with the power industry knows about the Ampertap
system. We have not mentioned it. I think we ought to say that there
are mechanical devices that will clean condenser tubes. There are
places where those things have done quite a satisfactory job for conden-
ser tubes. I am not sure that it meets the needs of the power plants,
because there are other things than condenser tubes. My answer is that
a mixture of things that include nonchemicals is also possible.
Tate. Is Point Beach once-through? And will the regulators recon-
sider the regulation with respect to once-through for site-specific
cases?
Draley. It is once-through. The answer of the question about the
regulators is going to depend partly upon what kind of pressure is put
on the regulators. I don't know the answer, but I think that they
ought to be site specific or an effort ought to be made to have them
site specific. I think we are doing a public disservice if we insist
that all regulations have to be the same for everybody.
-------
441
Gehrs. I think perhaps that this is a good time to bring this
together. I believe this latter point was, in my opinion, the most
important facet with respect to regulation of chlorination at power
plants. Is the real need for the regulations to be flexible and viable,
as you will, with respect to specific sites?
I want to thank each of you for your participation. I personally
think it was a very valuable conference bringing together scientists of
diverse backgrounds. The interactions started here will go on through
verbal and written interactions or future meetings.
The following represent the editor's summarization of two state-
ments too lengthy for inclusion in this report:
David H. Rosenblatt, U.S. Army Medical Bioengineering Research and
Development Laboratory. Dr. Rosenblatt summarized the complex equilibria
involved in the simple case of the hydrolysis of low concentrations of
chlorine in distilled water. He pointed out the simplifying assumptions
necessary for development of computer programs for calculation of equi-
librium concentrations of the chemical species of interest. The chemical
model and program developed will be adapted to other more complex equi-
librium and dynamic situations more closely approximating the real world.
Dr. David D. Woodbridge, Florida Institute of Technology. Dr.
Woodbridge summarized his experimental data on the use of irradiation
for disinfection and sewage treatment as an alternative to chlorine.
Experimental results were shown which indicated significant reductions
in concentration of chlorine and chloro-organics in water and sewage
effluents within time periods of several minutes using upwards to
80 kilorad doses.
The following are written comments or questions received after
the discussion had ended:
Jerome McKersie, Wisconsin Division of Natural Resources. The
proposed secondary treatment standards, published in the Federal Register
August 15, 1975, deletes disinfection requirements for municipal waste-
water treatment plants. Instead, disinfection requirements would be
based on state water quality standards. Will the EPA develop criteria
or guidance on where chlorination, chlorination with disinfection, no
disinfection or alternate methods of disinfection should be applied?
This type of information is vital to application of water quality
standards.
John R. J. Sorenson, QUAD Corporation. It is probably true^that,
as Mr. George C. White has put it, 'Ve have fowled our own nest."
Chlorination of waters has resulted in the introduction of too many
-------
442
unneeded chlorinated compounds. This has been the result of our misuse
of chlorine, which was only intended as a water disinfectant but has
been excessively used to the point that excess quantities of hypochlorous
acid are available for the unwanted reaction processes leading to the
production of undesirable chlorinated compounds. Since the rate of
disinfection is fast compared to the rate of formation of chlorine con-
taining compounds in the water chlorination process, efficient mixing
of water with chlorine should allow the reduction of the amount of chlo-
rine used for its intended purpose, disinfection. The use of chlorine
should not be allowed for the purpose of oxidation in the treatment of
wastewaters just to speed up the wastewater treatment. Wastewater treat-
ment should be done as it was in the past with natural processes employing
aeration and microbial decomposition of the components of wastewater.
Water can not be continually chlorinated on intake and discharge as it
passes from city to city along the waterway. It should be clear that
additions to this water by each city will become part of the water envi-
ronment of the next city. After a number of cities have made their
additions to this water it is no wonder that cities at the end of the
waterway have a grossly adulterated water supply.
Consistent with this is the need for the use of pure chlorine in
the water disinfection process. The chlorine used today is the chlorine
that no one else will buy. This chlorine is the dregs of the chlorine
production process. Chemical users of chlorine specify that it be
nearly 100% pure chlorine. This is done because they require that their
chemical syntheses be run as clean as possible to obtain good yields
of the desired products. To furnish this high quality chlorine the
chlorine produced in the electrolytic process is purified by allowing
the impure chlorine to pass through a fractionating tower which brings
about the removal of chlorinated compounds produced as a result of the
reaction of chlorine with lubricants and the graphite electrodes used in
the electrolytic production of chlorine. As I understand it, these
impurities are removed from the equipment along with the chlorine that
is used for sanitation purposes. Since these impurities are composed of
chloroform, methylene chloride, carbon tetrachloride in addition to
chlorinated aromatic and aliphatic hydrocarbons, this chlorine is con-
taminated with materials which are suspected as being responsible for
liver cancers seen in humans. Even if these impurities are added in
small quantities, if they are being added by each city along a waterway,
those individuals in cities along and at the end of that waterway may
be exposed to larger and larger quantities of chlorinated compounds which
may represent a substantial health risk.
At the moment and in the foreseeable future there will be no alter-
native to disinfection by chlorination. Chlorination is a remarkably
safe and effective method of disinfection and its applicability can be
sustained with preliminary natural treatment such as settling, filtra-
tion, and microbial removal of wastewater components prior to chlorina-
tion, by the efficient mixing of waters to be disinfected with pure
chlorine.
-------
443
Some have suggested that ozonization is an alternative to chlorina-
tion. However, ozonides and epoxides, which are known to be the active
species of the carcinogenic aromatic hydrocarbons, represent a clearly
defined cancer risk and should not be quickly adopted as an alternative.
No method of disinfection should be seriously considered until the
research has been done to show that that method produces water which has
been shown to be safer or less toxic than that produced by chlorination.
Research should continue with regard to the better utilization of
chlorine and the effects of compounds produced in this process. Research
on alternative methods should be done to establish that the alternative
method represents a lesser health risk.
Stephen A. Hubbs. Louisville Water Company. Drawing an analogy
between the concern of organic and heavy metal pollutants of ten years
ago and the thrust towards watershed and riverbasin planning today, I
wondered if it might be beneficial to compile an inventory of chlorine
discharged into receiving streams on an entire riverbasin level. Then,
as the kinetics and chemical mechanisms of chlorine chemistry unravel,
the aquatic effects of water chlorination may be immediately evaluated
based on known exposures to flora and fauna for extended time periods.
In essense, we would be gathering raw field data for an experiment which
has not yet been designed.
Chronologically, the situation for such data accumulation could not
be better. In Kentucky, all waste discharges of any quantity and quality
must provide information for NPDES permits (as in other states). Those
discharges utilizing chlorine have the capability to measure the amount
of chlorine in their effluent. Likewise, the total amount of chlorine
used in the process should be easily obtainable. These data, along
with a note concerning the method of chlorine analysis used, may provide
adequate information to evaluate bioaccumulation and bottom sediment
accumulation of chlorinated compounds. Thus, by requiring this infor-
mation on NPDES permits, a data base could be established that may
greatly aide the evaluation of long-term effects of water chlorination.
T. F. Craft, Georgia Institute of Technology. In spite of the work
that has already been carried out, it is apparent that we have only
nibbled at the edges of the problem. Our lack of knowledge of aqueous
chlorine reactions, their products, yields, and significance has been
brought into sharp focus by this meeting. A need for periodic reviews
of progress is obvious, and I hope that subsequent conferences at
appropriate intervals can be held.
------- |